{
  "best_global_step": 24200,
  "best_metric": 0.24659645557403564,
  "best_model_checkpoint": "saves/prompt-tuning/llama-3-8b-instruct/train_cola_1744902672/checkpoint-24200",
  "epoch": 83.16008316008316,
  "eval_steps": 200,
  "global_step": 40000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.010395010395010396,
      "grad_norm": 0.13502776622772217,
      "learning_rate": 0.29999999259779675,
      "loss": 11.1472,
      "num_input_tokens_seen": 3680,
      "step": 5
    },
    {
      "epoch": 0.02079002079002079,
      "grad_norm": 0.06767003238201141,
      "learning_rate": 0.29999996252634736,
      "loss": 8.1576,
      "num_input_tokens_seen": 7584,
      "step": 10
    },
    {
      "epoch": 0.031185031185031187,
      "grad_norm": 0.0790955126285553,
      "learning_rate": 0.2999999093230187,
      "loss": 6.3339,
      "num_input_tokens_seen": 11264,
      "step": 15
    },
    {
      "epoch": 0.04158004158004158,
      "grad_norm": 0.07418269664049149,
      "learning_rate": 0.299999832987819,
      "loss": 5.6189,
      "num_input_tokens_seen": 15232,
      "step": 20
    },
    {
      "epoch": 0.05197505197505198,
      "grad_norm": 0.020311672240495682,
      "learning_rate": 0.29999973352076004,
      "loss": 4.9144,
      "num_input_tokens_seen": 18976,
      "step": 25
    },
    {
      "epoch": 0.062370062370062374,
      "grad_norm": 0.1048390343785286,
      "learning_rate": 0.2999996109218572,
      "loss": 4.2665,
      "num_input_tokens_seen": 22784,
      "step": 30
    },
    {
      "epoch": 0.07276507276507277,
      "grad_norm": 0.02579992264509201,
      "learning_rate": 0.2999994651911293,
      "loss": 3.5678,
      "num_input_tokens_seen": 26560,
      "step": 35
    },
    {
      "epoch": 0.08316008316008316,
      "grad_norm": 0.02307146228849888,
      "learning_rate": 0.2999992963285989,
      "loss": 2.935,
      "num_input_tokens_seen": 30432,
      "step": 40
    },
    {
      "epoch": 0.09355509355509356,
      "grad_norm": 0.017982393503189087,
      "learning_rate": 0.29999910433429194,
      "loss": 2.2972,
      "num_input_tokens_seen": 34208,
      "step": 45
    },
    {
      "epoch": 0.10395010395010396,
      "grad_norm": 0.03599481284618378,
      "learning_rate": 0.29999888920823814,
      "loss": 1.7707,
      "num_input_tokens_seen": 37856,
      "step": 50
    },
    {
      "epoch": 0.11434511434511435,
      "grad_norm": 0.036551497876644135,
      "learning_rate": 0.29999865095047057,
      "loss": 1.3339,
      "num_input_tokens_seen": 41760,
      "step": 55
    },
    {
      "epoch": 0.12474012474012475,
      "grad_norm": 0.024816714227199554,
      "learning_rate": 0.29999838956102604,
      "loss": 0.7955,
      "num_input_tokens_seen": 45568,
      "step": 60
    },
    {
      "epoch": 0.13513513513513514,
      "grad_norm": 0.013269827701151371,
      "learning_rate": 0.29999810503994484,
      "loss": 0.4969,
      "num_input_tokens_seen": 49376,
      "step": 65
    },
    {
      "epoch": 0.14553014553014554,
      "grad_norm": 0.01420884020626545,
      "learning_rate": 0.29999779738727084,
      "loss": 0.3892,
      "num_input_tokens_seen": 53152,
      "step": 70
    },
    {
      "epoch": 0.15592515592515593,
      "grad_norm": 0.007514605764299631,
      "learning_rate": 0.29999746660305154,
      "loss": 0.3384,
      "num_input_tokens_seen": 57024,
      "step": 75
    },
    {
      "epoch": 0.16632016632016633,
      "grad_norm": 0.01797129213809967,
      "learning_rate": 0.2999971126873379,
      "loss": 0.2932,
      "num_input_tokens_seen": 60928,
      "step": 80
    },
    {
      "epoch": 0.17671517671517672,
      "grad_norm": 0.09406391531229019,
      "learning_rate": 0.2999967356401845,
      "loss": 0.3492,
      "num_input_tokens_seen": 64672,
      "step": 85
    },
    {
      "epoch": 0.18711018711018712,
      "grad_norm": 0.051161929965019226,
      "learning_rate": 0.29999633546164944,
      "loss": 2.7805,
      "num_input_tokens_seen": 68416,
      "step": 90
    },
    {
      "epoch": 0.19750519750519752,
      "grad_norm": 0.047030698508024216,
      "learning_rate": 0.29999591215179444,
      "loss": 2.9492,
      "num_input_tokens_seen": 72288,
      "step": 95
    },
    {
      "epoch": 0.2079002079002079,
      "grad_norm": 0.029817547649145126,
      "learning_rate": 0.2999954657106849,
      "loss": 1.4902,
      "num_input_tokens_seen": 76224,
      "step": 100
    },
    {
      "epoch": 0.2182952182952183,
      "grad_norm": 0.019339989870786667,
      "learning_rate": 0.2999949961383896,
      "loss": 0.8145,
      "num_input_tokens_seen": 80032,
      "step": 105
    },
    {
      "epoch": 0.2286902286902287,
      "grad_norm": 0.03233005106449127,
      "learning_rate": 0.2999945034349809,
      "loss": 0.5747,
      "num_input_tokens_seen": 84000,
      "step": 110
    },
    {
      "epoch": 0.2390852390852391,
      "grad_norm": 0.04136732965707779,
      "learning_rate": 0.2999939876005348,
      "loss": 0.4918,
      "num_input_tokens_seen": 87840,
      "step": 115
    },
    {
      "epoch": 0.2494802494802495,
      "grad_norm": 0.008512151427567005,
      "learning_rate": 0.29999344863513094,
      "loss": 0.4423,
      "num_input_tokens_seen": 91680,
      "step": 120
    },
    {
      "epoch": 0.2598752598752599,
      "grad_norm": 0.02633056789636612,
      "learning_rate": 0.2999928865388523,
      "loss": 0.3913,
      "num_input_tokens_seen": 95648,
      "step": 125
    },
    {
      "epoch": 0.2702702702702703,
      "grad_norm": 0.029025232419371605,
      "learning_rate": 0.29999230131178567,
      "loss": 0.4051,
      "num_input_tokens_seen": 99392,
      "step": 130
    },
    {
      "epoch": 0.2806652806652807,
      "grad_norm": 0.01705402508378029,
      "learning_rate": 0.2999916929540212,
      "loss": 0.397,
      "num_input_tokens_seen": 103424,
      "step": 135
    },
    {
      "epoch": 0.2910602910602911,
      "grad_norm": 0.03655802831053734,
      "learning_rate": 0.29999106146565285,
      "loss": 0.4193,
      "num_input_tokens_seen": 107168,
      "step": 140
    },
    {
      "epoch": 0.30145530145530147,
      "grad_norm": 0.021294977515935898,
      "learning_rate": 0.29999040684677786,
      "loss": 0.3571,
      "num_input_tokens_seen": 111104,
      "step": 145
    },
    {
      "epoch": 0.31185031185031187,
      "grad_norm": 0.007375025190412998,
      "learning_rate": 0.2999897290974972,
      "loss": 0.3344,
      "num_input_tokens_seen": 114944,
      "step": 150
    },
    {
      "epoch": 0.32224532224532226,
      "grad_norm": 0.009771570563316345,
      "learning_rate": 0.2999890282179155,
      "loss": 0.3405,
      "num_input_tokens_seen": 118944,
      "step": 155
    },
    {
      "epoch": 0.33264033264033266,
      "grad_norm": 0.033297546207904816,
      "learning_rate": 0.29998830420814077,
      "loss": 0.3741,
      "num_input_tokens_seen": 122848,
      "step": 160
    },
    {
      "epoch": 0.34303534303534305,
      "grad_norm": 0.027968930080533028,
      "learning_rate": 0.2999875570682846,
      "loss": 0.3409,
      "num_input_tokens_seen": 126656,
      "step": 165
    },
    {
      "epoch": 0.35343035343035345,
      "grad_norm": 0.006693217437714338,
      "learning_rate": 0.2999867867984623,
      "loss": 0.3116,
      "num_input_tokens_seen": 130272,
      "step": 170
    },
    {
      "epoch": 0.36382536382536385,
      "grad_norm": 0.025718048214912415,
      "learning_rate": 0.29998599339879267,
      "loss": 0.3441,
      "num_input_tokens_seen": 134144,
      "step": 175
    },
    {
      "epoch": 0.37422037422037424,
      "grad_norm": 0.02536095678806305,
      "learning_rate": 0.29998517686939796,
      "loss": 0.3092,
      "num_input_tokens_seen": 137984,
      "step": 180
    },
    {
      "epoch": 0.38461538461538464,
      "grad_norm": 0.02042613923549652,
      "learning_rate": 0.29998433721040413,
      "loss": 0.3889,
      "num_input_tokens_seen": 141728,
      "step": 185
    },
    {
      "epoch": 0.39501039501039503,
      "grad_norm": 0.03502725064754486,
      "learning_rate": 0.29998347442194073,
      "loss": 0.3761,
      "num_input_tokens_seen": 145504,
      "step": 190
    },
    {
      "epoch": 0.40540540540540543,
      "grad_norm": 0.010042614303529263,
      "learning_rate": 0.2999825885041407,
      "loss": 0.3479,
      "num_input_tokens_seen": 149216,
      "step": 195
    },
    {
      "epoch": 0.4158004158004158,
      "grad_norm": 0.004226129036396742,
      "learning_rate": 0.29998167945714077,
      "loss": 0.3225,
      "num_input_tokens_seen": 153120,
      "step": 200
    },
    {
      "epoch": 0.4158004158004158,
      "eval_loss": 0.30661338567733765,
      "eval_runtime": 13.3817,
      "eval_samples_per_second": 63.968,
      "eval_steps_per_second": 15.992,
      "num_input_tokens_seen": 153120,
      "step": 200
    },
    {
      "epoch": 0.4261954261954262,
      "grad_norm": 0.029882246628403664,
      "learning_rate": 0.2999807472810811,
      "loss": 0.3066,
      "num_input_tokens_seen": 156992,
      "step": 205
    },
    {
      "epoch": 0.4365904365904366,
      "grad_norm": 0.01448049210011959,
      "learning_rate": 0.29997979197610536,
      "loss": 0.3289,
      "num_input_tokens_seen": 160672,
      "step": 210
    },
    {
      "epoch": 0.446985446985447,
      "grad_norm": 0.008233333937823772,
      "learning_rate": 0.299978813542361,
      "loss": 0.3489,
      "num_input_tokens_seen": 164448,
      "step": 215
    },
    {
      "epoch": 0.4573804573804574,
      "grad_norm": 0.010366841219365597,
      "learning_rate": 0.2999778119799988,
      "loss": 0.3181,
      "num_input_tokens_seen": 168416,
      "step": 220
    },
    {
      "epoch": 0.4677754677754678,
      "grad_norm": 0.002933696610853076,
      "learning_rate": 0.29997678728917326,
      "loss": 0.3125,
      "num_input_tokens_seen": 172448,
      "step": 225
    },
    {
      "epoch": 0.4781704781704782,
      "grad_norm": 0.021501148119568825,
      "learning_rate": 0.2999757394700424,
      "loss": 0.3142,
      "num_input_tokens_seen": 176256,
      "step": 230
    },
    {
      "epoch": 0.4885654885654886,
      "grad_norm": 0.005177576560527086,
      "learning_rate": 0.29997466852276783,
      "loss": 0.3367,
      "num_input_tokens_seen": 179968,
      "step": 235
    },
    {
      "epoch": 0.498960498960499,
      "grad_norm": 0.020754391327500343,
      "learning_rate": 0.29997357444751466,
      "loss": 0.328,
      "num_input_tokens_seen": 183808,
      "step": 240
    },
    {
      "epoch": 0.5093555093555093,
      "grad_norm": 0.0068879518657922745,
      "learning_rate": 0.2999724572444516,
      "loss": 0.3212,
      "num_input_tokens_seen": 187648,
      "step": 245
    },
    {
      "epoch": 0.5197505197505198,
      "grad_norm": 0.020043136551976204,
      "learning_rate": 0.29997131691375095,
      "loss": 0.3052,
      "num_input_tokens_seen": 191488,
      "step": 250
    },
    {
      "epoch": 0.5301455301455301,
      "grad_norm": 0.013065611943602562,
      "learning_rate": 0.2999701534555886,
      "loss": 0.3754,
      "num_input_tokens_seen": 195328,
      "step": 255
    },
    {
      "epoch": 0.5405405405405406,
      "grad_norm": 0.03995097428560257,
      "learning_rate": 0.2999689668701439,
      "loss": 0.4291,
      "num_input_tokens_seen": 199008,
      "step": 260
    },
    {
      "epoch": 0.5509355509355509,
      "grad_norm": 0.028009936213493347,
      "learning_rate": 0.29996775715759993,
      "loss": 0.4939,
      "num_input_tokens_seen": 202848,
      "step": 265
    },
    {
      "epoch": 0.5613305613305614,
      "grad_norm": 0.01900005154311657,
      "learning_rate": 0.2999665243181432,
      "loss": 0.4212,
      "num_input_tokens_seen": 206816,
      "step": 270
    },
    {
      "epoch": 0.5717255717255717,
      "grad_norm": 0.02086745575070381,
      "learning_rate": 0.2999652683519638,
      "loss": 0.4318,
      "num_input_tokens_seen": 210496,
      "step": 275
    },
    {
      "epoch": 0.5821205821205822,
      "grad_norm": 0.03029743582010269,
      "learning_rate": 0.29996398925925544,
      "loss": 0.3949,
      "num_input_tokens_seen": 214432,
      "step": 280
    },
    {
      "epoch": 0.5925155925155925,
      "grad_norm": 0.009148284792900085,
      "learning_rate": 0.2999626870402154,
      "loss": 0.3509,
      "num_input_tokens_seen": 218144,
      "step": 285
    },
    {
      "epoch": 0.6029106029106029,
      "grad_norm": 0.003327825805172324,
      "learning_rate": 0.29996136169504445,
      "loss": 0.2992,
      "num_input_tokens_seen": 221920,
      "step": 290
    },
    {
      "epoch": 0.6133056133056133,
      "grad_norm": 0.011896105483174324,
      "learning_rate": 0.29996001322394694,
      "loss": 0.3451,
      "num_input_tokens_seen": 225728,
      "step": 295
    },
    {
      "epoch": 0.6237006237006237,
      "grad_norm": 0.012294461019337177,
      "learning_rate": 0.29995864162713093,
      "loss": 0.3668,
      "num_input_tokens_seen": 229376,
      "step": 300
    },
    {
      "epoch": 0.6340956340956341,
      "grad_norm": 0.01129826158285141,
      "learning_rate": 0.2999572469048079,
      "loss": 0.3196,
      "num_input_tokens_seen": 233408,
      "step": 305
    },
    {
      "epoch": 0.6444906444906445,
      "grad_norm": 0.022352736443281174,
      "learning_rate": 0.29995582905719287,
      "loss": 0.3337,
      "num_input_tokens_seen": 237152,
      "step": 310
    },
    {
      "epoch": 0.6548856548856549,
      "grad_norm": 0.018420910462737083,
      "learning_rate": 0.2999543880845046,
      "loss": 0.4146,
      "num_input_tokens_seen": 241024,
      "step": 315
    },
    {
      "epoch": 0.6652806652806653,
      "grad_norm": 0.007515036966651678,
      "learning_rate": 0.2999529239869652,
      "loss": 0.3348,
      "num_input_tokens_seen": 244768,
      "step": 320
    },
    {
      "epoch": 0.6756756756756757,
      "grad_norm": 0.008488037623465061,
      "learning_rate": 0.2999514367648005,
      "loss": 0.3123,
      "num_input_tokens_seen": 248832,
      "step": 325
    },
    {
      "epoch": 0.6860706860706861,
      "grad_norm": 0.01192617416381836,
      "learning_rate": 0.29994992641823987,
      "loss": 0.2993,
      "num_input_tokens_seen": 252704,
      "step": 330
    },
    {
      "epoch": 0.6964656964656964,
      "grad_norm": 0.006763458251953125,
      "learning_rate": 0.29994839294751613,
      "loss": 0.282,
      "num_input_tokens_seen": 256576,
      "step": 335
    },
    {
      "epoch": 0.7068607068607069,
      "grad_norm": 0.019589852541685104,
      "learning_rate": 0.29994683635286584,
      "loss": 0.3058,
      "num_input_tokens_seen": 260384,
      "step": 340
    },
    {
      "epoch": 0.7172557172557172,
      "grad_norm": 0.011946571990847588,
      "learning_rate": 0.2999452566345291,
      "loss": 0.3021,
      "num_input_tokens_seen": 264224,
      "step": 345
    },
    {
      "epoch": 0.7276507276507277,
      "grad_norm": 0.004684171173721552,
      "learning_rate": 0.2999436537927494,
      "loss": 0.3145,
      "num_input_tokens_seen": 268032,
      "step": 350
    },
    {
      "epoch": 0.738045738045738,
      "grad_norm": 0.008012689650058746,
      "learning_rate": 0.299942027827774,
      "loss": 0.3055,
      "num_input_tokens_seen": 271648,
      "step": 355
    },
    {
      "epoch": 0.7484407484407485,
      "grad_norm": 0.0197573471814394,
      "learning_rate": 0.29994037873985363,
      "loss": 0.3179,
      "num_input_tokens_seen": 275456,
      "step": 360
    },
    {
      "epoch": 0.7588357588357588,
      "grad_norm": 0.0026678459253162146,
      "learning_rate": 0.29993870652924254,
      "loss": 0.2814,
      "num_input_tokens_seen": 279232,
      "step": 365
    },
    {
      "epoch": 0.7692307692307693,
      "grad_norm": 0.0012401110725477338,
      "learning_rate": 0.29993701119619876,
      "loss": 0.3143,
      "num_input_tokens_seen": 282880,
      "step": 370
    },
    {
      "epoch": 0.7796257796257796,
      "grad_norm": 0.016871865838766098,
      "learning_rate": 0.2999352927409835,
      "loss": 0.3419,
      "num_input_tokens_seen": 286560,
      "step": 375
    },
    {
      "epoch": 0.7900207900207901,
      "grad_norm": 0.01477054599672556,
      "learning_rate": 0.29993355116386194,
      "loss": 0.374,
      "num_input_tokens_seen": 290400,
      "step": 380
    },
    {
      "epoch": 0.8004158004158004,
      "grad_norm": 0.018198197707533836,
      "learning_rate": 0.29993178646510266,
      "loss": 0.3913,
      "num_input_tokens_seen": 294176,
      "step": 385
    },
    {
      "epoch": 0.8108108108108109,
      "grad_norm": 0.012671583332121372,
      "learning_rate": 0.2999299986449777,
      "loss": 0.3133,
      "num_input_tokens_seen": 298016,
      "step": 390
    },
    {
      "epoch": 0.8212058212058212,
      "grad_norm": 0.011242128908634186,
      "learning_rate": 0.29992818770376284,
      "loss": 0.2792,
      "num_input_tokens_seen": 301760,
      "step": 395
    },
    {
      "epoch": 0.8316008316008316,
      "grad_norm": 0.007946498692035675,
      "learning_rate": 0.29992635364173725,
      "loss": 0.2547,
      "num_input_tokens_seen": 305504,
      "step": 400
    },
    {
      "epoch": 0.8316008316008316,
      "eval_loss": 0.2710922956466675,
      "eval_runtime": 13.4137,
      "eval_samples_per_second": 63.815,
      "eval_steps_per_second": 15.954,
      "num_input_tokens_seen": 305504,
      "step": 400
    },
    {
      "epoch": 0.841995841995842,
      "grad_norm": 0.006660632789134979,
      "learning_rate": 0.2999244964591839,
      "loss": 0.3168,
      "num_input_tokens_seen": 309280,
      "step": 405
    },
    {
      "epoch": 0.8523908523908524,
      "grad_norm": 0.02225935272872448,
      "learning_rate": 0.2999226161563891,
      "loss": 0.2954,
      "num_input_tokens_seen": 313184,
      "step": 410
    },
    {
      "epoch": 0.8627858627858628,
      "grad_norm": 0.004138450603932142,
      "learning_rate": 0.2999207127336429,
      "loss": 0.3142,
      "num_input_tokens_seen": 316960,
      "step": 415
    },
    {
      "epoch": 0.8731808731808732,
      "grad_norm": 0.018559657037258148,
      "learning_rate": 0.2999187861912387,
      "loss": 0.3172,
      "num_input_tokens_seen": 320704,
      "step": 420
    },
    {
      "epoch": 0.8835758835758836,
      "grad_norm": 0.010357167571783066,
      "learning_rate": 0.2999168365294737,
      "loss": 0.3592,
      "num_input_tokens_seen": 324512,
      "step": 425
    },
    {
      "epoch": 0.893970893970894,
      "grad_norm": 0.021069606766104698,
      "learning_rate": 0.29991486374864856,
      "loss": 0.339,
      "num_input_tokens_seen": 328448,
      "step": 430
    },
    {
      "epoch": 0.9043659043659044,
      "grad_norm": 0.018024256452918053,
      "learning_rate": 0.29991286784906745,
      "loss": 0.3239,
      "num_input_tokens_seen": 332192,
      "step": 435
    },
    {
      "epoch": 0.9147609147609148,
      "grad_norm": 0.003186495741829276,
      "learning_rate": 0.2999108488310382,
      "loss": 0.2863,
      "num_input_tokens_seen": 335936,
      "step": 440
    },
    {
      "epoch": 0.9251559251559252,
      "grad_norm": 0.026957403868436813,
      "learning_rate": 0.29990880669487213,
      "loss": 0.3209,
      "num_input_tokens_seen": 339744,
      "step": 445
    },
    {
      "epoch": 0.9355509355509356,
      "grad_norm": 0.0018293571192771196,
      "learning_rate": 0.29990674144088425,
      "loss": 0.2633,
      "num_input_tokens_seen": 343552,
      "step": 450
    },
    {
      "epoch": 0.9459459459459459,
      "grad_norm": 0.03124442882835865,
      "learning_rate": 0.299904653069393,
      "loss": 0.353,
      "num_input_tokens_seen": 347552,
      "step": 455
    },
    {
      "epoch": 0.9563409563409564,
      "grad_norm": 0.01398311834782362,
      "learning_rate": 0.29990254158072044,
      "loss": 0.3451,
      "num_input_tokens_seen": 351392,
      "step": 460
    },
    {
      "epoch": 0.9667359667359667,
      "grad_norm": 0.006803765427321196,
      "learning_rate": 0.2999004069751921,
      "loss": 0.2995,
      "num_input_tokens_seen": 355168,
      "step": 465
    },
    {
      "epoch": 0.9771309771309772,
      "grad_norm": 0.010371006093919277,
      "learning_rate": 0.2998982492531373,
      "loss": 0.2956,
      "num_input_tokens_seen": 358976,
      "step": 470
    },
    {
      "epoch": 0.9875259875259875,
      "grad_norm": 0.015438636764883995,
      "learning_rate": 0.2998960684148887,
      "loss": 0.3037,
      "num_input_tokens_seen": 362720,
      "step": 475
    },
    {
      "epoch": 0.997920997920998,
      "grad_norm": 0.008887200616300106,
      "learning_rate": 0.29989386446078264,
      "loss": 0.2824,
      "num_input_tokens_seen": 366592,
      "step": 480
    },
    {
      "epoch": 1.0083160083160083,
      "grad_norm": 0.01895778626203537,
      "learning_rate": 0.299891637391159,
      "loss": 0.3007,
      "num_input_tokens_seen": 370456,
      "step": 485
    },
    {
      "epoch": 1.0187110187110187,
      "grad_norm": 0.006097553763538599,
      "learning_rate": 0.2998893872063612,
      "loss": 0.3314,
      "num_input_tokens_seen": 374296,
      "step": 490
    },
    {
      "epoch": 1.0291060291060292,
      "grad_norm": 0.00845389161258936,
      "learning_rate": 0.2998871139067363,
      "loss": 0.3113,
      "num_input_tokens_seen": 378168,
      "step": 495
    },
    {
      "epoch": 1.0395010395010396,
      "grad_norm": 0.016880623996257782,
      "learning_rate": 0.2998848174926348,
      "loss": 0.3365,
      "num_input_tokens_seen": 381944,
      "step": 500
    },
    {
      "epoch": 1.04989604989605,
      "grad_norm": 0.013562766835093498,
      "learning_rate": 0.2998824979644109,
      "loss": 0.312,
      "num_input_tokens_seen": 385720,
      "step": 505
    },
    {
      "epoch": 1.0602910602910602,
      "grad_norm": 0.0033673536963760853,
      "learning_rate": 0.29988015532242224,
      "loss": 0.2701,
      "num_input_tokens_seen": 389688,
      "step": 510
    },
    {
      "epoch": 1.0706860706860706,
      "grad_norm": 0.0027311728335916996,
      "learning_rate": 0.29987778956703015,
      "loss": 0.3352,
      "num_input_tokens_seen": 393496,
      "step": 515
    },
    {
      "epoch": 1.0810810810810811,
      "grad_norm": 0.005902850069105625,
      "learning_rate": 0.2998754006985994,
      "loss": 0.3422,
      "num_input_tokens_seen": 397464,
      "step": 520
    },
    {
      "epoch": 1.0914760914760915,
      "grad_norm": 0.0026898488868027925,
      "learning_rate": 0.29987298871749846,
      "loss": 0.2888,
      "num_input_tokens_seen": 401144,
      "step": 525
    },
    {
      "epoch": 1.1018711018711018,
      "grad_norm": 0.0024025789462029934,
      "learning_rate": 0.2998705536240992,
      "loss": 0.2809,
      "num_input_tokens_seen": 404984,
      "step": 530
    },
    {
      "epoch": 1.1122661122661124,
      "grad_norm": 0.004416186828166246,
      "learning_rate": 0.2998680954187772,
      "loss": 0.2781,
      "num_input_tokens_seen": 408856,
      "step": 535
    },
    {
      "epoch": 1.1226611226611227,
      "grad_norm": 0.0009456760017201304,
      "learning_rate": 0.2998656141019115,
      "loss": 0.296,
      "num_input_tokens_seen": 412696,
      "step": 540
    },
    {
      "epoch": 1.133056133056133,
      "grad_norm": 0.019509410485625267,
      "learning_rate": 0.2998631096738848,
      "loss": 0.34,
      "num_input_tokens_seen": 416440,
      "step": 545
    },
    {
      "epoch": 1.1434511434511434,
      "grad_norm": 0.008890431374311447,
      "learning_rate": 0.29986058213508326,
      "loss": 0.2898,
      "num_input_tokens_seen": 420312,
      "step": 550
    },
    {
      "epoch": 1.1538461538461537,
      "grad_norm": 0.012911866419017315,
      "learning_rate": 0.29985803148589674,
      "loss": 0.2893,
      "num_input_tokens_seen": 424152,
      "step": 555
    },
    {
      "epoch": 1.1642411642411643,
      "grad_norm": 0.010390352457761765,
      "learning_rate": 0.2998554577267185,
      "loss": 0.2787,
      "num_input_tokens_seen": 428152,
      "step": 560
    },
    {
      "epoch": 1.1746361746361746,
      "grad_norm": 0.0026202909648418427,
      "learning_rate": 0.2998528608579455,
      "loss": 0.2808,
      "num_input_tokens_seen": 431896,
      "step": 565
    },
    {
      "epoch": 1.185031185031185,
      "grad_norm": 0.018476203083992004,
      "learning_rate": 0.2998502408799781,
      "loss": 0.3248,
      "num_input_tokens_seen": 435704,
      "step": 570
    },
    {
      "epoch": 1.1954261954261955,
      "grad_norm": 0.006836924236267805,
      "learning_rate": 0.2998475977932205,
      "loss": 0.3075,
      "num_input_tokens_seen": 439544,
      "step": 575
    },
    {
      "epoch": 1.2058212058212059,
      "grad_norm": 0.009780202060937881,
      "learning_rate": 0.29984493159808023,
      "loss": 0.2988,
      "num_input_tokens_seen": 443480,
      "step": 580
    },
    {
      "epoch": 1.2162162162162162,
      "grad_norm": 0.006580342072993517,
      "learning_rate": 0.29984224229496836,
      "loss": 0.2871,
      "num_input_tokens_seen": 447320,
      "step": 585
    },
    {
      "epoch": 1.2266112266112266,
      "grad_norm": 0.00594101520255208,
      "learning_rate": 0.2998395298842998,
      "loss": 0.4131,
      "num_input_tokens_seen": 451096,
      "step": 590
    },
    {
      "epoch": 1.237006237006237,
      "grad_norm": 0.012957035563886166,
      "learning_rate": 0.29983679436649263,
      "loss": 0.4489,
      "num_input_tokens_seen": 454872,
      "step": 595
    },
    {
      "epoch": 1.2474012474012475,
      "grad_norm": 0.010852313600480556,
      "learning_rate": 0.2998340357419689,
      "loss": 0.4244,
      "num_input_tokens_seen": 458648,
      "step": 600
    },
    {
      "epoch": 1.2474012474012475,
      "eval_loss": 0.2911844551563263,
      "eval_runtime": 13.4237,
      "eval_samples_per_second": 63.768,
      "eval_steps_per_second": 15.942,
      "num_input_tokens_seen": 458648,
      "step": 600
    },
    {
      "epoch": 1.2577962577962578,
      "grad_norm": 0.00274831079877913,
      "learning_rate": 0.29983125401115385,
      "loss": 0.3973,
      "num_input_tokens_seen": 462360,
      "step": 605
    },
    {
      "epoch": 1.2681912681912682,
      "grad_norm": 0.010852557606995106,
      "learning_rate": 0.29982844917447654,
      "loss": 0.3923,
      "num_input_tokens_seen": 466008,
      "step": 610
    },
    {
      "epoch": 1.2785862785862787,
      "grad_norm": 0.005563750863075256,
      "learning_rate": 0.2998256212323695,
      "loss": 0.3145,
      "num_input_tokens_seen": 469816,
      "step": 615
    },
    {
      "epoch": 1.288981288981289,
      "grad_norm": 0.009022928774356842,
      "learning_rate": 0.29982277018526887,
      "loss": 0.3037,
      "num_input_tokens_seen": 473592,
      "step": 620
    },
    {
      "epoch": 1.2993762993762994,
      "grad_norm": 0.0011146427132189274,
      "learning_rate": 0.2998198960336143,
      "loss": 0.3358,
      "num_input_tokens_seen": 477496,
      "step": 625
    },
    {
      "epoch": 1.3097713097713097,
      "grad_norm": 0.008859907276928425,
      "learning_rate": 0.299816998777849,
      "loss": 0.3219,
      "num_input_tokens_seen": 481304,
      "step": 630
    },
    {
      "epoch": 1.32016632016632,
      "grad_norm": 0.01543111726641655,
      "learning_rate": 0.2998140784184197,
      "loss": 0.3113,
      "num_input_tokens_seen": 485336,
      "step": 635
    },
    {
      "epoch": 1.3305613305613306,
      "grad_norm": 0.004169647116214037,
      "learning_rate": 0.2998111349557769,
      "loss": 0.3054,
      "num_input_tokens_seen": 489080,
      "step": 640
    },
    {
      "epoch": 1.340956340956341,
      "grad_norm": 0.008798204362392426,
      "learning_rate": 0.29980816839037444,
      "loss": 0.3016,
      "num_input_tokens_seen": 492760,
      "step": 645
    },
    {
      "epoch": 1.3513513513513513,
      "grad_norm": 0.0036174955312162638,
      "learning_rate": 0.2998051787226698,
      "loss": 0.3035,
      "num_input_tokens_seen": 496664,
      "step": 650
    },
    {
      "epoch": 1.3617463617463619,
      "grad_norm": 0.0030274689197540283,
      "learning_rate": 0.29980216595312403,
      "loss": 0.3252,
      "num_input_tokens_seen": 500376,
      "step": 655
    },
    {
      "epoch": 1.3721413721413722,
      "grad_norm": 0.007786626927554607,
      "learning_rate": 0.29979913008220177,
      "loss": 0.3237,
      "num_input_tokens_seen": 504056,
      "step": 660
    },
    {
      "epoch": 1.3825363825363826,
      "grad_norm": 0.006295106373727322,
      "learning_rate": 0.2997960711103711,
      "loss": 0.2821,
      "num_input_tokens_seen": 507864,
      "step": 665
    },
    {
      "epoch": 1.392931392931393,
      "grad_norm": 0.0015057717682793736,
      "learning_rate": 0.29979298903810386,
      "loss": 0.2874,
      "num_input_tokens_seen": 511608,
      "step": 670
    },
    {
      "epoch": 1.4033264033264032,
      "grad_norm": 0.0022668475285172462,
      "learning_rate": 0.29978988386587524,
      "loss": 0.3142,
      "num_input_tokens_seen": 515416,
      "step": 675
    },
    {
      "epoch": 1.4137214137214138,
      "grad_norm": 0.015244388952851295,
      "learning_rate": 0.2997867555941642,
      "loss": 0.3406,
      "num_input_tokens_seen": 519192,
      "step": 680
    },
    {
      "epoch": 1.4241164241164241,
      "grad_norm": 0.00899590365588665,
      "learning_rate": 0.299783604223453,
      "loss": 0.3567,
      "num_input_tokens_seen": 522968,
      "step": 685
    },
    {
      "epoch": 1.4345114345114345,
      "grad_norm": 0.002670561196282506,
      "learning_rate": 0.29978042975422786,
      "loss": 0.2691,
      "num_input_tokens_seen": 526712,
      "step": 690
    },
    {
      "epoch": 1.444906444906445,
      "grad_norm": 0.015272680670022964,
      "learning_rate": 0.29977723218697816,
      "loss": 0.3362,
      "num_input_tokens_seen": 530552,
      "step": 695
    },
    {
      "epoch": 1.4553014553014554,
      "grad_norm": 0.004360658582299948,
      "learning_rate": 0.299774011522197,
      "loss": 0.2537,
      "num_input_tokens_seen": 534360,
      "step": 700
    },
    {
      "epoch": 1.4656964656964657,
      "grad_norm": 0.012599441222846508,
      "learning_rate": 0.29977076776038114,
      "loss": 0.3003,
      "num_input_tokens_seen": 538232,
      "step": 705
    },
    {
      "epoch": 1.476091476091476,
      "grad_norm": 0.0012430218048393726,
      "learning_rate": 0.2997675009020307,
      "loss": 0.2616,
      "num_input_tokens_seen": 542008,
      "step": 710
    },
    {
      "epoch": 1.4864864864864864,
      "grad_norm": 0.02020493522286415,
      "learning_rate": 0.2997642109476496,
      "loss": 0.3375,
      "num_input_tokens_seen": 545848,
      "step": 715
    },
    {
      "epoch": 1.496881496881497,
      "grad_norm": 0.005804737564176321,
      "learning_rate": 0.299760897897745,
      "loss": 0.3179,
      "num_input_tokens_seen": 549688,
      "step": 720
    },
    {
      "epoch": 1.5072765072765073,
      "grad_norm": 0.008141480386257172,
      "learning_rate": 0.29975756175282803,
      "loss": 0.3099,
      "num_input_tokens_seen": 553400,
      "step": 725
    },
    {
      "epoch": 1.5176715176715176,
      "grad_norm": 0.012977621518075466,
      "learning_rate": 0.29975420251341306,
      "loss": 0.3475,
      "num_input_tokens_seen": 557368,
      "step": 730
    },
    {
      "epoch": 1.5280665280665282,
      "grad_norm": 0.0030478150583803654,
      "learning_rate": 0.29975082018001814,
      "loss": 0.275,
      "num_input_tokens_seen": 561176,
      "step": 735
    },
    {
      "epoch": 1.5384615384615383,
      "grad_norm": 0.0014059582026675344,
      "learning_rate": 0.2997474147531648,
      "loss": 0.2837,
      "num_input_tokens_seen": 565048,
      "step": 740
    },
    {
      "epoch": 1.5488565488565489,
      "grad_norm": 0.004261959809809923,
      "learning_rate": 0.29974398623337833,
      "loss": 0.2866,
      "num_input_tokens_seen": 568728,
      "step": 745
    },
    {
      "epoch": 1.5592515592515592,
      "grad_norm": 0.004058978520333767,
      "learning_rate": 0.2997405346211873,
      "loss": 0.2968,
      "num_input_tokens_seen": 572536,
      "step": 750
    },
    {
      "epoch": 1.5696465696465696,
      "grad_norm": 0.0016424914356321096,
      "learning_rate": 0.2997370599171241,
      "loss": 0.2722,
      "num_input_tokens_seen": 576408,
      "step": 755
    },
    {
      "epoch": 1.5800415800415801,
      "grad_norm": 0.00536632165312767,
      "learning_rate": 0.2997335621217246,
      "loss": 0.2821,
      "num_input_tokens_seen": 580216,
      "step": 760
    },
    {
      "epoch": 1.5904365904365905,
      "grad_norm": 0.0010197683004662395,
      "learning_rate": 0.29973004123552816,
      "loss": 0.3525,
      "num_input_tokens_seen": 584024,
      "step": 765
    },
    {
      "epoch": 1.6008316008316008,
      "grad_norm": 0.004009200260043144,
      "learning_rate": 0.2997264972590777,
      "loss": 0.3089,
      "num_input_tokens_seen": 587736,
      "step": 770
    },
    {
      "epoch": 1.6112266112266114,
      "grad_norm": 0.007063076365739107,
      "learning_rate": 0.29972293019291973,
      "loss": 0.3178,
      "num_input_tokens_seen": 591704,
      "step": 775
    },
    {
      "epoch": 1.6216216216216215,
      "grad_norm": 0.01016835868358612,
      "learning_rate": 0.2997193400376045,
      "loss": 0.3011,
      "num_input_tokens_seen": 595576,
      "step": 780
    },
    {
      "epoch": 1.632016632016632,
      "grad_norm": 0.006793530657887459,
      "learning_rate": 0.2997157267936854,
      "loss": 0.3123,
      "num_input_tokens_seen": 599384,
      "step": 785
    },
    {
      "epoch": 1.6424116424116424,
      "grad_norm": 0.006273724604398012,
      "learning_rate": 0.2997120904617199,
      "loss": 0.2827,
      "num_input_tokens_seen": 603032,
      "step": 790
    },
    {
      "epoch": 1.6528066528066527,
      "grad_norm": 0.006716644391417503,
      "learning_rate": 0.29970843104226863,
      "loss": 0.2628,
      "num_input_tokens_seen": 606904,
      "step": 795
    },
    {
      "epoch": 1.6632016632016633,
      "grad_norm": 0.0029990433249622583,
      "learning_rate": 0.2997047485358959,
      "loss": 0.36,
      "num_input_tokens_seen": 610680,
      "step": 800
    },
    {
      "epoch": 1.6632016632016633,
      "eval_loss": 0.2836262285709381,
      "eval_runtime": 13.3947,
      "eval_samples_per_second": 63.906,
      "eval_steps_per_second": 15.976,
      "num_input_tokens_seen": 610680,
      "step": 800
    },
    {
      "epoch": 1.6735966735966736,
      "grad_norm": 0.002816583961248398,
      "learning_rate": 0.2997010429431697,
      "loss": 0.3065,
      "num_input_tokens_seen": 614552,
      "step": 805
    },
    {
      "epoch": 1.683991683991684,
      "grad_norm": 0.006148617714643478,
      "learning_rate": 0.29969731426466134,
      "loss": 0.2967,
      "num_input_tokens_seen": 618328,
      "step": 810
    },
    {
      "epoch": 1.6943866943866945,
      "grad_norm": 0.004586759489029646,
      "learning_rate": 0.299693562500946,
      "loss": 0.319,
      "num_input_tokens_seen": 622072,
      "step": 815
    },
    {
      "epoch": 1.7047817047817047,
      "grad_norm": 0.001133651821874082,
      "learning_rate": 0.29968978765260207,
      "loss": 0.2803,
      "num_input_tokens_seen": 625912,
      "step": 820
    },
    {
      "epoch": 1.7151767151767152,
      "grad_norm": 0.005019965581595898,
      "learning_rate": 0.2996859897202118,
      "loss": 0.2548,
      "num_input_tokens_seen": 629560,
      "step": 825
    },
    {
      "epoch": 1.7255717255717256,
      "grad_norm": 0.0042550889775156975,
      "learning_rate": 0.2996821687043609,
      "loss": 0.3045,
      "num_input_tokens_seen": 633304,
      "step": 830
    },
    {
      "epoch": 1.735966735966736,
      "grad_norm": 0.00268346955999732,
      "learning_rate": 0.2996783246056384,
      "loss": 0.2763,
      "num_input_tokens_seen": 637112,
      "step": 835
    },
    {
      "epoch": 1.7463617463617465,
      "grad_norm": 0.0021750526502728462,
      "learning_rate": 0.29967445742463744,
      "loss": 0.2619,
      "num_input_tokens_seen": 640888,
      "step": 840
    },
    {
      "epoch": 1.7567567567567568,
      "grad_norm": 0.0048044403083622456,
      "learning_rate": 0.29967056716195417,
      "loss": 0.2555,
      "num_input_tokens_seen": 644632,
      "step": 845
    },
    {
      "epoch": 1.7671517671517671,
      "grad_norm": 0.004150399472564459,
      "learning_rate": 0.2996666538181885,
      "loss": 0.3068,
      "num_input_tokens_seen": 648440,
      "step": 850
    },
    {
      "epoch": 1.7775467775467777,
      "grad_norm": 0.008193307556211948,
      "learning_rate": 0.29966271739394407,
      "loss": 0.2639,
      "num_input_tokens_seen": 652088,
      "step": 855
    },
    {
      "epoch": 1.7879417879417878,
      "grad_norm": 0.009282145649194717,
      "learning_rate": 0.29965875788982776,
      "loss": 0.3415,
      "num_input_tokens_seen": 655992,
      "step": 860
    },
    {
      "epoch": 1.7983367983367984,
      "grad_norm": 0.007523451466113329,
      "learning_rate": 0.2996547753064503,
      "loss": 0.4057,
      "num_input_tokens_seen": 659800,
      "step": 865
    },
    {
      "epoch": 1.8087318087318087,
      "grad_norm": 0.010368818417191505,
      "learning_rate": 0.29965076964442583,
      "loss": 0.3145,
      "num_input_tokens_seen": 663672,
      "step": 870
    },
    {
      "epoch": 1.819126819126819,
      "grad_norm": 0.011550137773156166,
      "learning_rate": 0.299646740904372,
      "loss": 0.3179,
      "num_input_tokens_seen": 667512,
      "step": 875
    },
    {
      "epoch": 1.8295218295218296,
      "grad_norm": 0.008157351985573769,
      "learning_rate": 0.29964268908691016,
      "loss": 0.3393,
      "num_input_tokens_seen": 671480,
      "step": 880
    },
    {
      "epoch": 1.83991683991684,
      "grad_norm": 0.010317876003682613,
      "learning_rate": 0.29963861419266513,
      "loss": 0.3286,
      "num_input_tokens_seen": 675416,
      "step": 885
    },
    {
      "epoch": 1.8503118503118503,
      "grad_norm": 0.008963325060904026,
      "learning_rate": 0.29963451622226533,
      "loss": 0.3084,
      "num_input_tokens_seen": 679096,
      "step": 890
    },
    {
      "epoch": 1.8607068607068609,
      "grad_norm": 0.0024737264029681683,
      "learning_rate": 0.29963039517634277,
      "loss": 0.2849,
      "num_input_tokens_seen": 683000,
      "step": 895
    },
    {
      "epoch": 1.871101871101871,
      "grad_norm": 0.00416366895660758,
      "learning_rate": 0.2996262510555328,
      "loss": 0.2938,
      "num_input_tokens_seen": 686904,
      "step": 900
    },
    {
      "epoch": 1.8814968814968815,
      "grad_norm": 0.008106127381324768,
      "learning_rate": 0.2996220838604746,
      "loss": 0.2921,
      "num_input_tokens_seen": 690552,
      "step": 905
    },
    {
      "epoch": 1.8918918918918919,
      "grad_norm": 0.00660976255312562,
      "learning_rate": 0.29961789359181085,
      "loss": 0.2735,
      "num_input_tokens_seen": 694456,
      "step": 910
    },
    {
      "epoch": 1.9022869022869022,
      "grad_norm": 0.0022041278425604105,
      "learning_rate": 0.29961368025018764,
      "loss": 0.2903,
      "num_input_tokens_seen": 698200,
      "step": 915
    },
    {
      "epoch": 1.9126819126819128,
      "grad_norm": 0.003316605929285288,
      "learning_rate": 0.2996094438362548,
      "loss": 0.2906,
      "num_input_tokens_seen": 702200,
      "step": 920
    },
    {
      "epoch": 1.9230769230769231,
      "grad_norm": 0.0021033193916082382,
      "learning_rate": 0.2996051843506657,
      "loss": 0.2906,
      "num_input_tokens_seen": 706072,
      "step": 925
    },
    {
      "epoch": 1.9334719334719335,
      "grad_norm": 0.0011535674566403031,
      "learning_rate": 0.299600901794077,
      "loss": 0.3528,
      "num_input_tokens_seen": 709816,
      "step": 930
    },
    {
      "epoch": 1.943866943866944,
      "grad_norm": 0.003972208593040705,
      "learning_rate": 0.29959659616714923,
      "loss": 0.3141,
      "num_input_tokens_seen": 713784,
      "step": 935
    },
    {
      "epoch": 1.9542619542619541,
      "grad_norm": 0.0016242626588791609,
      "learning_rate": 0.2995922674705464,
      "loss": 0.2762,
      "num_input_tokens_seen": 717592,
      "step": 940
    },
    {
      "epoch": 1.9646569646569647,
      "grad_norm": 0.0049610259011387825,
      "learning_rate": 0.2995879157049361,
      "loss": 0.2725,
      "num_input_tokens_seen": 721400,
      "step": 945
    },
    {
      "epoch": 1.975051975051975,
      "grad_norm": 0.011431998573243618,
      "learning_rate": 0.2995835408709893,
      "loss": 0.2931,
      "num_input_tokens_seen": 725240,
      "step": 950
    },
    {
      "epoch": 1.9854469854469854,
      "grad_norm": 0.0026282817125320435,
      "learning_rate": 0.29957914296938076,
      "loss": 0.3078,
      "num_input_tokens_seen": 729176,
      "step": 955
    },
    {
      "epoch": 1.995841995841996,
      "grad_norm": 0.0013613426126539707,
      "learning_rate": 0.2995747220007886,
      "loss": 0.2721,
      "num_input_tokens_seen": 733080,
      "step": 960
    },
    {
      "epoch": 2.006237006237006,
      "grad_norm": 0.0009268623543903232,
      "learning_rate": 0.2995702779658947,
      "loss": 0.2691,
      "num_input_tokens_seen": 736776,
      "step": 965
    },
    {
      "epoch": 2.0166320166320166,
      "grad_norm": 0.0008056171354837716,
      "learning_rate": 0.29956581086538425,
      "loss": 0.2651,
      "num_input_tokens_seen": 740584,
      "step": 970
    },
    {
      "epoch": 2.027027027027027,
      "grad_norm": 0.0022148482967168093,
      "learning_rate": 0.2995613206999462,
      "loss": 0.2726,
      "num_input_tokens_seen": 744392,
      "step": 975
    },
    {
      "epoch": 2.0374220374220373,
      "grad_norm": 0.006720210425555706,
      "learning_rate": 0.29955680747027297,
      "loss": 0.2949,
      "num_input_tokens_seen": 748264,
      "step": 980
    },
    {
      "epoch": 2.047817047817048,
      "grad_norm": 0.005039123352617025,
      "learning_rate": 0.2995522711770607,
      "loss": 0.3024,
      "num_input_tokens_seen": 752296,
      "step": 985
    },
    {
      "epoch": 2.0582120582120584,
      "grad_norm": 0.0040133618749678135,
      "learning_rate": 0.2995477118210087,
      "loss": 0.2464,
      "num_input_tokens_seen": 756296,
      "step": 990
    },
    {
      "epoch": 2.0686070686070686,
      "grad_norm": 0.0010765615152195096,
      "learning_rate": 0.29954312940282024,
      "loss": 0.2711,
      "num_input_tokens_seen": 760072,
      "step": 995
    },
    {
      "epoch": 2.079002079002079,
      "grad_norm": 0.004728620406240225,
      "learning_rate": 0.29953852392320196,
      "loss": 0.3203,
      "num_input_tokens_seen": 763880,
      "step": 1000
    },
    {
      "epoch": 2.079002079002079,
      "eval_loss": 0.2618466913700104,
      "eval_runtime": 13.4863,
      "eval_samples_per_second": 63.472,
      "eval_steps_per_second": 15.868,
      "num_input_tokens_seen": 763880,
      "step": 1000
    },
    {
      "epoch": 2.0893970893970892,
      "grad_norm": 0.0026711730752140284,
      "learning_rate": 0.2995338953828641,
      "loss": 0.2836,
      "num_input_tokens_seen": 767656,
      "step": 1005
    },
    {
      "epoch": 2.0997920997921,
      "grad_norm": 0.0010943195084109902,
      "learning_rate": 0.2995292437825204,
      "loss": 0.2887,
      "num_input_tokens_seen": 771464,
      "step": 1010
    },
    {
      "epoch": 2.1101871101871104,
      "grad_norm": 0.003534775460138917,
      "learning_rate": 0.29952456912288816,
      "loss": 0.274,
      "num_input_tokens_seen": 775304,
      "step": 1015
    },
    {
      "epoch": 2.1205821205821205,
      "grad_norm": 0.0011056277435272932,
      "learning_rate": 0.2995198714046884,
      "loss": 0.2687,
      "num_input_tokens_seen": 779240,
      "step": 1020
    },
    {
      "epoch": 2.130977130977131,
      "grad_norm": 0.005262910388410091,
      "learning_rate": 0.2995151506286454,
      "loss": 0.2717,
      "num_input_tokens_seen": 783080,
      "step": 1025
    },
    {
      "epoch": 2.141372141372141,
      "grad_norm": 0.002627658424898982,
      "learning_rate": 0.2995104067954873,
      "loss": 0.2692,
      "num_input_tokens_seen": 786952,
      "step": 1030
    },
    {
      "epoch": 2.1517671517671517,
      "grad_norm": 0.0009026611223816872,
      "learning_rate": 0.2995056399059456,
      "loss": 0.2799,
      "num_input_tokens_seen": 790728,
      "step": 1035
    },
    {
      "epoch": 2.1621621621621623,
      "grad_norm": 0.001486190245486796,
      "learning_rate": 0.2995008499607554,
      "loss": 0.278,
      "num_input_tokens_seen": 794536,
      "step": 1040
    },
    {
      "epoch": 2.1725571725571724,
      "grad_norm": 0.0032881435472518206,
      "learning_rate": 0.2994960369606554,
      "loss": 0.326,
      "num_input_tokens_seen": 798472,
      "step": 1045
    },
    {
      "epoch": 2.182952182952183,
      "grad_norm": 0.003928574733436108,
      "learning_rate": 0.2994912009063878,
      "loss": 0.2935,
      "num_input_tokens_seen": 802440,
      "step": 1050
    },
    {
      "epoch": 2.1933471933471935,
      "grad_norm": 0.0009759695967659354,
      "learning_rate": 0.29948634179869843,
      "loss": 0.2807,
      "num_input_tokens_seen": 806376,
      "step": 1055
    },
    {
      "epoch": 2.2037422037422036,
      "grad_norm": 0.0009490824886597693,
      "learning_rate": 0.29948145963833656,
      "loss": 0.2577,
      "num_input_tokens_seen": 810184,
      "step": 1060
    },
    {
      "epoch": 2.214137214137214,
      "grad_norm": 0.001630600425414741,
      "learning_rate": 0.29947655442605514,
      "loss": 0.2781,
      "num_input_tokens_seen": 814056,
      "step": 1065
    },
    {
      "epoch": 2.2245322245322248,
      "grad_norm": 0.0023589676711708307,
      "learning_rate": 0.2994716261626106,
      "loss": 0.2535,
      "num_input_tokens_seen": 817864,
      "step": 1070
    },
    {
      "epoch": 2.234927234927235,
      "grad_norm": 0.0018771227914839983,
      "learning_rate": 0.2994666748487629,
      "loss": 0.2867,
      "num_input_tokens_seen": 821480,
      "step": 1075
    },
    {
      "epoch": 2.2453222453222454,
      "grad_norm": 0.0007826806395314634,
      "learning_rate": 0.2994617004852756,
      "loss": 0.2733,
      "num_input_tokens_seen": 825192,
      "step": 1080
    },
    {
      "epoch": 2.2557172557172556,
      "grad_norm": 0.006928688380867243,
      "learning_rate": 0.2994567030729159,
      "loss": 0.2671,
      "num_input_tokens_seen": 828936,
      "step": 1085
    },
    {
      "epoch": 2.266112266112266,
      "grad_norm": 0.00867519062012434,
      "learning_rate": 0.29945168261245436,
      "loss": 0.3084,
      "num_input_tokens_seen": 832744,
      "step": 1090
    },
    {
      "epoch": 2.2765072765072767,
      "grad_norm": 0.001305434969253838,
      "learning_rate": 0.29944663910466524,
      "loss": 0.2839,
      "num_input_tokens_seen": 836584,
      "step": 1095
    },
    {
      "epoch": 2.286902286902287,
      "grad_norm": 0.0025245025753974915,
      "learning_rate": 0.2994415725503263,
      "loss": 0.2291,
      "num_input_tokens_seen": 840456,
      "step": 1100
    },
    {
      "epoch": 2.2972972972972974,
      "grad_norm": 0.002111408393830061,
      "learning_rate": 0.29943648295021885,
      "loss": 0.2861,
      "num_input_tokens_seen": 844200,
      "step": 1105
    },
    {
      "epoch": 2.3076923076923075,
      "grad_norm": 0.0007129978621378541,
      "learning_rate": 0.2994313703051278,
      "loss": 0.3016,
      "num_input_tokens_seen": 847976,
      "step": 1110
    },
    {
      "epoch": 2.318087318087318,
      "grad_norm": 0.0009313568007200956,
      "learning_rate": 0.29942623461584156,
      "loss": 0.2704,
      "num_input_tokens_seen": 851944,
      "step": 1115
    },
    {
      "epoch": 2.3284823284823286,
      "grad_norm": 0.0012056262930855155,
      "learning_rate": 0.29942107588315214,
      "loss": 0.266,
      "num_input_tokens_seen": 855720,
      "step": 1120
    },
    {
      "epoch": 2.3388773388773387,
      "grad_norm": 0.0022411260288208723,
      "learning_rate": 0.29941589410785513,
      "loss": 0.2979,
      "num_input_tokens_seen": 859560,
      "step": 1125
    },
    {
      "epoch": 2.3492723492723493,
      "grad_norm": 0.0026391763240098953,
      "learning_rate": 0.29941068929074954,
      "loss": 0.2738,
      "num_input_tokens_seen": 863240,
      "step": 1130
    },
    {
      "epoch": 2.35966735966736,
      "grad_norm": 0.0035182605497539043,
      "learning_rate": 0.2994054614326381,
      "loss": 0.2716,
      "num_input_tokens_seen": 867048,
      "step": 1135
    },
    {
      "epoch": 2.37006237006237,
      "grad_norm": 0.0032562746200710535,
      "learning_rate": 0.29940021053432686,
      "loss": 0.2895,
      "num_input_tokens_seen": 870952,
      "step": 1140
    },
    {
      "epoch": 2.3804573804573805,
      "grad_norm": 0.0007953010499477386,
      "learning_rate": 0.29939493659662575,
      "loss": 0.2655,
      "num_input_tokens_seen": 874760,
      "step": 1145
    },
    {
      "epoch": 2.390852390852391,
      "grad_norm": 0.0014599555870518088,
      "learning_rate": 0.299389639620348,
      "loss": 0.2714,
      "num_input_tokens_seen": 878472,
      "step": 1150
    },
    {
      "epoch": 2.401247401247401,
      "grad_norm": 0.005799624603241682,
      "learning_rate": 0.29938431960631046,
      "loss": 0.2764,
      "num_input_tokens_seen": 882216,
      "step": 1155
    },
    {
      "epoch": 2.4116424116424118,
      "grad_norm": 0.011294803582131863,
      "learning_rate": 0.2993789765553335,
      "loss": 0.2795,
      "num_input_tokens_seen": 886152,
      "step": 1160
    },
    {
      "epoch": 2.422037422037422,
      "grad_norm": 0.0027259537018835545,
      "learning_rate": 0.2993736104682412,
      "loss": 0.29,
      "num_input_tokens_seen": 890024,
      "step": 1165
    },
    {
      "epoch": 2.4324324324324325,
      "grad_norm": 0.013517460785806179,
      "learning_rate": 0.299368221345861,
      "loss": 0.3209,
      "num_input_tokens_seen": 893896,
      "step": 1170
    },
    {
      "epoch": 2.442827442827443,
      "grad_norm": 0.002101968042552471,
      "learning_rate": 0.29936280918902397,
      "loss": 0.232,
      "num_input_tokens_seen": 897672,
      "step": 1175
    },
    {
      "epoch": 2.453222453222453,
      "grad_norm": 0.001841456163674593,
      "learning_rate": 0.2993573739985648,
      "loss": 0.3339,
      "num_input_tokens_seen": 901416,
      "step": 1180
    },
    {
      "epoch": 2.4636174636174637,
      "grad_norm": 0.0020527432207018137,
      "learning_rate": 0.2993519157753216,
      "loss": 0.2993,
      "num_input_tokens_seen": 905352,
      "step": 1185
    },
    {
      "epoch": 2.474012474012474,
      "grad_norm": 0.0009343986166641116,
      "learning_rate": 0.2993464345201361,
      "loss": 0.284,
      "num_input_tokens_seen": 909096,
      "step": 1190
    },
    {
      "epoch": 2.4844074844074844,
      "grad_norm": 0.0026284432969987392,
      "learning_rate": 0.2993409302338536,
      "loss": 0.2879,
      "num_input_tokens_seen": 913064,
      "step": 1195
    },
    {
      "epoch": 2.494802494802495,
      "grad_norm": 0.0015324263367801905,
      "learning_rate": 0.2993354029173229,
      "loss": 0.2574,
      "num_input_tokens_seen": 916648,
      "step": 1200
    },
    {
      "epoch": 2.494802494802495,
      "eval_loss": 0.28690236806869507,
      "eval_runtime": 13.4086,
      "eval_samples_per_second": 63.84,
      "eval_steps_per_second": 15.96,
      "num_input_tokens_seen": 916648,
      "step": 1200
    },
    {
      "epoch": 2.505197505197505,
      "grad_norm": 0.0026759032625705004,
      "learning_rate": 0.2993298525713965,
      "loss": 0.3513,
      "num_input_tokens_seen": 920520,
      "step": 1205
    },
    {
      "epoch": 2.5155925155925156,
      "grad_norm": 0.001370933372527361,
      "learning_rate": 0.29932427919693017,
      "loss": 0.3356,
      "num_input_tokens_seen": 924456,
      "step": 1210
    },
    {
      "epoch": 2.525987525987526,
      "grad_norm": 0.009586933068931103,
      "learning_rate": 0.2993186827947834,
      "loss": 0.4066,
      "num_input_tokens_seen": 928168,
      "step": 1215
    },
    {
      "epoch": 2.5363825363825363,
      "grad_norm": 0.0022334347013384104,
      "learning_rate": 0.2993130633658194,
      "loss": 0.2587,
      "num_input_tokens_seen": 931944,
      "step": 1220
    },
    {
      "epoch": 2.546777546777547,
      "grad_norm": 0.0017292560078203678,
      "learning_rate": 0.29930742091090456,
      "loss": 0.2827,
      "num_input_tokens_seen": 935656,
      "step": 1225
    },
    {
      "epoch": 2.5571725571725574,
      "grad_norm": 0.0005955182714387774,
      "learning_rate": 0.29930175543090914,
      "loss": 0.238,
      "num_input_tokens_seen": 939528,
      "step": 1230
    },
    {
      "epoch": 2.5675675675675675,
      "grad_norm": 0.00708902720361948,
      "learning_rate": 0.2992960669267068,
      "loss": 0.3014,
      "num_input_tokens_seen": 943240,
      "step": 1235
    },
    {
      "epoch": 2.577962577962578,
      "grad_norm": 0.0014608853962272406,
      "learning_rate": 0.29929035539917476,
      "loss": 0.2899,
      "num_input_tokens_seen": 947176,
      "step": 1240
    },
    {
      "epoch": 2.5883575883575882,
      "grad_norm": 0.0048211850225925446,
      "learning_rate": 0.2992846208491938,
      "loss": 0.2612,
      "num_input_tokens_seen": 950984,
      "step": 1245
    },
    {
      "epoch": 2.598752598752599,
      "grad_norm": 0.001732124132104218,
      "learning_rate": 0.2992788632776483,
      "loss": 0.2984,
      "num_input_tokens_seen": 954792,
      "step": 1250
    },
    {
      "epoch": 2.609147609147609,
      "grad_norm": 0.0011847848072648048,
      "learning_rate": 0.29927308268542613,
      "loss": 0.2928,
      "num_input_tokens_seen": 958632,
      "step": 1255
    },
    {
      "epoch": 2.6195426195426195,
      "grad_norm": 0.004880189895629883,
      "learning_rate": 0.2992672790734187,
      "loss": 0.2862,
      "num_input_tokens_seen": 962472,
      "step": 1260
    },
    {
      "epoch": 2.62993762993763,
      "grad_norm": 0.0031708574388176203,
      "learning_rate": 0.299261452442521,
      "loss": 0.2869,
      "num_input_tokens_seen": 966248,
      "step": 1265
    },
    {
      "epoch": 2.64033264033264,
      "grad_norm": 0.001979518448933959,
      "learning_rate": 0.29925560279363167,
      "loss": 0.2723,
      "num_input_tokens_seen": 970152,
      "step": 1270
    },
    {
      "epoch": 2.6507276507276507,
      "grad_norm": 0.006156404968351126,
      "learning_rate": 0.29924973012765266,
      "loss": 0.255,
      "num_input_tokens_seen": 973992,
      "step": 1275
    },
    {
      "epoch": 2.6611226611226613,
      "grad_norm": 0.003651246428489685,
      "learning_rate": 0.29924383444548974,
      "loss": 0.2463,
      "num_input_tokens_seen": 977704,
      "step": 1280
    },
    {
      "epoch": 2.6715176715176714,
      "grad_norm": 0.0013725819298997521,
      "learning_rate": 0.299237915748052,
      "loss": 0.2569,
      "num_input_tokens_seen": 981384,
      "step": 1285
    },
    {
      "epoch": 2.681912681912682,
      "grad_norm": 0.0019131888402625918,
      "learning_rate": 0.2992319740362522,
      "loss": 0.2642,
      "num_input_tokens_seen": 985288,
      "step": 1290
    },
    {
      "epoch": 2.6923076923076925,
      "grad_norm": 0.010615126229822636,
      "learning_rate": 0.2992260093110066,
      "loss": 0.2727,
      "num_input_tokens_seen": 988968,
      "step": 1295
    },
    {
      "epoch": 2.7027027027027026,
      "grad_norm": 0.0036567626520991325,
      "learning_rate": 0.2992200215732352,
      "loss": 0.2567,
      "num_input_tokens_seen": 992648,
      "step": 1300
    },
    {
      "epoch": 2.713097713097713,
      "grad_norm": 0.003921627998352051,
      "learning_rate": 0.2992140108238611,
      "loss": 0.2882,
      "num_input_tokens_seen": 996456,
      "step": 1305
    },
    {
      "epoch": 2.7234927234927238,
      "grad_norm": 0.008985362946987152,
      "learning_rate": 0.2992079770638115,
      "loss": 0.2483,
      "num_input_tokens_seen": 1000200,
      "step": 1310
    },
    {
      "epoch": 2.733887733887734,
      "grad_norm": 0.001910980325192213,
      "learning_rate": 0.29920192029401677,
      "loss": 0.2847,
      "num_input_tokens_seen": 1004168,
      "step": 1315
    },
    {
      "epoch": 2.7442827442827444,
      "grad_norm": 0.006979361176490784,
      "learning_rate": 0.2991958405154109,
      "loss": 0.3267,
      "num_input_tokens_seen": 1007976,
      "step": 1320
    },
    {
      "epoch": 2.7546777546777546,
      "grad_norm": 0.0023860628716647625,
      "learning_rate": 0.29918973772893154,
      "loss": 0.2462,
      "num_input_tokens_seen": 1011784,
      "step": 1325
    },
    {
      "epoch": 2.765072765072765,
      "grad_norm": 0.0018045055912807584,
      "learning_rate": 0.29918361193551973,
      "loss": 0.2794,
      "num_input_tokens_seen": 1015560,
      "step": 1330
    },
    {
      "epoch": 2.7754677754677752,
      "grad_norm": 0.004008368588984013,
      "learning_rate": 0.29917746313612026,
      "loss": 0.2971,
      "num_input_tokens_seen": 1019176,
      "step": 1335
    },
    {
      "epoch": 2.785862785862786,
      "grad_norm": 0.006423517595976591,
      "learning_rate": 0.29917129133168124,
      "loss": 0.3498,
      "num_input_tokens_seen": 1023016,
      "step": 1340
    },
    {
      "epoch": 2.7962577962577964,
      "grad_norm": 0.004084318410605192,
      "learning_rate": 0.2991650965231546,
      "loss": 0.3143,
      "num_input_tokens_seen": 1026824,
      "step": 1345
    },
    {
      "epoch": 2.8066528066528065,
      "grad_norm": 0.0011377106420695782,
      "learning_rate": 0.29915887871149544,
      "loss": 0.2945,
      "num_input_tokens_seen": 1030440,
      "step": 1350
    },
    {
      "epoch": 2.817047817047817,
      "grad_norm": 0.002099065575748682,
      "learning_rate": 0.2991526378976628,
      "loss": 0.2788,
      "num_input_tokens_seen": 1034248,
      "step": 1355
    },
    {
      "epoch": 2.8274428274428276,
      "grad_norm": 0.005464376416057348,
      "learning_rate": 0.29914637408261896,
      "loss": 0.2706,
      "num_input_tokens_seen": 1038152,
      "step": 1360
    },
    {
      "epoch": 2.8378378378378377,
      "grad_norm": 0.008425338193774223,
      "learning_rate": 0.29914008726733,
      "loss": 0.3003,
      "num_input_tokens_seen": 1041928,
      "step": 1365
    },
    {
      "epoch": 2.8482328482328483,
      "grad_norm": 0.0008336607133969665,
      "learning_rate": 0.2991337774527653,
      "loss": 0.2814,
      "num_input_tokens_seen": 1045608,
      "step": 1370
    },
    {
      "epoch": 2.858627858627859,
      "grad_norm": 0.001973898848518729,
      "learning_rate": 0.2991274446398981,
      "loss": 0.3095,
      "num_input_tokens_seen": 1049544,
      "step": 1375
    },
    {
      "epoch": 2.869022869022869,
      "grad_norm": 0.0015521857421845198,
      "learning_rate": 0.29912108882970484,
      "loss": 0.2808,
      "num_input_tokens_seen": 1053416,
      "step": 1380
    },
    {
      "epoch": 2.8794178794178795,
      "grad_norm": 0.0014273440465331078,
      "learning_rate": 0.2991147100231657,
      "loss": 0.276,
      "num_input_tokens_seen": 1057160,
      "step": 1385
    },
    {
      "epoch": 2.88981288981289,
      "grad_norm": 0.005998368840664625,
      "learning_rate": 0.2991083082212644,
      "loss": 0.2767,
      "num_input_tokens_seen": 1060936,
      "step": 1390
    },
    {
      "epoch": 2.9002079002079,
      "grad_norm": 0.0018427560571581125,
      "learning_rate": 0.2991018834249881,
      "loss": 0.3086,
      "num_input_tokens_seen": 1064808,
      "step": 1395
    },
    {
      "epoch": 2.9106029106029108,
      "grad_norm": 0.0029268977232277393,
      "learning_rate": 0.29909543563532764,
      "loss": 0.2787,
      "num_input_tokens_seen": 1068552,
      "step": 1400
    },
    {
      "epoch": 2.9106029106029108,
      "eval_loss": 0.2539924681186676,
      "eval_runtime": 13.4137,
      "eval_samples_per_second": 63.815,
      "eval_steps_per_second": 15.954,
      "num_input_tokens_seen": 1068552,
      "step": 1400
    },
    {
      "epoch": 2.920997920997921,
      "grad_norm": 0.004619130399078131,
      "learning_rate": 0.29908896485327746,
      "loss": 0.2901,
      "num_input_tokens_seen": 1072232,
      "step": 1405
    },
    {
      "epoch": 2.9313929313929314,
      "grad_norm": 0.0013719034614041448,
      "learning_rate": 0.29908247107983527,
      "loss": 0.2775,
      "num_input_tokens_seen": 1076072,
      "step": 1410
    },
    {
      "epoch": 2.9417879417879416,
      "grad_norm": 0.008435007184743881,
      "learning_rate": 0.29907595431600253,
      "loss": 0.2615,
      "num_input_tokens_seen": 1079880,
      "step": 1415
    },
    {
      "epoch": 2.952182952182952,
      "grad_norm": 0.0015799006214365363,
      "learning_rate": 0.29906941456278424,
      "loss": 0.2552,
      "num_input_tokens_seen": 1083752,
      "step": 1420
    },
    {
      "epoch": 2.9625779625779627,
      "grad_norm": 0.0013041372876614332,
      "learning_rate": 0.2990628518211889,
      "loss": 0.2795,
      "num_input_tokens_seen": 1087656,
      "step": 1425
    },
    {
      "epoch": 2.972972972972973,
      "grad_norm": 0.009832659736275673,
      "learning_rate": 0.2990562660922286,
      "loss": 0.3622,
      "num_input_tokens_seen": 1091560,
      "step": 1430
    },
    {
      "epoch": 2.9833679833679834,
      "grad_norm": 0.010085932910442352,
      "learning_rate": 0.2990496573769189,
      "loss": 0.3336,
      "num_input_tokens_seen": 1095336,
      "step": 1435
    },
    {
      "epoch": 2.993762993762994,
      "grad_norm": 0.0038716529961675406,
      "learning_rate": 0.29904302567627894,
      "loss": 0.3155,
      "num_input_tokens_seen": 1099016,
      "step": 1440
    },
    {
      "epoch": 3.004158004158004,
      "grad_norm": 0.003527118358761072,
      "learning_rate": 0.2990363709913314,
      "loss": 0.2757,
      "num_input_tokens_seen": 1102784,
      "step": 1445
    },
    {
      "epoch": 3.0145530145530146,
      "grad_norm": 0.003817586461082101,
      "learning_rate": 0.29902969332310264,
      "loss": 0.2956,
      "num_input_tokens_seen": 1106656,
      "step": 1450
    },
    {
      "epoch": 3.024948024948025,
      "grad_norm": 0.004686552565544844,
      "learning_rate": 0.2990229926726223,
      "loss": 0.3025,
      "num_input_tokens_seen": 1110464,
      "step": 1455
    },
    {
      "epoch": 3.0353430353430353,
      "grad_norm": 0.006018358748406172,
      "learning_rate": 0.29901626904092365,
      "loss": 0.3118,
      "num_input_tokens_seen": 1114400,
      "step": 1460
    },
    {
      "epoch": 3.045738045738046,
      "grad_norm": 0.002419866854324937,
      "learning_rate": 0.2990095224290438,
      "loss": 0.3199,
      "num_input_tokens_seen": 1118080,
      "step": 1465
    },
    {
      "epoch": 3.056133056133056,
      "grad_norm": 0.0040710545144975185,
      "learning_rate": 0.29900275283802297,
      "loss": 0.3269,
      "num_input_tokens_seen": 1121792,
      "step": 1470
    },
    {
      "epoch": 3.0665280665280665,
      "grad_norm": 0.002144722267985344,
      "learning_rate": 0.2989959602689051,
      "loss": 0.3098,
      "num_input_tokens_seen": 1125600,
      "step": 1475
    },
    {
      "epoch": 3.076923076923077,
      "grad_norm": 0.002598235383629799,
      "learning_rate": 0.2989891447227379,
      "loss": 0.3198,
      "num_input_tokens_seen": 1129472,
      "step": 1480
    },
    {
      "epoch": 3.087318087318087,
      "grad_norm": 0.009648177772760391,
      "learning_rate": 0.29898230620057215,
      "loss": 0.293,
      "num_input_tokens_seen": 1133408,
      "step": 1485
    },
    {
      "epoch": 3.0977130977130978,
      "grad_norm": 0.005802724976092577,
      "learning_rate": 0.2989754447034626,
      "loss": 0.2905,
      "num_input_tokens_seen": 1137280,
      "step": 1490
    },
    {
      "epoch": 3.108108108108108,
      "grad_norm": 0.010249296203255653,
      "learning_rate": 0.2989685602324673,
      "loss": 0.2925,
      "num_input_tokens_seen": 1141184,
      "step": 1495
    },
    {
      "epoch": 3.1185031185031185,
      "grad_norm": 0.00676561426371336,
      "learning_rate": 0.298961652788648,
      "loss": 0.2932,
      "num_input_tokens_seen": 1145056,
      "step": 1500
    },
    {
      "epoch": 3.128898128898129,
      "grad_norm": 0.00040619613719172776,
      "learning_rate": 0.29895472237306986,
      "loss": 0.2533,
      "num_input_tokens_seen": 1148832,
      "step": 1505
    },
    {
      "epoch": 3.139293139293139,
      "grad_norm": 0.0019757540430873632,
      "learning_rate": 0.29894776898680164,
      "loss": 0.2765,
      "num_input_tokens_seen": 1152640,
      "step": 1510
    },
    {
      "epoch": 3.1496881496881497,
      "grad_norm": 0.0011986758327111602,
      "learning_rate": 0.29894079263091566,
      "loss": 0.3014,
      "num_input_tokens_seen": 1156512,
      "step": 1515
    },
    {
      "epoch": 3.1600831600831603,
      "grad_norm": 0.006642407737672329,
      "learning_rate": 0.2989337933064877,
      "loss": 0.2755,
      "num_input_tokens_seen": 1160192,
      "step": 1520
    },
    {
      "epoch": 3.1704781704781704,
      "grad_norm": 0.0009561611223034561,
      "learning_rate": 0.29892677101459725,
      "loss": 0.2558,
      "num_input_tokens_seen": 1164128,
      "step": 1525
    },
    {
      "epoch": 3.180873180873181,
      "grad_norm": 0.003825856139883399,
      "learning_rate": 0.2989197257563272,
      "loss": 0.2643,
      "num_input_tokens_seen": 1167936,
      "step": 1530
    },
    {
      "epoch": 3.1912681912681915,
      "grad_norm": 0.0021503549069166183,
      "learning_rate": 0.2989126575327639,
      "loss": 0.287,
      "num_input_tokens_seen": 1171648,
      "step": 1535
    },
    {
      "epoch": 3.2016632016632016,
      "grad_norm": 0.005350227002054453,
      "learning_rate": 0.29890556634499754,
      "loss": 0.282,
      "num_input_tokens_seen": 1175456,
      "step": 1540
    },
    {
      "epoch": 3.212058212058212,
      "grad_norm": 0.012126617133617401,
      "learning_rate": 0.2988984521941216,
      "loss": 0.3208,
      "num_input_tokens_seen": 1179360,
      "step": 1545
    },
    {
      "epoch": 3.2224532224532223,
      "grad_norm": 0.004827439319342375,
      "learning_rate": 0.29889131508123307,
      "loss": 0.3076,
      "num_input_tokens_seen": 1183264,
      "step": 1550
    },
    {
      "epoch": 3.232848232848233,
      "grad_norm": 0.005428845528513193,
      "learning_rate": 0.2988841550074327,
      "loss": 0.3239,
      "num_input_tokens_seen": 1187040,
      "step": 1555
    },
    {
      "epoch": 3.2432432432432434,
      "grad_norm": 0.0013735729735344648,
      "learning_rate": 0.2988769719738246,
      "loss": 0.2934,
      "num_input_tokens_seen": 1190656,
      "step": 1560
    },
    {
      "epoch": 3.2536382536382535,
      "grad_norm": 0.0024517911951988935,
      "learning_rate": 0.29886976598151666,
      "loss": 0.2814,
      "num_input_tokens_seen": 1194400,
      "step": 1565
    },
    {
      "epoch": 3.264033264033264,
      "grad_norm": 0.001760362065397203,
      "learning_rate": 0.29886253703161986,
      "loss": 0.2562,
      "num_input_tokens_seen": 1198016,
      "step": 1570
    },
    {
      "epoch": 3.274428274428274,
      "grad_norm": 0.002623186679556966,
      "learning_rate": 0.29885528512524917,
      "loss": 0.2422,
      "num_input_tokens_seen": 1201792,
      "step": 1575
    },
    {
      "epoch": 3.284823284823285,
      "grad_norm": 0.0008549263002350926,
      "learning_rate": 0.29884801026352287,
      "loss": 0.2888,
      "num_input_tokens_seen": 1205728,
      "step": 1580
    },
    {
      "epoch": 3.2952182952182953,
      "grad_norm": 0.007070057559758425,
      "learning_rate": 0.2988407124475629,
      "loss": 0.2951,
      "num_input_tokens_seen": 1209664,
      "step": 1585
    },
    {
      "epoch": 3.3056133056133055,
      "grad_norm": 0.0041161710396409035,
      "learning_rate": 0.2988333916784945,
      "loss": 0.2755,
      "num_input_tokens_seen": 1213408,
      "step": 1590
    },
    {
      "epoch": 3.316008316008316,
      "grad_norm": 0.001973289530724287,
      "learning_rate": 0.2988260479574468,
      "loss": 0.2776,
      "num_input_tokens_seen": 1217152,
      "step": 1595
    },
    {
      "epoch": 3.3264033264033266,
      "grad_norm": 0.0008285017102025449,
      "learning_rate": 0.2988186812855523,
      "loss": 0.2469,
      "num_input_tokens_seen": 1220928,
      "step": 1600
    },
    {
      "epoch": 3.3264033264033266,
      "eval_loss": 0.25966429710388184,
      "eval_runtime": 13.4126,
      "eval_samples_per_second": 63.821,
      "eval_steps_per_second": 15.955,
      "num_input_tokens_seen": 1220928,
      "step": 1600
    },
    {
      "epoch": 3.3367983367983367,
      "grad_norm": 0.0073862276040017605,
      "learning_rate": 0.29881129166394693,
      "loss": 0.3005,
      "num_input_tokens_seen": 1224736,
      "step": 1605
    },
    {
      "epoch": 3.3471933471933473,
      "grad_norm": 0.001586039667017758,
      "learning_rate": 0.29880387909377026,
      "loss": 0.2683,
      "num_input_tokens_seen": 1228512,
      "step": 1610
    },
    {
      "epoch": 3.357588357588358,
      "grad_norm": 0.002718248637393117,
      "learning_rate": 0.2987964435761655,
      "loss": 0.3294,
      "num_input_tokens_seen": 1232416,
      "step": 1615
    },
    {
      "epoch": 3.367983367983368,
      "grad_norm": 0.0036075536627322435,
      "learning_rate": 0.29878898511227925,
      "loss": 0.2854,
      "num_input_tokens_seen": 1236096,
      "step": 1620
    },
    {
      "epoch": 3.3783783783783785,
      "grad_norm": 0.004828325472772121,
      "learning_rate": 0.2987815037032617,
      "loss": 0.2781,
      "num_input_tokens_seen": 1239840,
      "step": 1625
    },
    {
      "epoch": 3.3887733887733886,
      "grad_norm": 0.008463186211884022,
      "learning_rate": 0.29877399935026655,
      "loss": 0.2754,
      "num_input_tokens_seen": 1243648,
      "step": 1630
    },
    {
      "epoch": 3.399168399168399,
      "grad_norm": 0.002913767471909523,
      "learning_rate": 0.2987664720544511,
      "loss": 0.2612,
      "num_input_tokens_seen": 1247584,
      "step": 1635
    },
    {
      "epoch": 3.4095634095634098,
      "grad_norm": 0.004578461870551109,
      "learning_rate": 0.2987589218169761,
      "loss": 0.2937,
      "num_input_tokens_seen": 1251488,
      "step": 1640
    },
    {
      "epoch": 3.41995841995842,
      "grad_norm": 0.006026610732078552,
      "learning_rate": 0.29875134863900604,
      "loss": 0.2726,
      "num_input_tokens_seen": 1255328,
      "step": 1645
    },
    {
      "epoch": 3.4303534303534304,
      "grad_norm": 0.0063947043381631374,
      "learning_rate": 0.29874375252170865,
      "loss": 0.3089,
      "num_input_tokens_seen": 1259104,
      "step": 1650
    },
    {
      "epoch": 3.4407484407484406,
      "grad_norm": 0.006429970730096102,
      "learning_rate": 0.2987361334662553,
      "loss": 0.2674,
      "num_input_tokens_seen": 1262848,
      "step": 1655
    },
    {
      "epoch": 3.451143451143451,
      "grad_norm": 0.00415390869602561,
      "learning_rate": 0.29872849147382113,
      "loss": 0.2886,
      "num_input_tokens_seen": 1266624,
      "step": 1660
    },
    {
      "epoch": 3.4615384615384617,
      "grad_norm": 0.005241432227194309,
      "learning_rate": 0.2987208265455845,
      "loss": 0.2706,
      "num_input_tokens_seen": 1270400,
      "step": 1665
    },
    {
      "epoch": 3.471933471933472,
      "grad_norm": 0.0028793688397854567,
      "learning_rate": 0.29871313868272753,
      "loss": 0.269,
      "num_input_tokens_seen": 1274176,
      "step": 1670
    },
    {
      "epoch": 3.4823284823284824,
      "grad_norm": 0.0014138300903141499,
      "learning_rate": 0.29870542788643567,
      "loss": 0.2809,
      "num_input_tokens_seen": 1277792,
      "step": 1675
    },
    {
      "epoch": 3.492723492723493,
      "grad_norm": 0.008904961869120598,
      "learning_rate": 0.2986976941578981,
      "loss": 0.2915,
      "num_input_tokens_seen": 1281600,
      "step": 1680
    },
    {
      "epoch": 3.503118503118503,
      "grad_norm": 0.008952009491622448,
      "learning_rate": 0.29868993749830747,
      "loss": 0.2888,
      "num_input_tokens_seen": 1285376,
      "step": 1685
    },
    {
      "epoch": 3.5135135135135136,
      "grad_norm": 0.003298682626336813,
      "learning_rate": 0.2986821579088598,
      "loss": 0.2703,
      "num_input_tokens_seen": 1289216,
      "step": 1690
    },
    {
      "epoch": 3.523908523908524,
      "grad_norm": 0.0031915546860545874,
      "learning_rate": 0.29867435539075504,
      "loss": 0.2975,
      "num_input_tokens_seen": 1293056,
      "step": 1695
    },
    {
      "epoch": 3.5343035343035343,
      "grad_norm": 0.00769947050139308,
      "learning_rate": 0.2986665299451963,
      "loss": 0.284,
      "num_input_tokens_seen": 1296992,
      "step": 1700
    },
    {
      "epoch": 3.544698544698545,
      "grad_norm": 0.011347084306180477,
      "learning_rate": 0.29865868157339037,
      "loss": 0.3494,
      "num_input_tokens_seen": 1300832,
      "step": 1705
    },
    {
      "epoch": 3.555093555093555,
      "grad_norm": 0.009765577502548695,
      "learning_rate": 0.2986508102765476,
      "loss": 0.3548,
      "num_input_tokens_seen": 1304704,
      "step": 1710
    },
    {
      "epoch": 3.5654885654885655,
      "grad_norm": 0.0029405849054455757,
      "learning_rate": 0.2986429160558818,
      "loss": 0.294,
      "num_input_tokens_seen": 1308512,
      "step": 1715
    },
    {
      "epoch": 3.5758835758835756,
      "grad_norm": 0.001207957393489778,
      "learning_rate": 0.2986349989126104,
      "loss": 0.3166,
      "num_input_tokens_seen": 1312320,
      "step": 1720
    },
    {
      "epoch": 3.586278586278586,
      "grad_norm": 0.0012622280046343803,
      "learning_rate": 0.29862705884795426,
      "loss": 0.3154,
      "num_input_tokens_seen": 1316320,
      "step": 1725
    },
    {
      "epoch": 3.5966735966735968,
      "grad_norm": 0.0022347033955156803,
      "learning_rate": 0.2986190958631379,
      "loss": 0.2519,
      "num_input_tokens_seen": 1320192,
      "step": 1730
    },
    {
      "epoch": 3.607068607068607,
      "grad_norm": 0.0009959460003301501,
      "learning_rate": 0.29861110995938933,
      "loss": 0.2516,
      "num_input_tokens_seen": 1323936,
      "step": 1735
    },
    {
      "epoch": 3.6174636174636174,
      "grad_norm": 0.0010346631752327085,
      "learning_rate": 0.29860310113794,
      "loss": 0.2589,
      "num_input_tokens_seen": 1327904,
      "step": 1740
    },
    {
      "epoch": 3.627858627858628,
      "grad_norm": 0.0015725964913144708,
      "learning_rate": 0.29859506940002506,
      "loss": 0.2549,
      "num_input_tokens_seen": 1331616,
      "step": 1745
    },
    {
      "epoch": 3.638253638253638,
      "grad_norm": 0.0025410125963389874,
      "learning_rate": 0.298587014746883,
      "loss": 0.2893,
      "num_input_tokens_seen": 1335488,
      "step": 1750
    },
    {
      "epoch": 3.6486486486486487,
      "grad_norm": 0.0008813805179670453,
      "learning_rate": 0.298578937179756,
      "loss": 0.2843,
      "num_input_tokens_seen": 1339296,
      "step": 1755
    },
    {
      "epoch": 3.6590436590436592,
      "grad_norm": 0.0044625853188335896,
      "learning_rate": 0.29857083669988976,
      "loss": 0.2735,
      "num_input_tokens_seen": 1343040,
      "step": 1760
    },
    {
      "epoch": 3.6694386694386694,
      "grad_norm": 0.003523599123582244,
      "learning_rate": 0.29856271330853346,
      "loss": 0.3511,
      "num_input_tokens_seen": 1346720,
      "step": 1765
    },
    {
      "epoch": 3.67983367983368,
      "grad_norm": 0.004366528708487749,
      "learning_rate": 0.2985545670069398,
      "loss": 0.2915,
      "num_input_tokens_seen": 1350720,
      "step": 1770
    },
    {
      "epoch": 3.6902286902286905,
      "grad_norm": 0.005080975126475096,
      "learning_rate": 0.29854639779636505,
      "loss": 0.2513,
      "num_input_tokens_seen": 1354688,
      "step": 1775
    },
    {
      "epoch": 3.7006237006237006,
      "grad_norm": 0.002465195022523403,
      "learning_rate": 0.298538205678069,
      "loss": 0.2846,
      "num_input_tokens_seen": 1358560,
      "step": 1780
    },
    {
      "epoch": 3.711018711018711,
      "grad_norm": 0.005550271365791559,
      "learning_rate": 0.298529990653315,
      "loss": 0.2826,
      "num_input_tokens_seen": 1362528,
      "step": 1785
    },
    {
      "epoch": 3.7214137214137213,
      "grad_norm": 0.0009754779748618603,
      "learning_rate": 0.29852175272336984,
      "loss": 0.2833,
      "num_input_tokens_seen": 1366240,
      "step": 1790
    },
    {
      "epoch": 3.731808731808732,
      "grad_norm": 0.0019146411214023829,
      "learning_rate": 0.29851349188950405,
      "loss": 0.2937,
      "num_input_tokens_seen": 1370080,
      "step": 1795
    },
    {
      "epoch": 3.742203742203742,
      "grad_norm": 0.002858969150111079,
      "learning_rate": 0.2985052081529914,
      "loss": 0.2798,
      "num_input_tokens_seen": 1373952,
      "step": 1800
    },
    {
      "epoch": 3.742203742203742,
      "eval_loss": 0.2820360064506531,
      "eval_runtime": 13.4935,
      "eval_samples_per_second": 63.438,
      "eval_steps_per_second": 15.859,
      "num_input_tokens_seen": 1373952,
      "step": 1800
    },
    {
      "epoch": 3.7525987525987525,
      "grad_norm": 0.0032404414378106594,
      "learning_rate": 0.29849690151510944,
      "loss": 0.2722,
      "num_input_tokens_seen": 1377664,
      "step": 1805
    },
    {
      "epoch": 3.762993762993763,
      "grad_norm": 0.00905431155115366,
      "learning_rate": 0.2984885719771392,
      "loss": 0.3257,
      "num_input_tokens_seen": 1381376,
      "step": 1810
    },
    {
      "epoch": 3.773388773388773,
      "grad_norm": 0.004402908496558666,
      "learning_rate": 0.2984802195403651,
      "loss": 0.2728,
      "num_input_tokens_seen": 1385248,
      "step": 1815
    },
    {
      "epoch": 3.7837837837837838,
      "grad_norm": 0.0018533398397266865,
      "learning_rate": 0.2984718442060752,
      "loss": 0.3062,
      "num_input_tokens_seen": 1389056,
      "step": 1820
    },
    {
      "epoch": 3.7941787941787943,
      "grad_norm": 0.0006382323917932808,
      "learning_rate": 0.2984634459755611,
      "loss": 0.2994,
      "num_input_tokens_seen": 1392832,
      "step": 1825
    },
    {
      "epoch": 3.8045738045738045,
      "grad_norm": 0.004639607388526201,
      "learning_rate": 0.29845502485011793,
      "loss": 0.3047,
      "num_input_tokens_seen": 1396576,
      "step": 1830
    },
    {
      "epoch": 3.814968814968815,
      "grad_norm": 0.007282274309545755,
      "learning_rate": 0.2984465808310444,
      "loss": 0.2881,
      "num_input_tokens_seen": 1400384,
      "step": 1835
    },
    {
      "epoch": 3.8253638253638256,
      "grad_norm": 0.003735373029485345,
      "learning_rate": 0.29843811391964253,
      "loss": 0.3207,
      "num_input_tokens_seen": 1404256,
      "step": 1840
    },
    {
      "epoch": 3.8357588357588357,
      "grad_norm": 0.00577420461922884,
      "learning_rate": 0.2984296241172182,
      "loss": 0.2973,
      "num_input_tokens_seen": 1408064,
      "step": 1845
    },
    {
      "epoch": 3.8461538461538463,
      "grad_norm": 0.0009359292453154922,
      "learning_rate": 0.29842111142508043,
      "loss": 0.2966,
      "num_input_tokens_seen": 1411840,
      "step": 1850
    },
    {
      "epoch": 3.856548856548857,
      "grad_norm": 0.0036411641631275415,
      "learning_rate": 0.29841257584454217,
      "loss": 0.2721,
      "num_input_tokens_seen": 1415648,
      "step": 1855
    },
    {
      "epoch": 3.866943866943867,
      "grad_norm": 0.006938981357961893,
      "learning_rate": 0.29840401737691963,
      "loss": 0.3088,
      "num_input_tokens_seen": 1419328,
      "step": 1860
    },
    {
      "epoch": 3.8773388773388775,
      "grad_norm": 0.0011995135573670268,
      "learning_rate": 0.29839543602353263,
      "loss": 0.2921,
      "num_input_tokens_seen": 1423200,
      "step": 1865
    },
    {
      "epoch": 3.8877338877338876,
      "grad_norm": 0.0016656472580507398,
      "learning_rate": 0.2983868317857046,
      "loss": 0.2475,
      "num_input_tokens_seen": 1427072,
      "step": 1870
    },
    {
      "epoch": 3.898128898128898,
      "grad_norm": 0.002118919976055622,
      "learning_rate": 0.2983782046647623,
      "loss": 0.3024,
      "num_input_tokens_seen": 1430976,
      "step": 1875
    },
    {
      "epoch": 3.9085239085239083,
      "grad_norm": 0.005580019671469927,
      "learning_rate": 0.2983695546620362,
      "loss": 0.2788,
      "num_input_tokens_seen": 1434720,
      "step": 1880
    },
    {
      "epoch": 3.918918918918919,
      "grad_norm": 0.0051811556331813335,
      "learning_rate": 0.2983608817788603,
      "loss": 0.3212,
      "num_input_tokens_seen": 1438432,
      "step": 1885
    },
    {
      "epoch": 3.9293139293139294,
      "grad_norm": 0.0028633195906877518,
      "learning_rate": 0.29835218601657193,
      "loss": 0.2815,
      "num_input_tokens_seen": 1442272,
      "step": 1890
    },
    {
      "epoch": 3.9397089397089395,
      "grad_norm": 0.0019312766380608082,
      "learning_rate": 0.29834346737651224,
      "loss": 0.2888,
      "num_input_tokens_seen": 1446016,
      "step": 1895
    },
    {
      "epoch": 3.95010395010395,
      "grad_norm": 0.0031070923432707787,
      "learning_rate": 0.29833472586002563,
      "loss": 0.2875,
      "num_input_tokens_seen": 1449920,
      "step": 1900
    },
    {
      "epoch": 3.9604989604989607,
      "grad_norm": 0.00412020506337285,
      "learning_rate": 0.29832596146846024,
      "loss": 0.2841,
      "num_input_tokens_seen": 1453792,
      "step": 1905
    },
    {
      "epoch": 3.970893970893971,
      "grad_norm": 0.0054773990996181965,
      "learning_rate": 0.2983171742031676,
      "loss": 0.2991,
      "num_input_tokens_seen": 1457536,
      "step": 1910
    },
    {
      "epoch": 3.9812889812889813,
      "grad_norm": 0.002289661904796958,
      "learning_rate": 0.2983083640655028,
      "loss": 0.2544,
      "num_input_tokens_seen": 1461408,
      "step": 1915
    },
    {
      "epoch": 3.991683991683992,
      "grad_norm": 0.005125665571540594,
      "learning_rate": 0.29829953105682455,
      "loss": 0.2994,
      "num_input_tokens_seen": 1465376,
      "step": 1920
    },
    {
      "epoch": 4.002079002079002,
      "grad_norm": 0.002368838293477893,
      "learning_rate": 0.29829067517849495,
      "loss": 0.2736,
      "num_input_tokens_seen": 1469320,
      "step": 1925
    },
    {
      "epoch": 4.012474012474012,
      "grad_norm": 0.002574007725343108,
      "learning_rate": 0.2982817964318797,
      "loss": 0.2693,
      "num_input_tokens_seen": 1473192,
      "step": 1930
    },
    {
      "epoch": 4.022869022869023,
      "grad_norm": 0.0022572060115635395,
      "learning_rate": 0.298272894818348,
      "loss": 0.2654,
      "num_input_tokens_seen": 1476840,
      "step": 1935
    },
    {
      "epoch": 4.033264033264033,
      "grad_norm": 0.00462284404784441,
      "learning_rate": 0.2982639703392726,
      "loss": 0.3062,
      "num_input_tokens_seen": 1480584,
      "step": 1940
    },
    {
      "epoch": 4.043659043659043,
      "grad_norm": 0.000538617605343461,
      "learning_rate": 0.29825502299602974,
      "loss": 0.2708,
      "num_input_tokens_seen": 1484328,
      "step": 1945
    },
    {
      "epoch": 4.054054054054054,
      "grad_norm": 0.006655883975327015,
      "learning_rate": 0.2982460527899993,
      "loss": 0.2983,
      "num_input_tokens_seen": 1488328,
      "step": 1950
    },
    {
      "epoch": 4.0644490644490645,
      "grad_norm": 0.0034301530104130507,
      "learning_rate": 0.29823705972256453,
      "loss": 0.2851,
      "num_input_tokens_seen": 1492008,
      "step": 1955
    },
    {
      "epoch": 4.074844074844075,
      "grad_norm": 0.003470534924417734,
      "learning_rate": 0.2982280437951123,
      "loss": 0.2993,
      "num_input_tokens_seen": 1495880,
      "step": 1960
    },
    {
      "epoch": 4.085239085239086,
      "grad_norm": 0.0010923275258392096,
      "learning_rate": 0.298219005009033,
      "loss": 0.2746,
      "num_input_tokens_seen": 1499784,
      "step": 1965
    },
    {
      "epoch": 4.095634095634096,
      "grad_norm": 0.0030100836884230375,
      "learning_rate": 0.29820994336572043,
      "loss": 0.2968,
      "num_input_tokens_seen": 1503496,
      "step": 1970
    },
    {
      "epoch": 4.106029106029106,
      "grad_norm": 0.0008962377905845642,
      "learning_rate": 0.2982008588665721,
      "loss": 0.2863,
      "num_input_tokens_seen": 1507240,
      "step": 1975
    },
    {
      "epoch": 4.116424116424117,
      "grad_norm": 0.0038430250715464354,
      "learning_rate": 0.2981917515129889,
      "loss": 0.3135,
      "num_input_tokens_seen": 1511016,
      "step": 1980
    },
    {
      "epoch": 4.126819126819127,
      "grad_norm": 0.00257163611240685,
      "learning_rate": 0.2981826213063753,
      "loss": 0.2918,
      "num_input_tokens_seen": 1514792,
      "step": 1985
    },
    {
      "epoch": 4.137214137214137,
      "grad_norm": 0.0024485746398568153,
      "learning_rate": 0.2981734682481394,
      "loss": 0.2808,
      "num_input_tokens_seen": 1518600,
      "step": 1990
    },
    {
      "epoch": 4.147609147609147,
      "grad_norm": 0.0027945362962782383,
      "learning_rate": 0.29816429233969255,
      "loss": 0.2654,
      "num_input_tokens_seen": 1522408,
      "step": 1995
    },
    {
      "epoch": 4.158004158004158,
      "grad_norm": 0.0020251369569450617,
      "learning_rate": 0.2981550935824499,
      "loss": 0.2612,
      "num_input_tokens_seen": 1526312,
      "step": 2000
    },
    {
      "epoch": 4.158004158004158,
      "eval_loss": 0.2517845630645752,
      "eval_runtime": 13.4608,
      "eval_samples_per_second": 63.592,
      "eval_steps_per_second": 15.898,
      "num_input_tokens_seen": 1526312,
      "step": 2000
    },
    {
      "epoch": 4.168399168399168,
      "grad_norm": 0.006499717012047768,
      "learning_rate": 0.29814587197783,
      "loss": 0.2769,
      "num_input_tokens_seen": 1530184,
      "step": 2005
    },
    {
      "epoch": 4.1787941787941785,
      "grad_norm": 0.005391437094658613,
      "learning_rate": 0.29813662752725495,
      "loss": 0.2777,
      "num_input_tokens_seen": 1534024,
      "step": 2010
    },
    {
      "epoch": 4.1891891891891895,
      "grad_norm": 0.009302061051130295,
      "learning_rate": 0.29812736023215025,
      "loss": 0.3033,
      "num_input_tokens_seen": 1537672,
      "step": 2015
    },
    {
      "epoch": 4.1995841995842,
      "grad_norm": 0.007286230102181435,
      "learning_rate": 0.29811807009394514,
      "loss": 0.3335,
      "num_input_tokens_seen": 1541672,
      "step": 2020
    },
    {
      "epoch": 4.20997920997921,
      "grad_norm": 0.0037706149742007256,
      "learning_rate": 0.2981087571140723,
      "loss": 0.2943,
      "num_input_tokens_seen": 1545416,
      "step": 2025
    },
    {
      "epoch": 4.220374220374221,
      "grad_norm": 0.0032468908466398716,
      "learning_rate": 0.2980994212939678,
      "loss": 0.2753,
      "num_input_tokens_seen": 1549256,
      "step": 2030
    },
    {
      "epoch": 4.230769230769231,
      "grad_norm": 0.004540004767477512,
      "learning_rate": 0.2980900626350715,
      "loss": 0.2288,
      "num_input_tokens_seen": 1552968,
      "step": 2035
    },
    {
      "epoch": 4.241164241164241,
      "grad_norm": 0.0013811299577355385,
      "learning_rate": 0.29808068113882646,
      "loss": 0.2762,
      "num_input_tokens_seen": 1556808,
      "step": 2040
    },
    {
      "epoch": 4.251559251559252,
      "grad_norm": 0.0029561335686594248,
      "learning_rate": 0.2980712768066795,
      "loss": 0.3181,
      "num_input_tokens_seen": 1560552,
      "step": 2045
    },
    {
      "epoch": 4.261954261954262,
      "grad_norm": 0.0013340101577341557,
      "learning_rate": 0.2980618496400809,
      "loss": 0.2793,
      "num_input_tokens_seen": 1564392,
      "step": 2050
    },
    {
      "epoch": 4.272349272349272,
      "grad_norm": 0.003951576072722673,
      "learning_rate": 0.2980523996404844,
      "loss": 0.3006,
      "num_input_tokens_seen": 1568104,
      "step": 2055
    },
    {
      "epoch": 4.282744282744282,
      "grad_norm": 0.005777365528047085,
      "learning_rate": 0.2980429268093473,
      "loss": 0.2779,
      "num_input_tokens_seen": 1571880,
      "step": 2060
    },
    {
      "epoch": 4.293139293139293,
      "grad_norm": 0.0010954671306535602,
      "learning_rate": 0.29803343114813047,
      "loss": 0.2699,
      "num_input_tokens_seen": 1575624,
      "step": 2065
    },
    {
      "epoch": 4.303534303534303,
      "grad_norm": 0.003955200780183077,
      "learning_rate": 0.2980239126582983,
      "loss": 0.3168,
      "num_input_tokens_seen": 1579368,
      "step": 2070
    },
    {
      "epoch": 4.313929313929314,
      "grad_norm": 0.002212780062109232,
      "learning_rate": 0.2980143713413186,
      "loss": 0.3223,
      "num_input_tokens_seen": 1583176,
      "step": 2075
    },
    {
      "epoch": 4.324324324324325,
      "grad_norm": 0.004549290519207716,
      "learning_rate": 0.29800480719866274,
      "loss": 0.3243,
      "num_input_tokens_seen": 1586952,
      "step": 2080
    },
    {
      "epoch": 4.334719334719335,
      "grad_norm": 0.0031297740060836077,
      "learning_rate": 0.2979952202318057,
      "loss": 0.2759,
      "num_input_tokens_seen": 1590792,
      "step": 2085
    },
    {
      "epoch": 4.345114345114345,
      "grad_norm": 0.002063532592728734,
      "learning_rate": 0.2979856104422259,
      "loss": 0.2945,
      "num_input_tokens_seen": 1594504,
      "step": 2090
    },
    {
      "epoch": 4.355509355509356,
      "grad_norm": 0.004682120401412249,
      "learning_rate": 0.2979759778314052,
      "loss": 0.2836,
      "num_input_tokens_seen": 1598280,
      "step": 2095
    },
    {
      "epoch": 4.365904365904366,
      "grad_norm": 0.0016346347983926535,
      "learning_rate": 0.2979663224008292,
      "loss": 0.2672,
      "num_input_tokens_seen": 1601928,
      "step": 2100
    },
    {
      "epoch": 4.376299376299376,
      "grad_norm": 0.0013662363635376096,
      "learning_rate": 0.2979566441519868,
      "loss": 0.2608,
      "num_input_tokens_seen": 1605832,
      "step": 2105
    },
    {
      "epoch": 4.386694386694387,
      "grad_norm": 0.0035485406406223774,
      "learning_rate": 0.29794694308637054,
      "loss": 0.2626,
      "num_input_tokens_seen": 1609736,
      "step": 2110
    },
    {
      "epoch": 4.397089397089397,
      "grad_norm": 0.004933365620672703,
      "learning_rate": 0.2979372192054764,
      "loss": 0.2703,
      "num_input_tokens_seen": 1613512,
      "step": 2115
    },
    {
      "epoch": 4.407484407484407,
      "grad_norm": 0.0011550626950338483,
      "learning_rate": 0.297927472510804,
      "loss": 0.2841,
      "num_input_tokens_seen": 1617320,
      "step": 2120
    },
    {
      "epoch": 4.417879417879418,
      "grad_norm": 0.002945635002106428,
      "learning_rate": 0.29791770300385634,
      "loss": 0.2596,
      "num_input_tokens_seen": 1621096,
      "step": 2125
    },
    {
      "epoch": 4.428274428274428,
      "grad_norm": 0.006348751485347748,
      "learning_rate": 0.29790791068614003,
      "loss": 0.276,
      "num_input_tokens_seen": 1624936,
      "step": 2130
    },
    {
      "epoch": 4.4386694386694385,
      "grad_norm": 0.004368406254798174,
      "learning_rate": 0.2978980955591652,
      "loss": 0.2913,
      "num_input_tokens_seen": 1628904,
      "step": 2135
    },
    {
      "epoch": 4.4490644490644495,
      "grad_norm": 0.002178372349590063,
      "learning_rate": 0.2978882576244454,
      "loss": 0.2806,
      "num_input_tokens_seen": 1632712,
      "step": 2140
    },
    {
      "epoch": 4.45945945945946,
      "grad_norm": 0.0010187494335696101,
      "learning_rate": 0.2978783968834978,
      "loss": 0.2463,
      "num_input_tokens_seen": 1636712,
      "step": 2145
    },
    {
      "epoch": 4.46985446985447,
      "grad_norm": 0.003212394891306758,
      "learning_rate": 0.29786851333784303,
      "loss": 0.2748,
      "num_input_tokens_seen": 1640552,
      "step": 2150
    },
    {
      "epoch": 4.48024948024948,
      "grad_norm": 0.004137763287872076,
      "learning_rate": 0.2978586069890053,
      "loss": 0.3186,
      "num_input_tokens_seen": 1644328,
      "step": 2155
    },
    {
      "epoch": 4.490644490644491,
      "grad_norm": 0.0007519324426539242,
      "learning_rate": 0.29784867783851227,
      "loss": 0.2862,
      "num_input_tokens_seen": 1648104,
      "step": 2160
    },
    {
      "epoch": 4.501039501039501,
      "grad_norm": 0.004415489267557859,
      "learning_rate": 0.2978387258878951,
      "loss": 0.2679,
      "num_input_tokens_seen": 1651944,
      "step": 2165
    },
    {
      "epoch": 4.511434511434511,
      "grad_norm": 0.0046625323593616486,
      "learning_rate": 0.29782875113868856,
      "loss": 0.3172,
      "num_input_tokens_seen": 1655752,
      "step": 2170
    },
    {
      "epoch": 4.521829521829522,
      "grad_norm": 0.005025898106396198,
      "learning_rate": 0.2978187535924309,
      "loss": 0.316,
      "num_input_tokens_seen": 1659528,
      "step": 2175
    },
    {
      "epoch": 4.532224532224532,
      "grad_norm": 0.006388816982507706,
      "learning_rate": 0.29780873325066376,
      "loss": 0.2865,
      "num_input_tokens_seen": 1663336,
      "step": 2180
    },
    {
      "epoch": 4.542619542619542,
      "grad_norm": 0.004953952040523291,
      "learning_rate": 0.2977986901149325,
      "loss": 0.2819,
      "num_input_tokens_seen": 1667048,
      "step": 2185
    },
    {
      "epoch": 4.553014553014553,
      "grad_norm": 0.007029210217297077,
      "learning_rate": 0.29778862418678587,
      "loss": 0.2956,
      "num_input_tokens_seen": 1670728,
      "step": 2190
    },
    {
      "epoch": 4.5634095634095635,
      "grad_norm": 0.0015059924917295575,
      "learning_rate": 0.29777853546777616,
      "loss": 0.2669,
      "num_input_tokens_seen": 1674408,
      "step": 2195
    },
    {
      "epoch": 4.573804573804574,
      "grad_norm": 0.0033020901028066874,
      "learning_rate": 0.2977684239594592,
      "loss": 0.2863,
      "num_input_tokens_seen": 1678248,
      "step": 2200
    },
    {
      "epoch": 4.573804573804574,
      "eval_loss": 0.25246745347976685,
      "eval_runtime": 13.4113,
      "eval_samples_per_second": 63.827,
      "eval_steps_per_second": 15.957,
      "num_input_tokens_seen": 1678248,
      "step": 2200
    },
    {
      "epoch": 4.584199584199585,
      "grad_norm": 0.0023702087346464396,
      "learning_rate": 0.29775828966339424,
      "loss": 0.2615,
      "num_input_tokens_seen": 1682152,
      "step": 2205
    },
    {
      "epoch": 4.594594594594595,
      "grad_norm": 0.0032688830979168415,
      "learning_rate": 0.29774813258114424,
      "loss": 0.2823,
      "num_input_tokens_seen": 1685992,
      "step": 2210
    },
    {
      "epoch": 4.604989604989605,
      "grad_norm": 0.003169362898916006,
      "learning_rate": 0.29773795271427544,
      "loss": 0.2983,
      "num_input_tokens_seen": 1689672,
      "step": 2215
    },
    {
      "epoch": 4.615384615384615,
      "grad_norm": 0.0028951934073120356,
      "learning_rate": 0.2977277500643577,
      "loss": 0.2678,
      "num_input_tokens_seen": 1693352,
      "step": 2220
    },
    {
      "epoch": 4.625779625779626,
      "grad_norm": 0.003787314984947443,
      "learning_rate": 0.29771752463296447,
      "loss": 0.3028,
      "num_input_tokens_seen": 1697064,
      "step": 2225
    },
    {
      "epoch": 4.636174636174636,
      "grad_norm": 0.002762764459475875,
      "learning_rate": 0.29770727642167266,
      "loss": 0.276,
      "num_input_tokens_seen": 1700712,
      "step": 2230
    },
    {
      "epoch": 4.646569646569646,
      "grad_norm": 0.005911846645176411,
      "learning_rate": 0.29769700543206257,
      "loss": 0.2744,
      "num_input_tokens_seen": 1704648,
      "step": 2235
    },
    {
      "epoch": 4.656964656964657,
      "grad_norm": 0.006923735607415438,
      "learning_rate": 0.2976867116657182,
      "loss": 0.2973,
      "num_input_tokens_seen": 1708296,
      "step": 2240
    },
    {
      "epoch": 4.667359667359667,
      "grad_norm": 0.0024055743124336004,
      "learning_rate": 0.2976763951242269,
      "loss": 0.293,
      "num_input_tokens_seen": 1712040,
      "step": 2245
    },
    {
      "epoch": 4.6777546777546775,
      "grad_norm": 0.0024562478065490723,
      "learning_rate": 0.29766605580917965,
      "loss": 0.2825,
      "num_input_tokens_seen": 1715912,
      "step": 2250
    },
    {
      "epoch": 4.6881496881496885,
      "grad_norm": 0.00522797042503953,
      "learning_rate": 0.29765569372217093,
      "loss": 0.2786,
      "num_input_tokens_seen": 1719880,
      "step": 2255
    },
    {
      "epoch": 4.698544698544699,
      "grad_norm": 0.008458050899207592,
      "learning_rate": 0.2976453088647987,
      "loss": 0.3103,
      "num_input_tokens_seen": 1723720,
      "step": 2260
    },
    {
      "epoch": 4.708939708939709,
      "grad_norm": 0.0029204199090600014,
      "learning_rate": 0.2976349012386644,
      "loss": 0.2633,
      "num_input_tokens_seen": 1727688,
      "step": 2265
    },
    {
      "epoch": 4.71933471933472,
      "grad_norm": 0.0006647854461334646,
      "learning_rate": 0.29762447084537297,
      "loss": 0.2928,
      "num_input_tokens_seen": 1731656,
      "step": 2270
    },
    {
      "epoch": 4.72972972972973,
      "grad_norm": 0.000478226545965299,
      "learning_rate": 0.29761401768653306,
      "loss": 0.282,
      "num_input_tokens_seen": 1735464,
      "step": 2275
    },
    {
      "epoch": 4.74012474012474,
      "grad_norm": 0.0017165506724268198,
      "learning_rate": 0.29760354176375653,
      "loss": 0.2863,
      "num_input_tokens_seen": 1739432,
      "step": 2280
    },
    {
      "epoch": 4.75051975051975,
      "grad_norm": 0.0007923941011540592,
      "learning_rate": 0.29759304307865897,
      "loss": 0.2904,
      "num_input_tokens_seen": 1743272,
      "step": 2285
    },
    {
      "epoch": 4.760914760914761,
      "grad_norm": 0.0012859933776780963,
      "learning_rate": 0.2975825216328594,
      "loss": 0.297,
      "num_input_tokens_seen": 1747048,
      "step": 2290
    },
    {
      "epoch": 4.771309771309771,
      "grad_norm": 0.005715373437851667,
      "learning_rate": 0.2975719774279804,
      "loss": 0.3093,
      "num_input_tokens_seen": 1750984,
      "step": 2295
    },
    {
      "epoch": 4.781704781704782,
      "grad_norm": 0.002364710671827197,
      "learning_rate": 0.29756141046564794,
      "loss": 0.2623,
      "num_input_tokens_seen": 1754824,
      "step": 2300
    },
    {
      "epoch": 4.792099792099792,
      "grad_norm": 0.0005401739617809653,
      "learning_rate": 0.2975508207474916,
      "loss": 0.2799,
      "num_input_tokens_seen": 1758888,
      "step": 2305
    },
    {
      "epoch": 4.802494802494802,
      "grad_norm": 0.003149572992697358,
      "learning_rate": 0.2975402082751445,
      "loss": 0.2817,
      "num_input_tokens_seen": 1762728,
      "step": 2310
    },
    {
      "epoch": 4.8128898128898125,
      "grad_norm": 0.0023899332154542208,
      "learning_rate": 0.29752957305024313,
      "loss": 0.2836,
      "num_input_tokens_seen": 1766536,
      "step": 2315
    },
    {
      "epoch": 4.8232848232848236,
      "grad_norm": 0.0008628643117845058,
      "learning_rate": 0.2975189150744277,
      "loss": 0.2728,
      "num_input_tokens_seen": 1770440,
      "step": 2320
    },
    {
      "epoch": 4.833679833679834,
      "grad_norm": 0.001297208247706294,
      "learning_rate": 0.29750823434934165,
      "loss": 0.3134,
      "num_input_tokens_seen": 1774216,
      "step": 2325
    },
    {
      "epoch": 4.844074844074844,
      "grad_norm": 0.002594448858872056,
      "learning_rate": 0.29749753087663217,
      "loss": 0.2829,
      "num_input_tokens_seen": 1777960,
      "step": 2330
    },
    {
      "epoch": 4.854469854469855,
      "grad_norm": 0.00308404047973454,
      "learning_rate": 0.29748680465794985,
      "loss": 0.3007,
      "num_input_tokens_seen": 1781768,
      "step": 2335
    },
    {
      "epoch": 4.864864864864865,
      "grad_norm": 0.0015814410289749503,
      "learning_rate": 0.29747605569494884,
      "loss": 0.2741,
      "num_input_tokens_seen": 1785480,
      "step": 2340
    },
    {
      "epoch": 4.875259875259875,
      "grad_norm": 0.0005505598383024335,
      "learning_rate": 0.29746528398928673,
      "loss": 0.2768,
      "num_input_tokens_seen": 1789320,
      "step": 2345
    },
    {
      "epoch": 4.885654885654886,
      "grad_norm": 0.0033011562190949917,
      "learning_rate": 0.2974544895426247,
      "loss": 0.2726,
      "num_input_tokens_seen": 1793032,
      "step": 2350
    },
    {
      "epoch": 4.896049896049896,
      "grad_norm": 0.007996849715709686,
      "learning_rate": 0.29744367235662733,
      "loss": 0.2753,
      "num_input_tokens_seen": 1796872,
      "step": 2355
    },
    {
      "epoch": 4.906444906444906,
      "grad_norm": 0.0008668152149766684,
      "learning_rate": 0.29743283243296276,
      "loss": 0.2592,
      "num_input_tokens_seen": 1800680,
      "step": 2360
    },
    {
      "epoch": 4.916839916839917,
      "grad_norm": 0.000578886887524277,
      "learning_rate": 0.29742196977330276,
      "loss": 0.2792,
      "num_input_tokens_seen": 1804552,
      "step": 2365
    },
    {
      "epoch": 4.927234927234927,
      "grad_norm": 0.0006882133893668652,
      "learning_rate": 0.2974110843793223,
      "loss": 0.2769,
      "num_input_tokens_seen": 1808392,
      "step": 2370
    },
    {
      "epoch": 4.9376299376299375,
      "grad_norm": 0.002639227779582143,
      "learning_rate": 0.2974001762527002,
      "loss": 0.2504,
      "num_input_tokens_seen": 1812232,
      "step": 2375
    },
    {
      "epoch": 4.948024948024948,
      "grad_norm": 0.0014049126766622066,
      "learning_rate": 0.2973892453951186,
      "loss": 0.2921,
      "num_input_tokens_seen": 1815944,
      "step": 2380
    },
    {
      "epoch": 4.958419958419959,
      "grad_norm": 0.0032679408323019743,
      "learning_rate": 0.2973782918082631,
      "loss": 0.2953,
      "num_input_tokens_seen": 1819688,
      "step": 2385
    },
    {
      "epoch": 4.968814968814969,
      "grad_norm": 0.0038558205123990774,
      "learning_rate": 0.29736731549382295,
      "loss": 0.3072,
      "num_input_tokens_seen": 1823432,
      "step": 2390
    },
    {
      "epoch": 4.979209979209979,
      "grad_norm": 0.007292178925126791,
      "learning_rate": 0.2973563164534908,
      "loss": 0.322,
      "num_input_tokens_seen": 1827240,
      "step": 2395
    },
    {
      "epoch": 4.98960498960499,
      "grad_norm": 0.003858346026390791,
      "learning_rate": 0.29734529468896287,
      "loss": 0.3352,
      "num_input_tokens_seen": 1831112,
      "step": 2400
    },
    {
      "epoch": 4.98960498960499,
      "eval_loss": 0.2800733745098114,
      "eval_runtime": 13.4053,
      "eval_samples_per_second": 63.856,
      "eval_steps_per_second": 15.964,
      "num_input_tokens_seen": 1831112,
      "step": 2400
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.0040953257121145725,
      "learning_rate": 0.2973342502019388,
      "loss": 0.3131,
      "num_input_tokens_seen": 1834816,
      "step": 2405
    },
    {
      "epoch": 5.01039501039501,
      "grad_norm": 0.0005722068017348647,
      "learning_rate": 0.2973231829941219,
      "loss": 0.3025,
      "num_input_tokens_seen": 1838592,
      "step": 2410
    },
    {
      "epoch": 5.020790020790021,
      "grad_norm": 0.006669714115560055,
      "learning_rate": 0.2973120930672188,
      "loss": 0.3195,
      "num_input_tokens_seen": 1842368,
      "step": 2415
    },
    {
      "epoch": 5.031185031185031,
      "grad_norm": 0.0056261541321873665,
      "learning_rate": 0.2973009804229397,
      "loss": 0.3021,
      "num_input_tokens_seen": 1846208,
      "step": 2420
    },
    {
      "epoch": 5.041580041580041,
      "grad_norm": 0.001198282465338707,
      "learning_rate": 0.29728984506299827,
      "loss": 0.2817,
      "num_input_tokens_seen": 1850144,
      "step": 2425
    },
    {
      "epoch": 5.051975051975052,
      "grad_norm": 0.0013224403373897076,
      "learning_rate": 0.2972786869891118,
      "loss": 0.3173,
      "num_input_tokens_seen": 1854176,
      "step": 2430
    },
    {
      "epoch": 5.0623700623700625,
      "grad_norm": 0.0025285715237259865,
      "learning_rate": 0.29726750620300096,
      "loss": 0.2896,
      "num_input_tokens_seen": 1858048,
      "step": 2435
    },
    {
      "epoch": 5.072765072765073,
      "grad_norm": 0.004811470862478018,
      "learning_rate": 0.29725630270639003,
      "loss": 0.2944,
      "num_input_tokens_seen": 1861888,
      "step": 2440
    },
    {
      "epoch": 5.083160083160083,
      "grad_norm": 0.0016600388335064054,
      "learning_rate": 0.2972450765010067,
      "loss": 0.273,
      "num_input_tokens_seen": 1865760,
      "step": 2445
    },
    {
      "epoch": 5.093555093555094,
      "grad_norm": 0.003098409855738282,
      "learning_rate": 0.29723382758858213,
      "loss": 0.2806,
      "num_input_tokens_seen": 1869472,
      "step": 2450
    },
    {
      "epoch": 5.103950103950104,
      "grad_norm": 0.0032900734804570675,
      "learning_rate": 0.29722255597085107,
      "loss": 0.2622,
      "num_input_tokens_seen": 1873184,
      "step": 2455
    },
    {
      "epoch": 5.114345114345114,
      "grad_norm": 0.002768657635897398,
      "learning_rate": 0.2972112616495518,
      "loss": 0.2683,
      "num_input_tokens_seen": 1877088,
      "step": 2460
    },
    {
      "epoch": 5.124740124740125,
      "grad_norm": 0.004758150782436132,
      "learning_rate": 0.297199944626426,
      "loss": 0.2793,
      "num_input_tokens_seen": 1881056,
      "step": 2465
    },
    {
      "epoch": 5.135135135135135,
      "grad_norm": 0.002628802787512541,
      "learning_rate": 0.2971886049032189,
      "loss": 0.2707,
      "num_input_tokens_seen": 1884864,
      "step": 2470
    },
    {
      "epoch": 5.145530145530145,
      "grad_norm": 0.0006630115676671267,
      "learning_rate": 0.29717724248167926,
      "loss": 0.2628,
      "num_input_tokens_seen": 1888704,
      "step": 2475
    },
    {
      "epoch": 5.155925155925156,
      "grad_norm": 0.001930988859385252,
      "learning_rate": 0.29716585736355927,
      "loss": 0.2806,
      "num_input_tokens_seen": 1892608,
      "step": 2480
    },
    {
      "epoch": 5.166320166320166,
      "grad_norm": 0.0005219251615926623,
      "learning_rate": 0.2971544495506147,
      "loss": 0.2778,
      "num_input_tokens_seen": 1896448,
      "step": 2485
    },
    {
      "epoch": 5.1767151767151764,
      "grad_norm": 0.0009427057229913771,
      "learning_rate": 0.2971430190446048,
      "loss": 0.3103,
      "num_input_tokens_seen": 1900416,
      "step": 2490
    },
    {
      "epoch": 5.1871101871101875,
      "grad_norm": 0.0029217449482530355,
      "learning_rate": 0.2971315658472921,
      "loss": 0.3026,
      "num_input_tokens_seen": 1904192,
      "step": 2495
    },
    {
      "epoch": 5.197505197505198,
      "grad_norm": 0.0024619351606816053,
      "learning_rate": 0.2971200899604431,
      "loss": 0.285,
      "num_input_tokens_seen": 1907808,
      "step": 2500
    },
    {
      "epoch": 5.207900207900208,
      "grad_norm": 0.0010934073943644762,
      "learning_rate": 0.29710859138582735,
      "loss": 0.2577,
      "num_input_tokens_seen": 1911584,
      "step": 2505
    },
    {
      "epoch": 5.218295218295219,
      "grad_norm": 0.003578747157007456,
      "learning_rate": 0.29709707012521813,
      "loss": 0.2985,
      "num_input_tokens_seen": 1915264,
      "step": 2510
    },
    {
      "epoch": 5.228690228690229,
      "grad_norm": 0.0009640550124458969,
      "learning_rate": 0.29708552618039213,
      "loss": 0.3255,
      "num_input_tokens_seen": 1919072,
      "step": 2515
    },
    {
      "epoch": 5.239085239085239,
      "grad_norm": 0.0012125809444114566,
      "learning_rate": 0.2970739595531296,
      "loss": 0.2725,
      "num_input_tokens_seen": 1922720,
      "step": 2520
    },
    {
      "epoch": 5.24948024948025,
      "grad_norm": 0.0006683564861305058,
      "learning_rate": 0.2970623702452143,
      "loss": 0.2718,
      "num_input_tokens_seen": 1926496,
      "step": 2525
    },
    {
      "epoch": 5.25987525987526,
      "grad_norm": 0.0027965367771685123,
      "learning_rate": 0.2970507582584334,
      "loss": 0.2324,
      "num_input_tokens_seen": 1930304,
      "step": 2530
    },
    {
      "epoch": 5.27027027027027,
      "grad_norm": 0.001029761740937829,
      "learning_rate": 0.2970391235945776,
      "loss": 0.2776,
      "num_input_tokens_seen": 1934080,
      "step": 2535
    },
    {
      "epoch": 5.28066528066528,
      "grad_norm": 0.0016207931330427527,
      "learning_rate": 0.2970274662554412,
      "loss": 0.2822,
      "num_input_tokens_seen": 1937920,
      "step": 2540
    },
    {
      "epoch": 5.291060291060291,
      "grad_norm": 0.0030946426559239626,
      "learning_rate": 0.2970157862428218,
      "loss": 0.2942,
      "num_input_tokens_seen": 1941728,
      "step": 2545
    },
    {
      "epoch": 5.301455301455301,
      "grad_norm": 0.00411140127107501,
      "learning_rate": 0.2970040835585206,
      "loss": 0.2876,
      "num_input_tokens_seen": 1945696,
      "step": 2550
    },
    {
      "epoch": 5.3118503118503115,
      "grad_norm": 0.0006096501019783318,
      "learning_rate": 0.2969923582043424,
      "loss": 0.2862,
      "num_input_tokens_seen": 1949696,
      "step": 2555
    },
    {
      "epoch": 5.3222453222453225,
      "grad_norm": 0.0034708271268755198,
      "learning_rate": 0.2969806101820953,
      "loss": 0.2511,
      "num_input_tokens_seen": 1953376,
      "step": 2560
    },
    {
      "epoch": 5.332640332640333,
      "grad_norm": 0.009027624502778053,
      "learning_rate": 0.2969688394935911,
      "loss": 0.3366,
      "num_input_tokens_seen": 1957120,
      "step": 2565
    },
    {
      "epoch": 5.343035343035343,
      "grad_norm": 0.0005166474147699773,
      "learning_rate": 0.2969570461406449,
      "loss": 0.2772,
      "num_input_tokens_seen": 1960992,
      "step": 2570
    },
    {
      "epoch": 5.353430353430354,
      "grad_norm": 0.0013280416605994105,
      "learning_rate": 0.29694523012507534,
      "loss": 0.2569,
      "num_input_tokens_seen": 1964576,
      "step": 2575
    },
    {
      "epoch": 5.363825363825364,
      "grad_norm": 0.002501196227967739,
      "learning_rate": 0.2969333914487048,
      "loss": 0.2734,
      "num_input_tokens_seen": 1968224,
      "step": 2580
    },
    {
      "epoch": 5.374220374220374,
      "grad_norm": 0.0019239579560235143,
      "learning_rate": 0.2969215301133587,
      "loss": 0.2572,
      "num_input_tokens_seen": 1972096,
      "step": 2585
    },
    {
      "epoch": 5.384615384615385,
      "grad_norm": 0.0003502235922496766,
      "learning_rate": 0.29690964612086634,
      "loss": 0.2348,
      "num_input_tokens_seen": 1975872,
      "step": 2590
    },
    {
      "epoch": 5.395010395010395,
      "grad_norm": 0.00165514275431633,
      "learning_rate": 0.2968977394730604,
      "loss": 0.2919,
      "num_input_tokens_seen": 1979584,
      "step": 2595
    },
    {
      "epoch": 5.405405405405405,
      "grad_norm": 0.0012301926035434008,
      "learning_rate": 0.296885810171777,
      "loss": 0.2685,
      "num_input_tokens_seen": 1983296,
      "step": 2600
    },
    {
      "epoch": 5.405405405405405,
      "eval_loss": 0.26522520184516907,
      "eval_runtime": 13.3936,
      "eval_samples_per_second": 63.911,
      "eval_steps_per_second": 15.978,
      "num_input_tokens_seen": 1983296,
      "step": 2600
    },
    {
      "epoch": 5.415800415800415,
      "grad_norm": 0.004574560094624758,
      "learning_rate": 0.2968738582188558,
      "loss": 0.3089,
      "num_input_tokens_seen": 1986944,
      "step": 2605
    },
    {
      "epoch": 5.426195426195426,
      "grad_norm": 0.004007366485893726,
      "learning_rate": 0.2968618836161399,
      "loss": 0.2758,
      "num_input_tokens_seen": 1990816,
      "step": 2610
    },
    {
      "epoch": 5.4365904365904365,
      "grad_norm": 0.0022180613595992327,
      "learning_rate": 0.296849886365476,
      "loss": 0.2945,
      "num_input_tokens_seen": 1994688,
      "step": 2615
    },
    {
      "epoch": 5.446985446985447,
      "grad_norm": 0.0002435646892990917,
      "learning_rate": 0.2968378664687142,
      "loss": 0.2629,
      "num_input_tokens_seen": 1998560,
      "step": 2620
    },
    {
      "epoch": 5.457380457380458,
      "grad_norm": 0.0024614459834992886,
      "learning_rate": 0.296825823927708,
      "loss": 0.2667,
      "num_input_tokens_seen": 2002368,
      "step": 2625
    },
    {
      "epoch": 5.467775467775468,
      "grad_norm": 0.0045992545783519745,
      "learning_rate": 0.29681375874431476,
      "loss": 0.2709,
      "num_input_tokens_seen": 2006176,
      "step": 2630
    },
    {
      "epoch": 5.478170478170478,
      "grad_norm": 0.0017498929519206285,
      "learning_rate": 0.29680167092039483,
      "loss": 0.2509,
      "num_input_tokens_seen": 2010016,
      "step": 2635
    },
    {
      "epoch": 5.488565488565489,
      "grad_norm": 0.002731350716203451,
      "learning_rate": 0.2967895604578125,
      "loss": 0.2681,
      "num_input_tokens_seen": 2013760,
      "step": 2640
    },
    {
      "epoch": 5.498960498960499,
      "grad_norm": 0.0024128558579832315,
      "learning_rate": 0.2967774273584352,
      "loss": 0.2642,
      "num_input_tokens_seen": 2017632,
      "step": 2645
    },
    {
      "epoch": 5.509355509355509,
      "grad_norm": 0.0014635191764682531,
      "learning_rate": 0.2967652716241342,
      "loss": 0.2815,
      "num_input_tokens_seen": 2021600,
      "step": 2650
    },
    {
      "epoch": 5.51975051975052,
      "grad_norm": 0.00026850702124647796,
      "learning_rate": 0.29675309325678384,
      "loss": 0.2845,
      "num_input_tokens_seen": 2025408,
      "step": 2655
    },
    {
      "epoch": 5.53014553014553,
      "grad_norm": 0.0028420162852853537,
      "learning_rate": 0.29674089225826233,
      "loss": 0.2824,
      "num_input_tokens_seen": 2029280,
      "step": 2660
    },
    {
      "epoch": 5.54054054054054,
      "grad_norm": 0.0022541400976479053,
      "learning_rate": 0.29672866863045116,
      "loss": 0.2433,
      "num_input_tokens_seen": 2033120,
      "step": 2665
    },
    {
      "epoch": 5.5509355509355505,
      "grad_norm": 0.0012572930427268147,
      "learning_rate": 0.2967164223752354,
      "loss": 0.2603,
      "num_input_tokens_seen": 2036864,
      "step": 2670
    },
    {
      "epoch": 5.5613305613305615,
      "grad_norm": 0.0013288218760862947,
      "learning_rate": 0.2967041534945035,
      "loss": 0.2599,
      "num_input_tokens_seen": 2040864,
      "step": 2675
    },
    {
      "epoch": 5.571725571725572,
      "grad_norm": 0.002293024677783251,
      "learning_rate": 0.2966918619901476,
      "loss": 0.2546,
      "num_input_tokens_seen": 2044768,
      "step": 2680
    },
    {
      "epoch": 5.582120582120583,
      "grad_norm": 0.000882921798620373,
      "learning_rate": 0.2966795478640631,
      "loss": 0.2704,
      "num_input_tokens_seen": 2048544,
      "step": 2685
    },
    {
      "epoch": 5.592515592515593,
      "grad_norm": 0.004462075419723988,
      "learning_rate": 0.29666721111814903,
      "loss": 0.2929,
      "num_input_tokens_seen": 2052416,
      "step": 2690
    },
    {
      "epoch": 5.602910602910603,
      "grad_norm": 0.0016282545402646065,
      "learning_rate": 0.2966548517543079,
      "loss": 0.2663,
      "num_input_tokens_seen": 2056128,
      "step": 2695
    },
    {
      "epoch": 5.613305613305613,
      "grad_norm": 0.001500854385085404,
      "learning_rate": 0.29664246977444564,
      "loss": 0.2896,
      "num_input_tokens_seen": 2059968,
      "step": 2700
    },
    {
      "epoch": 5.623700623700624,
      "grad_norm": 0.0007602103287354112,
      "learning_rate": 0.2966300651804717,
      "loss": 0.2673,
      "num_input_tokens_seen": 2063744,
      "step": 2705
    },
    {
      "epoch": 5.634095634095634,
      "grad_norm": 0.0056661940179765224,
      "learning_rate": 0.296617637974299,
      "loss": 0.3036,
      "num_input_tokens_seen": 2067328,
      "step": 2710
    },
    {
      "epoch": 5.644490644490644,
      "grad_norm": 0.0006392626091837883,
      "learning_rate": 0.2966051881578441,
      "loss": 0.2784,
      "num_input_tokens_seen": 2071200,
      "step": 2715
    },
    {
      "epoch": 5.654885654885655,
      "grad_norm": 0.0009071764652617276,
      "learning_rate": 0.29659271573302676,
      "loss": 0.3175,
      "num_input_tokens_seen": 2075040,
      "step": 2720
    },
    {
      "epoch": 5.665280665280665,
      "grad_norm": 0.0009053181856870651,
      "learning_rate": 0.2965802207017705,
      "loss": 0.2811,
      "num_input_tokens_seen": 2078944,
      "step": 2725
    },
    {
      "epoch": 5.675675675675675,
      "grad_norm": 0.0002588829374872148,
      "learning_rate": 0.2965677030660021,
      "loss": 0.2849,
      "num_input_tokens_seen": 2082784,
      "step": 2730
    },
    {
      "epoch": 5.686070686070686,
      "grad_norm": 0.00215737521648407,
      "learning_rate": 0.2965551628276521,
      "loss": 0.2698,
      "num_input_tokens_seen": 2086560,
      "step": 2735
    },
    {
      "epoch": 5.696465696465697,
      "grad_norm": 0.0003844249586109072,
      "learning_rate": 0.29654259998865423,
      "loss": 0.273,
      "num_input_tokens_seen": 2090464,
      "step": 2740
    },
    {
      "epoch": 5.706860706860707,
      "grad_norm": 0.003719937987625599,
      "learning_rate": 0.2965300145509458,
      "loss": 0.2704,
      "num_input_tokens_seen": 2094240,
      "step": 2745
    },
    {
      "epoch": 5.717255717255718,
      "grad_norm": 0.0026224376633763313,
      "learning_rate": 0.2965174065164678,
      "loss": 0.2646,
      "num_input_tokens_seen": 2098080,
      "step": 2750
    },
    {
      "epoch": 5.727650727650728,
      "grad_norm": 0.0019469807157292962,
      "learning_rate": 0.2965047758871644,
      "loss": 0.2812,
      "num_input_tokens_seen": 2101856,
      "step": 2755
    },
    {
      "epoch": 5.738045738045738,
      "grad_norm": 0.0030355053022503853,
      "learning_rate": 0.2964921226649835,
      "loss": 0.2613,
      "num_input_tokens_seen": 2105568,
      "step": 2760
    },
    {
      "epoch": 5.748440748440748,
      "grad_norm": 0.0016812717076390982,
      "learning_rate": 0.2964794468518763,
      "loss": 0.2641,
      "num_input_tokens_seen": 2109376,
      "step": 2765
    },
    {
      "epoch": 5.758835758835759,
      "grad_norm": 0.0022579412907361984,
      "learning_rate": 0.2964667484497977,
      "loss": 0.2337,
      "num_input_tokens_seen": 2113056,
      "step": 2770
    },
    {
      "epoch": 5.769230769230769,
      "grad_norm": 0.001475276192650199,
      "learning_rate": 0.29645402746070587,
      "loss": 0.2851,
      "num_input_tokens_seen": 2116864,
      "step": 2775
    },
    {
      "epoch": 5.779625779625779,
      "grad_norm": 0.00034311559284105897,
      "learning_rate": 0.2964412838865625,
      "loss": 0.2585,
      "num_input_tokens_seen": 2120672,
      "step": 2780
    },
    {
      "epoch": 5.79002079002079,
      "grad_norm": 0.000345252308761701,
      "learning_rate": 0.29642851772933293,
      "loss": 0.2467,
      "num_input_tokens_seen": 2124448,
      "step": 2785
    },
    {
      "epoch": 5.8004158004158,
      "grad_norm": 0.004911020863801241,
      "learning_rate": 0.29641572899098567,
      "loss": 0.2839,
      "num_input_tokens_seen": 2128288,
      "step": 2790
    },
    {
      "epoch": 5.8108108108108105,
      "grad_norm": 0.0007516889600083232,
      "learning_rate": 0.29640291767349314,
      "loss": 0.2351,
      "num_input_tokens_seen": 2132192,
      "step": 2795
    },
    {
      "epoch": 5.8212058212058215,
      "grad_norm": 0.0025285689625889063,
      "learning_rate": 0.2963900837788308,
      "loss": 0.3158,
      "num_input_tokens_seen": 2135968,
      "step": 2800
    },
    {
      "epoch": 5.8212058212058215,
      "eval_loss": 0.2622532546520233,
      "eval_runtime": 13.3972,
      "eval_samples_per_second": 63.894,
      "eval_steps_per_second": 15.974,
      "num_input_tokens_seen": 2135968,
      "step": 2800
    },
    {
      "epoch": 5.831600831600832,
      "grad_norm": 0.003932430408895016,
      "learning_rate": 0.2963772273089779,
      "loss": 0.2827,
      "num_input_tokens_seen": 2139712,
      "step": 2805
    },
    {
      "epoch": 5.841995841995842,
      "grad_norm": 0.00220858515240252,
      "learning_rate": 0.2963643482659171,
      "loss": 0.2926,
      "num_input_tokens_seen": 2143552,
      "step": 2810
    },
    {
      "epoch": 5.852390852390853,
      "grad_norm": 0.0028687971644103527,
      "learning_rate": 0.2963514466516345,
      "loss": 0.2778,
      "num_input_tokens_seen": 2147264,
      "step": 2815
    },
    {
      "epoch": 5.862785862785863,
      "grad_norm": 0.004031349439173937,
      "learning_rate": 0.2963385224681196,
      "loss": 0.2788,
      "num_input_tokens_seen": 2151040,
      "step": 2820
    },
    {
      "epoch": 5.873180873180873,
      "grad_norm": 0.005414790008217096,
      "learning_rate": 0.29632557571736556,
      "loss": 0.2851,
      "num_input_tokens_seen": 2154784,
      "step": 2825
    },
    {
      "epoch": 5.883575883575883,
      "grad_norm": 0.001311891246587038,
      "learning_rate": 0.2963126064013689,
      "loss": 0.275,
      "num_input_tokens_seen": 2158464,
      "step": 2830
    },
    {
      "epoch": 5.893970893970894,
      "grad_norm": 0.005444258917123079,
      "learning_rate": 0.29629961452212966,
      "loss": 0.2729,
      "num_input_tokens_seen": 2162336,
      "step": 2835
    },
    {
      "epoch": 5.904365904365904,
      "grad_norm": 0.0008525390876457095,
      "learning_rate": 0.2962866000816513,
      "loss": 0.2796,
      "num_input_tokens_seen": 2166336,
      "step": 2840
    },
    {
      "epoch": 5.914760914760915,
      "grad_norm": 0.0009947591461241245,
      "learning_rate": 0.2962735630819409,
      "loss": 0.2658,
      "num_input_tokens_seen": 2170048,
      "step": 2845
    },
    {
      "epoch": 5.925155925155925,
      "grad_norm": 0.005027471575886011,
      "learning_rate": 0.2962605035250089,
      "loss": 0.2902,
      "num_input_tokens_seen": 2173952,
      "step": 2850
    },
    {
      "epoch": 5.9355509355509355,
      "grad_norm": 0.0036935494281351566,
      "learning_rate": 0.29624742141286914,
      "loss": 0.284,
      "num_input_tokens_seen": 2177728,
      "step": 2855
    },
    {
      "epoch": 5.945945945945946,
      "grad_norm": 0.0033489062916487455,
      "learning_rate": 0.29623431674753925,
      "loss": 0.2897,
      "num_input_tokens_seen": 2181632,
      "step": 2860
    },
    {
      "epoch": 5.956340956340957,
      "grad_norm": 0.0029152543283998966,
      "learning_rate": 0.29622118953103993,
      "loss": 0.2894,
      "num_input_tokens_seen": 2185568,
      "step": 2865
    },
    {
      "epoch": 5.966735966735967,
      "grad_norm": 0.003034088760614395,
      "learning_rate": 0.2962080397653957,
      "loss": 0.292,
      "num_input_tokens_seen": 2189376,
      "step": 2870
    },
    {
      "epoch": 5.977130977130977,
      "grad_norm": 0.0007119469228200614,
      "learning_rate": 0.29619486745263435,
      "loss": 0.29,
      "num_input_tokens_seen": 2193312,
      "step": 2875
    },
    {
      "epoch": 5.987525987525988,
      "grad_norm": 0.0004276215913705528,
      "learning_rate": 0.2961816725947873,
      "loss": 0.2983,
      "num_input_tokens_seen": 2197120,
      "step": 2880
    },
    {
      "epoch": 5.997920997920998,
      "grad_norm": 0.0034098317846655846,
      "learning_rate": 0.29616845519388924,
      "loss": 0.2898,
      "num_input_tokens_seen": 2200832,
      "step": 2885
    },
    {
      "epoch": 6.008316008316008,
      "grad_norm": 0.0021364279091358185,
      "learning_rate": 0.2961552152519785,
      "loss": 0.2149,
      "num_input_tokens_seen": 2204592,
      "step": 2890
    },
    {
      "epoch": 6.018711018711019,
      "grad_norm": 0.0005487263551913202,
      "learning_rate": 0.29614195277109695,
      "loss": 0.2731,
      "num_input_tokens_seen": 2208400,
      "step": 2895
    },
    {
      "epoch": 6.029106029106029,
      "grad_norm": 0.002146113896742463,
      "learning_rate": 0.2961286677532897,
      "loss": 0.2719,
      "num_input_tokens_seen": 2212208,
      "step": 2900
    },
    {
      "epoch": 6.039501039501039,
      "grad_norm": 0.002831290476024151,
      "learning_rate": 0.2961153602006055,
      "loss": 0.278,
      "num_input_tokens_seen": 2216016,
      "step": 2905
    },
    {
      "epoch": 6.04989604989605,
      "grad_norm": 0.0004945373511873186,
      "learning_rate": 0.29610203011509656,
      "loss": 0.2545,
      "num_input_tokens_seen": 2219760,
      "step": 2910
    },
    {
      "epoch": 6.0602910602910605,
      "grad_norm": 0.0011749834520742297,
      "learning_rate": 0.29608867749881856,
      "loss": 0.2568,
      "num_input_tokens_seen": 2223760,
      "step": 2915
    },
    {
      "epoch": 6.070686070686071,
      "grad_norm": 0.00045744189992547035,
      "learning_rate": 0.29607530235383067,
      "loss": 0.2561,
      "num_input_tokens_seen": 2227472,
      "step": 2920
    },
    {
      "epoch": 6.081081081081081,
      "grad_norm": 0.0016379895387217402,
      "learning_rate": 0.2960619046821954,
      "loss": 0.2692,
      "num_input_tokens_seen": 2231152,
      "step": 2925
    },
    {
      "epoch": 6.091476091476092,
      "grad_norm": 0.0038132204208523035,
      "learning_rate": 0.2960484844859789,
      "loss": 0.2721,
      "num_input_tokens_seen": 2234960,
      "step": 2930
    },
    {
      "epoch": 6.101871101871102,
      "grad_norm": 0.0005531749338842928,
      "learning_rate": 0.29603504176725076,
      "loss": 0.2831,
      "num_input_tokens_seen": 2238704,
      "step": 2935
    },
    {
      "epoch": 6.112266112266112,
      "grad_norm": 0.001137790153734386,
      "learning_rate": 0.296021576528084,
      "loss": 0.289,
      "num_input_tokens_seen": 2242768,
      "step": 2940
    },
    {
      "epoch": 6.122661122661123,
      "grad_norm": 0.0008443903061561286,
      "learning_rate": 0.29600808877055507,
      "loss": 0.272,
      "num_input_tokens_seen": 2246800,
      "step": 2945
    },
    {
      "epoch": 6.133056133056133,
      "grad_norm": 0.002689443761482835,
      "learning_rate": 0.29599457849674404,
      "loss": 0.2428,
      "num_input_tokens_seen": 2250608,
      "step": 2950
    },
    {
      "epoch": 6.143451143451143,
      "grad_norm": 0.002616341458633542,
      "learning_rate": 0.2959810457087343,
      "loss": 0.2855,
      "num_input_tokens_seen": 2254480,
      "step": 2955
    },
    {
      "epoch": 6.153846153846154,
      "grad_norm": 0.0014948769239708781,
      "learning_rate": 0.2959674904086128,
      "loss": 0.2705,
      "num_input_tokens_seen": 2258192,
      "step": 2960
    },
    {
      "epoch": 6.164241164241164,
      "grad_norm": 0.0017134882509708405,
      "learning_rate": 0.2959539125984699,
      "loss": 0.3003,
      "num_input_tokens_seen": 2262096,
      "step": 2965
    },
    {
      "epoch": 6.174636174636174,
      "grad_norm": 0.003938702866435051,
      "learning_rate": 0.2959403122803996,
      "loss": 0.2826,
      "num_input_tokens_seen": 2265968,
      "step": 2970
    },
    {
      "epoch": 6.185031185031185,
      "grad_norm": 0.0038301756139844656,
      "learning_rate": 0.2959266894564991,
      "loss": 0.3034,
      "num_input_tokens_seen": 2269712,
      "step": 2975
    },
    {
      "epoch": 6.1954261954261955,
      "grad_norm": 0.002198318485170603,
      "learning_rate": 0.2959130441288692,
      "loss": 0.3166,
      "num_input_tokens_seen": 2273584,
      "step": 2980
    },
    {
      "epoch": 6.205821205821206,
      "grad_norm": 0.0024367321748286486,
      "learning_rate": 0.2958993762996143,
      "loss": 0.2881,
      "num_input_tokens_seen": 2277488,
      "step": 2985
    },
    {
      "epoch": 6.216216216216216,
      "grad_norm": 0.0013135808985680342,
      "learning_rate": 0.2958856859708421,
      "loss": 0.2756,
      "num_input_tokens_seen": 2281584,
      "step": 2990
    },
    {
      "epoch": 6.226611226611227,
      "grad_norm": 0.0008482606499455869,
      "learning_rate": 0.2958719731446638,
      "loss": 0.2673,
      "num_input_tokens_seen": 2285360,
      "step": 2995
    },
    {
      "epoch": 6.237006237006237,
      "grad_norm": 0.00019554028403945267,
      "learning_rate": 0.29585823782319404,
      "loss": 0.2789,
      "num_input_tokens_seen": 2289200,
      "step": 3000
    },
    {
      "epoch": 6.237006237006237,
      "eval_loss": 0.24987664818763733,
      "eval_runtime": 13.4423,
      "eval_samples_per_second": 63.679,
      "eval_steps_per_second": 15.92,
      "num_input_tokens_seen": 2289200,
      "step": 3000
    },
    {
      "epoch": 6.247401247401247,
      "grad_norm": 0.0013581948587670922,
      "learning_rate": 0.2958444800085511,
      "loss": 0.2615,
      "num_input_tokens_seen": 2293104,
      "step": 3005
    },
    {
      "epoch": 6.257796257796258,
      "grad_norm": 0.0062060169875621796,
      "learning_rate": 0.2958306997028565,
      "loss": 0.3159,
      "num_input_tokens_seen": 2296912,
      "step": 3010
    },
    {
      "epoch": 6.268191268191268,
      "grad_norm": 0.0007732873782515526,
      "learning_rate": 0.2958168969082354,
      "loss": 0.2535,
      "num_input_tokens_seen": 2300720,
      "step": 3015
    },
    {
      "epoch": 6.278586278586278,
      "grad_norm": 0.0018295745830982924,
      "learning_rate": 0.2958030716268164,
      "loss": 0.3605,
      "num_input_tokens_seen": 2304592,
      "step": 3020
    },
    {
      "epoch": 6.288981288981289,
      "grad_norm": 0.003372908802703023,
      "learning_rate": 0.2957892238607314,
      "loss": 0.2894,
      "num_input_tokens_seen": 2308368,
      "step": 3025
    },
    {
      "epoch": 6.299376299376299,
      "grad_norm": 0.001269126427359879,
      "learning_rate": 0.2957753536121161,
      "loss": 0.2808,
      "num_input_tokens_seen": 2312080,
      "step": 3030
    },
    {
      "epoch": 6.3097713097713095,
      "grad_norm": 0.0008721768972463906,
      "learning_rate": 0.29576146088310923,
      "loss": 0.2833,
      "num_input_tokens_seen": 2315824,
      "step": 3035
    },
    {
      "epoch": 6.3201663201663205,
      "grad_norm": 0.004103085491806269,
      "learning_rate": 0.2957475456758533,
      "loss": 0.2859,
      "num_input_tokens_seen": 2319536,
      "step": 3040
    },
    {
      "epoch": 6.330561330561331,
      "grad_norm": 0.0017040999373421073,
      "learning_rate": 0.2957336079924944,
      "loss": 0.2791,
      "num_input_tokens_seen": 2323472,
      "step": 3045
    },
    {
      "epoch": 6.340956340956341,
      "grad_norm": 0.001673279912211001,
      "learning_rate": 0.2957196478351816,
      "loss": 0.3094,
      "num_input_tokens_seen": 2327248,
      "step": 3050
    },
    {
      "epoch": 6.351351351351352,
      "grad_norm": 0.0047913952730596066,
      "learning_rate": 0.295705665206068,
      "loss": 0.2627,
      "num_input_tokens_seen": 2330992,
      "step": 3055
    },
    {
      "epoch": 6.361746361746362,
      "grad_norm": 0.00355904595926404,
      "learning_rate": 0.2956916601073097,
      "loss": 0.2803,
      "num_input_tokens_seen": 2334736,
      "step": 3060
    },
    {
      "epoch": 6.372141372141372,
      "grad_norm": 0.0006098496378399432,
      "learning_rate": 0.29567763254106655,
      "loss": 0.2729,
      "num_input_tokens_seen": 2338608,
      "step": 3065
    },
    {
      "epoch": 6.382536382536383,
      "grad_norm": 0.000949639012105763,
      "learning_rate": 0.29566358250950175,
      "loss": 0.2689,
      "num_input_tokens_seen": 2342416,
      "step": 3070
    },
    {
      "epoch": 6.392931392931393,
      "grad_norm": 0.0008701894548721611,
      "learning_rate": 0.295649510014782,
      "loss": 0.279,
      "num_input_tokens_seen": 2346256,
      "step": 3075
    },
    {
      "epoch": 6.403326403326403,
      "grad_norm": 0.0036095604300498962,
      "learning_rate": 0.2956354150590775,
      "loss": 0.2884,
      "num_input_tokens_seen": 2350096,
      "step": 3080
    },
    {
      "epoch": 6.413721413721413,
      "grad_norm": 0.003506356617435813,
      "learning_rate": 0.2956212976445618,
      "loss": 0.2603,
      "num_input_tokens_seen": 2353904,
      "step": 3085
    },
    {
      "epoch": 6.424116424116424,
      "grad_norm": 0.0013488584663718939,
      "learning_rate": 0.295607157773412,
      "loss": 0.2847,
      "num_input_tokens_seen": 2357648,
      "step": 3090
    },
    {
      "epoch": 6.4345114345114345,
      "grad_norm": 0.001138151972554624,
      "learning_rate": 0.2955929954478087,
      "loss": 0.2736,
      "num_input_tokens_seen": 2361296,
      "step": 3095
    },
    {
      "epoch": 6.444906444906445,
      "grad_norm": 0.002459130482748151,
      "learning_rate": 0.29557881066993585,
      "loss": 0.2574,
      "num_input_tokens_seen": 2365136,
      "step": 3100
    },
    {
      "epoch": 6.455301455301456,
      "grad_norm": 0.0009467218187637627,
      "learning_rate": 0.29556460344198093,
      "loss": 0.2858,
      "num_input_tokens_seen": 2368880,
      "step": 3105
    },
    {
      "epoch": 6.465696465696466,
      "grad_norm": 0.0010017661843448877,
      "learning_rate": 0.29555037376613486,
      "loss": 0.3172,
      "num_input_tokens_seen": 2372624,
      "step": 3110
    },
    {
      "epoch": 6.476091476091476,
      "grad_norm": 0.00029432657174766064,
      "learning_rate": 0.29553612164459203,
      "loss": 0.3104,
      "num_input_tokens_seen": 2376560,
      "step": 3115
    },
    {
      "epoch": 6.486486486486487,
      "grad_norm": 0.0008937453385442495,
      "learning_rate": 0.29552184707955037,
      "loss": 0.2895,
      "num_input_tokens_seen": 2380496,
      "step": 3120
    },
    {
      "epoch": 6.496881496881497,
      "grad_norm": 0.0012816459638997912,
      "learning_rate": 0.29550755007321117,
      "loss": 0.2721,
      "num_input_tokens_seen": 2384080,
      "step": 3125
    },
    {
      "epoch": 6.507276507276507,
      "grad_norm": 0.002156519563868642,
      "learning_rate": 0.29549323062777916,
      "loss": 0.2664,
      "num_input_tokens_seen": 2387952,
      "step": 3130
    },
    {
      "epoch": 6.517671517671518,
      "grad_norm": 0.0029360484331846237,
      "learning_rate": 0.29547888874546263,
      "loss": 0.2703,
      "num_input_tokens_seen": 2391856,
      "step": 3135
    },
    {
      "epoch": 6.528066528066528,
      "grad_norm": 0.002531585283577442,
      "learning_rate": 0.2954645244284732,
      "loss": 0.2289,
      "num_input_tokens_seen": 2395760,
      "step": 3140
    },
    {
      "epoch": 6.538461538461538,
      "grad_norm": 0.0006937634316273034,
      "learning_rate": 0.2954501376790261,
      "loss": 0.2559,
      "num_input_tokens_seen": 2399600,
      "step": 3145
    },
    {
      "epoch": 6.548856548856548,
      "grad_norm": 0.0014957934617996216,
      "learning_rate": 0.29543572849933997,
      "loss": 0.2536,
      "num_input_tokens_seen": 2403312,
      "step": 3150
    },
    {
      "epoch": 6.5592515592515594,
      "grad_norm": 0.0010256493696942925,
      "learning_rate": 0.2954212968916368,
      "loss": 0.2818,
      "num_input_tokens_seen": 2407088,
      "step": 3155
    },
    {
      "epoch": 6.56964656964657,
      "grad_norm": 0.004824563395231962,
      "learning_rate": 0.29540684285814217,
      "loss": 0.2902,
      "num_input_tokens_seen": 2410960,
      "step": 3160
    },
    {
      "epoch": 6.58004158004158,
      "grad_norm": 0.0028179758228361607,
      "learning_rate": 0.2953923664010851,
      "loss": 0.2556,
      "num_input_tokens_seen": 2414864,
      "step": 3165
    },
    {
      "epoch": 6.590436590436591,
      "grad_norm": 0.0012605141382664442,
      "learning_rate": 0.295377867522698,
      "loss": 0.2871,
      "num_input_tokens_seen": 2418672,
      "step": 3170
    },
    {
      "epoch": 6.600831600831601,
      "grad_norm": 0.0008157733827829361,
      "learning_rate": 0.2953633462252168,
      "loss": 0.2458,
      "num_input_tokens_seen": 2422480,
      "step": 3175
    },
    {
      "epoch": 6.611226611226611,
      "grad_norm": 0.003589311381801963,
      "learning_rate": 0.2953488025108809,
      "loss": 0.3349,
      "num_input_tokens_seen": 2426192,
      "step": 3180
    },
    {
      "epoch": 6.621621621621622,
      "grad_norm": 0.004647602327167988,
      "learning_rate": 0.295334236381933,
      "loss": 0.3446,
      "num_input_tokens_seen": 2430064,
      "step": 3185
    },
    {
      "epoch": 6.632016632016632,
      "grad_norm": 0.0009125987417064607,
      "learning_rate": 0.29531964784061954,
      "loss": 0.2961,
      "num_input_tokens_seen": 2433936,
      "step": 3190
    },
    {
      "epoch": 6.642411642411642,
      "grad_norm": 0.0011381659423932433,
      "learning_rate": 0.2953050368891902,
      "loss": 0.2817,
      "num_input_tokens_seen": 2437744,
      "step": 3195
    },
    {
      "epoch": 6.652806652806653,
      "grad_norm": 0.0020581961143761873,
      "learning_rate": 0.29529040352989805,
      "loss": 0.2968,
      "num_input_tokens_seen": 2441648,
      "step": 3200
    },
    {
      "epoch": 6.652806652806653,
      "eval_loss": 0.2987210750579834,
      "eval_runtime": 13.441,
      "eval_samples_per_second": 63.686,
      "eval_steps_per_second": 15.921,
      "num_input_tokens_seen": 2441648,
      "step": 3200
    },
    {
      "epoch": 6.663201663201663,
      "grad_norm": 0.002160570817068219,
      "learning_rate": 0.29527574776499993,
      "loss": 0.2882,
      "num_input_tokens_seen": 2445296,
      "step": 3205
    },
    {
      "epoch": 6.673596673596673,
      "grad_norm": 0.0015864783199504018,
      "learning_rate": 0.2952610695967558,
      "loss": 0.2745,
      "num_input_tokens_seen": 2449072,
      "step": 3210
    },
    {
      "epoch": 6.6839916839916835,
      "grad_norm": 0.0049042971804738045,
      "learning_rate": 0.29524636902742935,
      "loss": 0.2593,
      "num_input_tokens_seen": 2452944,
      "step": 3215
    },
    {
      "epoch": 6.6943866943866945,
      "grad_norm": 0.004881016444414854,
      "learning_rate": 0.2952316460592875,
      "loss": 0.2545,
      "num_input_tokens_seen": 2456784,
      "step": 3220
    },
    {
      "epoch": 6.704781704781705,
      "grad_norm": 0.002456564689055085,
      "learning_rate": 0.29521690069460066,
      "loss": 0.3016,
      "num_input_tokens_seen": 2460656,
      "step": 3225
    },
    {
      "epoch": 6.715176715176716,
      "grad_norm": 0.0019492740975692868,
      "learning_rate": 0.29520213293564285,
      "loss": 0.3149,
      "num_input_tokens_seen": 2464464,
      "step": 3230
    },
    {
      "epoch": 6.725571725571726,
      "grad_norm": 0.002612174255773425,
      "learning_rate": 0.29518734278469144,
      "loss": 0.365,
      "num_input_tokens_seen": 2468304,
      "step": 3235
    },
    {
      "epoch": 6.735966735966736,
      "grad_norm": 0.001191766932606697,
      "learning_rate": 0.29517253024402723,
      "loss": 0.2907,
      "num_input_tokens_seen": 2472240,
      "step": 3240
    },
    {
      "epoch": 6.746361746361746,
      "grad_norm": 0.002067373599857092,
      "learning_rate": 0.2951576953159345,
      "loss": 0.2997,
      "num_input_tokens_seen": 2476144,
      "step": 3245
    },
    {
      "epoch": 6.756756756756757,
      "grad_norm": 0.003907051403075457,
      "learning_rate": 0.29514283800270097,
      "loss": 0.2957,
      "num_input_tokens_seen": 2479824,
      "step": 3250
    },
    {
      "epoch": 6.767151767151767,
      "grad_norm": 0.006945164408534765,
      "learning_rate": 0.2951279583066179,
      "loss": 0.2891,
      "num_input_tokens_seen": 2483344,
      "step": 3255
    },
    {
      "epoch": 6.777546777546777,
      "grad_norm": 0.00037364737363532186,
      "learning_rate": 0.2951130562299798,
      "loss": 0.2368,
      "num_input_tokens_seen": 2487120,
      "step": 3260
    },
    {
      "epoch": 6.787941787941788,
      "grad_norm": 0.0015776692889630795,
      "learning_rate": 0.29509813177508487,
      "loss": 0.249,
      "num_input_tokens_seen": 2490832,
      "step": 3265
    },
    {
      "epoch": 6.798336798336798,
      "grad_norm": 0.0016386404167860746,
      "learning_rate": 0.2950831849442346,
      "loss": 0.3019,
      "num_input_tokens_seen": 2494544,
      "step": 3270
    },
    {
      "epoch": 6.8087318087318085,
      "grad_norm": 0.0033174215350300074,
      "learning_rate": 0.2950682157397339,
      "loss": 0.296,
      "num_input_tokens_seen": 2498352,
      "step": 3275
    },
    {
      "epoch": 6.8191268191268195,
      "grad_norm": 0.0018637643661350012,
      "learning_rate": 0.2950532241638914,
      "loss": 0.2782,
      "num_input_tokens_seen": 2502128,
      "step": 3280
    },
    {
      "epoch": 6.82952182952183,
      "grad_norm": 0.0008951465133577585,
      "learning_rate": 0.2950382102190188,
      "loss": 0.255,
      "num_input_tokens_seen": 2505936,
      "step": 3285
    },
    {
      "epoch": 6.83991683991684,
      "grad_norm": 0.0025906923692673445,
      "learning_rate": 0.2950231739074316,
      "loss": 0.2779,
      "num_input_tokens_seen": 2509680,
      "step": 3290
    },
    {
      "epoch": 6.850311850311851,
      "grad_norm": 0.001334195607341826,
      "learning_rate": 0.29500811523144843,
      "loss": 0.2717,
      "num_input_tokens_seen": 2513744,
      "step": 3295
    },
    {
      "epoch": 6.860706860706861,
      "grad_norm": 0.0022794196847826242,
      "learning_rate": 0.2949930341933917,
      "loss": 0.3051,
      "num_input_tokens_seen": 2517488,
      "step": 3300
    },
    {
      "epoch": 6.871101871101871,
      "grad_norm": 0.001611954066902399,
      "learning_rate": 0.29497793079558693,
      "loss": 0.2277,
      "num_input_tokens_seen": 2521360,
      "step": 3305
    },
    {
      "epoch": 6.881496881496881,
      "grad_norm": 0.0005894139758311212,
      "learning_rate": 0.2949628050403633,
      "loss": 0.2974,
      "num_input_tokens_seen": 2525040,
      "step": 3310
    },
    {
      "epoch": 6.891891891891892,
      "grad_norm": 0.004006496164947748,
      "learning_rate": 0.2949476569300535,
      "loss": 0.2432,
      "num_input_tokens_seen": 2528816,
      "step": 3315
    },
    {
      "epoch": 6.902286902286902,
      "grad_norm": 0.0034496912267059088,
      "learning_rate": 0.29493248646699344,
      "loss": 0.2927,
      "num_input_tokens_seen": 2532560,
      "step": 3320
    },
    {
      "epoch": 6.912681912681912,
      "grad_norm": 0.0014164462918415666,
      "learning_rate": 0.29491729365352265,
      "loss": 0.2795,
      "num_input_tokens_seen": 2536432,
      "step": 3325
    },
    {
      "epoch": 6.923076923076923,
      "grad_norm": 0.0016756703844293952,
      "learning_rate": 0.29490207849198397,
      "loss": 0.2917,
      "num_input_tokens_seen": 2540272,
      "step": 3330
    },
    {
      "epoch": 6.9334719334719335,
      "grad_norm": 0.003685267176479101,
      "learning_rate": 0.29488684098472384,
      "loss": 0.3848,
      "num_input_tokens_seen": 2544112,
      "step": 3335
    },
    {
      "epoch": 6.943866943866944,
      "grad_norm": 0.00037594392779283226,
      "learning_rate": 0.2948715811340921,
      "loss": 0.2555,
      "num_input_tokens_seen": 2547984,
      "step": 3340
    },
    {
      "epoch": 6.954261954261955,
      "grad_norm": 0.0014869710430502892,
      "learning_rate": 0.294856298942442,
      "loss": 0.2918,
      "num_input_tokens_seen": 2551760,
      "step": 3345
    },
    {
      "epoch": 6.964656964656965,
      "grad_norm": 0.001702604815363884,
      "learning_rate": 0.2948409944121302,
      "loss": 0.2872,
      "num_input_tokens_seen": 2555536,
      "step": 3350
    },
    {
      "epoch": 6.975051975051975,
      "grad_norm": 0.0037030810490250587,
      "learning_rate": 0.29482566754551687,
      "loss": 0.2821,
      "num_input_tokens_seen": 2559312,
      "step": 3355
    },
    {
      "epoch": 6.985446985446986,
      "grad_norm": 0.0009733354090712965,
      "learning_rate": 0.2948103183449656,
      "loss": 0.261,
      "num_input_tokens_seen": 2563056,
      "step": 3360
    },
    {
      "epoch": 6.995841995841996,
      "grad_norm": 0.0014379547210410237,
      "learning_rate": 0.2947949468128435,
      "loss": 0.2887,
      "num_input_tokens_seen": 2566832,
      "step": 3365
    },
    {
      "epoch": 7.006237006237006,
      "grad_norm": 0.0029334097635000944,
      "learning_rate": 0.2947795529515209,
      "loss": 0.2605,
      "num_input_tokens_seen": 2570592,
      "step": 3370
    },
    {
      "epoch": 7.016632016632016,
      "grad_norm": 0.0014982214197516441,
      "learning_rate": 0.29476413676337193,
      "loss": 0.2564,
      "num_input_tokens_seen": 2574336,
      "step": 3375
    },
    {
      "epoch": 7.027027027027027,
      "grad_norm": 0.005589826963841915,
      "learning_rate": 0.2947486982507738,
      "loss": 0.2951,
      "num_input_tokens_seen": 2578080,
      "step": 3380
    },
    {
      "epoch": 7.037422037422037,
      "grad_norm": 0.0009938370203599334,
      "learning_rate": 0.29473323741610735,
      "loss": 0.2814,
      "num_input_tokens_seen": 2581920,
      "step": 3385
    },
    {
      "epoch": 7.047817047817047,
      "grad_norm": 0.001087399898096919,
      "learning_rate": 0.2947177542617569,
      "loss": 0.256,
      "num_input_tokens_seen": 2585760,
      "step": 3390
    },
    {
      "epoch": 7.058212058212058,
      "grad_norm": 0.0014066743897274137,
      "learning_rate": 0.2947022487901101,
      "loss": 0.2837,
      "num_input_tokens_seen": 2589568,
      "step": 3395
    },
    {
      "epoch": 7.0686070686070686,
      "grad_norm": 0.0015377472154796124,
      "learning_rate": 0.2946867210035581,
      "loss": 0.2783,
      "num_input_tokens_seen": 2593344,
      "step": 3400
    },
    {
      "epoch": 7.0686070686070686,
      "eval_loss": 0.2496897280216217,
      "eval_runtime": 13.4173,
      "eval_samples_per_second": 63.798,
      "eval_steps_per_second": 15.95,
      "num_input_tokens_seen": 2593344,
      "step": 3400
    },
    {
      "epoch": 7.079002079002079,
      "grad_norm": 0.0012426524190232158,
      "learning_rate": 0.2946711709044954,
      "loss": 0.2362,
      "num_input_tokens_seen": 2597056,
      "step": 3405
    },
    {
      "epoch": 7.08939708939709,
      "grad_norm": 0.0032184498850256205,
      "learning_rate": 0.2946555984953202,
      "loss": 0.3063,
      "num_input_tokens_seen": 2600896,
      "step": 3410
    },
    {
      "epoch": 7.0997920997921,
      "grad_norm": 0.003795068711042404,
      "learning_rate": 0.2946400037784338,
      "loss": 0.2877,
      "num_input_tokens_seen": 2604512,
      "step": 3415
    },
    {
      "epoch": 7.11018711018711,
      "grad_norm": 0.0013317748671397567,
      "learning_rate": 0.29462438675624114,
      "loss": 0.2838,
      "num_input_tokens_seen": 2608448,
      "step": 3420
    },
    {
      "epoch": 7.120582120582121,
      "grad_norm": 0.001264024875126779,
      "learning_rate": 0.2946087474311506,
      "loss": 0.2967,
      "num_input_tokens_seen": 2612128,
      "step": 3425
    },
    {
      "epoch": 7.130977130977131,
      "grad_norm": 0.0018578324234113097,
      "learning_rate": 0.294593085805574,
      "loss": 0.2696,
      "num_input_tokens_seen": 2615968,
      "step": 3430
    },
    {
      "epoch": 7.141372141372141,
      "grad_norm": 0.0028327356558293104,
      "learning_rate": 0.2945774018819264,
      "loss": 0.2713,
      "num_input_tokens_seen": 2619776,
      "step": 3435
    },
    {
      "epoch": 7.151767151767152,
      "grad_norm": 0.0003140619373880327,
      "learning_rate": 0.2945616956626266,
      "loss": 0.2403,
      "num_input_tokens_seen": 2623648,
      "step": 3440
    },
    {
      "epoch": 7.162162162162162,
      "grad_norm": 0.0021300730295479298,
      "learning_rate": 0.2945459671500966,
      "loss": 0.2527,
      "num_input_tokens_seen": 2627520,
      "step": 3445
    },
    {
      "epoch": 7.172557172557172,
      "grad_norm": 0.0008381623192690313,
      "learning_rate": 0.2945302163467621,
      "loss": 0.3084,
      "num_input_tokens_seen": 2631232,
      "step": 3450
    },
    {
      "epoch": 7.182952182952183,
      "grad_norm": 0.00034687481820583344,
      "learning_rate": 0.2945144432550519,
      "loss": 0.335,
      "num_input_tokens_seen": 2635008,
      "step": 3455
    },
    {
      "epoch": 7.1933471933471935,
      "grad_norm": 0.000970735854934901,
      "learning_rate": 0.29449864787739843,
      "loss": 0.3605,
      "num_input_tokens_seen": 2638656,
      "step": 3460
    },
    {
      "epoch": 7.203742203742204,
      "grad_norm": 0.0009498162544332445,
      "learning_rate": 0.2944828302162376,
      "loss": 0.2899,
      "num_input_tokens_seen": 2642304,
      "step": 3465
    },
    {
      "epoch": 7.214137214137214,
      "grad_norm": 0.000339907273883,
      "learning_rate": 0.2944669902740087,
      "loss": 0.2785,
      "num_input_tokens_seen": 2646144,
      "step": 3470
    },
    {
      "epoch": 7.224532224532225,
      "grad_norm": 0.0004050501447636634,
      "learning_rate": 0.2944511280531544,
      "loss": 0.2714,
      "num_input_tokens_seen": 2649856,
      "step": 3475
    },
    {
      "epoch": 7.234927234927235,
      "grad_norm": 0.0009633766603656113,
      "learning_rate": 0.29443524355612083,
      "loss": 0.2849,
      "num_input_tokens_seen": 2653568,
      "step": 3480
    },
    {
      "epoch": 7.245322245322245,
      "grad_norm": 0.0017693584086373448,
      "learning_rate": 0.29441933678535764,
      "loss": 0.2542,
      "num_input_tokens_seen": 2657440,
      "step": 3485
    },
    {
      "epoch": 7.255717255717256,
      "grad_norm": 0.0033374472986906767,
      "learning_rate": 0.29440340774331786,
      "loss": 0.2527,
      "num_input_tokens_seen": 2661248,
      "step": 3490
    },
    {
      "epoch": 7.266112266112266,
      "grad_norm": 0.0020857155323028564,
      "learning_rate": 0.2943874564324579,
      "loss": 0.3139,
      "num_input_tokens_seen": 2665088,
      "step": 3495
    },
    {
      "epoch": 7.276507276507276,
      "grad_norm": 0.002290264004841447,
      "learning_rate": 0.2943714828552376,
      "loss": 0.2983,
      "num_input_tokens_seen": 2668832,
      "step": 3500
    },
    {
      "epoch": 7.286902286902287,
      "grad_norm": 0.00248618028126657,
      "learning_rate": 0.29435548701412045,
      "loss": 0.2686,
      "num_input_tokens_seen": 2672608,
      "step": 3505
    },
    {
      "epoch": 7.297297297297297,
      "grad_norm": 0.0010215550428256392,
      "learning_rate": 0.2943394689115731,
      "loss": 0.3741,
      "num_input_tokens_seen": 2676288,
      "step": 3510
    },
    {
      "epoch": 7.3076923076923075,
      "grad_norm": 0.004506244324147701,
      "learning_rate": 0.29432342855006577,
      "loss": 0.3209,
      "num_input_tokens_seen": 2680032,
      "step": 3515
    },
    {
      "epoch": 7.3180873180873185,
      "grad_norm": 0.00153873220551759,
      "learning_rate": 0.294307365932072,
      "loss": 0.2855,
      "num_input_tokens_seen": 2683936,
      "step": 3520
    },
    {
      "epoch": 7.328482328482329,
      "grad_norm": 0.0015606615925207734,
      "learning_rate": 0.294291281060069,
      "loss": 0.2514,
      "num_input_tokens_seen": 2687776,
      "step": 3525
    },
    {
      "epoch": 7.338877338877339,
      "grad_norm": 0.0003289362066425383,
      "learning_rate": 0.29427517393653724,
      "loss": 0.2474,
      "num_input_tokens_seen": 2691584,
      "step": 3530
    },
    {
      "epoch": 7.349272349272349,
      "grad_norm": 0.0014647473581135273,
      "learning_rate": 0.29425904456396046,
      "loss": 0.2737,
      "num_input_tokens_seen": 2695520,
      "step": 3535
    },
    {
      "epoch": 7.35966735966736,
      "grad_norm": 0.0013536863261833787,
      "learning_rate": 0.2942428929448262,
      "loss": 0.2728,
      "num_input_tokens_seen": 2699392,
      "step": 3540
    },
    {
      "epoch": 7.37006237006237,
      "grad_norm": 0.001209497102536261,
      "learning_rate": 0.2942267190816252,
      "loss": 0.2741,
      "num_input_tokens_seen": 2703360,
      "step": 3545
    },
    {
      "epoch": 7.38045738045738,
      "grad_norm": 0.00017752284475136548,
      "learning_rate": 0.2942105229768516,
      "loss": 0.2493,
      "num_input_tokens_seen": 2707168,
      "step": 3550
    },
    {
      "epoch": 7.390852390852391,
      "grad_norm": 0.001956959953531623,
      "learning_rate": 0.29419430463300306,
      "loss": 0.3017,
      "num_input_tokens_seen": 2711040,
      "step": 3555
    },
    {
      "epoch": 7.401247401247401,
      "grad_norm": 0.00018391320190858096,
      "learning_rate": 0.2941780640525808,
      "loss": 0.294,
      "num_input_tokens_seen": 2714816,
      "step": 3560
    },
    {
      "epoch": 7.411642411642411,
      "grad_norm": 0.0013431685511022806,
      "learning_rate": 0.2941618012380891,
      "loss": 0.2956,
      "num_input_tokens_seen": 2718720,
      "step": 3565
    },
    {
      "epoch": 7.422037422037422,
      "grad_norm": 0.0035378821194171906,
      "learning_rate": 0.29414551619203605,
      "loss": 0.3033,
      "num_input_tokens_seen": 2722720,
      "step": 3570
    },
    {
      "epoch": 7.4324324324324325,
      "grad_norm": 0.0012397931423038244,
      "learning_rate": 0.29412920891693295,
      "loss": 0.2678,
      "num_input_tokens_seen": 2726560,
      "step": 3575
    },
    {
      "epoch": 7.442827442827443,
      "grad_norm": 0.0019612584728747606,
      "learning_rate": 0.2941128794152946,
      "loss": 0.2785,
      "num_input_tokens_seen": 2730368,
      "step": 3580
    },
    {
      "epoch": 7.453222453222454,
      "grad_norm": 0.0014950396725907922,
      "learning_rate": 0.2940965276896392,
      "loss": 0.2603,
      "num_input_tokens_seen": 2734048,
      "step": 3585
    },
    {
      "epoch": 7.463617463617464,
      "grad_norm": 0.0005553779774345458,
      "learning_rate": 0.2940801537424884,
      "loss": 0.2645,
      "num_input_tokens_seen": 2737888,
      "step": 3590
    },
    {
      "epoch": 7.474012474012474,
      "grad_norm": 0.0005526858731172979,
      "learning_rate": 0.2940637575763673,
      "loss": 0.2479,
      "num_input_tokens_seen": 2741728,
      "step": 3595
    },
    {
      "epoch": 7.484407484407485,
      "grad_norm": 0.002084500854834914,
      "learning_rate": 0.2940473391938043,
      "loss": 0.3116,
      "num_input_tokens_seen": 2745792,
      "step": 3600
    },
    {
      "epoch": 7.484407484407485,
      "eval_loss": 0.30892422795295715,
      "eval_runtime": 13.4337,
      "eval_samples_per_second": 63.72,
      "eval_steps_per_second": 15.93,
      "num_input_tokens_seen": 2745792,
      "step": 3600
    },
    {
      "epoch": 7.494802494802495,
      "grad_norm": 0.0005621399031952024,
      "learning_rate": 0.29403089859733145,
      "loss": 0.295,
      "num_input_tokens_seen": 2749760,
      "step": 3605
    },
    {
      "epoch": 7.505197505197505,
      "grad_norm": 0.002951402449980378,
      "learning_rate": 0.294014435789484,
      "loss": 0.2926,
      "num_input_tokens_seen": 2753440,
      "step": 3610
    },
    {
      "epoch": 7.515592515592516,
      "grad_norm": 0.0028559663332998753,
      "learning_rate": 0.2939979507728007,
      "loss": 0.264,
      "num_input_tokens_seen": 2757280,
      "step": 3615
    },
    {
      "epoch": 7.525987525987526,
      "grad_norm": 0.0005421764799393713,
      "learning_rate": 0.2939814435498239,
      "loss": 0.2761,
      "num_input_tokens_seen": 2761248,
      "step": 3620
    },
    {
      "epoch": 7.536382536382536,
      "grad_norm": 0.00045091958600096405,
      "learning_rate": 0.29396491412309905,
      "loss": 0.273,
      "num_input_tokens_seen": 2764992,
      "step": 3625
    },
    {
      "epoch": 7.546777546777546,
      "grad_norm": 0.0005407624994404614,
      "learning_rate": 0.2939483624951753,
      "loss": 0.2534,
      "num_input_tokens_seen": 2768896,
      "step": 3630
    },
    {
      "epoch": 7.557172557172557,
      "grad_norm": 0.0006500000017695129,
      "learning_rate": 0.2939317886686051,
      "loss": 0.2463,
      "num_input_tokens_seen": 2772768,
      "step": 3635
    },
    {
      "epoch": 7.5675675675675675,
      "grad_norm": 0.0007543907267972827,
      "learning_rate": 0.2939151926459443,
      "loss": 0.2548,
      "num_input_tokens_seen": 2776544,
      "step": 3640
    },
    {
      "epoch": 7.577962577962578,
      "grad_norm": 0.0010387434158474207,
      "learning_rate": 0.2938985744297522,
      "loss": 0.2815,
      "num_input_tokens_seen": 2780416,
      "step": 3645
    },
    {
      "epoch": 7.588357588357589,
      "grad_norm": 0.0008292027050629258,
      "learning_rate": 0.29388193402259166,
      "loss": 0.2733,
      "num_input_tokens_seen": 2784256,
      "step": 3650
    },
    {
      "epoch": 7.598752598752599,
      "grad_norm": 0.00036755899782292545,
      "learning_rate": 0.29386527142702873,
      "loss": 0.2312,
      "num_input_tokens_seen": 2788064,
      "step": 3655
    },
    {
      "epoch": 7.609147609147609,
      "grad_norm": 0.0016179765807464719,
      "learning_rate": 0.293848586645633,
      "loss": 0.2233,
      "num_input_tokens_seen": 2791712,
      "step": 3660
    },
    {
      "epoch": 7.61954261954262,
      "grad_norm": 0.00041799162863753736,
      "learning_rate": 0.2938318796809775,
      "loss": 0.2966,
      "num_input_tokens_seen": 2795456,
      "step": 3665
    },
    {
      "epoch": 7.62993762993763,
      "grad_norm": 0.0023691749665886164,
      "learning_rate": 0.29381515053563867,
      "loss": 0.2724,
      "num_input_tokens_seen": 2799168,
      "step": 3670
    },
    {
      "epoch": 7.64033264033264,
      "grad_norm": 0.002142353681847453,
      "learning_rate": 0.29379839921219636,
      "loss": 0.2782,
      "num_input_tokens_seen": 2802944,
      "step": 3675
    },
    {
      "epoch": 7.650727650727651,
      "grad_norm": 0.0021882164292037487,
      "learning_rate": 0.2937816257132338,
      "loss": 0.2738,
      "num_input_tokens_seen": 2806880,
      "step": 3680
    },
    {
      "epoch": 7.661122661122661,
      "grad_norm": 0.0012558167800307274,
      "learning_rate": 0.2937648300413376,
      "loss": 0.2815,
      "num_input_tokens_seen": 2810848,
      "step": 3685
    },
    {
      "epoch": 7.671517671517671,
      "grad_norm": 0.0031289570033550262,
      "learning_rate": 0.293748012199098,
      "loss": 0.2927,
      "num_input_tokens_seen": 2814592,
      "step": 3690
    },
    {
      "epoch": 7.6819126819126815,
      "grad_norm": 0.0004973011091351509,
      "learning_rate": 0.29373117218910844,
      "loss": 0.2799,
      "num_input_tokens_seen": 2818464,
      "step": 3695
    },
    {
      "epoch": 7.6923076923076925,
      "grad_norm": 0.0031013141851872206,
      "learning_rate": 0.2937143100139659,
      "loss": 0.2829,
      "num_input_tokens_seen": 2822368,
      "step": 3700
    },
    {
      "epoch": 7.702702702702703,
      "grad_norm": 0.0011804451933130622,
      "learning_rate": 0.29369742567627083,
      "loss": 0.2649,
      "num_input_tokens_seen": 2826304,
      "step": 3705
    },
    {
      "epoch": 7.713097713097713,
      "grad_norm": 0.005373354535549879,
      "learning_rate": 0.29368051917862675,
      "loss": 0.2741,
      "num_input_tokens_seen": 2830176,
      "step": 3710
    },
    {
      "epoch": 7.723492723492724,
      "grad_norm": 0.0027358466759324074,
      "learning_rate": 0.2936635905236411,
      "loss": 0.2761,
      "num_input_tokens_seen": 2833824,
      "step": 3715
    },
    {
      "epoch": 7.733887733887734,
      "grad_norm": 0.0011990158818662167,
      "learning_rate": 0.2936466397139244,
      "loss": 0.2605,
      "num_input_tokens_seen": 2837568,
      "step": 3720
    },
    {
      "epoch": 7.744282744282744,
      "grad_norm": 0.0020165364257991314,
      "learning_rate": 0.2936296667520907,
      "loss": 0.2798,
      "num_input_tokens_seen": 2841344,
      "step": 3725
    },
    {
      "epoch": 7.754677754677755,
      "grad_norm": 0.0003979332104790956,
      "learning_rate": 0.2936126716407574,
      "loss": 0.2664,
      "num_input_tokens_seen": 2845120,
      "step": 3730
    },
    {
      "epoch": 7.765072765072765,
      "grad_norm": 0.0034349027555435896,
      "learning_rate": 0.29359565438254537,
      "loss": 0.2747,
      "num_input_tokens_seen": 2848896,
      "step": 3735
    },
    {
      "epoch": 7.775467775467775,
      "grad_norm": 0.0005567230982705951,
      "learning_rate": 0.29357861498007887,
      "loss": 0.2665,
      "num_input_tokens_seen": 2852736,
      "step": 3740
    },
    {
      "epoch": 7.785862785862786,
      "grad_norm": 0.0022427982185035944,
      "learning_rate": 0.29356155343598567,
      "loss": 0.2629,
      "num_input_tokens_seen": 2856608,
      "step": 3745
    },
    {
      "epoch": 7.796257796257796,
      "grad_norm": 0.002074519172310829,
      "learning_rate": 0.2935444697528968,
      "loss": 0.2976,
      "num_input_tokens_seen": 2860448,
      "step": 3750
    },
    {
      "epoch": 7.8066528066528065,
      "grad_norm": 0.0010802532779052854,
      "learning_rate": 0.2935273639334468,
      "loss": 0.2817,
      "num_input_tokens_seen": 2864256,
      "step": 3755
    },
    {
      "epoch": 7.817047817047817,
      "grad_norm": 0.0012731625465676188,
      "learning_rate": 0.29351023598027365,
      "loss": 0.2544,
      "num_input_tokens_seen": 2867936,
      "step": 3760
    },
    {
      "epoch": 7.827442827442828,
      "grad_norm": 0.0017853660974651575,
      "learning_rate": 0.2934930858960186,
      "loss": 0.2812,
      "num_input_tokens_seen": 2871904,
      "step": 3765
    },
    {
      "epoch": 7.837837837837838,
      "grad_norm": 0.0004487301339395344,
      "learning_rate": 0.29347591368332643,
      "loss": 0.2822,
      "num_input_tokens_seen": 2875840,
      "step": 3770
    },
    {
      "epoch": 7.848232848232849,
      "grad_norm": 0.0007442218484356999,
      "learning_rate": 0.2934587193448454,
      "loss": 0.2929,
      "num_input_tokens_seen": 2879648,
      "step": 3775
    },
    {
      "epoch": 7.858627858627859,
      "grad_norm": 0.0013928577536717057,
      "learning_rate": 0.29344150288322696,
      "loss": 0.2816,
      "num_input_tokens_seen": 2883456,
      "step": 3780
    },
    {
      "epoch": 7.869022869022869,
      "grad_norm": 0.0022304090671241283,
      "learning_rate": 0.2934242643011263,
      "loss": 0.2832,
      "num_input_tokens_seen": 2887456,
      "step": 3785
    },
    {
      "epoch": 7.879417879417879,
      "grad_norm": 0.0011471620528027415,
      "learning_rate": 0.2934070036012016,
      "loss": 0.2489,
      "num_input_tokens_seen": 2891296,
      "step": 3790
    },
    {
      "epoch": 7.88981288981289,
      "grad_norm": 0.0011504165595397353,
      "learning_rate": 0.29338972078611475,
      "loss": 0.2446,
      "num_input_tokens_seen": 2895200,
      "step": 3795
    },
    {
      "epoch": 7.9002079002079,
      "grad_norm": 0.0012059705331921577,
      "learning_rate": 0.2933724158585311,
      "loss": 0.2829,
      "num_input_tokens_seen": 2898816,
      "step": 3800
    },
    {
      "epoch": 7.9002079002079,
      "eval_loss": 0.2539701759815216,
      "eval_runtime": 13.4268,
      "eval_samples_per_second": 63.753,
      "eval_steps_per_second": 15.938,
      "num_input_tokens_seen": 2898816,
      "step": 3800
    },
    {
      "epoch": 7.91060291060291,
      "grad_norm": 0.0031327817123383284,
      "learning_rate": 0.29335508882111916,
      "loss": 0.2795,
      "num_input_tokens_seen": 2902592,
      "step": 3805
    },
    {
      "epoch": 7.920997920997921,
      "grad_norm": 0.0013943853555247188,
      "learning_rate": 0.29333773967655097,
      "loss": 0.2698,
      "num_input_tokens_seen": 2906432,
      "step": 3810
    },
    {
      "epoch": 7.9313929313929314,
      "grad_norm": 0.001740603824146092,
      "learning_rate": 0.2933203684275021,
      "loss": 0.2827,
      "num_input_tokens_seen": 2910240,
      "step": 3815
    },
    {
      "epoch": 7.941787941787942,
      "grad_norm": 0.002541606780141592,
      "learning_rate": 0.2933029750766513,
      "loss": 0.2706,
      "num_input_tokens_seen": 2913952,
      "step": 3820
    },
    {
      "epoch": 7.952182952182953,
      "grad_norm": 0.002188492566347122,
      "learning_rate": 0.2932855596266809,
      "loss": 0.2795,
      "num_input_tokens_seen": 2917568,
      "step": 3825
    },
    {
      "epoch": 7.962577962577963,
      "grad_norm": 0.0017209012294188142,
      "learning_rate": 0.2932681220802765,
      "loss": 0.25,
      "num_input_tokens_seen": 2921344,
      "step": 3830
    },
    {
      "epoch": 7.972972972972973,
      "grad_norm": 0.0017609309870749712,
      "learning_rate": 0.2932506624401274,
      "loss": 0.3102,
      "num_input_tokens_seen": 2925184,
      "step": 3835
    },
    {
      "epoch": 7.983367983367984,
      "grad_norm": 0.004350841045379639,
      "learning_rate": 0.29323318070892584,
      "loss": 0.3085,
      "num_input_tokens_seen": 2928992,
      "step": 3840
    },
    {
      "epoch": 7.993762993762994,
      "grad_norm": 0.00346102355979383,
      "learning_rate": 0.29321567688936784,
      "loss": 0.2773,
      "num_input_tokens_seen": 2932896,
      "step": 3845
    },
    {
      "epoch": 8.004158004158004,
      "grad_norm": 0.00240493705496192,
      "learning_rate": 0.29319815098415275,
      "loss": 0.287,
      "num_input_tokens_seen": 2936528,
      "step": 3850
    },
    {
      "epoch": 8.014553014553014,
      "grad_norm": 0.00035926824784837663,
      "learning_rate": 0.2931806029959832,
      "loss": 0.2727,
      "num_input_tokens_seen": 2940336,
      "step": 3855
    },
    {
      "epoch": 8.024948024948024,
      "grad_norm": 0.0017733067506924272,
      "learning_rate": 0.29316303292756535,
      "loss": 0.269,
      "num_input_tokens_seen": 2944144,
      "step": 3860
    },
    {
      "epoch": 8.035343035343036,
      "grad_norm": 0.0013503055088222027,
      "learning_rate": 0.29314544078160876,
      "loss": 0.255,
      "num_input_tokens_seen": 2947888,
      "step": 3865
    },
    {
      "epoch": 8.045738045738046,
      "grad_norm": 0.0005201065796427429,
      "learning_rate": 0.2931278265608263,
      "loss": 0.2952,
      "num_input_tokens_seen": 2951664,
      "step": 3870
    },
    {
      "epoch": 8.056133056133056,
      "grad_norm": 0.0031352173537015915,
      "learning_rate": 0.29311019026793433,
      "loss": 0.2741,
      "num_input_tokens_seen": 2955632,
      "step": 3875
    },
    {
      "epoch": 8.066528066528067,
      "grad_norm": 0.002389635192230344,
      "learning_rate": 0.29309253190565254,
      "loss": 0.3125,
      "num_input_tokens_seen": 2959280,
      "step": 3880
    },
    {
      "epoch": 8.076923076923077,
      "grad_norm": 0.00141516060102731,
      "learning_rate": 0.2930748514767042,
      "loss": 0.276,
      "num_input_tokens_seen": 2962928,
      "step": 3885
    },
    {
      "epoch": 8.087318087318087,
      "grad_norm": 0.0027491434011608362,
      "learning_rate": 0.29305714898381574,
      "loss": 0.2798,
      "num_input_tokens_seen": 2966640,
      "step": 3890
    },
    {
      "epoch": 8.097713097713097,
      "grad_norm": 0.0011779993074014783,
      "learning_rate": 0.29303942442971714,
      "loss": 0.2699,
      "num_input_tokens_seen": 2970544,
      "step": 3895
    },
    {
      "epoch": 8.108108108108109,
      "grad_norm": 0.00158534268848598,
      "learning_rate": 0.2930216778171417,
      "loss": 0.2629,
      "num_input_tokens_seen": 2974288,
      "step": 3900
    },
    {
      "epoch": 8.118503118503119,
      "grad_norm": 0.00036598744918592274,
      "learning_rate": 0.2930039091488263,
      "loss": 0.28,
      "num_input_tokens_seen": 2978256,
      "step": 3905
    },
    {
      "epoch": 8.128898128898129,
      "grad_norm": 0.001971879508346319,
      "learning_rate": 0.29298611842751093,
      "loss": 0.2559,
      "num_input_tokens_seen": 2982064,
      "step": 3910
    },
    {
      "epoch": 8.13929313929314,
      "grad_norm": 0.002802992006763816,
      "learning_rate": 0.29296830565593923,
      "loss": 0.2706,
      "num_input_tokens_seen": 2985936,
      "step": 3915
    },
    {
      "epoch": 8.14968814968815,
      "grad_norm": 0.00185109069570899,
      "learning_rate": 0.2929504708368582,
      "loss": 0.2977,
      "num_input_tokens_seen": 2989872,
      "step": 3920
    },
    {
      "epoch": 8.16008316008316,
      "grad_norm": 0.000660832563880831,
      "learning_rate": 0.29293261397301806,
      "loss": 0.299,
      "num_input_tokens_seen": 2993648,
      "step": 3925
    },
    {
      "epoch": 8.170478170478171,
      "grad_norm": 0.000330803421093151,
      "learning_rate": 0.29291473506717275,
      "loss": 0.2802,
      "num_input_tokens_seen": 2997456,
      "step": 3930
    },
    {
      "epoch": 8.180873180873181,
      "grad_norm": 0.001893747365102172,
      "learning_rate": 0.29289683412207923,
      "loss": 0.2714,
      "num_input_tokens_seen": 3001296,
      "step": 3935
    },
    {
      "epoch": 8.191268191268192,
      "grad_norm": 0.0011223821202293038,
      "learning_rate": 0.29287891114049813,
      "loss": 0.2556,
      "num_input_tokens_seen": 3005168,
      "step": 3940
    },
    {
      "epoch": 8.201663201663202,
      "grad_norm": 0.0025791542138904333,
      "learning_rate": 0.29286096612519347,
      "loss": 0.2947,
      "num_input_tokens_seen": 3008912,
      "step": 3945
    },
    {
      "epoch": 8.212058212058212,
      "grad_norm": 0.0005244700005277991,
      "learning_rate": 0.2928429990789325,
      "loss": 0.2845,
      "num_input_tokens_seen": 3012688,
      "step": 3950
    },
    {
      "epoch": 8.222453222453222,
      "grad_norm": 0.0007702698931097984,
      "learning_rate": 0.29282501000448596,
      "loss": 0.2641,
      "num_input_tokens_seen": 3016496,
      "step": 3955
    },
    {
      "epoch": 8.232848232848234,
      "grad_norm": 0.0014458136865869164,
      "learning_rate": 0.2928069989046281,
      "loss": 0.2372,
      "num_input_tokens_seen": 3020400,
      "step": 3960
    },
    {
      "epoch": 8.243243243243244,
      "grad_norm": 0.0004499837523326278,
      "learning_rate": 0.2927889657821363,
      "loss": 0.2583,
      "num_input_tokens_seen": 3024144,
      "step": 3965
    },
    {
      "epoch": 8.253638253638254,
      "grad_norm": 0.0013547156704589725,
      "learning_rate": 0.2927709106397916,
      "loss": 0.2741,
      "num_input_tokens_seen": 3027824,
      "step": 3970
    },
    {
      "epoch": 8.264033264033264,
      "grad_norm": 0.0007258598925545812,
      "learning_rate": 0.29275283348037834,
      "loss": 0.2432,
      "num_input_tokens_seen": 3031536,
      "step": 3975
    },
    {
      "epoch": 8.274428274428274,
      "grad_norm": 0.00018851507047656924,
      "learning_rate": 0.29273473430668423,
      "loss": 0.3405,
      "num_input_tokens_seen": 3035248,
      "step": 3980
    },
    {
      "epoch": 8.284823284823284,
      "grad_norm": 0.0009012404480017722,
      "learning_rate": 0.2927166131215003,
      "loss": 0.298,
      "num_input_tokens_seen": 3038928,
      "step": 3985
    },
    {
      "epoch": 8.295218295218294,
      "grad_norm": 0.002488626865670085,
      "learning_rate": 0.2926984699276212,
      "loss": 0.2816,
      "num_input_tokens_seen": 3042832,
      "step": 3990
    },
    {
      "epoch": 8.305613305613306,
      "grad_norm": 0.0018329822923988104,
      "learning_rate": 0.29268030472784473,
      "loss": 0.2446,
      "num_input_tokens_seen": 3046544,
      "step": 3995
    },
    {
      "epoch": 8.316008316008316,
      "grad_norm": 0.002680974779650569,
      "learning_rate": 0.2926621175249723,
      "loss": 0.2801,
      "num_input_tokens_seen": 3050480,
      "step": 4000
    },
    {
      "epoch": 8.316008316008316,
      "eval_loss": 0.2654391825199127,
      "eval_runtime": 13.4682,
      "eval_samples_per_second": 63.557,
      "eval_steps_per_second": 15.889,
      "num_input_tokens_seen": 3050480,
      "step": 4000
    },
    {
      "epoch": 8.326403326403327,
      "grad_norm": 0.0003244733961764723,
      "learning_rate": 0.29264390832180853,
      "loss": 0.2722,
      "num_input_tokens_seen": 3054352,
      "step": 4005
    },
    {
      "epoch": 8.336798336798337,
      "grad_norm": 0.0008291470585390925,
      "learning_rate": 0.29262567712116144,
      "loss": 0.2867,
      "num_input_tokens_seen": 3058096,
      "step": 4010
    },
    {
      "epoch": 8.347193347193347,
      "grad_norm": 0.001917881309054792,
      "learning_rate": 0.29260742392584266,
      "loss": 0.2743,
      "num_input_tokens_seen": 3061776,
      "step": 4015
    },
    {
      "epoch": 8.357588357588357,
      "grad_norm": 0.00015456200344488025,
      "learning_rate": 0.292589148738667,
      "loss": 0.2811,
      "num_input_tokens_seen": 3065648,
      "step": 4020
    },
    {
      "epoch": 8.367983367983369,
      "grad_norm": 0.0018043442396447062,
      "learning_rate": 0.2925708515624527,
      "loss": 0.262,
      "num_input_tokens_seen": 3069456,
      "step": 4025
    },
    {
      "epoch": 8.378378378378379,
      "grad_norm": 0.001464469125494361,
      "learning_rate": 0.29255253240002144,
      "loss": 0.2627,
      "num_input_tokens_seen": 3073232,
      "step": 4030
    },
    {
      "epoch": 8.388773388773389,
      "grad_norm": 0.00012295882334001362,
      "learning_rate": 0.2925341912541983,
      "loss": 0.2699,
      "num_input_tokens_seen": 3077008,
      "step": 4035
    },
    {
      "epoch": 8.3991683991684,
      "grad_norm": 0.003647369332611561,
      "learning_rate": 0.2925158281278116,
      "loss": 0.283,
      "num_input_tokens_seen": 3080848,
      "step": 4040
    },
    {
      "epoch": 8.40956340956341,
      "grad_norm": 0.0002791236329358071,
      "learning_rate": 0.29249744302369324,
      "loss": 0.2689,
      "num_input_tokens_seen": 3084528,
      "step": 4045
    },
    {
      "epoch": 8.41995841995842,
      "grad_norm": 0.0015925226034596562,
      "learning_rate": 0.29247903594467844,
      "loss": 0.2485,
      "num_input_tokens_seen": 3088432,
      "step": 4050
    },
    {
      "epoch": 8.43035343035343,
      "grad_norm": 0.0017383011290803552,
      "learning_rate": 0.2924606068936058,
      "loss": 0.2554,
      "num_input_tokens_seen": 3092464,
      "step": 4055
    },
    {
      "epoch": 8.440748440748441,
      "grad_norm": 0.0008245429489761591,
      "learning_rate": 0.2924421558733173,
      "loss": 0.2538,
      "num_input_tokens_seen": 3096336,
      "step": 4060
    },
    {
      "epoch": 8.451143451143452,
      "grad_norm": 0.0013962909579277039,
      "learning_rate": 0.2924236828866583,
      "loss": 0.2372,
      "num_input_tokens_seen": 3100176,
      "step": 4065
    },
    {
      "epoch": 8.461538461538462,
      "grad_norm": 0.0009744802955538034,
      "learning_rate": 0.29240518793647763,
      "loss": 0.2861,
      "num_input_tokens_seen": 3103856,
      "step": 4070
    },
    {
      "epoch": 8.471933471933472,
      "grad_norm": 0.0014368684496730566,
      "learning_rate": 0.29238667102562743,
      "loss": 0.2711,
      "num_input_tokens_seen": 3107760,
      "step": 4075
    },
    {
      "epoch": 8.482328482328482,
      "grad_norm": 0.003547325963154435,
      "learning_rate": 0.29236813215696317,
      "loss": 0.2776,
      "num_input_tokens_seen": 3111504,
      "step": 4080
    },
    {
      "epoch": 8.492723492723492,
      "grad_norm": 0.0006229123682714999,
      "learning_rate": 0.2923495713333439,
      "loss": 0.2548,
      "num_input_tokens_seen": 3115344,
      "step": 4085
    },
    {
      "epoch": 8.503118503118504,
      "grad_norm": 0.0014495555078610778,
      "learning_rate": 0.29233098855763173,
      "loss": 0.2835,
      "num_input_tokens_seen": 3119152,
      "step": 4090
    },
    {
      "epoch": 8.513513513513514,
      "grad_norm": 0.0003765155270230025,
      "learning_rate": 0.29231238383269254,
      "loss": 0.2834,
      "num_input_tokens_seen": 3122992,
      "step": 4095
    },
    {
      "epoch": 8.523908523908524,
      "grad_norm": 0.001166557427495718,
      "learning_rate": 0.2922937571613954,
      "loss": 0.2828,
      "num_input_tokens_seen": 3126896,
      "step": 4100
    },
    {
      "epoch": 8.534303534303534,
      "grad_norm": 0.0004082966479472816,
      "learning_rate": 0.29227510854661265,
      "loss": 0.2615,
      "num_input_tokens_seen": 3130768,
      "step": 4105
    },
    {
      "epoch": 8.544698544698544,
      "grad_norm": 0.0010949320858344436,
      "learning_rate": 0.29225643799122025,
      "loss": 0.2904,
      "num_input_tokens_seen": 3134640,
      "step": 4110
    },
    {
      "epoch": 8.555093555093555,
      "grad_norm": 0.0008516350644640625,
      "learning_rate": 0.2922377454980974,
      "loss": 0.2736,
      "num_input_tokens_seen": 3138448,
      "step": 4115
    },
    {
      "epoch": 8.565488565488565,
      "grad_norm": 0.0012211723951622844,
      "learning_rate": 0.29221903107012676,
      "loss": 0.2595,
      "num_input_tokens_seen": 3142128,
      "step": 4120
    },
    {
      "epoch": 8.575883575883577,
      "grad_norm": 0.0018546601058915257,
      "learning_rate": 0.29220029471019426,
      "loss": 0.2411,
      "num_input_tokens_seen": 3145776,
      "step": 4125
    },
    {
      "epoch": 8.586278586278587,
      "grad_norm": 0.0007068297709338367,
      "learning_rate": 0.2921815364211893,
      "loss": 0.2713,
      "num_input_tokens_seen": 3149744,
      "step": 4130
    },
    {
      "epoch": 8.596673596673597,
      "grad_norm": 0.0033452603965997696,
      "learning_rate": 0.29216275620600474,
      "loss": 0.2792,
      "num_input_tokens_seen": 3153584,
      "step": 4135
    },
    {
      "epoch": 8.607068607068607,
      "grad_norm": 0.00016673705249559134,
      "learning_rate": 0.29214395406753657,
      "loss": 0.2522,
      "num_input_tokens_seen": 3157488,
      "step": 4140
    },
    {
      "epoch": 8.617463617463617,
      "grad_norm": 0.002090153517201543,
      "learning_rate": 0.2921251300086844,
      "loss": 0.2922,
      "num_input_tokens_seen": 3161232,
      "step": 4145
    },
    {
      "epoch": 8.627858627858627,
      "grad_norm": 0.0019381395541131496,
      "learning_rate": 0.2921062840323511,
      "loss": 0.2791,
      "num_input_tokens_seen": 3165008,
      "step": 4150
    },
    {
      "epoch": 8.638253638253639,
      "grad_norm": 0.0018437692197039723,
      "learning_rate": 0.29208741614144307,
      "loss": 0.2783,
      "num_input_tokens_seen": 3168784,
      "step": 4155
    },
    {
      "epoch": 8.64864864864865,
      "grad_norm": 0.00018089324294123799,
      "learning_rate": 0.2920685263388698,
      "loss": 0.2609,
      "num_input_tokens_seen": 3172496,
      "step": 4160
    },
    {
      "epoch": 8.65904365904366,
      "grad_norm": 0.00019504009105730802,
      "learning_rate": 0.2920496146275445,
      "loss": 0.2674,
      "num_input_tokens_seen": 3176432,
      "step": 4165
    },
    {
      "epoch": 8.66943866943867,
      "grad_norm": 0.0022795891854912043,
      "learning_rate": 0.29203068101038343,
      "loss": 0.2775,
      "num_input_tokens_seen": 3180240,
      "step": 4170
    },
    {
      "epoch": 8.67983367983368,
      "grad_norm": 0.0019669150933623314,
      "learning_rate": 0.2920117254903065,
      "loss": 0.2729,
      "num_input_tokens_seen": 3183984,
      "step": 4175
    },
    {
      "epoch": 8.69022869022869,
      "grad_norm": 0.0022099860943853855,
      "learning_rate": 0.29199274807023695,
      "loss": 0.2844,
      "num_input_tokens_seen": 3187792,
      "step": 4180
    },
    {
      "epoch": 8.700623700623701,
      "grad_norm": 0.0015788617311045527,
      "learning_rate": 0.29197374875310117,
      "loss": 0.2856,
      "num_input_tokens_seen": 3191664,
      "step": 4185
    },
    {
      "epoch": 8.711018711018712,
      "grad_norm": 0.0013265166198834777,
      "learning_rate": 0.2919547275418292,
      "loss": 0.283,
      "num_input_tokens_seen": 3195472,
      "step": 4190
    },
    {
      "epoch": 8.721413721413722,
      "grad_norm": 0.0020737501326948404,
      "learning_rate": 0.29193568443935436,
      "loss": 0.265,
      "num_input_tokens_seen": 3199120,
      "step": 4195
    },
    {
      "epoch": 8.731808731808732,
      "grad_norm": 0.0028038134332746267,
      "learning_rate": 0.2919166194486133,
      "loss": 0.2669,
      "num_input_tokens_seen": 3202864,
      "step": 4200
    },
    {
      "epoch": 8.731808731808732,
      "eval_loss": 0.25126686692237854,
      "eval_runtime": 13.4091,
      "eval_samples_per_second": 63.837,
      "eval_steps_per_second": 15.959,
      "num_input_tokens_seen": 3202864,
      "step": 4200
    },
    {
      "epoch": 8.742203742203742,
      "grad_norm": 0.0002503364812582731,
      "learning_rate": 0.2918975325725461,
      "loss": 0.2679,
      "num_input_tokens_seen": 3206640,
      "step": 4205
    },
    {
      "epoch": 8.752598752598752,
      "grad_norm": 0.0005465574213303626,
      "learning_rate": 0.29187842381409607,
      "loss": 0.2869,
      "num_input_tokens_seen": 3210512,
      "step": 4210
    },
    {
      "epoch": 8.762993762993762,
      "grad_norm": 0.0005704878713004291,
      "learning_rate": 0.29185929317621023,
      "loss": 0.2823,
      "num_input_tokens_seen": 3214256,
      "step": 4215
    },
    {
      "epoch": 8.773388773388774,
      "grad_norm": 0.0007609032327309251,
      "learning_rate": 0.29184014066183867,
      "loss": 0.2863,
      "num_input_tokens_seen": 3218128,
      "step": 4220
    },
    {
      "epoch": 8.783783783783784,
      "grad_norm": 0.0003097786975558847,
      "learning_rate": 0.2918209662739349,
      "loss": 0.2701,
      "num_input_tokens_seen": 3222000,
      "step": 4225
    },
    {
      "epoch": 8.794178794178794,
      "grad_norm": 0.0021032600197941065,
      "learning_rate": 0.29180177001545593,
      "loss": 0.2868,
      "num_input_tokens_seen": 3225808,
      "step": 4230
    },
    {
      "epoch": 8.804573804573804,
      "grad_norm": 0.0005216242861934006,
      "learning_rate": 0.29178255188936203,
      "loss": 0.2809,
      "num_input_tokens_seen": 3229840,
      "step": 4235
    },
    {
      "epoch": 8.814968814968815,
      "grad_norm": 0.0001477514742873609,
      "learning_rate": 0.2917633118986169,
      "loss": 0.2641,
      "num_input_tokens_seen": 3233648,
      "step": 4240
    },
    {
      "epoch": 8.825363825363825,
      "grad_norm": 0.00027837895322591066,
      "learning_rate": 0.2917440500461875,
      "loss": 0.2835,
      "num_input_tokens_seen": 3237680,
      "step": 4245
    },
    {
      "epoch": 8.835758835758837,
      "grad_norm": 0.0004899093182757497,
      "learning_rate": 0.29172476633504435,
      "loss": 0.2742,
      "num_input_tokens_seen": 3241648,
      "step": 4250
    },
    {
      "epoch": 8.846153846153847,
      "grad_norm": 0.00013976104673929513,
      "learning_rate": 0.2917054607681612,
      "loss": 0.2796,
      "num_input_tokens_seen": 3245392,
      "step": 4255
    },
    {
      "epoch": 8.856548856548857,
      "grad_norm": 0.002230289625003934,
      "learning_rate": 0.29168613334851523,
      "loss": 0.2758,
      "num_input_tokens_seen": 3249072,
      "step": 4260
    },
    {
      "epoch": 8.866943866943867,
      "grad_norm": 0.0009966047946363688,
      "learning_rate": 0.2916667840790869,
      "loss": 0.2501,
      "num_input_tokens_seen": 3253040,
      "step": 4265
    },
    {
      "epoch": 8.877338877338877,
      "grad_norm": 0.0005958973779343069,
      "learning_rate": 0.2916474129628603,
      "loss": 0.2715,
      "num_input_tokens_seen": 3256848,
      "step": 4270
    },
    {
      "epoch": 8.887733887733887,
      "grad_norm": 0.00016244281141553074,
      "learning_rate": 0.29162802000282245,
      "loss": 0.2485,
      "num_input_tokens_seen": 3260720,
      "step": 4275
    },
    {
      "epoch": 8.898128898128899,
      "grad_norm": 0.0021869537886232138,
      "learning_rate": 0.2916086052019642,
      "loss": 0.2599,
      "num_input_tokens_seen": 3264560,
      "step": 4280
    },
    {
      "epoch": 8.90852390852391,
      "grad_norm": 0.002946468535810709,
      "learning_rate": 0.2915891685632794,
      "loss": 0.2726,
      "num_input_tokens_seen": 3268368,
      "step": 4285
    },
    {
      "epoch": 8.91891891891892,
      "grad_norm": 0.0028296001255512238,
      "learning_rate": 0.29156971008976545,
      "loss": 0.2895,
      "num_input_tokens_seen": 3272112,
      "step": 4290
    },
    {
      "epoch": 8.92931392931393,
      "grad_norm": 0.0012245416874065995,
      "learning_rate": 0.2915502297844232,
      "loss": 0.2724,
      "num_input_tokens_seen": 3275952,
      "step": 4295
    },
    {
      "epoch": 8.93970893970894,
      "grad_norm": 0.0005565624451264739,
      "learning_rate": 0.2915307276502566,
      "loss": 0.2998,
      "num_input_tokens_seen": 3279792,
      "step": 4300
    },
    {
      "epoch": 8.95010395010395,
      "grad_norm": 0.0018246360123157501,
      "learning_rate": 0.29151120369027334,
      "loss": 0.2822,
      "num_input_tokens_seen": 3283632,
      "step": 4305
    },
    {
      "epoch": 8.96049896049896,
      "grad_norm": 0.002096001524478197,
      "learning_rate": 0.29149165790748405,
      "loss": 0.2792,
      "num_input_tokens_seen": 3287440,
      "step": 4310
    },
    {
      "epoch": 8.970893970893972,
      "grad_norm": 0.0006191023858264089,
      "learning_rate": 0.291472090304903,
      "loss": 0.2917,
      "num_input_tokens_seen": 3291120,
      "step": 4315
    },
    {
      "epoch": 8.981288981288982,
      "grad_norm": 0.0015177656896412373,
      "learning_rate": 0.2914525008855478,
      "loss": 0.2863,
      "num_input_tokens_seen": 3294960,
      "step": 4320
    },
    {
      "epoch": 8.991683991683992,
      "grad_norm": 0.0011767708929255605,
      "learning_rate": 0.2914328896524394,
      "loss": 0.2591,
      "num_input_tokens_seen": 3298768,
      "step": 4325
    },
    {
      "epoch": 9.002079002079002,
      "grad_norm": 0.0001497742923675105,
      "learning_rate": 0.291413256608602,
      "loss": 0.2477,
      "num_input_tokens_seen": 3302528,
      "step": 4330
    },
    {
      "epoch": 9.012474012474012,
      "grad_norm": 0.0015019926941022277,
      "learning_rate": 0.29139360175706336,
      "loss": 0.2695,
      "num_input_tokens_seen": 3306464,
      "step": 4335
    },
    {
      "epoch": 9.022869022869022,
      "grad_norm": 0.0022210744209587574,
      "learning_rate": 0.2913739251008544,
      "loss": 0.2802,
      "num_input_tokens_seen": 3310208,
      "step": 4340
    },
    {
      "epoch": 9.033264033264032,
      "grad_norm": 0.0024578555021435022,
      "learning_rate": 0.29135422664300964,
      "loss": 0.2803,
      "num_input_tokens_seen": 3314080,
      "step": 4345
    },
    {
      "epoch": 9.043659043659044,
      "grad_norm": 0.00010010333062382415,
      "learning_rate": 0.29133450638656677,
      "loss": 0.307,
      "num_input_tokens_seen": 3317952,
      "step": 4350
    },
    {
      "epoch": 9.054054054054054,
      "grad_norm": 0.0009152779821306467,
      "learning_rate": 0.2913147643345669,
      "loss": 0.2943,
      "num_input_tokens_seen": 3321728,
      "step": 4355
    },
    {
      "epoch": 9.064449064449065,
      "grad_norm": 0.0010701373685151339,
      "learning_rate": 0.29129500049005447,
      "loss": 0.2658,
      "num_input_tokens_seen": 3325600,
      "step": 4360
    },
    {
      "epoch": 9.074844074844075,
      "grad_norm": 0.0029930565506219864,
      "learning_rate": 0.2912752148560773,
      "loss": 0.2969,
      "num_input_tokens_seen": 3329248,
      "step": 4365
    },
    {
      "epoch": 9.085239085239085,
      "grad_norm": 0.0003962019691243768,
      "learning_rate": 0.2912554074356866,
      "loss": 0.2617,
      "num_input_tokens_seen": 3333024,
      "step": 4370
    },
    {
      "epoch": 9.095634095634095,
      "grad_norm": 0.00014814978931099176,
      "learning_rate": 0.2912355782319371,
      "loss": 0.2787,
      "num_input_tokens_seen": 3336768,
      "step": 4375
    },
    {
      "epoch": 9.106029106029107,
      "grad_norm": 0.00015518043073825538,
      "learning_rate": 0.2912157272478864,
      "loss": 0.2823,
      "num_input_tokens_seen": 3340768,
      "step": 4380
    },
    {
      "epoch": 9.116424116424117,
      "grad_norm": 0.0012748459121212363,
      "learning_rate": 0.291195854486596,
      "loss": 0.2712,
      "num_input_tokens_seen": 3344576,
      "step": 4385
    },
    {
      "epoch": 9.126819126819127,
      "grad_norm": 0.00031417200807482004,
      "learning_rate": 0.2911759599511305,
      "loss": 0.3012,
      "num_input_tokens_seen": 3348352,
      "step": 4390
    },
    {
      "epoch": 9.137214137214137,
      "grad_norm": 0.0007369376835413277,
      "learning_rate": 0.29115604364455777,
      "loss": 0.2809,
      "num_input_tokens_seen": 3352064,
      "step": 4395
    },
    {
      "epoch": 9.147609147609147,
      "grad_norm": 0.00012129736569477245,
      "learning_rate": 0.2911361055699493,
      "loss": 0.2617,
      "num_input_tokens_seen": 3355680,
      "step": 4400
    },
    {
      "epoch": 9.147609147609147,
      "eval_loss": 0.25105273723602295,
      "eval_runtime": 13.4218,
      "eval_samples_per_second": 63.777,
      "eval_steps_per_second": 15.944,
      "num_input_tokens_seen": 3355680,
      "step": 4400
    },
    {
      "epoch": 9.158004158004157,
      "grad_norm": 0.000679651799146086,
      "learning_rate": 0.2911161457303797,
      "loss": 0.2642,
      "num_input_tokens_seen": 3359360,
      "step": 4405
    },
    {
      "epoch": 9.16839916839917,
      "grad_norm": 0.0017391651635989547,
      "learning_rate": 0.291096164128927,
      "loss": 0.2793,
      "num_input_tokens_seen": 3363104,
      "step": 4410
    },
    {
      "epoch": 9.17879417879418,
      "grad_norm": 0.0006626248941756785,
      "learning_rate": 0.2910761607686727,
      "loss": 0.2821,
      "num_input_tokens_seen": 3366848,
      "step": 4415
    },
    {
      "epoch": 9.18918918918919,
      "grad_norm": 0.0018741786479949951,
      "learning_rate": 0.2910561356527016,
      "loss": 0.2822,
      "num_input_tokens_seen": 3370528,
      "step": 4420
    },
    {
      "epoch": 9.1995841995842,
      "grad_norm": 0.0007630917243659496,
      "learning_rate": 0.2910360887841017,
      "loss": 0.2904,
      "num_input_tokens_seen": 3374304,
      "step": 4425
    },
    {
      "epoch": 9.20997920997921,
      "grad_norm": 0.0006876765401102602,
      "learning_rate": 0.2910160201659645,
      "loss": 0.2793,
      "num_input_tokens_seen": 3378112,
      "step": 4430
    },
    {
      "epoch": 9.22037422037422,
      "grad_norm": 0.0033967948984354734,
      "learning_rate": 0.29099592980138494,
      "loss": 0.3019,
      "num_input_tokens_seen": 3381984,
      "step": 4435
    },
    {
      "epoch": 9.23076923076923,
      "grad_norm": 0.0001733355165924877,
      "learning_rate": 0.29097581769346115,
      "loss": 0.2511,
      "num_input_tokens_seen": 3385856,
      "step": 4440
    },
    {
      "epoch": 9.241164241164242,
      "grad_norm": 0.0008578650304116309,
      "learning_rate": 0.29095568384529463,
      "loss": 0.2849,
      "num_input_tokens_seen": 3389856,
      "step": 4445
    },
    {
      "epoch": 9.251559251559252,
      "grad_norm": 0.002634449163451791,
      "learning_rate": 0.2909355282599903,
      "loss": 0.2698,
      "num_input_tokens_seen": 3393728,
      "step": 4450
    },
    {
      "epoch": 9.261954261954262,
      "grad_norm": 0.0010786305647343397,
      "learning_rate": 0.29091535094065635,
      "loss": 0.2778,
      "num_input_tokens_seen": 3397600,
      "step": 4455
    },
    {
      "epoch": 9.272349272349272,
      "grad_norm": 0.0017043001716956496,
      "learning_rate": 0.2908951518904045,
      "loss": 0.2514,
      "num_input_tokens_seen": 3401504,
      "step": 4460
    },
    {
      "epoch": 9.282744282744282,
      "grad_norm": 0.00040898696170188487,
      "learning_rate": 0.29087493111234963,
      "loss": 0.2622,
      "num_input_tokens_seen": 3405376,
      "step": 4465
    },
    {
      "epoch": 9.293139293139292,
      "grad_norm": 0.0008852222817949951,
      "learning_rate": 0.29085468860961,
      "loss": 0.2916,
      "num_input_tokens_seen": 3409120,
      "step": 4470
    },
    {
      "epoch": 9.303534303534304,
      "grad_norm": 0.000428498926339671,
      "learning_rate": 0.2908344243853073,
      "loss": 0.2795,
      "num_input_tokens_seen": 3412992,
      "step": 4475
    },
    {
      "epoch": 9.313929313929314,
      "grad_norm": 0.00045505608431994915,
      "learning_rate": 0.2908141384425666,
      "loss": 0.2748,
      "num_input_tokens_seen": 3416736,
      "step": 4480
    },
    {
      "epoch": 9.324324324324325,
      "grad_norm": 0.0015887983608990908,
      "learning_rate": 0.2907938307845161,
      "loss": 0.2784,
      "num_input_tokens_seen": 3420320,
      "step": 4485
    },
    {
      "epoch": 9.334719334719335,
      "grad_norm": 0.0005984515300951898,
      "learning_rate": 0.2907735014142876,
      "loss": 0.2617,
      "num_input_tokens_seen": 3424160,
      "step": 4490
    },
    {
      "epoch": 9.345114345114345,
      "grad_norm": 0.0005569813656620681,
      "learning_rate": 0.2907531503350161,
      "loss": 0.2816,
      "num_input_tokens_seen": 3428192,
      "step": 4495
    },
    {
      "epoch": 9.355509355509355,
      "grad_norm": 0.00018043076852336526,
      "learning_rate": 0.29073277754983995,
      "loss": 0.2804,
      "num_input_tokens_seen": 3431904,
      "step": 4500
    },
    {
      "epoch": 9.365904365904367,
      "grad_norm": 0.0013101734220981598,
      "learning_rate": 0.290712383061901,
      "loss": 0.2865,
      "num_input_tokens_seen": 3435776,
      "step": 4505
    },
    {
      "epoch": 9.376299376299377,
      "grad_norm": 0.002785986987873912,
      "learning_rate": 0.2906919668743443,
      "loss": 0.2891,
      "num_input_tokens_seen": 3439616,
      "step": 4510
    },
    {
      "epoch": 9.386694386694387,
      "grad_norm": 0.0003400310524739325,
      "learning_rate": 0.29067152899031823,
      "loss": 0.2636,
      "num_input_tokens_seen": 3443360,
      "step": 4515
    },
    {
      "epoch": 9.397089397089397,
      "grad_norm": 0.0018372413469478488,
      "learning_rate": 0.2906510694129746,
      "loss": 0.2682,
      "num_input_tokens_seen": 3447168,
      "step": 4520
    },
    {
      "epoch": 9.407484407484407,
      "grad_norm": 0.0008135041571222246,
      "learning_rate": 0.2906305881454685,
      "loss": 0.261,
      "num_input_tokens_seen": 3450848,
      "step": 4525
    },
    {
      "epoch": 9.417879417879417,
      "grad_norm": 0.0006873209495097399,
      "learning_rate": 0.2906100851909585,
      "loss": 0.2778,
      "num_input_tokens_seen": 3454688,
      "step": 4530
    },
    {
      "epoch": 9.428274428274428,
      "grad_norm": 0.00018847943283617496,
      "learning_rate": 0.29058956055260626,
      "loss": 0.2477,
      "num_input_tokens_seen": 3458720,
      "step": 4535
    },
    {
      "epoch": 9.43866943866944,
      "grad_norm": 0.0006840723799541593,
      "learning_rate": 0.2905690142335771,
      "loss": 0.2659,
      "num_input_tokens_seen": 3462464,
      "step": 4540
    },
    {
      "epoch": 9.44906444906445,
      "grad_norm": 0.001151498407125473,
      "learning_rate": 0.29054844623703946,
      "loss": 0.2526,
      "num_input_tokens_seen": 3466112,
      "step": 4545
    },
    {
      "epoch": 9.45945945945946,
      "grad_norm": 0.0005568534252233803,
      "learning_rate": 0.2905278565661651,
      "loss": 0.2565,
      "num_input_tokens_seen": 3470016,
      "step": 4550
    },
    {
      "epoch": 9.46985446985447,
      "grad_norm": 0.00042003337875939906,
      "learning_rate": 0.2905072452241293,
      "loss": 0.2567,
      "num_input_tokens_seen": 3473952,
      "step": 4555
    },
    {
      "epoch": 9.48024948024948,
      "grad_norm": 0.0012943726032972336,
      "learning_rate": 0.2904866122141106,
      "loss": 0.2749,
      "num_input_tokens_seen": 3477920,
      "step": 4560
    },
    {
      "epoch": 9.49064449064449,
      "grad_norm": 0.0007788283401168883,
      "learning_rate": 0.2904659575392908,
      "loss": 0.2454,
      "num_input_tokens_seen": 3482016,
      "step": 4565
    },
    {
      "epoch": 9.5010395010395,
      "grad_norm": 0.0014733957359567285,
      "learning_rate": 0.2904452812028551,
      "loss": 0.2804,
      "num_input_tokens_seen": 3485728,
      "step": 4570
    },
    {
      "epoch": 9.511434511434512,
      "grad_norm": 0.0013959248317405581,
      "learning_rate": 0.2904245832079922,
      "loss": 0.251,
      "num_input_tokens_seen": 3489632,
      "step": 4575
    },
    {
      "epoch": 9.521829521829522,
      "grad_norm": 0.0006215545581653714,
      "learning_rate": 0.29040386355789377,
      "loss": 0.2819,
      "num_input_tokens_seen": 3493152,
      "step": 4580
    },
    {
      "epoch": 9.532224532224532,
      "grad_norm": 0.0006808478501625359,
      "learning_rate": 0.29038312225575524,
      "loss": 0.2677,
      "num_input_tokens_seen": 3496992,
      "step": 4585
    },
    {
      "epoch": 9.542619542619542,
      "grad_norm": 0.0037674764171242714,
      "learning_rate": 0.29036235930477505,
      "loss": 0.3004,
      "num_input_tokens_seen": 3500768,
      "step": 4590
    },
    {
      "epoch": 9.553014553014552,
      "grad_norm": 0.004937266930937767,
      "learning_rate": 0.29034157470815514,
      "loss": 0.3232,
      "num_input_tokens_seen": 3504640,
      "step": 4595
    },
    {
      "epoch": 9.563409563409563,
      "grad_norm": 0.00010515181202208623,
      "learning_rate": 0.2903207684691008,
      "loss": 0.2678,
      "num_input_tokens_seen": 3508192,
      "step": 4600
    },
    {
      "epoch": 9.563409563409563,
      "eval_loss": 0.2624172568321228,
      "eval_runtime": 13.4039,
      "eval_samples_per_second": 63.862,
      "eval_steps_per_second": 15.966,
      "num_input_tokens_seen": 3508192,
      "step": 4600
    },
    {
      "epoch": 9.573804573804575,
      "grad_norm": 0.0003542314516380429,
      "learning_rate": 0.29029994059082054,
      "loss": 0.2704,
      "num_input_tokens_seen": 3511936,
      "step": 4605
    },
    {
      "epoch": 9.584199584199585,
      "grad_norm": 0.0004584401031024754,
      "learning_rate": 0.2902790910765264,
      "loss": 0.2254,
      "num_input_tokens_seen": 3515840,
      "step": 4610
    },
    {
      "epoch": 9.594594594594595,
      "grad_norm": 0.002230600453913212,
      "learning_rate": 0.29025821992943346,
      "loss": 0.2854,
      "num_input_tokens_seen": 3519648,
      "step": 4615
    },
    {
      "epoch": 9.604989604989605,
      "grad_norm": 0.000521887035574764,
      "learning_rate": 0.29023732715276046,
      "loss": 0.2686,
      "num_input_tokens_seen": 3523520,
      "step": 4620
    },
    {
      "epoch": 9.615384615384615,
      "grad_norm": 0.00011615233233897015,
      "learning_rate": 0.2902164127497293,
      "loss": 0.2881,
      "num_input_tokens_seen": 3527232,
      "step": 4625
    },
    {
      "epoch": 9.625779625779625,
      "grad_norm": 0.00018442737928126007,
      "learning_rate": 0.2901954767235652,
      "loss": 0.2765,
      "num_input_tokens_seen": 3531040,
      "step": 4630
    },
    {
      "epoch": 9.636174636174637,
      "grad_norm": 0.0010019656037911773,
      "learning_rate": 0.2901745190774968,
      "loss": 0.2549,
      "num_input_tokens_seen": 3534976,
      "step": 4635
    },
    {
      "epoch": 9.646569646569647,
      "grad_norm": 0.0009959720773622394,
      "learning_rate": 0.290153539814756,
      "loss": 0.263,
      "num_input_tokens_seen": 3538752,
      "step": 4640
    },
    {
      "epoch": 9.656964656964657,
      "grad_norm": 0.000723016622941941,
      "learning_rate": 0.2901325389385781,
      "loss": 0.2362,
      "num_input_tokens_seen": 3542432,
      "step": 4645
    },
    {
      "epoch": 9.667359667359667,
      "grad_norm": 0.0018010125495493412,
      "learning_rate": 0.2901115164522016,
      "loss": 0.2763,
      "num_input_tokens_seen": 3546240,
      "step": 4650
    },
    {
      "epoch": 9.677754677754677,
      "grad_norm": 0.001170750823803246,
      "learning_rate": 0.29009047235886865,
      "loss": 0.2732,
      "num_input_tokens_seen": 3550208,
      "step": 4655
    },
    {
      "epoch": 9.688149688149688,
      "grad_norm": 0.0016040452755987644,
      "learning_rate": 0.2900694066618243,
      "loss": 0.2917,
      "num_input_tokens_seen": 3554080,
      "step": 4660
    },
    {
      "epoch": 9.698544698544698,
      "grad_norm": 0.001458129845559597,
      "learning_rate": 0.2900483193643172,
      "loss": 0.2793,
      "num_input_tokens_seen": 3557984,
      "step": 4665
    },
    {
      "epoch": 9.70893970893971,
      "grad_norm": 0.001126165152527392,
      "learning_rate": 0.29002721046959934,
      "loss": 0.2779,
      "num_input_tokens_seen": 3561760,
      "step": 4670
    },
    {
      "epoch": 9.71933471933472,
      "grad_norm": 0.001675745821557939,
      "learning_rate": 0.29000607998092587,
      "loss": 0.2662,
      "num_input_tokens_seen": 3565600,
      "step": 4675
    },
    {
      "epoch": 9.72972972972973,
      "grad_norm": 0.0008758932235650718,
      "learning_rate": 0.2899849279015555,
      "loss": 0.2738,
      "num_input_tokens_seen": 3569408,
      "step": 4680
    },
    {
      "epoch": 9.74012474012474,
      "grad_norm": 0.0011675817659124732,
      "learning_rate": 0.28996375423475007,
      "loss": 0.2649,
      "num_input_tokens_seen": 3573376,
      "step": 4685
    },
    {
      "epoch": 9.75051975051975,
      "grad_norm": 0.0006994726718403399,
      "learning_rate": 0.28994255898377486,
      "loss": 0.2558,
      "num_input_tokens_seen": 3577312,
      "step": 4690
    },
    {
      "epoch": 9.76091476091476,
      "grad_norm": 0.0017045224085450172,
      "learning_rate": 0.2899213421518984,
      "loss": 0.2265,
      "num_input_tokens_seen": 3580992,
      "step": 4695
    },
    {
      "epoch": 9.771309771309772,
      "grad_norm": 0.0012412710348144174,
      "learning_rate": 0.2899001037423926,
      "loss": 0.2731,
      "num_input_tokens_seen": 3584768,
      "step": 4700
    },
    {
      "epoch": 9.781704781704782,
      "grad_norm": 0.001517456374131143,
      "learning_rate": 0.28987884375853273,
      "loss": 0.2737,
      "num_input_tokens_seen": 3588544,
      "step": 4705
    },
    {
      "epoch": 9.792099792099792,
      "grad_norm": 0.0004324537003412843,
      "learning_rate": 0.2898575622035974,
      "loss": 0.2732,
      "num_input_tokens_seen": 3592448,
      "step": 4710
    },
    {
      "epoch": 9.802494802494802,
      "grad_norm": 0.001739976811222732,
      "learning_rate": 0.2898362590808683,
      "loss": 0.2711,
      "num_input_tokens_seen": 3596352,
      "step": 4715
    },
    {
      "epoch": 9.812889812889813,
      "grad_norm": 0.0031527949031442404,
      "learning_rate": 0.2898149343936308,
      "loss": 0.2866,
      "num_input_tokens_seen": 3600224,
      "step": 4720
    },
    {
      "epoch": 9.823284823284823,
      "grad_norm": 0.002125125378370285,
      "learning_rate": 0.2897935881451734,
      "loss": 0.2814,
      "num_input_tokens_seen": 3604032,
      "step": 4725
    },
    {
      "epoch": 9.833679833679835,
      "grad_norm": 0.0005428296863101423,
      "learning_rate": 0.28977222033878797,
      "loss": 0.2461,
      "num_input_tokens_seen": 3607776,
      "step": 4730
    },
    {
      "epoch": 9.844074844074845,
      "grad_norm": 0.0017898917431011796,
      "learning_rate": 0.28975083097776966,
      "loss": 0.2721,
      "num_input_tokens_seen": 3611648,
      "step": 4735
    },
    {
      "epoch": 9.854469854469855,
      "grad_norm": 0.0004148696025367826,
      "learning_rate": 0.28972942006541696,
      "loss": 0.2744,
      "num_input_tokens_seen": 3615488,
      "step": 4740
    },
    {
      "epoch": 9.864864864864865,
      "grad_norm": 0.0014468338340520859,
      "learning_rate": 0.2897079876050318,
      "loss": 0.266,
      "num_input_tokens_seen": 3619136,
      "step": 4745
    },
    {
      "epoch": 9.875259875259875,
      "grad_norm": 8.109946065815166e-05,
      "learning_rate": 0.2896865335999192,
      "loss": 0.2658,
      "num_input_tokens_seen": 3622976,
      "step": 4750
    },
    {
      "epoch": 9.885654885654885,
      "grad_norm": 0.003306103404611349,
      "learning_rate": 0.28966505805338777,
      "loss": 0.2865,
      "num_input_tokens_seen": 3626944,
      "step": 4755
    },
    {
      "epoch": 9.896049896049895,
      "grad_norm": 0.0022545347455888987,
      "learning_rate": 0.2896435609687492,
      "loss": 0.2836,
      "num_input_tokens_seen": 3630944,
      "step": 4760
    },
    {
      "epoch": 9.906444906444907,
      "grad_norm": 0.0022805219050496817,
      "learning_rate": 0.2896220423493187,
      "loss": 0.2459,
      "num_input_tokens_seen": 3634752,
      "step": 4765
    },
    {
      "epoch": 9.916839916839917,
      "grad_norm": 0.00028122018557041883,
      "learning_rate": 0.28960050219841466,
      "loss": 0.2398,
      "num_input_tokens_seen": 3638592,
      "step": 4770
    },
    {
      "epoch": 9.927234927234927,
      "grad_norm": 0.0010744949104264379,
      "learning_rate": 0.28957894051935884,
      "loss": 0.253,
      "num_input_tokens_seen": 3642464,
      "step": 4775
    },
    {
      "epoch": 9.937629937629938,
      "grad_norm": 0.0008272434351965785,
      "learning_rate": 0.2895573573154764,
      "loss": 0.2655,
      "num_input_tokens_seen": 3646336,
      "step": 4780
    },
    {
      "epoch": 9.948024948024948,
      "grad_norm": 0.00025309371994808316,
      "learning_rate": 0.28953575259009556,
      "loss": 0.2646,
      "num_input_tokens_seen": 3650048,
      "step": 4785
    },
    {
      "epoch": 9.958419958419958,
      "grad_norm": 0.0006306581781245768,
      "learning_rate": 0.2895141263465482,
      "loss": 0.286,
      "num_input_tokens_seen": 3653856,
      "step": 4790
    },
    {
      "epoch": 9.96881496881497,
      "grad_norm": 0.0010191835463047028,
      "learning_rate": 0.28949247858816934,
      "loss": 0.2711,
      "num_input_tokens_seen": 3657664,
      "step": 4795
    },
    {
      "epoch": 9.97920997920998,
      "grad_norm": 0.00030977826099842787,
      "learning_rate": 0.2894708093182973,
      "loss": 0.249,
      "num_input_tokens_seen": 3661568,
      "step": 4800
    },
    {
      "epoch": 9.97920997920998,
      "eval_loss": 0.2519153952598572,
      "eval_runtime": 13.4137,
      "eval_samples_per_second": 63.815,
      "eval_steps_per_second": 15.954,
      "num_input_tokens_seen": 3661568,
      "step": 4800
    },
    {
      "epoch": 9.98960498960499,
      "grad_norm": 0.0008100526756606996,
      "learning_rate": 0.2894491185402737,
      "loss": 0.2487,
      "num_input_tokens_seen": 3665312,
      "step": 4805
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.0007240594713948667,
      "learning_rate": 0.2894274062574437,
      "loss": 0.268,
      "num_input_tokens_seen": 3669168,
      "step": 4810
    },
    {
      "epoch": 10.01039501039501,
      "grad_norm": 0.0007532486924901605,
      "learning_rate": 0.2894056724731554,
      "loss": 0.2423,
      "num_input_tokens_seen": 3672976,
      "step": 4815
    },
    {
      "epoch": 10.02079002079002,
      "grad_norm": 0.0018373773200437427,
      "learning_rate": 0.28938391719076056,
      "loss": 0.3322,
      "num_input_tokens_seen": 3676848,
      "step": 4820
    },
    {
      "epoch": 10.03118503118503,
      "grad_norm": 0.0018719711806625128,
      "learning_rate": 0.28936214041361413,
      "loss": 0.2777,
      "num_input_tokens_seen": 3680624,
      "step": 4825
    },
    {
      "epoch": 10.041580041580042,
      "grad_norm": 0.0002641514874994755,
      "learning_rate": 0.2893403421450743,
      "loss": 0.2711,
      "num_input_tokens_seen": 3684528,
      "step": 4830
    },
    {
      "epoch": 10.051975051975052,
      "grad_norm": 0.0002581313601695001,
      "learning_rate": 0.2893185223885026,
      "loss": 0.3069,
      "num_input_tokens_seen": 3688368,
      "step": 4835
    },
    {
      "epoch": 10.062370062370062,
      "grad_norm": 0.0007623252458870411,
      "learning_rate": 0.289296681147264,
      "loss": 0.3089,
      "num_input_tokens_seen": 3692144,
      "step": 4840
    },
    {
      "epoch": 10.072765072765073,
      "grad_norm": 0.00021265962277539074,
      "learning_rate": 0.28927481842472663,
      "loss": 0.3059,
      "num_input_tokens_seen": 3695984,
      "step": 4845
    },
    {
      "epoch": 10.083160083160083,
      "grad_norm": 0.0001783665647963062,
      "learning_rate": 0.28925293422426207,
      "loss": 0.2566,
      "num_input_tokens_seen": 3699664,
      "step": 4850
    },
    {
      "epoch": 10.093555093555093,
      "grad_norm": 0.0008891077013686299,
      "learning_rate": 0.28923102854924504,
      "loss": 0.2569,
      "num_input_tokens_seen": 3703344,
      "step": 4855
    },
    {
      "epoch": 10.103950103950105,
      "grad_norm": 0.001206408953294158,
      "learning_rate": 0.2892091014030537,
      "loss": 0.3077,
      "num_input_tokens_seen": 3707248,
      "step": 4860
    },
    {
      "epoch": 10.114345114345115,
      "grad_norm": 0.004202185198664665,
      "learning_rate": 0.2891871527890696,
      "loss": 0.3181,
      "num_input_tokens_seen": 3710992,
      "step": 4865
    },
    {
      "epoch": 10.124740124740125,
      "grad_norm": 0.0012446991167962551,
      "learning_rate": 0.2891651827106773,
      "loss": 0.2739,
      "num_input_tokens_seen": 3714704,
      "step": 4870
    },
    {
      "epoch": 10.135135135135135,
      "grad_norm": 0.00018531220848672092,
      "learning_rate": 0.2891431911712651,
      "loss": 0.2738,
      "num_input_tokens_seen": 3718480,
      "step": 4875
    },
    {
      "epoch": 10.145530145530145,
      "grad_norm": 0.00020502042025327682,
      "learning_rate": 0.2891211781742241,
      "loss": 0.296,
      "num_input_tokens_seen": 3722320,
      "step": 4880
    },
    {
      "epoch": 10.155925155925155,
      "grad_norm": 0.0010973821626976132,
      "learning_rate": 0.2890991437229492,
      "loss": 0.2642,
      "num_input_tokens_seen": 3726128,
      "step": 4885
    },
    {
      "epoch": 10.166320166320165,
      "grad_norm": 0.0007894599111750722,
      "learning_rate": 0.2890770878208383,
      "loss": 0.2794,
      "num_input_tokens_seen": 3729872,
      "step": 4890
    },
    {
      "epoch": 10.176715176715177,
      "grad_norm": 0.0010003134375438094,
      "learning_rate": 0.28905501047129273,
      "loss": 0.2636,
      "num_input_tokens_seen": 3733776,
      "step": 4895
    },
    {
      "epoch": 10.187110187110187,
      "grad_norm": 0.002415601396933198,
      "learning_rate": 0.289032911677717,
      "loss": 0.3135,
      "num_input_tokens_seen": 3737616,
      "step": 4900
    },
    {
      "epoch": 10.197505197505198,
      "grad_norm": 0.0018411775818094611,
      "learning_rate": 0.28901079144351915,
      "loss": 0.2818,
      "num_input_tokens_seen": 3741328,
      "step": 4905
    },
    {
      "epoch": 10.207900207900208,
      "grad_norm": 0.0011528501054272056,
      "learning_rate": 0.2889886497721103,
      "loss": 0.2914,
      "num_input_tokens_seen": 3745136,
      "step": 4910
    },
    {
      "epoch": 10.218295218295218,
      "grad_norm": 8.849740697769448e-05,
      "learning_rate": 0.28896648666690505,
      "loss": 0.2766,
      "num_input_tokens_seen": 3748912,
      "step": 4915
    },
    {
      "epoch": 10.228690228690228,
      "grad_norm": 0.00015786263975314796,
      "learning_rate": 0.2889443021313212,
      "loss": 0.2747,
      "num_input_tokens_seen": 3752688,
      "step": 4920
    },
    {
      "epoch": 10.23908523908524,
      "grad_norm": 0.0003860458673443645,
      "learning_rate": 0.28892209616877984,
      "loss": 0.2804,
      "num_input_tokens_seen": 3756432,
      "step": 4925
    },
    {
      "epoch": 10.24948024948025,
      "grad_norm": 0.0012466453481465578,
      "learning_rate": 0.28889986878270546,
      "loss": 0.2688,
      "num_input_tokens_seen": 3760272,
      "step": 4930
    },
    {
      "epoch": 10.25987525987526,
      "grad_norm": 0.001825141254812479,
      "learning_rate": 0.28887761997652583,
      "loss": 0.2721,
      "num_input_tokens_seen": 3764144,
      "step": 4935
    },
    {
      "epoch": 10.27027027027027,
      "grad_norm": 0.0011676604626700282,
      "learning_rate": 0.2888553497536719,
      "loss": 0.265,
      "num_input_tokens_seen": 3767984,
      "step": 4940
    },
    {
      "epoch": 10.28066528066528,
      "grad_norm": 0.00023977972159627825,
      "learning_rate": 0.2888330581175781,
      "loss": 0.258,
      "num_input_tokens_seen": 3771760,
      "step": 4945
    },
    {
      "epoch": 10.29106029106029,
      "grad_norm": 0.0003496152930893004,
      "learning_rate": 0.28881074507168203,
      "loss": 0.2628,
      "num_input_tokens_seen": 3775664,
      "step": 4950
    },
    {
      "epoch": 10.301455301455302,
      "grad_norm": 0.0007483397494070232,
      "learning_rate": 0.2887884106194247,
      "loss": 0.2933,
      "num_input_tokens_seen": 3779408,
      "step": 4955
    },
    {
      "epoch": 10.311850311850312,
      "grad_norm": 0.0020368245895951986,
      "learning_rate": 0.28876605476425027,
      "loss": 0.2879,
      "num_input_tokens_seen": 3783248,
      "step": 4960
    },
    {
      "epoch": 10.322245322245323,
      "grad_norm": 0.0022097816690802574,
      "learning_rate": 0.2887436775096064,
      "loss": 0.3185,
      "num_input_tokens_seen": 3787024,
      "step": 4965
    },
    {
      "epoch": 10.332640332640333,
      "grad_norm": 0.0009934940608218312,
      "learning_rate": 0.2887212788589439,
      "loss": 0.2942,
      "num_input_tokens_seen": 3790800,
      "step": 4970
    },
    {
      "epoch": 10.343035343035343,
      "grad_norm": 0.0009917601710185409,
      "learning_rate": 0.2886988588157169,
      "loss": 0.2874,
      "num_input_tokens_seen": 3794672,
      "step": 4975
    },
    {
      "epoch": 10.353430353430353,
      "grad_norm": 0.0005199462757445872,
      "learning_rate": 0.28867641738338284,
      "loss": 0.2899,
      "num_input_tokens_seen": 3798352,
      "step": 4980
    },
    {
      "epoch": 10.363825363825363,
      "grad_norm": 0.0004383198684081435,
      "learning_rate": 0.2886539545654026,
      "loss": 0.2463,
      "num_input_tokens_seen": 3802000,
      "step": 4985
    },
    {
      "epoch": 10.374220374220375,
      "grad_norm": 0.001138963270932436,
      "learning_rate": 0.28863147036524006,
      "loss": 0.2847,
      "num_input_tokens_seen": 3805840,
      "step": 4990
    },
    {
      "epoch": 10.384615384615385,
      "grad_norm": 0.00042424467392265797,
      "learning_rate": 0.2886089647863626,
      "loss": 0.2859,
      "num_input_tokens_seen": 3809680,
      "step": 4995
    },
    {
      "epoch": 10.395010395010395,
      "grad_norm": 0.000569929659832269,
      "learning_rate": 0.288586437832241,
      "loss": 0.3018,
      "num_input_tokens_seen": 3813552,
      "step": 5000
    },
    {
      "epoch": 10.395010395010395,
      "eval_loss": 0.27405840158462524,
      "eval_runtime": 13.4322,
      "eval_samples_per_second": 63.727,
      "eval_steps_per_second": 15.932,
      "num_input_tokens_seen": 3813552,
      "step": 5000
    },
    {
      "epoch": 10.405405405405405,
      "grad_norm": 0.0006497822469100356,
      "learning_rate": 0.28856388950634904,
      "loss": 0.2645,
      "num_input_tokens_seen": 3817328,
      "step": 5005
    },
    {
      "epoch": 10.415800415800415,
      "grad_norm": 0.0019462925847619772,
      "learning_rate": 0.288541319812164,
      "loss": 0.3245,
      "num_input_tokens_seen": 3821136,
      "step": 5010
    },
    {
      "epoch": 10.426195426195425,
      "grad_norm": 0.0034500728361308575,
      "learning_rate": 0.2885187287531665,
      "loss": 0.3026,
      "num_input_tokens_seen": 3825168,
      "step": 5015
    },
    {
      "epoch": 10.436590436590437,
      "grad_norm": 0.0009363286662846804,
      "learning_rate": 0.2884961163328402,
      "loss": 0.2867,
      "num_input_tokens_seen": 3828848,
      "step": 5020
    },
    {
      "epoch": 10.446985446985448,
      "grad_norm": 0.0009797376114875078,
      "learning_rate": 0.28847348255467237,
      "loss": 0.3134,
      "num_input_tokens_seen": 3832720,
      "step": 5025
    },
    {
      "epoch": 10.457380457380458,
      "grad_norm": 0.0008015984203666449,
      "learning_rate": 0.28845082742215333,
      "loss": 0.2627,
      "num_input_tokens_seen": 3836528,
      "step": 5030
    },
    {
      "epoch": 10.467775467775468,
      "grad_norm": 0.0001685633324086666,
      "learning_rate": 0.2884281509387769,
      "loss": 0.274,
      "num_input_tokens_seen": 3840304,
      "step": 5035
    },
    {
      "epoch": 10.478170478170478,
      "grad_norm": 0.0004156519134994596,
      "learning_rate": 0.2884054531080399,
      "loss": 0.2737,
      "num_input_tokens_seen": 3844240,
      "step": 5040
    },
    {
      "epoch": 10.488565488565488,
      "grad_norm": 0.0005527977482415736,
      "learning_rate": 0.28838273393344277,
      "loss": 0.2226,
      "num_input_tokens_seen": 3848080,
      "step": 5045
    },
    {
      "epoch": 10.4989604989605,
      "grad_norm": 0.0012770200846716762,
      "learning_rate": 0.288359993418489,
      "loss": 0.3116,
      "num_input_tokens_seen": 3851856,
      "step": 5050
    },
    {
      "epoch": 10.50935550935551,
      "grad_norm": 0.0024924289900809526,
      "learning_rate": 0.28833723156668556,
      "loss": 0.3031,
      "num_input_tokens_seen": 3855760,
      "step": 5055
    },
    {
      "epoch": 10.51975051975052,
      "grad_norm": 0.0009867752669379115,
      "learning_rate": 0.2883144483815425,
      "loss": 0.294,
      "num_input_tokens_seen": 3859664,
      "step": 5060
    },
    {
      "epoch": 10.53014553014553,
      "grad_norm": 0.0009886941406875849,
      "learning_rate": 0.28829164386657335,
      "loss": 0.2859,
      "num_input_tokens_seen": 3863568,
      "step": 5065
    },
    {
      "epoch": 10.54054054054054,
      "grad_norm": 0.0011199767468497157,
      "learning_rate": 0.28826881802529486,
      "loss": 0.2485,
      "num_input_tokens_seen": 3867408,
      "step": 5070
    },
    {
      "epoch": 10.55093555093555,
      "grad_norm": 0.00019102166697848588,
      "learning_rate": 0.28824597086122705,
      "loss": 0.2301,
      "num_input_tokens_seen": 3871152,
      "step": 5075
    },
    {
      "epoch": 10.56133056133056,
      "grad_norm": 0.00044801447074860334,
      "learning_rate": 0.28822310237789317,
      "loss": 0.3111,
      "num_input_tokens_seen": 3874896,
      "step": 5080
    },
    {
      "epoch": 10.571725571725572,
      "grad_norm": 0.0004440463671926409,
      "learning_rate": 0.2882002125788199,
      "loss": 0.2822,
      "num_input_tokens_seen": 3878640,
      "step": 5085
    },
    {
      "epoch": 10.582120582120583,
      "grad_norm": 0.0011579656274989247,
      "learning_rate": 0.2881773014675371,
      "loss": 0.2793,
      "num_input_tokens_seen": 3882416,
      "step": 5090
    },
    {
      "epoch": 10.592515592515593,
      "grad_norm": 0.0009008412598632276,
      "learning_rate": 0.288154369047578,
      "loss": 0.2753,
      "num_input_tokens_seen": 3886288,
      "step": 5095
    },
    {
      "epoch": 10.602910602910603,
      "grad_norm": 0.0005793485324829817,
      "learning_rate": 0.28813141532247905,
      "loss": 0.2861,
      "num_input_tokens_seen": 3890192,
      "step": 5100
    },
    {
      "epoch": 10.613305613305613,
      "grad_norm": 0.00036320521030575037,
      "learning_rate": 0.28810844029578,
      "loss": 0.2742,
      "num_input_tokens_seen": 3894096,
      "step": 5105
    },
    {
      "epoch": 10.623700623700623,
      "grad_norm": 0.0015469376230612397,
      "learning_rate": 0.2880854439710238,
      "loss": 0.2659,
      "num_input_tokens_seen": 3897904,
      "step": 5110
    },
    {
      "epoch": 10.634095634095633,
      "grad_norm": 0.0001771929964888841,
      "learning_rate": 0.28806242635175694,
      "loss": 0.2747,
      "num_input_tokens_seen": 3901776,
      "step": 5115
    },
    {
      "epoch": 10.644490644490645,
      "grad_norm": 0.0004131075693294406,
      "learning_rate": 0.2880393874415289,
      "loss": 0.2845,
      "num_input_tokens_seen": 3905552,
      "step": 5120
    },
    {
      "epoch": 10.654885654885655,
      "grad_norm": 0.0006141449557617307,
      "learning_rate": 0.2880163272438926,
      "loss": 0.2829,
      "num_input_tokens_seen": 3909360,
      "step": 5125
    },
    {
      "epoch": 10.665280665280665,
      "grad_norm": 0.0007640831172466278,
      "learning_rate": 0.2879932457624042,
      "loss": 0.258,
      "num_input_tokens_seen": 3913200,
      "step": 5130
    },
    {
      "epoch": 10.675675675675675,
      "grad_norm": 0.0011102590942755342,
      "learning_rate": 0.2879701430006232,
      "loss": 0.2748,
      "num_input_tokens_seen": 3917104,
      "step": 5135
    },
    {
      "epoch": 10.686070686070686,
      "grad_norm": 0.0016640633111819625,
      "learning_rate": 0.28794701896211233,
      "loss": 0.287,
      "num_input_tokens_seen": 3920944,
      "step": 5140
    },
    {
      "epoch": 10.696465696465696,
      "grad_norm": 0.0001840692275436595,
      "learning_rate": 0.28792387365043753,
      "loss": 0.2702,
      "num_input_tokens_seen": 3924656,
      "step": 5145
    },
    {
      "epoch": 10.706860706860708,
      "grad_norm": 0.0005854679620824754,
      "learning_rate": 0.28790070706916815,
      "loss": 0.3082,
      "num_input_tokens_seen": 3928464,
      "step": 5150
    },
    {
      "epoch": 10.717255717255718,
      "grad_norm": 0.0020343204960227013,
      "learning_rate": 0.2878775192218768,
      "loss": 0.2626,
      "num_input_tokens_seen": 3932592,
      "step": 5155
    },
    {
      "epoch": 10.727650727650728,
      "grad_norm": 0.000736934831365943,
      "learning_rate": 0.2878543101121393,
      "loss": 0.2649,
      "num_input_tokens_seen": 3936464,
      "step": 5160
    },
    {
      "epoch": 10.738045738045738,
      "grad_norm": 0.00013142768875695765,
      "learning_rate": 0.28783107974353483,
      "loss": 0.256,
      "num_input_tokens_seen": 3940304,
      "step": 5165
    },
    {
      "epoch": 10.748440748440748,
      "grad_norm": 0.0011673502158373594,
      "learning_rate": 0.2878078281196457,
      "loss": 0.2795,
      "num_input_tokens_seen": 3944048,
      "step": 5170
    },
    {
      "epoch": 10.758835758835758,
      "grad_norm": 0.001666890224441886,
      "learning_rate": 0.28778455524405777,
      "loss": 0.2791,
      "num_input_tokens_seen": 3947824,
      "step": 5175
    },
    {
      "epoch": 10.76923076923077,
      "grad_norm": 0.0003241396916564554,
      "learning_rate": 0.2877612611203598,
      "loss": 0.2596,
      "num_input_tokens_seen": 3951568,
      "step": 5180
    },
    {
      "epoch": 10.77962577962578,
      "grad_norm": 0.00018004632147494704,
      "learning_rate": 0.28773794575214423,
      "loss": 0.27,
      "num_input_tokens_seen": 3955408,
      "step": 5185
    },
    {
      "epoch": 10.79002079002079,
      "grad_norm": 0.0014246640494093299,
      "learning_rate": 0.28771460914300645,
      "loss": 0.2948,
      "num_input_tokens_seen": 3959280,
      "step": 5190
    },
    {
      "epoch": 10.8004158004158,
      "grad_norm": 0.00027323196991346776,
      "learning_rate": 0.2876912512965454,
      "loss": 0.2778,
      "num_input_tokens_seen": 3963216,
      "step": 5195
    },
    {
      "epoch": 10.81081081081081,
      "grad_norm": 0.0006623450899496675,
      "learning_rate": 0.287667872216363,
      "loss": 0.2871,
      "num_input_tokens_seen": 3967024,
      "step": 5200
    },
    {
      "epoch": 10.81081081081081,
      "eval_loss": 0.2504214942455292,
      "eval_runtime": 13.433,
      "eval_samples_per_second": 63.723,
      "eval_steps_per_second": 15.931,
      "num_input_tokens_seen": 3967024,
      "step": 5200
    },
    {
      "epoch": 10.82120582120582,
      "grad_norm": 0.0001593843480804935,
      "learning_rate": 0.2876444719060647,
      "loss": 0.2564,
      "num_input_tokens_seen": 3970896,
      "step": 5205
    },
    {
      "epoch": 10.83160083160083,
      "grad_norm": 0.002191289095208049,
      "learning_rate": 0.287621050369259,
      "loss": 0.2633,
      "num_input_tokens_seen": 3974832,
      "step": 5210
    },
    {
      "epoch": 10.841995841995843,
      "grad_norm": 0.00039982551243156195,
      "learning_rate": 0.28759760760955794,
      "loss": 0.2646,
      "num_input_tokens_seen": 3978704,
      "step": 5215
    },
    {
      "epoch": 10.852390852390853,
      "grad_norm": 0.00032265440677292645,
      "learning_rate": 0.2875741436305766,
      "loss": 0.2671,
      "num_input_tokens_seen": 3982448,
      "step": 5220
    },
    {
      "epoch": 10.862785862785863,
      "grad_norm": 0.0006903450121171772,
      "learning_rate": 0.28755065843593347,
      "loss": 0.2647,
      "num_input_tokens_seen": 3986160,
      "step": 5225
    },
    {
      "epoch": 10.873180873180873,
      "grad_norm": 0.0008125486201606691,
      "learning_rate": 0.2875271520292502,
      "loss": 0.2784,
      "num_input_tokens_seen": 3990000,
      "step": 5230
    },
    {
      "epoch": 10.883575883575883,
      "grad_norm": 0.0006441565928980708,
      "learning_rate": 0.28750362441415184,
      "loss": 0.2813,
      "num_input_tokens_seen": 3993744,
      "step": 5235
    },
    {
      "epoch": 10.893970893970893,
      "grad_norm": 0.0004949583671987057,
      "learning_rate": 0.28748007559426664,
      "loss": 0.2646,
      "num_input_tokens_seen": 3997552,
      "step": 5240
    },
    {
      "epoch": 10.904365904365905,
      "grad_norm": 0.00015367277956102043,
      "learning_rate": 0.2874565055732261,
      "loss": 0.2807,
      "num_input_tokens_seen": 4001392,
      "step": 5245
    },
    {
      "epoch": 10.914760914760915,
      "grad_norm": 0.0006494914414361119,
      "learning_rate": 0.28743291435466495,
      "loss": 0.2587,
      "num_input_tokens_seen": 4005232,
      "step": 5250
    },
    {
      "epoch": 10.925155925155925,
      "grad_norm": 0.0008468227460980415,
      "learning_rate": 0.2874093019422214,
      "loss": 0.2684,
      "num_input_tokens_seen": 4009008,
      "step": 5255
    },
    {
      "epoch": 10.935550935550935,
      "grad_norm": 0.00025040569016709924,
      "learning_rate": 0.28738566833953666,
      "loss": 0.2502,
      "num_input_tokens_seen": 4012848,
      "step": 5260
    },
    {
      "epoch": 10.945945945945946,
      "grad_norm": 0.00017516358639113605,
      "learning_rate": 0.28736201355025537,
      "loss": 0.279,
      "num_input_tokens_seen": 4016560,
      "step": 5265
    },
    {
      "epoch": 10.956340956340956,
      "grad_norm": 0.0001797443110262975,
      "learning_rate": 0.28733833757802535,
      "loss": 0.2495,
      "num_input_tokens_seen": 4020400,
      "step": 5270
    },
    {
      "epoch": 10.966735966735968,
      "grad_norm": 8.787220576778054e-05,
      "learning_rate": 0.28731464042649785,
      "loss": 0.2736,
      "num_input_tokens_seen": 4024144,
      "step": 5275
    },
    {
      "epoch": 10.977130977130978,
      "grad_norm": 0.0016055147862061858,
      "learning_rate": 0.2872909220993271,
      "loss": 0.2709,
      "num_input_tokens_seen": 4027920,
      "step": 5280
    },
    {
      "epoch": 10.987525987525988,
      "grad_norm": 0.002014842117205262,
      "learning_rate": 0.287267182600171,
      "loss": 0.2791,
      "num_input_tokens_seen": 4031664,
      "step": 5285
    },
    {
      "epoch": 10.997920997920998,
      "grad_norm": 0.0012715302873402834,
      "learning_rate": 0.2872434219326902,
      "loss": 0.2464,
      "num_input_tokens_seen": 4035408,
      "step": 5290
    },
    {
      "epoch": 11.008316008316008,
      "grad_norm": 0.0002326093817828223,
      "learning_rate": 0.28721964010054907,
      "loss": 0.2573,
      "num_input_tokens_seen": 4039168,
      "step": 5295
    },
    {
      "epoch": 11.018711018711018,
      "grad_norm": 0.0003226780681870878,
      "learning_rate": 0.28719583710741503,
      "loss": 0.2841,
      "num_input_tokens_seen": 4043008,
      "step": 5300
    },
    {
      "epoch": 11.029106029106028,
      "grad_norm": 0.0005045076250098646,
      "learning_rate": 0.28717201295695877,
      "loss": 0.2898,
      "num_input_tokens_seen": 4046880,
      "step": 5305
    },
    {
      "epoch": 11.03950103950104,
      "grad_norm": 0.00042858265805989504,
      "learning_rate": 0.28714816765285434,
      "loss": 0.2714,
      "num_input_tokens_seen": 4050720,
      "step": 5310
    },
    {
      "epoch": 11.04989604989605,
      "grad_norm": 0.000739100796636194,
      "learning_rate": 0.28712430119877896,
      "loss": 0.2866,
      "num_input_tokens_seen": 4054560,
      "step": 5315
    },
    {
      "epoch": 11.06029106029106,
      "grad_norm": 0.0007063311641104519,
      "learning_rate": 0.28710041359841304,
      "loss": 0.2742,
      "num_input_tokens_seen": 4058304,
      "step": 5320
    },
    {
      "epoch": 11.07068607068607,
      "grad_norm": 0.0016039853217080235,
      "learning_rate": 0.28707650485544056,
      "loss": 0.2628,
      "num_input_tokens_seen": 4062272,
      "step": 5325
    },
    {
      "epoch": 11.08108108108108,
      "grad_norm": 0.0008827102719806135,
      "learning_rate": 0.28705257497354836,
      "loss": 0.2765,
      "num_input_tokens_seen": 4066112,
      "step": 5330
    },
    {
      "epoch": 11.09147609147609,
      "grad_norm": 0.0001345526980003342,
      "learning_rate": 0.28702862395642675,
      "loss": 0.2678,
      "num_input_tokens_seen": 4069952,
      "step": 5335
    },
    {
      "epoch": 11.101871101871103,
      "grad_norm": 0.0012599313631653786,
      "learning_rate": 0.28700465180776935,
      "loss": 0.2699,
      "num_input_tokens_seen": 4073824,
      "step": 5340
    },
    {
      "epoch": 11.112266112266113,
      "grad_norm": 0.0002967861364595592,
      "learning_rate": 0.2869806585312729,
      "loss": 0.2695,
      "num_input_tokens_seen": 4077664,
      "step": 5345
    },
    {
      "epoch": 11.122661122661123,
      "grad_norm": 0.001670153345912695,
      "learning_rate": 0.28695664413063754,
      "loss": 0.2905,
      "num_input_tokens_seen": 4081600,
      "step": 5350
    },
    {
      "epoch": 11.133056133056133,
      "grad_norm": 0.0021830438636243343,
      "learning_rate": 0.28693260860956654,
      "loss": 0.3019,
      "num_input_tokens_seen": 4085472,
      "step": 5355
    },
    {
      "epoch": 11.143451143451143,
      "grad_norm": 0.0011994813103228807,
      "learning_rate": 0.2869085519717665,
      "loss": 0.2787,
      "num_input_tokens_seen": 4089280,
      "step": 5360
    },
    {
      "epoch": 11.153846153846153,
      "grad_norm": 0.00024140537425410002,
      "learning_rate": 0.28688447422094726,
      "loss": 0.2739,
      "num_input_tokens_seen": 4093056,
      "step": 5365
    },
    {
      "epoch": 11.164241164241163,
      "grad_norm": 0.0010804218472912908,
      "learning_rate": 0.2868603753608219,
      "loss": 0.2628,
      "num_input_tokens_seen": 4096928,
      "step": 5370
    },
    {
      "epoch": 11.174636174636175,
      "grad_norm": 0.00030141184106469154,
      "learning_rate": 0.28683625539510665,
      "loss": 0.2738,
      "num_input_tokens_seen": 4100768,
      "step": 5375
    },
    {
      "epoch": 11.185031185031185,
      "grad_norm": 0.0006643576198257506,
      "learning_rate": 0.28681211432752135,
      "loss": 0.2672,
      "num_input_tokens_seen": 4104608,
      "step": 5380
    },
    {
      "epoch": 11.195426195426196,
      "grad_norm": 0.0011604147730395198,
      "learning_rate": 0.2867879521617887,
      "loss": 0.2773,
      "num_input_tokens_seen": 4108320,
      "step": 5385
    },
    {
      "epoch": 11.205821205821206,
      "grad_norm": 0.0005744689260609448,
      "learning_rate": 0.28676376890163485,
      "loss": 0.2748,
      "num_input_tokens_seen": 4112224,
      "step": 5390
    },
    {
      "epoch": 11.216216216216216,
      "grad_norm": 0.0015742257237434387,
      "learning_rate": 0.2867395645507891,
      "loss": 0.2811,
      "num_input_tokens_seen": 4116032,
      "step": 5395
    },
    {
      "epoch": 11.226611226611226,
      "grad_norm": 0.0003916473069693893,
      "learning_rate": 0.2867153391129842,
      "loss": 0.2634,
      "num_input_tokens_seen": 4120032,
      "step": 5400
    },
    {
      "epoch": 11.226611226611226,
      "eval_loss": 0.24936148524284363,
      "eval_runtime": 13.4616,
      "eval_samples_per_second": 63.589,
      "eval_steps_per_second": 15.897,
      "num_input_tokens_seen": 4120032,
      "step": 5400
    },
    {
      "epoch": 11.237006237006238,
      "grad_norm": 0.0012456130934879184,
      "learning_rate": 0.28669109259195585,
      "loss": 0.2167,
      "num_input_tokens_seen": 4123680,
      "step": 5405
    },
    {
      "epoch": 11.247401247401248,
      "grad_norm": 0.000903849839232862,
      "learning_rate": 0.2866668249914433,
      "loss": 0.2951,
      "num_input_tokens_seen": 4127712,
      "step": 5410
    },
    {
      "epoch": 11.257796257796258,
      "grad_norm": 0.0006740223034285009,
      "learning_rate": 0.2866425363151889,
      "loss": 0.2703,
      "num_input_tokens_seen": 4131360,
      "step": 5415
    },
    {
      "epoch": 11.268191268191268,
      "grad_norm": 0.00018547267245594412,
      "learning_rate": 0.2866182265669382,
      "loss": 0.2725,
      "num_input_tokens_seen": 4135200,
      "step": 5420
    },
    {
      "epoch": 11.278586278586278,
      "grad_norm": 0.0008092054631561041,
      "learning_rate": 0.28659389575044014,
      "loss": 0.2718,
      "num_input_tokens_seen": 4139040,
      "step": 5425
    },
    {
      "epoch": 11.288981288981288,
      "grad_norm": 0.0010923068039119244,
      "learning_rate": 0.28656954386944683,
      "loss": 0.296,
      "num_input_tokens_seen": 4142880,
      "step": 5430
    },
    {
      "epoch": 11.299376299376299,
      "grad_norm": 9.251588198821992e-05,
      "learning_rate": 0.28654517092771353,
      "loss": 0.2951,
      "num_input_tokens_seen": 4146496,
      "step": 5435
    },
    {
      "epoch": 11.30977130977131,
      "grad_norm": 0.0004181528929620981,
      "learning_rate": 0.286520776928999,
      "loss": 0.2586,
      "num_input_tokens_seen": 4150272,
      "step": 5440
    },
    {
      "epoch": 11.32016632016632,
      "grad_norm": 0.0017176421824842691,
      "learning_rate": 0.286496361877065,
      "loss": 0.2982,
      "num_input_tokens_seen": 4154080,
      "step": 5445
    },
    {
      "epoch": 11.33056133056133,
      "grad_norm": 0.001848685904406011,
      "learning_rate": 0.28647192577567676,
      "loss": 0.2637,
      "num_input_tokens_seen": 4157984,
      "step": 5450
    },
    {
      "epoch": 11.34095634095634,
      "grad_norm": 0.0011772282887250185,
      "learning_rate": 0.28644746862860254,
      "loss": 0.2825,
      "num_input_tokens_seen": 4161792,
      "step": 5455
    },
    {
      "epoch": 11.35135135135135,
      "grad_norm": 0.0002183064352720976,
      "learning_rate": 0.2864229904396139,
      "loss": 0.2804,
      "num_input_tokens_seen": 4165600,
      "step": 5460
    },
    {
      "epoch": 11.361746361746361,
      "grad_norm": 0.0003906710771843791,
      "learning_rate": 0.28639849121248573,
      "loss": 0.2626,
      "num_input_tokens_seen": 4169440,
      "step": 5465
    },
    {
      "epoch": 11.372141372141373,
      "grad_norm": 0.0009167538373731077,
      "learning_rate": 0.28637397095099615,
      "loss": 0.2665,
      "num_input_tokens_seen": 4173280,
      "step": 5470
    },
    {
      "epoch": 11.382536382536383,
      "grad_norm": 0.000585370697081089,
      "learning_rate": 0.28634942965892646,
      "loss": 0.2475,
      "num_input_tokens_seen": 4177120,
      "step": 5475
    },
    {
      "epoch": 11.392931392931393,
      "grad_norm": 0.0013894661096855998,
      "learning_rate": 0.28632486734006124,
      "loss": 0.2801,
      "num_input_tokens_seen": 4180960,
      "step": 5480
    },
    {
      "epoch": 11.403326403326403,
      "grad_norm": 0.0002715508744586259,
      "learning_rate": 0.28630028399818835,
      "loss": 0.2564,
      "num_input_tokens_seen": 4184832,
      "step": 5485
    },
    {
      "epoch": 11.413721413721413,
      "grad_norm": 0.0013098849449306726,
      "learning_rate": 0.2862756796370987,
      "loss": 0.2695,
      "num_input_tokens_seen": 4188736,
      "step": 5490
    },
    {
      "epoch": 11.424116424116423,
      "grad_norm": 9.128142846748233e-05,
      "learning_rate": 0.2862510542605868,
      "loss": 0.2454,
      "num_input_tokens_seen": 4192384,
      "step": 5495
    },
    {
      "epoch": 11.434511434511435,
      "grad_norm": 0.0007413947605527937,
      "learning_rate": 0.2862264078724501,
      "loss": 0.2604,
      "num_input_tokens_seen": 4196256,
      "step": 5500
    },
    {
      "epoch": 11.444906444906445,
      "grad_norm": 0.002100604586303234,
      "learning_rate": 0.28620174047648933,
      "loss": 0.2803,
      "num_input_tokens_seen": 4199904,
      "step": 5505
    },
    {
      "epoch": 11.455301455301456,
      "grad_norm": 0.0005267172819003463,
      "learning_rate": 0.2861770520765086,
      "loss": 0.2407,
      "num_input_tokens_seen": 4203648,
      "step": 5510
    },
    {
      "epoch": 11.465696465696466,
      "grad_norm": 0.0012598353205248713,
      "learning_rate": 0.2861523426763151,
      "loss": 0.3304,
      "num_input_tokens_seen": 4207488,
      "step": 5515
    },
    {
      "epoch": 11.476091476091476,
      "grad_norm": 0.0008330654818564653,
      "learning_rate": 0.2861276122797194,
      "loss": 0.2494,
      "num_input_tokens_seen": 4211200,
      "step": 5520
    },
    {
      "epoch": 11.486486486486486,
      "grad_norm": 0.0005036956281401217,
      "learning_rate": 0.28610286089053516,
      "loss": 0.2601,
      "num_input_tokens_seen": 4215072,
      "step": 5525
    },
    {
      "epoch": 11.496881496881496,
      "grad_norm": 0.00010328643838874996,
      "learning_rate": 0.28607808851257943,
      "loss": 0.2409,
      "num_input_tokens_seen": 4218944,
      "step": 5530
    },
    {
      "epoch": 11.507276507276508,
      "grad_norm": 0.0005854734918102622,
      "learning_rate": 0.28605329514967237,
      "loss": 0.2761,
      "num_input_tokens_seen": 4222816,
      "step": 5535
    },
    {
      "epoch": 11.517671517671518,
      "grad_norm": 0.0015499501023441553,
      "learning_rate": 0.2860284808056374,
      "loss": 0.2498,
      "num_input_tokens_seen": 4226688,
      "step": 5540
    },
    {
      "epoch": 11.528066528066528,
      "grad_norm": 0.0008388102287426591,
      "learning_rate": 0.28600364548430135,
      "loss": 0.2873,
      "num_input_tokens_seen": 4230496,
      "step": 5545
    },
    {
      "epoch": 11.538461538461538,
      "grad_norm": 0.0006467925850301981,
      "learning_rate": 0.28597878918949393,
      "loss": 0.2689,
      "num_input_tokens_seen": 4234272,
      "step": 5550
    },
    {
      "epoch": 11.548856548856548,
      "grad_norm": 7.578585064038634e-05,
      "learning_rate": 0.2859539119250485,
      "loss": 0.2648,
      "num_input_tokens_seen": 4238272,
      "step": 5555
    },
    {
      "epoch": 11.559251559251559,
      "grad_norm": 0.0008321469649672508,
      "learning_rate": 0.2859290136948013,
      "loss": 0.2671,
      "num_input_tokens_seen": 4241920,
      "step": 5560
    },
    {
      "epoch": 11.56964656964657,
      "grad_norm": 0.000922494160477072,
      "learning_rate": 0.28590409450259197,
      "loss": 0.2747,
      "num_input_tokens_seen": 4245632,
      "step": 5565
    },
    {
      "epoch": 11.58004158004158,
      "grad_norm": 0.0004655584925785661,
      "learning_rate": 0.28587915435226346,
      "loss": 0.271,
      "num_input_tokens_seen": 4249472,
      "step": 5570
    },
    {
      "epoch": 11.59043659043659,
      "grad_norm": 0.001390872523188591,
      "learning_rate": 0.2858541932476617,
      "loss": 0.2483,
      "num_input_tokens_seen": 4253376,
      "step": 5575
    },
    {
      "epoch": 11.6008316008316,
      "grad_norm": 0.0007518345373682678,
      "learning_rate": 0.2858292111926361,
      "loss": 0.3153,
      "num_input_tokens_seen": 4257152,
      "step": 5580
    },
    {
      "epoch": 11.611226611226611,
      "grad_norm": 0.0001840638869907707,
      "learning_rate": 0.28580420819103924,
      "loss": 0.2759,
      "num_input_tokens_seen": 4260960,
      "step": 5585
    },
    {
      "epoch": 11.621621621621621,
      "grad_norm": 0.0008532400825060904,
      "learning_rate": 0.2857791842467269,
      "loss": 0.2764,
      "num_input_tokens_seen": 4264832,
      "step": 5590
    },
    {
      "epoch": 11.632016632016633,
      "grad_norm": 0.00016213746857829392,
      "learning_rate": 0.2857541393635579,
      "loss": 0.2697,
      "num_input_tokens_seen": 4268704,
      "step": 5595
    },
    {
      "epoch": 11.642411642411643,
      "grad_norm": 0.0005190189694985747,
      "learning_rate": 0.2857290735453948,
      "loss": 0.2821,
      "num_input_tokens_seen": 4272608,
      "step": 5600
    },
    {
      "epoch": 11.642411642411643,
      "eval_loss": 0.2751775085926056,
      "eval_runtime": 13.4019,
      "eval_samples_per_second": 63.872,
      "eval_steps_per_second": 15.968,
      "num_input_tokens_seen": 4272608,
      "step": 5600
    },
    {
      "epoch": 11.652806652806653,
      "grad_norm": 0.0018004627199843526,
      "learning_rate": 0.28570398679610276,
      "loss": 0.2563,
      "num_input_tokens_seen": 4276480,
      "step": 5605
    },
    {
      "epoch": 11.663201663201663,
      "grad_norm": 0.0011417015921324492,
      "learning_rate": 0.2856788791195506,
      "loss": 0.2325,
      "num_input_tokens_seen": 4280160,
      "step": 5610
    },
    {
      "epoch": 11.673596673596673,
      "grad_norm": 0.0016824102494865656,
      "learning_rate": 0.28565375051961023,
      "loss": 0.3769,
      "num_input_tokens_seen": 4284032,
      "step": 5615
    },
    {
      "epoch": 11.683991683991684,
      "grad_norm": 0.0011130665661767125,
      "learning_rate": 0.28562860100015686,
      "loss": 0.2992,
      "num_input_tokens_seen": 4288000,
      "step": 5620
    },
    {
      "epoch": 11.694386694386694,
      "grad_norm": 0.0013533191522583365,
      "learning_rate": 0.2856034305650687,
      "loss": 0.3112,
      "num_input_tokens_seen": 4291776,
      "step": 5625
    },
    {
      "epoch": 11.704781704781706,
      "grad_norm": 0.0015348048182204366,
      "learning_rate": 0.28557823921822756,
      "loss": 0.2866,
      "num_input_tokens_seen": 4295552,
      "step": 5630
    },
    {
      "epoch": 11.715176715176716,
      "grad_norm": 0.002123083220794797,
      "learning_rate": 0.2855530269635181,
      "loss": 0.2612,
      "num_input_tokens_seen": 4299232,
      "step": 5635
    },
    {
      "epoch": 11.725571725571726,
      "grad_norm": 0.000828628137242049,
      "learning_rate": 0.2855277938048284,
      "loss": 0.2684,
      "num_input_tokens_seen": 4303168,
      "step": 5640
    },
    {
      "epoch": 11.735966735966736,
      "grad_norm": 0.00045358893112279475,
      "learning_rate": 0.2855025397460498,
      "loss": 0.2631,
      "num_input_tokens_seen": 4306848,
      "step": 5645
    },
    {
      "epoch": 11.746361746361746,
      "grad_norm": 0.001199671532958746,
      "learning_rate": 0.28547726479107666,
      "loss": 0.2802,
      "num_input_tokens_seen": 4310560,
      "step": 5650
    },
    {
      "epoch": 11.756756756756756,
      "grad_norm": 0.00013404253695625812,
      "learning_rate": 0.2854519689438068,
      "loss": 0.2811,
      "num_input_tokens_seen": 4314272,
      "step": 5655
    },
    {
      "epoch": 11.767151767151766,
      "grad_norm": 0.00024221248168032616,
      "learning_rate": 0.2854266522081412,
      "loss": 0.2832,
      "num_input_tokens_seen": 4318016,
      "step": 5660
    },
    {
      "epoch": 11.777546777546778,
      "grad_norm": 0.0004617544182110578,
      "learning_rate": 0.28540131458798385,
      "loss": 0.2517,
      "num_input_tokens_seen": 4321760,
      "step": 5665
    },
    {
      "epoch": 11.787941787941788,
      "grad_norm": 0.0011133091757073998,
      "learning_rate": 0.28537595608724226,
      "loss": 0.2815,
      "num_input_tokens_seen": 4325696,
      "step": 5670
    },
    {
      "epoch": 11.798336798336798,
      "grad_norm": 0.00035040912916883826,
      "learning_rate": 0.28535057670982705,
      "loss": 0.2786,
      "num_input_tokens_seen": 4329376,
      "step": 5675
    },
    {
      "epoch": 11.808731808731808,
      "grad_norm": 0.0004917478654533625,
      "learning_rate": 0.285325176459652,
      "loss": 0.2829,
      "num_input_tokens_seen": 4333184,
      "step": 5680
    },
    {
      "epoch": 11.819126819126819,
      "grad_norm": 0.0002756681642495096,
      "learning_rate": 0.28529975534063406,
      "loss": 0.275,
      "num_input_tokens_seen": 4337120,
      "step": 5685
    },
    {
      "epoch": 11.829521829521829,
      "grad_norm": 0.0005553779192268848,
      "learning_rate": 0.2852743133566936,
      "loss": 0.2245,
      "num_input_tokens_seen": 4340832,
      "step": 5690
    },
    {
      "epoch": 11.83991683991684,
      "grad_norm": 0.001064133713953197,
      "learning_rate": 0.2852488505117541,
      "loss": 0.3045,
      "num_input_tokens_seen": 4344672,
      "step": 5695
    },
    {
      "epoch": 11.85031185031185,
      "grad_norm": 0.0018903706222772598,
      "learning_rate": 0.28522336680974214,
      "loss": 0.2704,
      "num_input_tokens_seen": 4348416,
      "step": 5700
    },
    {
      "epoch": 11.86070686070686,
      "grad_norm": 0.00017086938896682113,
      "learning_rate": 0.2851978622545877,
      "loss": 0.2704,
      "num_input_tokens_seen": 4352192,
      "step": 5705
    },
    {
      "epoch": 11.871101871101871,
      "grad_norm": 0.001326584373600781,
      "learning_rate": 0.285172336850224,
      "loss": 0.2706,
      "num_input_tokens_seen": 4356032,
      "step": 5710
    },
    {
      "epoch": 11.881496881496881,
      "grad_norm": 0.0006213866290636361,
      "learning_rate": 0.2851467906005871,
      "loss": 0.2734,
      "num_input_tokens_seen": 4359872,
      "step": 5715
    },
    {
      "epoch": 11.891891891891891,
      "grad_norm": 0.00016239417891483754,
      "learning_rate": 0.28512122350961683,
      "loss": 0.2764,
      "num_input_tokens_seen": 4363712,
      "step": 5720
    },
    {
      "epoch": 11.902286902286903,
      "grad_norm": 0.0001600651303306222,
      "learning_rate": 0.2850956355812559,
      "loss": 0.2657,
      "num_input_tokens_seen": 4367552,
      "step": 5725
    },
    {
      "epoch": 11.912681912681913,
      "grad_norm": 0.00046010466758161783,
      "learning_rate": 0.28507002681945015,
      "loss": 0.2651,
      "num_input_tokens_seen": 4371264,
      "step": 5730
    },
    {
      "epoch": 11.923076923076923,
      "grad_norm": 0.0014085611328482628,
      "learning_rate": 0.28504439722814895,
      "loss": 0.2759,
      "num_input_tokens_seen": 4375008,
      "step": 5735
    },
    {
      "epoch": 11.933471933471933,
      "grad_norm": 0.00039940845454111695,
      "learning_rate": 0.28501874681130457,
      "loss": 0.2765,
      "num_input_tokens_seen": 4378976,
      "step": 5740
    },
    {
      "epoch": 11.943866943866944,
      "grad_norm": 0.00048410348244942725,
      "learning_rate": 0.2849930755728727,
      "loss": 0.2868,
      "num_input_tokens_seen": 4382624,
      "step": 5745
    },
    {
      "epoch": 11.954261954261954,
      "grad_norm": 0.0006358897080644965,
      "learning_rate": 0.28496738351681217,
      "loss": 0.2685,
      "num_input_tokens_seen": 4386624,
      "step": 5750
    },
    {
      "epoch": 11.964656964656964,
      "grad_norm": 0.00012058711581630632,
      "learning_rate": 0.284941670647085,
      "loss": 0.2684,
      "num_input_tokens_seen": 4390368,
      "step": 5755
    },
    {
      "epoch": 11.975051975051976,
      "grad_norm": 0.00016567543207202107,
      "learning_rate": 0.2849159369676563,
      "loss": 0.2604,
      "num_input_tokens_seen": 4394048,
      "step": 5760
    },
    {
      "epoch": 11.985446985446986,
      "grad_norm": 0.0013829134404659271,
      "learning_rate": 0.2848901824824948,
      "loss": 0.2799,
      "num_input_tokens_seen": 4397824,
      "step": 5765
    },
    {
      "epoch": 11.995841995841996,
      "grad_norm": 0.0005715001025237143,
      "learning_rate": 0.284864407195572,
      "loss": 0.2706,
      "num_input_tokens_seen": 4401696,
      "step": 5770
    },
    {
      "epoch": 12.006237006237006,
      "grad_norm": 0.0015061214799061418,
      "learning_rate": 0.28483861111086284,
      "loss": 0.2637,
      "num_input_tokens_seen": 4405368,
      "step": 5775
    },
    {
      "epoch": 12.016632016632016,
      "grad_norm": 0.0008287896635010839,
      "learning_rate": 0.2848127942323453,
      "loss": 0.2837,
      "num_input_tokens_seen": 4409304,
      "step": 5780
    },
    {
      "epoch": 12.027027027027026,
      "grad_norm": 0.0006331239128485322,
      "learning_rate": 0.2847869565640007,
      "loss": 0.2772,
      "num_input_tokens_seen": 4413016,
      "step": 5785
    },
    {
      "epoch": 12.037422037422038,
      "grad_norm": 0.0012379608815535903,
      "learning_rate": 0.2847610981098136,
      "loss": 0.2783,
      "num_input_tokens_seen": 4416696,
      "step": 5790
    },
    {
      "epoch": 12.047817047817048,
      "grad_norm": 0.0012926555937156081,
      "learning_rate": 0.2847352188737716,
      "loss": 0.2823,
      "num_input_tokens_seen": 4420440,
      "step": 5795
    },
    {
      "epoch": 12.058212058212058,
      "grad_norm": 0.0013617328368127346,
      "learning_rate": 0.2847093188598658,
      "loss": 0.2883,
      "num_input_tokens_seen": 4424280,
      "step": 5800
    },
    {
      "epoch": 12.058212058212058,
      "eval_loss": 0.26770585775375366,
      "eval_runtime": 13.4056,
      "eval_samples_per_second": 63.854,
      "eval_steps_per_second": 15.963,
      "num_input_tokens_seen": 4424280,
      "step": 5800
    },
    {
      "epoch": 12.068607068607069,
      "grad_norm": 0.00018180614279117435,
      "learning_rate": 0.28468339807209003,
      "loss": 0.2756,
      "num_input_tokens_seen": 4428088,
      "step": 5805
    },
    {
      "epoch": 12.079002079002079,
      "grad_norm": 0.00022024096688255668,
      "learning_rate": 0.2846574565144418,
      "loss": 0.2603,
      "num_input_tokens_seen": 4431928,
      "step": 5810
    },
    {
      "epoch": 12.089397089397089,
      "grad_norm": 9.740667883306742e-05,
      "learning_rate": 0.28463149419092154,
      "loss": 0.2587,
      "num_input_tokens_seen": 4435864,
      "step": 5815
    },
    {
      "epoch": 12.0997920997921,
      "grad_norm": 0.0004528888675849885,
      "learning_rate": 0.284605511105533,
      "loss": 0.265,
      "num_input_tokens_seen": 4439544,
      "step": 5820
    },
    {
      "epoch": 12.11018711018711,
      "grad_norm": 0.0007771481177769601,
      "learning_rate": 0.28457950726228315,
      "loss": 0.2695,
      "num_input_tokens_seen": 4443224,
      "step": 5825
    },
    {
      "epoch": 12.120582120582121,
      "grad_norm": 0.00012699983199127018,
      "learning_rate": 0.28455348266518193,
      "loss": 0.278,
      "num_input_tokens_seen": 4447000,
      "step": 5830
    },
    {
      "epoch": 12.130977130977131,
      "grad_norm": 0.0001094221806852147,
      "learning_rate": 0.28452743731824287,
      "loss": 0.2614,
      "num_input_tokens_seen": 4450744,
      "step": 5835
    },
    {
      "epoch": 12.141372141372141,
      "grad_norm": 0.0009366910089738667,
      "learning_rate": 0.28450137122548236,
      "loss": 0.2588,
      "num_input_tokens_seen": 4454456,
      "step": 5840
    },
    {
      "epoch": 12.151767151767151,
      "grad_norm": 0.0002853893965948373,
      "learning_rate": 0.2844752843909201,
      "loss": 0.2881,
      "num_input_tokens_seen": 4458200,
      "step": 5845
    },
    {
      "epoch": 12.162162162162161,
      "grad_norm": 0.00043805918539874256,
      "learning_rate": 0.28444917681857923,
      "loss": 0.2785,
      "num_input_tokens_seen": 4461912,
      "step": 5850
    },
    {
      "epoch": 12.172557172557173,
      "grad_norm": 0.0012598710600286722,
      "learning_rate": 0.28442304851248557,
      "loss": 0.288,
      "num_input_tokens_seen": 4465720,
      "step": 5855
    },
    {
      "epoch": 12.182952182952183,
      "grad_norm": 0.0009912147652357817,
      "learning_rate": 0.2843968994766686,
      "loss": 0.2593,
      "num_input_tokens_seen": 4469624,
      "step": 5860
    },
    {
      "epoch": 12.193347193347194,
      "grad_norm": 0.000685918377712369,
      "learning_rate": 0.28437072971516075,
      "loss": 0.2805,
      "num_input_tokens_seen": 4473464,
      "step": 5865
    },
    {
      "epoch": 12.203742203742204,
      "grad_norm": 0.001031465595588088,
      "learning_rate": 0.2843445392319979,
      "loss": 0.2691,
      "num_input_tokens_seen": 4477240,
      "step": 5870
    },
    {
      "epoch": 12.214137214137214,
      "grad_norm": 0.00011284740321571007,
      "learning_rate": 0.28431832803121865,
      "loss": 0.2729,
      "num_input_tokens_seen": 4480952,
      "step": 5875
    },
    {
      "epoch": 12.224532224532224,
      "grad_norm": 0.0010026207892224193,
      "learning_rate": 0.28429209611686534,
      "loss": 0.268,
      "num_input_tokens_seen": 4484824,
      "step": 5880
    },
    {
      "epoch": 12.234927234927236,
      "grad_norm": 0.00038795857108198106,
      "learning_rate": 0.28426584349298323,
      "loss": 0.2536,
      "num_input_tokens_seen": 4488536,
      "step": 5885
    },
    {
      "epoch": 12.245322245322246,
      "grad_norm": 0.0004827986122108996,
      "learning_rate": 0.2842395701636207,
      "loss": 0.2659,
      "num_input_tokens_seen": 4492472,
      "step": 5890
    },
    {
      "epoch": 12.255717255717256,
      "grad_norm": 0.0006701254169456661,
      "learning_rate": 0.28421327613282954,
      "loss": 0.2766,
      "num_input_tokens_seen": 4496056,
      "step": 5895
    },
    {
      "epoch": 12.266112266112266,
      "grad_norm": 0.000321400526445359,
      "learning_rate": 0.28418696140466454,
      "loss": 0.286,
      "num_input_tokens_seen": 4499864,
      "step": 5900
    },
    {
      "epoch": 12.276507276507276,
      "grad_norm": 0.0003939138550776988,
      "learning_rate": 0.2841606259831838,
      "loss": 0.2709,
      "num_input_tokens_seen": 4503512,
      "step": 5905
    },
    {
      "epoch": 12.286902286902286,
      "grad_norm": 0.001314492430537939,
      "learning_rate": 0.2841342698724486,
      "loss": 0.2688,
      "num_input_tokens_seen": 4507224,
      "step": 5910
    },
    {
      "epoch": 12.297297297297296,
      "grad_norm": 0.0020476027857512236,
      "learning_rate": 0.28410789307652334,
      "loss": 0.2821,
      "num_input_tokens_seen": 4511128,
      "step": 5915
    },
    {
      "epoch": 12.307692307692308,
      "grad_norm": 0.0008405789267271757,
      "learning_rate": 0.2840814955994756,
      "loss": 0.2699,
      "num_input_tokens_seen": 4514840,
      "step": 5920
    },
    {
      "epoch": 12.318087318087318,
      "grad_norm": 0.00016065254749264568,
      "learning_rate": 0.2840550774453763,
      "loss": 0.2848,
      "num_input_tokens_seen": 4518648,
      "step": 5925
    },
    {
      "epoch": 12.328482328482329,
      "grad_norm": 0.0002067928871838376,
      "learning_rate": 0.28402863861829947,
      "loss": 0.2772,
      "num_input_tokens_seen": 4522392,
      "step": 5930
    },
    {
      "epoch": 12.338877338877339,
      "grad_norm": 0.0013550748117268085,
      "learning_rate": 0.2840021791223222,
      "loss": 0.2733,
      "num_input_tokens_seen": 4526200,
      "step": 5935
    },
    {
      "epoch": 12.349272349272349,
      "grad_norm": 0.0005376901826821268,
      "learning_rate": 0.2839756989615249,
      "loss": 0.273,
      "num_input_tokens_seen": 4529912,
      "step": 5940
    },
    {
      "epoch": 12.359667359667359,
      "grad_norm": 0.0006089470698498189,
      "learning_rate": 0.28394919813999125,
      "loss": 0.2502,
      "num_input_tokens_seen": 4533720,
      "step": 5945
    },
    {
      "epoch": 12.37006237006237,
      "grad_norm": 0.0020214782562106848,
      "learning_rate": 0.28392267666180787,
      "loss": 0.2825,
      "num_input_tokens_seen": 4537528,
      "step": 5950
    },
    {
      "epoch": 12.380457380457381,
      "grad_norm": 0.000504746160004288,
      "learning_rate": 0.2838961345310648,
      "loss": 0.2656,
      "num_input_tokens_seen": 4541400,
      "step": 5955
    },
    {
      "epoch": 12.390852390852391,
      "grad_norm": 0.0003566770174074918,
      "learning_rate": 0.2838695717518552,
      "loss": 0.2994,
      "num_input_tokens_seen": 4545336,
      "step": 5960
    },
    {
      "epoch": 12.401247401247401,
      "grad_norm": 0.00028244126588106155,
      "learning_rate": 0.28384298832827526,
      "loss": 0.2663,
      "num_input_tokens_seen": 4549240,
      "step": 5965
    },
    {
      "epoch": 12.411642411642411,
      "grad_norm": 0.00011501672270242125,
      "learning_rate": 0.28381638426442457,
      "loss": 0.2678,
      "num_input_tokens_seen": 4552984,
      "step": 5970
    },
    {
      "epoch": 12.422037422037421,
      "grad_norm": 0.0004110147128812969,
      "learning_rate": 0.2837897595644057,
      "loss": 0.2756,
      "num_input_tokens_seen": 4556632,
      "step": 5975
    },
    {
      "epoch": 12.432432432432432,
      "grad_norm": 0.0007743166061118245,
      "learning_rate": 0.28376311423232475,
      "loss": 0.2669,
      "num_input_tokens_seen": 4560376,
      "step": 5980
    },
    {
      "epoch": 12.442827442827443,
      "grad_norm": 0.00017317116726189852,
      "learning_rate": 0.2837364482722905,
      "loss": 0.272,
      "num_input_tokens_seen": 4564120,
      "step": 5985
    },
    {
      "epoch": 12.453222453222454,
      "grad_norm": 0.0007913258741609752,
      "learning_rate": 0.28370976168841533,
      "loss": 0.2824,
      "num_input_tokens_seen": 4567864,
      "step": 5990
    },
    {
      "epoch": 12.463617463617464,
      "grad_norm": 0.0009628135594539344,
      "learning_rate": 0.2836830544848146,
      "loss": 0.2629,
      "num_input_tokens_seen": 4571576,
      "step": 5995
    },
    {
      "epoch": 12.474012474012474,
      "grad_norm": 7.43231939850375e-05,
      "learning_rate": 0.2836563266656069,
      "loss": 0.271,
      "num_input_tokens_seen": 4575480,
      "step": 6000
    },
    {
      "epoch": 12.474012474012474,
      "eval_loss": 0.24917590618133545,
      "eval_runtime": 13.3975,
      "eval_samples_per_second": 63.892,
      "eval_steps_per_second": 15.973,
      "num_input_tokens_seen": 4575480,
      "step": 6000
    },
    {
      "epoch": 12.484407484407484,
      "grad_norm": 9.255088662030175e-05,
      "learning_rate": 0.283629578234914,
      "loss": 0.2529,
      "num_input_tokens_seen": 4579224,
      "step": 6005
    },
    {
      "epoch": 12.494802494802494,
      "grad_norm": 0.0004901768988929689,
      "learning_rate": 0.2836028091968608,
      "loss": 0.2905,
      "num_input_tokens_seen": 4583192,
      "step": 6010
    },
    {
      "epoch": 12.505197505197506,
      "grad_norm": 0.0010116827907040715,
      "learning_rate": 0.28357601955557554,
      "loss": 0.2489,
      "num_input_tokens_seen": 4586872,
      "step": 6015
    },
    {
      "epoch": 12.515592515592516,
      "grad_norm": 0.0005950178601779044,
      "learning_rate": 0.2835492093151894,
      "loss": 0.3089,
      "num_input_tokens_seen": 4590712,
      "step": 6020
    },
    {
      "epoch": 12.525987525987526,
      "grad_norm": 0.0007853138376958668,
      "learning_rate": 0.2835223784798369,
      "loss": 0.282,
      "num_input_tokens_seen": 4594488,
      "step": 6025
    },
    {
      "epoch": 12.536382536382536,
      "grad_norm": 0.0018702605739235878,
      "learning_rate": 0.2834955270536557,
      "loss": 0.2796,
      "num_input_tokens_seen": 4598264,
      "step": 6030
    },
    {
      "epoch": 12.546777546777546,
      "grad_norm": 0.00075818708864972,
      "learning_rate": 0.2834686550407866,
      "loss": 0.2789,
      "num_input_tokens_seen": 4602104,
      "step": 6035
    },
    {
      "epoch": 12.557172557172557,
      "grad_norm": 0.00012929151125717908,
      "learning_rate": 0.28344176244537367,
      "loss": 0.2611,
      "num_input_tokens_seen": 4605976,
      "step": 6040
    },
    {
      "epoch": 12.567567567567568,
      "grad_norm": 0.0014215228147804737,
      "learning_rate": 0.28341484927156396,
      "loss": 0.2702,
      "num_input_tokens_seen": 4609848,
      "step": 6045
    },
    {
      "epoch": 12.577962577962579,
      "grad_norm": 0.0009078827570192516,
      "learning_rate": 0.28338791552350795,
      "loss": 0.2554,
      "num_input_tokens_seen": 4613656,
      "step": 6050
    },
    {
      "epoch": 12.588357588357589,
      "grad_norm": 0.0001375266001559794,
      "learning_rate": 0.28336096120535914,
      "loss": 0.2652,
      "num_input_tokens_seen": 4617464,
      "step": 6055
    },
    {
      "epoch": 12.598752598752599,
      "grad_norm": 0.00038482018862850964,
      "learning_rate": 0.2833339863212741,
      "loss": 0.2433,
      "num_input_tokens_seen": 4621336,
      "step": 6060
    },
    {
      "epoch": 12.609147609147609,
      "grad_norm": 0.0005166112096048892,
      "learning_rate": 0.28330699087541283,
      "loss": 0.2808,
      "num_input_tokens_seen": 4625208,
      "step": 6065
    },
    {
      "epoch": 12.619542619542619,
      "grad_norm": 0.0001253746886504814,
      "learning_rate": 0.2832799748719384,
      "loss": 0.269,
      "num_input_tokens_seen": 4629208,
      "step": 6070
    },
    {
      "epoch": 12.62993762993763,
      "grad_norm": 0.00010340005246689543,
      "learning_rate": 0.28325293831501686,
      "loss": 0.2782,
      "num_input_tokens_seen": 4632824,
      "step": 6075
    },
    {
      "epoch": 12.640332640332641,
      "grad_norm": 0.0002453316992614418,
      "learning_rate": 0.2832258812088177,
      "loss": 0.2574,
      "num_input_tokens_seen": 4636632,
      "step": 6080
    },
    {
      "epoch": 12.650727650727651,
      "grad_norm": 0.0009290928137488663,
      "learning_rate": 0.2831988035575134,
      "loss": 0.2524,
      "num_input_tokens_seen": 4640440,
      "step": 6085
    },
    {
      "epoch": 12.661122661122661,
      "grad_norm": 0.0016221727710217237,
      "learning_rate": 0.28317170536527975,
      "loss": 0.2982,
      "num_input_tokens_seen": 4644408,
      "step": 6090
    },
    {
      "epoch": 12.671517671517671,
      "grad_norm": 0.0012651460710912943,
      "learning_rate": 0.2831445866362956,
      "loss": 0.2707,
      "num_input_tokens_seen": 4648248,
      "step": 6095
    },
    {
      "epoch": 12.681912681912682,
      "grad_norm": 0.0007314207032322884,
      "learning_rate": 0.2831174473747429,
      "loss": 0.2838,
      "num_input_tokens_seen": 4651992,
      "step": 6100
    },
    {
      "epoch": 12.692307692307692,
      "grad_norm": 0.001460422296077013,
      "learning_rate": 0.2830902875848071,
      "loss": 0.2603,
      "num_input_tokens_seen": 4655576,
      "step": 6105
    },
    {
      "epoch": 12.702702702702704,
      "grad_norm": 0.00042679597390815616,
      "learning_rate": 0.28306310727067635,
      "loss": 0.2771,
      "num_input_tokens_seen": 4659416,
      "step": 6110
    },
    {
      "epoch": 12.713097713097714,
      "grad_norm": 0.000388226326322183,
      "learning_rate": 0.2830359064365423,
      "loss": 0.2669,
      "num_input_tokens_seen": 4663288,
      "step": 6115
    },
    {
      "epoch": 12.723492723492724,
      "grad_norm": 0.0001736678823363036,
      "learning_rate": 0.28300868508659965,
      "loss": 0.2785,
      "num_input_tokens_seen": 4667096,
      "step": 6120
    },
    {
      "epoch": 12.733887733887734,
      "grad_norm": 0.00040306313894689083,
      "learning_rate": 0.28298144322504626,
      "loss": 0.2582,
      "num_input_tokens_seen": 4670840,
      "step": 6125
    },
    {
      "epoch": 12.744282744282744,
      "grad_norm": 0.0004889412084594369,
      "learning_rate": 0.2829541808560832,
      "loss": 0.2695,
      "num_input_tokens_seen": 4674552,
      "step": 6130
    },
    {
      "epoch": 12.754677754677754,
      "grad_norm": 0.00038278326974250376,
      "learning_rate": 0.2829268979839146,
      "loss": 0.2436,
      "num_input_tokens_seen": 4678456,
      "step": 6135
    },
    {
      "epoch": 12.765072765072766,
      "grad_norm": 0.00026976349181495607,
      "learning_rate": 0.2828995946127479,
      "loss": 0.2698,
      "num_input_tokens_seen": 4682232,
      "step": 6140
    },
    {
      "epoch": 12.775467775467776,
      "grad_norm": 0.0020005572587251663,
      "learning_rate": 0.2828722707467936,
      "loss": 0.2864,
      "num_input_tokens_seen": 4686072,
      "step": 6145
    },
    {
      "epoch": 12.785862785862786,
      "grad_norm": 0.00045545899774879217,
      "learning_rate": 0.2828449263902653,
      "loss": 0.2595,
      "num_input_tokens_seen": 4689912,
      "step": 6150
    },
    {
      "epoch": 12.796257796257796,
      "grad_norm": 0.00031507579842582345,
      "learning_rate": 0.28281756154738,
      "loss": 0.3103,
      "num_input_tokens_seen": 4693688,
      "step": 6155
    },
    {
      "epoch": 12.806652806652806,
      "grad_norm": 0.0013923244550824165,
      "learning_rate": 0.28279017622235764,
      "loss": 0.2664,
      "num_input_tokens_seen": 4697496,
      "step": 6160
    },
    {
      "epoch": 12.817047817047817,
      "grad_norm": 0.0006714232731610537,
      "learning_rate": 0.28276277041942127,
      "loss": 0.2586,
      "num_input_tokens_seen": 4701368,
      "step": 6165
    },
    {
      "epoch": 12.827442827442827,
      "grad_norm": 0.0016950289718806744,
      "learning_rate": 0.2827353441427974,
      "loss": 0.2783,
      "num_input_tokens_seen": 4705304,
      "step": 6170
    },
    {
      "epoch": 12.837837837837839,
      "grad_norm": 0.0001451850839657709,
      "learning_rate": 0.2827078973967153,
      "loss": 0.2685,
      "num_input_tokens_seen": 4709240,
      "step": 6175
    },
    {
      "epoch": 12.848232848232849,
      "grad_norm": 0.00039057733374647796,
      "learning_rate": 0.2826804301854078,
      "loss": 0.2766,
      "num_input_tokens_seen": 4713080,
      "step": 6180
    },
    {
      "epoch": 12.858627858627859,
      "grad_norm": 0.0007726919720880687,
      "learning_rate": 0.2826529425131105,
      "loss": 0.2747,
      "num_input_tokens_seen": 4717112,
      "step": 6185
    },
    {
      "epoch": 12.869022869022869,
      "grad_norm": 0.0006966284126974642,
      "learning_rate": 0.2826254343840625,
      "loss": 0.2658,
      "num_input_tokens_seen": 4721112,
      "step": 6190
    },
    {
      "epoch": 12.879417879417879,
      "grad_norm": 0.00011544146400410682,
      "learning_rate": 0.2825979058025059,
      "loss": 0.2738,
      "num_input_tokens_seen": 4724856,
      "step": 6195
    },
    {
      "epoch": 12.88981288981289,
      "grad_norm": 0.0003345098521094769,
      "learning_rate": 0.2825703567726858,
      "loss": 0.2644,
      "num_input_tokens_seen": 4728792,
      "step": 6200
    },
    {
      "epoch": 12.88981288981289,
      "eval_loss": 0.24961183965206146,
      "eval_runtime": 13.4495,
      "eval_samples_per_second": 63.645,
      "eval_steps_per_second": 15.911,
      "num_input_tokens_seen": 4728792,
      "step": 6200
    },
    {
      "epoch": 12.9002079002079,
      "grad_norm": 0.00044291187077760696,
      "learning_rate": 0.2825427872988508,
      "loss": 0.2614,
      "num_input_tokens_seen": 4732600,
      "step": 6205
    },
    {
      "epoch": 12.910602910602911,
      "grad_norm": 0.0003708438016474247,
      "learning_rate": 0.28251519738525227,
      "loss": 0.2874,
      "num_input_tokens_seen": 4736472,
      "step": 6210
    },
    {
      "epoch": 12.920997920997921,
      "grad_norm": 0.00046882391325198114,
      "learning_rate": 0.28248758703614507,
      "loss": 0.2839,
      "num_input_tokens_seen": 4740248,
      "step": 6215
    },
    {
      "epoch": 12.931392931392931,
      "grad_norm": 0.00145276531111449,
      "learning_rate": 0.28245995625578696,
      "loss": 0.2751,
      "num_input_tokens_seen": 4744216,
      "step": 6220
    },
    {
      "epoch": 12.941787941787942,
      "grad_norm": 0.00027773386682383716,
      "learning_rate": 0.282432305048439,
      "loss": 0.2785,
      "num_input_tokens_seen": 4747960,
      "step": 6225
    },
    {
      "epoch": 12.952182952182952,
      "grad_norm": 0.0012273472966626287,
      "learning_rate": 0.28240463341836536,
      "loss": 0.2698,
      "num_input_tokens_seen": 4751704,
      "step": 6230
    },
    {
      "epoch": 12.962577962577962,
      "grad_norm": 0.0018733071628957987,
      "learning_rate": 0.2823769413698334,
      "loss": 0.2862,
      "num_input_tokens_seen": 4755480,
      "step": 6235
    },
    {
      "epoch": 12.972972972972974,
      "grad_norm": 0.0008979453705251217,
      "learning_rate": 0.2823492289071135,
      "loss": 0.2502,
      "num_input_tokens_seen": 4759128,
      "step": 6240
    },
    {
      "epoch": 12.983367983367984,
      "grad_norm": 0.00033327305573038757,
      "learning_rate": 0.2823214960344793,
      "loss": 0.2532,
      "num_input_tokens_seen": 4763160,
      "step": 6245
    },
    {
      "epoch": 12.993762993762994,
      "grad_norm": 9.502907778369263e-05,
      "learning_rate": 0.28229374275620756,
      "loss": 0.2426,
      "num_input_tokens_seen": 4767096,
      "step": 6250
    },
    {
      "epoch": 13.004158004158004,
      "grad_norm": 0.000791467959061265,
      "learning_rate": 0.28226596907657814,
      "loss": 0.275,
      "num_input_tokens_seen": 4770896,
      "step": 6255
    },
    {
      "epoch": 13.014553014553014,
      "grad_norm": 0.0010478816693648696,
      "learning_rate": 0.28223817499987414,
      "loss": 0.2809,
      "num_input_tokens_seen": 4774704,
      "step": 6260
    },
    {
      "epoch": 13.024948024948024,
      "grad_norm": 0.0010234445799142122,
      "learning_rate": 0.2822103605303818,
      "loss": 0.2708,
      "num_input_tokens_seen": 4778416,
      "step": 6265
    },
    {
      "epoch": 13.035343035343036,
      "grad_norm": 0.00024356390349566936,
      "learning_rate": 0.2821825256723903,
      "loss": 0.2847,
      "num_input_tokens_seen": 4782192,
      "step": 6270
    },
    {
      "epoch": 13.045738045738046,
      "grad_norm": 0.00011574855307117105,
      "learning_rate": 0.2821546704301923,
      "loss": 0.2983,
      "num_input_tokens_seen": 4786032,
      "step": 6275
    },
    {
      "epoch": 13.056133056133056,
      "grad_norm": 0.0010286258766427636,
      "learning_rate": 0.2821267948080834,
      "loss": 0.2702,
      "num_input_tokens_seen": 4789776,
      "step": 6280
    },
    {
      "epoch": 13.066528066528067,
      "grad_norm": 0.0003775153891183436,
      "learning_rate": 0.28209889881036226,
      "loss": 0.236,
      "num_input_tokens_seen": 4793648,
      "step": 6285
    },
    {
      "epoch": 13.076923076923077,
      "grad_norm": 0.0012265084078535438,
      "learning_rate": 0.28207098244133094,
      "loss": 0.3034,
      "num_input_tokens_seen": 4797648,
      "step": 6290
    },
    {
      "epoch": 13.087318087318087,
      "grad_norm": 0.0017759909387677908,
      "learning_rate": 0.2820430457052943,
      "loss": 0.2806,
      "num_input_tokens_seen": 4801456,
      "step": 6295
    },
    {
      "epoch": 13.097713097713097,
      "grad_norm": 0.0004505858232732862,
      "learning_rate": 0.28201508860656077,
      "loss": 0.2686,
      "num_input_tokens_seen": 4805264,
      "step": 6300
    },
    {
      "epoch": 13.108108108108109,
      "grad_norm": 0.0008845975971780717,
      "learning_rate": 0.2819871111494415,
      "loss": 0.3032,
      "num_input_tokens_seen": 4809008,
      "step": 6305
    },
    {
      "epoch": 13.118503118503119,
      "grad_norm": 0.001012252876535058,
      "learning_rate": 0.28195911333825113,
      "loss": 0.2753,
      "num_input_tokens_seen": 4812848,
      "step": 6310
    },
    {
      "epoch": 13.128898128898129,
      "grad_norm": 0.00036685154191218317,
      "learning_rate": 0.28193109517730713,
      "loss": 0.2701,
      "num_input_tokens_seen": 4816752,
      "step": 6315
    },
    {
      "epoch": 13.13929313929314,
      "grad_norm": 0.0002599575382191688,
      "learning_rate": 0.2819030566709303,
      "loss": 0.2786,
      "num_input_tokens_seen": 4820432,
      "step": 6320
    },
    {
      "epoch": 13.14968814968815,
      "grad_norm": 0.0006284094415605068,
      "learning_rate": 0.2818749978234445,
      "loss": 0.2745,
      "num_input_tokens_seen": 4824272,
      "step": 6325
    },
    {
      "epoch": 13.16008316008316,
      "grad_norm": 0.000569976749829948,
      "learning_rate": 0.2818469186391768,
      "loss": 0.2682,
      "num_input_tokens_seen": 4827984,
      "step": 6330
    },
    {
      "epoch": 13.170478170478171,
      "grad_norm": 0.0008672393742017448,
      "learning_rate": 0.28181881912245743,
      "loss": 0.2503,
      "num_input_tokens_seen": 4831664,
      "step": 6335
    },
    {
      "epoch": 13.180873180873181,
      "grad_norm": 0.0005721260677091777,
      "learning_rate": 0.2817906992776195,
      "loss": 0.2696,
      "num_input_tokens_seen": 4835280,
      "step": 6340
    },
    {
      "epoch": 13.191268191268192,
      "grad_norm": 0.0009665184188634157,
      "learning_rate": 0.28176255910899967,
      "loss": 0.2821,
      "num_input_tokens_seen": 4839088,
      "step": 6345
    },
    {
      "epoch": 13.201663201663202,
      "grad_norm": 0.0006864466122351587,
      "learning_rate": 0.2817343986209373,
      "loss": 0.2764,
      "num_input_tokens_seen": 4842960,
      "step": 6350
    },
    {
      "epoch": 13.212058212058212,
      "grad_norm": 0.0003391333157196641,
      "learning_rate": 0.2817062178177753,
      "loss": 0.2582,
      "num_input_tokens_seen": 4846800,
      "step": 6355
    },
    {
      "epoch": 13.222453222453222,
      "grad_norm": 0.0008498854003846645,
      "learning_rate": 0.2816780167038593,
      "loss": 0.3029,
      "num_input_tokens_seen": 4850640,
      "step": 6360
    },
    {
      "epoch": 13.232848232848234,
      "grad_norm": 0.001787275425158441,
      "learning_rate": 0.28164979528353834,
      "loss": 0.2902,
      "num_input_tokens_seen": 4854352,
      "step": 6365
    },
    {
      "epoch": 13.243243243243244,
      "grad_norm": 0.0008399019716307521,
      "learning_rate": 0.28162155356116453,
      "loss": 0.2688,
      "num_input_tokens_seen": 4858096,
      "step": 6370
    },
    {
      "epoch": 13.253638253638254,
      "grad_norm": 0.0014857238857075572,
      "learning_rate": 0.28159329154109314,
      "loss": 0.2366,
      "num_input_tokens_seen": 4861968,
      "step": 6375
    },
    {
      "epoch": 13.264033264033264,
      "grad_norm": 0.0009478324209339917,
      "learning_rate": 0.28156500922768246,
      "loss": 0.3054,
      "num_input_tokens_seen": 4865776,
      "step": 6380
    },
    {
      "epoch": 13.274428274428274,
      "grad_norm": 0.002258904976770282,
      "learning_rate": 0.28153670662529406,
      "loss": 0.2809,
      "num_input_tokens_seen": 4869552,
      "step": 6385
    },
    {
      "epoch": 13.284823284823284,
      "grad_norm": 0.0005595171242021024,
      "learning_rate": 0.28150838373829246,
      "loss": 0.2793,
      "num_input_tokens_seen": 4873488,
      "step": 6390
    },
    {
      "epoch": 13.295218295218294,
      "grad_norm": 0.0008201505988836288,
      "learning_rate": 0.2814800405710455,
      "loss": 0.2712,
      "num_input_tokens_seen": 4877200,
      "step": 6395
    },
    {
      "epoch": 13.305613305613306,
      "grad_norm": 0.000540904002264142,
      "learning_rate": 0.2814516771279239,
      "loss": 0.2439,
      "num_input_tokens_seen": 4880880,
      "step": 6400
    },
    {
      "epoch": 13.305613305613306,
      "eval_loss": 0.2496749460697174,
      "eval_runtime": 13.4111,
      "eval_samples_per_second": 63.828,
      "eval_steps_per_second": 15.957,
      "num_input_tokens_seen": 4880880,
      "step": 6400
    },
    {
      "epoch": 13.316008316008316,
      "grad_norm": 0.00034471412072889507,
      "learning_rate": 0.28142329341330186,
      "loss": 0.2857,
      "num_input_tokens_seen": 4884560,
      "step": 6405
    },
    {
      "epoch": 13.326403326403327,
      "grad_norm": 7.143099355744198e-05,
      "learning_rate": 0.2813948894315564,
      "loss": 0.2722,
      "num_input_tokens_seen": 4888400,
      "step": 6410
    },
    {
      "epoch": 13.336798336798337,
      "grad_norm": 0.0003664206597022712,
      "learning_rate": 0.2813664651870677,
      "loss": 0.29,
      "num_input_tokens_seen": 4892176,
      "step": 6415
    },
    {
      "epoch": 13.347193347193347,
      "grad_norm": 0.0004640058905351907,
      "learning_rate": 0.28133802068421926,
      "loss": 0.2594,
      "num_input_tokens_seen": 4895952,
      "step": 6420
    },
    {
      "epoch": 13.357588357588357,
      "grad_norm": 0.00030228865216486156,
      "learning_rate": 0.28130955592739754,
      "loss": 0.2802,
      "num_input_tokens_seen": 4899856,
      "step": 6425
    },
    {
      "epoch": 13.367983367983369,
      "grad_norm": 0.00037279140087775886,
      "learning_rate": 0.2812810709209922,
      "loss": 0.2667,
      "num_input_tokens_seen": 4903728,
      "step": 6430
    },
    {
      "epoch": 13.378378378378379,
      "grad_norm": 0.0009131809347309172,
      "learning_rate": 0.2812525656693959,
      "loss": 0.2634,
      "num_input_tokens_seen": 4907504,
      "step": 6435
    },
    {
      "epoch": 13.388773388773389,
      "grad_norm": 0.0001961613743333146,
      "learning_rate": 0.28122404017700453,
      "loss": 0.2531,
      "num_input_tokens_seen": 4911280,
      "step": 6440
    },
    {
      "epoch": 13.3991683991684,
      "grad_norm": 0.0010716636897996068,
      "learning_rate": 0.2811954944482171,
      "loss": 0.2886,
      "num_input_tokens_seen": 4915120,
      "step": 6445
    },
    {
      "epoch": 13.40956340956341,
      "grad_norm": 0.00133323366753757,
      "learning_rate": 0.2811669284874358,
      "loss": 0.3022,
      "num_input_tokens_seen": 4918928,
      "step": 6450
    },
    {
      "epoch": 13.41995841995842,
      "grad_norm": 0.003253166563808918,
      "learning_rate": 0.2811383422990657,
      "loss": 0.2374,
      "num_input_tokens_seen": 4922672,
      "step": 6455
    },
    {
      "epoch": 13.43035343035343,
      "grad_norm": 0.0007003267528489232,
      "learning_rate": 0.2811097358875152,
      "loss": 0.2337,
      "num_input_tokens_seen": 4926576,
      "step": 6460
    },
    {
      "epoch": 13.440748440748441,
      "grad_norm": 0.000843209505546838,
      "learning_rate": 0.2810811092571959,
      "loss": 0.327,
      "num_input_tokens_seen": 4930384,
      "step": 6465
    },
    {
      "epoch": 13.451143451143452,
      "grad_norm": 0.0005050098989158869,
      "learning_rate": 0.28105246241252224,
      "loss": 0.2666,
      "num_input_tokens_seen": 4934160,
      "step": 6470
    },
    {
      "epoch": 13.461538461538462,
      "grad_norm": 0.0003771214687731117,
      "learning_rate": 0.28102379535791194,
      "loss": 0.2751,
      "num_input_tokens_seen": 4937936,
      "step": 6475
    },
    {
      "epoch": 13.471933471933472,
      "grad_norm": 0.0005760493804700673,
      "learning_rate": 0.2809951080977859,
      "loss": 0.2444,
      "num_input_tokens_seen": 4941808,
      "step": 6480
    },
    {
      "epoch": 13.482328482328482,
      "grad_norm": 0.0004649146576412022,
      "learning_rate": 0.28096640063656797,
      "loss": 0.3131,
      "num_input_tokens_seen": 4945648,
      "step": 6485
    },
    {
      "epoch": 13.492723492723492,
      "grad_norm": 0.0007316998089663684,
      "learning_rate": 0.2809376729786852,
      "loss": 0.2846,
      "num_input_tokens_seen": 4949424,
      "step": 6490
    },
    {
      "epoch": 13.503118503118504,
      "grad_norm": 0.0005719223408959806,
      "learning_rate": 0.28090892512856785,
      "loss": 0.2812,
      "num_input_tokens_seen": 4953456,
      "step": 6495
    },
    {
      "epoch": 13.513513513513514,
      "grad_norm": 0.0005000099190510809,
      "learning_rate": 0.2808801570906491,
      "loss": 0.2645,
      "num_input_tokens_seen": 4957264,
      "step": 6500
    },
    {
      "epoch": 13.523908523908524,
      "grad_norm": 0.000599811552092433,
      "learning_rate": 0.2808513688693654,
      "loss": 0.2548,
      "num_input_tokens_seen": 4961072,
      "step": 6505
    },
    {
      "epoch": 13.534303534303534,
      "grad_norm": 0.0011316632153466344,
      "learning_rate": 0.28082256046915627,
      "loss": 0.2609,
      "num_input_tokens_seen": 4964944,
      "step": 6510
    },
    {
      "epoch": 13.544698544698544,
      "grad_norm": 0.0016541045624762774,
      "learning_rate": 0.28079373189446427,
      "loss": 0.2714,
      "num_input_tokens_seen": 4968848,
      "step": 6515
    },
    {
      "epoch": 13.555093555093555,
      "grad_norm": 0.0005101578426547348,
      "learning_rate": 0.28076488314973513,
      "loss": 0.2559,
      "num_input_tokens_seen": 4972656,
      "step": 6520
    },
    {
      "epoch": 13.565488565488565,
      "grad_norm": 0.00035254814429208636,
      "learning_rate": 0.28073601423941774,
      "loss": 0.2939,
      "num_input_tokens_seen": 4976368,
      "step": 6525
    },
    {
      "epoch": 13.575883575883577,
      "grad_norm": 0.0011359798954799771,
      "learning_rate": 0.28070712516796403,
      "loss": 0.2729,
      "num_input_tokens_seen": 4980144,
      "step": 6530
    },
    {
      "epoch": 13.586278586278587,
      "grad_norm": 0.0012922968016937375,
      "learning_rate": 0.28067821593982906,
      "loss": 0.2688,
      "num_input_tokens_seen": 4984080,
      "step": 6535
    },
    {
      "epoch": 13.596673596673597,
      "grad_norm": 0.0006646570400334895,
      "learning_rate": 0.28064928655947097,
      "loss": 0.3093,
      "num_input_tokens_seen": 4988016,
      "step": 6540
    },
    {
      "epoch": 13.607068607068607,
      "grad_norm": 0.0009308150038123131,
      "learning_rate": 0.28062033703135103,
      "loss": 0.2895,
      "num_input_tokens_seen": 4991920,
      "step": 6545
    },
    {
      "epoch": 13.617463617463617,
      "grad_norm": 0.00025679919053800404,
      "learning_rate": 0.2805913673599337,
      "loss": 0.2953,
      "num_input_tokens_seen": 4995760,
      "step": 6550
    },
    {
      "epoch": 13.627858627858627,
      "grad_norm": 0.0013019065372645855,
      "learning_rate": 0.2805623775496864,
      "loss": 0.2609,
      "num_input_tokens_seen": 4999632,
      "step": 6555
    },
    {
      "epoch": 13.638253638253639,
      "grad_norm": 0.0001876809837995097,
      "learning_rate": 0.2805333676050797,
      "loss": 0.2864,
      "num_input_tokens_seen": 5003536,
      "step": 6560
    },
    {
      "epoch": 13.64864864864865,
      "grad_norm": 0.000774115149397403,
      "learning_rate": 0.2805043375305873,
      "loss": 0.2753,
      "num_input_tokens_seen": 5007440,
      "step": 6565
    },
    {
      "epoch": 13.65904365904366,
      "grad_norm": 0.0013531945878639817,
      "learning_rate": 0.2804752873306861,
      "loss": 0.2791,
      "num_input_tokens_seen": 5011184,
      "step": 6570
    },
    {
      "epoch": 13.66943866943867,
      "grad_norm": 0.0001015575326164253,
      "learning_rate": 0.2804462170098559,
      "loss": 0.2796,
      "num_input_tokens_seen": 5015120,
      "step": 6575
    },
    {
      "epoch": 13.67983367983368,
      "grad_norm": 7.617779920110479e-05,
      "learning_rate": 0.2804171265725797,
      "loss": 0.2685,
      "num_input_tokens_seen": 5019024,
      "step": 6580
    },
    {
      "epoch": 13.69022869022869,
      "grad_norm": 0.00034285796573385596,
      "learning_rate": 0.28038801602334373,
      "loss": 0.2688,
      "num_input_tokens_seen": 5022960,
      "step": 6585
    },
    {
      "epoch": 13.700623700623701,
      "grad_norm": 0.0007522567757405341,
      "learning_rate": 0.28035888536663717,
      "loss": 0.2592,
      "num_input_tokens_seen": 5026864,
      "step": 6590
    },
    {
      "epoch": 13.711018711018712,
      "grad_norm": 0.0010528706479817629,
      "learning_rate": 0.2803297346069522,
      "loss": 0.2919,
      "num_input_tokens_seen": 5030800,
      "step": 6595
    },
    {
      "epoch": 13.721413721413722,
      "grad_norm": 0.0017097215168178082,
      "learning_rate": 0.28030056374878437,
      "loss": 0.2949,
      "num_input_tokens_seen": 5034608,
      "step": 6600
    },
    {
      "epoch": 13.721413721413722,
      "eval_loss": 0.2552819848060608,
      "eval_runtime": 13.4764,
      "eval_samples_per_second": 63.519,
      "eval_steps_per_second": 15.88,
      "num_input_tokens_seen": 5034608,
      "step": 6600
    },
    {
      "epoch": 13.731808731808732,
      "grad_norm": 0.0009890925139188766,
      "learning_rate": 0.2802713727966321,
      "loss": 0.2629,
      "num_input_tokens_seen": 5038480,
      "step": 6605
    },
    {
      "epoch": 13.742203742203742,
      "grad_norm": 0.0006453191163018346,
      "learning_rate": 0.28024216175499717,
      "loss": 0.2541,
      "num_input_tokens_seen": 5042160,
      "step": 6610
    },
    {
      "epoch": 13.752598752598752,
      "grad_norm": 0.00013424699136521667,
      "learning_rate": 0.2802129306283841,
      "loss": 0.244,
      "num_input_tokens_seen": 5046000,
      "step": 6615
    },
    {
      "epoch": 13.762993762993762,
      "grad_norm": 0.0012249441351741552,
      "learning_rate": 0.28018367942130074,
      "loss": 0.274,
      "num_input_tokens_seen": 5049648,
      "step": 6620
    },
    {
      "epoch": 13.773388773388774,
      "grad_norm": 0.0006099113961681724,
      "learning_rate": 0.28015440813825804,
      "loss": 0.2711,
      "num_input_tokens_seen": 5053456,
      "step": 6625
    },
    {
      "epoch": 13.783783783783784,
      "grad_norm": 0.0006076281424611807,
      "learning_rate": 0.28012511678377006,
      "loss": 0.2697,
      "num_input_tokens_seen": 5057200,
      "step": 6630
    },
    {
      "epoch": 13.794178794178794,
      "grad_norm": 0.0002612782991491258,
      "learning_rate": 0.28009580536235373,
      "loss": 0.268,
      "num_input_tokens_seen": 5060944,
      "step": 6635
    },
    {
      "epoch": 13.804573804573804,
      "grad_norm": 0.0005154876271262765,
      "learning_rate": 0.28006647387852934,
      "loss": 0.2761,
      "num_input_tokens_seen": 5064816,
      "step": 6640
    },
    {
      "epoch": 13.814968814968815,
      "grad_norm": 0.0008332670549862087,
      "learning_rate": 0.28003712233682015,
      "loss": 0.2613,
      "num_input_tokens_seen": 5068496,
      "step": 6645
    },
    {
      "epoch": 13.825363825363825,
      "grad_norm": 0.00019063743820879608,
      "learning_rate": 0.2800077507417526,
      "loss": 0.21,
      "num_input_tokens_seen": 5072240,
      "step": 6650
    },
    {
      "epoch": 13.835758835758837,
      "grad_norm": 0.0001325707562500611,
      "learning_rate": 0.2799783590978561,
      "loss": 0.3029,
      "num_input_tokens_seen": 5076080,
      "step": 6655
    },
    {
      "epoch": 13.846153846153847,
      "grad_norm": 0.001270609674975276,
      "learning_rate": 0.2799489474096632,
      "loss": 0.2918,
      "num_input_tokens_seen": 5080048,
      "step": 6660
    },
    {
      "epoch": 13.856548856548857,
      "grad_norm": 0.00011493675265228376,
      "learning_rate": 0.27991951568170953,
      "loss": 0.2847,
      "num_input_tokens_seen": 5083696,
      "step": 6665
    },
    {
      "epoch": 13.866943866943867,
      "grad_norm": 0.0009804422734305263,
      "learning_rate": 0.2798900639185339,
      "loss": 0.3059,
      "num_input_tokens_seen": 5087408,
      "step": 6670
    },
    {
      "epoch": 13.877338877338877,
      "grad_norm": 0.0011820865329355001,
      "learning_rate": 0.2798605921246781,
      "loss": 0.2821,
      "num_input_tokens_seen": 5091344,
      "step": 6675
    },
    {
      "epoch": 13.887733887733887,
      "grad_norm": 0.00038007143302820623,
      "learning_rate": 0.2798311003046871,
      "loss": 0.2785,
      "num_input_tokens_seen": 5095152,
      "step": 6680
    },
    {
      "epoch": 13.898128898128899,
      "grad_norm": 0.0001244920858880505,
      "learning_rate": 0.2798015884631089,
      "loss": 0.2688,
      "num_input_tokens_seen": 5099120,
      "step": 6685
    },
    {
      "epoch": 13.90852390852391,
      "grad_norm": 0.00029996802913956344,
      "learning_rate": 0.27977205660449445,
      "loss": 0.2646,
      "num_input_tokens_seen": 5103056,
      "step": 6690
    },
    {
      "epoch": 13.91891891891892,
      "grad_norm": 0.00040143224759958684,
      "learning_rate": 0.2797425047333981,
      "loss": 0.2527,
      "num_input_tokens_seen": 5106928,
      "step": 6695
    },
    {
      "epoch": 13.92931392931393,
      "grad_norm": 0.0004995149793103337,
      "learning_rate": 0.27971293285437715,
      "loss": 0.292,
      "num_input_tokens_seen": 5110704,
      "step": 6700
    },
    {
      "epoch": 13.93970893970894,
      "grad_norm": 0.0003274181508459151,
      "learning_rate": 0.2796833409719918,
      "loss": 0.2713,
      "num_input_tokens_seen": 5114640,
      "step": 6705
    },
    {
      "epoch": 13.95010395010395,
      "grad_norm": 0.001920558512210846,
      "learning_rate": 0.27965372909080566,
      "loss": 0.2807,
      "num_input_tokens_seen": 5118320,
      "step": 6710
    },
    {
      "epoch": 13.96049896049896,
      "grad_norm": 0.0006292856414802372,
      "learning_rate": 0.27962409721538506,
      "loss": 0.2717,
      "num_input_tokens_seen": 5122160,
      "step": 6715
    },
    {
      "epoch": 13.970893970893972,
      "grad_norm": 0.0004902294022031128,
      "learning_rate": 0.27959444535029976,
      "loss": 0.2631,
      "num_input_tokens_seen": 5125904,
      "step": 6720
    },
    {
      "epoch": 13.981288981288982,
      "grad_norm": 0.0005971001810394228,
      "learning_rate": 0.27956477350012243,
      "loss": 0.2839,
      "num_input_tokens_seen": 5129680,
      "step": 6725
    },
    {
      "epoch": 13.991683991683992,
      "grad_norm": 0.00025451448163948953,
      "learning_rate": 0.27953508166942875,
      "loss": 0.2829,
      "num_input_tokens_seen": 5133392,
      "step": 6730
    },
    {
      "epoch": 14.002079002079002,
      "grad_norm": 0.00016253946523647755,
      "learning_rate": 0.27950536986279767,
      "loss": 0.3089,
      "num_input_tokens_seen": 5137056,
      "step": 6735
    },
    {
      "epoch": 14.012474012474012,
      "grad_norm": 0.0005154295940883458,
      "learning_rate": 0.2794756380848111,
      "loss": 0.26,
      "num_input_tokens_seen": 5140832,
      "step": 6740
    },
    {
      "epoch": 14.022869022869022,
      "grad_norm": 0.0002365779655519873,
      "learning_rate": 0.279445886340054,
      "loss": 0.2481,
      "num_input_tokens_seen": 5144640,
      "step": 6745
    },
    {
      "epoch": 14.033264033264032,
      "grad_norm": 0.00011351846478646621,
      "learning_rate": 0.27941611463311455,
      "loss": 0.2522,
      "num_input_tokens_seen": 5148480,
      "step": 6750
    },
    {
      "epoch": 14.043659043659044,
      "grad_norm": 0.0008666909416206181,
      "learning_rate": 0.2793863229685839,
      "loss": 0.2844,
      "num_input_tokens_seen": 5152320,
      "step": 6755
    },
    {
      "epoch": 14.054054054054054,
      "grad_norm": 0.0004768530197907239,
      "learning_rate": 0.27935651135105627,
      "loss": 0.26,
      "num_input_tokens_seen": 5156032,
      "step": 6760
    },
    {
      "epoch": 14.064449064449065,
      "grad_norm": 0.0001247638720087707,
      "learning_rate": 0.279326679785129,
      "loss": 0.2442,
      "num_input_tokens_seen": 5159808,
      "step": 6765
    },
    {
      "epoch": 14.074844074844075,
      "grad_norm": 0.0009793314384296536,
      "learning_rate": 0.2792968282754024,
      "loss": 0.3086,
      "num_input_tokens_seen": 5163520,
      "step": 6770
    },
    {
      "epoch": 14.085239085239085,
      "grad_norm": 0.0006179669871926308,
      "learning_rate": 0.2792669568264801,
      "loss": 0.2798,
      "num_input_tokens_seen": 5167200,
      "step": 6775
    },
    {
      "epoch": 14.095634095634095,
      "grad_norm": 0.0003137282619718462,
      "learning_rate": 0.27923706544296856,
      "loss": 0.2547,
      "num_input_tokens_seen": 5170944,
      "step": 6780
    },
    {
      "epoch": 14.106029106029107,
      "grad_norm": 0.0006193777662701905,
      "learning_rate": 0.2792071541294775,
      "loss": 0.2805,
      "num_input_tokens_seen": 5174816,
      "step": 6785
    },
    {
      "epoch": 14.116424116424117,
      "grad_norm": 0.0009216408943757415,
      "learning_rate": 0.27917722289061947,
      "loss": 0.2944,
      "num_input_tokens_seen": 5178784,
      "step": 6790
    },
    {
      "epoch": 14.126819126819127,
      "grad_norm": 0.00031459599267691374,
      "learning_rate": 0.27914727173101034,
      "loss": 0.2726,
      "num_input_tokens_seen": 5182624,
      "step": 6795
    },
    {
      "epoch": 14.137214137214137,
      "grad_norm": 0.00016076047904789448,
      "learning_rate": 0.279117300655269,
      "loss": 0.2525,
      "num_input_tokens_seen": 5186400,
      "step": 6800
    },
    {
      "epoch": 14.137214137214137,
      "eval_loss": 0.2500765025615692,
      "eval_runtime": 13.4269,
      "eval_samples_per_second": 63.753,
      "eval_steps_per_second": 15.938,
      "num_input_tokens_seen": 5186400,
      "step": 6800
    },
    {
      "epoch": 14.147609147609147,
      "grad_norm": 0.0005756820901297033,
      "learning_rate": 0.2790873096680173,
      "loss": 0.2931,
      "num_input_tokens_seen": 5190208,
      "step": 6805
    },
    {
      "epoch": 14.158004158004157,
      "grad_norm": 0.0004946394474245608,
      "learning_rate": 0.2790572987738802,
      "loss": 0.2771,
      "num_input_tokens_seen": 5194176,
      "step": 6810
    },
    {
      "epoch": 14.16839916839917,
      "grad_norm": 0.0003213364689145237,
      "learning_rate": 0.27902726797748584,
      "loss": 0.2653,
      "num_input_tokens_seen": 5197952,
      "step": 6815
    },
    {
      "epoch": 14.17879417879418,
      "grad_norm": 0.000698927731718868,
      "learning_rate": 0.2789972172834652,
      "loss": 0.2301,
      "num_input_tokens_seen": 5201728,
      "step": 6820
    },
    {
      "epoch": 14.18918918918919,
      "grad_norm": 0.0005051936605013907,
      "learning_rate": 0.2789671466964527,
      "loss": 0.289,
      "num_input_tokens_seen": 5205600,
      "step": 6825
    },
    {
      "epoch": 14.1995841995842,
      "grad_norm": 0.00032667024061083794,
      "learning_rate": 0.2789370562210854,
      "loss": 0.2637,
      "num_input_tokens_seen": 5209312,
      "step": 6830
    },
    {
      "epoch": 14.20997920997921,
      "grad_norm": 0.00027768072322942317,
      "learning_rate": 0.27890694586200376,
      "loss": 0.2727,
      "num_input_tokens_seen": 5213184,
      "step": 6835
    },
    {
      "epoch": 14.22037422037422,
      "grad_norm": 0.0005688949604518712,
      "learning_rate": 0.2788768156238511,
      "loss": 0.2481,
      "num_input_tokens_seen": 5217024,
      "step": 6840
    },
    {
      "epoch": 14.23076923076923,
      "grad_norm": 0.00036395780625753105,
      "learning_rate": 0.27884666551127385,
      "loss": 0.2526,
      "num_input_tokens_seen": 5220832,
      "step": 6845
    },
    {
      "epoch": 14.241164241164242,
      "grad_norm": 0.0005436564679257572,
      "learning_rate": 0.2788164955289217,
      "loss": 0.275,
      "num_input_tokens_seen": 5224736,
      "step": 6850
    },
    {
      "epoch": 14.251559251559252,
      "grad_norm": 7.632367487531155e-05,
      "learning_rate": 0.27878630568144697,
      "loss": 0.2688,
      "num_input_tokens_seen": 5228736,
      "step": 6855
    },
    {
      "epoch": 14.261954261954262,
      "grad_norm": 0.00011083899153163657,
      "learning_rate": 0.2787560959735056,
      "loss": 0.2664,
      "num_input_tokens_seen": 5232512,
      "step": 6860
    },
    {
      "epoch": 14.272349272349272,
      "grad_norm": 0.001330072176642716,
      "learning_rate": 0.27872586640975616,
      "loss": 0.2793,
      "num_input_tokens_seen": 5236352,
      "step": 6865
    },
    {
      "epoch": 14.282744282744282,
      "grad_norm": 0.0016209499444812536,
      "learning_rate": 0.27869561699486045,
      "loss": 0.285,
      "num_input_tokens_seen": 5240224,
      "step": 6870
    },
    {
      "epoch": 14.293139293139292,
      "grad_norm": 0.0009718404617160559,
      "learning_rate": 0.2786653477334833,
      "loss": 0.2843,
      "num_input_tokens_seen": 5243968,
      "step": 6875
    },
    {
      "epoch": 14.303534303534304,
      "grad_norm": 0.0006024660542607307,
      "learning_rate": 0.2786350586302926,
      "loss": 0.2396,
      "num_input_tokens_seen": 5247648,
      "step": 6880
    },
    {
      "epoch": 14.313929313929314,
      "grad_norm": 0.0013160171220079064,
      "learning_rate": 0.27860474968995935,
      "loss": 0.2923,
      "num_input_tokens_seen": 5251328,
      "step": 6885
    },
    {
      "epoch": 14.324324324324325,
      "grad_norm": 0.00010290257341694087,
      "learning_rate": 0.27857442091715756,
      "loss": 0.2544,
      "num_input_tokens_seen": 5255008,
      "step": 6890
    },
    {
      "epoch": 14.334719334719335,
      "grad_norm": 0.000662333273794502,
      "learning_rate": 0.27854407231656425,
      "loss": 0.2655,
      "num_input_tokens_seen": 5258720,
      "step": 6895
    },
    {
      "epoch": 14.345114345114345,
      "grad_norm": 0.0007001158664934337,
      "learning_rate": 0.2785137038928596,
      "loss": 0.2789,
      "num_input_tokens_seen": 5262528,
      "step": 6900
    },
    {
      "epoch": 14.355509355509355,
      "grad_norm": 0.0003008940548170358,
      "learning_rate": 0.27848331565072687,
      "loss": 0.2867,
      "num_input_tokens_seen": 5266240,
      "step": 6905
    },
    {
      "epoch": 14.365904365904367,
      "grad_norm": 0.00010646719601936638,
      "learning_rate": 0.27845290759485225,
      "loss": 0.2696,
      "num_input_tokens_seen": 5270080,
      "step": 6910
    },
    {
      "epoch": 14.376299376299377,
      "grad_norm": 0.00040050657116807997,
      "learning_rate": 0.278422479729925,
      "loss": 0.2413,
      "num_input_tokens_seen": 5273856,
      "step": 6915
    },
    {
      "epoch": 14.386694386694387,
      "grad_norm": 0.0005284688086248934,
      "learning_rate": 0.2783920320606375,
      "loss": 0.2846,
      "num_input_tokens_seen": 5277792,
      "step": 6920
    },
    {
      "epoch": 14.397089397089397,
      "grad_norm": 0.0003488158399704844,
      "learning_rate": 0.2783615645916852,
      "loss": 0.2499,
      "num_input_tokens_seen": 5281504,
      "step": 6925
    },
    {
      "epoch": 14.407484407484407,
      "grad_norm": 0.0012154263677075505,
      "learning_rate": 0.2783310773277666,
      "loss": 0.2734,
      "num_input_tokens_seen": 5285280,
      "step": 6930
    },
    {
      "epoch": 14.417879417879417,
      "grad_norm": 0.00025995244504883885,
      "learning_rate": 0.2783005702735831,
      "loss": 0.269,
      "num_input_tokens_seen": 5289312,
      "step": 6935
    },
    {
      "epoch": 14.428274428274428,
      "grad_norm": 0.000442017539171502,
      "learning_rate": 0.2782700434338394,
      "loss": 0.2717,
      "num_input_tokens_seen": 5293248,
      "step": 6940
    },
    {
      "epoch": 14.43866943866944,
      "grad_norm": 0.0014491007896140218,
      "learning_rate": 0.278239496813243,
      "loss": 0.2769,
      "num_input_tokens_seen": 5296992,
      "step": 6945
    },
    {
      "epoch": 14.44906444906445,
      "grad_norm": 5.974783198325895e-05,
      "learning_rate": 0.27820893041650463,
      "loss": 0.2615,
      "num_input_tokens_seen": 5300704,
      "step": 6950
    },
    {
      "epoch": 14.45945945945946,
      "grad_norm": 0.00010668839240679517,
      "learning_rate": 0.27817834424833804,
      "loss": 0.2766,
      "num_input_tokens_seen": 5304416,
      "step": 6955
    },
    {
      "epoch": 14.46985446985447,
      "grad_norm": 0.0001003067081910558,
      "learning_rate": 0.27814773831345996,
      "loss": 0.2709,
      "num_input_tokens_seen": 5308192,
      "step": 6960
    },
    {
      "epoch": 14.48024948024948,
      "grad_norm": 0.00045547171612270176,
      "learning_rate": 0.2781171126165902,
      "loss": 0.2867,
      "num_input_tokens_seen": 5311968,
      "step": 6965
    },
    {
      "epoch": 14.49064449064449,
      "grad_norm": 0.0005999349523335695,
      "learning_rate": 0.2780864671624517,
      "loss": 0.2887,
      "num_input_tokens_seen": 5315808,
      "step": 6970
    },
    {
      "epoch": 14.5010395010395,
      "grad_norm": 0.0009601388592272997,
      "learning_rate": 0.27805580195577034,
      "loss": 0.2742,
      "num_input_tokens_seen": 5319552,
      "step": 6975
    },
    {
      "epoch": 14.511434511434512,
      "grad_norm": 0.0010883768554776907,
      "learning_rate": 0.2780251170012751,
      "loss": 0.2862,
      "num_input_tokens_seen": 5323328,
      "step": 6980
    },
    {
      "epoch": 14.521829521829522,
      "grad_norm": 0.0006590487901121378,
      "learning_rate": 0.27799441230369787,
      "loss": 0.2781,
      "num_input_tokens_seen": 5327296,
      "step": 6985
    },
    {
      "epoch": 14.532224532224532,
      "grad_norm": 0.001135352998971939,
      "learning_rate": 0.27796368786777387,
      "loss": 0.2862,
      "num_input_tokens_seen": 5331168,
      "step": 6990
    },
    {
      "epoch": 14.542619542619542,
      "grad_norm": 0.0002866137365344912,
      "learning_rate": 0.277932943698241,
      "loss": 0.2378,
      "num_input_tokens_seen": 5335072,
      "step": 6995
    },
    {
      "epoch": 14.553014553014552,
      "grad_norm": 0.00010941985237877816,
      "learning_rate": 0.2779021797998406,
      "loss": 0.297,
      "num_input_tokens_seen": 5339008,
      "step": 7000
    },
    {
      "epoch": 14.553014553014552,
      "eval_loss": 0.25118035078048706,
      "eval_runtime": 13.4163,
      "eval_samples_per_second": 63.803,
      "eval_steps_per_second": 15.951,
      "num_input_tokens_seen": 5339008,
      "step": 7000
    },
    {
      "epoch": 14.563409563409563,
      "grad_norm": 0.00032486781128682196,
      "learning_rate": 0.2778713961773167,
      "loss": 0.2902,
      "num_input_tokens_seen": 5343072,
      "step": 7005
    },
    {
      "epoch": 14.573804573804575,
      "grad_norm": 0.0010244246805086732,
      "learning_rate": 0.2778405928354166,
      "loss": 0.2826,
      "num_input_tokens_seen": 5346944,
      "step": 7010
    },
    {
      "epoch": 14.584199584199585,
      "grad_norm": 0.0006725835264660418,
      "learning_rate": 0.27780976977889055,
      "loss": 0.2716,
      "num_input_tokens_seen": 5350720,
      "step": 7015
    },
    {
      "epoch": 14.594594594594595,
      "grad_norm": 0.001178294070996344,
      "learning_rate": 0.27777892701249185,
      "loss": 0.268,
      "num_input_tokens_seen": 5354624,
      "step": 7020
    },
    {
      "epoch": 14.604989604989605,
      "grad_norm": 9.888135537039489e-05,
      "learning_rate": 0.2777480645409768,
      "loss": 0.2664,
      "num_input_tokens_seen": 5358432,
      "step": 7025
    },
    {
      "epoch": 14.615384615384615,
      "grad_norm": 0.001524461666122079,
      "learning_rate": 0.27771718236910486,
      "loss": 0.2794,
      "num_input_tokens_seen": 5362272,
      "step": 7030
    },
    {
      "epoch": 14.625779625779625,
      "grad_norm": 0.0018248377600684762,
      "learning_rate": 0.27768628050163835,
      "loss": 0.285,
      "num_input_tokens_seen": 5366080,
      "step": 7035
    },
    {
      "epoch": 14.636174636174637,
      "grad_norm": 0.0005425760755315423,
      "learning_rate": 0.2776553589433428,
      "loss": 0.2868,
      "num_input_tokens_seen": 5369952,
      "step": 7040
    },
    {
      "epoch": 14.646569646569647,
      "grad_norm": 0.0002823318645823747,
      "learning_rate": 0.27762441769898666,
      "loss": 0.269,
      "num_input_tokens_seen": 5373792,
      "step": 7045
    },
    {
      "epoch": 14.656964656964657,
      "grad_norm": 0.0007993949111551046,
      "learning_rate": 0.2775934567733415,
      "loss": 0.2653,
      "num_input_tokens_seen": 5377696,
      "step": 7050
    },
    {
      "epoch": 14.667359667359667,
      "grad_norm": 0.0005036405054852366,
      "learning_rate": 0.2775624761711819,
      "loss": 0.2945,
      "num_input_tokens_seen": 5381568,
      "step": 7055
    },
    {
      "epoch": 14.677754677754677,
      "grad_norm": 0.0004559840017464012,
      "learning_rate": 0.2775314758972854,
      "loss": 0.2559,
      "num_input_tokens_seen": 5385248,
      "step": 7060
    },
    {
      "epoch": 14.688149688149688,
      "grad_norm": 0.00017250121163669974,
      "learning_rate": 0.2775004559564327,
      "loss": 0.2511,
      "num_input_tokens_seen": 5389024,
      "step": 7065
    },
    {
      "epoch": 14.698544698544698,
      "grad_norm": 0.0004886345122940838,
      "learning_rate": 0.2774694163534073,
      "loss": 0.2766,
      "num_input_tokens_seen": 5392896,
      "step": 7070
    },
    {
      "epoch": 14.70893970893971,
      "grad_norm": 0.00036456025554798543,
      "learning_rate": 0.27743835709299614,
      "loss": 0.2798,
      "num_input_tokens_seen": 5396768,
      "step": 7075
    },
    {
      "epoch": 14.71933471933472,
      "grad_norm": 0.0020755392033606768,
      "learning_rate": 0.2774072781799888,
      "loss": 0.2768,
      "num_input_tokens_seen": 5400640,
      "step": 7080
    },
    {
      "epoch": 14.72972972972973,
      "grad_norm": 0.0006276414496824145,
      "learning_rate": 0.27737617961917804,
      "loss": 0.282,
      "num_input_tokens_seen": 5404448,
      "step": 7085
    },
    {
      "epoch": 14.74012474012474,
      "grad_norm": 0.0009041255689226091,
      "learning_rate": 0.27734506141535964,
      "loss": 0.2608,
      "num_input_tokens_seen": 5408224,
      "step": 7090
    },
    {
      "epoch": 14.75051975051975,
      "grad_norm": 0.0009239327046088874,
      "learning_rate": 0.2773139235733325,
      "loss": 0.2736,
      "num_input_tokens_seen": 5412064,
      "step": 7095
    },
    {
      "epoch": 14.76091476091476,
      "grad_norm": 0.0006969145033508539,
      "learning_rate": 0.2772827660978984,
      "loss": 0.2683,
      "num_input_tokens_seen": 5415808,
      "step": 7100
    },
    {
      "epoch": 14.771309771309772,
      "grad_norm": 0.0007697590626776218,
      "learning_rate": 0.27725158899386226,
      "loss": 0.2809,
      "num_input_tokens_seen": 5419552,
      "step": 7105
    },
    {
      "epoch": 14.781704781704782,
      "grad_norm": 0.0008874837076291442,
      "learning_rate": 0.27722039226603196,
      "loss": 0.2784,
      "num_input_tokens_seen": 5423296,
      "step": 7110
    },
    {
      "epoch": 14.792099792099792,
      "grad_norm": 0.0005333812441676855,
      "learning_rate": 0.2771891759192184,
      "loss": 0.2919,
      "num_input_tokens_seen": 5427008,
      "step": 7115
    },
    {
      "epoch": 14.802494802494802,
      "grad_norm": 0.0011189348297193646,
      "learning_rate": 0.2771579399582355,
      "loss": 0.2493,
      "num_input_tokens_seen": 5430784,
      "step": 7120
    },
    {
      "epoch": 14.812889812889813,
      "grad_norm": 9.423021401744336e-05,
      "learning_rate": 0.2771266843879004,
      "loss": 0.3178,
      "num_input_tokens_seen": 5434528,
      "step": 7125
    },
    {
      "epoch": 14.823284823284823,
      "grad_norm": 0.0003147637180518359,
      "learning_rate": 0.2770954092130329,
      "loss": 0.2551,
      "num_input_tokens_seen": 5438304,
      "step": 7130
    },
    {
      "epoch": 14.833679833679835,
      "grad_norm": 0.000748441438190639,
      "learning_rate": 0.27706411443845613,
      "loss": 0.2665,
      "num_input_tokens_seen": 5442080,
      "step": 7135
    },
    {
      "epoch": 14.844074844074845,
      "grad_norm": 0.00019327978952787817,
      "learning_rate": 0.27703280006899617,
      "loss": 0.2796,
      "num_input_tokens_seen": 5445952,
      "step": 7140
    },
    {
      "epoch": 14.854469854469855,
      "grad_norm": 0.00251404638402164,
      "learning_rate": 0.277001466109482,
      "loss": 0.2741,
      "num_input_tokens_seen": 5449728,
      "step": 7145
    },
    {
      "epoch": 14.864864864864865,
      "grad_norm": 0.0012029935605823994,
      "learning_rate": 0.2769701125647458,
      "loss": 0.2416,
      "num_input_tokens_seen": 5453440,
      "step": 7150
    },
    {
      "epoch": 14.875259875259875,
      "grad_norm": 0.00038265660987235606,
      "learning_rate": 0.27693873943962266,
      "loss": 0.2996,
      "num_input_tokens_seen": 5457440,
      "step": 7155
    },
    {
      "epoch": 14.885654885654885,
      "grad_norm": 0.00014777277829125524,
      "learning_rate": 0.2769073467389506,
      "loss": 0.2727,
      "num_input_tokens_seen": 5461344,
      "step": 7160
    },
    {
      "epoch": 14.896049896049895,
      "grad_norm": 0.0017229628283530474,
      "learning_rate": 0.2768759344675709,
      "loss": 0.2927,
      "num_input_tokens_seen": 5465280,
      "step": 7165
    },
    {
      "epoch": 14.906444906444907,
      "grad_norm": 0.000883397413417697,
      "learning_rate": 0.27684450263032767,
      "loss": 0.2671,
      "num_input_tokens_seen": 5468896,
      "step": 7170
    },
    {
      "epoch": 14.916839916839917,
      "grad_norm": 0.0008071648189797997,
      "learning_rate": 0.2768130512320682,
      "loss": 0.2594,
      "num_input_tokens_seen": 5472608,
      "step": 7175
    },
    {
      "epoch": 14.927234927234927,
      "grad_norm": 0.0005227037472650409,
      "learning_rate": 0.27678158027764244,
      "loss": 0.2937,
      "num_input_tokens_seen": 5476352,
      "step": 7180
    },
    {
      "epoch": 14.937629937629938,
      "grad_norm": 0.00016350824444089085,
      "learning_rate": 0.27675008977190385,
      "loss": 0.2686,
      "num_input_tokens_seen": 5480032,
      "step": 7185
    },
    {
      "epoch": 14.948024948024948,
      "grad_norm": 0.00031125274836085737,
      "learning_rate": 0.2767185797197086,
      "loss": 0.2768,
      "num_input_tokens_seen": 5483808,
      "step": 7190
    },
    {
      "epoch": 14.958419958419958,
      "grad_norm": 0.00030609374516643584,
      "learning_rate": 0.2766870501259159,
      "loss": 0.2628,
      "num_input_tokens_seen": 5487584,
      "step": 7195
    },
    {
      "epoch": 14.96881496881497,
      "grad_norm": 0.00020283118647057563,
      "learning_rate": 0.276655500995388,
      "loss": 0.2724,
      "num_input_tokens_seen": 5491424,
      "step": 7200
    },
    {
      "epoch": 14.96881496881497,
      "eval_loss": 0.24888208508491516,
      "eval_runtime": 13.4766,
      "eval_samples_per_second": 63.517,
      "eval_steps_per_second": 15.879,
      "num_input_tokens_seen": 5491424,
      "step": 7200
    },
    {
      "epoch": 14.97920997920998,
      "grad_norm": 0.000280176434898749,
      "learning_rate": 0.27662393233299015,
      "loss": 0.2685,
      "num_input_tokens_seen": 5495296,
      "step": 7205
    },
    {
      "epoch": 14.98960498960499,
      "grad_norm": 0.0021873991936445236,
      "learning_rate": 0.27659234414359074,
      "loss": 0.2794,
      "num_input_tokens_seen": 5499232,
      "step": 7210
    },
    {
      "epoch": 15.0,
      "grad_norm": 0.0002343564701732248,
      "learning_rate": 0.27656073643206097,
      "loss": 0.2657,
      "num_input_tokens_seen": 5503048,
      "step": 7215
    },
    {
      "epoch": 15.01039501039501,
      "grad_norm": 0.00028844503685832024,
      "learning_rate": 0.27652910920327517,
      "loss": 0.2634,
      "num_input_tokens_seen": 5507048,
      "step": 7220
    },
    {
      "epoch": 15.02079002079002,
      "grad_norm": 0.0009478189167566597,
      "learning_rate": 0.2764974624621107,
      "loss": 0.2657,
      "num_input_tokens_seen": 5510952,
      "step": 7225
    },
    {
      "epoch": 15.03118503118503,
      "grad_norm": 0.0002082430146401748,
      "learning_rate": 0.2764657962134479,
      "loss": 0.2751,
      "num_input_tokens_seen": 5514792,
      "step": 7230
    },
    {
      "epoch": 15.041580041580042,
      "grad_norm": 0.0013066637329757214,
      "learning_rate": 0.27643411046217,
      "loss": 0.2744,
      "num_input_tokens_seen": 5518600,
      "step": 7235
    },
    {
      "epoch": 15.051975051975052,
      "grad_norm": 0.0003306009457446635,
      "learning_rate": 0.27640240521316334,
      "loss": 0.2549,
      "num_input_tokens_seen": 5522344,
      "step": 7240
    },
    {
      "epoch": 15.062370062370062,
      "grad_norm": 0.0006251315353438258,
      "learning_rate": 0.2763706804713174,
      "loss": 0.2798,
      "num_input_tokens_seen": 5526152,
      "step": 7245
    },
    {
      "epoch": 15.072765072765073,
      "grad_norm": 0.0002411407040199265,
      "learning_rate": 0.2763389362415245,
      "loss": 0.2819,
      "num_input_tokens_seen": 5529832,
      "step": 7250
    },
    {
      "epoch": 15.083160083160083,
      "grad_norm": 0.000795442727394402,
      "learning_rate": 0.27630717252867987,
      "loss": 0.2973,
      "num_input_tokens_seen": 5533576,
      "step": 7255
    },
    {
      "epoch": 15.093555093555093,
      "grad_norm": 0.0003539549943525344,
      "learning_rate": 0.276275389337682,
      "loss": 0.273,
      "num_input_tokens_seen": 5537512,
      "step": 7260
    },
    {
      "epoch": 15.103950103950105,
      "grad_norm": 0.0010497792391106486,
      "learning_rate": 0.2762435866734322,
      "loss": 0.3076,
      "num_input_tokens_seen": 5541256,
      "step": 7265
    },
    {
      "epoch": 15.114345114345115,
      "grad_norm": 0.00047377869486808777,
      "learning_rate": 0.27621176454083485,
      "loss": 0.2641,
      "num_input_tokens_seen": 5545160,
      "step": 7270
    },
    {
      "epoch": 15.124740124740125,
      "grad_norm": 0.00047894002636894584,
      "learning_rate": 0.2761799229447973,
      "loss": 0.2676,
      "num_input_tokens_seen": 5549032,
      "step": 7275
    },
    {
      "epoch": 15.135135135135135,
      "grad_norm": 0.0009624660015106201,
      "learning_rate": 0.27614806189023006,
      "loss": 0.3008,
      "num_input_tokens_seen": 5552936,
      "step": 7280
    },
    {
      "epoch": 15.145530145530145,
      "grad_norm": 0.0003097022126894444,
      "learning_rate": 0.27611618138204636,
      "loss": 0.2847,
      "num_input_tokens_seen": 5556744,
      "step": 7285
    },
    {
      "epoch": 15.155925155925155,
      "grad_norm": 0.0007015501614660025,
      "learning_rate": 0.2760842814251626,
      "loss": 0.2779,
      "num_input_tokens_seen": 5560424,
      "step": 7290
    },
    {
      "epoch": 15.166320166320165,
      "grad_norm": 0.0013820198364555836,
      "learning_rate": 0.2760523620244982,
      "loss": 0.2677,
      "num_input_tokens_seen": 5564296,
      "step": 7295
    },
    {
      "epoch": 15.176715176715177,
      "grad_norm": 0.00037687172880396247,
      "learning_rate": 0.27602042318497544,
      "loss": 0.2634,
      "num_input_tokens_seen": 5568264,
      "step": 7300
    },
    {
      "epoch": 15.187110187110187,
      "grad_norm": 0.0014727680245414376,
      "learning_rate": 0.2759884649115198,
      "loss": 0.2779,
      "num_input_tokens_seen": 5572008,
      "step": 7305
    },
    {
      "epoch": 15.197505197505198,
      "grad_norm": 0.00024315298651345074,
      "learning_rate": 0.2759564872090596,
      "loss": 0.2565,
      "num_input_tokens_seen": 5575880,
      "step": 7310
    },
    {
      "epoch": 15.207900207900208,
      "grad_norm": 0.0002517875109333545,
      "learning_rate": 0.2759244900825262,
      "loss": 0.2737,
      "num_input_tokens_seen": 5579592,
      "step": 7315
    },
    {
      "epoch": 15.218295218295218,
      "grad_norm": 0.00039618040318600833,
      "learning_rate": 0.2758924735368539,
      "loss": 0.2653,
      "num_input_tokens_seen": 5583304,
      "step": 7320
    },
    {
      "epoch": 15.228690228690228,
      "grad_norm": 0.00029686916968785226,
      "learning_rate": 0.27586043757698014,
      "loss": 0.2701,
      "num_input_tokens_seen": 5587208,
      "step": 7325
    },
    {
      "epoch": 15.23908523908524,
      "grad_norm": 0.0010557174682617188,
      "learning_rate": 0.27582838220784534,
      "loss": 0.2777,
      "num_input_tokens_seen": 5591016,
      "step": 7330
    },
    {
      "epoch": 15.24948024948025,
      "grad_norm": 0.0010066907852888107,
      "learning_rate": 0.27579630743439265,
      "loss": 0.2736,
      "num_input_tokens_seen": 5594824,
      "step": 7335
    },
    {
      "epoch": 15.25987525987526,
      "grad_norm": 0.00014965262380428612,
      "learning_rate": 0.2757642132615686,
      "loss": 0.2507,
      "num_input_tokens_seen": 5598600,
      "step": 7340
    },
    {
      "epoch": 15.27027027027027,
      "grad_norm": 0.0001960218942258507,
      "learning_rate": 0.2757320996943223,
      "loss": 0.2588,
      "num_input_tokens_seen": 5602280,
      "step": 7345
    },
    {
      "epoch": 15.28066528066528,
      "grad_norm": 0.0004989554290659726,
      "learning_rate": 0.2756999667376062,
      "loss": 0.254,
      "num_input_tokens_seen": 5606312,
      "step": 7350
    },
    {
      "epoch": 15.29106029106029,
      "grad_norm": 0.0005768209230154753,
      "learning_rate": 0.2756678143963756,
      "loss": 0.2364,
      "num_input_tokens_seen": 5610120,
      "step": 7355
    },
    {
      "epoch": 15.301455301455302,
      "grad_norm": 0.00041111421887762845,
      "learning_rate": 0.2756356426755888,
      "loss": 0.2546,
      "num_input_tokens_seen": 5614024,
      "step": 7360
    },
    {
      "epoch": 15.311850311850312,
      "grad_norm": 0.0007249697227962315,
      "learning_rate": 0.27560345158020705,
      "loss": 0.2577,
      "num_input_tokens_seen": 5617896,
      "step": 7365
    },
    {
      "epoch": 15.322245322245323,
      "grad_norm": 0.0007824570639058948,
      "learning_rate": 0.27557124111519465,
      "loss": 0.2714,
      "num_input_tokens_seen": 5621832,
      "step": 7370
    },
    {
      "epoch": 15.332640332640333,
      "grad_norm": 0.00031998445047065616,
      "learning_rate": 0.27553901128551883,
      "loss": 0.2661,
      "num_input_tokens_seen": 5625640,
      "step": 7375
    },
    {
      "epoch": 15.343035343035343,
      "grad_norm": 0.0008927280432544649,
      "learning_rate": 0.2755067620961498,
      "loss": 0.2418,
      "num_input_tokens_seen": 5629480,
      "step": 7380
    },
    {
      "epoch": 15.353430353430353,
      "grad_norm": 0.00022247085871640593,
      "learning_rate": 0.27547449355206094,
      "loss": 0.2928,
      "num_input_tokens_seen": 5633224,
      "step": 7385
    },
    {
      "epoch": 15.363825363825363,
      "grad_norm": 0.0015979583840817213,
      "learning_rate": 0.2754422056582283,
      "loss": 0.2797,
      "num_input_tokens_seen": 5636936,
      "step": 7390
    },
    {
      "epoch": 15.374220374220375,
      "grad_norm": 0.00013116904301568866,
      "learning_rate": 0.27540989841963115,
      "loss": 0.273,
      "num_input_tokens_seen": 5640712,
      "step": 7395
    },
    {
      "epoch": 15.384615384615385,
      "grad_norm": 0.00010890357953030616,
      "learning_rate": 0.27537757184125167,
      "loss": 0.2623,
      "num_input_tokens_seen": 5644520,
      "step": 7400
    },
    {
      "epoch": 15.384615384615385,
      "eval_loss": 0.24839846789836884,
      "eval_runtime": 13.4346,
      "eval_samples_per_second": 63.716,
      "eval_steps_per_second": 15.929,
      "num_input_tokens_seen": 5644520,
      "step": 7400
    },
    {
      "epoch": 15.395010395010395,
      "grad_norm": 0.00021131696121301502,
      "learning_rate": 0.275345225928075,
      "loss": 0.2648,
      "num_input_tokens_seen": 5648264,
      "step": 7405
    },
    {
      "epoch": 15.405405405405405,
      "grad_norm": 0.0008523308206349611,
      "learning_rate": 0.2753128606850893,
      "loss": 0.2736,
      "num_input_tokens_seen": 5652072,
      "step": 7410
    },
    {
      "epoch": 15.415800415800415,
      "grad_norm": 0.0011783391237258911,
      "learning_rate": 0.2752804761172858,
      "loss": 0.2763,
      "num_input_tokens_seen": 5656008,
      "step": 7415
    },
    {
      "epoch": 15.426195426195425,
      "grad_norm": 0.0001437998580513522,
      "learning_rate": 0.27524807222965836,
      "loss": 0.2707,
      "num_input_tokens_seen": 5659848,
      "step": 7420
    },
    {
      "epoch": 15.436590436590437,
      "grad_norm": 0.0004677558899857104,
      "learning_rate": 0.27521564902720436,
      "loss": 0.2402,
      "num_input_tokens_seen": 5663592,
      "step": 7425
    },
    {
      "epoch": 15.446985446985448,
      "grad_norm": 0.0001335811975877732,
      "learning_rate": 0.2751832065149236,
      "loss": 0.2776,
      "num_input_tokens_seen": 5667272,
      "step": 7430
    },
    {
      "epoch": 15.457380457380458,
      "grad_norm": 0.00037285161670297384,
      "learning_rate": 0.2751507446978193,
      "loss": 0.2617,
      "num_input_tokens_seen": 5671208,
      "step": 7435
    },
    {
      "epoch": 15.467775467775468,
      "grad_norm": 0.00011653076944639906,
      "learning_rate": 0.2751182635808974,
      "loss": 0.2621,
      "num_input_tokens_seen": 5674984,
      "step": 7440
    },
    {
      "epoch": 15.478170478170478,
      "grad_norm": 0.0005675626453012228,
      "learning_rate": 0.27508576316916694,
      "loss": 0.2913,
      "num_input_tokens_seen": 5678856,
      "step": 7445
    },
    {
      "epoch": 15.488565488565488,
      "grad_norm": 8.135367534123361e-05,
      "learning_rate": 0.2750532434676399,
      "loss": 0.2808,
      "num_input_tokens_seen": 5682696,
      "step": 7450
    },
    {
      "epoch": 15.4989604989605,
      "grad_norm": 0.0005800941144116223,
      "learning_rate": 0.27502070448133115,
      "loss": 0.2593,
      "num_input_tokens_seen": 5686472,
      "step": 7455
    },
    {
      "epoch": 15.50935550935551,
      "grad_norm": 0.0006041646702215075,
      "learning_rate": 0.2749881462152587,
      "loss": 0.2907,
      "num_input_tokens_seen": 5690312,
      "step": 7460
    },
    {
      "epoch": 15.51975051975052,
      "grad_norm": 0.0007312985835596919,
      "learning_rate": 0.2749555686744434,
      "loss": 0.2664,
      "num_input_tokens_seen": 5694120,
      "step": 7465
    },
    {
      "epoch": 15.53014553014553,
      "grad_norm": 0.00031196887721307576,
      "learning_rate": 0.2749229718639091,
      "loss": 0.2269,
      "num_input_tokens_seen": 5697992,
      "step": 7470
    },
    {
      "epoch": 15.54054054054054,
      "grad_norm": 0.0002736408787313849,
      "learning_rate": 0.27489035578868265,
      "loss": 0.2817,
      "num_input_tokens_seen": 5701800,
      "step": 7475
    },
    {
      "epoch": 15.55093555093555,
      "grad_norm": 0.0003088978410232812,
      "learning_rate": 0.2748577204537939,
      "loss": 0.2589,
      "num_input_tokens_seen": 5705672,
      "step": 7480
    },
    {
      "epoch": 15.56133056133056,
      "grad_norm": 0.0001431064010830596,
      "learning_rate": 0.2748250658642756,
      "loss": 0.2604,
      "num_input_tokens_seen": 5709352,
      "step": 7485
    },
    {
      "epoch": 15.571725571725572,
      "grad_norm": 0.00029914206243120134,
      "learning_rate": 0.2747923920251634,
      "loss": 0.2788,
      "num_input_tokens_seen": 5713160,
      "step": 7490
    },
    {
      "epoch": 15.582120582120583,
      "grad_norm": 0.00027012015925720334,
      "learning_rate": 0.27475969894149627,
      "loss": 0.2761,
      "num_input_tokens_seen": 5716872,
      "step": 7495
    },
    {
      "epoch": 15.592515592515593,
      "grad_norm": 0.0009495631093159318,
      "learning_rate": 0.2747269866183156,
      "loss": 0.2775,
      "num_input_tokens_seen": 5720648,
      "step": 7500
    },
    {
      "epoch": 15.602910602910603,
      "grad_norm": 0.0007115293992683291,
      "learning_rate": 0.27469425506066625,
      "loss": 0.2586,
      "num_input_tokens_seen": 5724456,
      "step": 7505
    },
    {
      "epoch": 15.613305613305613,
      "grad_norm": 0.00032747091609053314,
      "learning_rate": 0.27466150427359576,
      "loss": 0.2771,
      "num_input_tokens_seen": 5728392,
      "step": 7510
    },
    {
      "epoch": 15.623700623700623,
      "grad_norm": 0.00010101830412168056,
      "learning_rate": 0.2746287342621547,
      "loss": 0.2651,
      "num_input_tokens_seen": 5732168,
      "step": 7515
    },
    {
      "epoch": 15.634095634095633,
      "grad_norm": 0.0008337240433320403,
      "learning_rate": 0.2745959450313966,
      "loss": 0.2629,
      "num_input_tokens_seen": 5735848,
      "step": 7520
    },
    {
      "epoch": 15.644490644490645,
      "grad_norm": 0.0012070474913343787,
      "learning_rate": 0.27456313658637804,
      "loss": 0.277,
      "num_input_tokens_seen": 5739720,
      "step": 7525
    },
    {
      "epoch": 15.654885654885655,
      "grad_norm": 0.0004778236907441169,
      "learning_rate": 0.27453030893215846,
      "loss": 0.2817,
      "num_input_tokens_seen": 5743368,
      "step": 7530
    },
    {
      "epoch": 15.665280665280665,
      "grad_norm": 0.00017180699796881527,
      "learning_rate": 0.2744974620738003,
      "loss": 0.2627,
      "num_input_tokens_seen": 5747176,
      "step": 7535
    },
    {
      "epoch": 15.675675675675675,
      "grad_norm": 0.0009562495979480445,
      "learning_rate": 0.27446459601636897,
      "loss": 0.2576,
      "num_input_tokens_seen": 5751144,
      "step": 7540
    },
    {
      "epoch": 15.686070686070686,
      "grad_norm": 0.00031361792935058475,
      "learning_rate": 0.2744317107649328,
      "loss": 0.2596,
      "num_input_tokens_seen": 5754952,
      "step": 7545
    },
    {
      "epoch": 15.696465696465696,
      "grad_norm": 0.00020459649385884404,
      "learning_rate": 0.2743988063245631,
      "loss": 0.2636,
      "num_input_tokens_seen": 5758792,
      "step": 7550
    },
    {
      "epoch": 15.706860706860708,
      "grad_norm": 0.0005400121444836259,
      "learning_rate": 0.2743658827003342,
      "loss": 0.2946,
      "num_input_tokens_seen": 5762600,
      "step": 7555
    },
    {
      "epoch": 15.717255717255718,
      "grad_norm": 0.00036298486520536244,
      "learning_rate": 0.27433293989732327,
      "loss": 0.2715,
      "num_input_tokens_seen": 5766376,
      "step": 7560
    },
    {
      "epoch": 15.727650727650728,
      "grad_norm": 0.0008316165767610073,
      "learning_rate": 0.27429997792061056,
      "loss": 0.2731,
      "num_input_tokens_seen": 5770056,
      "step": 7565
    },
    {
      "epoch": 15.738045738045738,
      "grad_norm": 0.0005599592113867402,
      "learning_rate": 0.27426699677527927,
      "loss": 0.2803,
      "num_input_tokens_seen": 5773992,
      "step": 7570
    },
    {
      "epoch": 15.748440748440748,
      "grad_norm": 0.0001371019461657852,
      "learning_rate": 0.2742339964664154,
      "loss": 0.2791,
      "num_input_tokens_seen": 5777736,
      "step": 7575
    },
    {
      "epoch": 15.758835758835758,
      "grad_norm": 0.000719712523277849,
      "learning_rate": 0.274200976999108,
      "loss": 0.2826,
      "num_input_tokens_seen": 5781608,
      "step": 7580
    },
    {
      "epoch": 15.76923076923077,
      "grad_norm": 0.00038290329393930733,
      "learning_rate": 0.27416793837844916,
      "loss": 0.271,
      "num_input_tokens_seen": 5785288,
      "step": 7585
    },
    {
      "epoch": 15.77962577962578,
      "grad_norm": 0.0005798368365503848,
      "learning_rate": 0.27413488060953384,
      "loss": 0.2681,
      "num_input_tokens_seen": 5789192,
      "step": 7590
    },
    {
      "epoch": 15.79002079002079,
      "grad_norm": 0.0005446581053547561,
      "learning_rate": 0.27410180369745996,
      "loss": 0.2505,
      "num_input_tokens_seen": 5792968,
      "step": 7595
    },
    {
      "epoch": 15.8004158004158,
      "grad_norm": 0.0006107335793785751,
      "learning_rate": 0.27406870764732844,
      "loss": 0.2063,
      "num_input_tokens_seen": 5796744,
      "step": 7600
    },
    {
      "epoch": 15.8004158004158,
      "eval_loss": 0.2708362638950348,
      "eval_runtime": 13.4023,
      "eval_samples_per_second": 63.87,
      "eval_steps_per_second": 15.967,
      "num_input_tokens_seen": 5796744,
      "step": 7600
    },
    {
      "epoch": 15.81081081081081,
      "grad_norm": 0.0011588734341785312,
      "learning_rate": 0.27403559246424297,
      "loss": 0.289,
      "num_input_tokens_seen": 5800552,
      "step": 7605
    },
    {
      "epoch": 15.82120582120582,
      "grad_norm": 0.00027321293600834906,
      "learning_rate": 0.2740024581533105,
      "loss": 0.2921,
      "num_input_tokens_seen": 5804360,
      "step": 7610
    },
    {
      "epoch": 15.83160083160083,
      "grad_norm": 0.0006256555789150298,
      "learning_rate": 0.2739693047196406,
      "loss": 0.2967,
      "num_input_tokens_seen": 5808328,
      "step": 7615
    },
    {
      "epoch": 15.841995841995843,
      "grad_norm": 0.00046540325274690986,
      "learning_rate": 0.27393613216834606,
      "loss": 0.2725,
      "num_input_tokens_seen": 5812136,
      "step": 7620
    },
    {
      "epoch": 15.852390852390853,
      "grad_norm": 0.002030816627666354,
      "learning_rate": 0.2739029405045424,
      "loss": 0.2905,
      "num_input_tokens_seen": 5815912,
      "step": 7625
    },
    {
      "epoch": 15.862785862785863,
      "grad_norm": 0.00039951325743459165,
      "learning_rate": 0.2738697297333483,
      "loss": 0.2606,
      "num_input_tokens_seen": 5819624,
      "step": 7630
    },
    {
      "epoch": 15.873180873180873,
      "grad_norm": 0.00013634399510920048,
      "learning_rate": 0.2738364998598852,
      "loss": 0.2707,
      "num_input_tokens_seen": 5823400,
      "step": 7635
    },
    {
      "epoch": 15.883575883575883,
      "grad_norm": 0.0010050374548882246,
      "learning_rate": 0.27380325088927765,
      "loss": 0.2715,
      "num_input_tokens_seen": 5827272,
      "step": 7640
    },
    {
      "epoch": 15.893970893970893,
      "grad_norm": 0.0004820208705496043,
      "learning_rate": 0.27376998282665294,
      "loss": 0.2729,
      "num_input_tokens_seen": 5831080,
      "step": 7645
    },
    {
      "epoch": 15.904365904365905,
      "grad_norm": 0.0012906744377687573,
      "learning_rate": 0.27373669567714154,
      "loss": 0.2669,
      "num_input_tokens_seen": 5834760,
      "step": 7650
    },
    {
      "epoch": 15.914760914760915,
      "grad_norm": 0.0008492125198245049,
      "learning_rate": 0.27370338944587663,
      "loss": 0.2603,
      "num_input_tokens_seen": 5838472,
      "step": 7655
    },
    {
      "epoch": 15.925155925155925,
      "grad_norm": 0.0007136305794119835,
      "learning_rate": 0.27367006413799455,
      "loss": 0.2649,
      "num_input_tokens_seen": 5842248,
      "step": 7660
    },
    {
      "epoch": 15.935550935550935,
      "grad_norm": 0.0005543690058402717,
      "learning_rate": 0.2736367197586345,
      "loss": 0.263,
      "num_input_tokens_seen": 5846056,
      "step": 7665
    },
    {
      "epoch": 15.945945945945946,
      "grad_norm": 0.0001712001539999619,
      "learning_rate": 0.2736033563129385,
      "loss": 0.2767,
      "num_input_tokens_seen": 5849704,
      "step": 7670
    },
    {
      "epoch": 15.956340956340956,
      "grad_norm": 0.0004004684742540121,
      "learning_rate": 0.27356997380605164,
      "loss": 0.2759,
      "num_input_tokens_seen": 5853512,
      "step": 7675
    },
    {
      "epoch": 15.966735966735968,
      "grad_norm": 0.00021644837397616357,
      "learning_rate": 0.27353657224312194,
      "loss": 0.2459,
      "num_input_tokens_seen": 5857544,
      "step": 7680
    },
    {
      "epoch": 15.977130977130978,
      "grad_norm": 0.00010654184006853029,
      "learning_rate": 0.2735031516293004,
      "loss": 0.2886,
      "num_input_tokens_seen": 5861384,
      "step": 7685
    },
    {
      "epoch": 15.987525987525988,
      "grad_norm": 0.00025882787303999066,
      "learning_rate": 0.2734697119697408,
      "loss": 0.2721,
      "num_input_tokens_seen": 5865128,
      "step": 7690
    },
    {
      "epoch": 15.997920997920998,
      "grad_norm": 0.0003930949023924768,
      "learning_rate": 0.27343625326959997,
      "loss": 0.2756,
      "num_input_tokens_seen": 5869000,
      "step": 7695
    },
    {
      "epoch": 16.008316008316008,
      "grad_norm": 0.0003782971471082419,
      "learning_rate": 0.27340277553403775,
      "loss": 0.2788,
      "num_input_tokens_seen": 5872928,
      "step": 7700
    },
    {
      "epoch": 16.01871101871102,
      "grad_norm": 7.335251575568691e-05,
      "learning_rate": 0.2733692787682167,
      "loss": 0.2612,
      "num_input_tokens_seen": 5877024,
      "step": 7705
    },
    {
      "epoch": 16.02910602910603,
      "grad_norm": 0.0006023618625476956,
      "learning_rate": 0.27333576297730255,
      "loss": 0.276,
      "num_input_tokens_seen": 5880928,
      "step": 7710
    },
    {
      "epoch": 16.03950103950104,
      "grad_norm": 0.0004253872612025589,
      "learning_rate": 0.2733022281664638,
      "loss": 0.2591,
      "num_input_tokens_seen": 5884704,
      "step": 7715
    },
    {
      "epoch": 16.04989604989605,
      "grad_norm": 0.0008588171331211925,
      "learning_rate": 0.273268674340872,
      "loss": 0.2762,
      "num_input_tokens_seen": 5888544,
      "step": 7720
    },
    {
      "epoch": 16.06029106029106,
      "grad_norm": 0.0009492579847574234,
      "learning_rate": 0.27323510150570146,
      "loss": 0.285,
      "num_input_tokens_seen": 5892352,
      "step": 7725
    },
    {
      "epoch": 16.070686070686072,
      "grad_norm": 0.0006792630301788449,
      "learning_rate": 0.27320150966612966,
      "loss": 0.2829,
      "num_input_tokens_seen": 5896128,
      "step": 7730
    },
    {
      "epoch": 16.08108108108108,
      "grad_norm": 0.0004948316491208971,
      "learning_rate": 0.2731678988273368,
      "loss": 0.2592,
      "num_input_tokens_seen": 5900064,
      "step": 7735
    },
    {
      "epoch": 16.091476091476093,
      "grad_norm": 0.0008550684433430433,
      "learning_rate": 0.27313426899450605,
      "loss": 0.3345,
      "num_input_tokens_seen": 5903712,
      "step": 7740
    },
    {
      "epoch": 16.1018711018711,
      "grad_norm": 0.00038054073229432106,
      "learning_rate": 0.27310062017282366,
      "loss": 0.2934,
      "num_input_tokens_seen": 5907552,
      "step": 7745
    },
    {
      "epoch": 16.112266112266113,
      "grad_norm": 0.0004169708117842674,
      "learning_rate": 0.2730669523674787,
      "loss": 0.2648,
      "num_input_tokens_seen": 5911264,
      "step": 7750
    },
    {
      "epoch": 16.12266112266112,
      "grad_norm": 0.0009244192042388022,
      "learning_rate": 0.2730332655836631,
      "loss": 0.2928,
      "num_input_tokens_seen": 5915008,
      "step": 7755
    },
    {
      "epoch": 16.133056133056133,
      "grad_norm": 0.0001721863227430731,
      "learning_rate": 0.2729995598265718,
      "loss": 0.2555,
      "num_input_tokens_seen": 5918816,
      "step": 7760
    },
    {
      "epoch": 16.143451143451145,
      "grad_norm": 0.0008631807286292315,
      "learning_rate": 0.2729658351014027,
      "loss": 0.2653,
      "num_input_tokens_seen": 5922560,
      "step": 7765
    },
    {
      "epoch": 16.153846153846153,
      "grad_norm": 0.0011688689701259136,
      "learning_rate": 0.27293209141335656,
      "loss": 0.272,
      "num_input_tokens_seen": 5926304,
      "step": 7770
    },
    {
      "epoch": 16.164241164241165,
      "grad_norm": 7.41495969123207e-05,
      "learning_rate": 0.27289832876763703,
      "loss": 0.2706,
      "num_input_tokens_seen": 5930176,
      "step": 7775
    },
    {
      "epoch": 16.174636174636174,
      "grad_norm": 0.000559632433578372,
      "learning_rate": 0.27286454716945074,
      "loss": 0.2639,
      "num_input_tokens_seen": 5934080,
      "step": 7780
    },
    {
      "epoch": 16.185031185031185,
      "grad_norm": 0.00014519080286845565,
      "learning_rate": 0.27283074662400725,
      "loss": 0.274,
      "num_input_tokens_seen": 5938080,
      "step": 7785
    },
    {
      "epoch": 16.195426195426194,
      "grad_norm": 0.0015915262047201395,
      "learning_rate": 0.2727969271365191,
      "loss": 0.2936,
      "num_input_tokens_seen": 5941792,
      "step": 7790
    },
    {
      "epoch": 16.205821205821206,
      "grad_norm": 0.00047141005052253604,
      "learning_rate": 0.2727630887122016,
      "loss": 0.2788,
      "num_input_tokens_seen": 5945600,
      "step": 7795
    },
    {
      "epoch": 16.216216216216218,
      "grad_norm": 0.0008207892533391714,
      "learning_rate": 0.27272923135627314,
      "loss": 0.276,
      "num_input_tokens_seen": 5949536,
      "step": 7800
    },
    {
      "epoch": 16.216216216216218,
      "eval_loss": 0.24984216690063477,
      "eval_runtime": 13.4177,
      "eval_samples_per_second": 63.796,
      "eval_steps_per_second": 15.949,
      "num_input_tokens_seen": 5949536,
      "step": 7800
    },
    {
      "epoch": 16.226611226611226,
      "grad_norm": 0.00035605349694378674,
      "learning_rate": 0.2726953550739548,
      "loss": 0.2676,
      "num_input_tokens_seen": 5953504,
      "step": 7805
    },
    {
      "epoch": 16.237006237006238,
      "grad_norm": 0.0010749581269919872,
      "learning_rate": 0.27266145987047086,
      "loss": 0.2728,
      "num_input_tokens_seen": 5957216,
      "step": 7810
    },
    {
      "epoch": 16.247401247401246,
      "grad_norm": 0.0012606856180354953,
      "learning_rate": 0.27262754575104836,
      "loss": 0.2844,
      "num_input_tokens_seen": 5961088,
      "step": 7815
    },
    {
      "epoch": 16.257796257796258,
      "grad_norm": 0.00040339259430766106,
      "learning_rate": 0.27259361272091726,
      "loss": 0.2683,
      "num_input_tokens_seen": 5964896,
      "step": 7820
    },
    {
      "epoch": 16.26819126819127,
      "grad_norm": 0.00016065733507275581,
      "learning_rate": 0.27255966078531046,
      "loss": 0.3041,
      "num_input_tokens_seen": 5968768,
      "step": 7825
    },
    {
      "epoch": 16.27858627858628,
      "grad_norm": 0.0006202685763128102,
      "learning_rate": 0.2725256899494638,
      "loss": 0.2618,
      "num_input_tokens_seen": 5972800,
      "step": 7830
    },
    {
      "epoch": 16.28898128898129,
      "grad_norm": 0.000727097736671567,
      "learning_rate": 0.272491700218616,
      "loss": 0.2975,
      "num_input_tokens_seen": 5976608,
      "step": 7835
    },
    {
      "epoch": 16.2993762993763,
      "grad_norm": 0.0005357848131097853,
      "learning_rate": 0.27245769159800876,
      "loss": 0.2763,
      "num_input_tokens_seen": 5980480,
      "step": 7840
    },
    {
      "epoch": 16.30977130977131,
      "grad_norm": 0.0002505619777366519,
      "learning_rate": 0.2724236640928865,
      "loss": 0.3001,
      "num_input_tokens_seen": 5984128,
      "step": 7845
    },
    {
      "epoch": 16.32016632016632,
      "grad_norm": 0.00027863041032105684,
      "learning_rate": 0.27238961770849673,
      "loss": 0.2463,
      "num_input_tokens_seen": 5988064,
      "step": 7850
    },
    {
      "epoch": 16.33056133056133,
      "grad_norm": 4.527588316705078e-05,
      "learning_rate": 0.27235555245008997,
      "loss": 0.2369,
      "num_input_tokens_seen": 5991808,
      "step": 7855
    },
    {
      "epoch": 16.340956340956343,
      "grad_norm": 0.00029998397803865373,
      "learning_rate": 0.2723214683229193,
      "loss": 0.2976,
      "num_input_tokens_seen": 5995616,
      "step": 7860
    },
    {
      "epoch": 16.35135135135135,
      "grad_norm": 0.001439696061424911,
      "learning_rate": 0.27228736533224107,
      "loss": 0.2644,
      "num_input_tokens_seen": 5999488,
      "step": 7865
    },
    {
      "epoch": 16.361746361746363,
      "grad_norm": 0.0008670384413562715,
      "learning_rate": 0.27225324348331437,
      "loss": 0.2881,
      "num_input_tokens_seen": 6003232,
      "step": 7870
    },
    {
      "epoch": 16.37214137214137,
      "grad_norm": 0.0005674163694493473,
      "learning_rate": 0.27221910278140116,
      "loss": 0.2788,
      "num_input_tokens_seen": 6007104,
      "step": 7875
    },
    {
      "epoch": 16.382536382536383,
      "grad_norm": 0.0002761151408776641,
      "learning_rate": 0.2721849432317664,
      "loss": 0.2599,
      "num_input_tokens_seen": 6010880,
      "step": 7880
    },
    {
      "epoch": 16.39293139293139,
      "grad_norm": 0.0007668761536478996,
      "learning_rate": 0.2721507648396779,
      "loss": 0.277,
      "num_input_tokens_seen": 6014560,
      "step": 7885
    },
    {
      "epoch": 16.403326403326403,
      "grad_norm": 0.00029539118986576796,
      "learning_rate": 0.27211656761040653,
      "loss": 0.2843,
      "num_input_tokens_seen": 6018272,
      "step": 7890
    },
    {
      "epoch": 16.413721413721415,
      "grad_norm": 0.0005167144699953496,
      "learning_rate": 0.2720823515492257,
      "loss": 0.2809,
      "num_input_tokens_seen": 6022176,
      "step": 7895
    },
    {
      "epoch": 16.424116424116423,
      "grad_norm": 0.0008906522998586297,
      "learning_rate": 0.27204811666141215,
      "loss": 0.2531,
      "num_input_tokens_seen": 6025984,
      "step": 7900
    },
    {
      "epoch": 16.434511434511435,
      "grad_norm": 0.0008986930479295552,
      "learning_rate": 0.2720138629522452,
      "loss": 0.2458,
      "num_input_tokens_seen": 6029952,
      "step": 7905
    },
    {
      "epoch": 16.444906444906444,
      "grad_norm": 0.00016851077089086175,
      "learning_rate": 0.2719795904270073,
      "loss": 0.2899,
      "num_input_tokens_seen": 6033792,
      "step": 7910
    },
    {
      "epoch": 16.455301455301456,
      "grad_norm": 0.0010905377566814423,
      "learning_rate": 0.2719452990909837,
      "loss": 0.2609,
      "num_input_tokens_seen": 6037664,
      "step": 7915
    },
    {
      "epoch": 16.465696465696467,
      "grad_norm": 0.0002057001693174243,
      "learning_rate": 0.2719109889494625,
      "loss": 0.2727,
      "num_input_tokens_seen": 6041632,
      "step": 7920
    },
    {
      "epoch": 16.476091476091476,
      "grad_norm": 0.0011404691031202674,
      "learning_rate": 0.27187666000773475,
      "loss": 0.2702,
      "num_input_tokens_seen": 6045408,
      "step": 7925
    },
    {
      "epoch": 16.486486486486488,
      "grad_norm": 0.0003962080227211118,
      "learning_rate": 0.2718423122710944,
      "loss": 0.2657,
      "num_input_tokens_seen": 6049152,
      "step": 7930
    },
    {
      "epoch": 16.496881496881496,
      "grad_norm": 0.00016637657245155424,
      "learning_rate": 0.2718079457448384,
      "loss": 0.2666,
      "num_input_tokens_seen": 6053024,
      "step": 7935
    },
    {
      "epoch": 16.507276507276508,
      "grad_norm": 0.0007608078303746879,
      "learning_rate": 0.27177356043426637,
      "loss": 0.2789,
      "num_input_tokens_seen": 6056800,
      "step": 7940
    },
    {
      "epoch": 16.517671517671516,
      "grad_norm": 0.001341905095614493,
      "learning_rate": 0.27173915634468104,
      "loss": 0.2705,
      "num_input_tokens_seen": 6060672,
      "step": 7945
    },
    {
      "epoch": 16.528066528066528,
      "grad_norm": 0.0006301902467384934,
      "learning_rate": 0.27170473348138796,
      "loss": 0.2423,
      "num_input_tokens_seen": 6064480,
      "step": 7950
    },
    {
      "epoch": 16.53846153846154,
      "grad_norm": 0.0005650501698255539,
      "learning_rate": 0.27167029184969554,
      "loss": 0.2648,
      "num_input_tokens_seen": 6068160,
      "step": 7955
    },
    {
      "epoch": 16.54885654885655,
      "grad_norm": 0.00035487956483848393,
      "learning_rate": 0.27163583145491504,
      "loss": 0.2455,
      "num_input_tokens_seen": 6071968,
      "step": 7960
    },
    {
      "epoch": 16.55925155925156,
      "grad_norm": 0.0006121479091234505,
      "learning_rate": 0.2716013523023608,
      "loss": 0.2575,
      "num_input_tokens_seen": 6075712,
      "step": 7965
    },
    {
      "epoch": 16.56964656964657,
      "grad_norm": 0.0007161162793636322,
      "learning_rate": 0.27156685439734995,
      "loss": 0.2725,
      "num_input_tokens_seen": 6079520,
      "step": 7970
    },
    {
      "epoch": 16.58004158004158,
      "grad_norm": 0.0010961029911413789,
      "learning_rate": 0.2715323377452024,
      "loss": 0.2676,
      "num_input_tokens_seen": 6083424,
      "step": 7975
    },
    {
      "epoch": 16.59043659043659,
      "grad_norm": 0.0007766846683807671,
      "learning_rate": 0.2714978023512411,
      "loss": 0.286,
      "num_input_tokens_seen": 6087264,
      "step": 7980
    },
    {
      "epoch": 16.6008316008316,
      "grad_norm": 0.000907827285118401,
      "learning_rate": 0.2714632482207918,
      "loss": 0.2839,
      "num_input_tokens_seen": 6091008,
      "step": 7985
    },
    {
      "epoch": 16.611226611226613,
      "grad_norm": 0.0006716377683915198,
      "learning_rate": 0.2714286753591833,
      "loss": 0.2888,
      "num_input_tokens_seen": 6094816,
      "step": 7990
    },
    {
      "epoch": 16.62162162162162,
      "grad_norm": 0.0005515064112842083,
      "learning_rate": 0.27139408377174706,
      "loss": 0.2532,
      "num_input_tokens_seen": 6098624,
      "step": 7995
    },
    {
      "epoch": 16.632016632016633,
      "grad_norm": 0.0005116141401231289,
      "learning_rate": 0.27135947346381756,
      "loss": 0.3006,
      "num_input_tokens_seen": 6102304,
      "step": 8000
    },
    {
      "epoch": 16.632016632016633,
      "eval_loss": 0.2607586085796356,
      "eval_runtime": 13.4437,
      "eval_samples_per_second": 63.673,
      "eval_steps_per_second": 15.918,
      "num_input_tokens_seen": 6102304,
      "step": 8000
    },
    {
      "epoch": 16.64241164241164,
      "grad_norm": 0.0005792885785922408,
      "learning_rate": 0.2713248444407322,
      "loss": 0.2745,
      "num_input_tokens_seen": 6106176,
      "step": 8005
    },
    {
      "epoch": 16.652806652806653,
      "grad_norm": 0.0004844397772103548,
      "learning_rate": 0.27129019670783106,
      "loss": 0.2792,
      "num_input_tokens_seen": 6110048,
      "step": 8010
    },
    {
      "epoch": 16.66320166320166,
      "grad_norm": 0.0007052401197142899,
      "learning_rate": 0.27125553027045746,
      "loss": 0.2722,
      "num_input_tokens_seen": 6113888,
      "step": 8015
    },
    {
      "epoch": 16.673596673596673,
      "grad_norm": 0.00034177390625700355,
      "learning_rate": 0.2712208451339572,
      "loss": 0.2939,
      "num_input_tokens_seen": 6117632,
      "step": 8020
    },
    {
      "epoch": 16.683991683991685,
      "grad_norm": 0.00014036981156095862,
      "learning_rate": 0.27118614130367935,
      "loss": 0.2338,
      "num_input_tokens_seen": 6121376,
      "step": 8025
    },
    {
      "epoch": 16.694386694386694,
      "grad_norm": 0.00019832489488180727,
      "learning_rate": 0.2711514187849756,
      "loss": 0.2479,
      "num_input_tokens_seen": 6124992,
      "step": 8030
    },
    {
      "epoch": 16.704781704781706,
      "grad_norm": 0.0009102554759010673,
      "learning_rate": 0.27111667758320057,
      "loss": 0.2694,
      "num_input_tokens_seen": 6128736,
      "step": 8035
    },
    {
      "epoch": 16.715176715176714,
      "grad_norm": 0.00021431792993098497,
      "learning_rate": 0.27108191770371176,
      "loss": 0.2581,
      "num_input_tokens_seen": 6132608,
      "step": 8040
    },
    {
      "epoch": 16.725571725571726,
      "grad_norm": 0.000998622621409595,
      "learning_rate": 0.2710471391518697,
      "loss": 0.2702,
      "num_input_tokens_seen": 6136448,
      "step": 8045
    },
    {
      "epoch": 16.735966735966738,
      "grad_norm": 0.00026419892674311996,
      "learning_rate": 0.2710123419330375,
      "loss": 0.264,
      "num_input_tokens_seen": 6140448,
      "step": 8050
    },
    {
      "epoch": 16.746361746361746,
      "grad_norm": 0.0005204705521464348,
      "learning_rate": 0.2709775260525816,
      "loss": 0.2498,
      "num_input_tokens_seen": 6144128,
      "step": 8055
    },
    {
      "epoch": 16.756756756756758,
      "grad_norm": 0.0010125788394361734,
      "learning_rate": 0.27094269151587075,
      "loss": 0.2903,
      "num_input_tokens_seen": 6148000,
      "step": 8060
    },
    {
      "epoch": 16.767151767151766,
      "grad_norm": 0.0008944572764448822,
      "learning_rate": 0.27090783832827703,
      "loss": 0.2695,
      "num_input_tokens_seen": 6151840,
      "step": 8065
    },
    {
      "epoch": 16.777546777546778,
      "grad_norm": 8.207056816900149e-05,
      "learning_rate": 0.2708729664951753,
      "loss": 0.2824,
      "num_input_tokens_seen": 6155392,
      "step": 8070
    },
    {
      "epoch": 16.787941787941786,
      "grad_norm": 0.0005768160335719585,
      "learning_rate": 0.27083807602194304,
      "loss": 0.2624,
      "num_input_tokens_seen": 6159328,
      "step": 8075
    },
    {
      "epoch": 16.7983367983368,
      "grad_norm": 0.0002121137804351747,
      "learning_rate": 0.270803166913961,
      "loss": 0.2763,
      "num_input_tokens_seen": 6163136,
      "step": 8080
    },
    {
      "epoch": 16.80873180873181,
      "grad_norm": 0.0006614199955947697,
      "learning_rate": 0.27076823917661247,
      "loss": 0.263,
      "num_input_tokens_seen": 6167008,
      "step": 8085
    },
    {
      "epoch": 16.81912681912682,
      "grad_norm": 0.0002814335748553276,
      "learning_rate": 0.2707332928152838,
      "loss": 0.268,
      "num_input_tokens_seen": 6170816,
      "step": 8090
    },
    {
      "epoch": 16.82952182952183,
      "grad_norm": 0.00060169497737661,
      "learning_rate": 0.2706983278353641,
      "loss": 0.27,
      "num_input_tokens_seen": 6174656,
      "step": 8095
    },
    {
      "epoch": 16.83991683991684,
      "grad_norm": 5.6001288612606004e-05,
      "learning_rate": 0.27066334424224553,
      "loss": 0.2714,
      "num_input_tokens_seen": 6178432,
      "step": 8100
    },
    {
      "epoch": 16.85031185031185,
      "grad_norm": 0.000641416700091213,
      "learning_rate": 0.27062834204132297,
      "loss": 0.2545,
      "num_input_tokens_seen": 6182080,
      "step": 8105
    },
    {
      "epoch": 16.86070686070686,
      "grad_norm": 0.0007159411325119436,
      "learning_rate": 0.27059332123799407,
      "loss": 0.269,
      "num_input_tokens_seen": 6185984,
      "step": 8110
    },
    {
      "epoch": 16.87110187110187,
      "grad_norm": 0.0010314378887414932,
      "learning_rate": 0.27055828183765956,
      "loss": 0.253,
      "num_input_tokens_seen": 6189760,
      "step": 8115
    },
    {
      "epoch": 16.881496881496883,
      "grad_norm": 6.857199332443997e-05,
      "learning_rate": 0.270523223845723,
      "loss": 0.2312,
      "num_input_tokens_seen": 6193440,
      "step": 8120
    },
    {
      "epoch": 16.89189189189189,
      "grad_norm": 0.00027284023235552013,
      "learning_rate": 0.2704881472675907,
      "loss": 0.2688,
      "num_input_tokens_seen": 6197120,
      "step": 8125
    },
    {
      "epoch": 16.902286902286903,
      "grad_norm": 0.001017877133563161,
      "learning_rate": 0.270453052108672,
      "loss": 0.2583,
      "num_input_tokens_seen": 6201120,
      "step": 8130
    },
    {
      "epoch": 16.91268191268191,
      "grad_norm": 0.0009476360864937305,
      "learning_rate": 0.2704179383743789,
      "loss": 0.2787,
      "num_input_tokens_seen": 6204960,
      "step": 8135
    },
    {
      "epoch": 16.923076923076923,
      "grad_norm": 0.00039831383037380874,
      "learning_rate": 0.27038280607012644,
      "loss": 0.2585,
      "num_input_tokens_seen": 6208768,
      "step": 8140
    },
    {
      "epoch": 16.933471933471935,
      "grad_norm": 0.0005661526811309159,
      "learning_rate": 0.27034765520133247,
      "loss": 0.2789,
      "num_input_tokens_seen": 6212544,
      "step": 8145
    },
    {
      "epoch": 16.943866943866944,
      "grad_norm": 0.00018347745935898274,
      "learning_rate": 0.2703124857734177,
      "loss": 0.2604,
      "num_input_tokens_seen": 6216320,
      "step": 8150
    },
    {
      "epoch": 16.954261954261955,
      "grad_norm": 0.0001623460848350078,
      "learning_rate": 0.27027729779180565,
      "loss": 0.2784,
      "num_input_tokens_seen": 6220064,
      "step": 8155
    },
    {
      "epoch": 16.964656964656964,
      "grad_norm": 0.00019272784993518144,
      "learning_rate": 0.27024209126192283,
      "loss": 0.2705,
      "num_input_tokens_seen": 6223808,
      "step": 8160
    },
    {
      "epoch": 16.975051975051976,
      "grad_norm": 0.001076703192666173,
      "learning_rate": 0.2702068661891984,
      "loss": 0.2757,
      "num_input_tokens_seen": 6227712,
      "step": 8165
    },
    {
      "epoch": 16.985446985446984,
      "grad_norm": 0.0003786480228882283,
      "learning_rate": 0.2701716225790647,
      "loss": 0.2784,
      "num_input_tokens_seen": 6231488,
      "step": 8170
    },
    {
      "epoch": 16.995841995841996,
      "grad_norm": 0.0003053348627872765,
      "learning_rate": 0.27013636043695655,
      "loss": 0.2735,
      "num_input_tokens_seen": 6235296,
      "step": 8175
    },
    {
      "epoch": 17.006237006237008,
      "grad_norm": 8.648289076518267e-05,
      "learning_rate": 0.27010107976831194,
      "loss": 0.2439,
      "num_input_tokens_seen": 6238992,
      "step": 8180
    },
    {
      "epoch": 17.016632016632016,
      "grad_norm": 0.00018706229457166046,
      "learning_rate": 0.2700657805785715,
      "loss": 0.2835,
      "num_input_tokens_seen": 6242928,
      "step": 8185
    },
    {
      "epoch": 17.027027027027028,
      "grad_norm": 0.00018215205636806786,
      "learning_rate": 0.2700304628731789,
      "loss": 0.2769,
      "num_input_tokens_seen": 6246672,
      "step": 8190
    },
    {
      "epoch": 17.037422037422036,
      "grad_norm": 0.00028125691460445523,
      "learning_rate": 0.26999512665758046,
      "loss": 0.284,
      "num_input_tokens_seen": 6250544,
      "step": 8195
    },
    {
      "epoch": 17.04781704781705,
      "grad_norm": 0.000305652356473729,
      "learning_rate": 0.2699597719372256,
      "loss": 0.2718,
      "num_input_tokens_seen": 6254288,
      "step": 8200
    },
    {
      "epoch": 17.04781704781705,
      "eval_loss": 0.2478809356689453,
      "eval_runtime": 13.4489,
      "eval_samples_per_second": 63.649,
      "eval_steps_per_second": 15.912,
      "num_input_tokens_seen": 6254288,
      "step": 8200
    },
    {
      "epoch": 17.058212058212057,
      "grad_norm": 0.00012966342910658568,
      "learning_rate": 0.26992439871756635,
      "loss": 0.2274,
      "num_input_tokens_seen": 6258128,
      "step": 8205
    },
    {
      "epoch": 17.06860706860707,
      "grad_norm": 6.038767969585024e-05,
      "learning_rate": 0.2698890070040578,
      "loss": 0.2561,
      "num_input_tokens_seen": 6262000,
      "step": 8210
    },
    {
      "epoch": 17.07900207900208,
      "grad_norm": 0.0008076019585132599,
      "learning_rate": 0.2698535968021577,
      "loss": 0.3054,
      "num_input_tokens_seen": 6265840,
      "step": 8215
    },
    {
      "epoch": 17.08939708939709,
      "grad_norm": 0.0015476435655727983,
      "learning_rate": 0.26981816811732684,
      "loss": 0.2983,
      "num_input_tokens_seen": 6269744,
      "step": 8220
    },
    {
      "epoch": 17.0997920997921,
      "grad_norm": 0.0005021239048801363,
      "learning_rate": 0.26978272095502875,
      "loss": 0.2605,
      "num_input_tokens_seen": 6273552,
      "step": 8225
    },
    {
      "epoch": 17.11018711018711,
      "grad_norm": 0.0007540836813859642,
      "learning_rate": 0.26974725532072974,
      "loss": 0.3453,
      "num_input_tokens_seen": 6277296,
      "step": 8230
    },
    {
      "epoch": 17.12058212058212,
      "grad_norm": 0.0003067753277719021,
      "learning_rate": 0.26971177121989914,
      "loss": 0.2624,
      "num_input_tokens_seen": 6281168,
      "step": 8235
    },
    {
      "epoch": 17.13097713097713,
      "grad_norm": 0.00023701327154412866,
      "learning_rate": 0.2696762686580091,
      "loss": 0.2911,
      "num_input_tokens_seen": 6285072,
      "step": 8240
    },
    {
      "epoch": 17.14137214137214,
      "grad_norm": 0.0005202029133215547,
      "learning_rate": 0.26964074764053436,
      "loss": 0.2678,
      "num_input_tokens_seen": 6288944,
      "step": 8245
    },
    {
      "epoch": 17.151767151767153,
      "grad_norm": 0.00014841160736978054,
      "learning_rate": 0.2696052081729529,
      "loss": 0.2911,
      "num_input_tokens_seen": 6292816,
      "step": 8250
    },
    {
      "epoch": 17.16216216216216,
      "grad_norm": 0.0014726329827681184,
      "learning_rate": 0.2695696502607453,
      "loss": 0.2773,
      "num_input_tokens_seen": 6296560,
      "step": 8255
    },
    {
      "epoch": 17.172557172557173,
      "grad_norm": 0.00014644402835983783,
      "learning_rate": 0.26953407390939504,
      "loss": 0.2748,
      "num_input_tokens_seen": 6300496,
      "step": 8260
    },
    {
      "epoch": 17.18295218295218,
      "grad_norm": 0.00020524821593426168,
      "learning_rate": 0.26949847912438835,
      "loss": 0.278,
      "num_input_tokens_seen": 6304368,
      "step": 8265
    },
    {
      "epoch": 17.193347193347194,
      "grad_norm": 0.0004535742336884141,
      "learning_rate": 0.26946286591121454,
      "loss": 0.2807,
      "num_input_tokens_seen": 6308176,
      "step": 8270
    },
    {
      "epoch": 17.203742203742205,
      "grad_norm": 0.00041877449257299304,
      "learning_rate": 0.2694272342753655,
      "loss": 0.2715,
      "num_input_tokens_seen": 6311984,
      "step": 8275
    },
    {
      "epoch": 17.214137214137214,
      "grad_norm": 0.00035122406552545726,
      "learning_rate": 0.26939158422233617,
      "loss": 0.2698,
      "num_input_tokens_seen": 6315888,
      "step": 8280
    },
    {
      "epoch": 17.224532224532226,
      "grad_norm": 0.06246982514858246,
      "learning_rate": 0.26935591575762413,
      "loss": 1.5278,
      "num_input_tokens_seen": 6319536,
      "step": 8285
    },
    {
      "epoch": 17.234927234927234,
      "grad_norm": 0.001877827220596373,
      "learning_rate": 0.26932022888672996,
      "loss": 0.268,
      "num_input_tokens_seen": 6323440,
      "step": 8290
    },
    {
      "epoch": 17.245322245322246,
      "grad_norm": 0.000290247262455523,
      "learning_rate": 0.26928452361515703,
      "loss": 0.2952,
      "num_input_tokens_seen": 6327344,
      "step": 8295
    },
    {
      "epoch": 17.255717255717254,
      "grad_norm": 0.0005606574704870582,
      "learning_rate": 0.26924879994841155,
      "loss": 0.2703,
      "num_input_tokens_seen": 6331120,
      "step": 8300
    },
    {
      "epoch": 17.266112266112266,
      "grad_norm": 0.0005547350156120956,
      "learning_rate": 0.2692130578920025,
      "loss": 0.2909,
      "num_input_tokens_seen": 6334832,
      "step": 8305
    },
    {
      "epoch": 17.276507276507278,
      "grad_norm": 0.002565674716606736,
      "learning_rate": 0.26917729745144187,
      "loss": 0.2818,
      "num_input_tokens_seen": 6338544,
      "step": 8310
    },
    {
      "epoch": 17.286902286902286,
      "grad_norm": 0.0018962292233482003,
      "learning_rate": 0.2691415186322443,
      "loss": 0.2926,
      "num_input_tokens_seen": 6342512,
      "step": 8315
    },
    {
      "epoch": 17.2972972972973,
      "grad_norm": 0.0005084316362626851,
      "learning_rate": 0.2691057214399273,
      "loss": 0.2885,
      "num_input_tokens_seen": 6346448,
      "step": 8320
    },
    {
      "epoch": 17.307692307692307,
      "grad_norm": 0.0011925476137548685,
      "learning_rate": 0.2690699058800113,
      "loss": 0.2757,
      "num_input_tokens_seen": 6350352,
      "step": 8325
    },
    {
      "epoch": 17.31808731808732,
      "grad_norm": 0.0004931745934300125,
      "learning_rate": 0.2690340719580194,
      "loss": 0.2636,
      "num_input_tokens_seen": 6354032,
      "step": 8330
    },
    {
      "epoch": 17.328482328482327,
      "grad_norm": 0.00015508159413002431,
      "learning_rate": 0.2689982196794778,
      "loss": 0.2568,
      "num_input_tokens_seen": 6357872,
      "step": 8335
    },
    {
      "epoch": 17.33887733887734,
      "grad_norm": 0.00019498851906973869,
      "learning_rate": 0.2689623490499153,
      "loss": 0.2772,
      "num_input_tokens_seen": 6361648,
      "step": 8340
    },
    {
      "epoch": 17.34927234927235,
      "grad_norm": 0.00045442607370205224,
      "learning_rate": 0.2689264600748636,
      "loss": 0.2547,
      "num_input_tokens_seen": 6365616,
      "step": 8345
    },
    {
      "epoch": 17.35966735966736,
      "grad_norm": 0.0021386456210166216,
      "learning_rate": 0.26889055275985724,
      "loss": 0.2911,
      "num_input_tokens_seen": 6369360,
      "step": 8350
    },
    {
      "epoch": 17.37006237006237,
      "grad_norm": 0.0002435635105939582,
      "learning_rate": 0.2688546271104335,
      "loss": 0.2772,
      "num_input_tokens_seen": 6373264,
      "step": 8355
    },
    {
      "epoch": 17.38045738045738,
      "grad_norm": 0.0006914423429407179,
      "learning_rate": 0.26881868313213275,
      "loss": 0.2872,
      "num_input_tokens_seen": 6376976,
      "step": 8360
    },
    {
      "epoch": 17.39085239085239,
      "grad_norm": 0.0010851302649825811,
      "learning_rate": 0.2687827208304978,
      "loss": 0.2756,
      "num_input_tokens_seen": 6380816,
      "step": 8365
    },
    {
      "epoch": 17.401247401247403,
      "grad_norm": 0.0008882470428943634,
      "learning_rate": 0.26874674021107464,
      "loss": 0.2485,
      "num_input_tokens_seen": 6384624,
      "step": 8370
    },
    {
      "epoch": 17.41164241164241,
      "grad_norm": 0.0010572298197075725,
      "learning_rate": 0.2687107412794118,
      "loss": 0.3055,
      "num_input_tokens_seen": 6388528,
      "step": 8375
    },
    {
      "epoch": 17.422037422037423,
      "grad_norm": 0.0014390433207154274,
      "learning_rate": 0.26867472404106096,
      "loss": 0.2798,
      "num_input_tokens_seen": 6392400,
      "step": 8380
    },
    {
      "epoch": 17.43243243243243,
      "grad_norm": 0.000769271980971098,
      "learning_rate": 0.26863868850157624,
      "loss": 0.2627,
      "num_input_tokens_seen": 6396144,
      "step": 8385
    },
    {
      "epoch": 17.442827442827443,
      "grad_norm": 0.001266435720026493,
      "learning_rate": 0.26860263466651485,
      "loss": 0.2194,
      "num_input_tokens_seen": 6399984,
      "step": 8390
    },
    {
      "epoch": 17.453222453222452,
      "grad_norm": 0.00041078811045736074,
      "learning_rate": 0.26856656254143674,
      "loss": 0.2418,
      "num_input_tokens_seen": 6403696,
      "step": 8395
    },
    {
      "epoch": 17.463617463617464,
      "grad_norm": 0.00017173706146422774,
      "learning_rate": 0.2685304721319047,
      "loss": 0.2789,
      "num_input_tokens_seen": 6407504,
      "step": 8400
    },
    {
      "epoch": 17.463617463617464,
      "eval_loss": 0.2646408677101135,
      "eval_runtime": 13.421,
      "eval_samples_per_second": 63.781,
      "eval_steps_per_second": 15.945,
      "num_input_tokens_seen": 6407504,
      "step": 8400
    },
    {
      "epoch": 17.474012474012476,
      "grad_norm": 0.0004560579836834222,
      "learning_rate": 0.2684943634434843,
      "loss": 0.2693,
      "num_input_tokens_seen": 6411440,
      "step": 8405
    },
    {
      "epoch": 17.484407484407484,
      "grad_norm": 0.0002050929906545207,
      "learning_rate": 0.268458236481744,
      "loss": 0.2705,
      "num_input_tokens_seen": 6415152,
      "step": 8410
    },
    {
      "epoch": 17.494802494802496,
      "grad_norm": 0.0006892663077451289,
      "learning_rate": 0.2684220912522549,
      "loss": 0.2904,
      "num_input_tokens_seen": 6418864,
      "step": 8415
    },
    {
      "epoch": 17.505197505197504,
      "grad_norm": 0.0005261699552647769,
      "learning_rate": 0.2683859277605913,
      "loss": 0.2858,
      "num_input_tokens_seen": 6422768,
      "step": 8420
    },
    {
      "epoch": 17.515592515592516,
      "grad_norm": 0.0010613807244226336,
      "learning_rate": 0.2683497460123298,
      "loss": 0.3047,
      "num_input_tokens_seen": 6426608,
      "step": 8425
    },
    {
      "epoch": 17.525987525987524,
      "grad_norm": 0.0001666166353970766,
      "learning_rate": 0.26831354601305013,
      "loss": 0.2852,
      "num_input_tokens_seen": 6430416,
      "step": 8430
    },
    {
      "epoch": 17.536382536382536,
      "grad_norm": 0.00035167779424227774,
      "learning_rate": 0.26827732776833496,
      "loss": 0.2776,
      "num_input_tokens_seen": 6434224,
      "step": 8435
    },
    {
      "epoch": 17.546777546777548,
      "grad_norm": 0.00012222674558870494,
      "learning_rate": 0.26824109128376944,
      "loss": 0.2765,
      "num_input_tokens_seen": 6437840,
      "step": 8440
    },
    {
      "epoch": 17.557172557172557,
      "grad_norm": 0.001379015389829874,
      "learning_rate": 0.2682048365649417,
      "loss": 0.2897,
      "num_input_tokens_seen": 6441680,
      "step": 8445
    },
    {
      "epoch": 17.56756756756757,
      "grad_norm": 0.001555951894260943,
      "learning_rate": 0.2681685636174428,
      "loss": 0.2629,
      "num_input_tokens_seen": 6445584,
      "step": 8450
    },
    {
      "epoch": 17.577962577962577,
      "grad_norm": 0.0002624621265567839,
      "learning_rate": 0.2681322724468663,
      "loss": 0.2609,
      "num_input_tokens_seen": 6449328,
      "step": 8455
    },
    {
      "epoch": 17.58835758835759,
      "grad_norm": 0.00029405258828774095,
      "learning_rate": 0.2680959630588089,
      "loss": 0.2491,
      "num_input_tokens_seen": 6453008,
      "step": 8460
    },
    {
      "epoch": 17.598752598752597,
      "grad_norm": 0.0002858947846107185,
      "learning_rate": 0.26805963545886985,
      "loss": 0.2783,
      "num_input_tokens_seen": 6456784,
      "step": 8465
    },
    {
      "epoch": 17.60914760914761,
      "grad_norm": 0.000558042956981808,
      "learning_rate": 0.26802328965265143,
      "loss": 0.2625,
      "num_input_tokens_seen": 6460560,
      "step": 8470
    },
    {
      "epoch": 17.61954261954262,
      "grad_norm": 0.00012233134475536644,
      "learning_rate": 0.26798692564575854,
      "loss": 0.2762,
      "num_input_tokens_seen": 6464432,
      "step": 8475
    },
    {
      "epoch": 17.62993762993763,
      "grad_norm": 0.0005754560115747154,
      "learning_rate": 0.26795054344379904,
      "loss": 0.2805,
      "num_input_tokens_seen": 6468208,
      "step": 8480
    },
    {
      "epoch": 17.64033264033264,
      "grad_norm": 0.0006691869348287582,
      "learning_rate": 0.2679141430523835,
      "loss": 0.2718,
      "num_input_tokens_seen": 6471952,
      "step": 8485
    },
    {
      "epoch": 17.65072765072765,
      "grad_norm": 0.0010407933732494712,
      "learning_rate": 0.2678777244771252,
      "loss": 0.276,
      "num_input_tokens_seen": 6475984,
      "step": 8490
    },
    {
      "epoch": 17.66112266112266,
      "grad_norm": 0.0005314350128173828,
      "learning_rate": 0.2678412877236405,
      "loss": 0.2679,
      "num_input_tokens_seen": 6479760,
      "step": 8495
    },
    {
      "epoch": 17.671517671517673,
      "grad_norm": 0.002024328801780939,
      "learning_rate": 0.2678048327975484,
      "loss": 0.2955,
      "num_input_tokens_seen": 6483600,
      "step": 8500
    },
    {
      "epoch": 17.68191268191268,
      "grad_norm": 0.0006551137194037437,
      "learning_rate": 0.2677683597044706,
      "loss": 0.2871,
      "num_input_tokens_seen": 6487408,
      "step": 8505
    },
    {
      "epoch": 17.692307692307693,
      "grad_norm": 0.000241446090512909,
      "learning_rate": 0.2677318684500318,
      "loss": 0.2678,
      "num_input_tokens_seen": 6490960,
      "step": 8510
    },
    {
      "epoch": 17.7027027027027,
      "grad_norm": 0.00017587537877261639,
      "learning_rate": 0.2676953590398593,
      "loss": 0.2662,
      "num_input_tokens_seen": 6494864,
      "step": 8515
    },
    {
      "epoch": 17.713097713097714,
      "grad_norm": 0.0007013955037109554,
      "learning_rate": 0.2676588314795834,
      "loss": 0.2684,
      "num_input_tokens_seen": 6498896,
      "step": 8520
    },
    {
      "epoch": 17.723492723492722,
      "grad_norm": 0.00043093739077448845,
      "learning_rate": 0.26762228577483715,
      "loss": 0.2562,
      "num_input_tokens_seen": 6502640,
      "step": 8525
    },
    {
      "epoch": 17.733887733887734,
      "grad_norm": 0.00014598824782297015,
      "learning_rate": 0.2675857219312563,
      "loss": 0.2764,
      "num_input_tokens_seen": 6506448,
      "step": 8530
    },
    {
      "epoch": 17.744282744282746,
      "grad_norm": 0.0002772712614387274,
      "learning_rate": 0.2675491399544794,
      "loss": 0.2812,
      "num_input_tokens_seen": 6510224,
      "step": 8535
    },
    {
      "epoch": 17.754677754677754,
      "grad_norm": 0.0004319855070207268,
      "learning_rate": 0.2675125398501479,
      "loss": 0.2915,
      "num_input_tokens_seen": 6513936,
      "step": 8540
    },
    {
      "epoch": 17.765072765072766,
      "grad_norm": 0.0006322558037936687,
      "learning_rate": 0.26747592162390604,
      "loss": 0.2231,
      "num_input_tokens_seen": 6517712,
      "step": 8545
    },
    {
      "epoch": 17.775467775467774,
      "grad_norm": 0.0010780047159641981,
      "learning_rate": 0.26743928528140076,
      "loss": 0.2958,
      "num_input_tokens_seen": 6521392,
      "step": 8550
    },
    {
      "epoch": 17.785862785862786,
      "grad_norm": 0.0008319044718518853,
      "learning_rate": 0.26740263082828186,
      "loss": 0.3161,
      "num_input_tokens_seen": 6525200,
      "step": 8555
    },
    {
      "epoch": 17.796257796257795,
      "grad_norm": 0.0001987960422411561,
      "learning_rate": 0.2673659582702019,
      "loss": 0.2785,
      "num_input_tokens_seen": 6529136,
      "step": 8560
    },
    {
      "epoch": 17.806652806652806,
      "grad_norm": 0.0006443793536163867,
      "learning_rate": 0.2673292676128163,
      "loss": 0.2685,
      "num_input_tokens_seen": 6533072,
      "step": 8565
    },
    {
      "epoch": 17.81704781704782,
      "grad_norm": 0.0005758479237556458,
      "learning_rate": 0.2672925588617831,
      "loss": 0.2659,
      "num_input_tokens_seen": 6536848,
      "step": 8570
    },
    {
      "epoch": 17.827442827442827,
      "grad_norm": 0.0006438523414544761,
      "learning_rate": 0.2672558320227634,
      "loss": 0.2873,
      "num_input_tokens_seen": 6540688,
      "step": 8575
    },
    {
      "epoch": 17.83783783783784,
      "grad_norm": 0.0018779652891680598,
      "learning_rate": 0.2672190871014209,
      "loss": 0.2756,
      "num_input_tokens_seen": 6544464,
      "step": 8580
    },
    {
      "epoch": 17.848232848232847,
      "grad_norm": 0.0008292693528346717,
      "learning_rate": 0.267182324103422,
      "loss": 0.2708,
      "num_input_tokens_seen": 6548400,
      "step": 8585
    },
    {
      "epoch": 17.85862785862786,
      "grad_norm": 0.00012196957686683163,
      "learning_rate": 0.2671455430344362,
      "loss": 0.297,
      "num_input_tokens_seen": 6552208,
      "step": 8590
    },
    {
      "epoch": 17.86902286902287,
      "grad_norm": 0.0006085454951971769,
      "learning_rate": 0.2671087439001355,
      "loss": 0.2732,
      "num_input_tokens_seen": 6555984,
      "step": 8595
    },
    {
      "epoch": 17.87941787941788,
      "grad_norm": 0.00011699511378537863,
      "learning_rate": 0.2670719267061948,
      "loss": 0.2784,
      "num_input_tokens_seen": 6559760,
      "step": 8600
    },
    {
      "epoch": 17.87941787941788,
      "eval_loss": 0.25499433279037476,
      "eval_runtime": 13.4542,
      "eval_samples_per_second": 63.623,
      "eval_steps_per_second": 15.906,
      "num_input_tokens_seen": 6559760,
      "step": 8600
    },
    {
      "epoch": 17.88981288981289,
      "grad_norm": 0.0009677985799498856,
      "learning_rate": 0.2670350914582918,
      "loss": 0.2762,
      "num_input_tokens_seen": 6563472,
      "step": 8605
    },
    {
      "epoch": 17.9002079002079,
      "grad_norm": 0.00014068270684219897,
      "learning_rate": 0.26699823816210694,
      "loss": 0.2666,
      "num_input_tokens_seen": 6567216,
      "step": 8610
    },
    {
      "epoch": 17.91060291060291,
      "grad_norm": 0.0009792582131922245,
      "learning_rate": 0.26696136682332344,
      "loss": 0.2785,
      "num_input_tokens_seen": 6571024,
      "step": 8615
    },
    {
      "epoch": 17.92099792099792,
      "grad_norm": 0.0001228635519510135,
      "learning_rate": 0.2669244774476274,
      "loss": 0.2752,
      "num_input_tokens_seen": 6574832,
      "step": 8620
    },
    {
      "epoch": 17.93139293139293,
      "grad_norm": 0.00035585102159529924,
      "learning_rate": 0.2668875700407075,
      "loss": 0.262,
      "num_input_tokens_seen": 6578576,
      "step": 8625
    },
    {
      "epoch": 17.941787941787943,
      "grad_norm": 0.0007766150520183146,
      "learning_rate": 0.26685064460825547,
      "loss": 0.2501,
      "num_input_tokens_seen": 6582512,
      "step": 8630
    },
    {
      "epoch": 17.95218295218295,
      "grad_norm": 0.0016311418730765581,
      "learning_rate": 0.26681370115596553,
      "loss": 0.2986,
      "num_input_tokens_seen": 6586224,
      "step": 8635
    },
    {
      "epoch": 17.962577962577964,
      "grad_norm": 0.0008403054671362042,
      "learning_rate": 0.26677673968953497,
      "loss": 0.2684,
      "num_input_tokens_seen": 6590160,
      "step": 8640
    },
    {
      "epoch": 17.972972972972972,
      "grad_norm": 0.00011512546916492283,
      "learning_rate": 0.2667397602146636,
      "loss": 0.2552,
      "num_input_tokens_seen": 6593936,
      "step": 8645
    },
    {
      "epoch": 17.983367983367984,
      "grad_norm": 5.480998879647814e-05,
      "learning_rate": 0.2667027627370542,
      "loss": 0.2682,
      "num_input_tokens_seen": 6597776,
      "step": 8650
    },
    {
      "epoch": 17.993762993762992,
      "grad_norm": 0.00033503398299217224,
      "learning_rate": 0.26666574726241216,
      "loss": 0.2073,
      "num_input_tokens_seen": 6601584,
      "step": 8655
    },
    {
      "epoch": 18.004158004158004,
      "grad_norm": 0.0008311098208650947,
      "learning_rate": 0.2666287137964458,
      "loss": 0.3243,
      "num_input_tokens_seen": 6605376,
      "step": 8660
    },
    {
      "epoch": 18.014553014553016,
      "grad_norm": 0.0022813421674072742,
      "learning_rate": 0.26659166234486614,
      "loss": 0.2788,
      "num_input_tokens_seen": 6609216,
      "step": 8665
    },
    {
      "epoch": 18.024948024948024,
      "grad_norm": 0.00010725743777584285,
      "learning_rate": 0.2665545929133869,
      "loss": 0.2594,
      "num_input_tokens_seen": 6613024,
      "step": 8670
    },
    {
      "epoch": 18.035343035343036,
      "grad_norm": 0.0017440845258533955,
      "learning_rate": 0.2665175055077248,
      "loss": 0.2974,
      "num_input_tokens_seen": 6616672,
      "step": 8675
    },
    {
      "epoch": 18.045738045738045,
      "grad_norm": 0.00010363061301177368,
      "learning_rate": 0.2664804001335991,
      "loss": 0.2815,
      "num_input_tokens_seen": 6620576,
      "step": 8680
    },
    {
      "epoch": 18.056133056133056,
      "grad_norm": 0.0005967674660496414,
      "learning_rate": 0.26644327679673185,
      "loss": 0.2956,
      "num_input_tokens_seen": 6624416,
      "step": 8685
    },
    {
      "epoch": 18.066528066528065,
      "grad_norm": 0.0006394773372448981,
      "learning_rate": 0.26640613550284803,
      "loss": 0.2895,
      "num_input_tokens_seen": 6628352,
      "step": 8690
    },
    {
      "epoch": 18.076923076923077,
      "grad_norm": 0.0001469774724682793,
      "learning_rate": 0.26636897625767525,
      "loss": 0.2603,
      "num_input_tokens_seen": 6632192,
      "step": 8695
    },
    {
      "epoch": 18.08731808731809,
      "grad_norm": 0.00020608387421816587,
      "learning_rate": 0.266331799066944,
      "loss": 0.2621,
      "num_input_tokens_seen": 6635968,
      "step": 8700
    },
    {
      "epoch": 18.097713097713097,
      "grad_norm": 0.0005856282077729702,
      "learning_rate": 0.2662946039363874,
      "loss": 0.2861,
      "num_input_tokens_seen": 6639616,
      "step": 8705
    },
    {
      "epoch": 18.10810810810811,
      "grad_norm": 0.000237450614804402,
      "learning_rate": 0.2662573908717414,
      "loss": 0.2684,
      "num_input_tokens_seen": 6643392,
      "step": 8710
    },
    {
      "epoch": 18.118503118503117,
      "grad_norm": 0.00040499100578017533,
      "learning_rate": 0.2662201598787447,
      "loss": 0.2808,
      "num_input_tokens_seen": 6647264,
      "step": 8715
    },
    {
      "epoch": 18.12889812889813,
      "grad_norm": 0.0005211139796301723,
      "learning_rate": 0.2661829109631389,
      "loss": 0.2748,
      "num_input_tokens_seen": 6651008,
      "step": 8720
    },
    {
      "epoch": 18.13929313929314,
      "grad_norm": 0.0001843898353399709,
      "learning_rate": 0.26614564413066816,
      "loss": 0.2776,
      "num_input_tokens_seen": 6654944,
      "step": 8725
    },
    {
      "epoch": 18.14968814968815,
      "grad_norm": 0.0003061479073949158,
      "learning_rate": 0.2661083593870795,
      "loss": 0.2611,
      "num_input_tokens_seen": 6658816,
      "step": 8730
    },
    {
      "epoch": 18.16008316008316,
      "grad_norm": 0.0004203488351777196,
      "learning_rate": 0.26607105673812276,
      "loss": 0.2641,
      "num_input_tokens_seen": 6662528,
      "step": 8735
    },
    {
      "epoch": 18.17047817047817,
      "grad_norm": 0.0005340935895219445,
      "learning_rate": 0.2660337361895504,
      "loss": 0.2724,
      "num_input_tokens_seen": 6666336,
      "step": 8740
    },
    {
      "epoch": 18.18087318087318,
      "grad_norm": 0.0003870767541229725,
      "learning_rate": 0.26599639774711775,
      "loss": 0.263,
      "num_input_tokens_seen": 6670272,
      "step": 8745
    },
    {
      "epoch": 18.19126819126819,
      "grad_norm": 0.0005318223848007619,
      "learning_rate": 0.2659590414165829,
      "loss": 0.2776,
      "num_input_tokens_seen": 6674112,
      "step": 8750
    },
    {
      "epoch": 18.2016632016632,
      "grad_norm": 0.000512822880409658,
      "learning_rate": 0.2659216672037066,
      "loss": 0.2741,
      "num_input_tokens_seen": 6677952,
      "step": 8755
    },
    {
      "epoch": 18.212058212058214,
      "grad_norm": 0.0001758085418259725,
      "learning_rate": 0.26588427511425244,
      "loss": 0.2342,
      "num_input_tokens_seen": 6681728,
      "step": 8760
    },
    {
      "epoch": 18.222453222453222,
      "grad_norm": 0.0004862227360717952,
      "learning_rate": 0.26584686515398676,
      "loss": 0.2872,
      "num_input_tokens_seen": 6685504,
      "step": 8765
    },
    {
      "epoch": 18.232848232848234,
      "grad_norm": 0.001122916815802455,
      "learning_rate": 0.2658094373286787,
      "loss": 0.282,
      "num_input_tokens_seen": 6689280,
      "step": 8770
    },
    {
      "epoch": 18.243243243243242,
      "grad_norm": 0.001084034563973546,
      "learning_rate": 0.2657719916441,
      "loss": 0.2795,
      "num_input_tokens_seen": 6693088,
      "step": 8775
    },
    {
      "epoch": 18.253638253638254,
      "grad_norm": 0.0008276728331111372,
      "learning_rate": 0.2657345281060253,
      "loss": 0.2644,
      "num_input_tokens_seen": 6696736,
      "step": 8780
    },
    {
      "epoch": 18.264033264033262,
      "grad_norm": 0.000359440193278715,
      "learning_rate": 0.26569704672023203,
      "loss": 0.2474,
      "num_input_tokens_seen": 6700512,
      "step": 8785
    },
    {
      "epoch": 18.274428274428274,
      "grad_norm": 0.0006516285357065499,
      "learning_rate": 0.26565954749250015,
      "loss": 0.3119,
      "num_input_tokens_seen": 6704288,
      "step": 8790
    },
    {
      "epoch": 18.284823284823286,
      "grad_norm": 0.00048808593419380486,
      "learning_rate": 0.2656220304286126,
      "loss": 0.2913,
      "num_input_tokens_seen": 6708064,
      "step": 8795
    },
    {
      "epoch": 18.295218295218294,
      "grad_norm": 5.839101504534483e-05,
      "learning_rate": 0.265584495534355,
      "loss": 0.2703,
      "num_input_tokens_seen": 6711968,
      "step": 8800
    },
    {
      "epoch": 18.295218295218294,
      "eval_loss": 0.24927042424678802,
      "eval_runtime": 13.4681,
      "eval_samples_per_second": 63.558,
      "eval_steps_per_second": 15.889,
      "num_input_tokens_seen": 6711968,
      "step": 8800
    },
    {
      "epoch": 18.305613305613306,
      "grad_norm": 0.00012995293946005404,
      "learning_rate": 0.2655469428155156,
      "loss": 0.2737,
      "num_input_tokens_seen": 6715776,
      "step": 8805
    },
    {
      "epoch": 18.316008316008315,
      "grad_norm": 0.0002193444815929979,
      "learning_rate": 0.2655093722778856,
      "loss": 0.2988,
      "num_input_tokens_seen": 6719616,
      "step": 8810
    },
    {
      "epoch": 18.326403326403327,
      "grad_norm": 0.0019608838483691216,
      "learning_rate": 0.2654717839272588,
      "loss": 0.2965,
      "num_input_tokens_seen": 6723360,
      "step": 8815
    },
    {
      "epoch": 18.33679833679834,
      "grad_norm": 7.834819552954286e-05,
      "learning_rate": 0.2654341777694318,
      "loss": 0.2791,
      "num_input_tokens_seen": 6727136,
      "step": 8820
    },
    {
      "epoch": 18.347193347193347,
      "grad_norm": 0.0009709492442198098,
      "learning_rate": 0.265396553810204,
      "loss": 0.2687,
      "num_input_tokens_seen": 6730944,
      "step": 8825
    },
    {
      "epoch": 18.35758835758836,
      "grad_norm": 0.000904196291230619,
      "learning_rate": 0.26535891205537737,
      "loss": 0.3192,
      "num_input_tokens_seen": 6734880,
      "step": 8830
    },
    {
      "epoch": 18.367983367983367,
      "grad_norm": 9.113308624364436e-05,
      "learning_rate": 0.26532125251075683,
      "loss": 0.2812,
      "num_input_tokens_seen": 6738688,
      "step": 8835
    },
    {
      "epoch": 18.37837837837838,
      "grad_norm": 0.00037280612741596997,
      "learning_rate": 0.26528357518214996,
      "loss": 0.2732,
      "num_input_tokens_seen": 6742592,
      "step": 8840
    },
    {
      "epoch": 18.388773388773387,
      "grad_norm": 0.0012490348890423775,
      "learning_rate": 0.26524588007536704,
      "loss": 0.2933,
      "num_input_tokens_seen": 6746432,
      "step": 8845
    },
    {
      "epoch": 18.3991683991684,
      "grad_norm": 0.0004500504001043737,
      "learning_rate": 0.26520816719622115,
      "loss": 0.2797,
      "num_input_tokens_seen": 6750240,
      "step": 8850
    },
    {
      "epoch": 18.40956340956341,
      "grad_norm": 0.0007879060576669872,
      "learning_rate": 0.2651704365505281,
      "loss": 0.2724,
      "num_input_tokens_seen": 6754016,
      "step": 8855
    },
    {
      "epoch": 18.41995841995842,
      "grad_norm": 0.0003343670687172562,
      "learning_rate": 0.26513268814410634,
      "loss": 0.264,
      "num_input_tokens_seen": 6757856,
      "step": 8860
    },
    {
      "epoch": 18.43035343035343,
      "grad_norm": 0.00012655530008487403,
      "learning_rate": 0.2650949219827773,
      "loss": 0.2844,
      "num_input_tokens_seen": 6761728,
      "step": 8865
    },
    {
      "epoch": 18.44074844074844,
      "grad_norm": 0.00011438181536504999,
      "learning_rate": 0.26505713807236486,
      "loss": 0.2695,
      "num_input_tokens_seen": 6765600,
      "step": 8870
    },
    {
      "epoch": 18.45114345114345,
      "grad_norm": 0.0012365286238491535,
      "learning_rate": 0.26501933641869585,
      "loss": 0.2434,
      "num_input_tokens_seen": 6769344,
      "step": 8875
    },
    {
      "epoch": 18.46153846153846,
      "grad_norm": 0.00010640593245625496,
      "learning_rate": 0.26498151702759976,
      "loss": 0.2694,
      "num_input_tokens_seen": 6773184,
      "step": 8880
    },
    {
      "epoch": 18.471933471933472,
      "grad_norm": 0.00030520043219439685,
      "learning_rate": 0.2649436799049088,
      "loss": 0.257,
      "num_input_tokens_seen": 6777024,
      "step": 8885
    },
    {
      "epoch": 18.482328482328484,
      "grad_norm": 0.0005573865491896868,
      "learning_rate": 0.2649058250564579,
      "loss": 0.2832,
      "num_input_tokens_seen": 6780832,
      "step": 8890
    },
    {
      "epoch": 18.492723492723492,
      "grad_norm": 0.0004710531502496451,
      "learning_rate": 0.26486795248808476,
      "loss": 0.2699,
      "num_input_tokens_seen": 6784576,
      "step": 8895
    },
    {
      "epoch": 18.503118503118504,
      "grad_norm": 0.00010646675218595192,
      "learning_rate": 0.2648300622056298,
      "loss": 0.2869,
      "num_input_tokens_seen": 6788384,
      "step": 8900
    },
    {
      "epoch": 18.513513513513512,
      "grad_norm": 0.0007506481488235295,
      "learning_rate": 0.2647921542149363,
      "loss": 0.2755,
      "num_input_tokens_seen": 6792128,
      "step": 8905
    },
    {
      "epoch": 18.523908523908524,
      "grad_norm": 0.0007890819106251001,
      "learning_rate": 0.26475422852185,
      "loss": 0.2538,
      "num_input_tokens_seen": 6796128,
      "step": 8910
    },
    {
      "epoch": 18.534303534303533,
      "grad_norm": 0.0008609460201114416,
      "learning_rate": 0.2647162851322196,
      "loss": 0.2943,
      "num_input_tokens_seen": 6799840,
      "step": 8915
    },
    {
      "epoch": 18.544698544698544,
      "grad_norm": 0.0003641176735982299,
      "learning_rate": 0.2646783240518964,
      "loss": 0.2921,
      "num_input_tokens_seen": 6803520,
      "step": 8920
    },
    {
      "epoch": 18.555093555093556,
      "grad_norm": 0.000554008933249861,
      "learning_rate": 0.26464034528673447,
      "loss": 0.2832,
      "num_input_tokens_seen": 6807520,
      "step": 8925
    },
    {
      "epoch": 18.565488565488565,
      "grad_norm": 0.00026493993937037885,
      "learning_rate": 0.26460234884259065,
      "loss": 0.2728,
      "num_input_tokens_seen": 6811328,
      "step": 8930
    },
    {
      "epoch": 18.575883575883577,
      "grad_norm": 0.0005672450643032789,
      "learning_rate": 0.2645643347253245,
      "loss": 0.2624,
      "num_input_tokens_seen": 6815008,
      "step": 8935
    },
    {
      "epoch": 18.586278586278585,
      "grad_norm": 0.0002318425540579483,
      "learning_rate": 0.2645263029407982,
      "loss": 0.2755,
      "num_input_tokens_seen": 6818848,
      "step": 8940
    },
    {
      "epoch": 18.596673596673597,
      "grad_norm": 0.0004432741552591324,
      "learning_rate": 0.2644882534948767,
      "loss": 0.2792,
      "num_input_tokens_seen": 6822688,
      "step": 8945
    },
    {
      "epoch": 18.60706860706861,
      "grad_norm": 0.00048561059520579875,
      "learning_rate": 0.2644501863934278,
      "loss": 0.2747,
      "num_input_tokens_seen": 6826688,
      "step": 8950
    },
    {
      "epoch": 18.617463617463617,
      "grad_norm": 0.000606888032052666,
      "learning_rate": 0.26441210164232193,
      "loss": 0.3011,
      "num_input_tokens_seen": 6830528,
      "step": 8955
    },
    {
      "epoch": 18.62785862785863,
      "grad_norm": 0.0005163914756849408,
      "learning_rate": 0.26437399924743216,
      "loss": 0.2517,
      "num_input_tokens_seen": 6834240,
      "step": 8960
    },
    {
      "epoch": 18.638253638253637,
      "grad_norm": 6.785208097426221e-05,
      "learning_rate": 0.26433587921463436,
      "loss": 0.2817,
      "num_input_tokens_seen": 6838048,
      "step": 8965
    },
    {
      "epoch": 18.64864864864865,
      "grad_norm": 0.0003469529328867793,
      "learning_rate": 0.2642977415498072,
      "loss": 0.2594,
      "num_input_tokens_seen": 6841888,
      "step": 8970
    },
    {
      "epoch": 18.659043659043657,
      "grad_norm": 0.0003062535251956433,
      "learning_rate": 0.26425958625883195,
      "loss": 0.2825,
      "num_input_tokens_seen": 6845600,
      "step": 8975
    },
    {
      "epoch": 18.66943866943867,
      "grad_norm": 0.0006538403104059398,
      "learning_rate": 0.2642214133475926,
      "loss": 0.2804,
      "num_input_tokens_seen": 6849440,
      "step": 8980
    },
    {
      "epoch": 18.67983367983368,
      "grad_norm": 0.0005928523023612797,
      "learning_rate": 0.26418322282197587,
      "loss": 0.2745,
      "num_input_tokens_seen": 6853312,
      "step": 8985
    },
    {
      "epoch": 18.69022869022869,
      "grad_norm": 0.0002444878628011793,
      "learning_rate": 0.2641450146878714,
      "loss": 0.27,
      "num_input_tokens_seen": 6857152,
      "step": 8990
    },
    {
      "epoch": 18.7006237006237,
      "grad_norm": 0.00040578245534561574,
      "learning_rate": 0.26410678895117107,
      "loss": 0.2709,
      "num_input_tokens_seen": 6860992,
      "step": 8995
    },
    {
      "epoch": 18.71101871101871,
      "grad_norm": 0.0005521972198039293,
      "learning_rate": 0.26406854561777,
      "loss": 0.2681,
      "num_input_tokens_seen": 6864736,
      "step": 9000
    },
    {
      "epoch": 18.71101871101871,
      "eval_loss": 0.2508901059627533,
      "eval_runtime": 13.4323,
      "eval_samples_per_second": 63.727,
      "eval_steps_per_second": 15.932,
      "num_input_tokens_seen": 6864736,
      "step": 9000
    },
    {
      "epoch": 18.72141372141372,
      "grad_norm": 0.00047054438618943095,
      "learning_rate": 0.26403028469356576,
      "loss": 0.2499,
      "num_input_tokens_seen": 6868448,
      "step": 9005
    },
    {
      "epoch": 18.731808731808734,
      "grad_norm": 0.0003133145801257342,
      "learning_rate": 0.2639920061844585,
      "loss": 0.2431,
      "num_input_tokens_seen": 6872256,
      "step": 9010
    },
    {
      "epoch": 18.742203742203742,
      "grad_norm": 0.00085276341997087,
      "learning_rate": 0.2639537100963515,
      "loss": 0.2633,
      "num_input_tokens_seen": 6876000,
      "step": 9015
    },
    {
      "epoch": 18.752598752598754,
      "grad_norm": 0.0002503904397599399,
      "learning_rate": 0.26391539643515033,
      "loss": 0.2895,
      "num_input_tokens_seen": 6879744,
      "step": 9020
    },
    {
      "epoch": 18.762993762993762,
      "grad_norm": 0.0019540635403245687,
      "learning_rate": 0.26387706520676346,
      "loss": 0.2955,
      "num_input_tokens_seen": 6883616,
      "step": 9025
    },
    {
      "epoch": 18.773388773388774,
      "grad_norm": 0.0002149842621292919,
      "learning_rate": 0.26383871641710205,
      "loss": 0.2712,
      "num_input_tokens_seen": 6887392,
      "step": 9030
    },
    {
      "epoch": 18.783783783783782,
      "grad_norm": 0.0004719561547972262,
      "learning_rate": 0.26380035007208,
      "loss": 0.2685,
      "num_input_tokens_seen": 6891232,
      "step": 9035
    },
    {
      "epoch": 18.794178794178794,
      "grad_norm": 0.00020663491159211844,
      "learning_rate": 0.26376196617761394,
      "loss": 0.2877,
      "num_input_tokens_seen": 6895136,
      "step": 9040
    },
    {
      "epoch": 18.804573804573806,
      "grad_norm": 0.0012713199248537421,
      "learning_rate": 0.263723564739623,
      "loss": 0.2961,
      "num_input_tokens_seen": 6899040,
      "step": 9045
    },
    {
      "epoch": 18.814968814968815,
      "grad_norm": 0.0002601715677883476,
      "learning_rate": 0.2636851457640293,
      "loss": 0.2586,
      "num_input_tokens_seen": 6902784,
      "step": 9050
    },
    {
      "epoch": 18.825363825363826,
      "grad_norm": 0.00018605576769914478,
      "learning_rate": 0.26364670925675737,
      "loss": 0.2886,
      "num_input_tokens_seen": 6906496,
      "step": 9055
    },
    {
      "epoch": 18.835758835758835,
      "grad_norm": 0.0005511005292646587,
      "learning_rate": 0.2636082552237347,
      "loss": 0.2451,
      "num_input_tokens_seen": 6910464,
      "step": 9060
    },
    {
      "epoch": 18.846153846153847,
      "grad_norm": 0.00011461966641945764,
      "learning_rate": 0.26356978367089146,
      "loss": 0.2569,
      "num_input_tokens_seen": 6914368,
      "step": 9065
    },
    {
      "epoch": 18.856548856548855,
      "grad_norm": 0.0006976470467634499,
      "learning_rate": 0.26353129460416036,
      "loss": 0.2446,
      "num_input_tokens_seen": 6918176,
      "step": 9070
    },
    {
      "epoch": 18.866943866943867,
      "grad_norm": 0.00028620948432944715,
      "learning_rate": 0.2634927880294769,
      "loss": 0.2619,
      "num_input_tokens_seen": 6921920,
      "step": 9075
    },
    {
      "epoch": 18.87733887733888,
      "grad_norm": 0.0001477905607316643,
      "learning_rate": 0.26345426395277927,
      "loss": 0.2735,
      "num_input_tokens_seen": 6925728,
      "step": 9080
    },
    {
      "epoch": 18.887733887733887,
      "grad_norm": 0.00023691196111030877,
      "learning_rate": 0.2634157223800084,
      "loss": 0.2552,
      "num_input_tokens_seen": 6929632,
      "step": 9085
    },
    {
      "epoch": 18.8981288981289,
      "grad_norm": 0.0005094673251733184,
      "learning_rate": 0.26337716331710787,
      "loss": 0.2477,
      "num_input_tokens_seen": 6933472,
      "step": 9090
    },
    {
      "epoch": 18.908523908523907,
      "grad_norm": 0.0001158994491561316,
      "learning_rate": 0.2633385867700239,
      "loss": 0.2712,
      "num_input_tokens_seen": 6937248,
      "step": 9095
    },
    {
      "epoch": 18.91891891891892,
      "grad_norm": 0.000534303777385503,
      "learning_rate": 0.2632999927447056,
      "loss": 0.2652,
      "num_input_tokens_seen": 6940992,
      "step": 9100
    },
    {
      "epoch": 18.929313929313928,
      "grad_norm": 0.00021672413276974112,
      "learning_rate": 0.2632613812471046,
      "loss": 0.273,
      "num_input_tokens_seen": 6944768,
      "step": 9105
    },
    {
      "epoch": 18.93970893970894,
      "grad_norm": 0.00037548094405792654,
      "learning_rate": 0.2632227522831753,
      "loss": 0.2566,
      "num_input_tokens_seen": 6948608,
      "step": 9110
    },
    {
      "epoch": 18.95010395010395,
      "grad_norm": 0.00047370526590384543,
      "learning_rate": 0.26318410585887475,
      "loss": 0.2544,
      "num_input_tokens_seen": 6952576,
      "step": 9115
    },
    {
      "epoch": 18.96049896049896,
      "grad_norm": 7.9651981650386e-05,
      "learning_rate": 0.2631454419801627,
      "loss": 0.2246,
      "num_input_tokens_seen": 6956352,
      "step": 9120
    },
    {
      "epoch": 18.97089397089397,
      "grad_norm": 0.0004938665078952909,
      "learning_rate": 0.2631067606530016,
      "loss": 0.2546,
      "num_input_tokens_seen": 6960288,
      "step": 9125
    },
    {
      "epoch": 18.98128898128898,
      "grad_norm": 0.0008581780130043626,
      "learning_rate": 0.2630680618833567,
      "loss": 0.2666,
      "num_input_tokens_seen": 6964032,
      "step": 9130
    },
    {
      "epoch": 18.991683991683992,
      "grad_norm": 0.0009217361221089959,
      "learning_rate": 0.26302934567719566,
      "loss": 0.2866,
      "num_input_tokens_seen": 6967712,
      "step": 9135
    },
    {
      "epoch": 19.002079002079004,
      "grad_norm": 0.00028948872932232916,
      "learning_rate": 0.2629906120404892,
      "loss": 0.2791,
      "num_input_tokens_seen": 6971504,
      "step": 9140
    },
    {
      "epoch": 19.012474012474012,
      "grad_norm": 0.000636327953543514,
      "learning_rate": 0.26295186097921036,
      "loss": 0.2738,
      "num_input_tokens_seen": 6975248,
      "step": 9145
    },
    {
      "epoch": 19.022869022869024,
      "grad_norm": 0.0005886437138542533,
      "learning_rate": 0.2629130924993351,
      "loss": 0.2748,
      "num_input_tokens_seen": 6978960,
      "step": 9150
    },
    {
      "epoch": 19.033264033264032,
      "grad_norm": 0.0005098135443404317,
      "learning_rate": 0.2628743066068421,
      "loss": 0.2498,
      "num_input_tokens_seen": 6982768,
      "step": 9155
    },
    {
      "epoch": 19.043659043659044,
      "grad_norm": 0.000199946211068891,
      "learning_rate": 0.26283550330771244,
      "loss": 0.2905,
      "num_input_tokens_seen": 6986608,
      "step": 9160
    },
    {
      "epoch": 19.054054054054053,
      "grad_norm": 0.001007182989269495,
      "learning_rate": 0.2627966826079303,
      "loss": 0.2544,
      "num_input_tokens_seen": 6990320,
      "step": 9165
    },
    {
      "epoch": 19.064449064449065,
      "grad_norm": 0.00046881879097782075,
      "learning_rate": 0.26275784451348216,
      "loss": 0.2618,
      "num_input_tokens_seen": 6994128,
      "step": 9170
    },
    {
      "epoch": 19.074844074844076,
      "grad_norm": 0.00011254283162998036,
      "learning_rate": 0.2627189890303574,
      "loss": 0.2752,
      "num_input_tokens_seen": 6997840,
      "step": 9175
    },
    {
      "epoch": 19.085239085239085,
      "grad_norm": 0.0006154330330900848,
      "learning_rate": 0.262680116164548,
      "loss": 0.255,
      "num_input_tokens_seen": 7001712,
      "step": 9180
    },
    {
      "epoch": 19.095634095634097,
      "grad_norm": 0.0003279992379248142,
      "learning_rate": 0.2626412259220487,
      "loss": 0.2617,
      "num_input_tokens_seen": 7005680,
      "step": 9185
    },
    {
      "epoch": 19.106029106029105,
      "grad_norm": 0.0008878106600604951,
      "learning_rate": 0.2626023183088568,
      "loss": 0.2342,
      "num_input_tokens_seen": 7009520,
      "step": 9190
    },
    {
      "epoch": 19.116424116424117,
      "grad_norm": 0.0005408648867160082,
      "learning_rate": 0.26256339333097234,
      "loss": 0.2626,
      "num_input_tokens_seen": 7013232,
      "step": 9195
    },
    {
      "epoch": 19.126819126819125,
      "grad_norm": 0.00024181387561839074,
      "learning_rate": 0.2625244509943981,
      "loss": 0.286,
      "num_input_tokens_seen": 7016944,
      "step": 9200
    },
    {
      "epoch": 19.126819126819125,
      "eval_loss": 0.261119544506073,
      "eval_runtime": 13.4245,
      "eval_samples_per_second": 63.764,
      "eval_steps_per_second": 15.941,
      "num_input_tokens_seen": 7016944,
      "step": 9200
    },
    {
      "epoch": 19.137214137214137,
      "grad_norm": 0.0011105149751529098,
      "learning_rate": 0.2624854913051395,
      "loss": 0.2593,
      "num_input_tokens_seen": 7020912,
      "step": 9205
    },
    {
      "epoch": 19.14760914760915,
      "grad_norm": 0.00030829329625703394,
      "learning_rate": 0.26244651426920446,
      "loss": 0.2514,
      "num_input_tokens_seen": 7024656,
      "step": 9210
    },
    {
      "epoch": 19.158004158004157,
      "grad_norm": 0.0009486337658017874,
      "learning_rate": 0.26240751989260386,
      "loss": 0.2905,
      "num_input_tokens_seen": 7028496,
      "step": 9215
    },
    {
      "epoch": 19.16839916839917,
      "grad_norm": 0.00012095565762137994,
      "learning_rate": 0.2623685081813511,
      "loss": 0.2648,
      "num_input_tokens_seen": 7032368,
      "step": 9220
    },
    {
      "epoch": 19.178794178794178,
      "grad_norm": 0.00048102933214977384,
      "learning_rate": 0.2623294791414623,
      "loss": 0.2724,
      "num_input_tokens_seen": 7036016,
      "step": 9225
    },
    {
      "epoch": 19.18918918918919,
      "grad_norm": 0.0005834060139022768,
      "learning_rate": 0.26229043277895614,
      "loss": 0.276,
      "num_input_tokens_seen": 7039696,
      "step": 9230
    },
    {
      "epoch": 19.1995841995842,
      "grad_norm": 0.0004141333338338882,
      "learning_rate": 0.2622513690998542,
      "loss": 0.2648,
      "num_input_tokens_seen": 7043472,
      "step": 9235
    },
    {
      "epoch": 19.20997920997921,
      "grad_norm": 0.0008386893314309418,
      "learning_rate": 0.26221228811018044,
      "loss": 0.2842,
      "num_input_tokens_seen": 7047280,
      "step": 9240
    },
    {
      "epoch": 19.22037422037422,
      "grad_norm": 0.0005181029555387795,
      "learning_rate": 0.2621731898159617,
      "loss": 0.2813,
      "num_input_tokens_seen": 7051120,
      "step": 9245
    },
    {
      "epoch": 19.23076923076923,
      "grad_norm": 0.00014839498908258975,
      "learning_rate": 0.26213407422322743,
      "loss": 0.2604,
      "num_input_tokens_seen": 7054928,
      "step": 9250
    },
    {
      "epoch": 19.241164241164242,
      "grad_norm": 0.0004908874398097396,
      "learning_rate": 0.2620949413380098,
      "loss": 0.2759,
      "num_input_tokens_seen": 7058608,
      "step": 9255
    },
    {
      "epoch": 19.25155925155925,
      "grad_norm": 0.00044023379450663924,
      "learning_rate": 0.26205579116634353,
      "loss": 0.2544,
      "num_input_tokens_seen": 7062640,
      "step": 9260
    },
    {
      "epoch": 19.261954261954262,
      "grad_norm": 0.0010700172279030085,
      "learning_rate": 0.26201662371426604,
      "loss": 0.2648,
      "num_input_tokens_seen": 7066256,
      "step": 9265
    },
    {
      "epoch": 19.272349272349274,
      "grad_norm": 0.00025858485605567694,
      "learning_rate": 0.2619774389878175,
      "loss": 0.2869,
      "num_input_tokens_seen": 7069872,
      "step": 9270
    },
    {
      "epoch": 19.282744282744282,
      "grad_norm": 0.00018654775340110064,
      "learning_rate": 0.2619382369930407,
      "loss": 0.2719,
      "num_input_tokens_seen": 7073808,
      "step": 9275
    },
    {
      "epoch": 19.293139293139294,
      "grad_norm": 0.00035718854633159935,
      "learning_rate": 0.261899017735981,
      "loss": 0.2781,
      "num_input_tokens_seen": 7077712,
      "step": 9280
    },
    {
      "epoch": 19.303534303534303,
      "grad_norm": 0.00031909672543406487,
      "learning_rate": 0.2618597812226866,
      "loss": 0.2708,
      "num_input_tokens_seen": 7081424,
      "step": 9285
    },
    {
      "epoch": 19.313929313929314,
      "grad_norm": 0.00016799983859527856,
      "learning_rate": 0.2618205274592082,
      "loss": 0.249,
      "num_input_tokens_seen": 7085264,
      "step": 9290
    },
    {
      "epoch": 19.324324324324323,
      "grad_norm": 0.0005648421938531101,
      "learning_rate": 0.2617812564515992,
      "loss": 0.2411,
      "num_input_tokens_seen": 7089136,
      "step": 9295
    },
    {
      "epoch": 19.334719334719335,
      "grad_norm": 0.00031264216522686183,
      "learning_rate": 0.2617419682059158,
      "loss": 0.2935,
      "num_input_tokens_seen": 7093040,
      "step": 9300
    },
    {
      "epoch": 19.345114345114347,
      "grad_norm": 0.0014313278952613473,
      "learning_rate": 0.26170266272821663,
      "loss": 0.2583,
      "num_input_tokens_seen": 7096848,
      "step": 9305
    },
    {
      "epoch": 19.355509355509355,
      "grad_norm": 0.0004390967369545251,
      "learning_rate": 0.26166334002456315,
      "loss": 0.2752,
      "num_input_tokens_seen": 7100592,
      "step": 9310
    },
    {
      "epoch": 19.365904365904367,
      "grad_norm": 0.00035188955371268094,
      "learning_rate": 0.2616240001010194,
      "loss": 0.2691,
      "num_input_tokens_seen": 7104368,
      "step": 9315
    },
    {
      "epoch": 19.376299376299375,
      "grad_norm": 0.00023063593835104257,
      "learning_rate": 0.26158464296365197,
      "loss": 0.2851,
      "num_input_tokens_seen": 7108176,
      "step": 9320
    },
    {
      "epoch": 19.386694386694387,
      "grad_norm": 8.219943265430629e-05,
      "learning_rate": 0.2615452686185304,
      "loss": 0.2777,
      "num_input_tokens_seen": 7112080,
      "step": 9325
    },
    {
      "epoch": 19.397089397089395,
      "grad_norm": 9.776716615306213e-05,
      "learning_rate": 0.26150587707172673,
      "loss": 0.2817,
      "num_input_tokens_seen": 7115952,
      "step": 9330
    },
    {
      "epoch": 19.407484407484407,
      "grad_norm": 0.00042810491868294775,
      "learning_rate": 0.2614664683293154,
      "loss": 0.2818,
      "num_input_tokens_seen": 7119856,
      "step": 9335
    },
    {
      "epoch": 19.41787941787942,
      "grad_norm": 0.0006230822764337063,
      "learning_rate": 0.26142704239737397,
      "loss": 0.2647,
      "num_input_tokens_seen": 7123792,
      "step": 9340
    },
    {
      "epoch": 19.428274428274428,
      "grad_norm": 0.00037900966708548367,
      "learning_rate": 0.26138759928198235,
      "loss": 0.2594,
      "num_input_tokens_seen": 7127568,
      "step": 9345
    },
    {
      "epoch": 19.43866943866944,
      "grad_norm": 0.0005252771079540253,
      "learning_rate": 0.26134813898922304,
      "loss": 0.2884,
      "num_input_tokens_seen": 7131344,
      "step": 9350
    },
    {
      "epoch": 19.449064449064448,
      "grad_norm": 0.001312321750447154,
      "learning_rate": 0.26130866152518145,
      "loss": 0.2711,
      "num_input_tokens_seen": 7135344,
      "step": 9355
    },
    {
      "epoch": 19.45945945945946,
      "grad_norm": 0.0001656927925068885,
      "learning_rate": 0.2612691668959455,
      "loss": 0.2614,
      "num_input_tokens_seen": 7139024,
      "step": 9360
    },
    {
      "epoch": 19.46985446985447,
      "grad_norm": 0.00027781364042311907,
      "learning_rate": 0.2612296551076057,
      "loss": 0.2944,
      "num_input_tokens_seen": 7142864,
      "step": 9365
    },
    {
      "epoch": 19.48024948024948,
      "grad_norm": 0.0007671174826100469,
      "learning_rate": 0.26119012616625525,
      "loss": 0.2647,
      "num_input_tokens_seen": 7146480,
      "step": 9370
    },
    {
      "epoch": 19.490644490644492,
      "grad_norm": 8.600906585343182e-05,
      "learning_rate": 0.26115058007799,
      "loss": 0.2617,
      "num_input_tokens_seen": 7150288,
      "step": 9375
    },
    {
      "epoch": 19.5010395010395,
      "grad_norm": 0.0005850433371961117,
      "learning_rate": 0.26111101684890864,
      "loss": 0.2728,
      "num_input_tokens_seen": 7154192,
      "step": 9380
    },
    {
      "epoch": 19.511434511434512,
      "grad_norm": 0.00041858365875668824,
      "learning_rate": 0.26107143648511205,
      "loss": 0.2783,
      "num_input_tokens_seen": 7158032,
      "step": 9385
    },
    {
      "epoch": 19.52182952182952,
      "grad_norm": 0.00014683879271615297,
      "learning_rate": 0.2610318389927042,
      "loss": 0.2644,
      "num_input_tokens_seen": 7162032,
      "step": 9390
    },
    {
      "epoch": 19.532224532224532,
      "grad_norm": 0.0009800773113965988,
      "learning_rate": 0.26099222437779146,
      "loss": 0.2617,
      "num_input_tokens_seen": 7165744,
      "step": 9395
    },
    {
      "epoch": 19.542619542619544,
      "grad_norm": 0.0006388739566318691,
      "learning_rate": 0.26095259264648285,
      "loss": 0.2682,
      "num_input_tokens_seen": 7169456,
      "step": 9400
    },
    {
      "epoch": 19.542619542619544,
      "eval_loss": 0.2484639585018158,
      "eval_runtime": 13.5127,
      "eval_samples_per_second": 63.348,
      "eval_steps_per_second": 15.837,
      "num_input_tokens_seen": 7169456,
      "step": 9400
    },
    {
      "epoch": 19.553014553014552,
      "grad_norm": 0.00011533289944054559,
      "learning_rate": 0.2609129438048902,
      "loss": 0.2732,
      "num_input_tokens_seen": 7173200,
      "step": 9405
    },
    {
      "epoch": 19.563409563409564,
      "grad_norm": 0.0007667131721973419,
      "learning_rate": 0.2608732778591278,
      "loss": 0.2647,
      "num_input_tokens_seen": 7177072,
      "step": 9410
    },
    {
      "epoch": 19.573804573804573,
      "grad_norm": 0.0008894977509044111,
      "learning_rate": 0.2608335948153126,
      "loss": 0.2836,
      "num_input_tokens_seen": 7180784,
      "step": 9415
    },
    {
      "epoch": 19.584199584199585,
      "grad_norm": 0.0003119780740235001,
      "learning_rate": 0.26079389467956426,
      "loss": 0.2856,
      "num_input_tokens_seen": 7184656,
      "step": 9420
    },
    {
      "epoch": 19.594594594594593,
      "grad_norm": 0.0007998279179446399,
      "learning_rate": 0.26075417745800505,
      "loss": 0.2803,
      "num_input_tokens_seen": 7188528,
      "step": 9425
    },
    {
      "epoch": 19.604989604989605,
      "grad_norm": 0.0005445594433695078,
      "learning_rate": 0.26071444315675985,
      "loss": 0.28,
      "num_input_tokens_seen": 7192528,
      "step": 9430
    },
    {
      "epoch": 19.615384615384617,
      "grad_norm": 0.00012255639012437314,
      "learning_rate": 0.2606746917819562,
      "loss": 0.2746,
      "num_input_tokens_seen": 7196336,
      "step": 9435
    },
    {
      "epoch": 19.625779625779625,
      "grad_norm": 0.0008745240047574043,
      "learning_rate": 0.2606349233397242,
      "loss": 0.2558,
      "num_input_tokens_seen": 7200240,
      "step": 9440
    },
    {
      "epoch": 19.636174636174637,
      "grad_norm": 0.00013615828356705606,
      "learning_rate": 0.26059513783619676,
      "loss": 0.2483,
      "num_input_tokens_seen": 7203984,
      "step": 9445
    },
    {
      "epoch": 19.646569646569645,
      "grad_norm": 0.0005787672125734389,
      "learning_rate": 0.26055533527750924,
      "loss": 0.316,
      "num_input_tokens_seen": 7207792,
      "step": 9450
    },
    {
      "epoch": 19.656964656964657,
      "grad_norm": 6.797454261686653e-05,
      "learning_rate": 0.26051551566979964,
      "loss": 0.2696,
      "num_input_tokens_seen": 7211504,
      "step": 9455
    },
    {
      "epoch": 19.66735966735967,
      "grad_norm": 0.0009281965903937817,
      "learning_rate": 0.26047567901920876,
      "loss": 0.2733,
      "num_input_tokens_seen": 7215408,
      "step": 9460
    },
    {
      "epoch": 19.677754677754677,
      "grad_norm": 0.00013899813347961754,
      "learning_rate": 0.2604358253318798,
      "loss": 0.2603,
      "num_input_tokens_seen": 7219312,
      "step": 9465
    },
    {
      "epoch": 19.68814968814969,
      "grad_norm": 0.00020733648852910846,
      "learning_rate": 0.26039595461395876,
      "loss": 0.2515,
      "num_input_tokens_seen": 7223088,
      "step": 9470
    },
    {
      "epoch": 19.698544698544698,
      "grad_norm": 6.337475497275591e-05,
      "learning_rate": 0.26035606687159424,
      "loss": 0.2115,
      "num_input_tokens_seen": 7226896,
      "step": 9475
    },
    {
      "epoch": 19.70893970893971,
      "grad_norm": 0.00015888236521277577,
      "learning_rate": 0.26031616211093733,
      "loss": 0.2753,
      "num_input_tokens_seen": 7230608,
      "step": 9480
    },
    {
      "epoch": 19.719334719334718,
      "grad_norm": 0.00015440689458046108,
      "learning_rate": 0.26027624033814195,
      "loss": 0.2827,
      "num_input_tokens_seen": 7234544,
      "step": 9485
    },
    {
      "epoch": 19.72972972972973,
      "grad_norm": 0.0013223260175436735,
      "learning_rate": 0.2602363015593645,
      "loss": 0.3068,
      "num_input_tokens_seen": 7238256,
      "step": 9490
    },
    {
      "epoch": 19.74012474012474,
      "grad_norm": 0.0003211788134649396,
      "learning_rate": 0.26019634578076395,
      "loss": 0.2836,
      "num_input_tokens_seen": 7242192,
      "step": 9495
    },
    {
      "epoch": 19.75051975051975,
      "grad_norm": 0.0006311276811175048,
      "learning_rate": 0.26015637300850214,
      "loss": 0.2358,
      "num_input_tokens_seen": 7245968,
      "step": 9500
    },
    {
      "epoch": 19.760914760914762,
      "grad_norm": 0.0005032388726249337,
      "learning_rate": 0.26011638324874325,
      "loss": 0.2884,
      "num_input_tokens_seen": 7249872,
      "step": 9505
    },
    {
      "epoch": 19.77130977130977,
      "grad_norm": 0.0005025675054639578,
      "learning_rate": 0.2600763765076543,
      "loss": 0.2675,
      "num_input_tokens_seen": 7253680,
      "step": 9510
    },
    {
      "epoch": 19.781704781704782,
      "grad_norm": 0.03965955972671509,
      "learning_rate": 0.2600363527914048,
      "loss": 0.4439,
      "num_input_tokens_seen": 7257424,
      "step": 9515
    },
    {
      "epoch": 19.79209979209979,
      "grad_norm": 0.005118821747601032,
      "learning_rate": 0.25999631210616686,
      "loss": 0.2639,
      "num_input_tokens_seen": 7261232,
      "step": 9520
    },
    {
      "epoch": 19.802494802494802,
      "grad_norm": 0.00065595842897892,
      "learning_rate": 0.25995625445811527,
      "loss": 0.2578,
      "num_input_tokens_seen": 7265040,
      "step": 9525
    },
    {
      "epoch": 19.812889812889814,
      "grad_norm": 0.0015273848548531532,
      "learning_rate": 0.2599161798534275,
      "loss": 0.2256,
      "num_input_tokens_seen": 7268784,
      "step": 9530
    },
    {
      "epoch": 19.823284823284823,
      "grad_norm": 0.0006779061513952911,
      "learning_rate": 0.25987608829828346,
      "loss": 0.2673,
      "num_input_tokens_seen": 7272656,
      "step": 9535
    },
    {
      "epoch": 19.833679833679835,
      "grad_norm": 0.0011744765797629952,
      "learning_rate": 0.25983597979886586,
      "loss": 0.2496,
      "num_input_tokens_seen": 7276464,
      "step": 9540
    },
    {
      "epoch": 19.844074844074843,
      "grad_norm": 0.0007340550655499101,
      "learning_rate": 0.2597958543613599,
      "loss": 0.2774,
      "num_input_tokens_seen": 7280240,
      "step": 9545
    },
    {
      "epoch": 19.854469854469855,
      "grad_norm": 0.000658744596876204,
      "learning_rate": 0.25975571199195335,
      "loss": 0.2739,
      "num_input_tokens_seen": 7283952,
      "step": 9550
    },
    {
      "epoch": 19.864864864864863,
      "grad_norm": 0.0026952181942760944,
      "learning_rate": 0.25971555269683677,
      "loss": 0.2896,
      "num_input_tokens_seen": 7287728,
      "step": 9555
    },
    {
      "epoch": 19.875259875259875,
      "grad_norm": 0.04383690282702446,
      "learning_rate": 0.25967537648220324,
      "loss": 0.6043,
      "num_input_tokens_seen": 7291856,
      "step": 9560
    },
    {
      "epoch": 19.885654885654887,
      "grad_norm": 0.0022812476381659508,
      "learning_rate": 0.2596351833542483,
      "loss": 0.3452,
      "num_input_tokens_seen": 7295600,
      "step": 9565
    },
    {
      "epoch": 19.896049896049895,
      "grad_norm": 0.001968629891052842,
      "learning_rate": 0.25959497331917036,
      "loss": 0.3131,
      "num_input_tokens_seen": 7299472,
      "step": 9570
    },
    {
      "epoch": 19.906444906444907,
      "grad_norm": 0.000367986096534878,
      "learning_rate": 0.2595547463831703,
      "loss": 0.2942,
      "num_input_tokens_seen": 7303344,
      "step": 9575
    },
    {
      "epoch": 19.916839916839916,
      "grad_norm": 0.0011965053854510188,
      "learning_rate": 0.25951450255245156,
      "loss": 0.2876,
      "num_input_tokens_seen": 7307152,
      "step": 9580
    },
    {
      "epoch": 19.927234927234927,
      "grad_norm": 0.0003532941045705229,
      "learning_rate": 0.2594742418332203,
      "loss": 0.2766,
      "num_input_tokens_seen": 7311056,
      "step": 9585
    },
    {
      "epoch": 19.93762993762994,
      "grad_norm": 0.0007409484242089093,
      "learning_rate": 0.2594339642316852,
      "loss": 0.2898,
      "num_input_tokens_seen": 7315088,
      "step": 9590
    },
    {
      "epoch": 19.948024948024948,
      "grad_norm": 0.0010140087688341737,
      "learning_rate": 0.2593936697540576,
      "loss": 0.2852,
      "num_input_tokens_seen": 7319024,
      "step": 9595
    },
    {
      "epoch": 19.95841995841996,
      "grad_norm": 0.001010441337712109,
      "learning_rate": 0.2593533584065514,
      "loss": 0.2833,
      "num_input_tokens_seen": 7322736,
      "step": 9600
    },
    {
      "epoch": 19.95841995841996,
      "eval_loss": 0.2534381151199341,
      "eval_runtime": 13.4488,
      "eval_samples_per_second": 63.649,
      "eval_steps_per_second": 15.912,
      "num_input_tokens_seen": 7322736,
      "step": 9600
    },
    {
      "epoch": 19.968814968814968,
      "grad_norm": 0.0007372833206318319,
      "learning_rate": 0.2593130301953831,
      "loss": 0.2804,
      "num_input_tokens_seen": 7326448,
      "step": 9605
    },
    {
      "epoch": 19.97920997920998,
      "grad_norm": 0.0008373813470825553,
      "learning_rate": 0.2592726851267718,
      "loss": 0.2646,
      "num_input_tokens_seen": 7330320,
      "step": 9610
    },
    {
      "epoch": 19.989604989604988,
      "grad_norm": 0.00010396361176390201,
      "learning_rate": 0.2592323232069393,
      "loss": 0.2772,
      "num_input_tokens_seen": 7334192,
      "step": 9615
    },
    {
      "epoch": 20.0,
      "grad_norm": 0.0002812615130096674,
      "learning_rate": 0.25919194444210986,
      "loss": 0.2622,
      "num_input_tokens_seen": 7337920,
      "step": 9620
    },
    {
      "epoch": 20.010395010395012,
      "grad_norm": 0.00020592918735928833,
      "learning_rate": 0.2591515488385103,
      "loss": 0.2429,
      "num_input_tokens_seen": 7341696,
      "step": 9625
    },
    {
      "epoch": 20.02079002079002,
      "grad_norm": 0.0008334507001563907,
      "learning_rate": 0.2591111364023704,
      "loss": 0.3063,
      "num_input_tokens_seen": 7345504,
      "step": 9630
    },
    {
      "epoch": 20.031185031185032,
      "grad_norm": 0.00033460030681453645,
      "learning_rate": 0.259070707139922,
      "loss": 0.2648,
      "num_input_tokens_seen": 7349216,
      "step": 9635
    },
    {
      "epoch": 20.04158004158004,
      "grad_norm": 0.00011835694749606773,
      "learning_rate": 0.25903026105739985,
      "loss": 0.2671,
      "num_input_tokens_seen": 7352960,
      "step": 9640
    },
    {
      "epoch": 20.051975051975052,
      "grad_norm": 0.0007295488030649722,
      "learning_rate": 0.2589897981610413,
      "loss": 0.2547,
      "num_input_tokens_seen": 7356672,
      "step": 9645
    },
    {
      "epoch": 20.06237006237006,
      "grad_norm": 0.0009349461179226637,
      "learning_rate": 0.2589493184570863,
      "loss": 0.2629,
      "num_input_tokens_seen": 7360608,
      "step": 9650
    },
    {
      "epoch": 20.072765072765073,
      "grad_norm": 0.0007338064024224877,
      "learning_rate": 0.25890882195177717,
      "loss": 0.2556,
      "num_input_tokens_seen": 7364512,
      "step": 9655
    },
    {
      "epoch": 20.083160083160084,
      "grad_norm": 0.00016088932170532644,
      "learning_rate": 0.25886830865135907,
      "loss": 0.2756,
      "num_input_tokens_seen": 7368128,
      "step": 9660
    },
    {
      "epoch": 20.093555093555093,
      "grad_norm": 0.0022302772849798203,
      "learning_rate": 0.25882777856207967,
      "loss": 0.2782,
      "num_input_tokens_seen": 7371872,
      "step": 9665
    },
    {
      "epoch": 20.103950103950105,
      "grad_norm": 0.0006010463112033904,
      "learning_rate": 0.2587872316901892,
      "loss": 0.266,
      "num_input_tokens_seen": 7375776,
      "step": 9670
    },
    {
      "epoch": 20.114345114345113,
      "grad_norm": 0.00026487052673473954,
      "learning_rate": 0.25874666804194046,
      "loss": 0.2673,
      "num_input_tokens_seen": 7379552,
      "step": 9675
    },
    {
      "epoch": 20.124740124740125,
      "grad_norm": 0.0009183789370581508,
      "learning_rate": 0.258706087623589,
      "loss": 0.2788,
      "num_input_tokens_seen": 7383328,
      "step": 9680
    },
    {
      "epoch": 20.135135135135137,
      "grad_norm": 0.0006324736168608069,
      "learning_rate": 0.25866549044139264,
      "loss": 0.2627,
      "num_input_tokens_seen": 7387040,
      "step": 9685
    },
    {
      "epoch": 20.145530145530145,
      "grad_norm": 0.0010920120403170586,
      "learning_rate": 0.25862487650161214,
      "loss": 0.2796,
      "num_input_tokens_seen": 7390912,
      "step": 9690
    },
    {
      "epoch": 20.155925155925157,
      "grad_norm": 0.0003527570515871048,
      "learning_rate": 0.2585842458105106,
      "loss": 0.2667,
      "num_input_tokens_seen": 7394720,
      "step": 9695
    },
    {
      "epoch": 20.166320166320165,
      "grad_norm": 0.0004457251343410462,
      "learning_rate": 0.2585435983743538,
      "loss": 0.2679,
      "num_input_tokens_seen": 7398432,
      "step": 9700
    },
    {
      "epoch": 20.176715176715177,
      "grad_norm": 0.00019624625565484166,
      "learning_rate": 0.2585029341994101,
      "loss": 0.2579,
      "num_input_tokens_seen": 7402176,
      "step": 9705
    },
    {
      "epoch": 20.187110187110186,
      "grad_norm": 0.0003071998944506049,
      "learning_rate": 0.2584622532919504,
      "loss": 0.2727,
      "num_input_tokens_seen": 7406016,
      "step": 9710
    },
    {
      "epoch": 20.197505197505198,
      "grad_norm": 0.0002194322441937402,
      "learning_rate": 0.2584215556582482,
      "loss": 0.2724,
      "num_input_tokens_seen": 7409760,
      "step": 9715
    },
    {
      "epoch": 20.20790020790021,
      "grad_norm": 9.876451804302633e-05,
      "learning_rate": 0.25838084130457967,
      "loss": 0.2872,
      "num_input_tokens_seen": 7413568,
      "step": 9720
    },
    {
      "epoch": 20.218295218295218,
      "grad_norm": 0.0002197983703808859,
      "learning_rate": 0.2583401102372234,
      "loss": 0.2761,
      "num_input_tokens_seen": 7417440,
      "step": 9725
    },
    {
      "epoch": 20.22869022869023,
      "grad_norm": 0.00025207240832969546,
      "learning_rate": 0.2582993624624606,
      "loss": 0.2601,
      "num_input_tokens_seen": 7421184,
      "step": 9730
    },
    {
      "epoch": 20.239085239085238,
      "grad_norm": 0.0005256514414213598,
      "learning_rate": 0.25825859798657513,
      "loss": 0.2699,
      "num_input_tokens_seen": 7425024,
      "step": 9735
    },
    {
      "epoch": 20.24948024948025,
      "grad_norm": 0.00022195992642082274,
      "learning_rate": 0.25821781681585343,
      "loss": 0.2753,
      "num_input_tokens_seen": 7428896,
      "step": 9740
    },
    {
      "epoch": 20.25987525987526,
      "grad_norm": 0.0005951013299636543,
      "learning_rate": 0.2581770189565844,
      "loss": 0.2621,
      "num_input_tokens_seen": 7432736,
      "step": 9745
    },
    {
      "epoch": 20.27027027027027,
      "grad_norm": 0.0007667667814530432,
      "learning_rate": 0.25813620441505963,
      "loss": 0.2847,
      "num_input_tokens_seen": 7436576,
      "step": 9750
    },
    {
      "epoch": 20.280665280665282,
      "grad_norm": 0.0010342283640056849,
      "learning_rate": 0.2580953731975732,
      "loss": 0.2742,
      "num_input_tokens_seen": 7440416,
      "step": 9755
    },
    {
      "epoch": 20.29106029106029,
      "grad_norm": 0.0021008229814469814,
      "learning_rate": 0.2580545253104218,
      "loss": 0.2748,
      "num_input_tokens_seen": 7444512,
      "step": 9760
    },
    {
      "epoch": 20.301455301455302,
      "grad_norm": 0.0007322630845010281,
      "learning_rate": 0.2580136607599047,
      "loss": 0.2675,
      "num_input_tokens_seen": 7448384,
      "step": 9765
    },
    {
      "epoch": 20.31185031185031,
      "grad_norm": 0.00025445473147556186,
      "learning_rate": 0.2579727795523238,
      "loss": 0.2903,
      "num_input_tokens_seen": 7452256,
      "step": 9770
    },
    {
      "epoch": 20.322245322245323,
      "grad_norm": 0.00046907784417271614,
      "learning_rate": 0.25793188169398334,
      "loss": 0.2735,
      "num_input_tokens_seen": 7456192,
      "step": 9775
    },
    {
      "epoch": 20.33264033264033,
      "grad_norm": 0.0007322622113861144,
      "learning_rate": 0.25789096719119037,
      "loss": 0.2819,
      "num_input_tokens_seen": 7460000,
      "step": 9780
    },
    {
      "epoch": 20.343035343035343,
      "grad_norm": 0.0002992659283336252,
      "learning_rate": 0.2578500360502544,
      "loss": 0.2549,
      "num_input_tokens_seen": 7463712,
      "step": 9785
    },
    {
      "epoch": 20.353430353430355,
      "grad_norm": 0.000893943477421999,
      "learning_rate": 0.2578090882774876,
      "loss": 0.2948,
      "num_input_tokens_seen": 7467488,
      "step": 9790
    },
    {
      "epoch": 20.363825363825363,
      "grad_norm": 0.0009516054415144026,
      "learning_rate": 0.25776812387920456,
      "loss": 0.2806,
      "num_input_tokens_seen": 7471168,
      "step": 9795
    },
    {
      "epoch": 20.374220374220375,
      "grad_norm": 0.00033791764872148633,
      "learning_rate": 0.2577271428617225,
      "loss": 0.2695,
      "num_input_tokens_seen": 7474848,
      "step": 9800
    },
    {
      "epoch": 20.374220374220375,
      "eval_loss": 0.25159597396850586,
      "eval_runtime": 13.4199,
      "eval_samples_per_second": 63.786,
      "eval_steps_per_second": 15.946,
      "num_input_tokens_seen": 7474848,
      "step": 9800
    },
    {
      "epoch": 20.384615384615383,
      "grad_norm": 8.797948248684406e-05,
      "learning_rate": 0.25768614523136124,
      "loss": 0.2517,
      "num_input_tokens_seen": 7478560,
      "step": 9805
    },
    {
      "epoch": 20.395010395010395,
      "grad_norm": 0.0013992168242111802,
      "learning_rate": 0.25764513099444314,
      "loss": 0.2617,
      "num_input_tokens_seen": 7482400,
      "step": 9810
    },
    {
      "epoch": 20.405405405405407,
      "grad_norm": 0.0007539973012171686,
      "learning_rate": 0.25760410015729307,
      "loss": 0.2801,
      "num_input_tokens_seen": 7486272,
      "step": 9815
    },
    {
      "epoch": 20.415800415800415,
      "grad_norm": 0.0001553182810312137,
      "learning_rate": 0.2575630527262385,
      "loss": 0.2691,
      "num_input_tokens_seen": 7490144,
      "step": 9820
    },
    {
      "epoch": 20.426195426195427,
      "grad_norm": 0.0006334925419650972,
      "learning_rate": 0.25752198870760945,
      "loss": 0.264,
      "num_input_tokens_seen": 7493856,
      "step": 9825
    },
    {
      "epoch": 20.436590436590436,
      "grad_norm": 0.00036544946487993,
      "learning_rate": 0.2574809081077386,
      "loss": 0.2717,
      "num_input_tokens_seen": 7497632,
      "step": 9830
    },
    {
      "epoch": 20.446985446985448,
      "grad_norm": 0.0003197744663339108,
      "learning_rate": 0.257439810932961,
      "loss": 0.2719,
      "num_input_tokens_seen": 7501408,
      "step": 9835
    },
    {
      "epoch": 20.457380457380456,
      "grad_norm": 0.0008498377283103764,
      "learning_rate": 0.2573986971896144,
      "loss": 0.2618,
      "num_input_tokens_seen": 7505344,
      "step": 9840
    },
    {
      "epoch": 20.467775467775468,
      "grad_norm": 0.00033469058689661324,
      "learning_rate": 0.257357566884039,
      "loss": 0.2736,
      "num_input_tokens_seen": 7509376,
      "step": 9845
    },
    {
      "epoch": 20.47817047817048,
      "grad_norm": 0.00034822215093299747,
      "learning_rate": 0.25731642002257765,
      "loss": 0.2763,
      "num_input_tokens_seen": 7513248,
      "step": 9850
    },
    {
      "epoch": 20.488565488565488,
      "grad_norm": 0.0014490610919892788,
      "learning_rate": 0.25727525661157574,
      "loss": 0.2894,
      "num_input_tokens_seen": 7517184,
      "step": 9855
    },
    {
      "epoch": 20.4989604989605,
      "grad_norm": 0.0005810630973428488,
      "learning_rate": 0.2572340766573811,
      "loss": 0.2576,
      "num_input_tokens_seen": 7521056,
      "step": 9860
    },
    {
      "epoch": 20.509355509355508,
      "grad_norm": 0.0001826606021495536,
      "learning_rate": 0.25719288016634434,
      "loss": 0.2497,
      "num_input_tokens_seen": 7524960,
      "step": 9865
    },
    {
      "epoch": 20.51975051975052,
      "grad_norm": 0.0007309364737011492,
      "learning_rate": 0.25715166714481835,
      "loss": 0.2728,
      "num_input_tokens_seen": 7528704,
      "step": 9870
    },
    {
      "epoch": 20.53014553014553,
      "grad_norm": 0.00028674094937741756,
      "learning_rate": 0.2571104375991587,
      "loss": 0.2209,
      "num_input_tokens_seen": 7532576,
      "step": 9875
    },
    {
      "epoch": 20.54054054054054,
      "grad_norm": 0.00024698779452592134,
      "learning_rate": 0.2570691915357236,
      "loss": 0.2693,
      "num_input_tokens_seen": 7536416,
      "step": 9880
    },
    {
      "epoch": 20.550935550935552,
      "grad_norm": 0.00039102367009036243,
      "learning_rate": 0.2570279289608736,
      "loss": 0.272,
      "num_input_tokens_seen": 7540096,
      "step": 9885
    },
    {
      "epoch": 20.56133056133056,
      "grad_norm": 0.00012924833572469652,
      "learning_rate": 0.256986649880972,
      "loss": 0.2837,
      "num_input_tokens_seen": 7543872,
      "step": 9890
    },
    {
      "epoch": 20.571725571725572,
      "grad_norm": 0.0007486614631488919,
      "learning_rate": 0.25694535430238447,
      "loss": 0.2788,
      "num_input_tokens_seen": 7547680,
      "step": 9895
    },
    {
      "epoch": 20.58212058212058,
      "grad_norm": 0.0007160227396525443,
      "learning_rate": 0.25690404223147933,
      "loss": 0.2522,
      "num_input_tokens_seen": 7551424,
      "step": 9900
    },
    {
      "epoch": 20.592515592515593,
      "grad_norm": 0.001348499208688736,
      "learning_rate": 0.2568627136746275,
      "loss": 0.2796,
      "num_input_tokens_seen": 7555232,
      "step": 9905
    },
    {
      "epoch": 20.602910602910605,
      "grad_norm": 0.0009232627926394343,
      "learning_rate": 0.25682136863820226,
      "loss": 0.2903,
      "num_input_tokens_seen": 7559008,
      "step": 9910
    },
    {
      "epoch": 20.613305613305613,
      "grad_norm": 0.0003160377382300794,
      "learning_rate": 0.25678000712857957,
      "loss": 0.2805,
      "num_input_tokens_seen": 7562848,
      "step": 9915
    },
    {
      "epoch": 20.623700623700625,
      "grad_norm": 0.000509598758071661,
      "learning_rate": 0.2567386291521379,
      "loss": 0.2229,
      "num_input_tokens_seen": 7566624,
      "step": 9920
    },
    {
      "epoch": 20.634095634095633,
      "grad_norm": 0.00032215958344750106,
      "learning_rate": 0.2566972347152583,
      "loss": 0.2889,
      "num_input_tokens_seen": 7570464,
      "step": 9925
    },
    {
      "epoch": 20.644490644490645,
      "grad_norm": 0.0004113535978831351,
      "learning_rate": 0.2566558238243242,
      "loss": 0.3072,
      "num_input_tokens_seen": 7574400,
      "step": 9930
    },
    {
      "epoch": 20.654885654885653,
      "grad_norm": 0.00024217707687057555,
      "learning_rate": 0.25661439648572176,
      "loss": 0.2576,
      "num_input_tokens_seen": 7578176,
      "step": 9935
    },
    {
      "epoch": 20.665280665280665,
      "grad_norm": 0.0017230167286470532,
      "learning_rate": 0.25657295270583963,
      "loss": 0.2495,
      "num_input_tokens_seen": 7581888,
      "step": 9940
    },
    {
      "epoch": 20.675675675675677,
      "grad_norm": 0.0003539897734299302,
      "learning_rate": 0.25653149249106894,
      "loss": 0.2644,
      "num_input_tokens_seen": 7585760,
      "step": 9945
    },
    {
      "epoch": 20.686070686070686,
      "grad_norm": 0.000334468757500872,
      "learning_rate": 0.25649001584780323,
      "loss": 0.2343,
      "num_input_tokens_seen": 7589568,
      "step": 9950
    },
    {
      "epoch": 20.696465696465697,
      "grad_norm": 0.0006285231211222708,
      "learning_rate": 0.2564485227824389,
      "loss": 0.2659,
      "num_input_tokens_seen": 7593440,
      "step": 9955
    },
    {
      "epoch": 20.706860706860706,
      "grad_norm": 0.0005435169441625476,
      "learning_rate": 0.25640701330137466,
      "loss": 0.273,
      "num_input_tokens_seen": 7597312,
      "step": 9960
    },
    {
      "epoch": 20.717255717255718,
      "grad_norm": 0.0008654603152535856,
      "learning_rate": 0.2563654874110117,
      "loss": 0.2667,
      "num_input_tokens_seen": 7601152,
      "step": 9965
    },
    {
      "epoch": 20.727650727650726,
      "grad_norm": 0.00017654300609137863,
      "learning_rate": 0.256323945117754,
      "loss": 0.2629,
      "num_input_tokens_seen": 7604832,
      "step": 9970
    },
    {
      "epoch": 20.738045738045738,
      "grad_norm": 0.0002609273069538176,
      "learning_rate": 0.2562823864280078,
      "loss": 0.2307,
      "num_input_tokens_seen": 7608544,
      "step": 9975
    },
    {
      "epoch": 20.74844074844075,
      "grad_norm": 0.00043342800927348435,
      "learning_rate": 0.25624081134818194,
      "loss": 0.2535,
      "num_input_tokens_seen": 7612320,
      "step": 9980
    },
    {
      "epoch": 20.758835758835758,
      "grad_norm": 0.00020509760361164808,
      "learning_rate": 0.2561992198846879,
      "loss": 0.2662,
      "num_input_tokens_seen": 7616096,
      "step": 9985
    },
    {
      "epoch": 20.76923076923077,
      "grad_norm": 0.0001900003699120134,
      "learning_rate": 0.25615761204393955,
      "loss": 0.2493,
      "num_input_tokens_seen": 7619872,
      "step": 9990
    },
    {
      "epoch": 20.77962577962578,
      "grad_norm": 0.000363758736057207,
      "learning_rate": 0.2561159878323534,
      "loss": 0.2442,
      "num_input_tokens_seen": 7623552,
      "step": 9995
    },
    {
      "epoch": 20.79002079002079,
      "grad_norm": 0.0004418054304551333,
      "learning_rate": 0.2560743472563483,
      "loss": 0.2552,
      "num_input_tokens_seen": 7627360,
      "step": 10000
    },
    {
      "epoch": 20.79002079002079,
      "eval_loss": 0.24894678592681885,
      "eval_runtime": 13.4271,
      "eval_samples_per_second": 63.751,
      "eval_steps_per_second": 15.938,
      "num_input_tokens_seen": 7627360,
      "step": 10000
    },
    {
      "epoch": 20.8004158004158,
      "grad_norm": 0.0006407376495189965,
      "learning_rate": 0.25603269032234593,
      "loss": 0.2771,
      "num_input_tokens_seen": 7631232,
      "step": 10005
    },
    {
      "epoch": 20.81081081081081,
      "grad_norm": 0.0007491199648939073,
      "learning_rate": 0.2559910170367702,
      "loss": 0.2777,
      "num_input_tokens_seen": 7635072,
      "step": 10010
    },
    {
      "epoch": 20.821205821205822,
      "grad_norm": 0.0012122910702601075,
      "learning_rate": 0.2559493274060477,
      "loss": 0.2741,
      "num_input_tokens_seen": 7638848,
      "step": 10015
    },
    {
      "epoch": 20.83160083160083,
      "grad_norm": 0.0010487971594557166,
      "learning_rate": 0.2559076214366074,
      "loss": 0.274,
      "num_input_tokens_seen": 7642752,
      "step": 10020
    },
    {
      "epoch": 20.841995841995843,
      "grad_norm": 0.00025001520407386124,
      "learning_rate": 0.25586589913488106,
      "loss": 0.2621,
      "num_input_tokens_seen": 7646528,
      "step": 10025
    },
    {
      "epoch": 20.85239085239085,
      "grad_norm": 0.001138955820351839,
      "learning_rate": 0.2558241605073026,
      "loss": 0.2703,
      "num_input_tokens_seen": 7650368,
      "step": 10030
    },
    {
      "epoch": 20.862785862785863,
      "grad_norm": 0.00033714191522449255,
      "learning_rate": 0.25578240556030873,
      "loss": 0.2699,
      "num_input_tokens_seen": 7654272,
      "step": 10035
    },
    {
      "epoch": 20.873180873180875,
      "grad_norm": 0.0009100576862692833,
      "learning_rate": 0.2557406343003386,
      "loss": 0.2823,
      "num_input_tokens_seen": 7658048,
      "step": 10040
    },
    {
      "epoch": 20.883575883575883,
      "grad_norm": 0.0009798434330150485,
      "learning_rate": 0.25569884673383375,
      "loss": 0.2898,
      "num_input_tokens_seen": 7661984,
      "step": 10045
    },
    {
      "epoch": 20.893970893970895,
      "grad_norm": 0.00022918320610187948,
      "learning_rate": 0.25565704286723856,
      "loss": 0.2781,
      "num_input_tokens_seen": 7665728,
      "step": 10050
    },
    {
      "epoch": 20.904365904365903,
      "grad_norm": 0.0010545308468863368,
      "learning_rate": 0.25561522270699955,
      "loss": 0.2716,
      "num_input_tokens_seen": 7669536,
      "step": 10055
    },
    {
      "epoch": 20.914760914760915,
      "grad_norm": 0.0003725242568179965,
      "learning_rate": 0.25557338625956594,
      "loss": 0.2825,
      "num_input_tokens_seen": 7673344,
      "step": 10060
    },
    {
      "epoch": 20.925155925155924,
      "grad_norm": 0.0005269638495519757,
      "learning_rate": 0.25553153353138947,
      "loss": 0.2602,
      "num_input_tokens_seen": 7677312,
      "step": 10065
    },
    {
      "epoch": 20.935550935550935,
      "grad_norm": 0.0002566588227637112,
      "learning_rate": 0.2554896645289243,
      "loss": 0.276,
      "num_input_tokens_seen": 7681024,
      "step": 10070
    },
    {
      "epoch": 20.945945945945947,
      "grad_norm": 0.0005664086784236133,
      "learning_rate": 0.2554477792586272,
      "loss": 0.2685,
      "num_input_tokens_seen": 7684960,
      "step": 10075
    },
    {
      "epoch": 20.956340956340956,
      "grad_norm": 0.000668698747176677,
      "learning_rate": 0.25540587772695744,
      "loss": 0.2683,
      "num_input_tokens_seen": 7688736,
      "step": 10080
    },
    {
      "epoch": 20.966735966735968,
      "grad_norm": 0.00025064373039640486,
      "learning_rate": 0.2553639599403767,
      "loss": 0.2804,
      "num_input_tokens_seen": 7692704,
      "step": 10085
    },
    {
      "epoch": 20.977130977130976,
      "grad_norm": 0.00024317228235304356,
      "learning_rate": 0.2553220259053493,
      "loss": 0.2681,
      "num_input_tokens_seen": 7696416,
      "step": 10090
    },
    {
      "epoch": 20.987525987525988,
      "grad_norm": 0.0006572874845005572,
      "learning_rate": 0.2552800756283419,
      "loss": 0.2749,
      "num_input_tokens_seen": 7700256,
      "step": 10095
    },
    {
      "epoch": 20.997920997921,
      "grad_norm": 9.075023262994364e-05,
      "learning_rate": 0.25523810911582373,
      "loss": 0.2799,
      "num_input_tokens_seen": 7704096,
      "step": 10100
    },
    {
      "epoch": 21.008316008316008,
      "grad_norm": 0.0001262922160094604,
      "learning_rate": 0.25519612637426675,
      "loss": 0.2768,
      "num_input_tokens_seen": 7707792,
      "step": 10105
    },
    {
      "epoch": 21.01871101871102,
      "grad_norm": 0.0014157682890072465,
      "learning_rate": 0.25515412741014504,
      "loss": 0.2899,
      "num_input_tokens_seen": 7711440,
      "step": 10110
    },
    {
      "epoch": 21.02910602910603,
      "grad_norm": 0.0004617314843926579,
      "learning_rate": 0.2551121122299355,
      "loss": 0.2706,
      "num_input_tokens_seen": 7715472,
      "step": 10115
    },
    {
      "epoch": 21.03950103950104,
      "grad_norm": 0.0007789027877151966,
      "learning_rate": 0.2550700808401173,
      "loss": 0.2766,
      "num_input_tokens_seen": 7719376,
      "step": 10120
    },
    {
      "epoch": 21.04989604989605,
      "grad_norm": 0.00013219054380897433,
      "learning_rate": 0.2550280332471722,
      "loss": 0.2775,
      "num_input_tokens_seen": 7723152,
      "step": 10125
    },
    {
      "epoch": 21.06029106029106,
      "grad_norm": 0.0005019409582018852,
      "learning_rate": 0.2549859694575845,
      "loss": 0.2737,
      "num_input_tokens_seen": 7726960,
      "step": 10130
    },
    {
      "epoch": 21.070686070686072,
      "grad_norm": 0.0005093032377772033,
      "learning_rate": 0.254943889477841,
      "loss": 0.2408,
      "num_input_tokens_seen": 7730704,
      "step": 10135
    },
    {
      "epoch": 21.08108108108108,
      "grad_norm": 0.0010023302165791392,
      "learning_rate": 0.25490179331443097,
      "loss": 0.2966,
      "num_input_tokens_seen": 7734512,
      "step": 10140
    },
    {
      "epoch": 21.091476091476093,
      "grad_norm": 0.0007999064400792122,
      "learning_rate": 0.25485968097384615,
      "loss": 0.3068,
      "num_input_tokens_seen": 7738288,
      "step": 10145
    },
    {
      "epoch": 21.1018711018711,
      "grad_norm": 0.0008813382592052221,
      "learning_rate": 0.25481755246258075,
      "loss": 0.2654,
      "num_input_tokens_seen": 7742032,
      "step": 10150
    },
    {
      "epoch": 21.112266112266113,
      "grad_norm": 0.0009257479105144739,
      "learning_rate": 0.2547754077871315,
      "loss": 0.261,
      "num_input_tokens_seen": 7745936,
      "step": 10155
    },
    {
      "epoch": 21.12266112266112,
      "grad_norm": 0.000368928856914863,
      "learning_rate": 0.25473324695399774,
      "loss": 0.2379,
      "num_input_tokens_seen": 7749712,
      "step": 10160
    },
    {
      "epoch": 21.133056133056133,
      "grad_norm": 0.0005834809853695333,
      "learning_rate": 0.25469106996968105,
      "loss": 0.2869,
      "num_input_tokens_seen": 7753616,
      "step": 10165
    },
    {
      "epoch": 21.143451143451145,
      "grad_norm": 0.0004701579746324569,
      "learning_rate": 0.2546488768406858,
      "loss": 0.2059,
      "num_input_tokens_seen": 7757392,
      "step": 10170
    },
    {
      "epoch": 21.153846153846153,
      "grad_norm": 0.0005686631775461137,
      "learning_rate": 0.25460666757351863,
      "loss": 0.2578,
      "num_input_tokens_seen": 7761168,
      "step": 10175
    },
    {
      "epoch": 21.164241164241165,
      "grad_norm": 0.00017975068476516753,
      "learning_rate": 0.25456444217468877,
      "loss": 0.2558,
      "num_input_tokens_seen": 7765008,
      "step": 10180
    },
    {
      "epoch": 21.174636174636174,
      "grad_norm": 0.0004805494681932032,
      "learning_rate": 0.25452220065070785,
      "loss": 0.2751,
      "num_input_tokens_seen": 7768784,
      "step": 10185
    },
    {
      "epoch": 21.185031185031185,
      "grad_norm": 0.00020410734578035772,
      "learning_rate": 0.2544799430080901,
      "loss": 0.2718,
      "num_input_tokens_seen": 7772496,
      "step": 10190
    },
    {
      "epoch": 21.195426195426194,
      "grad_norm": 0.00018421626009512693,
      "learning_rate": 0.2544376692533522,
      "loss": 0.2871,
      "num_input_tokens_seen": 7776208,
      "step": 10195
    },
    {
      "epoch": 21.205821205821206,
      "grad_norm": 0.0006492980755865574,
      "learning_rate": 0.2543953793930132,
      "loss": 0.2899,
      "num_input_tokens_seen": 7779952,
      "step": 10200
    },
    {
      "epoch": 21.205821205821206,
      "eval_loss": 0.28288534283638,
      "eval_runtime": 13.409,
      "eval_samples_per_second": 63.838,
      "eval_steps_per_second": 15.959,
      "num_input_tokens_seen": 7779952,
      "step": 10200
    },
    {
      "epoch": 21.216216216216218,
      "grad_norm": 0.0006508665392175317,
      "learning_rate": 0.2543530734335948,
      "loss": 0.273,
      "num_input_tokens_seen": 7783856,
      "step": 10205
    },
    {
      "epoch": 21.226611226611226,
      "grad_norm": 0.0001785898202797398,
      "learning_rate": 0.2543107513816211,
      "loss": 0.2718,
      "num_input_tokens_seen": 7787728,
      "step": 10210
    },
    {
      "epoch": 21.237006237006238,
      "grad_norm": 6.725174171151593e-05,
      "learning_rate": 0.25426841324361865,
      "loss": 0.2359,
      "num_input_tokens_seen": 7791664,
      "step": 10215
    },
    {
      "epoch": 21.247401247401246,
      "grad_norm": 0.00042148795910179615,
      "learning_rate": 0.2542260590261166,
      "loss": 0.2726,
      "num_input_tokens_seen": 7795472,
      "step": 10220
    },
    {
      "epoch": 21.257796257796258,
      "grad_norm": 0.00023086188593879342,
      "learning_rate": 0.2541836887356465,
      "loss": 0.273,
      "num_input_tokens_seen": 7799280,
      "step": 10225
    },
    {
      "epoch": 21.26819126819127,
      "grad_norm": 0.00020786735694855452,
      "learning_rate": 0.2541413023787423,
      "loss": 0.2751,
      "num_input_tokens_seen": 7802992,
      "step": 10230
    },
    {
      "epoch": 21.27858627858628,
      "grad_norm": 0.0010975359473377466,
      "learning_rate": 0.2540988999619405,
      "loss": 0.2831,
      "num_input_tokens_seen": 7806832,
      "step": 10235
    },
    {
      "epoch": 21.28898128898129,
      "grad_norm": 0.00010360888700233772,
      "learning_rate": 0.25405648149178023,
      "loss": 0.2701,
      "num_input_tokens_seen": 7810544,
      "step": 10240
    },
    {
      "epoch": 21.2993762993763,
      "grad_norm": 0.00020738496095873415,
      "learning_rate": 0.2540140469748028,
      "loss": 0.268,
      "num_input_tokens_seen": 7814192,
      "step": 10245
    },
    {
      "epoch": 21.30977130977131,
      "grad_norm": 0.0003965446085203439,
      "learning_rate": 0.25397159641755224,
      "loss": 0.2814,
      "num_input_tokens_seen": 7818064,
      "step": 10250
    },
    {
      "epoch": 21.32016632016632,
      "grad_norm": 0.00012908197822980583,
      "learning_rate": 0.2539291298265749,
      "loss": 0.2769,
      "num_input_tokens_seen": 7821904,
      "step": 10255
    },
    {
      "epoch": 21.33056133056133,
      "grad_norm": 0.0003077381697949022,
      "learning_rate": 0.2538866472084197,
      "loss": 0.2785,
      "num_input_tokens_seen": 7825616,
      "step": 10260
    },
    {
      "epoch": 21.340956340956343,
      "grad_norm": 0.000493488390929997,
      "learning_rate": 0.25384414856963794,
      "loss": 0.2562,
      "num_input_tokens_seen": 7829360,
      "step": 10265
    },
    {
      "epoch": 21.35135135135135,
      "grad_norm": 0.00016070547280833125,
      "learning_rate": 0.25380163391678356,
      "loss": 0.2738,
      "num_input_tokens_seen": 7833200,
      "step": 10270
    },
    {
      "epoch": 21.361746361746363,
      "grad_norm": 0.0006622174405492842,
      "learning_rate": 0.2537591032564127,
      "loss": 0.2475,
      "num_input_tokens_seen": 7837040,
      "step": 10275
    },
    {
      "epoch": 21.37214137214137,
      "grad_norm": 0.0009966210927814245,
      "learning_rate": 0.25371655659508424,
      "loss": 0.2871,
      "num_input_tokens_seen": 7840720,
      "step": 10280
    },
    {
      "epoch": 21.382536382536383,
      "grad_norm": 0.0007196349324658513,
      "learning_rate": 0.25367399393935935,
      "loss": 0.2812,
      "num_input_tokens_seen": 7844496,
      "step": 10285
    },
    {
      "epoch": 21.39293139293139,
      "grad_norm": 0.0010075728641822934,
      "learning_rate": 0.25363141529580174,
      "loss": 0.2871,
      "num_input_tokens_seen": 7848624,
      "step": 10290
    },
    {
      "epoch": 21.403326403326403,
      "grad_norm": 0.0006293815677054226,
      "learning_rate": 0.2535888206709776,
      "loss": 0.2491,
      "num_input_tokens_seen": 7852304,
      "step": 10295
    },
    {
      "epoch": 21.413721413721415,
      "grad_norm": 0.000508844677824527,
      "learning_rate": 0.2535462100714555,
      "loss": 0.2878,
      "num_input_tokens_seen": 7856208,
      "step": 10300
    },
    {
      "epoch": 21.424116424116423,
      "grad_norm": 0.000340888713253662,
      "learning_rate": 0.2535035835038066,
      "loss": 0.2555,
      "num_input_tokens_seen": 7859856,
      "step": 10305
    },
    {
      "epoch": 21.434511434511435,
      "grad_norm": 0.00020129710901528597,
      "learning_rate": 0.2534609409746044,
      "loss": 0.2759,
      "num_input_tokens_seen": 7863664,
      "step": 10310
    },
    {
      "epoch": 21.444906444906444,
      "grad_norm": 0.0003353781357873231,
      "learning_rate": 0.253418282490425,
      "loss": 0.2764,
      "num_input_tokens_seen": 7867440,
      "step": 10315
    },
    {
      "epoch": 21.455301455301456,
      "grad_norm": 0.0006508993683382869,
      "learning_rate": 0.2533756080578467,
      "loss": 0.2649,
      "num_input_tokens_seen": 7871280,
      "step": 10320
    },
    {
      "epoch": 21.465696465696467,
      "grad_norm": 0.00023910297022666782,
      "learning_rate": 0.25333291768345056,
      "loss": 0.2258,
      "num_input_tokens_seen": 7875344,
      "step": 10325
    },
    {
      "epoch": 21.476091476091476,
      "grad_norm": 5.987012991681695e-05,
      "learning_rate": 0.25329021137381996,
      "loss": 0.2548,
      "num_input_tokens_seen": 7879312,
      "step": 10330
    },
    {
      "epoch": 21.486486486486488,
      "grad_norm": 0.0001869567931862548,
      "learning_rate": 0.25324748913554074,
      "loss": 0.2932,
      "num_input_tokens_seen": 7883152,
      "step": 10335
    },
    {
      "epoch": 21.496881496881496,
      "grad_norm": 0.000450716121122241,
      "learning_rate": 0.2532047509752013,
      "loss": 0.2614,
      "num_input_tokens_seen": 7887056,
      "step": 10340
    },
    {
      "epoch": 21.507276507276508,
      "grad_norm": 0.0002124433813150972,
      "learning_rate": 0.25316199689939217,
      "loss": 0.2742,
      "num_input_tokens_seen": 7891056,
      "step": 10345
    },
    {
      "epoch": 21.517671517671516,
      "grad_norm": 0.0001818554155761376,
      "learning_rate": 0.2531192269147068,
      "loss": 0.2602,
      "num_input_tokens_seen": 7894928,
      "step": 10350
    },
    {
      "epoch": 21.528066528066528,
      "grad_norm": 0.00017955267685465515,
      "learning_rate": 0.2530764410277407,
      "loss": 0.2612,
      "num_input_tokens_seen": 7898768,
      "step": 10355
    },
    {
      "epoch": 21.53846153846154,
      "grad_norm": 0.0003220679936930537,
      "learning_rate": 0.25303363924509203,
      "loss": 0.2595,
      "num_input_tokens_seen": 7902576,
      "step": 10360
    },
    {
      "epoch": 21.54885654885655,
      "grad_norm": 0.0006659325445070863,
      "learning_rate": 0.25299082157336145,
      "loss": 0.2495,
      "num_input_tokens_seen": 7906480,
      "step": 10365
    },
    {
      "epoch": 21.55925155925156,
      "grad_norm": 0.00011034584167646244,
      "learning_rate": 0.2529479880191519,
      "loss": 0.273,
      "num_input_tokens_seen": 7910320,
      "step": 10370
    },
    {
      "epoch": 21.56964656964657,
      "grad_norm": 0.00046427754568867385,
      "learning_rate": 0.2529051385890689,
      "loss": 0.28,
      "num_input_tokens_seen": 7914000,
      "step": 10375
    },
    {
      "epoch": 21.58004158004158,
      "grad_norm": 0.0007128691649995744,
      "learning_rate": 0.2528622732897203,
      "loss": 0.2708,
      "num_input_tokens_seen": 7917840,
      "step": 10380
    },
    {
      "epoch": 21.59043659043659,
      "grad_norm": 0.00014065760478843004,
      "learning_rate": 0.25281939212771654,
      "loss": 0.2712,
      "num_input_tokens_seen": 7921648,
      "step": 10385
    },
    {
      "epoch": 21.6008316008316,
      "grad_norm": 7.154179911594838e-05,
      "learning_rate": 0.2527764951096704,
      "loss": 0.2709,
      "num_input_tokens_seen": 7925456,
      "step": 10390
    },
    {
      "epoch": 21.611226611226613,
      "grad_norm": 0.0003487508511170745,
      "learning_rate": 0.2527335822421971,
      "loss": 0.2784,
      "num_input_tokens_seen": 7929168,
      "step": 10395
    },
    {
      "epoch": 21.62162162162162,
      "grad_norm": 0.0002465861034579575,
      "learning_rate": 0.25269065353191444,
      "loss": 0.2748,
      "num_input_tokens_seen": 7932848,
      "step": 10400
    },
    {
      "epoch": 21.62162162162162,
      "eval_loss": 0.2521704435348511,
      "eval_runtime": 13.4291,
      "eval_samples_per_second": 63.742,
      "eval_steps_per_second": 15.936,
      "num_input_tokens_seen": 7932848,
      "step": 10400
    },
    {
      "epoch": 21.632016632016633,
      "grad_norm": 0.0005314700538292527,
      "learning_rate": 0.2526477089854425,
      "loss": 0.2804,
      "num_input_tokens_seen": 7936816,
      "step": 10405
    },
    {
      "epoch": 21.64241164241164,
      "grad_norm": 0.00019947731925640255,
      "learning_rate": 0.25260474860940385,
      "loss": 0.2572,
      "num_input_tokens_seen": 7940656,
      "step": 10410
    },
    {
      "epoch": 21.652806652806653,
      "grad_norm": 0.0004532934108283371,
      "learning_rate": 0.2525617724104236,
      "loss": 0.295,
      "num_input_tokens_seen": 7944592,
      "step": 10415
    },
    {
      "epoch": 21.66320166320166,
      "grad_norm": 0.0006383038125932217,
      "learning_rate": 0.25251878039512915,
      "loss": 0.2543,
      "num_input_tokens_seen": 7948432,
      "step": 10420
    },
    {
      "epoch": 21.673596673596673,
      "grad_norm": 0.0005806440021842718,
      "learning_rate": 0.25247577257015047,
      "loss": 0.2594,
      "num_input_tokens_seen": 7952112,
      "step": 10425
    },
    {
      "epoch": 21.683991683991685,
      "grad_norm": 0.00012706445704679936,
      "learning_rate": 0.2524327489421198,
      "loss": 0.2646,
      "num_input_tokens_seen": 7955920,
      "step": 10430
    },
    {
      "epoch": 21.694386694386694,
      "grad_norm": 0.0002419205120531842,
      "learning_rate": 0.25238970951767203,
      "loss": 0.252,
      "num_input_tokens_seen": 7959664,
      "step": 10435
    },
    {
      "epoch": 21.704781704781706,
      "grad_norm": 0.00037064144271425903,
      "learning_rate": 0.25234665430344433,
      "loss": 0.2793,
      "num_input_tokens_seen": 7963472,
      "step": 10440
    },
    {
      "epoch": 21.715176715176714,
      "grad_norm": 0.0004733344539999962,
      "learning_rate": 0.2523035833060764,
      "loss": 0.2702,
      "num_input_tokens_seen": 7967184,
      "step": 10445
    },
    {
      "epoch": 21.725571725571726,
      "grad_norm": 0.00035818744800053537,
      "learning_rate": 0.2522604965322103,
      "loss": 0.274,
      "num_input_tokens_seen": 7971152,
      "step": 10450
    },
    {
      "epoch": 21.735966735966738,
      "grad_norm": 0.0007519974024035037,
      "learning_rate": 0.25221739398849047,
      "loss": 0.2924,
      "num_input_tokens_seen": 7974896,
      "step": 10455
    },
    {
      "epoch": 21.746361746361746,
      "grad_norm": 0.001055193948559463,
      "learning_rate": 0.252174275681564,
      "loss": 0.279,
      "num_input_tokens_seen": 7978640,
      "step": 10460
    },
    {
      "epoch": 21.756756756756758,
      "grad_norm": 0.0007141061942093074,
      "learning_rate": 0.2521311416180802,
      "loss": 0.2514,
      "num_input_tokens_seen": 7982288,
      "step": 10465
    },
    {
      "epoch": 21.767151767151766,
      "grad_norm": 0.0005833365139551461,
      "learning_rate": 0.25208799180469094,
      "loss": 0.2757,
      "num_input_tokens_seen": 7986128,
      "step": 10470
    },
    {
      "epoch": 21.777546777546778,
      "grad_norm": 0.000380914454581216,
      "learning_rate": 0.2520448262480504,
      "loss": 0.2948,
      "num_input_tokens_seen": 7989968,
      "step": 10475
    },
    {
      "epoch": 21.787941787941786,
      "grad_norm": 0.0007166217546910048,
      "learning_rate": 0.25200164495481525,
      "loss": 0.2682,
      "num_input_tokens_seen": 7993776,
      "step": 10480
    },
    {
      "epoch": 21.7983367983368,
      "grad_norm": 0.00014241099415812641,
      "learning_rate": 0.25195844793164474,
      "loss": 0.2781,
      "num_input_tokens_seen": 7997648,
      "step": 10485
    },
    {
      "epoch": 21.80873180873181,
      "grad_norm": 8.604626054875553e-05,
      "learning_rate": 0.2519152351852001,
      "loss": 0.2536,
      "num_input_tokens_seen": 8001232,
      "step": 10490
    },
    {
      "epoch": 21.81912681912682,
      "grad_norm": 0.0009744454291649163,
      "learning_rate": 0.25187200672214555,
      "loss": 0.2761,
      "num_input_tokens_seen": 8005136,
      "step": 10495
    },
    {
      "epoch": 21.82952182952183,
      "grad_norm": 0.0005549754132516682,
      "learning_rate": 0.2518287625491473,
      "loss": 0.2852,
      "num_input_tokens_seen": 8008912,
      "step": 10500
    },
    {
      "epoch": 21.83991683991684,
      "grad_norm": 0.0008176462142728269,
      "learning_rate": 0.25178550267287425,
      "loss": 0.2812,
      "num_input_tokens_seen": 8012816,
      "step": 10505
    },
    {
      "epoch": 21.85031185031185,
      "grad_norm": 0.00022608082508668303,
      "learning_rate": 0.2517422270999976,
      "loss": 0.2662,
      "num_input_tokens_seen": 8016752,
      "step": 10510
    },
    {
      "epoch": 21.86070686070686,
      "grad_norm": 0.0005152887897565961,
      "learning_rate": 0.2516989358371909,
      "loss": 0.2868,
      "num_input_tokens_seen": 8020656,
      "step": 10515
    },
    {
      "epoch": 21.87110187110187,
      "grad_norm": 0.000178591872099787,
      "learning_rate": 0.25165562889113025,
      "loss": 0.2719,
      "num_input_tokens_seen": 8024304,
      "step": 10520
    },
    {
      "epoch": 21.881496881496883,
      "grad_norm": 0.0011281645856797695,
      "learning_rate": 0.2516123062684942,
      "loss": 0.2791,
      "num_input_tokens_seen": 8028208,
      "step": 10525
    },
    {
      "epoch": 21.89189189189189,
      "grad_norm": 9.733244951348752e-05,
      "learning_rate": 0.25156896797596356,
      "loss": 0.2685,
      "num_input_tokens_seen": 8032112,
      "step": 10530
    },
    {
      "epoch": 21.902286902286903,
      "grad_norm": 0.00015054376854095608,
      "learning_rate": 0.2515256140202216,
      "loss": 0.2576,
      "num_input_tokens_seen": 8035824,
      "step": 10535
    },
    {
      "epoch": 21.91268191268191,
      "grad_norm": 0.000295458099571988,
      "learning_rate": 0.25148224440795425,
      "loss": 0.2895,
      "num_input_tokens_seen": 8039696,
      "step": 10540
    },
    {
      "epoch": 21.923076923076923,
      "grad_norm": 0.0014142782893031836,
      "learning_rate": 0.2514388591458494,
      "loss": 0.2574,
      "num_input_tokens_seen": 8043440,
      "step": 10545
    },
    {
      "epoch": 21.933471933471935,
      "grad_norm": 0.0007092607556842268,
      "learning_rate": 0.2513954582405977,
      "loss": 0.2599,
      "num_input_tokens_seen": 8047248,
      "step": 10550
    },
    {
      "epoch": 21.943866943866944,
      "grad_norm": 0.00017317199672106653,
      "learning_rate": 0.2513520416988922,
      "loss": 0.2618,
      "num_input_tokens_seen": 8051152,
      "step": 10555
    },
    {
      "epoch": 21.954261954261955,
      "grad_norm": 0.0002898134116549045,
      "learning_rate": 0.2513086095274281,
      "loss": 0.2821,
      "num_input_tokens_seen": 8054992,
      "step": 10560
    },
    {
      "epoch": 21.964656964656964,
      "grad_norm": 0.0005926968879066408,
      "learning_rate": 0.25126516173290336,
      "loss": 0.2752,
      "num_input_tokens_seen": 8058832,
      "step": 10565
    },
    {
      "epoch": 21.975051975051976,
      "grad_norm": 9.112121188081801e-05,
      "learning_rate": 0.2512216983220181,
      "loss": 0.229,
      "num_input_tokens_seen": 8062608,
      "step": 10570
    },
    {
      "epoch": 21.985446985446984,
      "grad_norm": 0.0007004551007412374,
      "learning_rate": 0.25117821930147494,
      "loss": 0.2907,
      "num_input_tokens_seen": 8066352,
      "step": 10575
    },
    {
      "epoch": 21.995841995841996,
      "grad_norm": 8.978394907899201e-05,
      "learning_rate": 0.2511347246779788,
      "loss": 0.2694,
      "num_input_tokens_seen": 8070096,
      "step": 10580
    },
    {
      "epoch": 22.006237006237008,
      "grad_norm": 0.0003931598912458867,
      "learning_rate": 0.25109121445823723,
      "loss": 0.2648,
      "num_input_tokens_seen": 8073960,
      "step": 10585
    },
    {
      "epoch": 22.016632016632016,
      "grad_norm": 0.0006780650583095849,
      "learning_rate": 0.25104768864896004,
      "loss": 0.2781,
      "num_input_tokens_seen": 8077736,
      "step": 10590
    },
    {
      "epoch": 22.027027027027028,
      "grad_norm": 0.0006108471425250173,
      "learning_rate": 0.2510041472568594,
      "loss": 0.2637,
      "num_input_tokens_seen": 8081416,
      "step": 10595
    },
    {
      "epoch": 22.037422037422036,
      "grad_norm": 0.00024073536042124033,
      "learning_rate": 0.25096059028864987,
      "loss": 0.2799,
      "num_input_tokens_seen": 8085448,
      "step": 10600
    },
    {
      "epoch": 22.037422037422036,
      "eval_loss": 0.2572762966156006,
      "eval_runtime": 13.4248,
      "eval_samples_per_second": 63.763,
      "eval_steps_per_second": 15.941,
      "num_input_tokens_seen": 8085448,
      "step": 10600
    },
    {
      "epoch": 22.04781704781705,
      "grad_norm": 0.0017821715446189046,
      "learning_rate": 0.25091701775104863,
      "loss": 0.2837,
      "num_input_tokens_seen": 8089320,
      "step": 10605
    },
    {
      "epoch": 22.058212058212057,
      "grad_norm": 0.0003902337048202753,
      "learning_rate": 0.250873429650775,
      "loss": 0.2756,
      "num_input_tokens_seen": 8092904,
      "step": 10610
    },
    {
      "epoch": 22.06860706860707,
      "grad_norm": 0.0001144780617323704,
      "learning_rate": 0.25082982599455095,
      "loss": 0.2529,
      "num_input_tokens_seen": 8096680,
      "step": 10615
    },
    {
      "epoch": 22.07900207900208,
      "grad_norm": 0.00048730819253250957,
      "learning_rate": 0.2507862067891006,
      "loss": 0.2651,
      "num_input_tokens_seen": 8100488,
      "step": 10620
    },
    {
      "epoch": 22.08939708939709,
      "grad_norm": 9.896356641547754e-05,
      "learning_rate": 0.25074257204115064,
      "loss": 0.286,
      "num_input_tokens_seen": 8104456,
      "step": 10625
    },
    {
      "epoch": 22.0997920997921,
      "grad_norm": 0.0005414117476902902,
      "learning_rate": 0.25069892175742997,
      "loss": 0.2675,
      "num_input_tokens_seen": 8108168,
      "step": 10630
    },
    {
      "epoch": 22.11018711018711,
      "grad_norm": 0.0002482913841959089,
      "learning_rate": 0.25065525594467014,
      "loss": 0.2635,
      "num_input_tokens_seen": 8111912,
      "step": 10635
    },
    {
      "epoch": 22.12058212058212,
      "grad_norm": 0.00011848725989693776,
      "learning_rate": 0.2506115746096049,
      "loss": 0.2229,
      "num_input_tokens_seen": 8115848,
      "step": 10640
    },
    {
      "epoch": 22.13097713097713,
      "grad_norm": 0.0006396223325282335,
      "learning_rate": 0.25056787775897055,
      "loss": 0.3073,
      "num_input_tokens_seen": 8119592,
      "step": 10645
    },
    {
      "epoch": 22.14137214137214,
      "grad_norm": 0.00043715719948522747,
      "learning_rate": 0.2505241653995056,
      "loss": 0.2772,
      "num_input_tokens_seen": 8123336,
      "step": 10650
    },
    {
      "epoch": 22.151767151767153,
      "grad_norm": 0.0002397741045570001,
      "learning_rate": 0.25048043753795113,
      "loss": 0.2833,
      "num_input_tokens_seen": 8127016,
      "step": 10655
    },
    {
      "epoch": 22.16216216216216,
      "grad_norm": 0.0006872433004900813,
      "learning_rate": 0.2504366941810504,
      "loss": 0.2638,
      "num_input_tokens_seen": 8130824,
      "step": 10660
    },
    {
      "epoch": 22.172557172557173,
      "grad_norm": 0.0006273009348660707,
      "learning_rate": 0.2503929353355493,
      "loss": 0.2949,
      "num_input_tokens_seen": 8134536,
      "step": 10665
    },
    {
      "epoch": 22.18295218295218,
      "grad_norm": 0.0009745433926582336,
      "learning_rate": 0.250349161008196,
      "loss": 0.2546,
      "num_input_tokens_seen": 8138344,
      "step": 10670
    },
    {
      "epoch": 22.193347193347194,
      "grad_norm": 0.00014149340859148651,
      "learning_rate": 0.2503053712057409,
      "loss": 0.2646,
      "num_input_tokens_seen": 8142152,
      "step": 10675
    },
    {
      "epoch": 22.203742203742205,
      "grad_norm": 0.00045154045801609755,
      "learning_rate": 0.25026156593493715,
      "loss": 0.2546,
      "num_input_tokens_seen": 8145960,
      "step": 10680
    },
    {
      "epoch": 22.214137214137214,
      "grad_norm": 9.074409899767488e-05,
      "learning_rate": 0.2502177452025399,
      "loss": 0.2938,
      "num_input_tokens_seen": 8149768,
      "step": 10685
    },
    {
      "epoch": 22.224532224532226,
      "grad_norm": 0.00021942477906122804,
      "learning_rate": 0.25017390901530695,
      "loss": 0.2605,
      "num_input_tokens_seen": 8153544,
      "step": 10690
    },
    {
      "epoch": 22.234927234927234,
      "grad_norm": 0.00031422381289303303,
      "learning_rate": 0.2501300573799984,
      "loss": 0.2528,
      "num_input_tokens_seen": 8157288,
      "step": 10695
    },
    {
      "epoch": 22.245322245322246,
      "grad_norm": 0.0004385676875244826,
      "learning_rate": 0.2500861903033766,
      "loss": 0.2441,
      "num_input_tokens_seen": 8160936,
      "step": 10700
    },
    {
      "epoch": 22.255717255717254,
      "grad_norm": 0.0007493385346606374,
      "learning_rate": 0.25004230779220654,
      "loss": 0.2647,
      "num_input_tokens_seen": 8164776,
      "step": 10705
    },
    {
      "epoch": 22.266112266112266,
      "grad_norm": 0.0003376492240931839,
      "learning_rate": 0.24999840985325542,
      "loss": 0.2726,
      "num_input_tokens_seen": 8168584,
      "step": 10710
    },
    {
      "epoch": 22.276507276507278,
      "grad_norm": 0.0004350741219241172,
      "learning_rate": 0.24995449649329285,
      "loss": 0.2726,
      "num_input_tokens_seen": 8172360,
      "step": 10715
    },
    {
      "epoch": 22.286902286902286,
      "grad_norm": 0.000273155135801062,
      "learning_rate": 0.2499105677190908,
      "loss": 0.2665,
      "num_input_tokens_seen": 8176264,
      "step": 10720
    },
    {
      "epoch": 22.2972972972973,
      "grad_norm": 0.0009184036171063781,
      "learning_rate": 0.24986662353742364,
      "loss": 0.2738,
      "num_input_tokens_seen": 8179944,
      "step": 10725
    },
    {
      "epoch": 22.307692307692307,
      "grad_norm": 0.000325720408000052,
      "learning_rate": 0.24982266395506814,
      "loss": 0.2811,
      "num_input_tokens_seen": 8183752,
      "step": 10730
    },
    {
      "epoch": 22.31808731808732,
      "grad_norm": 0.00035290457890369,
      "learning_rate": 0.2497786889788034,
      "loss": 0.2788,
      "num_input_tokens_seen": 8187720,
      "step": 10735
    },
    {
      "epoch": 22.328482328482327,
      "grad_norm": 0.00017647678032517433,
      "learning_rate": 0.24973469861541095,
      "loss": 0.28,
      "num_input_tokens_seen": 8191624,
      "step": 10740
    },
    {
      "epoch": 22.33887733887734,
      "grad_norm": 0.0009805572917684913,
      "learning_rate": 0.24969069287167456,
      "loss": 0.2778,
      "num_input_tokens_seen": 8195368,
      "step": 10745
    },
    {
      "epoch": 22.34927234927235,
      "grad_norm": 0.0004555070190690458,
      "learning_rate": 0.2496466717543806,
      "loss": 0.2788,
      "num_input_tokens_seen": 8199240,
      "step": 10750
    },
    {
      "epoch": 22.35966735966736,
      "grad_norm": 0.00021970391389913857,
      "learning_rate": 0.24960263527031762,
      "loss": 0.2623,
      "num_input_tokens_seen": 8203048,
      "step": 10755
    },
    {
      "epoch": 22.37006237006237,
      "grad_norm": 0.0008720891200937331,
      "learning_rate": 0.24955858342627657,
      "loss": 0.2662,
      "num_input_tokens_seen": 8206760,
      "step": 10760
    },
    {
      "epoch": 22.38045738045738,
      "grad_norm": 0.00029956205980852246,
      "learning_rate": 0.24951451622905083,
      "loss": 0.2703,
      "num_input_tokens_seen": 8210632,
      "step": 10765
    },
    {
      "epoch": 22.39085239085239,
      "grad_norm": 0.00047952140448614955,
      "learning_rate": 0.24947043368543612,
      "loss": 0.2759,
      "num_input_tokens_seen": 8214472,
      "step": 10770
    },
    {
      "epoch": 22.401247401247403,
      "grad_norm": 0.0010888060787692666,
      "learning_rate": 0.2494263358022305,
      "loss": 0.2806,
      "num_input_tokens_seen": 8218440,
      "step": 10775
    },
    {
      "epoch": 22.41164241164241,
      "grad_norm": 0.00041226629400625825,
      "learning_rate": 0.24938222258623444,
      "loss": 0.3006,
      "num_input_tokens_seen": 8222312,
      "step": 10780
    },
    {
      "epoch": 22.422037422037423,
      "grad_norm": 0.0010342691093683243,
      "learning_rate": 0.24933809404425075,
      "loss": 0.2554,
      "num_input_tokens_seen": 8226152,
      "step": 10785
    },
    {
      "epoch": 22.43243243243243,
      "grad_norm": 0.0007386417128145695,
      "learning_rate": 0.24929395018308453,
      "loss": 0.2692,
      "num_input_tokens_seen": 8230024,
      "step": 10790
    },
    {
      "epoch": 22.442827442827443,
      "grad_norm": 0.00018341877148486674,
      "learning_rate": 0.24924979100954348,
      "loss": 0.292,
      "num_input_tokens_seen": 8233992,
      "step": 10795
    },
    {
      "epoch": 22.453222453222452,
      "grad_norm": 0.0004787735524587333,
      "learning_rate": 0.24920561653043735,
      "loss": 0.2724,
      "num_input_tokens_seen": 8237768,
      "step": 10800
    },
    {
      "epoch": 22.453222453222452,
      "eval_loss": 0.2622421383857727,
      "eval_runtime": 13.4181,
      "eval_samples_per_second": 63.794,
      "eval_steps_per_second": 15.949,
      "num_input_tokens_seen": 8237768,
      "step": 10800
    },
    {
      "epoch": 22.463617463617464,
      "grad_norm": 0.0007335016271099448,
      "learning_rate": 0.24916142675257846,
      "loss": 0.2642,
      "num_input_tokens_seen": 8241608,
      "step": 10805
    },
    {
      "epoch": 22.474012474012476,
      "grad_norm": 0.0007495825411751866,
      "learning_rate": 0.24911722168278144,
      "loss": 0.2505,
      "num_input_tokens_seen": 8245512,
      "step": 10810
    },
    {
      "epoch": 22.484407484407484,
      "grad_norm": 0.0004769162624143064,
      "learning_rate": 0.24907300132786328,
      "loss": 0.2689,
      "num_input_tokens_seen": 8249480,
      "step": 10815
    },
    {
      "epoch": 22.494802494802496,
      "grad_norm": 0.0006044330657459795,
      "learning_rate": 0.24902876569464322,
      "loss": 0.2618,
      "num_input_tokens_seen": 8253352,
      "step": 10820
    },
    {
      "epoch": 22.505197505197504,
      "grad_norm": 0.00044620202970691025,
      "learning_rate": 0.24898451478994305,
      "loss": 0.2788,
      "num_input_tokens_seen": 8257096,
      "step": 10825
    },
    {
      "epoch": 22.515592515592516,
      "grad_norm": 0.0002337045007152483,
      "learning_rate": 0.2489402486205868,
      "loss": 0.2742,
      "num_input_tokens_seen": 8261000,
      "step": 10830
    },
    {
      "epoch": 22.525987525987524,
      "grad_norm": 0.00014708917296957225,
      "learning_rate": 0.24889596719340085,
      "loss": 0.2469,
      "num_input_tokens_seen": 8264744,
      "step": 10835
    },
    {
      "epoch": 22.536382536382536,
      "grad_norm": 0.0004472598375286907,
      "learning_rate": 0.24885167051521392,
      "loss": 0.2702,
      "num_input_tokens_seen": 8268648,
      "step": 10840
    },
    {
      "epoch": 22.546777546777548,
      "grad_norm": 0.0010834215208888054,
      "learning_rate": 0.24880735859285716,
      "loss": 0.2689,
      "num_input_tokens_seen": 8272264,
      "step": 10845
    },
    {
      "epoch": 22.557172557172557,
      "grad_norm": 0.0001862489734776318,
      "learning_rate": 0.24876303143316406,
      "loss": 0.2574,
      "num_input_tokens_seen": 8276104,
      "step": 10850
    },
    {
      "epoch": 22.56756756756757,
      "grad_norm": 0.0005557636031880975,
      "learning_rate": 0.24871868904297031,
      "loss": 0.2988,
      "num_input_tokens_seen": 8279912,
      "step": 10855
    },
    {
      "epoch": 22.577962577962577,
      "grad_norm": 0.00017783230578061193,
      "learning_rate": 0.24867433142911416,
      "loss": 0.2831,
      "num_input_tokens_seen": 8283784,
      "step": 10860
    },
    {
      "epoch": 22.58835758835759,
      "grad_norm": 0.0013421468902379274,
      "learning_rate": 0.24862995859843612,
      "loss": 0.2909,
      "num_input_tokens_seen": 8287496,
      "step": 10865
    },
    {
      "epoch": 22.598752598752597,
      "grad_norm": 0.00035398625186644495,
      "learning_rate": 0.24858557055777897,
      "loss": 0.2671,
      "num_input_tokens_seen": 8291272,
      "step": 10870
    },
    {
      "epoch": 22.60914760914761,
      "grad_norm": 0.0004928258713334799,
      "learning_rate": 0.24854116731398793,
      "loss": 0.2533,
      "num_input_tokens_seen": 8294984,
      "step": 10875
    },
    {
      "epoch": 22.61954261954262,
      "grad_norm": 8.907296432880685e-05,
      "learning_rate": 0.24849674887391052,
      "loss": 0.2273,
      "num_input_tokens_seen": 8298760,
      "step": 10880
    },
    {
      "epoch": 22.62993762993763,
      "grad_norm": 0.00015555019490420818,
      "learning_rate": 0.2484523152443967,
      "loss": 0.2996,
      "num_input_tokens_seen": 8302600,
      "step": 10885
    },
    {
      "epoch": 22.64033264033264,
      "grad_norm": 0.0002797841443680227,
      "learning_rate": 0.24840786643229862,
      "loss": 0.2746,
      "num_input_tokens_seen": 8306248,
      "step": 10890
    },
    {
      "epoch": 22.65072765072765,
      "grad_norm": 5.0095448386855423e-05,
      "learning_rate": 0.2483634024444709,
      "loss": 0.2634,
      "num_input_tokens_seen": 8310088,
      "step": 10895
    },
    {
      "epoch": 22.66112266112266,
      "grad_norm": 0.0008946954039856791,
      "learning_rate": 0.24831892328777033,
      "loss": 0.2885,
      "num_input_tokens_seen": 8314056,
      "step": 10900
    },
    {
      "epoch": 22.671517671517673,
      "grad_norm": 0.0001729989453451708,
      "learning_rate": 0.2482744289690563,
      "loss": 0.2724,
      "num_input_tokens_seen": 8317832,
      "step": 10905
    },
    {
      "epoch": 22.68191268191268,
      "grad_norm": 0.0004232304636389017,
      "learning_rate": 0.2482299194951903,
      "loss": 0.2625,
      "num_input_tokens_seen": 8321576,
      "step": 10910
    },
    {
      "epoch": 22.692307692307693,
      "grad_norm": 0.00023504953423980623,
      "learning_rate": 0.2481853948730363,
      "loss": 0.281,
      "num_input_tokens_seen": 8325448,
      "step": 10915
    },
    {
      "epoch": 22.7027027027027,
      "grad_norm": 0.0006898775463923812,
      "learning_rate": 0.24814085510946052,
      "loss": 0.2712,
      "num_input_tokens_seen": 8329192,
      "step": 10920
    },
    {
      "epoch": 22.713097713097714,
      "grad_norm": 0.0007428527460433543,
      "learning_rate": 0.24809630021133158,
      "loss": 0.2682,
      "num_input_tokens_seen": 8333128,
      "step": 10925
    },
    {
      "epoch": 22.723492723492722,
      "grad_norm": 0.0003109718963969499,
      "learning_rate": 0.24805173018552037,
      "loss": 0.2618,
      "num_input_tokens_seen": 8336968,
      "step": 10930
    },
    {
      "epoch": 22.733887733887734,
      "grad_norm": 0.0007977247005328536,
      "learning_rate": 0.2480071450389002,
      "loss": 0.2667,
      "num_input_tokens_seen": 8340712,
      "step": 10935
    },
    {
      "epoch": 22.744282744282746,
      "grad_norm": 0.00029371693381108344,
      "learning_rate": 0.24796254477834662,
      "loss": 0.273,
      "num_input_tokens_seen": 8344424,
      "step": 10940
    },
    {
      "epoch": 22.754677754677754,
      "grad_norm": 0.0003132218844257295,
      "learning_rate": 0.24791792941073754,
      "loss": 0.2712,
      "num_input_tokens_seen": 8348360,
      "step": 10945
    },
    {
      "epoch": 22.765072765072766,
      "grad_norm": 0.0008319218759424984,
      "learning_rate": 0.2478732989429533,
      "loss": 0.2616,
      "num_input_tokens_seen": 8352232,
      "step": 10950
    },
    {
      "epoch": 22.775467775467774,
      "grad_norm": 0.0007291169022209942,
      "learning_rate": 0.24782865338187632,
      "loss": 0.2632,
      "num_input_tokens_seen": 8356136,
      "step": 10955
    },
    {
      "epoch": 22.785862785862786,
      "grad_norm": 0.0003768869210034609,
      "learning_rate": 0.2477839927343916,
      "loss": 0.3096,
      "num_input_tokens_seen": 8359912,
      "step": 10960
    },
    {
      "epoch": 22.796257796257795,
      "grad_norm": 0.0004930169088765979,
      "learning_rate": 0.2477393170073864,
      "loss": 0.2733,
      "num_input_tokens_seen": 8363656,
      "step": 10965
    },
    {
      "epoch": 22.806652806652806,
      "grad_norm": 0.0010685250163078308,
      "learning_rate": 0.2476946262077503,
      "loss": 0.272,
      "num_input_tokens_seen": 8367592,
      "step": 10970
    },
    {
      "epoch": 22.81704781704782,
      "grad_norm": 0.0009760446846485138,
      "learning_rate": 0.24764992034237507,
      "loss": 0.2965,
      "num_input_tokens_seen": 8371464,
      "step": 10975
    },
    {
      "epoch": 22.827442827442827,
      "grad_norm": 0.00044547097058966756,
      "learning_rate": 0.24760519941815498,
      "loss": 0.2619,
      "num_input_tokens_seen": 8375208,
      "step": 10980
    },
    {
      "epoch": 22.83783783783784,
      "grad_norm": 0.0008346649119630456,
      "learning_rate": 0.2475604634419866,
      "loss": 0.2865,
      "num_input_tokens_seen": 8379080,
      "step": 10985
    },
    {
      "epoch": 22.848232848232847,
      "grad_norm": 0.0001985155831789598,
      "learning_rate": 0.24751571242076872,
      "loss": 0.2763,
      "num_input_tokens_seen": 8382856,
      "step": 10990
    },
    {
      "epoch": 22.85862785862786,
      "grad_norm": 0.00023423205129802227,
      "learning_rate": 0.2474709463614025,
      "loss": 0.2828,
      "num_input_tokens_seen": 8386920,
      "step": 10995
    },
    {
      "epoch": 22.86902286902287,
      "grad_norm": 0.00041810987750068307,
      "learning_rate": 0.24742616527079145,
      "loss": 0.2279,
      "num_input_tokens_seen": 8390664,
      "step": 11000
    },
    {
      "epoch": 22.86902286902287,
      "eval_loss": 0.2504565417766571,
      "eval_runtime": 13.4057,
      "eval_samples_per_second": 63.854,
      "eval_steps_per_second": 15.963,
      "num_input_tokens_seen": 8390664,
      "step": 11000
    },
    {
      "epoch": 22.87941787941788,
      "grad_norm": 0.0005937424139119685,
      "learning_rate": 0.24738136915584139,
      "loss": 0.3025,
      "num_input_tokens_seen": 8394504,
      "step": 11005
    },
    {
      "epoch": 22.88981288981289,
      "grad_norm": 0.000998524483293295,
      "learning_rate": 0.24733655802346047,
      "loss": 0.2674,
      "num_input_tokens_seen": 8398312,
      "step": 11010
    },
    {
      "epoch": 22.9002079002079,
      "grad_norm": 0.00020585885795298964,
      "learning_rate": 0.24729173188055906,
      "loss": 0.2725,
      "num_input_tokens_seen": 8402088,
      "step": 11015
    },
    {
      "epoch": 22.91060291060291,
      "grad_norm": 0.0009814698714762926,
      "learning_rate": 0.24724689073404996,
      "loss": 0.2724,
      "num_input_tokens_seen": 8405928,
      "step": 11020
    },
    {
      "epoch": 22.92099792099792,
      "grad_norm": 0.0003191218711435795,
      "learning_rate": 0.24720203459084822,
      "loss": 0.2721,
      "num_input_tokens_seen": 8409576,
      "step": 11025
    },
    {
      "epoch": 22.93139293139293,
      "grad_norm": 0.00011392906890250742,
      "learning_rate": 0.24715716345787123,
      "loss": 0.2535,
      "num_input_tokens_seen": 8413384,
      "step": 11030
    },
    {
      "epoch": 22.941787941787943,
      "grad_norm": 0.00046227345592342317,
      "learning_rate": 0.2471122773420387,
      "loss": 0.2346,
      "num_input_tokens_seen": 8417256,
      "step": 11035
    },
    {
      "epoch": 22.95218295218295,
      "grad_norm": 0.00015766163414809853,
      "learning_rate": 0.24706737625027259,
      "loss": 0.2415,
      "num_input_tokens_seen": 8421096,
      "step": 11040
    },
    {
      "epoch": 22.962577962577964,
      "grad_norm": 0.00012834918743465096,
      "learning_rate": 0.24702246018949725,
      "loss": 0.2935,
      "num_input_tokens_seen": 8424872,
      "step": 11045
    },
    {
      "epoch": 22.972972972972972,
      "grad_norm": 0.00014894672494847327,
      "learning_rate": 0.2469775291666393,
      "loss": 0.2685,
      "num_input_tokens_seen": 8428648,
      "step": 11050
    },
    {
      "epoch": 22.983367983367984,
      "grad_norm": 0.0003273191978223622,
      "learning_rate": 0.24693258318862765,
      "loss": 0.2626,
      "num_input_tokens_seen": 8432616,
      "step": 11055
    },
    {
      "epoch": 22.993762993762992,
      "grad_norm": 0.0008077337988652289,
      "learning_rate": 0.2468876222623935,
      "loss": 0.265,
      "num_input_tokens_seen": 8436616,
      "step": 11060
    },
    {
      "epoch": 23.004158004158004,
      "grad_norm": 0.0002598306746222079,
      "learning_rate": 0.2468426463948705,
      "loss": 0.3022,
      "num_input_tokens_seen": 8440208,
      "step": 11065
    },
    {
      "epoch": 23.014553014553016,
      "grad_norm": 0.000348987290635705,
      "learning_rate": 0.24679765559299438,
      "loss": 0.2744,
      "num_input_tokens_seen": 8444112,
      "step": 11070
    },
    {
      "epoch": 23.024948024948024,
      "grad_norm": 0.000650043657515198,
      "learning_rate": 0.24675264986370332,
      "loss": 0.2776,
      "num_input_tokens_seen": 8447920,
      "step": 11075
    },
    {
      "epoch": 23.035343035343036,
      "grad_norm": 0.00030308536952361465,
      "learning_rate": 0.2467076292139378,
      "loss": 0.2516,
      "num_input_tokens_seen": 8451792,
      "step": 11080
    },
    {
      "epoch": 23.045738045738045,
      "grad_norm": 0.00021021509019192308,
      "learning_rate": 0.24666259365064055,
      "loss": 0.2305,
      "num_input_tokens_seen": 8455504,
      "step": 11085
    },
    {
      "epoch": 23.056133056133056,
      "grad_norm": 0.00029759813332930207,
      "learning_rate": 0.24661754318075663,
      "loss": 0.2575,
      "num_input_tokens_seen": 8459408,
      "step": 11090
    },
    {
      "epoch": 23.066528066528065,
      "grad_norm": 0.000461197312688455,
      "learning_rate": 0.2465724778112334,
      "loss": 0.2767,
      "num_input_tokens_seen": 8463280,
      "step": 11095
    },
    {
      "epoch": 23.076923076923077,
      "grad_norm": 0.0005435289931483567,
      "learning_rate": 0.24652739754902042,
      "loss": 0.2807,
      "num_input_tokens_seen": 8467120,
      "step": 11100
    },
    {
      "epoch": 23.08731808731809,
      "grad_norm": 0.0008591612568125129,
      "learning_rate": 0.24648230240106975,
      "loss": 0.2692,
      "num_input_tokens_seen": 8470960,
      "step": 11105
    },
    {
      "epoch": 23.097713097713097,
      "grad_norm": 0.0005267381202429533,
      "learning_rate": 0.2464371923743356,
      "loss": 0.2306,
      "num_input_tokens_seen": 8474736,
      "step": 11110
    },
    {
      "epoch": 23.10810810810811,
      "grad_norm": 0.00031147280242294073,
      "learning_rate": 0.24639206747577444,
      "loss": 0.262,
      "num_input_tokens_seen": 8478672,
      "step": 11115
    },
    {
      "epoch": 23.118503118503117,
      "grad_norm": 0.0003172263677697629,
      "learning_rate": 0.24634692771234515,
      "loss": 0.2708,
      "num_input_tokens_seen": 8482608,
      "step": 11120
    },
    {
      "epoch": 23.12889812889813,
      "grad_norm": 0.0001414748840034008,
      "learning_rate": 0.2463017730910088,
      "loss": 0.2639,
      "num_input_tokens_seen": 8486480,
      "step": 11125
    },
    {
      "epoch": 23.13929313929314,
      "grad_norm": 0.000844207766931504,
      "learning_rate": 0.2462566036187289,
      "loss": 0.2635,
      "num_input_tokens_seen": 8490224,
      "step": 11130
    },
    {
      "epoch": 23.14968814968815,
      "grad_norm": 0.0010493883164599538,
      "learning_rate": 0.24621141930247106,
      "loss": 0.2633,
      "num_input_tokens_seen": 8493968,
      "step": 11135
    },
    {
      "epoch": 23.16008316008316,
      "grad_norm": 0.00019993902242276818,
      "learning_rate": 0.2461662201492033,
      "loss": 0.2601,
      "num_input_tokens_seen": 8497712,
      "step": 11140
    },
    {
      "epoch": 23.17047817047817,
      "grad_norm": 0.00027402755222283304,
      "learning_rate": 0.24612100616589586,
      "loss": 0.2663,
      "num_input_tokens_seen": 8501328,
      "step": 11145
    },
    {
      "epoch": 23.18087318087318,
      "grad_norm": 0.0005629007937386632,
      "learning_rate": 0.24607577735952135,
      "loss": 0.2785,
      "num_input_tokens_seen": 8505136,
      "step": 11150
    },
    {
      "epoch": 23.19126819126819,
      "grad_norm": 0.0006390362395904958,
      "learning_rate": 0.24603053373705464,
      "loss": 0.2748,
      "num_input_tokens_seen": 8508912,
      "step": 11155
    },
    {
      "epoch": 23.2016632016632,
      "grad_norm": 0.0005824014660902321,
      "learning_rate": 0.2459852753054728,
      "loss": 0.2833,
      "num_input_tokens_seen": 8512784,
      "step": 11160
    },
    {
      "epoch": 23.212058212058214,
      "grad_norm": 0.0002615506818983704,
      "learning_rate": 0.24594000207175526,
      "loss": 0.2548,
      "num_input_tokens_seen": 8516560,
      "step": 11165
    },
    {
      "epoch": 23.222453222453222,
      "grad_norm": 0.00040190783329308033,
      "learning_rate": 0.2458947140428838,
      "loss": 0.2687,
      "num_input_tokens_seen": 8520400,
      "step": 11170
    },
    {
      "epoch": 23.232848232848234,
      "grad_norm": 0.0004516924964264035,
      "learning_rate": 0.24584941122584233,
      "loss": 0.2866,
      "num_input_tokens_seen": 8524208,
      "step": 11175
    },
    {
      "epoch": 23.243243243243242,
      "grad_norm": 0.00045322030200622976,
      "learning_rate": 0.24580409362761713,
      "loss": 0.2819,
      "num_input_tokens_seen": 8528048,
      "step": 11180
    },
    {
      "epoch": 23.253638253638254,
      "grad_norm": 9.581913036527112e-05,
      "learning_rate": 0.2457587612551967,
      "loss": 0.2712,
      "num_input_tokens_seen": 8531920,
      "step": 11185
    },
    {
      "epoch": 23.264033264033262,
      "grad_norm": 0.0005042356206104159,
      "learning_rate": 0.24571341411557193,
      "loss": 0.2512,
      "num_input_tokens_seen": 8535696,
      "step": 11190
    },
    {
      "epoch": 23.274428274428274,
      "grad_norm": 0.00011410010483814403,
      "learning_rate": 0.2456680522157359,
      "loss": 0.2696,
      "num_input_tokens_seen": 8539568,
      "step": 11195
    },
    {
      "epoch": 23.284823284823286,
      "grad_norm": 0.0006625992245972157,
      "learning_rate": 0.245622675562684,
      "loss": 0.266,
      "num_input_tokens_seen": 8543280,
      "step": 11200
    },
    {
      "epoch": 23.284823284823286,
      "eval_loss": 0.2535899877548218,
      "eval_runtime": 13.4784,
      "eval_samples_per_second": 63.509,
      "eval_steps_per_second": 15.877,
      "num_input_tokens_seen": 8543280,
      "step": 11200
    },
    {
      "epoch": 23.295218295218294,
      "grad_norm": 0.0005809700815007091,
      "learning_rate": 0.24557728416341384,
      "loss": 0.2772,
      "num_input_tokens_seen": 8547184,
      "step": 11205
    },
    {
      "epoch": 23.305613305613306,
      "grad_norm": 0.0003124606446363032,
      "learning_rate": 0.24553187802492538,
      "loss": 0.2534,
      "num_input_tokens_seen": 8551024,
      "step": 11210
    },
    {
      "epoch": 23.316008316008315,
      "grad_norm": 0.0001347609650110826,
      "learning_rate": 0.24548645715422074,
      "loss": 0.2506,
      "num_input_tokens_seen": 8554864,
      "step": 11215
    },
    {
      "epoch": 23.326403326403327,
      "grad_norm": 0.00015566272486466914,
      "learning_rate": 0.2454410215583045,
      "loss": 0.2262,
      "num_input_tokens_seen": 8558704,
      "step": 11220
    },
    {
      "epoch": 23.33679833679834,
      "grad_norm": 0.000481221271911636,
      "learning_rate": 0.24539557124418332,
      "loss": 0.2455,
      "num_input_tokens_seen": 8562448,
      "step": 11225
    },
    {
      "epoch": 23.347193347193347,
      "grad_norm": 0.0002525653981138021,
      "learning_rate": 0.24535010621886624,
      "loss": 0.2774,
      "num_input_tokens_seen": 8566256,
      "step": 11230
    },
    {
      "epoch": 23.35758835758836,
      "grad_norm": 0.0013254560763016343,
      "learning_rate": 0.2453046264893646,
      "loss": 0.2895,
      "num_input_tokens_seen": 8569968,
      "step": 11235
    },
    {
      "epoch": 23.367983367983367,
      "grad_norm": 0.0009029508219100535,
      "learning_rate": 0.24525913206269184,
      "loss": 0.2852,
      "num_input_tokens_seen": 8574000,
      "step": 11240
    },
    {
      "epoch": 23.37837837837838,
      "grad_norm": 0.0008508173050358891,
      "learning_rate": 0.2452136229458638,
      "loss": 0.2937,
      "num_input_tokens_seen": 8577840,
      "step": 11245
    },
    {
      "epoch": 23.388773388773387,
      "grad_norm": 0.0013674009824171662,
      "learning_rate": 0.24516809914589857,
      "loss": 0.2486,
      "num_input_tokens_seen": 8581648,
      "step": 11250
    },
    {
      "epoch": 23.3991683991684,
      "grad_norm": 0.0006002847221679986,
      "learning_rate": 0.2451225606698165,
      "loss": 0.2697,
      "num_input_tokens_seen": 8585680,
      "step": 11255
    },
    {
      "epoch": 23.40956340956341,
      "grad_norm": 0.000300979329040274,
      "learning_rate": 0.2450770075246402,
      "loss": 0.2789,
      "num_input_tokens_seen": 8589392,
      "step": 11260
    },
    {
      "epoch": 23.41995841995842,
      "grad_norm": 0.0005880179814994335,
      "learning_rate": 0.24503143971739455,
      "loss": 0.2747,
      "num_input_tokens_seen": 8593328,
      "step": 11265
    },
    {
      "epoch": 23.43035343035343,
      "grad_norm": 0.0005758588085882366,
      "learning_rate": 0.24498585725510663,
      "loss": 0.2753,
      "num_input_tokens_seen": 8597296,
      "step": 11270
    },
    {
      "epoch": 23.44074844074844,
      "grad_norm": 0.00017160146671812981,
      "learning_rate": 0.24494026014480583,
      "loss": 0.2649,
      "num_input_tokens_seen": 8601264,
      "step": 11275
    },
    {
      "epoch": 23.45114345114345,
      "grad_norm": 0.0007787612266838551,
      "learning_rate": 0.24489464839352387,
      "loss": 0.2883,
      "num_input_tokens_seen": 8605040,
      "step": 11280
    },
    {
      "epoch": 23.46153846153846,
      "grad_norm": 0.0004692948132287711,
      "learning_rate": 0.2448490220082946,
      "loss": 0.2557,
      "num_input_tokens_seen": 8608752,
      "step": 11285
    },
    {
      "epoch": 23.471933471933472,
      "grad_norm": 0.00012573765707202256,
      "learning_rate": 0.24480338099615415,
      "loss": 0.2581,
      "num_input_tokens_seen": 8612464,
      "step": 11290
    },
    {
      "epoch": 23.482328482328484,
      "grad_norm": 0.000173399384948425,
      "learning_rate": 0.244757725364141,
      "loss": 0.2545,
      "num_input_tokens_seen": 8616240,
      "step": 11295
    },
    {
      "epoch": 23.492723492723492,
      "grad_norm": 0.0003453368553891778,
      "learning_rate": 0.24471205511929583,
      "loss": 0.2901,
      "num_input_tokens_seen": 8620048,
      "step": 11300
    },
    {
      "epoch": 23.503118503118504,
      "grad_norm": 0.00042090981150977314,
      "learning_rate": 0.24466637026866145,
      "loss": 0.2741,
      "num_input_tokens_seen": 8623824,
      "step": 11305
    },
    {
      "epoch": 23.513513513513512,
      "grad_norm": 0.00013157064677216113,
      "learning_rate": 0.2446206708192832,
      "loss": 0.2674,
      "num_input_tokens_seen": 8627664,
      "step": 11310
    },
    {
      "epoch": 23.523908523908524,
      "grad_norm": 0.00015192497812677175,
      "learning_rate": 0.2445749567782084,
      "loss": 0.2688,
      "num_input_tokens_seen": 8631504,
      "step": 11315
    },
    {
      "epoch": 23.534303534303533,
      "grad_norm": 0.0004945768741890788,
      "learning_rate": 0.2445292281524868,
      "loss": 0.2697,
      "num_input_tokens_seen": 8635472,
      "step": 11320
    },
    {
      "epoch": 23.544698544698544,
      "grad_norm": 0.00030011392664164305,
      "learning_rate": 0.24448348494917022,
      "loss": 0.2665,
      "num_input_tokens_seen": 8639376,
      "step": 11325
    },
    {
      "epoch": 23.555093555093556,
      "grad_norm": 0.00026054843328893185,
      "learning_rate": 0.24443772717531295,
      "loss": 0.2785,
      "num_input_tokens_seen": 8643120,
      "step": 11330
    },
    {
      "epoch": 23.565488565488565,
      "grad_norm": 0.00013831326214130968,
      "learning_rate": 0.24439195483797138,
      "loss": 0.2847,
      "num_input_tokens_seen": 8646896,
      "step": 11335
    },
    {
      "epoch": 23.575883575883577,
      "grad_norm": 0.0006735522183589637,
      "learning_rate": 0.24434616794420416,
      "loss": 0.2752,
      "num_input_tokens_seen": 8650768,
      "step": 11340
    },
    {
      "epoch": 23.586278586278585,
      "grad_norm": 0.0002660851750988513,
      "learning_rate": 0.24430036650107223,
      "loss": 0.2653,
      "num_input_tokens_seen": 8654576,
      "step": 11345
    },
    {
      "epoch": 23.596673596673597,
      "grad_norm": 0.0005186255439184606,
      "learning_rate": 0.2442545505156387,
      "loss": 0.2764,
      "num_input_tokens_seen": 8658448,
      "step": 11350
    },
    {
      "epoch": 23.60706860706861,
      "grad_norm": 0.0005912862252444029,
      "learning_rate": 0.24420871999496904,
      "loss": 0.2757,
      "num_input_tokens_seen": 8662352,
      "step": 11355
    },
    {
      "epoch": 23.617463617463617,
      "grad_norm": 0.0006086938083171844,
      "learning_rate": 0.24416287494613084,
      "loss": 0.2787,
      "num_input_tokens_seen": 8666192,
      "step": 11360
    },
    {
      "epoch": 23.62785862785863,
      "grad_norm": 0.00010927175753749907,
      "learning_rate": 0.24411701537619399,
      "loss": 0.2588,
      "num_input_tokens_seen": 8670032,
      "step": 11365
    },
    {
      "epoch": 23.638253638253637,
      "grad_norm": 0.0005654322449117899,
      "learning_rate": 0.24407114129223062,
      "loss": 0.2489,
      "num_input_tokens_seen": 8673840,
      "step": 11370
    },
    {
      "epoch": 23.64864864864865,
      "grad_norm": 7.375364657491446e-05,
      "learning_rate": 0.2440252527013151,
      "loss": 0.2614,
      "num_input_tokens_seen": 8677456,
      "step": 11375
    },
    {
      "epoch": 23.659043659043657,
      "grad_norm": 3.828847184195183e-05,
      "learning_rate": 0.24397934961052403,
      "loss": 0.2649,
      "num_input_tokens_seen": 8681136,
      "step": 11380
    },
    {
      "epoch": 23.66943866943867,
      "grad_norm": 0.0005979792331345379,
      "learning_rate": 0.24393343202693618,
      "loss": 0.235,
      "num_input_tokens_seen": 8684912,
      "step": 11385
    },
    {
      "epoch": 23.67983367983368,
      "grad_norm": 0.00038252881495282054,
      "learning_rate": 0.2438874999576327,
      "loss": 0.2851,
      "num_input_tokens_seen": 8688816,
      "step": 11390
    },
    {
      "epoch": 23.69022869022869,
      "grad_norm": 8.783923840383068e-05,
      "learning_rate": 0.24384155340969688,
      "loss": 0.2697,
      "num_input_tokens_seen": 8692720,
      "step": 11395
    },
    {
      "epoch": 23.7006237006237,
      "grad_norm": 0.0006030662334524095,
      "learning_rate": 0.24379559239021423,
      "loss": 0.2712,
      "num_input_tokens_seen": 8696432,
      "step": 11400
    },
    {
      "epoch": 23.7006237006237,
      "eval_loss": 0.25823086500167847,
      "eval_runtime": 13.4282,
      "eval_samples_per_second": 63.747,
      "eval_steps_per_second": 15.937,
      "num_input_tokens_seen": 8696432,
      "step": 11400
    },
    {
      "epoch": 23.71101871101871,
      "grad_norm": 0.00014482211554422975,
      "learning_rate": 0.2437496169062725,
      "loss": 0.2818,
      "num_input_tokens_seen": 8700272,
      "step": 11405
    },
    {
      "epoch": 23.72141372141372,
      "grad_norm": 0.0003956988512072712,
      "learning_rate": 0.24370362696496176,
      "loss": 0.2845,
      "num_input_tokens_seen": 8704112,
      "step": 11410
    },
    {
      "epoch": 23.731808731808734,
      "grad_norm": 0.00016388812218792737,
      "learning_rate": 0.24365762257337417,
      "loss": 0.2573,
      "num_input_tokens_seen": 8707952,
      "step": 11415
    },
    {
      "epoch": 23.742203742203742,
      "grad_norm": 0.001138166873715818,
      "learning_rate": 0.2436116037386042,
      "loss": 0.3075,
      "num_input_tokens_seen": 8711696,
      "step": 11420
    },
    {
      "epoch": 23.752598752598754,
      "grad_norm": 0.0001774019910953939,
      "learning_rate": 0.24356557046774852,
      "loss": 0.2921,
      "num_input_tokens_seen": 8715728,
      "step": 11425
    },
    {
      "epoch": 23.762993762993762,
      "grad_norm": 0.0004788069927599281,
      "learning_rate": 0.24351952276790606,
      "loss": 0.295,
      "num_input_tokens_seen": 8719536,
      "step": 11430
    },
    {
      "epoch": 23.773388773388774,
      "grad_norm": 0.0005475771613419056,
      "learning_rate": 0.24347346064617797,
      "loss": 0.275,
      "num_input_tokens_seen": 8723344,
      "step": 11435
    },
    {
      "epoch": 23.783783783783782,
      "grad_norm": 0.00013213448983151466,
      "learning_rate": 0.24342738410966758,
      "loss": 0.2439,
      "num_input_tokens_seen": 8727184,
      "step": 11440
    },
    {
      "epoch": 23.794178794178794,
      "grad_norm": 0.00032229532371275127,
      "learning_rate": 0.24338129316548046,
      "loss": 0.251,
      "num_input_tokens_seen": 8730992,
      "step": 11445
    },
    {
      "epoch": 23.804573804573806,
      "grad_norm": 0.0009105876088142395,
      "learning_rate": 0.24333518782072444,
      "loss": 0.2757,
      "num_input_tokens_seen": 8734704,
      "step": 11450
    },
    {
      "epoch": 23.814968814968815,
      "grad_norm": 0.0008148581255227327,
      "learning_rate": 0.24328906808250952,
      "loss": 0.2849,
      "num_input_tokens_seen": 8738512,
      "step": 11455
    },
    {
      "epoch": 23.825363825363826,
      "grad_norm": 0.0008986066095530987,
      "learning_rate": 0.243242933957948,
      "loss": 0.2796,
      "num_input_tokens_seen": 8742256,
      "step": 11460
    },
    {
      "epoch": 23.835758835758835,
      "grad_norm": 0.0007245682063512504,
      "learning_rate": 0.24319678545415427,
      "loss": 0.297,
      "num_input_tokens_seen": 8746064,
      "step": 11465
    },
    {
      "epoch": 23.846153846153847,
      "grad_norm": 0.0003781277628149837,
      "learning_rate": 0.24315062257824507,
      "loss": 0.275,
      "num_input_tokens_seen": 8749840,
      "step": 11470
    },
    {
      "epoch": 23.856548856548855,
      "grad_norm": 8.64934190758504e-05,
      "learning_rate": 0.24310444533733921,
      "loss": 0.2807,
      "num_input_tokens_seen": 8753616,
      "step": 11475
    },
    {
      "epoch": 23.866943866943867,
      "grad_norm": 0.00019423867342993617,
      "learning_rate": 0.2430582537385579,
      "loss": 0.2953,
      "num_input_tokens_seen": 8757264,
      "step": 11480
    },
    {
      "epoch": 23.87733887733888,
      "grad_norm": 0.0004998295335099101,
      "learning_rate": 0.2430120477890244,
      "loss": 0.2753,
      "num_input_tokens_seen": 8761104,
      "step": 11485
    },
    {
      "epoch": 23.887733887733887,
      "grad_norm": 0.00035150619805790484,
      "learning_rate": 0.24296582749586426,
      "loss": 0.2569,
      "num_input_tokens_seen": 8764848,
      "step": 11490
    },
    {
      "epoch": 23.8981288981289,
      "grad_norm": 0.00010583133553154767,
      "learning_rate": 0.24291959286620526,
      "loss": 0.2734,
      "num_input_tokens_seen": 8768624,
      "step": 11495
    },
    {
      "epoch": 23.908523908523907,
      "grad_norm": 0.00037805380998179317,
      "learning_rate": 0.24287334390717738,
      "loss": 0.241,
      "num_input_tokens_seen": 8772464,
      "step": 11500
    },
    {
      "epoch": 23.91891891891892,
      "grad_norm": 0.00040706092840991914,
      "learning_rate": 0.24282708062591268,
      "loss": 0.2647,
      "num_input_tokens_seen": 8776368,
      "step": 11505
    },
    {
      "epoch": 23.929313929313928,
      "grad_norm": 8.585912291891873e-05,
      "learning_rate": 0.24278080302954563,
      "loss": 0.2689,
      "num_input_tokens_seen": 8780208,
      "step": 11510
    },
    {
      "epoch": 23.93970893970894,
      "grad_norm": 0.0005100580747239292,
      "learning_rate": 0.24273451112521283,
      "loss": 0.2674,
      "num_input_tokens_seen": 8783984,
      "step": 11515
    },
    {
      "epoch": 23.95010395010395,
      "grad_norm": 0.00020650606893468648,
      "learning_rate": 0.242688204920053,
      "loss": 0.2573,
      "num_input_tokens_seen": 8787856,
      "step": 11520
    },
    {
      "epoch": 23.96049896049896,
      "grad_norm": 0.0003401414433028549,
      "learning_rate": 0.24264188442120715,
      "loss": 0.2629,
      "num_input_tokens_seen": 8791600,
      "step": 11525
    },
    {
      "epoch": 23.97089397089397,
      "grad_norm": 0.0001000309202936478,
      "learning_rate": 0.24259554963581853,
      "loss": 0.2295,
      "num_input_tokens_seen": 8795312,
      "step": 11530
    },
    {
      "epoch": 23.98128898128898,
      "grad_norm": 0.000363660859875381,
      "learning_rate": 0.24254920057103257,
      "loss": 0.273,
      "num_input_tokens_seen": 8799184,
      "step": 11535
    },
    {
      "epoch": 23.991683991683992,
      "grad_norm": 0.00042738919728435576,
      "learning_rate": 0.24250283723399685,
      "loss": 0.2189,
      "num_input_tokens_seen": 8802960,
      "step": 11540
    },
    {
      "epoch": 24.002079002079004,
      "grad_norm": 0.0002777199260890484,
      "learning_rate": 0.24245645963186108,
      "loss": 0.2775,
      "num_input_tokens_seen": 8806816,
      "step": 11545
    },
    {
      "epoch": 24.012474012474012,
      "grad_norm": 0.0008512705680914223,
      "learning_rate": 0.2424100677717774,
      "loss": 0.3022,
      "num_input_tokens_seen": 8810720,
      "step": 11550
    },
    {
      "epoch": 24.022869022869024,
      "grad_norm": 0.00034911895636469126,
      "learning_rate": 0.24236366166090004,
      "loss": 0.282,
      "num_input_tokens_seen": 8814816,
      "step": 11555
    },
    {
      "epoch": 24.033264033264032,
      "grad_norm": 0.0013787307543680072,
      "learning_rate": 0.24231724130638527,
      "loss": 0.2805,
      "num_input_tokens_seen": 8818688,
      "step": 11560
    },
    {
      "epoch": 24.043659043659044,
      "grad_norm": 0.0007107564015313983,
      "learning_rate": 0.2422708067153917,
      "loss": 0.2755,
      "num_input_tokens_seen": 8822528,
      "step": 11565
    },
    {
      "epoch": 24.054054054054053,
      "grad_norm": 0.0005409236764535308,
      "learning_rate": 0.24222435789508026,
      "loss": 0.2691,
      "num_input_tokens_seen": 8826528,
      "step": 11570
    },
    {
      "epoch": 24.064449064449065,
      "grad_norm": 0.0004387928347568959,
      "learning_rate": 0.24217789485261387,
      "loss": 0.2698,
      "num_input_tokens_seen": 8830368,
      "step": 11575
    },
    {
      "epoch": 24.074844074844076,
      "grad_norm": 0.0003292463661637157,
      "learning_rate": 0.2421314175951577,
      "loss": 0.2717,
      "num_input_tokens_seen": 8834176,
      "step": 11580
    },
    {
      "epoch": 24.085239085239085,
      "grad_norm": 0.00030610713292844594,
      "learning_rate": 0.2420849261298791,
      "loss": 0.2943,
      "num_input_tokens_seen": 8838144,
      "step": 11585
    },
    {
      "epoch": 24.095634095634097,
      "grad_norm": 0.0003340231196489185,
      "learning_rate": 0.24203842046394775,
      "loss": 0.2839,
      "num_input_tokens_seen": 8841952,
      "step": 11590
    },
    {
      "epoch": 24.106029106029105,
      "grad_norm": 0.0007471473072655499,
      "learning_rate": 0.24199190060453535,
      "loss": 0.2716,
      "num_input_tokens_seen": 8845632,
      "step": 11595
    },
    {
      "epoch": 24.116424116424117,
      "grad_norm": 7.737662235740572e-05,
      "learning_rate": 0.2419453665588158,
      "loss": 0.2664,
      "num_input_tokens_seen": 8849408,
      "step": 11600
    },
    {
      "epoch": 24.116424116424117,
      "eval_loss": 0.24997226893901825,
      "eval_runtime": 13.4528,
      "eval_samples_per_second": 63.63,
      "eval_steps_per_second": 15.907,
      "num_input_tokens_seen": 8849408,
      "step": 11600
    },
    {
      "epoch": 24.126819126819125,
      "grad_norm": 0.00013961437798570842,
      "learning_rate": 0.24189881833396523,
      "loss": 0.259,
      "num_input_tokens_seen": 8853184,
      "step": 11605
    },
    {
      "epoch": 24.137214137214137,
      "grad_norm": 0.00033145881025120616,
      "learning_rate": 0.24185225593716203,
      "loss": 0.2518,
      "num_input_tokens_seen": 8856896,
      "step": 11610
    },
    {
      "epoch": 24.14760914760915,
      "grad_norm": 0.00014620805450249463,
      "learning_rate": 0.2418056793755867,
      "loss": 0.2595,
      "num_input_tokens_seen": 8860672,
      "step": 11615
    },
    {
      "epoch": 24.158004158004157,
      "grad_norm": 0.0008421497186645865,
      "learning_rate": 0.24175908865642187,
      "loss": 0.2813,
      "num_input_tokens_seen": 8864544,
      "step": 11620
    },
    {
      "epoch": 24.16839916839917,
      "grad_norm": 0.0003480222949292511,
      "learning_rate": 0.24171248378685248,
      "loss": 0.2805,
      "num_input_tokens_seen": 8868352,
      "step": 11625
    },
    {
      "epoch": 24.178794178794178,
      "grad_norm": 0.0004227267927490175,
      "learning_rate": 0.24166586477406554,
      "loss": 0.2793,
      "num_input_tokens_seen": 8872128,
      "step": 11630
    },
    {
      "epoch": 24.18918918918919,
      "grad_norm": 5.2247785788495094e-05,
      "learning_rate": 0.24161923162525034,
      "loss": 0.2808,
      "num_input_tokens_seen": 8875840,
      "step": 11635
    },
    {
      "epoch": 24.1995841995842,
      "grad_norm": 0.000608329544775188,
      "learning_rate": 0.2415725843475982,
      "loss": 0.2802,
      "num_input_tokens_seen": 8879648,
      "step": 11640
    },
    {
      "epoch": 24.20997920997921,
      "grad_norm": 3.049782208108809e-05,
      "learning_rate": 0.24152592294830286,
      "loss": 0.2641,
      "num_input_tokens_seen": 8883648,
      "step": 11645
    },
    {
      "epoch": 24.22037422037422,
      "grad_norm": 0.0006523304618895054,
      "learning_rate": 0.24147924743455995,
      "loss": 0.2821,
      "num_input_tokens_seen": 8887552,
      "step": 11650
    },
    {
      "epoch": 24.23076923076923,
      "grad_norm": 8.20455388748087e-05,
      "learning_rate": 0.24143255781356754,
      "loss": 0.2531,
      "num_input_tokens_seen": 8891424,
      "step": 11655
    },
    {
      "epoch": 24.241164241164242,
      "grad_norm": 0.0001101844827644527,
      "learning_rate": 0.24138585409252566,
      "loss": 0.2807,
      "num_input_tokens_seen": 8895264,
      "step": 11660
    },
    {
      "epoch": 24.25155925155925,
      "grad_norm": 0.0004913162556476891,
      "learning_rate": 0.24133913627863662,
      "loss": 0.2643,
      "num_input_tokens_seen": 8898944,
      "step": 11665
    },
    {
      "epoch": 24.261954261954262,
      "grad_norm": 0.0002816529886331409,
      "learning_rate": 0.241292404379105,
      "loss": 0.2619,
      "num_input_tokens_seen": 8902848,
      "step": 11670
    },
    {
      "epoch": 24.272349272349274,
      "grad_norm": 0.00026957373484037817,
      "learning_rate": 0.24124565840113735,
      "loss": 0.2499,
      "num_input_tokens_seen": 8906624,
      "step": 11675
    },
    {
      "epoch": 24.282744282744282,
      "grad_norm": 0.00021040673891548067,
      "learning_rate": 0.2411988983519425,
      "loss": 0.2637,
      "num_input_tokens_seen": 8910432,
      "step": 11680
    },
    {
      "epoch": 24.293139293139294,
      "grad_norm": 0.00018961050955113024,
      "learning_rate": 0.24115212423873145,
      "loss": 0.2671,
      "num_input_tokens_seen": 8914208,
      "step": 11685
    },
    {
      "epoch": 24.303534303534303,
      "grad_norm": 0.00011383459786884487,
      "learning_rate": 0.24110533606871737,
      "loss": 0.2785,
      "num_input_tokens_seen": 8918112,
      "step": 11690
    },
    {
      "epoch": 24.313929313929314,
      "grad_norm": 0.00026835172320716083,
      "learning_rate": 0.24105853384911552,
      "loss": 0.2727,
      "num_input_tokens_seen": 8921920,
      "step": 11695
    },
    {
      "epoch": 24.324324324324323,
      "grad_norm": 0.00013970643340144306,
      "learning_rate": 0.24101171758714346,
      "loss": 0.2615,
      "num_input_tokens_seen": 8925632,
      "step": 11700
    },
    {
      "epoch": 24.334719334719335,
      "grad_norm": 0.0005839199875481427,
      "learning_rate": 0.24096488729002086,
      "loss": 0.2357,
      "num_input_tokens_seen": 8929344,
      "step": 11705
    },
    {
      "epoch": 24.345114345114347,
      "grad_norm": 0.0003611715801525861,
      "learning_rate": 0.24091804296496946,
      "loss": 0.2977,
      "num_input_tokens_seen": 8933056,
      "step": 11710
    },
    {
      "epoch": 24.355509355509355,
      "grad_norm": 0.0005658221780322492,
      "learning_rate": 0.2408711846192133,
      "loss": 0.2586,
      "num_input_tokens_seen": 8936768,
      "step": 11715
    },
    {
      "epoch": 24.365904365904367,
      "grad_norm": 0.0004351457755547017,
      "learning_rate": 0.24082431225997855,
      "loss": 0.2817,
      "num_input_tokens_seen": 8940672,
      "step": 11720
    },
    {
      "epoch": 24.376299376299375,
      "grad_norm": 0.0003449256473686546,
      "learning_rate": 0.24077742589449344,
      "loss": 0.262,
      "num_input_tokens_seen": 8944512,
      "step": 11725
    },
    {
      "epoch": 24.386694386694387,
      "grad_norm": 0.00021875255333725363,
      "learning_rate": 0.24073052552998844,
      "loss": 0.2464,
      "num_input_tokens_seen": 8948256,
      "step": 11730
    },
    {
      "epoch": 24.397089397089395,
      "grad_norm": 0.00021527346689254045,
      "learning_rate": 0.2406836111736963,
      "loss": 0.2698,
      "num_input_tokens_seen": 8952032,
      "step": 11735
    },
    {
      "epoch": 24.407484407484407,
      "grad_norm": 0.00019175326451659203,
      "learning_rate": 0.2406366828328517,
      "loss": 0.2813,
      "num_input_tokens_seen": 8955872,
      "step": 11740
    },
    {
      "epoch": 24.41787941787942,
      "grad_norm": 0.00028428129735402763,
      "learning_rate": 0.2405897405146915,
      "loss": 0.2885,
      "num_input_tokens_seen": 8959680,
      "step": 11745
    },
    {
      "epoch": 24.428274428274428,
      "grad_norm": 0.0006020207656547427,
      "learning_rate": 0.240542784226455,
      "loss": 0.2801,
      "num_input_tokens_seen": 8963584,
      "step": 11750
    },
    {
      "epoch": 24.43866943866944,
      "grad_norm": 0.0001576445938553661,
      "learning_rate": 0.24049581397538328,
      "loss": 0.2067,
      "num_input_tokens_seen": 8967232,
      "step": 11755
    },
    {
      "epoch": 24.449064449064448,
      "grad_norm": 0.0008181019220501184,
      "learning_rate": 0.24044882976871984,
      "loss": 0.2633,
      "num_input_tokens_seen": 8971008,
      "step": 11760
    },
    {
      "epoch": 24.45945945945946,
      "grad_norm": 0.00036928593181073666,
      "learning_rate": 0.2404018316137102,
      "loss": 0.2393,
      "num_input_tokens_seen": 8974944,
      "step": 11765
    },
    {
      "epoch": 24.46985446985447,
      "grad_norm": 0.0008843201212584972,
      "learning_rate": 0.24035481951760204,
      "loss": 0.2743,
      "num_input_tokens_seen": 8978688,
      "step": 11770
    },
    {
      "epoch": 24.48024948024948,
      "grad_norm": 0.000123951758723706,
      "learning_rate": 0.2403077934876452,
      "loss": 0.2674,
      "num_input_tokens_seen": 8982432,
      "step": 11775
    },
    {
      "epoch": 24.490644490644492,
      "grad_norm": 0.0010666167363524437,
      "learning_rate": 0.2402607535310918,
      "loss": 0.2826,
      "num_input_tokens_seen": 8986272,
      "step": 11780
    },
    {
      "epoch": 24.5010395010395,
      "grad_norm": 9.37069344217889e-05,
      "learning_rate": 0.2402136996551959,
      "loss": 0.2666,
      "num_input_tokens_seen": 8990016,
      "step": 11785
    },
    {
      "epoch": 24.511434511434512,
      "grad_norm": 0.00044376743608154356,
      "learning_rate": 0.24016663186721376,
      "loss": 0.3267,
      "num_input_tokens_seen": 8993728,
      "step": 11790
    },
    {
      "epoch": 24.52182952182952,
      "grad_norm": 0.0006809383048675954,
      "learning_rate": 0.24011955017440395,
      "loss": 0.2668,
      "num_input_tokens_seen": 8997568,
      "step": 11795
    },
    {
      "epoch": 24.532224532224532,
      "grad_norm": 0.0003921407333109528,
      "learning_rate": 0.24007245458402696,
      "loss": 0.2668,
      "num_input_tokens_seen": 9001408,
      "step": 11800
    },
    {
      "epoch": 24.532224532224532,
      "eval_loss": 0.2539379298686981,
      "eval_runtime": 13.4097,
      "eval_samples_per_second": 63.834,
      "eval_steps_per_second": 15.959,
      "num_input_tokens_seen": 9001408,
      "step": 11800
    },
    {
      "epoch": 24.542619542619544,
      "grad_norm": 0.00016116558981593698,
      "learning_rate": 0.2400253451033456,
      "loss": 0.2579,
      "num_input_tokens_seen": 9005152,
      "step": 11805
    },
    {
      "epoch": 24.553014553014552,
      "grad_norm": 0.00011709408863680437,
      "learning_rate": 0.23997822173962463,
      "loss": 0.2544,
      "num_input_tokens_seen": 9009056,
      "step": 11810
    },
    {
      "epoch": 24.563409563409564,
      "grad_norm": 0.0003134087019134313,
      "learning_rate": 0.23993108450013118,
      "loss": 0.2778,
      "num_input_tokens_seen": 9012832,
      "step": 11815
    },
    {
      "epoch": 24.573804573804573,
      "grad_norm": 9.469052019994706e-05,
      "learning_rate": 0.2398839333921343,
      "loss": 0.2647,
      "num_input_tokens_seen": 9016576,
      "step": 11820
    },
    {
      "epoch": 24.584199584199585,
      "grad_norm": 0.00024380724062211812,
      "learning_rate": 0.23983676842290536,
      "loss": 0.2607,
      "num_input_tokens_seen": 9020224,
      "step": 11825
    },
    {
      "epoch": 24.594594594594593,
      "grad_norm": 0.0002000228123506531,
      "learning_rate": 0.2397895895997178,
      "loss": 0.2735,
      "num_input_tokens_seen": 9024000,
      "step": 11830
    },
    {
      "epoch": 24.604989604989605,
      "grad_norm": 0.0010783580364659429,
      "learning_rate": 0.23974239692984714,
      "loss": 0.278,
      "num_input_tokens_seen": 9027776,
      "step": 11835
    },
    {
      "epoch": 24.615384615384617,
      "grad_norm": 0.0009282429818995297,
      "learning_rate": 0.2396951904205711,
      "loss": 0.2704,
      "num_input_tokens_seen": 9031520,
      "step": 11840
    },
    {
      "epoch": 24.625779625779625,
      "grad_norm": 0.0003448710194788873,
      "learning_rate": 0.23964797007916952,
      "loss": 0.2776,
      "num_input_tokens_seen": 9035360,
      "step": 11845
    },
    {
      "epoch": 24.636174636174637,
      "grad_norm": 0.0006336988299153745,
      "learning_rate": 0.23960073591292436,
      "loss": 0.2617,
      "num_input_tokens_seen": 9039104,
      "step": 11850
    },
    {
      "epoch": 24.646569646569645,
      "grad_norm": 0.0003483004111330956,
      "learning_rate": 0.2395534879291197,
      "loss": 0.2606,
      "num_input_tokens_seen": 9043008,
      "step": 11855
    },
    {
      "epoch": 24.656964656964657,
      "grad_norm": 0.00030497240368276834,
      "learning_rate": 0.23950622613504186,
      "loss": 0.245,
      "num_input_tokens_seen": 9046912,
      "step": 11860
    },
    {
      "epoch": 24.66735966735967,
      "grad_norm": 0.00017458910588175058,
      "learning_rate": 0.2394589505379791,
      "loss": 0.2615,
      "num_input_tokens_seen": 9050656,
      "step": 11865
    },
    {
      "epoch": 24.677754677754677,
      "grad_norm": 0.0002103852020809427,
      "learning_rate": 0.23941166114522197,
      "loss": 0.2662,
      "num_input_tokens_seen": 9054400,
      "step": 11870
    },
    {
      "epoch": 24.68814968814969,
      "grad_norm": 0.00023987595341168344,
      "learning_rate": 0.23936435796406308,
      "loss": 0.2771,
      "num_input_tokens_seen": 9058208,
      "step": 11875
    },
    {
      "epoch": 24.698544698544698,
      "grad_norm": 0.0006227315752767026,
      "learning_rate": 0.23931704100179715,
      "loss": 0.267,
      "num_input_tokens_seen": 9061952,
      "step": 11880
    },
    {
      "epoch": 24.70893970893971,
      "grad_norm": 0.0004599630774464458,
      "learning_rate": 0.2392697102657211,
      "loss": 0.2836,
      "num_input_tokens_seen": 9065664,
      "step": 11885
    },
    {
      "epoch": 24.719334719334718,
      "grad_norm": 0.0006134499562904239,
      "learning_rate": 0.23922236576313388,
      "loss": 0.2832,
      "num_input_tokens_seen": 9069344,
      "step": 11890
    },
    {
      "epoch": 24.72972972972973,
      "grad_norm": 0.0002611627278383821,
      "learning_rate": 0.2391750075013366,
      "loss": 0.2769,
      "num_input_tokens_seen": 9073184,
      "step": 11895
    },
    {
      "epoch": 24.74012474012474,
      "grad_norm": 0.0002813110768329352,
      "learning_rate": 0.2391276354876326,
      "loss": 0.2613,
      "num_input_tokens_seen": 9077088,
      "step": 11900
    },
    {
      "epoch": 24.75051975051975,
      "grad_norm": 0.0008468155283480883,
      "learning_rate": 0.23908024972932707,
      "loss": 0.2888,
      "num_input_tokens_seen": 9081024,
      "step": 11905
    },
    {
      "epoch": 24.760914760914762,
      "grad_norm": 0.00011971390631515533,
      "learning_rate": 0.2390328502337276,
      "loss": 0.2704,
      "num_input_tokens_seen": 9084896,
      "step": 11910
    },
    {
      "epoch": 24.77130977130977,
      "grad_norm": 8.70052317623049e-05,
      "learning_rate": 0.23898543700814376,
      "loss": 0.2683,
      "num_input_tokens_seen": 9088704,
      "step": 11915
    },
    {
      "epoch": 24.781704781704782,
      "grad_norm": 0.00014342175563797355,
      "learning_rate": 0.2389380100598873,
      "loss": 0.2706,
      "num_input_tokens_seen": 9092416,
      "step": 11920
    },
    {
      "epoch": 24.79209979209979,
      "grad_norm": 0.00019683194113895297,
      "learning_rate": 0.23889056939627207,
      "loss": 0.2795,
      "num_input_tokens_seen": 9096128,
      "step": 11925
    },
    {
      "epoch": 24.802494802494802,
      "grad_norm": 0.0006490256637334824,
      "learning_rate": 0.23884311502461386,
      "loss": 0.2779,
      "num_input_tokens_seen": 9099840,
      "step": 11930
    },
    {
      "epoch": 24.812889812889814,
      "grad_norm": 0.00020424398826435208,
      "learning_rate": 0.23879564695223088,
      "loss": 0.2545,
      "num_input_tokens_seen": 9103712,
      "step": 11935
    },
    {
      "epoch": 24.823284823284823,
      "grad_norm": 0.00016514574235770851,
      "learning_rate": 0.23874816518644332,
      "loss": 0.2767,
      "num_input_tokens_seen": 9107520,
      "step": 11940
    },
    {
      "epoch": 24.833679833679835,
      "grad_norm": 0.00014729250688105822,
      "learning_rate": 0.23870066973457335,
      "loss": 0.2736,
      "num_input_tokens_seen": 9111392,
      "step": 11945
    },
    {
      "epoch": 24.844074844074843,
      "grad_norm": 7.94215957284905e-05,
      "learning_rate": 0.23865316060394545,
      "loss": 0.2413,
      "num_input_tokens_seen": 9115168,
      "step": 11950
    },
    {
      "epoch": 24.854469854469855,
      "grad_norm": 0.001129868789575994,
      "learning_rate": 0.2386056378018861,
      "loss": 0.2707,
      "num_input_tokens_seen": 9118848,
      "step": 11955
    },
    {
      "epoch": 24.864864864864863,
      "grad_norm": 0.000570828327909112,
      "learning_rate": 0.2385581013357239,
      "loss": 0.1991,
      "num_input_tokens_seen": 9122880,
      "step": 11960
    },
    {
      "epoch": 24.875259875259875,
      "grad_norm": 0.0002486240118741989,
      "learning_rate": 0.23851055121278958,
      "loss": 0.2431,
      "num_input_tokens_seen": 9126688,
      "step": 11965
    },
    {
      "epoch": 24.885654885654887,
      "grad_norm": 0.0006722926045767963,
      "learning_rate": 0.23846298744041594,
      "loss": 0.2749,
      "num_input_tokens_seen": 9130464,
      "step": 11970
    },
    {
      "epoch": 24.896049896049895,
      "grad_norm": 0.0006204298115335405,
      "learning_rate": 0.23841541002593802,
      "loss": 0.225,
      "num_input_tokens_seen": 9134080,
      "step": 11975
    },
    {
      "epoch": 24.906444906444907,
      "grad_norm": 0.00014286961231846362,
      "learning_rate": 0.23836781897669276,
      "loss": 0.2916,
      "num_input_tokens_seen": 9137824,
      "step": 11980
    },
    {
      "epoch": 24.916839916839916,
      "grad_norm": 0.00010744509927462786,
      "learning_rate": 0.23832021430001926,
      "loss": 0.2515,
      "num_input_tokens_seen": 9141760,
      "step": 11985
    },
    {
      "epoch": 24.927234927234927,
      "grad_norm": 0.00016330509970430285,
      "learning_rate": 0.2382725960032588,
      "loss": 0.2861,
      "num_input_tokens_seen": 9145696,
      "step": 11990
    },
    {
      "epoch": 24.93762993762994,
      "grad_norm": 0.000831234036013484,
      "learning_rate": 0.23822496409375482,
      "loss": 0.2813,
      "num_input_tokens_seen": 9149760,
      "step": 11995
    },
    {
      "epoch": 24.948024948024948,
      "grad_norm": 0.00016153437900356948,
      "learning_rate": 0.2381773185788526,
      "loss": 0.2822,
      "num_input_tokens_seen": 9153696,
      "step": 12000
    },
    {
      "epoch": 24.948024948024948,
      "eval_loss": 0.2769145369529724,
      "eval_runtime": 13.4503,
      "eval_samples_per_second": 63.642,
      "eval_steps_per_second": 15.91,
      "num_input_tokens_seen": 9153696,
      "step": 12000
    },
    {
      "epoch": 24.95841995841996,
      "grad_norm": 0.0004421588673721999,
      "learning_rate": 0.2381296594658998,
      "loss": 0.2743,
      "num_input_tokens_seen": 9157440,
      "step": 12005
    },
    {
      "epoch": 24.968814968814968,
      "grad_norm": 0.00010778033902170137,
      "learning_rate": 0.238081986762246,
      "loss": 0.26,
      "num_input_tokens_seen": 9161248,
      "step": 12010
    },
    {
      "epoch": 24.97920997920998,
      "grad_norm": 0.0003534658462740481,
      "learning_rate": 0.23803430047524293,
      "loss": 0.2621,
      "num_input_tokens_seen": 9164960,
      "step": 12015
    },
    {
      "epoch": 24.989604989604988,
      "grad_norm": 0.0002858824154827744,
      "learning_rate": 0.23798660061224441,
      "loss": 0.2564,
      "num_input_tokens_seen": 9168704,
      "step": 12020
    },
    {
      "epoch": 25.0,
      "grad_norm": 0.000365712505299598,
      "learning_rate": 0.23793888718060632,
      "loss": 0.2494,
      "num_input_tokens_seen": 9172624,
      "step": 12025
    },
    {
      "epoch": 25.010395010395012,
      "grad_norm": 0.00014775973977521062,
      "learning_rate": 0.23789116018768675,
      "loss": 0.2672,
      "num_input_tokens_seen": 9176400,
      "step": 12030
    },
    {
      "epoch": 25.02079002079002,
      "grad_norm": 0.0006968433735892177,
      "learning_rate": 0.2378434196408458,
      "loss": 0.2701,
      "num_input_tokens_seen": 9180272,
      "step": 12035
    },
    {
      "epoch": 25.031185031185032,
      "grad_norm": 7.712693331995979e-05,
      "learning_rate": 0.23779566554744563,
      "loss": 0.2756,
      "num_input_tokens_seen": 9184080,
      "step": 12040
    },
    {
      "epoch": 25.04158004158004,
      "grad_norm": 0.0003730101161636412,
      "learning_rate": 0.23774789791485051,
      "loss": 0.2769,
      "num_input_tokens_seen": 9187824,
      "step": 12045
    },
    {
      "epoch": 25.051975051975052,
      "grad_norm": 0.0001803341438062489,
      "learning_rate": 0.2377001167504268,
      "loss": 0.2576,
      "num_input_tokens_seen": 9191536,
      "step": 12050
    },
    {
      "epoch": 25.06237006237006,
      "grad_norm": 0.0001336833811365068,
      "learning_rate": 0.23765232206154302,
      "loss": 0.2448,
      "num_input_tokens_seen": 9195408,
      "step": 12055
    },
    {
      "epoch": 25.072765072765073,
      "grad_norm": 0.00039558636490255594,
      "learning_rate": 0.23760451385556966,
      "loss": 0.2943,
      "num_input_tokens_seen": 9199152,
      "step": 12060
    },
    {
      "epoch": 25.083160083160084,
      "grad_norm": 0.00046535010915249586,
      "learning_rate": 0.23755669213987932,
      "loss": 0.2656,
      "num_input_tokens_seen": 9202960,
      "step": 12065
    },
    {
      "epoch": 25.093555093555093,
      "grad_norm": 0.0004747312341351062,
      "learning_rate": 0.23750885692184676,
      "loss": 0.2673,
      "num_input_tokens_seen": 9206832,
      "step": 12070
    },
    {
      "epoch": 25.103950103950105,
      "grad_norm": 0.0007404095958918333,
      "learning_rate": 0.23746100820884875,
      "loss": 0.2673,
      "num_input_tokens_seen": 9210736,
      "step": 12075
    },
    {
      "epoch": 25.114345114345113,
      "grad_norm": 7.103076495695859e-05,
      "learning_rate": 0.23741314600826421,
      "loss": 0.2747,
      "num_input_tokens_seen": 9214576,
      "step": 12080
    },
    {
      "epoch": 25.124740124740125,
      "grad_norm": 0.0006793164066039026,
      "learning_rate": 0.23736527032747406,
      "loss": 0.2906,
      "num_input_tokens_seen": 9218256,
      "step": 12085
    },
    {
      "epoch": 25.135135135135137,
      "grad_norm": 0.0004902869113720953,
      "learning_rate": 0.23731738117386128,
      "loss": 0.2769,
      "num_input_tokens_seen": 9221968,
      "step": 12090
    },
    {
      "epoch": 25.145530145530145,
      "grad_norm": 0.0008474405622109771,
      "learning_rate": 0.237269478554811,
      "loss": 0.2653,
      "num_input_tokens_seen": 9225872,
      "step": 12095
    },
    {
      "epoch": 25.155925155925157,
      "grad_norm": 0.00038160482654348016,
      "learning_rate": 0.23722156247771053,
      "loss": 0.2949,
      "num_input_tokens_seen": 9229808,
      "step": 12100
    },
    {
      "epoch": 25.166320166320165,
      "grad_norm": 4.230980630381964e-05,
      "learning_rate": 0.23717363294994895,
      "loss": 0.2766,
      "num_input_tokens_seen": 9233456,
      "step": 12105
    },
    {
      "epoch": 25.176715176715177,
      "grad_norm": 0.0004975068150088191,
      "learning_rate": 0.2371256899789177,
      "loss": 0.2682,
      "num_input_tokens_seen": 9237360,
      "step": 12110
    },
    {
      "epoch": 25.187110187110186,
      "grad_norm": 0.00021768540318589658,
      "learning_rate": 0.23707773357201017,
      "loss": 0.2703,
      "num_input_tokens_seen": 9241360,
      "step": 12115
    },
    {
      "epoch": 25.197505197505198,
      "grad_norm": 0.0001609144965186715,
      "learning_rate": 0.2370297637366218,
      "loss": 0.2596,
      "num_input_tokens_seen": 9245264,
      "step": 12120
    },
    {
      "epoch": 25.20790020790021,
      "grad_norm": 0.0005136209656484425,
      "learning_rate": 0.23698178048015026,
      "loss": 0.2835,
      "num_input_tokens_seen": 9249072,
      "step": 12125
    },
    {
      "epoch": 25.218295218295218,
      "grad_norm": 0.0002003306581173092,
      "learning_rate": 0.236933783809995,
      "loss": 0.2702,
      "num_input_tokens_seen": 9252880,
      "step": 12130
    },
    {
      "epoch": 25.22869022869023,
      "grad_norm": 3.456450212979689e-05,
      "learning_rate": 0.23688577373355785,
      "loss": 0.2677,
      "num_input_tokens_seen": 9256656,
      "step": 12135
    },
    {
      "epoch": 25.239085239085238,
      "grad_norm": 0.0002603501779958606,
      "learning_rate": 0.23683775025824247,
      "loss": 0.275,
      "num_input_tokens_seen": 9260400,
      "step": 12140
    },
    {
      "epoch": 25.24948024948025,
      "grad_norm": 0.0003529172099661082,
      "learning_rate": 0.2367897133914548,
      "loss": 0.258,
      "num_input_tokens_seen": 9264336,
      "step": 12145
    },
    {
      "epoch": 25.25987525987526,
      "grad_norm": 0.000611013441812247,
      "learning_rate": 0.2367416631406026,
      "loss": 0.2657,
      "num_input_tokens_seen": 9268400,
      "step": 12150
    },
    {
      "epoch": 25.27027027027027,
      "grad_norm": 0.0006457085255533457,
      "learning_rate": 0.23669359951309588,
      "loss": 0.2756,
      "num_input_tokens_seen": 9272208,
      "step": 12155
    },
    {
      "epoch": 25.280665280665282,
      "grad_norm": 0.0004691470239777118,
      "learning_rate": 0.23664552251634666,
      "loss": 0.2663,
      "num_input_tokens_seen": 9276176,
      "step": 12160
    },
    {
      "epoch": 25.29106029106029,
      "grad_norm": 0.00025904655922204256,
      "learning_rate": 0.23659743215776907,
      "loss": 0.2796,
      "num_input_tokens_seen": 9280048,
      "step": 12165
    },
    {
      "epoch": 25.301455301455302,
      "grad_norm": 7.312667730730027e-05,
      "learning_rate": 0.23654932844477908,
      "loss": 0.264,
      "num_input_tokens_seen": 9284016,
      "step": 12170
    },
    {
      "epoch": 25.31185031185031,
      "grad_norm": 0.0007526806439273059,
      "learning_rate": 0.23650121138479507,
      "loss": 0.2645,
      "num_input_tokens_seen": 9287920,
      "step": 12175
    },
    {
      "epoch": 25.322245322245323,
      "grad_norm": 0.0001987564901355654,
      "learning_rate": 0.23645308098523724,
      "loss": 0.2382,
      "num_input_tokens_seen": 9291824,
      "step": 12180
    },
    {
      "epoch": 25.33264033264033,
      "grad_norm": 0.00020318591850809753,
      "learning_rate": 0.23640493725352785,
      "loss": 0.2647,
      "num_input_tokens_seen": 9295632,
      "step": 12185
    },
    {
      "epoch": 25.343035343035343,
      "grad_norm": 0.00010177848162129521,
      "learning_rate": 0.2363567801970913,
      "loss": 0.2636,
      "num_input_tokens_seen": 9299472,
      "step": 12190
    },
    {
      "epoch": 25.353430353430355,
      "grad_norm": 0.00040496455039829016,
      "learning_rate": 0.236308609823354,
      "loss": 0.269,
      "num_input_tokens_seen": 9303344,
      "step": 12195
    },
    {
      "epoch": 25.363825363825363,
      "grad_norm": 0.0003969160025008023,
      "learning_rate": 0.23626042613974452,
      "loss": 0.2745,
      "num_input_tokens_seen": 9307088,
      "step": 12200
    },
    {
      "epoch": 25.363825363825363,
      "eval_loss": 0.2504133880138397,
      "eval_runtime": 13.4638,
      "eval_samples_per_second": 63.578,
      "eval_steps_per_second": 15.894,
      "num_input_tokens_seen": 9307088,
      "step": 12200
    },
    {
      "epoch": 25.374220374220375,
      "grad_norm": 0.00026065215934067965,
      "learning_rate": 0.23621222915369325,
      "loss": 0.2615,
      "num_input_tokens_seen": 9310928,
      "step": 12205
    },
    {
      "epoch": 25.384615384615383,
      "grad_norm": 0.0005557139520533383,
      "learning_rate": 0.23616401887263283,
      "loss": 0.2371,
      "num_input_tokens_seen": 9314672,
      "step": 12210
    },
    {
      "epoch": 25.395010395010395,
      "grad_norm": 5.077619061921723e-05,
      "learning_rate": 0.23611579530399793,
      "loss": 0.2606,
      "num_input_tokens_seen": 9318416,
      "step": 12215
    },
    {
      "epoch": 25.405405405405407,
      "grad_norm": 0.00044916279148310423,
      "learning_rate": 0.23606755845522517,
      "loss": 0.2587,
      "num_input_tokens_seen": 9322128,
      "step": 12220
    },
    {
      "epoch": 25.415800415800415,
      "grad_norm": 0.00030319116194732487,
      "learning_rate": 0.23601930833375329,
      "loss": 0.2487,
      "num_input_tokens_seen": 9326000,
      "step": 12225
    },
    {
      "epoch": 25.426195426195427,
      "grad_norm": 0.0004046520043630153,
      "learning_rate": 0.23597104494702312,
      "loss": 0.2696,
      "num_input_tokens_seen": 9329904,
      "step": 12230
    },
    {
      "epoch": 25.436590436590436,
      "grad_norm": 0.000509279256220907,
      "learning_rate": 0.23592276830247744,
      "loss": 0.2629,
      "num_input_tokens_seen": 9333648,
      "step": 12235
    },
    {
      "epoch": 25.446985446985448,
      "grad_norm": 0.0005246573709882796,
      "learning_rate": 0.2358744784075611,
      "loss": 0.2783,
      "num_input_tokens_seen": 9337360,
      "step": 12240
    },
    {
      "epoch": 25.457380457380456,
      "grad_norm": 0.000660305842757225,
      "learning_rate": 0.235826175269721,
      "loss": 0.273,
      "num_input_tokens_seen": 9341232,
      "step": 12245
    },
    {
      "epoch": 25.467775467775468,
      "grad_norm": 0.00025600846856832504,
      "learning_rate": 0.23577785889640612,
      "loss": 0.2568,
      "num_input_tokens_seen": 9345104,
      "step": 12250
    },
    {
      "epoch": 25.47817047817048,
      "grad_norm": 0.00022463449568022043,
      "learning_rate": 0.23572952929506744,
      "loss": 0.2548,
      "num_input_tokens_seen": 9348912,
      "step": 12255
    },
    {
      "epoch": 25.488565488565488,
      "grad_norm": 0.0007313215173780918,
      "learning_rate": 0.23568118647315803,
      "loss": 0.2761,
      "num_input_tokens_seen": 9352784,
      "step": 12260
    },
    {
      "epoch": 25.4989604989605,
      "grad_norm": 5.804266038467176e-05,
      "learning_rate": 0.23563283043813296,
      "loss": 0.2666,
      "num_input_tokens_seen": 9356528,
      "step": 12265
    },
    {
      "epoch": 25.509355509355508,
      "grad_norm": 0.00037733608041889966,
      "learning_rate": 0.23558446119744922,
      "loss": 0.2586,
      "num_input_tokens_seen": 9360432,
      "step": 12270
    },
    {
      "epoch": 25.51975051975052,
      "grad_norm": 0.00020216168195474893,
      "learning_rate": 0.23553607875856608,
      "loss": 0.2632,
      "num_input_tokens_seen": 9364144,
      "step": 12275
    },
    {
      "epoch": 25.53014553014553,
      "grad_norm": 0.0003941018076147884,
      "learning_rate": 0.2354876831289447,
      "loss": 0.2653,
      "num_input_tokens_seen": 9367888,
      "step": 12280
    },
    {
      "epoch": 25.54054054054054,
      "grad_norm": 0.0004069653805345297,
      "learning_rate": 0.23543927431604827,
      "loss": 0.2553,
      "num_input_tokens_seen": 9371792,
      "step": 12285
    },
    {
      "epoch": 25.550935550935552,
      "grad_norm": 0.00014301339979283512,
      "learning_rate": 0.23539085232734203,
      "loss": 0.2451,
      "num_input_tokens_seen": 9375504,
      "step": 12290
    },
    {
      "epoch": 25.56133056133056,
      "grad_norm": 0.00013156987552065402,
      "learning_rate": 0.2353424171702933,
      "loss": 0.2976,
      "num_input_tokens_seen": 9379152,
      "step": 12295
    },
    {
      "epoch": 25.571725571725572,
      "grad_norm": 0.0006048521609045565,
      "learning_rate": 0.23529396885237133,
      "loss": 0.2722,
      "num_input_tokens_seen": 9382960,
      "step": 12300
    },
    {
      "epoch": 25.58212058212058,
      "grad_norm": 9.623035293770954e-05,
      "learning_rate": 0.2352455073810475,
      "loss": 0.2744,
      "num_input_tokens_seen": 9386864,
      "step": 12305
    },
    {
      "epoch": 25.592515592515593,
      "grad_norm": 0.00025088590336963534,
      "learning_rate": 0.23519703276379517,
      "loss": 0.2298,
      "num_input_tokens_seen": 9390736,
      "step": 12310
    },
    {
      "epoch": 25.602910602910605,
      "grad_norm": 0.0004715774266514927,
      "learning_rate": 0.2351485450080897,
      "loss": 0.2988,
      "num_input_tokens_seen": 9394576,
      "step": 12315
    },
    {
      "epoch": 25.613305613305613,
      "grad_norm": 0.00013741556904278696,
      "learning_rate": 0.2351000441214086,
      "loss": 0.2709,
      "num_input_tokens_seen": 9398384,
      "step": 12320
    },
    {
      "epoch": 25.623700623700625,
      "grad_norm": 0.000391049514291808,
      "learning_rate": 0.23505153011123125,
      "loss": 0.2734,
      "num_input_tokens_seen": 9402256,
      "step": 12325
    },
    {
      "epoch": 25.634095634095633,
      "grad_norm": 0.0010021297493949533,
      "learning_rate": 0.23500300298503912,
      "loss": 0.2923,
      "num_input_tokens_seen": 9406224,
      "step": 12330
    },
    {
      "epoch": 25.644490644490645,
      "grad_norm": 0.00020896481873933226,
      "learning_rate": 0.23495446275031576,
      "loss": 0.2692,
      "num_input_tokens_seen": 9410160,
      "step": 12335
    },
    {
      "epoch": 25.654885654885653,
      "grad_norm": 0.0005280310288071632,
      "learning_rate": 0.2349059094145466,
      "loss": 0.2783,
      "num_input_tokens_seen": 9413936,
      "step": 12340
    },
    {
      "epoch": 25.665280665280665,
      "grad_norm": 0.00010060082422569394,
      "learning_rate": 0.2348573429852192,
      "loss": 0.2325,
      "num_input_tokens_seen": 9417744,
      "step": 12345
    },
    {
      "epoch": 25.675675675675677,
      "grad_norm": 0.00010296085383743048,
      "learning_rate": 0.23480876346982313,
      "loss": 0.2518,
      "num_input_tokens_seen": 9421520,
      "step": 12350
    },
    {
      "epoch": 25.686070686070686,
      "grad_norm": 0.00018870858184527606,
      "learning_rate": 0.23476017087585,
      "loss": 0.2844,
      "num_input_tokens_seen": 9425456,
      "step": 12355
    },
    {
      "epoch": 25.696465696465697,
      "grad_norm": 0.00025114015443250537,
      "learning_rate": 0.23471156521079334,
      "loss": 0.2705,
      "num_input_tokens_seen": 9429200,
      "step": 12360
    },
    {
      "epoch": 25.706860706860706,
      "grad_norm": 0.000310354633256793,
      "learning_rate": 0.23466294648214875,
      "loss": 0.2762,
      "num_input_tokens_seen": 9433040,
      "step": 12365
    },
    {
      "epoch": 25.717255717255718,
      "grad_norm": 0.00012162828352302313,
      "learning_rate": 0.2346143146974139,
      "loss": 0.2281,
      "num_input_tokens_seen": 9436848,
      "step": 12370
    },
    {
      "epoch": 25.727650727650726,
      "grad_norm": 0.00043991600978188217,
      "learning_rate": 0.23456566986408836,
      "loss": 0.3021,
      "num_input_tokens_seen": 9440752,
      "step": 12375
    },
    {
      "epoch": 25.738045738045738,
      "grad_norm": 0.0002981819270644337,
      "learning_rate": 0.23451701198967384,
      "loss": 0.2587,
      "num_input_tokens_seen": 9444592,
      "step": 12380
    },
    {
      "epoch": 25.74844074844075,
      "grad_norm": 0.00028357660630717874,
      "learning_rate": 0.23446834108167397,
      "loss": 0.2792,
      "num_input_tokens_seen": 9448400,
      "step": 12385
    },
    {
      "epoch": 25.758835758835758,
      "grad_norm": 0.00033950238139368594,
      "learning_rate": 0.23441965714759438,
      "loss": 0.2724,
      "num_input_tokens_seen": 9452304,
      "step": 12390
    },
    {
      "epoch": 25.76923076923077,
      "grad_norm": 0.0009178064065054059,
      "learning_rate": 0.23437096019494277,
      "loss": 0.2677,
      "num_input_tokens_seen": 9455984,
      "step": 12395
    },
    {
      "epoch": 25.77962577962578,
      "grad_norm": 0.0008330817800015211,
      "learning_rate": 0.23432225023122885,
      "loss": 0.2536,
      "num_input_tokens_seen": 9459824,
      "step": 12400
    },
    {
      "epoch": 25.77962577962578,
      "eval_loss": 0.24817043542861938,
      "eval_runtime": 13.4113,
      "eval_samples_per_second": 63.827,
      "eval_steps_per_second": 15.957,
      "num_input_tokens_seen": 9459824,
      "step": 12400
    },
    {
      "epoch": 25.79002079002079,
      "grad_norm": 0.0006262730457819998,
      "learning_rate": 0.23427352726396428,
      "loss": 0.274,
      "num_input_tokens_seen": 9463664,
      "step": 12405
    },
    {
      "epoch": 25.8004158004158,
      "grad_norm": 0.00023454749316442758,
      "learning_rate": 0.2342247913006628,
      "loss": 0.272,
      "num_input_tokens_seen": 9467472,
      "step": 12410
    },
    {
      "epoch": 25.81081081081081,
      "grad_norm": 0.00041848912951536477,
      "learning_rate": 0.23417604234883999,
      "loss": 0.2767,
      "num_input_tokens_seen": 9471088,
      "step": 12415
    },
    {
      "epoch": 25.821205821205822,
      "grad_norm": 0.00023458755458705127,
      "learning_rate": 0.23412728041601363,
      "loss": 0.2786,
      "num_input_tokens_seen": 9474864,
      "step": 12420
    },
    {
      "epoch": 25.83160083160083,
      "grad_norm": 0.00040733005153015256,
      "learning_rate": 0.23407850550970347,
      "loss": 0.277,
      "num_input_tokens_seen": 9478576,
      "step": 12425
    },
    {
      "epoch": 25.841995841995843,
      "grad_norm": 0.00040534278377890587,
      "learning_rate": 0.23402971763743116,
      "loss": 0.2601,
      "num_input_tokens_seen": 9482384,
      "step": 12430
    },
    {
      "epoch": 25.85239085239085,
      "grad_norm": 0.0006408787448890507,
      "learning_rate": 0.23398091680672037,
      "loss": 0.2786,
      "num_input_tokens_seen": 9486288,
      "step": 12435
    },
    {
      "epoch": 25.862785862785863,
      "grad_norm": 0.00020789298287127167,
      "learning_rate": 0.23393210302509687,
      "loss": 0.2902,
      "num_input_tokens_seen": 9490032,
      "step": 12440
    },
    {
      "epoch": 25.873180873180875,
      "grad_norm": 0.00016575797053519636,
      "learning_rate": 0.23388327630008832,
      "loss": 0.2932,
      "num_input_tokens_seen": 9493808,
      "step": 12445
    },
    {
      "epoch": 25.883575883575883,
      "grad_norm": 5.5754659115336835e-05,
      "learning_rate": 0.23383443663922443,
      "loss": 0.2488,
      "num_input_tokens_seen": 9497616,
      "step": 12450
    },
    {
      "epoch": 25.893970893970895,
      "grad_norm": 0.0003425349132157862,
      "learning_rate": 0.23378558405003685,
      "loss": 0.3026,
      "num_input_tokens_seen": 9501520,
      "step": 12455
    },
    {
      "epoch": 25.904365904365903,
      "grad_norm": 0.0005627562059089541,
      "learning_rate": 0.2337367185400593,
      "loss": 0.2786,
      "num_input_tokens_seen": 9505520,
      "step": 12460
    },
    {
      "epoch": 25.914760914760915,
      "grad_norm": 0.0004024420923087746,
      "learning_rate": 0.23368784011682747,
      "loss": 0.2746,
      "num_input_tokens_seen": 9509328,
      "step": 12465
    },
    {
      "epoch": 25.925155925155924,
      "grad_norm": 9.865907486528158e-05,
      "learning_rate": 0.23363894878787902,
      "loss": 0.2612,
      "num_input_tokens_seen": 9513168,
      "step": 12470
    },
    {
      "epoch": 25.935550935550935,
      "grad_norm": 0.0007322691380977631,
      "learning_rate": 0.23359004456075352,
      "loss": 0.2718,
      "num_input_tokens_seen": 9516976,
      "step": 12475
    },
    {
      "epoch": 25.945945945945947,
      "grad_norm": 0.00012380574480630457,
      "learning_rate": 0.23354112744299277,
      "loss": 0.2899,
      "num_input_tokens_seen": 9520816,
      "step": 12480
    },
    {
      "epoch": 25.956340956340956,
      "grad_norm": 0.0006794001092202961,
      "learning_rate": 0.2334921974421403,
      "loss": 0.2575,
      "num_input_tokens_seen": 9524592,
      "step": 12485
    },
    {
      "epoch": 25.966735966735968,
      "grad_norm": 0.0005183862522244453,
      "learning_rate": 0.23344325456574178,
      "loss": 0.2586,
      "num_input_tokens_seen": 9528336,
      "step": 12490
    },
    {
      "epoch": 25.977130977130976,
      "grad_norm": 0.0009788044262677431,
      "learning_rate": 0.23339429882134477,
      "loss": 0.2903,
      "num_input_tokens_seen": 9532080,
      "step": 12495
    },
    {
      "epoch": 25.987525987525988,
      "grad_norm": 0.00023358248290605843,
      "learning_rate": 0.23334533021649884,
      "loss": 0.2761,
      "num_input_tokens_seen": 9535760,
      "step": 12500
    },
    {
      "epoch": 25.997920997921,
      "grad_norm": 0.00012860505376011133,
      "learning_rate": 0.23329634875875566,
      "loss": 0.2841,
      "num_input_tokens_seen": 9539440,
      "step": 12505
    },
    {
      "epoch": 26.008316008316008,
      "grad_norm": 0.00021904618188273162,
      "learning_rate": 0.23324735445566874,
      "loss": 0.2708,
      "num_input_tokens_seen": 9543000,
      "step": 12510
    },
    {
      "epoch": 26.01871101871102,
      "grad_norm": 0.00015079065633472055,
      "learning_rate": 0.2331983473147936,
      "loss": 0.2574,
      "num_input_tokens_seen": 9546840,
      "step": 12515
    },
    {
      "epoch": 26.02910602910603,
      "grad_norm": 0.00047588636516593397,
      "learning_rate": 0.23314932734368776,
      "loss": 0.3206,
      "num_input_tokens_seen": 9550808,
      "step": 12520
    },
    {
      "epoch": 26.03950103950104,
      "grad_norm": 0.0009797702077776194,
      "learning_rate": 0.2331002945499107,
      "loss": 0.2406,
      "num_input_tokens_seen": 9554520,
      "step": 12525
    },
    {
      "epoch": 26.04989604989605,
      "grad_norm": 0.0001158341474365443,
      "learning_rate": 0.23305124894102397,
      "loss": 0.2722,
      "num_input_tokens_seen": 9558328,
      "step": 12530
    },
    {
      "epoch": 26.06029106029106,
      "grad_norm": 0.0003536955046001822,
      "learning_rate": 0.23300219052459092,
      "loss": 0.2813,
      "num_input_tokens_seen": 9562104,
      "step": 12535
    },
    {
      "epoch": 26.070686070686072,
      "grad_norm": 0.0009927182691171765,
      "learning_rate": 0.23295311930817708,
      "loss": 0.2656,
      "num_input_tokens_seen": 9565976,
      "step": 12540
    },
    {
      "epoch": 26.08108108108108,
      "grad_norm": 0.00011536591046024114,
      "learning_rate": 0.23290403529934972,
      "loss": 0.264,
      "num_input_tokens_seen": 9569912,
      "step": 12545
    },
    {
      "epoch": 26.091476091476093,
      "grad_norm": 0.0006257393979467452,
      "learning_rate": 0.23285493850567832,
      "loss": 0.2761,
      "num_input_tokens_seen": 9573816,
      "step": 12550
    },
    {
      "epoch": 26.1018711018711,
      "grad_norm": 0.0006008932250551879,
      "learning_rate": 0.23280582893473414,
      "loss": 0.2807,
      "num_input_tokens_seen": 9577656,
      "step": 12555
    },
    {
      "epoch": 26.112266112266113,
      "grad_norm": 0.0003209675778634846,
      "learning_rate": 0.2327567065940906,
      "loss": 0.2795,
      "num_input_tokens_seen": 9581368,
      "step": 12560
    },
    {
      "epoch": 26.12266112266112,
      "grad_norm": 0.00016421207692474127,
      "learning_rate": 0.23270757149132285,
      "loss": 0.2745,
      "num_input_tokens_seen": 9585176,
      "step": 12565
    },
    {
      "epoch": 26.133056133056133,
      "grad_norm": 8.873855404090136e-05,
      "learning_rate": 0.23265842363400827,
      "loss": 0.2714,
      "num_input_tokens_seen": 9588952,
      "step": 12570
    },
    {
      "epoch": 26.143451143451145,
      "grad_norm": 0.0004526890115812421,
      "learning_rate": 0.23260926302972595,
      "loss": 0.25,
      "num_input_tokens_seen": 9592568,
      "step": 12575
    },
    {
      "epoch": 26.153846153846153,
      "grad_norm": 0.000568743736948818,
      "learning_rate": 0.2325600896860572,
      "loss": 0.2757,
      "num_input_tokens_seen": 9596408,
      "step": 12580
    },
    {
      "epoch": 26.164241164241165,
      "grad_norm": 7.831402035662904e-05,
      "learning_rate": 0.23251090361058505,
      "loss": 0.277,
      "num_input_tokens_seen": 9600152,
      "step": 12585
    },
    {
      "epoch": 26.174636174636174,
      "grad_norm": 0.0004367585643194616,
      "learning_rate": 0.23246170481089476,
      "loss": 0.2763,
      "num_input_tokens_seen": 9603992,
      "step": 12590
    },
    {
      "epoch": 26.185031185031185,
      "grad_norm": 0.00038059434155002236,
      "learning_rate": 0.23241249329457317,
      "loss": 0.2798,
      "num_input_tokens_seen": 9607800,
      "step": 12595
    },
    {
      "epoch": 26.195426195426194,
      "grad_norm": 9.448687342228368e-05,
      "learning_rate": 0.23236326906920957,
      "loss": 0.2723,
      "num_input_tokens_seen": 9611704,
      "step": 12600
    },
    {
      "epoch": 26.195426195426194,
      "eval_loss": 0.24951080977916718,
      "eval_runtime": 13.4167,
      "eval_samples_per_second": 63.801,
      "eval_steps_per_second": 15.95,
      "num_input_tokens_seen": 9611704,
      "step": 12600
    },
    {
      "epoch": 26.205821205821206,
      "grad_norm": 0.0008909075404517353,
      "learning_rate": 0.2323140321423948,
      "loss": 0.2801,
      "num_input_tokens_seen": 9615352,
      "step": 12605
    },
    {
      "epoch": 26.216216216216218,
      "grad_norm": 0.0003795934026129544,
      "learning_rate": 0.23226478252172184,
      "loss": 0.2354,
      "num_input_tokens_seen": 9619064,
      "step": 12610
    },
    {
      "epoch": 26.226611226611226,
      "grad_norm": 0.0008331681019626558,
      "learning_rate": 0.23221552021478561,
      "loss": 0.2773,
      "num_input_tokens_seen": 9622904,
      "step": 12615
    },
    {
      "epoch": 26.237006237006238,
      "grad_norm": 7.846447988413274e-05,
      "learning_rate": 0.232166245229183,
      "loss": 0.2573,
      "num_input_tokens_seen": 9626776,
      "step": 12620
    },
    {
      "epoch": 26.247401247401246,
      "grad_norm": 0.00021828798344358802,
      "learning_rate": 0.2321169575725128,
      "loss": 0.271,
      "num_input_tokens_seen": 9630712,
      "step": 12625
    },
    {
      "epoch": 26.257796257796258,
      "grad_norm": 0.0002588498464319855,
      "learning_rate": 0.23206765725237577,
      "loss": 0.2802,
      "num_input_tokens_seen": 9634520,
      "step": 12630
    },
    {
      "epoch": 26.26819126819127,
      "grad_norm": 6.29895948804915e-05,
      "learning_rate": 0.2320183442763747,
      "loss": 0.2517,
      "num_input_tokens_seen": 9638296,
      "step": 12635
    },
    {
      "epoch": 26.27858627858628,
      "grad_norm": 0.00019017315935343504,
      "learning_rate": 0.23196901865211422,
      "loss": 0.277,
      "num_input_tokens_seen": 9642328,
      "step": 12640
    },
    {
      "epoch": 26.28898128898129,
      "grad_norm": 0.0003713190380949527,
      "learning_rate": 0.231919680387201,
      "loss": 0.2782,
      "num_input_tokens_seen": 9646008,
      "step": 12645
    },
    {
      "epoch": 26.2993762993763,
      "grad_norm": 0.0006231119041331112,
      "learning_rate": 0.23187032948924358,
      "loss": 0.2599,
      "num_input_tokens_seen": 9649848,
      "step": 12650
    },
    {
      "epoch": 26.30977130977131,
      "grad_norm": 0.00022300405544228852,
      "learning_rate": 0.23182096596585247,
      "loss": 0.2363,
      "num_input_tokens_seen": 9653592,
      "step": 12655
    },
    {
      "epoch": 26.32016632016632,
      "grad_norm": 0.001334618660621345,
      "learning_rate": 0.23177158982464025,
      "loss": 0.3357,
      "num_input_tokens_seen": 9657464,
      "step": 12660
    },
    {
      "epoch": 26.33056133056133,
      "grad_norm": 0.0001812281843740493,
      "learning_rate": 0.23172220107322122,
      "loss": 0.262,
      "num_input_tokens_seen": 9661144,
      "step": 12665
    },
    {
      "epoch": 26.340956340956343,
      "grad_norm": 0.0002520125708542764,
      "learning_rate": 0.23167279971921184,
      "loss": 0.283,
      "num_input_tokens_seen": 9665048,
      "step": 12670
    },
    {
      "epoch": 26.35135135135135,
      "grad_norm": 0.00042743273661471903,
      "learning_rate": 0.23162338577023034,
      "loss": 0.2877,
      "num_input_tokens_seen": 9668728,
      "step": 12675
    },
    {
      "epoch": 26.361746361746363,
      "grad_norm": 0.0008316849125549197,
      "learning_rate": 0.23157395923389704,
      "loss": 0.2605,
      "num_input_tokens_seen": 9672664,
      "step": 12680
    },
    {
      "epoch": 26.37214137214137,
      "grad_norm": 0.0007197829545475543,
      "learning_rate": 0.2315245201178341,
      "loss": 0.3093,
      "num_input_tokens_seen": 9676760,
      "step": 12685
    },
    {
      "epoch": 26.382536382536383,
      "grad_norm": 0.0007787124486640096,
      "learning_rate": 0.23147506842966564,
      "loss": 0.2814,
      "num_input_tokens_seen": 9680728,
      "step": 12690
    },
    {
      "epoch": 26.39293139293139,
      "grad_norm": 9.129822865361348e-05,
      "learning_rate": 0.23142560417701774,
      "loss": 0.2679,
      "num_input_tokens_seen": 9684440,
      "step": 12695
    },
    {
      "epoch": 26.403326403326403,
      "grad_norm": 0.0009400932467542589,
      "learning_rate": 0.23137612736751845,
      "loss": 0.2754,
      "num_input_tokens_seen": 9688248,
      "step": 12700
    },
    {
      "epoch": 26.413721413721415,
      "grad_norm": 0.0003207334375474602,
      "learning_rate": 0.23132663800879766,
      "loss": 0.234,
      "num_input_tokens_seen": 9692152,
      "step": 12705
    },
    {
      "epoch": 26.424116424116423,
      "grad_norm": 8.331677236128598e-05,
      "learning_rate": 0.2312771361084873,
      "loss": 0.2579,
      "num_input_tokens_seen": 9695800,
      "step": 12710
    },
    {
      "epoch": 26.434511434511435,
      "grad_norm": 0.0001733617827994749,
      "learning_rate": 0.23122762167422112,
      "loss": 0.2591,
      "num_input_tokens_seen": 9699512,
      "step": 12715
    },
    {
      "epoch": 26.444906444906444,
      "grad_norm": 0.0002785191172733903,
      "learning_rate": 0.23117809471363493,
      "loss": 0.2672,
      "num_input_tokens_seen": 9703384,
      "step": 12720
    },
    {
      "epoch": 26.455301455301456,
      "grad_norm": 0.0005105927702970803,
      "learning_rate": 0.23112855523436637,
      "loss": 0.2628,
      "num_input_tokens_seen": 9707032,
      "step": 12725
    },
    {
      "epoch": 26.465696465696467,
      "grad_norm": 0.00028112190193496644,
      "learning_rate": 0.23107900324405511,
      "loss": 0.2648,
      "num_input_tokens_seen": 9711000,
      "step": 12730
    },
    {
      "epoch": 26.476091476091476,
      "grad_norm": 0.0001392694830428809,
      "learning_rate": 0.2310294387503426,
      "loss": 0.2557,
      "num_input_tokens_seen": 9714744,
      "step": 12735
    },
    {
      "epoch": 26.486486486486488,
      "grad_norm": 0.00015754684864077717,
      "learning_rate": 0.23097986176087237,
      "loss": 0.2304,
      "num_input_tokens_seen": 9718712,
      "step": 12740
    },
    {
      "epoch": 26.496881496881496,
      "grad_norm": 0.0005974313244223595,
      "learning_rate": 0.23093027228328986,
      "loss": 0.3189,
      "num_input_tokens_seen": 9722456,
      "step": 12745
    },
    {
      "epoch": 26.507276507276508,
      "grad_norm": 0.0004982036189176142,
      "learning_rate": 0.23088067032524226,
      "loss": 0.2692,
      "num_input_tokens_seen": 9726296,
      "step": 12750
    },
    {
      "epoch": 26.517671517671516,
      "grad_norm": 0.00012270480510778725,
      "learning_rate": 0.23083105589437888,
      "loss": 0.2662,
      "num_input_tokens_seen": 9729976,
      "step": 12755
    },
    {
      "epoch": 26.528066528066528,
      "grad_norm": 0.00018452580843586475,
      "learning_rate": 0.23078142899835094,
      "loss": 0.2667,
      "num_input_tokens_seen": 9733784,
      "step": 12760
    },
    {
      "epoch": 26.53846153846154,
      "grad_norm": 0.00037734053330495954,
      "learning_rate": 0.23073178964481147,
      "loss": 0.2489,
      "num_input_tokens_seen": 9737528,
      "step": 12765
    },
    {
      "epoch": 26.54885654885655,
      "grad_norm": 0.0004967832937836647,
      "learning_rate": 0.2306821378414155,
      "loss": 0.2487,
      "num_input_tokens_seen": 9741368,
      "step": 12770
    },
    {
      "epoch": 26.55925155925156,
      "grad_norm": 0.0004886838141828775,
      "learning_rate": 0.2306324735958199,
      "loss": 0.2509,
      "num_input_tokens_seen": 9745144,
      "step": 12775
    },
    {
      "epoch": 26.56964656964657,
      "grad_norm": 0.0005577904521487653,
      "learning_rate": 0.23058279691568362,
      "loss": 0.2528,
      "num_input_tokens_seen": 9748984,
      "step": 12780
    },
    {
      "epoch": 26.58004158004158,
      "grad_norm": 0.00034506028168834746,
      "learning_rate": 0.23053310780866745,
      "loss": 0.2582,
      "num_input_tokens_seen": 9752728,
      "step": 12785
    },
    {
      "epoch": 26.59043659043659,
      "grad_norm": 0.00026751452242024243,
      "learning_rate": 0.23048340628243397,
      "loss": 0.2529,
      "num_input_tokens_seen": 9756568,
      "step": 12790
    },
    {
      "epoch": 26.6008316008316,
      "grad_norm": 0.00030433418578468263,
      "learning_rate": 0.23043369234464783,
      "loss": 0.2541,
      "num_input_tokens_seen": 9760440,
      "step": 12795
    },
    {
      "epoch": 26.611226611226613,
      "grad_norm": 0.00019172928296029568,
      "learning_rate": 0.2303839660029755,
      "loss": 0.2621,
      "num_input_tokens_seen": 9764344,
      "step": 12800
    },
    {
      "epoch": 26.611226611226613,
      "eval_loss": 0.24915602803230286,
      "eval_runtime": 13.5127,
      "eval_samples_per_second": 63.348,
      "eval_steps_per_second": 15.837,
      "num_input_tokens_seen": 9764344,
      "step": 12800
    },
    {
      "epoch": 26.62162162162162,
      "grad_norm": 0.00030297087505459785,
      "learning_rate": 0.23033422726508548,
      "loss": 0.2427,
      "num_input_tokens_seen": 9768280,
      "step": 12805
    },
    {
      "epoch": 26.632016632016633,
      "grad_norm": 0.0012150837574154139,
      "learning_rate": 0.23028447613864808,
      "loss": 0.26,
      "num_input_tokens_seen": 9772024,
      "step": 12810
    },
    {
      "epoch": 26.64241164241164,
      "grad_norm": 0.0001291695807594806,
      "learning_rate": 0.2302347126313355,
      "loss": 0.276,
      "num_input_tokens_seen": 9775960,
      "step": 12815
    },
    {
      "epoch": 26.652806652806653,
      "grad_norm": 0.0006311475881375372,
      "learning_rate": 0.23018493675082197,
      "loss": 0.2812,
      "num_input_tokens_seen": 9779672,
      "step": 12820
    },
    {
      "epoch": 26.66320166320166,
      "grad_norm": 0.0001517148339189589,
      "learning_rate": 0.2301351485047835,
      "loss": 0.2699,
      "num_input_tokens_seen": 9783512,
      "step": 12825
    },
    {
      "epoch": 26.673596673596673,
      "grad_norm": 0.00045714061707258224,
      "learning_rate": 0.23008534790089813,
      "loss": 0.2958,
      "num_input_tokens_seen": 9787256,
      "step": 12830
    },
    {
      "epoch": 26.683991683991685,
      "grad_norm": 0.0007368313963524997,
      "learning_rate": 0.2300355349468457,
      "loss": 0.2443,
      "num_input_tokens_seen": 9791032,
      "step": 12835
    },
    {
      "epoch": 26.694386694386694,
      "grad_norm": 0.0004565578419715166,
      "learning_rate": 0.22998570965030793,
      "loss": 0.2571,
      "num_input_tokens_seen": 9794808,
      "step": 12840
    },
    {
      "epoch": 26.704781704781706,
      "grad_norm": 9.455796680413187e-05,
      "learning_rate": 0.22993587201896862,
      "loss": 0.2644,
      "num_input_tokens_seen": 9798616,
      "step": 12845
    },
    {
      "epoch": 26.715176715176714,
      "grad_norm": 0.0002607906353659928,
      "learning_rate": 0.2298860220605133,
      "loss": 0.2785,
      "num_input_tokens_seen": 9802584,
      "step": 12850
    },
    {
      "epoch": 26.725571725571726,
      "grad_norm": 0.0001764680491760373,
      "learning_rate": 0.22983615978262942,
      "loss": 0.2805,
      "num_input_tokens_seen": 9806488,
      "step": 12855
    },
    {
      "epoch": 26.735966735966738,
      "grad_norm": 0.0003495899436529726,
      "learning_rate": 0.22978628519300648,
      "loss": 0.2658,
      "num_input_tokens_seen": 9810520,
      "step": 12860
    },
    {
      "epoch": 26.746361746361746,
      "grad_norm": 0.0006536575965583324,
      "learning_rate": 0.22973639829933568,
      "loss": 0.2675,
      "num_input_tokens_seen": 9814328,
      "step": 12865
    },
    {
      "epoch": 26.756756756756758,
      "grad_norm": 0.00024923565797507763,
      "learning_rate": 0.22968649910931027,
      "loss": 0.2925,
      "num_input_tokens_seen": 9818232,
      "step": 12870
    },
    {
      "epoch": 26.767151767151766,
      "grad_norm": 0.0004650003102142364,
      "learning_rate": 0.22963658763062528,
      "loss": 0.2707,
      "num_input_tokens_seen": 9822072,
      "step": 12875
    },
    {
      "epoch": 26.777546777546778,
      "grad_norm": 0.0009208290721289814,
      "learning_rate": 0.22958666387097765,
      "loss": 0.2629,
      "num_input_tokens_seen": 9825816,
      "step": 12880
    },
    {
      "epoch": 26.787941787941786,
      "grad_norm": 9.822161518968642e-05,
      "learning_rate": 0.22953672783806633,
      "loss": 0.2524,
      "num_input_tokens_seen": 9829528,
      "step": 12885
    },
    {
      "epoch": 26.7983367983368,
      "grad_norm": 0.0002308362745679915,
      "learning_rate": 0.22948677953959207,
      "loss": 0.297,
      "num_input_tokens_seen": 9833400,
      "step": 12890
    },
    {
      "epoch": 26.80873180873181,
      "grad_norm": 0.0004099719226360321,
      "learning_rate": 0.2294368189832575,
      "loss": 0.2599,
      "num_input_tokens_seen": 9837176,
      "step": 12895
    },
    {
      "epoch": 26.81912681912682,
      "grad_norm": 0.00016004088683985174,
      "learning_rate": 0.2293868461767672,
      "loss": 0.2691,
      "num_input_tokens_seen": 9840856,
      "step": 12900
    },
    {
      "epoch": 26.82952182952183,
      "grad_norm": 0.00046312747872434556,
      "learning_rate": 0.22933686112782758,
      "loss": 0.2744,
      "num_input_tokens_seen": 9844792,
      "step": 12905
    },
    {
      "epoch": 26.83991683991684,
      "grad_norm": 0.0005662898765876889,
      "learning_rate": 0.22928686384414698,
      "loss": 0.2773,
      "num_input_tokens_seen": 9848600,
      "step": 12910
    },
    {
      "epoch": 26.85031185031185,
      "grad_norm": 0.00025487018865533173,
      "learning_rate": 0.22923685433343552,
      "loss": 0.2741,
      "num_input_tokens_seen": 9852248,
      "step": 12915
    },
    {
      "epoch": 26.86070686070686,
      "grad_norm": 0.00038627488538622856,
      "learning_rate": 0.22918683260340542,
      "loss": 0.279,
      "num_input_tokens_seen": 9856216,
      "step": 12920
    },
    {
      "epoch": 26.87110187110187,
      "grad_norm": 0.00028804840985685587,
      "learning_rate": 0.2291367986617706,
      "loss": 0.2785,
      "num_input_tokens_seen": 9859992,
      "step": 12925
    },
    {
      "epoch": 26.881496881496883,
      "grad_norm": 0.0003801829880103469,
      "learning_rate": 0.22908675251624697,
      "loss": 0.2818,
      "num_input_tokens_seen": 9863992,
      "step": 12930
    },
    {
      "epoch": 26.89189189189189,
      "grad_norm": 8.225213241530582e-05,
      "learning_rate": 0.22903669417455216,
      "loss": 0.2544,
      "num_input_tokens_seen": 9867928,
      "step": 12935
    },
    {
      "epoch": 26.902286902286903,
      "grad_norm": 0.00010724605090217665,
      "learning_rate": 0.22898662364440592,
      "loss": 0.2732,
      "num_input_tokens_seen": 9871672,
      "step": 12940
    },
    {
      "epoch": 26.91268191268191,
      "grad_norm": 0.00022391023230738938,
      "learning_rate": 0.2289365409335297,
      "loss": 0.2939,
      "num_input_tokens_seen": 9875448,
      "step": 12945
    },
    {
      "epoch": 26.923076923076923,
      "grad_norm": 0.0007258779369294643,
      "learning_rate": 0.2288864460496469,
      "loss": 0.2779,
      "num_input_tokens_seen": 9879256,
      "step": 12950
    },
    {
      "epoch": 26.933471933471935,
      "grad_norm": 0.0004381911421660334,
      "learning_rate": 0.22883633900048272,
      "loss": 0.2625,
      "num_input_tokens_seen": 9883000,
      "step": 12955
    },
    {
      "epoch": 26.943866943866944,
      "grad_norm": 0.0005304153892211616,
      "learning_rate": 0.2287862197937644,
      "loss": 0.2793,
      "num_input_tokens_seen": 9886680,
      "step": 12960
    },
    {
      "epoch": 26.954261954261955,
      "grad_norm": 0.00025625419220887125,
      "learning_rate": 0.2287360884372209,
      "loss": 0.2455,
      "num_input_tokens_seen": 9890584,
      "step": 12965
    },
    {
      "epoch": 26.964656964656964,
      "grad_norm": 0.00018429012561682612,
      "learning_rate": 0.22868594493858307,
      "loss": 0.2798,
      "num_input_tokens_seen": 9894360,
      "step": 12970
    },
    {
      "epoch": 26.975051975051976,
      "grad_norm": 0.000454979483038187,
      "learning_rate": 0.2286357893055837,
      "loss": 0.2829,
      "num_input_tokens_seen": 9898200,
      "step": 12975
    },
    {
      "epoch": 26.985446985446984,
      "grad_norm": 0.0005798538913950324,
      "learning_rate": 0.22858562154595746,
      "loss": 0.2734,
      "num_input_tokens_seen": 9902008,
      "step": 12980
    },
    {
      "epoch": 26.995841995841996,
      "grad_norm": 0.0006047403439879417,
      "learning_rate": 0.22853544166744078,
      "loss": 0.2668,
      "num_input_tokens_seen": 9905688,
      "step": 12985
    },
    {
      "epoch": 27.006237006237008,
      "grad_norm": 0.0003344781289342791,
      "learning_rate": 0.22848524967777206,
      "loss": 0.26,
      "num_input_tokens_seen": 9909512,
      "step": 12990
    },
    {
      "epoch": 27.016632016632016,
      "grad_norm": 0.00031634216429665685,
      "learning_rate": 0.22843504558469152,
      "loss": 0.2729,
      "num_input_tokens_seen": 9913160,
      "step": 12995
    },
    {
      "epoch": 27.027027027027028,
      "grad_norm": 0.0003347745514474809,
      "learning_rate": 0.2283848293959413,
      "loss": 0.2696,
      "num_input_tokens_seen": 9917064,
      "step": 13000
    },
    {
      "epoch": 27.027027027027028,
      "eval_loss": 0.25639572739601135,
      "eval_runtime": 13.4265,
      "eval_samples_per_second": 63.755,
      "eval_steps_per_second": 15.939,
      "num_input_tokens_seen": 9917064,
      "step": 13000
    },
    {
      "epoch": 27.037422037422036,
      "grad_norm": 0.0002382049133302644,
      "learning_rate": 0.22833460111926532,
      "loss": 0.2637,
      "num_input_tokens_seen": 9920744,
      "step": 13005
    },
    {
      "epoch": 27.04781704781705,
      "grad_norm": 0.0004500375653151423,
      "learning_rate": 0.22828436076240946,
      "loss": 0.2537,
      "num_input_tokens_seen": 9924552,
      "step": 13010
    },
    {
      "epoch": 27.058212058212057,
      "grad_norm": 0.0002097799879265949,
      "learning_rate": 0.22823410833312135,
      "loss": 0.2492,
      "num_input_tokens_seen": 9928296,
      "step": 13015
    },
    {
      "epoch": 27.06860706860707,
      "grad_norm": 0.00050159334205091,
      "learning_rate": 0.2281838438391506,
      "loss": 0.2639,
      "num_input_tokens_seen": 9932136,
      "step": 13020
    },
    {
      "epoch": 27.07900207900208,
      "grad_norm": 0.0001650533522479236,
      "learning_rate": 0.22813356728824863,
      "loss": 0.2499,
      "num_input_tokens_seen": 9935784,
      "step": 13025
    },
    {
      "epoch": 27.08939708939709,
      "grad_norm": 0.00020827454864047468,
      "learning_rate": 0.2280832786881687,
      "loss": 0.2349,
      "num_input_tokens_seen": 9939464,
      "step": 13030
    },
    {
      "epoch": 27.0997920997921,
      "grad_norm": 0.0002079051046166569,
      "learning_rate": 0.22803297804666592,
      "loss": 0.2995,
      "num_input_tokens_seen": 9943336,
      "step": 13035
    },
    {
      "epoch": 27.11018711018711,
      "grad_norm": 0.0003163835790473968,
      "learning_rate": 0.22798266537149728,
      "loss": 0.2871,
      "num_input_tokens_seen": 9947112,
      "step": 13040
    },
    {
      "epoch": 27.12058212058212,
      "grad_norm": 0.0006993371061980724,
      "learning_rate": 0.22793234067042167,
      "loss": 0.2769,
      "num_input_tokens_seen": 9950792,
      "step": 13045
    },
    {
      "epoch": 27.13097713097713,
      "grad_norm": 0.000580761581659317,
      "learning_rate": 0.22788200395119979,
      "loss": 0.2916,
      "num_input_tokens_seen": 9954824,
      "step": 13050
    },
    {
      "epoch": 27.14137214137214,
      "grad_norm": 0.0004978504148311913,
      "learning_rate": 0.2278316552215942,
      "loss": 0.2711,
      "num_input_tokens_seen": 9958600,
      "step": 13055
    },
    {
      "epoch": 27.151767151767153,
      "grad_norm": 0.0003250467998441309,
      "learning_rate": 0.22778129448936918,
      "loss": 0.2676,
      "num_input_tokens_seen": 9962312,
      "step": 13060
    },
    {
      "epoch": 27.16216216216216,
      "grad_norm": 0.00018614131840877235,
      "learning_rate": 0.22773092176229118,
      "loss": 0.2689,
      "num_input_tokens_seen": 9966056,
      "step": 13065
    },
    {
      "epoch": 27.172557172557173,
      "grad_norm": 7.626097794855013e-05,
      "learning_rate": 0.22768053704812816,
      "loss": 0.2613,
      "num_input_tokens_seen": 9969928,
      "step": 13070
    },
    {
      "epoch": 27.18295218295218,
      "grad_norm": 0.0005300771445035934,
      "learning_rate": 0.22763014035465018,
      "loss": 0.2463,
      "num_input_tokens_seen": 9973736,
      "step": 13075
    },
    {
      "epoch": 27.193347193347194,
      "grad_norm": 0.00015333088231272995,
      "learning_rate": 0.22757973168962892,
      "loss": 0.2422,
      "num_input_tokens_seen": 9977448,
      "step": 13080
    },
    {
      "epoch": 27.203742203742205,
      "grad_norm": 0.00033755411277525127,
      "learning_rate": 0.22752931106083818,
      "loss": 0.2873,
      "num_input_tokens_seen": 9981512,
      "step": 13085
    },
    {
      "epoch": 27.214137214137214,
      "grad_norm": 0.00017177635163534433,
      "learning_rate": 0.22747887847605341,
      "loss": 0.262,
      "num_input_tokens_seen": 9985384,
      "step": 13090
    },
    {
      "epoch": 27.224532224532226,
      "grad_norm": 0.0002843157562892884,
      "learning_rate": 0.22742843394305184,
      "loss": 0.2592,
      "num_input_tokens_seen": 9989256,
      "step": 13095
    },
    {
      "epoch": 27.234927234927234,
      "grad_norm": 0.0003438879793975502,
      "learning_rate": 0.22737797746961272,
      "loss": 0.2713,
      "num_input_tokens_seen": 9993000,
      "step": 13100
    },
    {
      "epoch": 27.245322245322246,
      "grad_norm": 0.0005375894252210855,
      "learning_rate": 0.22732750906351712,
      "loss": 0.2554,
      "num_input_tokens_seen": 9996872,
      "step": 13105
    },
    {
      "epoch": 27.255717255717254,
      "grad_norm": 0.00018800931866280735,
      "learning_rate": 0.22727702873254785,
      "loss": 0.2754,
      "num_input_tokens_seen": 10000648,
      "step": 13110
    },
    {
      "epoch": 27.266112266112266,
      "grad_norm": 0.0009014391689561307,
      "learning_rate": 0.22722653648448968,
      "loss": 0.2667,
      "num_input_tokens_seen": 10004264,
      "step": 13115
    },
    {
      "epoch": 27.276507276507278,
      "grad_norm": 0.0004581070097628981,
      "learning_rate": 0.22717603232712902,
      "loss": 0.2675,
      "num_input_tokens_seen": 10008136,
      "step": 13120
    },
    {
      "epoch": 27.286902286902286,
      "grad_norm": 0.00024369248421862721,
      "learning_rate": 0.22712551626825436,
      "loss": 0.2313,
      "num_input_tokens_seen": 10012072,
      "step": 13125
    },
    {
      "epoch": 27.2972972972973,
      "grad_norm": 0.00024073383247014135,
      "learning_rate": 0.2270749883156559,
      "loss": 0.3025,
      "num_input_tokens_seen": 10015848,
      "step": 13130
    },
    {
      "epoch": 27.307692307692307,
      "grad_norm": 0.0002042127016466111,
      "learning_rate": 0.22702444847712563,
      "loss": 0.2685,
      "num_input_tokens_seen": 10019560,
      "step": 13135
    },
    {
      "epoch": 27.31808731808732,
      "grad_norm": 0.0004922857624478638,
      "learning_rate": 0.22697389676045743,
      "loss": 0.2796,
      "num_input_tokens_seen": 10023496,
      "step": 13140
    },
    {
      "epoch": 27.328482328482327,
      "grad_norm": 0.0010805350029841065,
      "learning_rate": 0.22692333317344704,
      "loss": 0.2942,
      "num_input_tokens_seen": 10027304,
      "step": 13145
    },
    {
      "epoch": 27.33887733887734,
      "grad_norm": 0.0006932936958037317,
      "learning_rate": 0.22687275772389198,
      "loss": 0.2934,
      "num_input_tokens_seen": 10031048,
      "step": 13150
    },
    {
      "epoch": 27.34927234927235,
      "grad_norm": 8.872978651197627e-05,
      "learning_rate": 0.22682217041959168,
      "loss": 0.2689,
      "num_input_tokens_seen": 10034792,
      "step": 13155
    },
    {
      "epoch": 27.35966735966736,
      "grad_norm": 0.0003817516262643039,
      "learning_rate": 0.2267715712683473,
      "loss": 0.2648,
      "num_input_tokens_seen": 10038568,
      "step": 13160
    },
    {
      "epoch": 27.37006237006237,
      "grad_norm": 0.0007749811629764736,
      "learning_rate": 0.22672096027796182,
      "loss": 0.2695,
      "num_input_tokens_seen": 10042280,
      "step": 13165
    },
    {
      "epoch": 27.38045738045738,
      "grad_norm": 0.00026218913262709975,
      "learning_rate": 0.22667033745624016,
      "loss": 0.2647,
      "num_input_tokens_seen": 10046088,
      "step": 13170
    },
    {
      "epoch": 27.39085239085239,
      "grad_norm": 0.0003157384635414928,
      "learning_rate": 0.22661970281098895,
      "loss": 0.2825,
      "num_input_tokens_seen": 10049864,
      "step": 13175
    },
    {
      "epoch": 27.401247401247403,
      "grad_norm": 0.00024935463443398476,
      "learning_rate": 0.22656905635001667,
      "loss": 0.2646,
      "num_input_tokens_seen": 10053576,
      "step": 13180
    },
    {
      "epoch": 27.41164241164241,
      "grad_norm": 0.000323864835081622,
      "learning_rate": 0.2265183980811337,
      "loss": 0.2819,
      "num_input_tokens_seen": 10057320,
      "step": 13185
    },
    {
      "epoch": 27.422037422037423,
      "grad_norm": 0.0004008161195088178,
      "learning_rate": 0.22646772801215218,
      "loss": 0.2521,
      "num_input_tokens_seen": 10061000,
      "step": 13190
    },
    {
      "epoch": 27.43243243243243,
      "grad_norm": 0.0002013692574109882,
      "learning_rate": 0.22641704615088598,
      "loss": 0.237,
      "num_input_tokens_seen": 10064712,
      "step": 13195
    },
    {
      "epoch": 27.442827442827443,
      "grad_norm": 0.00044108665315434337,
      "learning_rate": 0.22636635250515103,
      "loss": 0.2466,
      "num_input_tokens_seen": 10068520,
      "step": 13200
    },
    {
      "epoch": 27.442827442827443,
      "eval_loss": 0.25269845128059387,
      "eval_runtime": 13.4039,
      "eval_samples_per_second": 63.862,
      "eval_steps_per_second": 15.966,
      "num_input_tokens_seen": 10068520,
      "step": 13200
    },
    {
      "epoch": 27.453222453222452,
      "grad_norm": 0.00015968404477462173,
      "learning_rate": 0.2263156470827648,
      "loss": 0.267,
      "num_input_tokens_seen": 10072328,
      "step": 13205
    },
    {
      "epoch": 27.463617463617464,
      "grad_norm": 0.0003435423714108765,
      "learning_rate": 0.22626492989154678,
      "loss": 0.2779,
      "num_input_tokens_seen": 10076136,
      "step": 13210
    },
    {
      "epoch": 27.474012474012476,
      "grad_norm": 0.0005049800965934992,
      "learning_rate": 0.22621420093931813,
      "loss": 0.2752,
      "num_input_tokens_seen": 10079912,
      "step": 13215
    },
    {
      "epoch": 27.484407484407484,
      "grad_norm": 0.0006863938760943711,
      "learning_rate": 0.22616346023390194,
      "loss": 0.2753,
      "num_input_tokens_seen": 10083816,
      "step": 13220
    },
    {
      "epoch": 27.494802494802496,
      "grad_norm": 0.00016316442633979023,
      "learning_rate": 0.22611270778312306,
      "loss": 0.2788,
      "num_input_tokens_seen": 10087560,
      "step": 13225
    },
    {
      "epoch": 27.505197505197504,
      "grad_norm": 5.5427837651222944e-05,
      "learning_rate": 0.2260619435948081,
      "loss": 0.2638,
      "num_input_tokens_seen": 10091432,
      "step": 13230
    },
    {
      "epoch": 27.515592515592516,
      "grad_norm": 8.78640275914222e-05,
      "learning_rate": 0.22601116767678567,
      "loss": 0.2536,
      "num_input_tokens_seen": 10095368,
      "step": 13235
    },
    {
      "epoch": 27.525987525987524,
      "grad_norm": 0.00015075386909302324,
      "learning_rate": 0.2259603800368859,
      "loss": 0.2693,
      "num_input_tokens_seen": 10099176,
      "step": 13240
    },
    {
      "epoch": 27.536382536382536,
      "grad_norm": 0.00029906249255873263,
      "learning_rate": 0.22590958068294098,
      "loss": 0.2655,
      "num_input_tokens_seen": 10102824,
      "step": 13245
    },
    {
      "epoch": 27.546777546777548,
      "grad_norm": 0.00022729103511665016,
      "learning_rate": 0.22585876962278478,
      "loss": 0.2421,
      "num_input_tokens_seen": 10106696,
      "step": 13250
    },
    {
      "epoch": 27.557172557172557,
      "grad_norm": 0.0001592343469383195,
      "learning_rate": 0.22580794686425298,
      "loss": 0.2426,
      "num_input_tokens_seen": 10110568,
      "step": 13255
    },
    {
      "epoch": 27.56756756756757,
      "grad_norm": 0.0001673367660259828,
      "learning_rate": 0.22575711241518312,
      "loss": 0.2883,
      "num_input_tokens_seen": 10114376,
      "step": 13260
    },
    {
      "epoch": 27.577962577962577,
      "grad_norm": 0.0004015630984213203,
      "learning_rate": 0.22570626628341453,
      "loss": 0.2657,
      "num_input_tokens_seen": 10118152,
      "step": 13265
    },
    {
      "epoch": 27.58835758835759,
      "grad_norm": 9.43072373047471e-05,
      "learning_rate": 0.22565540847678828,
      "loss": 0.2819,
      "num_input_tokens_seen": 10121896,
      "step": 13270
    },
    {
      "epoch": 27.598752598752597,
      "grad_norm": 0.00030626184889115393,
      "learning_rate": 0.2256045390031473,
      "loss": 0.2721,
      "num_input_tokens_seen": 10125608,
      "step": 13275
    },
    {
      "epoch": 27.60914760914761,
      "grad_norm": 0.00025307227042503655,
      "learning_rate": 0.22555365787033627,
      "loss": 0.2644,
      "num_input_tokens_seen": 10129512,
      "step": 13280
    },
    {
      "epoch": 27.61954261954262,
      "grad_norm": 0.00017517949163448066,
      "learning_rate": 0.22550276508620173,
      "loss": 0.2628,
      "num_input_tokens_seen": 10133352,
      "step": 13285
    },
    {
      "epoch": 27.62993762993763,
      "grad_norm": 6.868541822768748e-05,
      "learning_rate": 0.22545186065859202,
      "loss": 0.2735,
      "num_input_tokens_seen": 10137064,
      "step": 13290
    },
    {
      "epoch": 27.64033264033264,
      "grad_norm": 0.0003172597207594663,
      "learning_rate": 0.2254009445953572,
      "loss": 0.2764,
      "num_input_tokens_seen": 10140968,
      "step": 13295
    },
    {
      "epoch": 27.65072765072765,
      "grad_norm": 0.00017866762937046587,
      "learning_rate": 0.22535001690434917,
      "loss": 0.2689,
      "num_input_tokens_seen": 10144872,
      "step": 13300
    },
    {
      "epoch": 27.66112266112266,
      "grad_norm": 0.0004021819040644914,
      "learning_rate": 0.22529907759342163,
      "loss": 0.2643,
      "num_input_tokens_seen": 10148680,
      "step": 13305
    },
    {
      "epoch": 27.671517671517673,
      "grad_norm": 0.00021828350145369768,
      "learning_rate": 0.22524812667043007,
      "loss": 0.2824,
      "num_input_tokens_seen": 10152712,
      "step": 13310
    },
    {
      "epoch": 27.68191268191268,
      "grad_norm": 0.0008664872730150819,
      "learning_rate": 0.22519716414323177,
      "loss": 0.2692,
      "num_input_tokens_seen": 10156456,
      "step": 13315
    },
    {
      "epoch": 27.692307692307693,
      "grad_norm": 9.309778397437185e-05,
      "learning_rate": 0.22514619001968567,
      "loss": 0.2727,
      "num_input_tokens_seen": 10160168,
      "step": 13320
    },
    {
      "epoch": 27.7027027027027,
      "grad_norm": 0.0008709369576536119,
      "learning_rate": 0.2250952043076528,
      "loss": 0.2531,
      "num_input_tokens_seen": 10164104,
      "step": 13325
    },
    {
      "epoch": 27.713097713097714,
      "grad_norm": 0.0003518385929055512,
      "learning_rate": 0.2250442070149957,
      "loss": 0.2565,
      "num_input_tokens_seen": 10167720,
      "step": 13330
    },
    {
      "epoch": 27.723492723492722,
      "grad_norm": 0.0005521838320419192,
      "learning_rate": 0.22499319814957885,
      "loss": 0.2522,
      "num_input_tokens_seen": 10171400,
      "step": 13335
    },
    {
      "epoch": 27.733887733887734,
      "grad_norm": 6.336794467642903e-05,
      "learning_rate": 0.2249421777192684,
      "loss": 0.3081,
      "num_input_tokens_seen": 10175208,
      "step": 13340
    },
    {
      "epoch": 27.744282744282746,
      "grad_norm": 0.0005448810989037156,
      "learning_rate": 0.22489114573193236,
      "loss": 0.2875,
      "num_input_tokens_seen": 10179112,
      "step": 13345
    },
    {
      "epoch": 27.754677754677754,
      "grad_norm": 6.213696178747341e-05,
      "learning_rate": 0.2248401021954405,
      "loss": 0.2788,
      "num_input_tokens_seen": 10182984,
      "step": 13350
    },
    {
      "epoch": 27.765072765072766,
      "grad_norm": 0.00028042099438607693,
      "learning_rate": 0.22478904711766443,
      "loss": 0.2714,
      "num_input_tokens_seen": 10186920,
      "step": 13355
    },
    {
      "epoch": 27.775467775467774,
      "grad_norm": 0.00021156042930670083,
      "learning_rate": 0.22473798050647734,
      "loss": 0.2738,
      "num_input_tokens_seen": 10190824,
      "step": 13360
    },
    {
      "epoch": 27.785862785862786,
      "grad_norm": 9.774912905413657e-05,
      "learning_rate": 0.22468690236975453,
      "loss": 0.2632,
      "num_input_tokens_seen": 10194664,
      "step": 13365
    },
    {
      "epoch": 27.796257796257795,
      "grad_norm": 0.0005764098023064435,
      "learning_rate": 0.22463581271537272,
      "loss": 0.2857,
      "num_input_tokens_seen": 10198440,
      "step": 13370
    },
    {
      "epoch": 27.806652806652806,
      "grad_norm": 0.0001233753573615104,
      "learning_rate": 0.22458471155121076,
      "loss": 0.2803,
      "num_input_tokens_seen": 10202280,
      "step": 13375
    },
    {
      "epoch": 27.81704781704782,
      "grad_norm": 0.00018434102821629494,
      "learning_rate": 0.2245335988851489,
      "loss": 0.2679,
      "num_input_tokens_seen": 10206056,
      "step": 13380
    },
    {
      "epoch": 27.827442827442827,
      "grad_norm": 0.0005305905360728502,
      "learning_rate": 0.2244824747250695,
      "loss": 0.2676,
      "num_input_tokens_seen": 10209800,
      "step": 13385
    },
    {
      "epoch": 27.83783783783784,
      "grad_norm": 0.00022734608501195908,
      "learning_rate": 0.22443133907885646,
      "loss": 0.2796,
      "num_input_tokens_seen": 10213576,
      "step": 13390
    },
    {
      "epoch": 27.848232848232847,
      "grad_norm": 0.0002604887413326651,
      "learning_rate": 0.22438019195439557,
      "loss": 0.2626,
      "num_input_tokens_seen": 10217352,
      "step": 13395
    },
    {
      "epoch": 27.85862785862786,
      "grad_norm": 0.00034578944905661047,
      "learning_rate": 0.22432903335957435,
      "loss": 0.283,
      "num_input_tokens_seen": 10221224,
      "step": 13400
    },
    {
      "epoch": 27.85862785862786,
      "eval_loss": 0.2568046748638153,
      "eval_runtime": 13.4539,
      "eval_samples_per_second": 63.625,
      "eval_steps_per_second": 15.906,
      "num_input_tokens_seen": 10221224,
      "step": 13400
    },
    {
      "epoch": 27.86902286902287,
      "grad_norm": 0.0003184888046234846,
      "learning_rate": 0.22427786330228214,
      "loss": 0.2681,
      "num_input_tokens_seen": 10225096,
      "step": 13405
    },
    {
      "epoch": 27.87941787941788,
      "grad_norm": 0.00016574117762502283,
      "learning_rate": 0.22422668179040997,
      "loss": 0.2694,
      "num_input_tokens_seen": 10228840,
      "step": 13410
    },
    {
      "epoch": 27.88981288981289,
      "grad_norm": 0.0006649466813541949,
      "learning_rate": 0.2241754888318507,
      "loss": 0.2347,
      "num_input_tokens_seen": 10232648,
      "step": 13415
    },
    {
      "epoch": 27.9002079002079,
      "grad_norm": 0.0010515139438211918,
      "learning_rate": 0.22412428443449886,
      "loss": 0.2817,
      "num_input_tokens_seen": 10236680,
      "step": 13420
    },
    {
      "epoch": 27.91060291060291,
      "grad_norm": 0.00011317091411910951,
      "learning_rate": 0.22407306860625087,
      "loss": 0.2779,
      "num_input_tokens_seen": 10240392,
      "step": 13425
    },
    {
      "epoch": 27.92099792099792,
      "grad_norm": 0.00038695023977197707,
      "learning_rate": 0.22402184135500483,
      "loss": 0.2738,
      "num_input_tokens_seen": 10244232,
      "step": 13430
    },
    {
      "epoch": 27.93139293139293,
      "grad_norm": 0.0007046873797662556,
      "learning_rate": 0.22397060268866067,
      "loss": 0.2757,
      "num_input_tokens_seen": 10247976,
      "step": 13435
    },
    {
      "epoch": 27.941787941787943,
      "grad_norm": 0.000474357046186924,
      "learning_rate": 0.22391935261511994,
      "loss": 0.2508,
      "num_input_tokens_seen": 10251784,
      "step": 13440
    },
    {
      "epoch": 27.95218295218295,
      "grad_norm": 0.00031123822554945946,
      "learning_rate": 0.22386809114228615,
      "loss": 0.2569,
      "num_input_tokens_seen": 10255656,
      "step": 13445
    },
    {
      "epoch": 27.962577962577964,
      "grad_norm": 0.00018421078857500106,
      "learning_rate": 0.22381681827806446,
      "loss": 0.2427,
      "num_input_tokens_seen": 10259400,
      "step": 13450
    },
    {
      "epoch": 27.972972972972972,
      "grad_norm": 4.176851871307008e-05,
      "learning_rate": 0.22376553403036173,
      "loss": 0.2832,
      "num_input_tokens_seen": 10263272,
      "step": 13455
    },
    {
      "epoch": 27.983367983367984,
      "grad_norm": 0.0004083602107129991,
      "learning_rate": 0.22371423840708662,
      "loss": 0.255,
      "num_input_tokens_seen": 10267144,
      "step": 13460
    },
    {
      "epoch": 27.993762993762992,
      "grad_norm": 0.00045082971337251365,
      "learning_rate": 0.22366293141614962,
      "loss": 0.277,
      "num_input_tokens_seen": 10271048,
      "step": 13465
    },
    {
      "epoch": 28.004158004158004,
      "grad_norm": 0.0003402531147003174,
      "learning_rate": 0.22361161306546287,
      "loss": 0.2765,
      "num_input_tokens_seen": 10274840,
      "step": 13470
    },
    {
      "epoch": 28.014553014553016,
      "grad_norm": 0.0006915408303029835,
      "learning_rate": 0.22356028336294037,
      "loss": 0.2627,
      "num_input_tokens_seen": 10278488,
      "step": 13475
    },
    {
      "epoch": 28.024948024948024,
      "grad_norm": 0.00031890012905932963,
      "learning_rate": 0.2235089423164977,
      "loss": 0.2674,
      "num_input_tokens_seen": 10282328,
      "step": 13480
    },
    {
      "epoch": 28.035343035343036,
      "grad_norm": 0.00023372222494799644,
      "learning_rate": 0.22345758993405243,
      "loss": 0.2429,
      "num_input_tokens_seen": 10286008,
      "step": 13485
    },
    {
      "epoch": 28.045738045738045,
      "grad_norm": 0.00025254394859075546,
      "learning_rate": 0.2234062262235236,
      "loss": 0.2541,
      "num_input_tokens_seen": 10289720,
      "step": 13490
    },
    {
      "epoch": 28.056133056133056,
      "grad_norm": 6.45907421130687e-05,
      "learning_rate": 0.22335485119283222,
      "loss": 0.2674,
      "num_input_tokens_seen": 10293400,
      "step": 13495
    },
    {
      "epoch": 28.066528066528065,
      "grad_norm": 0.0006500897579826415,
      "learning_rate": 0.22330346484990093,
      "loss": 0.2743,
      "num_input_tokens_seen": 10297144,
      "step": 13500
    },
    {
      "epoch": 28.076923076923077,
      "grad_norm": 0.00016336972475983202,
      "learning_rate": 0.22325206720265425,
      "loss": 0.2645,
      "num_input_tokens_seen": 10301208,
      "step": 13505
    },
    {
      "epoch": 28.08731808731809,
      "grad_norm": 0.00020031094027217478,
      "learning_rate": 0.2232006582590182,
      "loss": 0.2799,
      "num_input_tokens_seen": 10305016,
      "step": 13510
    },
    {
      "epoch": 28.097713097713097,
      "grad_norm": 0.00023100509133655578,
      "learning_rate": 0.22314923802692077,
      "loss": 0.2546,
      "num_input_tokens_seen": 10308792,
      "step": 13515
    },
    {
      "epoch": 28.10810810810811,
      "grad_norm": 0.000627058616373688,
      "learning_rate": 0.22309780651429156,
      "loss": 0.2682,
      "num_input_tokens_seen": 10312600,
      "step": 13520
    },
    {
      "epoch": 28.118503118503117,
      "grad_norm": 0.00032175221713259816,
      "learning_rate": 0.22304636372906203,
      "loss": 0.2452,
      "num_input_tokens_seen": 10316408,
      "step": 13525
    },
    {
      "epoch": 28.12889812889813,
      "grad_norm": 0.0003232719318475574,
      "learning_rate": 0.22299490967916522,
      "loss": 0.2879,
      "num_input_tokens_seen": 10320408,
      "step": 13530
    },
    {
      "epoch": 28.13929313929314,
      "grad_norm": 0.00021448101324494928,
      "learning_rate": 0.22294344437253602,
      "loss": 0.2762,
      "num_input_tokens_seen": 10324280,
      "step": 13535
    },
    {
      "epoch": 28.14968814968815,
      "grad_norm": 0.0006172165740281343,
      "learning_rate": 0.22289196781711101,
      "loss": 0.2744,
      "num_input_tokens_seen": 10328248,
      "step": 13540
    },
    {
      "epoch": 28.16008316008316,
      "grad_norm": 0.00017873563047032803,
      "learning_rate": 0.2228404800208286,
      "loss": 0.2747,
      "num_input_tokens_seen": 10331896,
      "step": 13545
    },
    {
      "epoch": 28.17047817047817,
      "grad_norm": 0.00047356702270917594,
      "learning_rate": 0.22278898099162875,
      "loss": 0.2649,
      "num_input_tokens_seen": 10335640,
      "step": 13550
    },
    {
      "epoch": 28.18087318087318,
      "grad_norm": 0.0002550750505179167,
      "learning_rate": 0.22273747073745337,
      "loss": 0.2821,
      "num_input_tokens_seen": 10339608,
      "step": 13555
    },
    {
      "epoch": 28.19126819126819,
      "grad_norm": 0.0002169163926737383,
      "learning_rate": 0.22268594926624588,
      "loss": 0.261,
      "num_input_tokens_seen": 10343384,
      "step": 13560
    },
    {
      "epoch": 28.2016632016632,
      "grad_norm": 0.00020018444047309458,
      "learning_rate": 0.22263441658595162,
      "loss": 0.2755,
      "num_input_tokens_seen": 10347288,
      "step": 13565
    },
    {
      "epoch": 28.212058212058214,
      "grad_norm": 0.00021457232651300728,
      "learning_rate": 0.2225828727045175,
      "loss": 0.257,
      "num_input_tokens_seen": 10351064,
      "step": 13570
    },
    {
      "epoch": 28.222453222453222,
      "grad_norm": 0.000828791584353894,
      "learning_rate": 0.22253131762989228,
      "loss": 0.2812,
      "num_input_tokens_seen": 10354904,
      "step": 13575
    },
    {
      "epoch": 28.232848232848234,
      "grad_norm": 0.00044811420957557857,
      "learning_rate": 0.2224797513700264,
      "loss": 0.2427,
      "num_input_tokens_seen": 10358680,
      "step": 13580
    },
    {
      "epoch": 28.243243243243242,
      "grad_norm": 8.97665013326332e-05,
      "learning_rate": 0.22242817393287204,
      "loss": 0.2806,
      "num_input_tokens_seen": 10362520,
      "step": 13585
    },
    {
      "epoch": 28.253638253638254,
      "grad_norm": 0.0005119292181916535,
      "learning_rate": 0.22237658532638305,
      "loss": 0.2634,
      "num_input_tokens_seen": 10366392,
      "step": 13590
    },
    {
      "epoch": 28.264033264033262,
      "grad_norm": 0.0004910401185043156,
      "learning_rate": 0.22232498555851513,
      "loss": 0.2645,
      "num_input_tokens_seen": 10370136,
      "step": 13595
    },
    {
      "epoch": 28.274428274428274,
      "grad_norm": 9.295009658671916e-05,
      "learning_rate": 0.22227337463722546,
      "loss": 0.2742,
      "num_input_tokens_seen": 10373912,
      "step": 13600
    },
    {
      "epoch": 28.274428274428274,
      "eval_loss": 0.24840591847896576,
      "eval_runtime": 13.4017,
      "eval_samples_per_second": 63.873,
      "eval_steps_per_second": 15.968,
      "num_input_tokens_seen": 10373912,
      "step": 13600
    },
    {
      "epoch": 28.284823284823286,
      "grad_norm": 0.0004236153035890311,
      "learning_rate": 0.2222217525704732,
      "loss": 0.268,
      "num_input_tokens_seen": 10377688,
      "step": 13605
    },
    {
      "epoch": 28.295218295218294,
      "grad_norm": 0.00020817089534830302,
      "learning_rate": 0.22217011936621908,
      "loss": 0.278,
      "num_input_tokens_seen": 10381464,
      "step": 13610
    },
    {
      "epoch": 28.305613305613306,
      "grad_norm": 0.0009431368671357632,
      "learning_rate": 0.22211847503242566,
      "loss": 0.2661,
      "num_input_tokens_seen": 10385336,
      "step": 13615
    },
    {
      "epoch": 28.316008316008315,
      "grad_norm": 0.00024732365272939205,
      "learning_rate": 0.22206681957705704,
      "loss": 0.2672,
      "num_input_tokens_seen": 10389208,
      "step": 13620
    },
    {
      "epoch": 28.326403326403327,
      "grad_norm": 0.00034605455584824085,
      "learning_rate": 0.2220151530080792,
      "loss": 0.2895,
      "num_input_tokens_seen": 10393048,
      "step": 13625
    },
    {
      "epoch": 28.33679833679834,
      "grad_norm": 0.00024305492115672678,
      "learning_rate": 0.2219634753334598,
      "loss": 0.2513,
      "num_input_tokens_seen": 10396888,
      "step": 13630
    },
    {
      "epoch": 28.347193347193347,
      "grad_norm": 0.00016207690350711346,
      "learning_rate": 0.22191178656116817,
      "loss": 0.2446,
      "num_input_tokens_seen": 10400696,
      "step": 13635
    },
    {
      "epoch": 28.35758835758836,
      "grad_norm": 0.0005197438294999301,
      "learning_rate": 0.2218600866991753,
      "loss": 0.2822,
      "num_input_tokens_seen": 10404504,
      "step": 13640
    },
    {
      "epoch": 28.367983367983367,
      "grad_norm": 0.0002536851679906249,
      "learning_rate": 0.221808375755454,
      "loss": 0.2358,
      "num_input_tokens_seen": 10408472,
      "step": 13645
    },
    {
      "epoch": 28.37837837837838,
      "grad_norm": 0.00018766659195534885,
      "learning_rate": 0.22175665373797881,
      "loss": 0.3053,
      "num_input_tokens_seen": 10412344,
      "step": 13650
    },
    {
      "epoch": 28.388773388773387,
      "grad_norm": 0.0006162977661006153,
      "learning_rate": 0.22170492065472583,
      "loss": 0.2757,
      "num_input_tokens_seen": 10416248,
      "step": 13655
    },
    {
      "epoch": 28.3991683991684,
      "grad_norm": 0.0006687522400170565,
      "learning_rate": 0.221653176513673,
      "loss": 0.2617,
      "num_input_tokens_seen": 10420056,
      "step": 13660
    },
    {
      "epoch": 28.40956340956341,
      "grad_norm": 0.0001514917385065928,
      "learning_rate": 0.2216014213227999,
      "loss": 0.2663,
      "num_input_tokens_seen": 10423800,
      "step": 13665
    },
    {
      "epoch": 28.41995841995842,
      "grad_norm": 3.96703981095925e-05,
      "learning_rate": 0.22154965509008784,
      "loss": 0.25,
      "num_input_tokens_seen": 10427544,
      "step": 13670
    },
    {
      "epoch": 28.43035343035343,
      "grad_norm": 0.00032525864662602544,
      "learning_rate": 0.2214978778235198,
      "loss": 0.2847,
      "num_input_tokens_seen": 10431320,
      "step": 13675
    },
    {
      "epoch": 28.44074844074844,
      "grad_norm": 0.0009113075793720782,
      "learning_rate": 0.2214460895310805,
      "loss": 0.2556,
      "num_input_tokens_seen": 10435192,
      "step": 13680
    },
    {
      "epoch": 28.45114345114345,
      "grad_norm": 0.0002059585676761344,
      "learning_rate": 0.22139429022075635,
      "loss": 0.2411,
      "num_input_tokens_seen": 10438872,
      "step": 13685
    },
    {
      "epoch": 28.46153846153846,
      "grad_norm": 0.0006226692348718643,
      "learning_rate": 0.22134247990053546,
      "loss": 0.2778,
      "num_input_tokens_seen": 10442744,
      "step": 13690
    },
    {
      "epoch": 28.471933471933472,
      "grad_norm": 0.0007250407943502069,
      "learning_rate": 0.2212906585784076,
      "loss": 0.2591,
      "num_input_tokens_seen": 10446552,
      "step": 13695
    },
    {
      "epoch": 28.482328482328484,
      "grad_norm": 0.00031673398916609585,
      "learning_rate": 0.22123882626236432,
      "loss": 0.2693,
      "num_input_tokens_seen": 10450392,
      "step": 13700
    },
    {
      "epoch": 28.492723492723492,
      "grad_norm": 0.0001294183894060552,
      "learning_rate": 0.2211869829603988,
      "loss": 0.2741,
      "num_input_tokens_seen": 10454200,
      "step": 13705
    },
    {
      "epoch": 28.503118503118504,
      "grad_norm": 0.0002462098782416433,
      "learning_rate": 0.22113512868050592,
      "loss": 0.2515,
      "num_input_tokens_seen": 10457944,
      "step": 13710
    },
    {
      "epoch": 28.513513513513512,
      "grad_norm": 0.00043582136277109385,
      "learning_rate": 0.2210832634306822,
      "loss": 0.2284,
      "num_input_tokens_seen": 10461816,
      "step": 13715
    },
    {
      "epoch": 28.523908523908524,
      "grad_norm": 0.00022136872576083988,
      "learning_rate": 0.22103138721892598,
      "loss": 0.3063,
      "num_input_tokens_seen": 10465816,
      "step": 13720
    },
    {
      "epoch": 28.534303534303533,
      "grad_norm": 3.260079029132612e-05,
      "learning_rate": 0.22097950005323724,
      "loss": 0.2756,
      "num_input_tokens_seen": 10469496,
      "step": 13725
    },
    {
      "epoch": 28.544698544698544,
      "grad_norm": 6.137879972811788e-05,
      "learning_rate": 0.22092760194161762,
      "loss": 0.2795,
      "num_input_tokens_seen": 10473368,
      "step": 13730
    },
    {
      "epoch": 28.555093555093556,
      "grad_norm": 0.0002104137820424512,
      "learning_rate": 0.2208756928920704,
      "loss": 0.2812,
      "num_input_tokens_seen": 10477144,
      "step": 13735
    },
    {
      "epoch": 28.565488565488565,
      "grad_norm": 3.5347711673239246e-05,
      "learning_rate": 0.22082377291260072,
      "loss": 0.2711,
      "num_input_tokens_seen": 10480856,
      "step": 13740
    },
    {
      "epoch": 28.575883575883577,
      "grad_norm": 0.00022669402824249119,
      "learning_rate": 0.2207718420112152,
      "loss": 0.2506,
      "num_input_tokens_seen": 10484696,
      "step": 13745
    },
    {
      "epoch": 28.586278586278585,
      "grad_norm": 0.0003218489873688668,
      "learning_rate": 0.22071990019592228,
      "loss": 0.2412,
      "num_input_tokens_seen": 10488472,
      "step": 13750
    },
    {
      "epoch": 28.596673596673597,
      "grad_norm": 0.00044010457349941134,
      "learning_rate": 0.22066794747473198,
      "loss": 0.2697,
      "num_input_tokens_seen": 10492312,
      "step": 13755
    },
    {
      "epoch": 28.60706860706861,
      "grad_norm": 0.00035100351669825613,
      "learning_rate": 0.2206159838556562,
      "loss": 0.2526,
      "num_input_tokens_seen": 10496216,
      "step": 13760
    },
    {
      "epoch": 28.617463617463617,
      "grad_norm": 0.00087849295232445,
      "learning_rate": 0.2205640093467082,
      "loss": 0.283,
      "num_input_tokens_seen": 10500024,
      "step": 13765
    },
    {
      "epoch": 28.62785862785863,
      "grad_norm": 0.00043509266106411815,
      "learning_rate": 0.22051202395590322,
      "loss": 0.2847,
      "num_input_tokens_seen": 10503800,
      "step": 13770
    },
    {
      "epoch": 28.638253638253637,
      "grad_norm": 0.00039357872446998954,
      "learning_rate": 0.22046002769125808,
      "loss": 0.2867,
      "num_input_tokens_seen": 10507608,
      "step": 13775
    },
    {
      "epoch": 28.64864864864865,
      "grad_norm": 0.00011844849359476939,
      "learning_rate": 0.2204080205607912,
      "loss": 0.2858,
      "num_input_tokens_seen": 10511608,
      "step": 13780
    },
    {
      "epoch": 28.659043659043657,
      "grad_norm": 6.902393943164498e-05,
      "learning_rate": 0.22035600257252272,
      "loss": 0.2789,
      "num_input_tokens_seen": 10515320,
      "step": 13785
    },
    {
      "epoch": 28.66943866943867,
      "grad_norm": 0.0005170084768906236,
      "learning_rate": 0.2203039737344745,
      "loss": 0.2836,
      "num_input_tokens_seen": 10519160,
      "step": 13790
    },
    {
      "epoch": 28.67983367983368,
      "grad_norm": 4.670147973229177e-05,
      "learning_rate": 0.22025193405467003,
      "loss": 0.2647,
      "num_input_tokens_seen": 10523000,
      "step": 13795
    },
    {
      "epoch": 28.69022869022869,
      "grad_norm": 0.0003400059649720788,
      "learning_rate": 0.2201998835411345,
      "loss": 0.2789,
      "num_input_tokens_seen": 10526808,
      "step": 13800
    },
    {
      "epoch": 28.69022869022869,
      "eval_loss": 0.2496807724237442,
      "eval_runtime": 13.4138,
      "eval_samples_per_second": 63.815,
      "eval_steps_per_second": 15.954,
      "num_input_tokens_seen": 10526808,
      "step": 13800
    },
    {
      "epoch": 28.7006237006237,
      "grad_norm": 8.334021549671888e-05,
      "learning_rate": 0.22014782220189474,
      "loss": 0.2616,
      "num_input_tokens_seen": 10530584,
      "step": 13805
    },
    {
      "epoch": 28.71101871101871,
      "grad_norm": 0.00017785215459298342,
      "learning_rate": 0.2200957500449793,
      "loss": 0.259,
      "num_input_tokens_seen": 10534488,
      "step": 13810
    },
    {
      "epoch": 28.72141372141372,
      "grad_norm": 0.0004352965042926371,
      "learning_rate": 0.22004366707841827,
      "loss": 0.2642,
      "num_input_tokens_seen": 10538264,
      "step": 13815
    },
    {
      "epoch": 28.731808731808734,
      "grad_norm": 0.0003644624666776508,
      "learning_rate": 0.21999157331024358,
      "loss": 0.2577,
      "num_input_tokens_seen": 10541912,
      "step": 13820
    },
    {
      "epoch": 28.742203742203742,
      "grad_norm": 0.0003521043690852821,
      "learning_rate": 0.21993946874848871,
      "loss": 0.2608,
      "num_input_tokens_seen": 10545720,
      "step": 13825
    },
    {
      "epoch": 28.752598752598754,
      "grad_norm": 0.0003447293129283935,
      "learning_rate": 0.2198873534011888,
      "loss": 0.2725,
      "num_input_tokens_seen": 10549624,
      "step": 13830
    },
    {
      "epoch": 28.762993762993762,
      "grad_norm": 0.00019379006698727608,
      "learning_rate": 0.2198352272763808,
      "loss": 0.2613,
      "num_input_tokens_seen": 10553368,
      "step": 13835
    },
    {
      "epoch": 28.773388773388774,
      "grad_norm": 0.00016407031216658652,
      "learning_rate": 0.2197830903821031,
      "loss": 0.2623,
      "num_input_tokens_seen": 10557144,
      "step": 13840
    },
    {
      "epoch": 28.783783783783782,
      "grad_norm": 0.0004928882699459791,
      "learning_rate": 0.21973094272639598,
      "loss": 0.2589,
      "num_input_tokens_seen": 10560984,
      "step": 13845
    },
    {
      "epoch": 28.794178794178794,
      "grad_norm": 0.00047189852921292186,
      "learning_rate": 0.21967878431730117,
      "loss": 0.2779,
      "num_input_tokens_seen": 10564824,
      "step": 13850
    },
    {
      "epoch": 28.804573804573806,
      "grad_norm": 0.00022804511536378413,
      "learning_rate": 0.21962661516286217,
      "loss": 0.2681,
      "num_input_tokens_seen": 10568664,
      "step": 13855
    },
    {
      "epoch": 28.814968814968815,
      "grad_norm": 5.3561681852443144e-05,
      "learning_rate": 0.21957443527112414,
      "loss": 0.2734,
      "num_input_tokens_seen": 10572376,
      "step": 13860
    },
    {
      "epoch": 28.825363825363826,
      "grad_norm": 0.00013434651191346347,
      "learning_rate": 0.21952224465013384,
      "loss": 0.2639,
      "num_input_tokens_seen": 10576088,
      "step": 13865
    },
    {
      "epoch": 28.835758835758835,
      "grad_norm": 0.0008785526151768863,
      "learning_rate": 0.21947004330793976,
      "loss": 0.2856,
      "num_input_tokens_seen": 10579960,
      "step": 13870
    },
    {
      "epoch": 28.846153846153847,
      "grad_norm": 0.0004586655995808542,
      "learning_rate": 0.21941783125259198,
      "loss": 0.2255,
      "num_input_tokens_seen": 10583544,
      "step": 13875
    },
    {
      "epoch": 28.856548856548855,
      "grad_norm": 0.0003975137078668922,
      "learning_rate": 0.21936560849214226,
      "loss": 0.2606,
      "num_input_tokens_seen": 10587288,
      "step": 13880
    },
    {
      "epoch": 28.866943866943867,
      "grad_norm": 0.00015181452909018844,
      "learning_rate": 0.21931337503464404,
      "loss": 0.2874,
      "num_input_tokens_seen": 10591160,
      "step": 13885
    },
    {
      "epoch": 28.87733887733888,
      "grad_norm": 0.00035508748260326684,
      "learning_rate": 0.21926113088815233,
      "loss": 0.2657,
      "num_input_tokens_seen": 10595032,
      "step": 13890
    },
    {
      "epoch": 28.887733887733887,
      "grad_norm": 0.0006310552125796676,
      "learning_rate": 0.2192088760607238,
      "loss": 0.2897,
      "num_input_tokens_seen": 10598968,
      "step": 13895
    },
    {
      "epoch": 28.8981288981289,
      "grad_norm": 0.0010681103449314833,
      "learning_rate": 0.2191566105604169,
      "loss": 0.2591,
      "num_input_tokens_seen": 10602680,
      "step": 13900
    },
    {
      "epoch": 28.908523908523907,
      "grad_norm": 0.0005205380730330944,
      "learning_rate": 0.21910433439529153,
      "loss": 0.2653,
      "num_input_tokens_seen": 10606488,
      "step": 13905
    },
    {
      "epoch": 28.91891891891892,
      "grad_norm": 0.00010886816744459793,
      "learning_rate": 0.2190520475734094,
      "loss": 0.2227,
      "num_input_tokens_seen": 10610328,
      "step": 13910
    },
    {
      "epoch": 28.929313929313928,
      "grad_norm": 0.00032817869214341044,
      "learning_rate": 0.2189997501028338,
      "loss": 0.2586,
      "num_input_tokens_seen": 10614104,
      "step": 13915
    },
    {
      "epoch": 28.93970893970894,
      "grad_norm": 0.00036502041621133685,
      "learning_rate": 0.2189474419916296,
      "loss": 0.2733,
      "num_input_tokens_seen": 10617976,
      "step": 13920
    },
    {
      "epoch": 28.95010395010395,
      "grad_norm": 0.0002529247140046209,
      "learning_rate": 0.21889512324786342,
      "loss": 0.2694,
      "num_input_tokens_seen": 10621688,
      "step": 13925
    },
    {
      "epoch": 28.96049896049896,
      "grad_norm": 0.00014868068683426827,
      "learning_rate": 0.21884279387960345,
      "loss": 0.2727,
      "num_input_tokens_seen": 10625368,
      "step": 13930
    },
    {
      "epoch": 28.97089397089397,
      "grad_norm": 0.000550108146853745,
      "learning_rate": 0.2187904538949195,
      "loss": 0.2802,
      "num_input_tokens_seen": 10629112,
      "step": 13935
    },
    {
      "epoch": 28.98128898128898,
      "grad_norm": 0.00043433523387648165,
      "learning_rate": 0.2187381033018831,
      "loss": 0.2756,
      "num_input_tokens_seen": 10633048,
      "step": 13940
    },
    {
      "epoch": 28.991683991683992,
      "grad_norm": 0.0002551070647314191,
      "learning_rate": 0.2186857421085673,
      "loss": 0.2707,
      "num_input_tokens_seen": 10636824,
      "step": 13945
    },
    {
      "epoch": 29.002079002079004,
      "grad_norm": 0.0006948256050236523,
      "learning_rate": 0.21863337032304697,
      "loss": 0.2716,
      "num_input_tokens_seen": 10640672,
      "step": 13950
    },
    {
      "epoch": 29.012474012474012,
      "grad_norm": 0.00031978514743968844,
      "learning_rate": 0.21858098795339845,
      "loss": 0.2705,
      "num_input_tokens_seen": 10644512,
      "step": 13955
    },
    {
      "epoch": 29.022869022869024,
      "grad_norm": 0.00045618126750923693,
      "learning_rate": 0.21852859500769975,
      "loss": 0.2534,
      "num_input_tokens_seen": 10648384,
      "step": 13960
    },
    {
      "epoch": 29.033264033264032,
      "grad_norm": 0.00010564591502770782,
      "learning_rate": 0.21847619149403044,
      "loss": 0.2418,
      "num_input_tokens_seen": 10652288,
      "step": 13965
    },
    {
      "epoch": 29.043659043659044,
      "grad_norm": 0.00027221511118113995,
      "learning_rate": 0.21842377742047195,
      "loss": 0.2495,
      "num_input_tokens_seen": 10656000,
      "step": 13970
    },
    {
      "epoch": 29.054054054054053,
      "grad_norm": 0.00017686275532469153,
      "learning_rate": 0.21837135279510705,
      "loss": 0.2643,
      "num_input_tokens_seen": 10659872,
      "step": 13975
    },
    {
      "epoch": 29.064449064449065,
      "grad_norm": 0.00022571589215658605,
      "learning_rate": 0.21831891762602038,
      "loss": 0.2604,
      "num_input_tokens_seen": 10663648,
      "step": 13980
    },
    {
      "epoch": 29.074844074844076,
      "grad_norm": 0.00042226180084981024,
      "learning_rate": 0.21826647192129806,
      "loss": 0.2637,
      "num_input_tokens_seen": 10667456,
      "step": 13985
    },
    {
      "epoch": 29.085239085239085,
      "grad_norm": 5.466501534101553e-05,
      "learning_rate": 0.21821401568902787,
      "loss": 0.2717,
      "num_input_tokens_seen": 10671264,
      "step": 13990
    },
    {
      "epoch": 29.095634095634097,
      "grad_norm": 3.547640881151892e-05,
      "learning_rate": 0.21816154893729925,
      "loss": 0.288,
      "num_input_tokens_seen": 10675264,
      "step": 13995
    },
    {
      "epoch": 29.106029106029105,
      "grad_norm": 8.872819307725877e-05,
      "learning_rate": 0.2181090716742032,
      "loss": 0.273,
      "num_input_tokens_seen": 10678976,
      "step": 14000
    },
    {
      "epoch": 29.106029106029105,
      "eval_loss": 0.25908181071281433,
      "eval_runtime": 13.4319,
      "eval_samples_per_second": 63.729,
      "eval_steps_per_second": 15.932,
      "num_input_tokens_seen": 10678976,
      "step": 14000
    },
    {
      "epoch": 29.116424116424117,
      "grad_norm": 0.0005775134195573628,
      "learning_rate": 0.21805658390783236,
      "loss": 0.2599,
      "num_input_tokens_seen": 10682720,
      "step": 14005
    },
    {
      "epoch": 29.126819126819125,
      "grad_norm": 0.0008774096495471895,
      "learning_rate": 0.21800408564628107,
      "loss": 0.26,
      "num_input_tokens_seen": 10686592,
      "step": 14010
    },
    {
      "epoch": 29.137214137214137,
      "grad_norm": 0.00013676991511601955,
      "learning_rate": 0.21795157689764516,
      "loss": 0.2927,
      "num_input_tokens_seen": 10690464,
      "step": 14015
    },
    {
      "epoch": 29.14760914760915,
      "grad_norm": 0.0003483971522655338,
      "learning_rate": 0.21789905767002216,
      "loss": 0.2726,
      "num_input_tokens_seen": 10694304,
      "step": 14020
    },
    {
      "epoch": 29.158004158004157,
      "grad_norm": 8.366950351046398e-05,
      "learning_rate": 0.2178465279715112,
      "loss": 0.2579,
      "num_input_tokens_seen": 10698208,
      "step": 14025
    },
    {
      "epoch": 29.16839916839917,
      "grad_norm": 0.00020679635053966194,
      "learning_rate": 0.21779398781021303,
      "loss": 0.2474,
      "num_input_tokens_seen": 10701888,
      "step": 14030
    },
    {
      "epoch": 29.178794178794178,
      "grad_norm": 0.0008092079660855234,
      "learning_rate": 0.21774143719422998,
      "loss": 0.3004,
      "num_input_tokens_seen": 10705792,
      "step": 14035
    },
    {
      "epoch": 29.18918918918919,
      "grad_norm": 0.0009311113390140235,
      "learning_rate": 0.21768887613166601,
      "loss": 0.2824,
      "num_input_tokens_seen": 10709632,
      "step": 14040
    },
    {
      "epoch": 29.1995841995842,
      "grad_norm": 0.0006800752598792315,
      "learning_rate": 0.2176363046306267,
      "loss": 0.2897,
      "num_input_tokens_seen": 10713536,
      "step": 14045
    },
    {
      "epoch": 29.20997920997921,
      "grad_norm": 0.00022434488346334547,
      "learning_rate": 0.21758372269921925,
      "loss": 0.2876,
      "num_input_tokens_seen": 10717376,
      "step": 14050
    },
    {
      "epoch": 29.22037422037422,
      "grad_norm": 0.00045676474110223353,
      "learning_rate": 0.21753113034555244,
      "loss": 0.2497,
      "num_input_tokens_seen": 10721344,
      "step": 14055
    },
    {
      "epoch": 29.23076923076923,
      "grad_norm": 0.00011392116721253842,
      "learning_rate": 0.2174785275777367,
      "loss": 0.3248,
      "num_input_tokens_seen": 10725312,
      "step": 14060
    },
    {
      "epoch": 29.241164241164242,
      "grad_norm": 0.00035922828828915954,
      "learning_rate": 0.21742591440388404,
      "loss": 0.2263,
      "num_input_tokens_seen": 10728864,
      "step": 14065
    },
    {
      "epoch": 29.25155925155925,
      "grad_norm": 0.00013645851868204772,
      "learning_rate": 0.21737329083210802,
      "loss": 0.2233,
      "num_input_tokens_seen": 10732640,
      "step": 14070
    },
    {
      "epoch": 29.261954261954262,
      "grad_norm": 0.0001173399796243757,
      "learning_rate": 0.2173206568705239,
      "loss": 0.2297,
      "num_input_tokens_seen": 10736416,
      "step": 14075
    },
    {
      "epoch": 29.272349272349274,
      "grad_norm": 0.00047181517584249377,
      "learning_rate": 0.2172680125272485,
      "loss": 0.2609,
      "num_input_tokens_seen": 10740096,
      "step": 14080
    },
    {
      "epoch": 29.282744282744282,
      "grad_norm": 0.00013942638179287314,
      "learning_rate": 0.2172153578104002,
      "loss": 0.299,
      "num_input_tokens_seen": 10743936,
      "step": 14085
    },
    {
      "epoch": 29.293139293139294,
      "grad_norm": 0.00021279761858750135,
      "learning_rate": 0.21716269272809902,
      "loss": 0.2727,
      "num_input_tokens_seen": 10747680,
      "step": 14090
    },
    {
      "epoch": 29.303534303534303,
      "grad_norm": 0.00019477716705296189,
      "learning_rate": 0.21711001728846666,
      "loss": 0.2689,
      "num_input_tokens_seen": 10751392,
      "step": 14095
    },
    {
      "epoch": 29.313929313929314,
      "grad_norm": 0.0002950727939605713,
      "learning_rate": 0.21705733149962628,
      "loss": 0.2491,
      "num_input_tokens_seen": 10755232,
      "step": 14100
    },
    {
      "epoch": 29.324324324324323,
      "grad_norm": 0.000299810926662758,
      "learning_rate": 0.21700463536970263,
      "loss": 0.2718,
      "num_input_tokens_seen": 10759232,
      "step": 14105
    },
    {
      "epoch": 29.334719334719335,
      "grad_norm": 0.0007824341882951558,
      "learning_rate": 0.21695192890682222,
      "loss": 0.252,
      "num_input_tokens_seen": 10763136,
      "step": 14110
    },
    {
      "epoch": 29.345114345114347,
      "grad_norm": 0.0002583378809504211,
      "learning_rate": 0.21689921211911298,
      "loss": 0.2369,
      "num_input_tokens_seen": 10766912,
      "step": 14115
    },
    {
      "epoch": 29.355509355509355,
      "grad_norm": 7.649027975276113e-05,
      "learning_rate": 0.21684648501470452,
      "loss": 0.2736,
      "num_input_tokens_seen": 10770880,
      "step": 14120
    },
    {
      "epoch": 29.365904365904367,
      "grad_norm": 0.00036276126047596335,
      "learning_rate": 0.216793747601728,
      "loss": 0.2902,
      "num_input_tokens_seen": 10774688,
      "step": 14125
    },
    {
      "epoch": 29.376299376299375,
      "grad_norm": 0.0001871962595032528,
      "learning_rate": 0.21674099988831627,
      "loss": 0.2616,
      "num_input_tokens_seen": 10778560,
      "step": 14130
    },
    {
      "epoch": 29.386694386694387,
      "grad_norm": 0.0005337464972399175,
      "learning_rate": 0.21668824188260363,
      "loss": 0.2632,
      "num_input_tokens_seen": 10782304,
      "step": 14135
    },
    {
      "epoch": 29.397089397089395,
      "grad_norm": 0.0007402925402857363,
      "learning_rate": 0.21663547359272606,
      "loss": 0.2654,
      "num_input_tokens_seen": 10786304,
      "step": 14140
    },
    {
      "epoch": 29.407484407484407,
      "grad_norm": 0.0001457414764445275,
      "learning_rate": 0.216582695026821,
      "loss": 0.2546,
      "num_input_tokens_seen": 10790112,
      "step": 14145
    },
    {
      "epoch": 29.41787941787942,
      "grad_norm": 0.00019454641733318567,
      "learning_rate": 0.21652990619302767,
      "loss": 0.2586,
      "num_input_tokens_seen": 10793984,
      "step": 14150
    },
    {
      "epoch": 29.428274428274428,
      "grad_norm": 0.0005052004707977176,
      "learning_rate": 0.21647710709948673,
      "loss": 0.2815,
      "num_input_tokens_seen": 10797632,
      "step": 14155
    },
    {
      "epoch": 29.43866943866944,
      "grad_norm": 0.00010229657345917076,
      "learning_rate": 0.2164242977543405,
      "loss": 0.2479,
      "num_input_tokens_seen": 10801344,
      "step": 14160
    },
    {
      "epoch": 29.449064449064448,
      "grad_norm": 0.0007540743099525571,
      "learning_rate": 0.21637147816573277,
      "loss": 0.313,
      "num_input_tokens_seen": 10805152,
      "step": 14165
    },
    {
      "epoch": 29.45945945945946,
      "grad_norm": 0.0003524782950989902,
      "learning_rate": 0.21631864834180908,
      "loss": 0.2595,
      "num_input_tokens_seen": 10809024,
      "step": 14170
    },
    {
      "epoch": 29.46985446985447,
      "grad_norm": 6.010181095916778e-05,
      "learning_rate": 0.21626580829071637,
      "loss": 0.2634,
      "num_input_tokens_seen": 10812672,
      "step": 14175
    },
    {
      "epoch": 29.48024948024948,
      "grad_norm": 0.0005448845331557095,
      "learning_rate": 0.21621295802060328,
      "loss": 0.282,
      "num_input_tokens_seen": 10816480,
      "step": 14180
    },
    {
      "epoch": 29.490644490644492,
      "grad_norm": 0.0006666610715910792,
      "learning_rate": 0.21616009753961996,
      "loss": 0.262,
      "num_input_tokens_seen": 10820224,
      "step": 14185
    },
    {
      "epoch": 29.5010395010395,
      "grad_norm": 0.0002556899271439761,
      "learning_rate": 0.2161072268559182,
      "loss": 0.2724,
      "num_input_tokens_seen": 10823872,
      "step": 14190
    },
    {
      "epoch": 29.511434511434512,
      "grad_norm": 0.0002947613247670233,
      "learning_rate": 0.21605434597765133,
      "loss": 0.2601,
      "num_input_tokens_seen": 10827648,
      "step": 14195
    },
    {
      "epoch": 29.52182952182952,
      "grad_norm": 5.8358047681394964e-05,
      "learning_rate": 0.21600145491297418,
      "loss": 0.2554,
      "num_input_tokens_seen": 10831520,
      "step": 14200
    },
    {
      "epoch": 29.52182952182952,
      "eval_loss": 0.24808000028133392,
      "eval_runtime": 13.4332,
      "eval_samples_per_second": 63.723,
      "eval_steps_per_second": 15.931,
      "num_input_tokens_seen": 10831520,
      "step": 14200
    },
    {
      "epoch": 29.532224532224532,
      "grad_norm": 0.0004730260989163071,
      "learning_rate": 0.21594855367004326,
      "loss": 0.2488,
      "num_input_tokens_seen": 10835584,
      "step": 14205
    },
    {
      "epoch": 29.542619542619544,
      "grad_norm": 0.0008272915729321539,
      "learning_rate": 0.21589564225701663,
      "loss": 0.2317,
      "num_input_tokens_seen": 10839392,
      "step": 14210
    },
    {
      "epoch": 29.553014553014552,
      "grad_norm": 6.040620428393595e-05,
      "learning_rate": 0.21584272068205385,
      "loss": 0.2575,
      "num_input_tokens_seen": 10843200,
      "step": 14215
    },
    {
      "epoch": 29.563409563409564,
      "grad_norm": 0.00018372329941485077,
      "learning_rate": 0.2157897889533161,
      "loss": 0.2745,
      "num_input_tokens_seen": 10846944,
      "step": 14220
    },
    {
      "epoch": 29.573804573804573,
      "grad_norm": 0.00029755596187897027,
      "learning_rate": 0.21573684707896612,
      "loss": 0.2656,
      "num_input_tokens_seen": 10850816,
      "step": 14225
    },
    {
      "epoch": 29.584199584199585,
      "grad_norm": 0.0002608706708997488,
      "learning_rate": 0.21568389506716826,
      "loss": 0.2686,
      "num_input_tokens_seen": 10854592,
      "step": 14230
    },
    {
      "epoch": 29.594594594594593,
      "grad_norm": 0.00029237096896395087,
      "learning_rate": 0.21563093292608831,
      "loss": 0.2798,
      "num_input_tokens_seen": 10858368,
      "step": 14235
    },
    {
      "epoch": 29.604989604989605,
      "grad_norm": 0.0002047140005743131,
      "learning_rate": 0.21557796066389376,
      "loss": 0.2774,
      "num_input_tokens_seen": 10862272,
      "step": 14240
    },
    {
      "epoch": 29.615384615384617,
      "grad_norm": 0.0006563168717548251,
      "learning_rate": 0.21552497828875353,
      "loss": 0.273,
      "num_input_tokens_seen": 10865984,
      "step": 14245
    },
    {
      "epoch": 29.625779625779625,
      "grad_norm": 0.0001679753913776949,
      "learning_rate": 0.21547198580883828,
      "loss": 0.2607,
      "num_input_tokens_seen": 10869856,
      "step": 14250
    },
    {
      "epoch": 29.636174636174637,
      "grad_norm": 0.00012551013787742704,
      "learning_rate": 0.21541898323232,
      "loss": 0.2215,
      "num_input_tokens_seen": 10873792,
      "step": 14255
    },
    {
      "epoch": 29.646569646569645,
      "grad_norm": 0.0003813078219536692,
      "learning_rate": 0.2153659705673724,
      "loss": 0.3193,
      "num_input_tokens_seen": 10877632,
      "step": 14260
    },
    {
      "epoch": 29.656964656964657,
      "grad_norm": 0.0005680934991687536,
      "learning_rate": 0.2153129478221707,
      "loss": 0.2757,
      "num_input_tokens_seen": 10881536,
      "step": 14265
    },
    {
      "epoch": 29.66735966735967,
      "grad_norm": 0.00011374261521268636,
      "learning_rate": 0.21525991500489164,
      "loss": 0.2805,
      "num_input_tokens_seen": 10885312,
      "step": 14270
    },
    {
      "epoch": 29.677754677754677,
      "grad_norm": 3.348291284055449e-05,
      "learning_rate": 0.21520687212371362,
      "loss": 0.2726,
      "num_input_tokens_seen": 10889056,
      "step": 14275
    },
    {
      "epoch": 29.68814968814969,
      "grad_norm": 5.475520811160095e-05,
      "learning_rate": 0.21515381918681648,
      "loss": 0.2594,
      "num_input_tokens_seen": 10892928,
      "step": 14280
    },
    {
      "epoch": 29.698544698544698,
      "grad_norm": 9.333314665127546e-05,
      "learning_rate": 0.21510075620238167,
      "loss": 0.2632,
      "num_input_tokens_seen": 10896960,
      "step": 14285
    },
    {
      "epoch": 29.70893970893971,
      "grad_norm": 0.00015351705951616168,
      "learning_rate": 0.21504768317859208,
      "loss": 0.2798,
      "num_input_tokens_seen": 10900704,
      "step": 14290
    },
    {
      "epoch": 29.719334719334718,
      "grad_norm": 0.00033292261650785804,
      "learning_rate": 0.2149946001236323,
      "loss": 0.2461,
      "num_input_tokens_seen": 10904544,
      "step": 14295
    },
    {
      "epoch": 29.72972972972973,
      "grad_norm": 0.00029600097332149744,
      "learning_rate": 0.21494150704568848,
      "loss": 0.2871,
      "num_input_tokens_seen": 10908096,
      "step": 14300
    },
    {
      "epoch": 29.74012474012474,
      "grad_norm": 0.0004019803600385785,
      "learning_rate": 0.21488840395294811,
      "loss": 0.2626,
      "num_input_tokens_seen": 10911936,
      "step": 14305
    },
    {
      "epoch": 29.75051975051975,
      "grad_norm": 0.0001952204038389027,
      "learning_rate": 0.21483529085360042,
      "loss": 0.2774,
      "num_input_tokens_seen": 10915776,
      "step": 14310
    },
    {
      "epoch": 29.760914760914762,
      "grad_norm": 0.00013958119961898774,
      "learning_rate": 0.2147821677558361,
      "loss": 0.2566,
      "num_input_tokens_seen": 10919680,
      "step": 14315
    },
    {
      "epoch": 29.77130977130977,
      "grad_norm": 0.0002875577483791858,
      "learning_rate": 0.2147290346678475,
      "loss": 0.2941,
      "num_input_tokens_seen": 10923488,
      "step": 14320
    },
    {
      "epoch": 29.781704781704782,
      "grad_norm": 0.0005459499661810696,
      "learning_rate": 0.21467589159782827,
      "loss": 0.2386,
      "num_input_tokens_seen": 10927392,
      "step": 14325
    },
    {
      "epoch": 29.79209979209979,
      "grad_norm": 0.00017256855790037662,
      "learning_rate": 0.21462273855397374,
      "loss": 0.2661,
      "num_input_tokens_seen": 10931168,
      "step": 14330
    },
    {
      "epoch": 29.802494802494802,
      "grad_norm": 0.0001086880947696045,
      "learning_rate": 0.21456957554448083,
      "loss": 0.2901,
      "num_input_tokens_seen": 10934912,
      "step": 14335
    },
    {
      "epoch": 29.812889812889814,
      "grad_norm": 0.00013465310621540993,
      "learning_rate": 0.21451640257754795,
      "loss": 0.2663,
      "num_input_tokens_seen": 10938656,
      "step": 14340
    },
    {
      "epoch": 29.823284823284823,
      "grad_norm": 0.0007124594412744045,
      "learning_rate": 0.21446321966137508,
      "loss": 0.2719,
      "num_input_tokens_seen": 10942304,
      "step": 14345
    },
    {
      "epoch": 29.833679833679835,
      "grad_norm": 0.00045124292955733836,
      "learning_rate": 0.21441002680416354,
      "loss": 0.2463,
      "num_input_tokens_seen": 10945952,
      "step": 14350
    },
    {
      "epoch": 29.844074844074843,
      "grad_norm": 0.0005798981292173266,
      "learning_rate": 0.21435682401411654,
      "loss": 0.2653,
      "num_input_tokens_seen": 10949536,
      "step": 14355
    },
    {
      "epoch": 29.854469854469855,
      "grad_norm": 9.99642361421138e-05,
      "learning_rate": 0.2143036112994385,
      "loss": 0.2816,
      "num_input_tokens_seen": 10953408,
      "step": 14360
    },
    {
      "epoch": 29.864864864864863,
      "grad_norm": 0.00011309548426652327,
      "learning_rate": 0.21425038866833548,
      "loss": 0.2679,
      "num_input_tokens_seen": 10957280,
      "step": 14365
    },
    {
      "epoch": 29.875259875259875,
      "grad_norm": 9.759383829077706e-05,
      "learning_rate": 0.21419715612901508,
      "loss": 0.2794,
      "num_input_tokens_seen": 10961216,
      "step": 14370
    },
    {
      "epoch": 29.885654885654887,
      "grad_norm": 0.00030324768158607185,
      "learning_rate": 0.21414391368968652,
      "loss": 0.2731,
      "num_input_tokens_seen": 10965056,
      "step": 14375
    },
    {
      "epoch": 29.896049896049895,
      "grad_norm": 0.00017458086949773133,
      "learning_rate": 0.21409066135856034,
      "loss": 0.2837,
      "num_input_tokens_seen": 10968960,
      "step": 14380
    },
    {
      "epoch": 29.906444906444907,
      "grad_norm": 0.000135528709506616,
      "learning_rate": 0.21403739914384878,
      "loss": 0.2751,
      "num_input_tokens_seen": 10972864,
      "step": 14385
    },
    {
      "epoch": 29.916839916839916,
      "grad_norm": 0.00046915566781535745,
      "learning_rate": 0.21398412705376554,
      "loss": 0.2817,
      "num_input_tokens_seen": 10976736,
      "step": 14390
    },
    {
      "epoch": 29.927234927234927,
      "grad_norm": 0.00025480909971520305,
      "learning_rate": 0.2139308450965258,
      "loss": 0.2656,
      "num_input_tokens_seen": 10980512,
      "step": 14395
    },
    {
      "epoch": 29.93762993762994,
      "grad_norm": 0.0002910270995926112,
      "learning_rate": 0.21387755328034638,
      "loss": 0.2605,
      "num_input_tokens_seen": 10984224,
      "step": 14400
    },
    {
      "epoch": 29.93762993762994,
      "eval_loss": 0.24851953983306885,
      "eval_runtime": 13.4113,
      "eval_samples_per_second": 63.827,
      "eval_steps_per_second": 15.957,
      "num_input_tokens_seen": 10984224,
      "step": 14400
    },
    {
      "epoch": 29.948024948024948,
      "grad_norm": 0.00014919557725079358,
      "learning_rate": 0.2138242516134455,
      "loss": 0.2688,
      "num_input_tokens_seen": 10988000,
      "step": 14405
    },
    {
      "epoch": 29.95841995841996,
      "grad_norm": 0.0001066965633071959,
      "learning_rate": 0.2137709401040429,
      "loss": 0.2667,
      "num_input_tokens_seen": 10991808,
      "step": 14410
    },
    {
      "epoch": 29.968814968814968,
      "grad_norm": 0.00021898998238611966,
      "learning_rate": 0.21371761876036,
      "loss": 0.2726,
      "num_input_tokens_seen": 10995616,
      "step": 14415
    },
    {
      "epoch": 29.97920997920998,
      "grad_norm": 0.0002709201944526285,
      "learning_rate": 0.21366428759061956,
      "loss": 0.2842,
      "num_input_tokens_seen": 10999360,
      "step": 14420
    },
    {
      "epoch": 29.989604989604988,
      "grad_norm": 0.0007954153697937727,
      "learning_rate": 0.2136109466030459,
      "loss": 0.2774,
      "num_input_tokens_seen": 11003104,
      "step": 14425
    },
    {
      "epoch": 30.0,
      "grad_norm": 0.0003685402625706047,
      "learning_rate": 0.2135575958058649,
      "loss": 0.2779,
      "num_input_tokens_seen": 11006712,
      "step": 14430
    },
    {
      "epoch": 30.010395010395012,
      "grad_norm": 0.00014657160500064492,
      "learning_rate": 0.2135042352073039,
      "loss": 0.2699,
      "num_input_tokens_seen": 11010584,
      "step": 14435
    },
    {
      "epoch": 30.02079002079002,
      "grad_norm": 3.5852124710800126e-05,
      "learning_rate": 0.2134508648155918,
      "loss": 0.2682,
      "num_input_tokens_seen": 11014328,
      "step": 14440
    },
    {
      "epoch": 30.031185031185032,
      "grad_norm": 0.0004663629806600511,
      "learning_rate": 0.213397484638959,
      "loss": 0.262,
      "num_input_tokens_seen": 11017944,
      "step": 14445
    },
    {
      "epoch": 30.04158004158004,
      "grad_norm": 8.454402268398553e-05,
      "learning_rate": 0.21334409468563728,
      "loss": 0.2558,
      "num_input_tokens_seen": 11021720,
      "step": 14450
    },
    {
      "epoch": 30.051975051975052,
      "grad_norm": 0.00010607652802718803,
      "learning_rate": 0.2132906949638602,
      "loss": 0.2653,
      "num_input_tokens_seen": 11025688,
      "step": 14455
    },
    {
      "epoch": 30.06237006237006,
      "grad_norm": 0.00010131805174751207,
      "learning_rate": 0.21323728548186255,
      "loss": 0.2648,
      "num_input_tokens_seen": 11029464,
      "step": 14460
    },
    {
      "epoch": 30.072765072765073,
      "grad_norm": 0.00019035505829378963,
      "learning_rate": 0.21318386624788088,
      "loss": 0.2554,
      "num_input_tokens_seen": 11033368,
      "step": 14465
    },
    {
      "epoch": 30.083160083160084,
      "grad_norm": 0.00047763189650140703,
      "learning_rate": 0.21313043727015288,
      "loss": 0.2415,
      "num_input_tokens_seen": 11037112,
      "step": 14470
    },
    {
      "epoch": 30.093555093555093,
      "grad_norm": 0.0001339707669103518,
      "learning_rate": 0.2130769985569182,
      "loss": 0.2781,
      "num_input_tokens_seen": 11040952,
      "step": 14475
    },
    {
      "epoch": 30.103950103950105,
      "grad_norm": 0.00038092880276963115,
      "learning_rate": 0.21302355011641766,
      "loss": 0.2668,
      "num_input_tokens_seen": 11044888,
      "step": 14480
    },
    {
      "epoch": 30.114345114345113,
      "grad_norm": 0.0001321235322393477,
      "learning_rate": 0.21297009195689365,
      "loss": 0.2725,
      "num_input_tokens_seen": 11048760,
      "step": 14485
    },
    {
      "epoch": 30.124740124740125,
      "grad_norm": 0.00017710465181153268,
      "learning_rate": 0.21291662408659015,
      "loss": 0.2488,
      "num_input_tokens_seen": 11052472,
      "step": 14490
    },
    {
      "epoch": 30.135135135135137,
      "grad_norm": 0.0002573048113845289,
      "learning_rate": 0.21286314651375254,
      "loss": 0.2469,
      "num_input_tokens_seen": 11056216,
      "step": 14495
    },
    {
      "epoch": 30.145530145530145,
      "grad_norm": 0.0006025422480888665,
      "learning_rate": 0.2128096592466278,
      "loss": 0.2795,
      "num_input_tokens_seen": 11059928,
      "step": 14500
    },
    {
      "epoch": 30.155925155925157,
      "grad_norm": 0.00015312687901314348,
      "learning_rate": 0.21275616229346428,
      "loss": 0.272,
      "num_input_tokens_seen": 11063768,
      "step": 14505
    },
    {
      "epoch": 30.166320166320165,
      "grad_norm": 0.00030205590883269906,
      "learning_rate": 0.21270265566251184,
      "loss": 0.279,
      "num_input_tokens_seen": 11067512,
      "step": 14510
    },
    {
      "epoch": 30.176715176715177,
      "grad_norm": 1.925172909977846e-05,
      "learning_rate": 0.21264913936202193,
      "loss": 0.2766,
      "num_input_tokens_seen": 11071416,
      "step": 14515
    },
    {
      "epoch": 30.187110187110186,
      "grad_norm": 0.00011313604773022234,
      "learning_rate": 0.2125956134002475,
      "loss": 0.2699,
      "num_input_tokens_seen": 11075160,
      "step": 14520
    },
    {
      "epoch": 30.197505197505198,
      "grad_norm": 0.00010148439469048753,
      "learning_rate": 0.2125420777854428,
      "loss": 0.2596,
      "num_input_tokens_seen": 11079064,
      "step": 14525
    },
    {
      "epoch": 30.20790020790021,
      "grad_norm": 0.0004043885273858905,
      "learning_rate": 0.21248853252586372,
      "loss": 0.291,
      "num_input_tokens_seen": 11082808,
      "step": 14530
    },
    {
      "epoch": 30.218295218295218,
      "grad_norm": 0.00021684229432139546,
      "learning_rate": 0.21243497762976774,
      "loss": 0.2237,
      "num_input_tokens_seen": 11086680,
      "step": 14535
    },
    {
      "epoch": 30.22869022869023,
      "grad_norm": 0.0008704495849087834,
      "learning_rate": 0.21238141310541356,
      "loss": 0.2942,
      "num_input_tokens_seen": 11090456,
      "step": 14540
    },
    {
      "epoch": 30.239085239085238,
      "grad_norm": 0.0005876815412193537,
      "learning_rate": 0.21232783896106153,
      "loss": 0.2703,
      "num_input_tokens_seen": 11094264,
      "step": 14545
    },
    {
      "epoch": 30.24948024948025,
      "grad_norm": 0.0007548096473328769,
      "learning_rate": 0.21227425520497345,
      "loss": 0.271,
      "num_input_tokens_seen": 11098104,
      "step": 14550
    },
    {
      "epoch": 30.25987525987526,
      "grad_norm": 0.00036280223866924644,
      "learning_rate": 0.2122206618454127,
      "loss": 0.2861,
      "num_input_tokens_seen": 11102008,
      "step": 14555
    },
    {
      "epoch": 30.27027027027027,
      "grad_norm": 0.00014951849880162627,
      "learning_rate": 0.2121670588906439,
      "loss": 0.2715,
      "num_input_tokens_seen": 11105784,
      "step": 14560
    },
    {
      "epoch": 30.280665280665282,
      "grad_norm": 5.922588024986908e-05,
      "learning_rate": 0.21211344634893345,
      "loss": 0.2498,
      "num_input_tokens_seen": 11109528,
      "step": 14565
    },
    {
      "epoch": 30.29106029106029,
      "grad_norm": 0.00031049034441821277,
      "learning_rate": 0.21205982422854897,
      "loss": 0.2563,
      "num_input_tokens_seen": 11113272,
      "step": 14570
    },
    {
      "epoch": 30.301455301455302,
      "grad_norm": 0.00013043484068475664,
      "learning_rate": 0.21200619253775974,
      "loss": 0.2424,
      "num_input_tokens_seen": 11117048,
      "step": 14575
    },
    {
      "epoch": 30.31185031185031,
      "grad_norm": 0.00012856299872510135,
      "learning_rate": 0.21195255128483637,
      "loss": 0.2657,
      "num_input_tokens_seen": 11120792,
      "step": 14580
    },
    {
      "epoch": 30.322245322245323,
      "grad_norm": 0.00017962511628866196,
      "learning_rate": 0.21189890047805102,
      "loss": 0.2805,
      "num_input_tokens_seen": 11124600,
      "step": 14585
    },
    {
      "epoch": 30.33264033264033,
      "grad_norm": 0.00015870440984144807,
      "learning_rate": 0.21184524012567735,
      "loss": 0.2646,
      "num_input_tokens_seen": 11128408,
      "step": 14590
    },
    {
      "epoch": 30.343035343035343,
      "grad_norm": 0.0002069677138933912,
      "learning_rate": 0.2117915702359905,
      "loss": 0.2592,
      "num_input_tokens_seen": 11132120,
      "step": 14595
    },
    {
      "epoch": 30.353430353430355,
      "grad_norm": 0.000353456474840641,
      "learning_rate": 0.211737890817267,
      "loss": 0.2916,
      "num_input_tokens_seen": 11135896,
      "step": 14600
    },
    {
      "epoch": 30.353430353430355,
      "eval_loss": 0.2486879974603653,
      "eval_runtime": 13.4204,
      "eval_samples_per_second": 63.783,
      "eval_steps_per_second": 15.946,
      "num_input_tokens_seen": 11135896,
      "step": 14600
    },
    {
      "epoch": 30.363825363825363,
      "grad_norm": 0.0002123817102983594,
      "learning_rate": 0.21168420187778483,
      "loss": 0.2508,
      "num_input_tokens_seen": 11139672,
      "step": 14605
    },
    {
      "epoch": 30.374220374220375,
      "grad_norm": 0.00018537460709922016,
      "learning_rate": 0.21163050342582362,
      "loss": 0.2418,
      "num_input_tokens_seen": 11143480,
      "step": 14610
    },
    {
      "epoch": 30.384615384615383,
      "grad_norm": 0.00010919541091425344,
      "learning_rate": 0.21157679546966426,
      "loss": 0.2786,
      "num_input_tokens_seen": 11147416,
      "step": 14615
    },
    {
      "epoch": 30.395010395010395,
      "grad_norm": 0.0005122404545545578,
      "learning_rate": 0.2115230780175892,
      "loss": 0.2537,
      "num_input_tokens_seen": 11151128,
      "step": 14620
    },
    {
      "epoch": 30.405405405405407,
      "grad_norm": 0.0005644683260470629,
      "learning_rate": 0.21146935107788237,
      "loss": 0.2852,
      "num_input_tokens_seen": 11154840,
      "step": 14625
    },
    {
      "epoch": 30.415800415800415,
      "grad_norm": 0.0005030115135014057,
      "learning_rate": 0.21141561465882916,
      "loss": 0.2782,
      "num_input_tokens_seen": 11158584,
      "step": 14630
    },
    {
      "epoch": 30.426195426195427,
      "grad_norm": 0.00010961908265016973,
      "learning_rate": 0.21136186876871635,
      "loss": 0.2733,
      "num_input_tokens_seen": 11162712,
      "step": 14635
    },
    {
      "epoch": 30.436590436590436,
      "grad_norm": 5.42264933756087e-05,
      "learning_rate": 0.21130811341583225,
      "loss": 0.2578,
      "num_input_tokens_seen": 11166520,
      "step": 14640
    },
    {
      "epoch": 30.446985446985448,
      "grad_norm": 0.0004980212543159723,
      "learning_rate": 0.21125434860846667,
      "loss": 0.2776,
      "num_input_tokens_seen": 11170264,
      "step": 14645
    },
    {
      "epoch": 30.457380457380456,
      "grad_norm": 0.0004620563704520464,
      "learning_rate": 0.2112005743549107,
      "loss": 0.2673,
      "num_input_tokens_seen": 11174072,
      "step": 14650
    },
    {
      "epoch": 30.467775467775468,
      "grad_norm": 0.0003817227843683213,
      "learning_rate": 0.21114679066345707,
      "loss": 0.283,
      "num_input_tokens_seen": 11177880,
      "step": 14655
    },
    {
      "epoch": 30.47817047817048,
      "grad_norm": 0.00012710904411505908,
      "learning_rate": 0.21109299754239993,
      "loss": 0.2833,
      "num_input_tokens_seen": 11181720,
      "step": 14660
    },
    {
      "epoch": 30.488565488565488,
      "grad_norm": 0.00015433464432135224,
      "learning_rate": 0.21103919500003482,
      "loss": 0.2677,
      "num_input_tokens_seen": 11185560,
      "step": 14665
    },
    {
      "epoch": 30.4989604989605,
      "grad_norm": 0.0003585397789720446,
      "learning_rate": 0.21098538304465872,
      "loss": 0.2797,
      "num_input_tokens_seen": 11189368,
      "step": 14670
    },
    {
      "epoch": 30.509355509355508,
      "grad_norm": 0.0005723215290345252,
      "learning_rate": 0.2109315616845702,
      "loss": 0.2888,
      "num_input_tokens_seen": 11193112,
      "step": 14675
    },
    {
      "epoch": 30.51975051975052,
      "grad_norm": 0.0002040408580796793,
      "learning_rate": 0.21087773092806925,
      "loss": 0.2808,
      "num_input_tokens_seen": 11197016,
      "step": 14680
    },
    {
      "epoch": 30.53014553014553,
      "grad_norm": 8.161534060491249e-05,
      "learning_rate": 0.21082389078345704,
      "loss": 0.2717,
      "num_input_tokens_seen": 11200888,
      "step": 14685
    },
    {
      "epoch": 30.54054054054054,
      "grad_norm": 0.00043451960664242506,
      "learning_rate": 0.2107700412590365,
      "loss": 0.277,
      "num_input_tokens_seen": 11204632,
      "step": 14690
    },
    {
      "epoch": 30.550935550935552,
      "grad_norm": 0.0005331505089998245,
      "learning_rate": 0.210716182363112,
      "loss": 0.2622,
      "num_input_tokens_seen": 11208536,
      "step": 14695
    },
    {
      "epoch": 30.56133056133056,
      "grad_norm": 0.00018044031457975507,
      "learning_rate": 0.2106623141039891,
      "loss": 0.2398,
      "num_input_tokens_seen": 11212440,
      "step": 14700
    },
    {
      "epoch": 30.571725571725572,
      "grad_norm": 0.00010524292156333104,
      "learning_rate": 0.21060843648997507,
      "loss": 0.2758,
      "num_input_tokens_seen": 11216376,
      "step": 14705
    },
    {
      "epoch": 30.58212058212058,
      "grad_norm": 0.0001528459251858294,
      "learning_rate": 0.21055454952937844,
      "loss": 0.2693,
      "num_input_tokens_seen": 11220184,
      "step": 14710
    },
    {
      "epoch": 30.592515592515593,
      "grad_norm": 0.00018082439783029258,
      "learning_rate": 0.21050065323050937,
      "loss": 0.2674,
      "num_input_tokens_seen": 11223928,
      "step": 14715
    },
    {
      "epoch": 30.602910602910605,
      "grad_norm": 0.00011943643767153844,
      "learning_rate": 0.21044674760167928,
      "loss": 0.2776,
      "num_input_tokens_seen": 11227736,
      "step": 14720
    },
    {
      "epoch": 30.613305613305613,
      "grad_norm": 0.0004058046324644238,
      "learning_rate": 0.210392832651201,
      "loss": 0.2646,
      "num_input_tokens_seen": 11231736,
      "step": 14725
    },
    {
      "epoch": 30.623700623700625,
      "grad_norm": 0.0002450610918458551,
      "learning_rate": 0.210338908387389,
      "loss": 0.2446,
      "num_input_tokens_seen": 11235480,
      "step": 14730
    },
    {
      "epoch": 30.634095634095633,
      "grad_norm": 0.0001955005864147097,
      "learning_rate": 0.21028497481855912,
      "loss": 0.2746,
      "num_input_tokens_seen": 11239160,
      "step": 14735
    },
    {
      "epoch": 30.644490644490645,
      "grad_norm": 0.0002759649360086769,
      "learning_rate": 0.21023103195302847,
      "loss": 0.2292,
      "num_input_tokens_seen": 11242968,
      "step": 14740
    },
    {
      "epoch": 30.654885654885653,
      "grad_norm": 0.0003272598551120609,
      "learning_rate": 0.21017707979911582,
      "loss": 0.2753,
      "num_input_tokens_seen": 11246840,
      "step": 14745
    },
    {
      "epoch": 30.665280665280665,
      "grad_norm": 0.00018998980522155762,
      "learning_rate": 0.21012311836514122,
      "loss": 0.243,
      "num_input_tokens_seen": 11250648,
      "step": 14750
    },
    {
      "epoch": 30.675675675675677,
      "grad_norm": 0.0002959202975034714,
      "learning_rate": 0.21006914765942622,
      "loss": 0.2692,
      "num_input_tokens_seen": 11254424,
      "step": 14755
    },
    {
      "epoch": 30.686070686070686,
      "grad_norm": 0.0002520527341403067,
      "learning_rate": 0.2100151676902938,
      "loss": 0.2704,
      "num_input_tokens_seen": 11258136,
      "step": 14760
    },
    {
      "epoch": 30.696465696465697,
      "grad_norm": 0.0004911919822916389,
      "learning_rate": 0.2099611784660683,
      "loss": 0.2802,
      "num_input_tokens_seen": 11262104,
      "step": 14765
    },
    {
      "epoch": 30.706860706860706,
      "grad_norm": 0.00017487841250840575,
      "learning_rate": 0.20990717999507552,
      "loss": 0.2659,
      "num_input_tokens_seen": 11265816,
      "step": 14770
    },
    {
      "epoch": 30.717255717255718,
      "grad_norm": 0.0005485157016664743,
      "learning_rate": 0.20985317228564276,
      "loss": 0.2616,
      "num_input_tokens_seen": 11269656,
      "step": 14775
    },
    {
      "epoch": 30.727650727650726,
      "grad_norm": 0.0004643997235689312,
      "learning_rate": 0.20979915534609872,
      "loss": 0.3221,
      "num_input_tokens_seen": 11273400,
      "step": 14780
    },
    {
      "epoch": 30.738045738045738,
      "grad_norm": 7.572733011329547e-05,
      "learning_rate": 0.20974512918477342,
      "loss": 0.2686,
      "num_input_tokens_seen": 11277240,
      "step": 14785
    },
    {
      "epoch": 30.74844074844075,
      "grad_norm": 0.0007108752615749836,
      "learning_rate": 0.2096910938099984,
      "loss": 0.2893,
      "num_input_tokens_seen": 11281080,
      "step": 14790
    },
    {
      "epoch": 30.758835758835758,
      "grad_norm": 9.185164526570588e-05,
      "learning_rate": 0.2096370492301066,
      "loss": 0.2728,
      "num_input_tokens_seen": 11284920,
      "step": 14795
    },
    {
      "epoch": 30.76923076923077,
      "grad_norm": 0.00011459348752396181,
      "learning_rate": 0.2095829954534323,
      "loss": 0.2517,
      "num_input_tokens_seen": 11288728,
      "step": 14800
    },
    {
      "epoch": 30.76923076923077,
      "eval_loss": 0.25030773878097534,
      "eval_runtime": 13.5245,
      "eval_samples_per_second": 63.292,
      "eval_steps_per_second": 15.823,
      "num_input_tokens_seen": 11288728,
      "step": 14800
    },
    {
      "epoch": 30.77962577962578,
      "grad_norm": 5.75693447899539e-05,
      "learning_rate": 0.2095289324883114,
      "loss": 0.2984,
      "num_input_tokens_seen": 11292728,
      "step": 14805
    },
    {
      "epoch": 30.79002079002079,
      "grad_norm": 0.00034482585033401847,
      "learning_rate": 0.20947486034308097,
      "loss": 0.2827,
      "num_input_tokens_seen": 11296440,
      "step": 14810
    },
    {
      "epoch": 30.8004158004158,
      "grad_norm": 0.00020812460570596159,
      "learning_rate": 0.2094207790260797,
      "loss": 0.2655,
      "num_input_tokens_seen": 11300120,
      "step": 14815
    },
    {
      "epoch": 30.81081081081081,
      "grad_norm": 0.00023784405493643135,
      "learning_rate": 0.20936668854564758,
      "loss": 0.2549,
      "num_input_tokens_seen": 11303928,
      "step": 14820
    },
    {
      "epoch": 30.821205821205822,
      "grad_norm": 0.0002435097994748503,
      "learning_rate": 0.20931258891012602,
      "loss": 0.2623,
      "num_input_tokens_seen": 11307672,
      "step": 14825
    },
    {
      "epoch": 30.83160083160083,
      "grad_norm": 0.00035023706732317805,
      "learning_rate": 0.20925848012785792,
      "loss": 0.2715,
      "num_input_tokens_seen": 11311448,
      "step": 14830
    },
    {
      "epoch": 30.841995841995843,
      "grad_norm": 0.0004942963132634759,
      "learning_rate": 0.20920436220718747,
      "loss": 0.2658,
      "num_input_tokens_seen": 11315192,
      "step": 14835
    },
    {
      "epoch": 30.85239085239085,
      "grad_norm": 0.00016547893756069243,
      "learning_rate": 0.20915023515646033,
      "loss": 0.2668,
      "num_input_tokens_seen": 11319128,
      "step": 14840
    },
    {
      "epoch": 30.862785862785863,
      "grad_norm": 0.0004132789617870003,
      "learning_rate": 0.20909609898402368,
      "loss": 0.243,
      "num_input_tokens_seen": 11322840,
      "step": 14845
    },
    {
      "epoch": 30.873180873180875,
      "grad_norm": 0.00033745967084541917,
      "learning_rate": 0.2090419536982258,
      "loss": 0.2712,
      "num_input_tokens_seen": 11326584,
      "step": 14850
    },
    {
      "epoch": 30.883575883575883,
      "grad_norm": 0.000425832491600886,
      "learning_rate": 0.2089877993074168,
      "loss": 0.2436,
      "num_input_tokens_seen": 11330616,
      "step": 14855
    },
    {
      "epoch": 30.893970893970895,
      "grad_norm": 7.139249646570534e-05,
      "learning_rate": 0.20893363581994784,
      "loss": 0.2432,
      "num_input_tokens_seen": 11334488,
      "step": 14860
    },
    {
      "epoch": 30.904365904365903,
      "grad_norm": 0.000588972819969058,
      "learning_rate": 0.2088794632441716,
      "loss": 0.2519,
      "num_input_tokens_seen": 11338200,
      "step": 14865
    },
    {
      "epoch": 30.914760914760915,
      "grad_norm": 6.970312097109854e-05,
      "learning_rate": 0.20882528158844219,
      "loss": 0.276,
      "num_input_tokens_seen": 11342072,
      "step": 14870
    },
    {
      "epoch": 30.925155925155924,
      "grad_norm": 0.0005155346007086337,
      "learning_rate": 0.20877109086111514,
      "loss": 0.2696,
      "num_input_tokens_seen": 11345688,
      "step": 14875
    },
    {
      "epoch": 30.935550935550935,
      "grad_norm": 0.000572209304664284,
      "learning_rate": 0.2087168910705473,
      "loss": 0.2827,
      "num_input_tokens_seen": 11349528,
      "step": 14880
    },
    {
      "epoch": 30.945945945945947,
      "grad_norm": 5.021830293117091e-05,
      "learning_rate": 0.208662682225097,
      "loss": 0.2746,
      "num_input_tokens_seen": 11353272,
      "step": 14885
    },
    {
      "epoch": 30.956340956340956,
      "grad_norm": 0.0005160231376066804,
      "learning_rate": 0.2086084643331239,
      "loss": 0.2774,
      "num_input_tokens_seen": 11357080,
      "step": 14890
    },
    {
      "epoch": 30.966735966735968,
      "grad_norm": 0.00018559713498689234,
      "learning_rate": 0.20855423740298906,
      "loss": 0.2839,
      "num_input_tokens_seen": 11360984,
      "step": 14895
    },
    {
      "epoch": 30.977130977130976,
      "grad_norm": 7.044704398140311e-05,
      "learning_rate": 0.208500001443055,
      "loss": 0.2645,
      "num_input_tokens_seen": 11364696,
      "step": 14900
    },
    {
      "epoch": 30.987525987525988,
      "grad_norm": 0.00016494252486154437,
      "learning_rate": 0.20844575646168553,
      "loss": 0.2448,
      "num_input_tokens_seen": 11368568,
      "step": 14905
    },
    {
      "epoch": 30.997920997921,
      "grad_norm": 0.00010968758579110727,
      "learning_rate": 0.20839150246724594,
      "loss": 0.2737,
      "num_input_tokens_seen": 11372376,
      "step": 14910
    },
    {
      "epoch": 31.008316008316008,
      "grad_norm": 0.00016372599930036813,
      "learning_rate": 0.20833723946810287,
      "loss": 0.2649,
      "num_input_tokens_seen": 11376080,
      "step": 14915
    },
    {
      "epoch": 31.01871101871102,
      "grad_norm": 0.0007427406380884349,
      "learning_rate": 0.20828296747262437,
      "loss": 0.2638,
      "num_input_tokens_seen": 11379856,
      "step": 14920
    },
    {
      "epoch": 31.02910602910603,
      "grad_norm": 0.0005963361472822726,
      "learning_rate": 0.20822868648917986,
      "loss": 0.2714,
      "num_input_tokens_seen": 11383664,
      "step": 14925
    },
    {
      "epoch": 31.03950103950104,
      "grad_norm": 0.0003358885005582124,
      "learning_rate": 0.20817439652614017,
      "loss": 0.2511,
      "num_input_tokens_seen": 11387440,
      "step": 14930
    },
    {
      "epoch": 31.04989604989605,
      "grad_norm": 0.00011668734805425629,
      "learning_rate": 0.20812009759187744,
      "loss": 0.2909,
      "num_input_tokens_seen": 11391280,
      "step": 14935
    },
    {
      "epoch": 31.06029106029106,
      "grad_norm": 0.00025887053925544024,
      "learning_rate": 0.2080657896947653,
      "loss": 0.2524,
      "num_input_tokens_seen": 11395152,
      "step": 14940
    },
    {
      "epoch": 31.070686070686072,
      "grad_norm": 0.00021860796550754458,
      "learning_rate": 0.2080114728431787,
      "loss": 0.2712,
      "num_input_tokens_seen": 11399024,
      "step": 14945
    },
    {
      "epoch": 31.08108108108108,
      "grad_norm": 0.0001636134402360767,
      "learning_rate": 0.20795714704549392,
      "loss": 0.2585,
      "num_input_tokens_seen": 11402960,
      "step": 14950
    },
    {
      "epoch": 31.091476091476093,
      "grad_norm": 0.000590287905652076,
      "learning_rate": 0.20790281231008875,
      "loss": 0.2704,
      "num_input_tokens_seen": 11406736,
      "step": 14955
    },
    {
      "epoch": 31.1018711018711,
      "grad_norm": 0.000144617966725491,
      "learning_rate": 0.20784846864534226,
      "loss": 0.2705,
      "num_input_tokens_seen": 11410704,
      "step": 14960
    },
    {
      "epoch": 31.112266112266113,
      "grad_norm": 0.0011210811790078878,
      "learning_rate": 0.20779411605963496,
      "loss": 0.2793,
      "num_input_tokens_seen": 11414480,
      "step": 14965
    },
    {
      "epoch": 31.12266112266112,
      "grad_norm": 6.359563121804968e-05,
      "learning_rate": 0.2077397545613487,
      "loss": 0.2526,
      "num_input_tokens_seen": 11418320,
      "step": 14970
    },
    {
      "epoch": 31.133056133056133,
      "grad_norm": 0.0005163986934348941,
      "learning_rate": 0.20768538415886661,
      "loss": 0.272,
      "num_input_tokens_seen": 11422128,
      "step": 14975
    },
    {
      "epoch": 31.143451143451145,
      "grad_norm": 0.00028234333149157465,
      "learning_rate": 0.20763100486057343,
      "loss": 0.2513,
      "num_input_tokens_seen": 11426096,
      "step": 14980
    },
    {
      "epoch": 31.153846153846153,
      "grad_norm": 0.00027606377261690795,
      "learning_rate": 0.20757661667485502,
      "loss": 0.2627,
      "num_input_tokens_seen": 11429808,
      "step": 14985
    },
    {
      "epoch": 31.164241164241165,
      "grad_norm": 0.0005827684071846306,
      "learning_rate": 0.2075222196100988,
      "loss": 0.2672,
      "num_input_tokens_seen": 11433520,
      "step": 14990
    },
    {
      "epoch": 31.174636174636174,
      "grad_norm": 0.00015686712868046016,
      "learning_rate": 0.20746781367469344,
      "loss": 0.2604,
      "num_input_tokens_seen": 11437264,
      "step": 14995
    },
    {
      "epoch": 31.185031185031185,
      "grad_norm": 0.00039666733937337995,
      "learning_rate": 0.207413398877029,
      "loss": 0.2847,
      "num_input_tokens_seen": 11441040,
      "step": 15000
    },
    {
      "epoch": 31.185031185031185,
      "eval_loss": 0.24743162095546722,
      "eval_runtime": 13.4157,
      "eval_samples_per_second": 63.806,
      "eval_steps_per_second": 15.951,
      "num_input_tokens_seen": 11441040,
      "step": 15000
    },
    {
      "epoch": 31.195426195426194,
      "grad_norm": 0.00020095422223675996,
      "learning_rate": 0.20735897522549698,
      "loss": 0.2329,
      "num_input_tokens_seen": 11444976,
      "step": 15005
    },
    {
      "epoch": 31.205821205821206,
      "grad_norm": 0.00011427220306359231,
      "learning_rate": 0.2073045427284902,
      "loss": 0.278,
      "num_input_tokens_seen": 11448720,
      "step": 15010
    },
    {
      "epoch": 31.216216216216218,
      "grad_norm": 0.0003669828874990344,
      "learning_rate": 0.2072501013944027,
      "loss": 0.2476,
      "num_input_tokens_seen": 11452624,
      "step": 15015
    },
    {
      "epoch": 31.226611226611226,
      "grad_norm": 0.0003313622728455812,
      "learning_rate": 0.20719565123163017,
      "loss": 0.2786,
      "num_input_tokens_seen": 11456400,
      "step": 15020
    },
    {
      "epoch": 31.237006237006238,
      "grad_norm": 0.0005790401482954621,
      "learning_rate": 0.20714119224856944,
      "loss": 0.2588,
      "num_input_tokens_seen": 11460208,
      "step": 15025
    },
    {
      "epoch": 31.247401247401246,
      "grad_norm": 0.00020884226250927895,
      "learning_rate": 0.2070867244536188,
      "loss": 0.2597,
      "num_input_tokens_seen": 11464112,
      "step": 15030
    },
    {
      "epoch": 31.257796257796258,
      "grad_norm": 0.00036518240813165903,
      "learning_rate": 0.20703224785517785,
      "loss": 0.2772,
      "num_input_tokens_seen": 11467856,
      "step": 15035
    },
    {
      "epoch": 31.26819126819127,
      "grad_norm": 0.0004984860424883664,
      "learning_rate": 0.20697776246164754,
      "loss": 0.2683,
      "num_input_tokens_seen": 11471632,
      "step": 15040
    },
    {
      "epoch": 31.27858627858628,
      "grad_norm": 0.00014108205505181104,
      "learning_rate": 0.2069232682814303,
      "loss": 0.2393,
      "num_input_tokens_seen": 11475536,
      "step": 15045
    },
    {
      "epoch": 31.28898128898129,
      "grad_norm": 0.0004306963237468153,
      "learning_rate": 0.20686876532292972,
      "loss": 0.2755,
      "num_input_tokens_seen": 11479376,
      "step": 15050
    },
    {
      "epoch": 31.2993762993763,
      "grad_norm": 0.0003535844443831593,
      "learning_rate": 0.20681425359455083,
      "loss": 0.2786,
      "num_input_tokens_seen": 11483056,
      "step": 15055
    },
    {
      "epoch": 31.30977130977131,
      "grad_norm": 0.0004925647517666221,
      "learning_rate": 0.20675973310470008,
      "loss": 0.2736,
      "num_input_tokens_seen": 11486864,
      "step": 15060
    },
    {
      "epoch": 31.32016632016632,
      "grad_norm": 0.0006180040654726326,
      "learning_rate": 0.2067052038617852,
      "loss": 0.2758,
      "num_input_tokens_seen": 11490480,
      "step": 15065
    },
    {
      "epoch": 31.33056133056133,
      "grad_norm": 0.0008966726600192487,
      "learning_rate": 0.2066506658742153,
      "loss": 0.285,
      "num_input_tokens_seen": 11494288,
      "step": 15070
    },
    {
      "epoch": 31.340956340956343,
      "grad_norm": 0.00029795506270602345,
      "learning_rate": 0.20659611915040077,
      "loss": 0.254,
      "num_input_tokens_seen": 11498000,
      "step": 15075
    },
    {
      "epoch": 31.35135135135135,
      "grad_norm": 0.0002672720293048769,
      "learning_rate": 0.20654156369875348,
      "loss": 0.2572,
      "num_input_tokens_seen": 11501744,
      "step": 15080
    },
    {
      "epoch": 31.361746361746363,
      "grad_norm": 0.0005981008871458471,
      "learning_rate": 0.20648699952768648,
      "loss": 0.2419,
      "num_input_tokens_seen": 11505520,
      "step": 15085
    },
    {
      "epoch": 31.37214137214137,
      "grad_norm": 6.665072578471154e-05,
      "learning_rate": 0.20643242664561437,
      "loss": 0.26,
      "num_input_tokens_seen": 11509360,
      "step": 15090
    },
    {
      "epoch": 31.382536382536383,
      "grad_norm": 0.00021264624956529588,
      "learning_rate": 0.20637784506095277,
      "loss": 0.2713,
      "num_input_tokens_seen": 11513200,
      "step": 15095
    },
    {
      "epoch": 31.39293139293139,
      "grad_norm": 0.0002655001007951796,
      "learning_rate": 0.20632325478211908,
      "loss": 0.2741,
      "num_input_tokens_seen": 11516880,
      "step": 15100
    },
    {
      "epoch": 31.403326403326403,
      "grad_norm": 0.0005713010323233902,
      "learning_rate": 0.20626865581753165,
      "loss": 0.2606,
      "num_input_tokens_seen": 11520560,
      "step": 15105
    },
    {
      "epoch": 31.413721413721415,
      "grad_norm": 0.0005730713601224124,
      "learning_rate": 0.2062140481756104,
      "loss": 0.2967,
      "num_input_tokens_seen": 11524368,
      "step": 15110
    },
    {
      "epoch": 31.424116424116423,
      "grad_norm": 0.0005838734796270728,
      "learning_rate": 0.20615943186477648,
      "loss": 0.2826,
      "num_input_tokens_seen": 11528176,
      "step": 15115
    },
    {
      "epoch": 31.434511434511435,
      "grad_norm": 0.0001749437506077811,
      "learning_rate": 0.20610480689345242,
      "loss": 0.272,
      "num_input_tokens_seen": 11532016,
      "step": 15120
    },
    {
      "epoch": 31.444906444906444,
      "grad_norm": 0.0004875649174209684,
      "learning_rate": 0.2060501732700621,
      "loss": 0.2556,
      "num_input_tokens_seen": 11535920,
      "step": 15125
    },
    {
      "epoch": 31.455301455301456,
      "grad_norm": 0.00014699202438350767,
      "learning_rate": 0.20599553100303067,
      "loss": 0.2763,
      "num_input_tokens_seen": 11539664,
      "step": 15130
    },
    {
      "epoch": 31.465696465696467,
      "grad_norm": 0.001029477920383215,
      "learning_rate": 0.20594088010078465,
      "loss": 0.2565,
      "num_input_tokens_seen": 11543568,
      "step": 15135
    },
    {
      "epoch": 31.476091476091476,
      "grad_norm": 0.00029580388218164444,
      "learning_rate": 0.20588622057175196,
      "loss": 0.2474,
      "num_input_tokens_seen": 11547472,
      "step": 15140
    },
    {
      "epoch": 31.486486486486488,
      "grad_norm": 0.00011070615437347442,
      "learning_rate": 0.20583155242436177,
      "loss": 0.2729,
      "num_input_tokens_seen": 11551312,
      "step": 15145
    },
    {
      "epoch": 31.496881496881496,
      "grad_norm": 0.0001253844820894301,
      "learning_rate": 0.20577687566704453,
      "loss": 0.2581,
      "num_input_tokens_seen": 11555184,
      "step": 15150
    },
    {
      "epoch": 31.507276507276508,
      "grad_norm": 0.00015230536519084126,
      "learning_rate": 0.20572219030823213,
      "loss": 0.2554,
      "num_input_tokens_seen": 11559056,
      "step": 15155
    },
    {
      "epoch": 31.517671517671516,
      "grad_norm": 0.0002877760271076113,
      "learning_rate": 0.20566749635635775,
      "loss": 0.2568,
      "num_input_tokens_seen": 11562864,
      "step": 15160
    },
    {
      "epoch": 31.528066528066528,
      "grad_norm": 0.0002018535597017035,
      "learning_rate": 0.20561279381985587,
      "loss": 0.289,
      "num_input_tokens_seen": 11566704,
      "step": 15165
    },
    {
      "epoch": 31.53846153846154,
      "grad_norm": 0.00036334648029878736,
      "learning_rate": 0.2055580827071623,
      "loss": 0.2707,
      "num_input_tokens_seen": 11570512,
      "step": 15170
    },
    {
      "epoch": 31.54885654885655,
      "grad_norm": 0.0006616413593292236,
      "learning_rate": 0.20550336302671418,
      "loss": 0.2698,
      "num_input_tokens_seen": 11574320,
      "step": 15175
    },
    {
      "epoch": 31.55925155925156,
      "grad_norm": 0.00012892454105895013,
      "learning_rate": 0.20544863478695,
      "loss": 0.2399,
      "num_input_tokens_seen": 11578160,
      "step": 15180
    },
    {
      "epoch": 31.56964656964657,
      "grad_norm": 0.0002216225693700835,
      "learning_rate": 0.20539389799630953,
      "loss": 0.2881,
      "num_input_tokens_seen": 11582000,
      "step": 15185
    },
    {
      "epoch": 31.58004158004158,
      "grad_norm": 0.00015289570728782564,
      "learning_rate": 0.20533915266323388,
      "loss": 0.2751,
      "num_input_tokens_seen": 11585776,
      "step": 15190
    },
    {
      "epoch": 31.59043659043659,
      "grad_norm": 0.0011692427797243,
      "learning_rate": 0.20528439879616542,
      "loss": 0.2604,
      "num_input_tokens_seen": 11589552,
      "step": 15195
    },
    {
      "epoch": 31.6008316008316,
      "grad_norm": 0.000329432834405452,
      "learning_rate": 0.20522963640354794,
      "loss": 0.2556,
      "num_input_tokens_seen": 11593456,
      "step": 15200
    },
    {
      "epoch": 31.6008316008316,
      "eval_loss": 0.25349685549736023,
      "eval_runtime": 13.4167,
      "eval_samples_per_second": 63.801,
      "eval_steps_per_second": 15.95,
      "num_input_tokens_seen": 11593456,
      "step": 15200
    },
    {
      "epoch": 31.611226611226613,
      "grad_norm": 0.00016098761989269406,
      "learning_rate": 0.20517486549382644,
      "loss": 0.2759,
      "num_input_tokens_seen": 11597232,
      "step": 15205
    },
    {
      "epoch": 31.62162162162162,
      "grad_norm": 0.0003965785726904869,
      "learning_rate": 0.20512008607544735,
      "loss": 0.2576,
      "num_input_tokens_seen": 11600816,
      "step": 15210
    },
    {
      "epoch": 31.632016632016633,
      "grad_norm": 0.00016998022329062223,
      "learning_rate": 0.20506529815685826,
      "loss": 0.285,
      "num_input_tokens_seen": 11604688,
      "step": 15215
    },
    {
      "epoch": 31.64241164241164,
      "grad_norm": 0.0002748713013716042,
      "learning_rate": 0.2050105017465082,
      "loss": 0.2781,
      "num_input_tokens_seen": 11608528,
      "step": 15220
    },
    {
      "epoch": 31.652806652806653,
      "grad_norm": 2.1838486645719968e-05,
      "learning_rate": 0.20495569685284754,
      "loss": 0.2684,
      "num_input_tokens_seen": 11612304,
      "step": 15225
    },
    {
      "epoch": 31.66320166320166,
      "grad_norm": 0.0005906136939302087,
      "learning_rate": 0.20490088348432778,
      "loss": 0.2442,
      "num_input_tokens_seen": 11616272,
      "step": 15230
    },
    {
      "epoch": 31.673596673596673,
      "grad_norm": 0.00036783976247534156,
      "learning_rate": 0.2048460616494018,
      "loss": 0.2841,
      "num_input_tokens_seen": 11620272,
      "step": 15235
    },
    {
      "epoch": 31.683991683991685,
      "grad_norm": 0.0005082294810563326,
      "learning_rate": 0.2047912313565239,
      "loss": 0.2736,
      "num_input_tokens_seen": 11624144,
      "step": 15240
    },
    {
      "epoch": 31.694386694386694,
      "grad_norm": 8.031754987314343e-05,
      "learning_rate": 0.20473639261414958,
      "loss": 0.2546,
      "num_input_tokens_seen": 11627888,
      "step": 15245
    },
    {
      "epoch": 31.704781704781706,
      "grad_norm": 0.000283773522824049,
      "learning_rate": 0.2046815454307357,
      "loss": 0.2678,
      "num_input_tokens_seen": 11631696,
      "step": 15250
    },
    {
      "epoch": 31.715176715176714,
      "grad_norm": 7.317338895518333e-05,
      "learning_rate": 0.20462668981474028,
      "loss": 0.2565,
      "num_input_tokens_seen": 11635472,
      "step": 15255
    },
    {
      "epoch": 31.725571725571726,
      "grad_norm": 0.00017082841077353805,
      "learning_rate": 0.20457182577462288,
      "loss": 0.2636,
      "num_input_tokens_seen": 11639248,
      "step": 15260
    },
    {
      "epoch": 31.735966735966738,
      "grad_norm": 6.815803499193862e-05,
      "learning_rate": 0.2045169533188441,
      "loss": 0.2543,
      "num_input_tokens_seen": 11643024,
      "step": 15265
    },
    {
      "epoch": 31.746361746361746,
      "grad_norm": 6.431232031900436e-05,
      "learning_rate": 0.20446207245586603,
      "loss": 0.2748,
      "num_input_tokens_seen": 11646832,
      "step": 15270
    },
    {
      "epoch": 31.756756756756758,
      "grad_norm": 0.00018241336510982364,
      "learning_rate": 0.20440718319415196,
      "loss": 0.2701,
      "num_input_tokens_seen": 11650480,
      "step": 15275
    },
    {
      "epoch": 31.767151767151766,
      "grad_norm": 0.0007018996984697878,
      "learning_rate": 0.20435228554216653,
      "loss": 0.27,
      "num_input_tokens_seen": 11654480,
      "step": 15280
    },
    {
      "epoch": 31.777546777546778,
      "grad_norm": 0.0004037163162138313,
      "learning_rate": 0.20429737950837565,
      "loss": 0.2568,
      "num_input_tokens_seen": 11658224,
      "step": 15285
    },
    {
      "epoch": 31.787941787941786,
      "grad_norm": 0.00017245157505385578,
      "learning_rate": 0.20424246510124647,
      "loss": 0.2476,
      "num_input_tokens_seen": 11662160,
      "step": 15290
    },
    {
      "epoch": 31.7983367983368,
      "grad_norm": 0.0002721804194152355,
      "learning_rate": 0.20418754232924755,
      "loss": 0.2747,
      "num_input_tokens_seen": 11665904,
      "step": 15295
    },
    {
      "epoch": 31.80873180873181,
      "grad_norm": 0.00017300549370702356,
      "learning_rate": 0.20413261120084863,
      "loss": 0.2964,
      "num_input_tokens_seen": 11669616,
      "step": 15300
    },
    {
      "epoch": 31.81912681912682,
      "grad_norm": 0.0009127103257924318,
      "learning_rate": 0.2040776717245208,
      "loss": 0.2895,
      "num_input_tokens_seen": 11673392,
      "step": 15305
    },
    {
      "epoch": 31.82952182952183,
      "grad_norm": 0.0006942368927411735,
      "learning_rate": 0.2040227239087364,
      "loss": 0.2778,
      "num_input_tokens_seen": 11677264,
      "step": 15310
    },
    {
      "epoch": 31.83991683991684,
      "grad_norm": 0.0005713171558454633,
      "learning_rate": 0.20396776776196904,
      "loss": 0.2931,
      "num_input_tokens_seen": 11680912,
      "step": 15315
    },
    {
      "epoch": 31.85031185031185,
      "grad_norm": 0.000258445244980976,
      "learning_rate": 0.20391280329269373,
      "loss": 0.2958,
      "num_input_tokens_seen": 11684752,
      "step": 15320
    },
    {
      "epoch": 31.86070686070686,
      "grad_norm": 0.0003681774251163006,
      "learning_rate": 0.20385783050938663,
      "loss": 0.2811,
      "num_input_tokens_seen": 11688496,
      "step": 15325
    },
    {
      "epoch": 31.87110187110187,
      "grad_norm": 0.00021278730127960443,
      "learning_rate": 0.20380284942052526,
      "loss": 0.28,
      "num_input_tokens_seen": 11692368,
      "step": 15330
    },
    {
      "epoch": 31.881496881496883,
      "grad_norm": 0.00010583264520391822,
      "learning_rate": 0.2037478600345884,
      "loss": 0.2759,
      "num_input_tokens_seen": 11696304,
      "step": 15335
    },
    {
      "epoch": 31.89189189189189,
      "grad_norm": 0.0002900429244618863,
      "learning_rate": 0.20369286236005604,
      "loss": 0.2822,
      "num_input_tokens_seen": 11700048,
      "step": 15340
    },
    {
      "epoch": 31.902286902286903,
      "grad_norm": 9.922471508616582e-05,
      "learning_rate": 0.20363785640540957,
      "loss": 0.2824,
      "num_input_tokens_seen": 11703856,
      "step": 15345
    },
    {
      "epoch": 31.91268191268191,
      "grad_norm": 0.0008042150875553489,
      "learning_rate": 0.2035828421791316,
      "loss": 0.2742,
      "num_input_tokens_seen": 11707792,
      "step": 15350
    },
    {
      "epoch": 31.923076923076923,
      "grad_norm": 0.0001235621457453817,
      "learning_rate": 0.20352781968970599,
      "loss": 0.2751,
      "num_input_tokens_seen": 11711536,
      "step": 15355
    },
    {
      "epoch": 31.933471933471935,
      "grad_norm": 0.00010859921167138964,
      "learning_rate": 0.2034727889456179,
      "loss": 0.2441,
      "num_input_tokens_seen": 11715280,
      "step": 15360
    },
    {
      "epoch": 31.943866943866944,
      "grad_norm": 0.0003521858307067305,
      "learning_rate": 0.2034177499553538,
      "loss": 0.227,
      "num_input_tokens_seen": 11719120,
      "step": 15365
    },
    {
      "epoch": 31.954261954261955,
      "grad_norm": 0.00040172928129322827,
      "learning_rate": 0.2033627027274014,
      "loss": 0.3021,
      "num_input_tokens_seen": 11722960,
      "step": 15370
    },
    {
      "epoch": 31.964656964656964,
      "grad_norm": 0.0002521440328564495,
      "learning_rate": 0.20330764727024955,
      "loss": 0.2668,
      "num_input_tokens_seen": 11726928,
      "step": 15375
    },
    {
      "epoch": 31.975051975051976,
      "grad_norm": 0.00024207847309298813,
      "learning_rate": 0.20325258359238868,
      "loss": 0.2853,
      "num_input_tokens_seen": 11730768,
      "step": 15380
    },
    {
      "epoch": 31.985446985446984,
      "grad_norm": 0.0007389273960143328,
      "learning_rate": 0.20319751170231018,
      "loss": 0.2853,
      "num_input_tokens_seen": 11734608,
      "step": 15385
    },
    {
      "epoch": 31.995841995841996,
      "grad_norm": 0.0003433914389461279,
      "learning_rate": 0.2031424316085068,
      "loss": 0.2678,
      "num_input_tokens_seen": 11738192,
      "step": 15390
    },
    {
      "epoch": 32.00623700623701,
      "grad_norm": 0.0002961848513223231,
      "learning_rate": 0.20308734331947265,
      "loss": 0.2892,
      "num_input_tokens_seen": 11742032,
      "step": 15395
    },
    {
      "epoch": 32.016632016632016,
      "grad_norm": 0.00013911539281252772,
      "learning_rate": 0.20303224684370305,
      "loss": 0.2735,
      "num_input_tokens_seen": 11745744,
      "step": 15400
    },
    {
      "epoch": 32.016632016632016,
      "eval_loss": 0.2534793019294739,
      "eval_runtime": 13.4433,
      "eval_samples_per_second": 63.675,
      "eval_steps_per_second": 15.919,
      "num_input_tokens_seen": 11745744,
      "step": 15400
    },
    {
      "epoch": 32.027027027027025,
      "grad_norm": 0.0009543925989419222,
      "learning_rate": 0.20297714218969456,
      "loss": 0.248,
      "num_input_tokens_seen": 11749392,
      "step": 15405
    },
    {
      "epoch": 32.03742203742204,
      "grad_norm": 0.0006277076900005341,
      "learning_rate": 0.20292202936594497,
      "loss": 0.2533,
      "num_input_tokens_seen": 11753168,
      "step": 15410
    },
    {
      "epoch": 32.04781704781705,
      "grad_norm": 0.00030200681067071855,
      "learning_rate": 0.2028669083809534,
      "loss": 0.2815,
      "num_input_tokens_seen": 11756912,
      "step": 15415
    },
    {
      "epoch": 32.05821205821206,
      "grad_norm": 0.00011790819553425536,
      "learning_rate": 0.20281177924322016,
      "loss": 0.2587,
      "num_input_tokens_seen": 11760688,
      "step": 15420
    },
    {
      "epoch": 32.06860706860707,
      "grad_norm": 0.00010677181126084179,
      "learning_rate": 0.2027566419612469,
      "loss": 0.2824,
      "num_input_tokens_seen": 11764656,
      "step": 15425
    },
    {
      "epoch": 32.07900207900208,
      "grad_norm": 0.0005335345631465316,
      "learning_rate": 0.20270149654353647,
      "loss": 0.2707,
      "num_input_tokens_seen": 11768592,
      "step": 15430
    },
    {
      "epoch": 32.08939708939709,
      "grad_norm": 0.0002658010052982718,
      "learning_rate": 0.202646342998593,
      "loss": 0.271,
      "num_input_tokens_seen": 11772400,
      "step": 15435
    },
    {
      "epoch": 32.0997920997921,
      "grad_norm": 0.0003651408478617668,
      "learning_rate": 0.20259118133492185,
      "loss": 0.2264,
      "num_input_tokens_seen": 11776336,
      "step": 15440
    },
    {
      "epoch": 32.11018711018711,
      "grad_norm": 0.00028817771817557514,
      "learning_rate": 0.20253601156102966,
      "loss": 0.2816,
      "num_input_tokens_seen": 11780080,
      "step": 15445
    },
    {
      "epoch": 32.12058212058212,
      "grad_norm": 0.00010571320308372378,
      "learning_rate": 0.20248083368542422,
      "loss": 0.281,
      "num_input_tokens_seen": 11783792,
      "step": 15450
    },
    {
      "epoch": 32.13097713097713,
      "grad_norm": 5.7641031162347645e-05,
      "learning_rate": 0.2024256477166147,
      "loss": 0.2666,
      "num_input_tokens_seen": 11787568,
      "step": 15455
    },
    {
      "epoch": 32.141372141372145,
      "grad_norm": 0.00021898458362556994,
      "learning_rate": 0.2023704536631115,
      "loss": 0.24,
      "num_input_tokens_seen": 11791440,
      "step": 15460
    },
    {
      "epoch": 32.15176715176715,
      "grad_norm": 0.0003960388421546668,
      "learning_rate": 0.20231525153342625,
      "loss": 0.3096,
      "num_input_tokens_seen": 11795216,
      "step": 15465
    },
    {
      "epoch": 32.16216216216216,
      "grad_norm": 5.404434341471642e-05,
      "learning_rate": 0.20226004133607173,
      "loss": 0.2396,
      "num_input_tokens_seen": 11799088,
      "step": 15470
    },
    {
      "epoch": 32.17255717255717,
      "grad_norm": 0.00022807875939179212,
      "learning_rate": 0.20220482307956214,
      "loss": 0.2699,
      "num_input_tokens_seen": 11802928,
      "step": 15475
    },
    {
      "epoch": 32.182952182952185,
      "grad_norm": 0.000115324801299721,
      "learning_rate": 0.20214959677241276,
      "loss": 0.262,
      "num_input_tokens_seen": 11806768,
      "step": 15480
    },
    {
      "epoch": 32.19334719334719,
      "grad_norm": 0.00021686071704607457,
      "learning_rate": 0.20209436242314022,
      "loss": 0.271,
      "num_input_tokens_seen": 11810640,
      "step": 15485
    },
    {
      "epoch": 32.2037422037422,
      "grad_norm": 4.4735646952176467e-05,
      "learning_rate": 0.2020391200402623,
      "loss": 0.2635,
      "num_input_tokens_seen": 11814352,
      "step": 15490
    },
    {
      "epoch": 32.21413721413722,
      "grad_norm": 0.00015896240074653178,
      "learning_rate": 0.2019838696322981,
      "loss": 0.2733,
      "num_input_tokens_seen": 11818160,
      "step": 15495
    },
    {
      "epoch": 32.224532224532226,
      "grad_norm": 0.00027150404639542103,
      "learning_rate": 0.20192861120776798,
      "loss": 0.2821,
      "num_input_tokens_seen": 11821904,
      "step": 15500
    },
    {
      "epoch": 32.234927234927234,
      "grad_norm": 0.0004229968471918255,
      "learning_rate": 0.20187334477519345,
      "loss": 0.26,
      "num_input_tokens_seen": 11825776,
      "step": 15505
    },
    {
      "epoch": 32.24532224532224,
      "grad_norm": 3.722673864103854e-05,
      "learning_rate": 0.20181807034309726,
      "loss": 0.2624,
      "num_input_tokens_seen": 11829584,
      "step": 15510
    },
    {
      "epoch": 32.25571725571726,
      "grad_norm": 0.0002588228671811521,
      "learning_rate": 0.2017627879200034,
      "loss": 0.2364,
      "num_input_tokens_seen": 11833360,
      "step": 15515
    },
    {
      "epoch": 32.266112266112266,
      "grad_norm": 0.00022604198602493852,
      "learning_rate": 0.2017074975144372,
      "loss": 0.2469,
      "num_input_tokens_seen": 11837264,
      "step": 15520
    },
    {
      "epoch": 32.276507276507274,
      "grad_norm": 0.0003075867425650358,
      "learning_rate": 0.20165219913492508,
      "loss": 0.2455,
      "num_input_tokens_seen": 11841168,
      "step": 15525
    },
    {
      "epoch": 32.28690228690229,
      "grad_norm": 0.00024371856125071645,
      "learning_rate": 0.20159689278999468,
      "loss": 0.2517,
      "num_input_tokens_seen": 11845168,
      "step": 15530
    },
    {
      "epoch": 32.2972972972973,
      "grad_norm": 0.0008227110374718904,
      "learning_rate": 0.20154157848817508,
      "loss": 0.2804,
      "num_input_tokens_seen": 11849104,
      "step": 15535
    },
    {
      "epoch": 32.30769230769231,
      "grad_norm": 4.614240970113315e-05,
      "learning_rate": 0.20148625623799632,
      "loss": 0.2647,
      "num_input_tokens_seen": 11852816,
      "step": 15540
    },
    {
      "epoch": 32.318087318087315,
      "grad_norm": 5.6565393606433645e-05,
      "learning_rate": 0.20143092604798984,
      "loss": 0.288,
      "num_input_tokens_seen": 11856816,
      "step": 15545
    },
    {
      "epoch": 32.32848232848233,
      "grad_norm": 0.0008120402344502509,
      "learning_rate": 0.2013755879266883,
      "loss": 0.2788,
      "num_input_tokens_seen": 11860496,
      "step": 15550
    },
    {
      "epoch": 32.33887733887734,
      "grad_norm": 0.00021874764934182167,
      "learning_rate": 0.20132024188262543,
      "loss": 0.243,
      "num_input_tokens_seen": 11864400,
      "step": 15555
    },
    {
      "epoch": 32.34927234927235,
      "grad_norm": 0.0005347630358301103,
      "learning_rate": 0.2012648879243363,
      "loss": 0.3496,
      "num_input_tokens_seen": 11868432,
      "step": 15560
    },
    {
      "epoch": 32.35966735966736,
      "grad_norm": 0.0001643247960601002,
      "learning_rate": 0.20120952606035725,
      "loss": 0.2784,
      "num_input_tokens_seen": 11872496,
      "step": 15565
    },
    {
      "epoch": 32.37006237006237,
      "grad_norm": 0.00032321701291948557,
      "learning_rate": 0.20115415629922576,
      "loss": 0.278,
      "num_input_tokens_seen": 11876240,
      "step": 15570
    },
    {
      "epoch": 32.38045738045738,
      "grad_norm": 0.00019577352213673294,
      "learning_rate": 0.20109877864948048,
      "loss": 0.2826,
      "num_input_tokens_seen": 11880144,
      "step": 15575
    },
    {
      "epoch": 32.39085239085239,
      "grad_norm": 0.00010374561679782346,
      "learning_rate": 0.20104339311966138,
      "loss": 0.2743,
      "num_input_tokens_seen": 11883920,
      "step": 15580
    },
    {
      "epoch": 32.4012474012474,
      "grad_norm": 3.9047976315487176e-05,
      "learning_rate": 0.2009879997183097,
      "loss": 0.2534,
      "num_input_tokens_seen": 11887664,
      "step": 15585
    },
    {
      "epoch": 32.41164241164241,
      "grad_norm": 0.00010887360258493572,
      "learning_rate": 0.20093259845396763,
      "loss": 0.2867,
      "num_input_tokens_seen": 11891312,
      "step": 15590
    },
    {
      "epoch": 32.42203742203742,
      "grad_norm": 0.00020036347268614918,
      "learning_rate": 0.20087718933517884,
      "loss": 0.2589,
      "num_input_tokens_seen": 11895024,
      "step": 15595
    },
    {
      "epoch": 32.432432432432435,
      "grad_norm": 0.00023753606365062296,
      "learning_rate": 0.20082177237048807,
      "loss": 0.2402,
      "num_input_tokens_seen": 11898672,
      "step": 15600
    },
    {
      "epoch": 32.432432432432435,
      "eval_loss": 0.24875333905220032,
      "eval_runtime": 13.4591,
      "eval_samples_per_second": 63.6,
      "eval_steps_per_second": 15.9,
      "num_input_tokens_seen": 11898672,
      "step": 15600
    },
    {
      "epoch": 32.44282744282744,
      "grad_norm": 0.0006049730582162738,
      "learning_rate": 0.20076634756844133,
      "loss": 0.273,
      "num_input_tokens_seen": 11902576,
      "step": 15605
    },
    {
      "epoch": 32.45322245322245,
      "grad_norm": 3.909133010893129e-05,
      "learning_rate": 0.20071091493758586,
      "loss": 0.2845,
      "num_input_tokens_seen": 11906352,
      "step": 15610
    },
    {
      "epoch": 32.46361746361746,
      "grad_norm": 0.00044083455577492714,
      "learning_rate": 0.20065547448647003,
      "loss": 0.2658,
      "num_input_tokens_seen": 11910224,
      "step": 15615
    },
    {
      "epoch": 32.474012474012476,
      "grad_norm": 0.0003770831972360611,
      "learning_rate": 0.20060002622364348,
      "loss": 0.2748,
      "num_input_tokens_seen": 11914032,
      "step": 15620
    },
    {
      "epoch": 32.484407484407484,
      "grad_norm": 9.661898366175592e-05,
      "learning_rate": 0.20054457015765695,
      "loss": 0.2807,
      "num_input_tokens_seen": 11917872,
      "step": 15625
    },
    {
      "epoch": 32.49480249480249,
      "grad_norm": 0.00025012672995217144,
      "learning_rate": 0.20048910629706254,
      "loss": 0.256,
      "num_input_tokens_seen": 11921744,
      "step": 15630
    },
    {
      "epoch": 32.50519750519751,
      "grad_norm": 0.00043226455454714596,
      "learning_rate": 0.20043363465041347,
      "loss": 0.2669,
      "num_input_tokens_seen": 11925520,
      "step": 15635
    },
    {
      "epoch": 32.515592515592516,
      "grad_norm": 0.0002737355825956911,
      "learning_rate": 0.2003781552262641,
      "loss": 0.2364,
      "num_input_tokens_seen": 11929328,
      "step": 15640
    },
    {
      "epoch": 32.525987525987524,
      "grad_norm": 5.1621307648019865e-05,
      "learning_rate": 0.20032266803317014,
      "loss": 0.2543,
      "num_input_tokens_seen": 11932944,
      "step": 15645
    },
    {
      "epoch": 32.53638253638254,
      "grad_norm": 0.0003747539594769478,
      "learning_rate": 0.2002671730796884,
      "loss": 0.2612,
      "num_input_tokens_seen": 11936784,
      "step": 15650
    },
    {
      "epoch": 32.54677754677755,
      "grad_norm": 0.0002851150056812912,
      "learning_rate": 0.20021167037437684,
      "loss": 0.2771,
      "num_input_tokens_seen": 11940688,
      "step": 15655
    },
    {
      "epoch": 32.55717255717256,
      "grad_norm": 0.00021103833569213748,
      "learning_rate": 0.20015615992579472,
      "loss": 0.256,
      "num_input_tokens_seen": 11944464,
      "step": 15660
    },
    {
      "epoch": 32.567567567567565,
      "grad_norm": 0.000189159472938627,
      "learning_rate": 0.20010064174250244,
      "loss": 0.2711,
      "num_input_tokens_seen": 11948272,
      "step": 15665
    },
    {
      "epoch": 32.57796257796258,
      "grad_norm": 0.00024098555149976164,
      "learning_rate": 0.2000451158330616,
      "loss": 0.2668,
      "num_input_tokens_seen": 11952048,
      "step": 15670
    },
    {
      "epoch": 32.58835758835759,
      "grad_norm": 0.0008048078743740916,
      "learning_rate": 0.199989582206035,
      "loss": 0.2857,
      "num_input_tokens_seen": 11955952,
      "step": 15675
    },
    {
      "epoch": 32.5987525987526,
      "grad_norm": 0.00024033478985074908,
      "learning_rate": 0.1999340408699866,
      "loss": 0.287,
      "num_input_tokens_seen": 11959792,
      "step": 15680
    },
    {
      "epoch": 32.60914760914761,
      "grad_norm": 0.00038019142812117934,
      "learning_rate": 0.19987849183348155,
      "loss": 0.3033,
      "num_input_tokens_seen": 11963536,
      "step": 15685
    },
    {
      "epoch": 32.61954261954262,
      "grad_norm": 9.447165939491242e-05,
      "learning_rate": 0.19982293510508628,
      "loss": 0.2773,
      "num_input_tokens_seen": 11967280,
      "step": 15690
    },
    {
      "epoch": 32.62993762993763,
      "grad_norm": 0.00020353558647911996,
      "learning_rate": 0.19976737069336833,
      "loss": 0.2607,
      "num_input_tokens_seen": 11971152,
      "step": 15695
    },
    {
      "epoch": 32.64033264033264,
      "grad_norm": 0.00034135454916395247,
      "learning_rate": 0.1997117986068964,
      "loss": 0.2845,
      "num_input_tokens_seen": 11974896,
      "step": 15700
    },
    {
      "epoch": 32.65072765072765,
      "grad_norm": 5.22316295246128e-05,
      "learning_rate": 0.19965621885424037,
      "loss": 0.2509,
      "num_input_tokens_seen": 11978800,
      "step": 15705
    },
    {
      "epoch": 32.66112266112266,
      "grad_norm": 0.00016469572437927127,
      "learning_rate": 0.19960063144397142,
      "loss": 0.2807,
      "num_input_tokens_seen": 11982608,
      "step": 15710
    },
    {
      "epoch": 32.67151767151767,
      "grad_norm": 0.00048100523417815566,
      "learning_rate": 0.19954503638466176,
      "loss": 0.2685,
      "num_input_tokens_seen": 11986416,
      "step": 15715
    },
    {
      "epoch": 32.681912681912685,
      "grad_norm": 0.00011516644735820591,
      "learning_rate": 0.1994894336848848,
      "loss": 0.2641,
      "num_input_tokens_seen": 11990448,
      "step": 15720
    },
    {
      "epoch": 32.69230769230769,
      "grad_norm": 9.864517778623849e-05,
      "learning_rate": 0.1994338233532153,
      "loss": 0.2601,
      "num_input_tokens_seen": 11994288,
      "step": 15725
    },
    {
      "epoch": 32.7027027027027,
      "grad_norm": 0.00024225196102634072,
      "learning_rate": 0.19937820539822904,
      "loss": 0.2788,
      "num_input_tokens_seen": 11998064,
      "step": 15730
    },
    {
      "epoch": 32.71309771309771,
      "grad_norm": 0.0001712695084279403,
      "learning_rate": 0.199322579828503,
      "loss": 0.2751,
      "num_input_tokens_seen": 12001936,
      "step": 15735
    },
    {
      "epoch": 32.723492723492726,
      "grad_norm": 9.223847882822156e-05,
      "learning_rate": 0.19926694665261527,
      "loss": 0.2713,
      "num_input_tokens_seen": 12005648,
      "step": 15740
    },
    {
      "epoch": 32.733887733887734,
      "grad_norm": 5.836924538016319e-05,
      "learning_rate": 0.19921130587914526,
      "loss": 0.2653,
      "num_input_tokens_seen": 12009328,
      "step": 15745
    },
    {
      "epoch": 32.74428274428274,
      "grad_norm": 7.56628141971305e-05,
      "learning_rate": 0.19915565751667344,
      "loss": 0.245,
      "num_input_tokens_seen": 12013104,
      "step": 15750
    },
    {
      "epoch": 32.75467775467776,
      "grad_norm": 0.0001046677862177603,
      "learning_rate": 0.19910000157378152,
      "loss": 0.2104,
      "num_input_tokens_seen": 12017168,
      "step": 15755
    },
    {
      "epoch": 32.765072765072766,
      "grad_norm": 0.0005254389834590256,
      "learning_rate": 0.1990443380590523,
      "loss": 0.302,
      "num_input_tokens_seen": 12021040,
      "step": 15760
    },
    {
      "epoch": 32.775467775467774,
      "grad_norm": 0.0003612090367823839,
      "learning_rate": 0.19898866698106984,
      "loss": 0.2889,
      "num_input_tokens_seen": 12024720,
      "step": 15765
    },
    {
      "epoch": 32.78586278586278,
      "grad_norm": 0.0006334806093946099,
      "learning_rate": 0.19893298834841933,
      "loss": 0.2735,
      "num_input_tokens_seen": 12028528,
      "step": 15770
    },
    {
      "epoch": 32.7962577962578,
      "grad_norm": 0.000236947278608568,
      "learning_rate": 0.19887730216968705,
      "loss": 0.2539,
      "num_input_tokens_seen": 12032336,
      "step": 15775
    },
    {
      "epoch": 32.80665280665281,
      "grad_norm": 0.00017690783715806901,
      "learning_rate": 0.19882160845346053,
      "loss": 0.2571,
      "num_input_tokens_seen": 12036016,
      "step": 15780
    },
    {
      "epoch": 32.817047817047815,
      "grad_norm": 5.5965025239856914e-05,
      "learning_rate": 0.1987659072083285,
      "loss": 0.2886,
      "num_input_tokens_seen": 12039664,
      "step": 15785
    },
    {
      "epoch": 32.82744282744283,
      "grad_norm": 0.0001633249339647591,
      "learning_rate": 0.1987101984428807,
      "loss": 0.2701,
      "num_input_tokens_seen": 12043376,
      "step": 15790
    },
    {
      "epoch": 32.83783783783784,
      "grad_norm": 0.0004956405027769506,
      "learning_rate": 0.19865448216570822,
      "loss": 0.2699,
      "num_input_tokens_seen": 12047184,
      "step": 15795
    },
    {
      "epoch": 32.84823284823285,
      "grad_norm": 0.00033626865479163826,
      "learning_rate": 0.19859875838540317,
      "loss": 0.2652,
      "num_input_tokens_seen": 12050992,
      "step": 15800
    },
    {
      "epoch": 32.84823284823285,
      "eval_loss": 0.25051096081733704,
      "eval_runtime": 13.4256,
      "eval_samples_per_second": 63.759,
      "eval_steps_per_second": 15.94,
      "num_input_tokens_seen": 12050992,
      "step": 15800
    },
    {
      "epoch": 32.858627858627855,
      "grad_norm": 0.0002648533845786005,
      "learning_rate": 0.1985430271105588,
      "loss": 0.2834,
      "num_input_tokens_seen": 12054768,
      "step": 15805
    },
    {
      "epoch": 32.86902286902287,
      "grad_norm": 0.0002831656893249601,
      "learning_rate": 0.19848728834976961,
      "loss": 0.271,
      "num_input_tokens_seen": 12058576,
      "step": 15810
    },
    {
      "epoch": 32.87941787941788,
      "grad_norm": 0.00017496984219178557,
      "learning_rate": 0.19843154211163128,
      "loss": 0.2614,
      "num_input_tokens_seen": 12062352,
      "step": 15815
    },
    {
      "epoch": 32.88981288981289,
      "grad_norm": 0.0007717504049651325,
      "learning_rate": 0.1983757884047405,
      "loss": 0.2706,
      "num_input_tokens_seen": 12066288,
      "step": 15820
    },
    {
      "epoch": 32.9002079002079,
      "grad_norm": 0.0007676717359572649,
      "learning_rate": 0.1983200272376952,
      "loss": 0.2932,
      "num_input_tokens_seen": 12070096,
      "step": 15825
    },
    {
      "epoch": 32.91060291060291,
      "grad_norm": 0.0004217039095237851,
      "learning_rate": 0.1982642586190945,
      "loss": 0.2688,
      "num_input_tokens_seen": 12073712,
      "step": 15830
    },
    {
      "epoch": 32.92099792099792,
      "grad_norm": 0.0005239245947450399,
      "learning_rate": 0.1982084825575386,
      "loss": 0.2668,
      "num_input_tokens_seen": 12077680,
      "step": 15835
    },
    {
      "epoch": 32.931392931392935,
      "grad_norm": 5.8531808463158086e-05,
      "learning_rate": 0.19815269906162883,
      "loss": 0.2697,
      "num_input_tokens_seen": 12081456,
      "step": 15840
    },
    {
      "epoch": 32.94178794178794,
      "grad_norm": 0.0001866299717221409,
      "learning_rate": 0.19809690813996775,
      "loss": 0.2928,
      "num_input_tokens_seen": 12085200,
      "step": 15845
    },
    {
      "epoch": 32.95218295218295,
      "grad_norm": 0.00029696914134547114,
      "learning_rate": 0.19804110980115905,
      "loss": 0.2737,
      "num_input_tokens_seen": 12088912,
      "step": 15850
    },
    {
      "epoch": 32.96257796257796,
      "grad_norm": 0.00037159433122724295,
      "learning_rate": 0.19798530405380746,
      "loss": 0.2744,
      "num_input_tokens_seen": 12092656,
      "step": 15855
    },
    {
      "epoch": 32.972972972972975,
      "grad_norm": 0.00012550615065265447,
      "learning_rate": 0.19792949090651893,
      "loss": 0.2785,
      "num_input_tokens_seen": 12096464,
      "step": 15860
    },
    {
      "epoch": 32.983367983367984,
      "grad_norm": 0.00012656039325520396,
      "learning_rate": 0.19787367036790066,
      "loss": 0.2527,
      "num_input_tokens_seen": 12100464,
      "step": 15865
    },
    {
      "epoch": 32.99376299376299,
      "grad_norm": 0.00039187405491247773,
      "learning_rate": 0.19781784244656075,
      "loss": 0.2645,
      "num_input_tokens_seen": 12104176,
      "step": 15870
    },
    {
      "epoch": 33.00415800415801,
      "grad_norm": 0.00022895254369359463,
      "learning_rate": 0.19776200715110864,
      "loss": 0.2509,
      "num_input_tokens_seen": 12108064,
      "step": 15875
    },
    {
      "epoch": 33.014553014553016,
      "grad_norm": 0.0001250521745532751,
      "learning_rate": 0.1977061644901548,
      "loss": 0.2638,
      "num_input_tokens_seen": 12111936,
      "step": 15880
    },
    {
      "epoch": 33.024948024948024,
      "grad_norm": 0.0001927202829392627,
      "learning_rate": 0.1976503144723109,
      "loss": 0.2559,
      "num_input_tokens_seen": 12115936,
      "step": 15885
    },
    {
      "epoch": 33.03534303534303,
      "grad_norm": 0.0002640201710164547,
      "learning_rate": 0.19759445710618967,
      "loss": 0.2782,
      "num_input_tokens_seen": 12119744,
      "step": 15890
    },
    {
      "epoch": 33.04573804573805,
      "grad_norm": 6.730109453201294e-05,
      "learning_rate": 0.19753859240040508,
      "loss": 0.2609,
      "num_input_tokens_seen": 12123456,
      "step": 15895
    },
    {
      "epoch": 33.056133056133056,
      "grad_norm": 0.0004990263259969652,
      "learning_rate": 0.1974827203635721,
      "loss": 0.2742,
      "num_input_tokens_seen": 12127328,
      "step": 15900
    },
    {
      "epoch": 33.066528066528065,
      "grad_norm": 0.00037125975359231234,
      "learning_rate": 0.19742684100430694,
      "loss": 0.2644,
      "num_input_tokens_seen": 12130944,
      "step": 15905
    },
    {
      "epoch": 33.07692307692308,
      "grad_norm": 0.0002494606014806777,
      "learning_rate": 0.19737095433122692,
      "loss": 0.2755,
      "num_input_tokens_seen": 12134880,
      "step": 15910
    },
    {
      "epoch": 33.08731808731809,
      "grad_norm": 0.00015916707343421876,
      "learning_rate": 0.19731506035295046,
      "loss": 0.2821,
      "num_input_tokens_seen": 12138752,
      "step": 15915
    },
    {
      "epoch": 33.0977130977131,
      "grad_norm": 0.00013659498654305935,
      "learning_rate": 0.19725915907809702,
      "loss": 0.2696,
      "num_input_tokens_seen": 12142784,
      "step": 15920
    },
    {
      "epoch": 33.108108108108105,
      "grad_norm": 0.0005090643535368145,
      "learning_rate": 0.1972032505152874,
      "loss": 0.2726,
      "num_input_tokens_seen": 12146592,
      "step": 15925
    },
    {
      "epoch": 33.11850311850312,
      "grad_norm": 0.00010413699055789039,
      "learning_rate": 0.19714733467314338,
      "loss": 0.2501,
      "num_input_tokens_seen": 12150496,
      "step": 15930
    },
    {
      "epoch": 33.12889812889813,
      "grad_norm": 0.00035530258901417255,
      "learning_rate": 0.19709141156028784,
      "loss": 0.2694,
      "num_input_tokens_seen": 12154304,
      "step": 15935
    },
    {
      "epoch": 33.13929313929314,
      "grad_norm": 9.427600161870942e-05,
      "learning_rate": 0.1970354811853448,
      "loss": 0.2683,
      "num_input_tokens_seen": 12158208,
      "step": 15940
    },
    {
      "epoch": 33.14968814968815,
      "grad_norm": 9.188500553136691e-05,
      "learning_rate": 0.19697954355693953,
      "loss": 0.2702,
      "num_input_tokens_seen": 12162048,
      "step": 15945
    },
    {
      "epoch": 33.16008316008316,
      "grad_norm": 0.00042223732452839613,
      "learning_rate": 0.19692359868369827,
      "loss": 0.2654,
      "num_input_tokens_seen": 12165824,
      "step": 15950
    },
    {
      "epoch": 33.17047817047817,
      "grad_norm": 0.0004130545712541789,
      "learning_rate": 0.1968676465742484,
      "loss": 0.2708,
      "num_input_tokens_seen": 12169696,
      "step": 15955
    },
    {
      "epoch": 33.18087318087318,
      "grad_norm": 3.997397652710788e-05,
      "learning_rate": 0.19681168723721845,
      "loss": 0.2678,
      "num_input_tokens_seen": 12173632,
      "step": 15960
    },
    {
      "epoch": 33.19126819126819,
      "grad_norm": 0.0002635691489558667,
      "learning_rate": 0.19675572068123803,
      "loss": 0.2668,
      "num_input_tokens_seen": 12177408,
      "step": 15965
    },
    {
      "epoch": 33.2016632016632,
      "grad_norm": 0.00045208234223537147,
      "learning_rate": 0.19669974691493794,
      "loss": 0.2528,
      "num_input_tokens_seen": 12181440,
      "step": 15970
    },
    {
      "epoch": 33.21205821205821,
      "grad_norm": 0.0003312194312456995,
      "learning_rate": 0.19664376594695002,
      "loss": 0.27,
      "num_input_tokens_seen": 12185312,
      "step": 15975
    },
    {
      "epoch": 33.222453222453225,
      "grad_norm": 0.0004204189754091203,
      "learning_rate": 0.19658777778590722,
      "loss": 0.2541,
      "num_input_tokens_seen": 12189120,
      "step": 15980
    },
    {
      "epoch": 33.232848232848234,
      "grad_norm": 0.0001763259497238323,
      "learning_rate": 0.19653178244044364,
      "loss": 0.273,
      "num_input_tokens_seen": 12192928,
      "step": 15985
    },
    {
      "epoch": 33.24324324324324,
      "grad_norm": 0.00014383697998709977,
      "learning_rate": 0.19647577991919443,
      "loss": 0.2636,
      "num_input_tokens_seen": 12196800,
      "step": 15990
    },
    {
      "epoch": 33.25363825363825,
      "grad_norm": 0.00033120953594334424,
      "learning_rate": 0.1964197702307959,
      "loss": 0.2732,
      "num_input_tokens_seen": 12200736,
      "step": 15995
    },
    {
      "epoch": 33.264033264033266,
      "grad_norm": 0.0004515945620369166,
      "learning_rate": 0.19636375338388545,
      "loss": 0.246,
      "num_input_tokens_seen": 12204352,
      "step": 16000
    },
    {
      "epoch": 33.264033264033266,
      "eval_loss": 0.24759477376937866,
      "eval_runtime": 13.413,
      "eval_samples_per_second": 63.819,
      "eval_steps_per_second": 15.955,
      "num_input_tokens_seen": 12204352,
      "step": 16000
    },
    {
      "epoch": 33.274428274428274,
      "grad_norm": 0.00024289079010486603,
      "learning_rate": 0.1963077293871016,
      "loss": 0.237,
      "num_input_tokens_seen": 12208096,
      "step": 16005
    },
    {
      "epoch": 33.28482328482328,
      "grad_norm": 0.00017072846821974963,
      "learning_rate": 0.19625169824908395,
      "loss": 0.3022,
      "num_input_tokens_seen": 12211872,
      "step": 16010
    },
    {
      "epoch": 33.2952182952183,
      "grad_norm": 0.0003880539152305573,
      "learning_rate": 0.19619565997847319,
      "loss": 0.2555,
      "num_input_tokens_seen": 12215552,
      "step": 16015
    },
    {
      "epoch": 33.305613305613306,
      "grad_norm": 0.0007254658848978579,
      "learning_rate": 0.19613961458391113,
      "loss": 0.2775,
      "num_input_tokens_seen": 12219328,
      "step": 16020
    },
    {
      "epoch": 33.316008316008315,
      "grad_norm": 0.00028293346986174583,
      "learning_rate": 0.19608356207404065,
      "loss": 0.2609,
      "num_input_tokens_seen": 12223136,
      "step": 16025
    },
    {
      "epoch": 33.32640332640332,
      "grad_norm": 0.0003068010846618563,
      "learning_rate": 0.1960275024575058,
      "loss": 0.2805,
      "num_input_tokens_seen": 12226976,
      "step": 16030
    },
    {
      "epoch": 33.33679833679834,
      "grad_norm": 0.00012259873619768769,
      "learning_rate": 0.19597143574295164,
      "loss": 0.2849,
      "num_input_tokens_seen": 12230688,
      "step": 16035
    },
    {
      "epoch": 33.34719334719335,
      "grad_norm": 0.00020578320254571736,
      "learning_rate": 0.1959153619390244,
      "loss": 0.2757,
      "num_input_tokens_seen": 12234368,
      "step": 16040
    },
    {
      "epoch": 33.357588357588355,
      "grad_norm": 0.0004884267109446228,
      "learning_rate": 0.1958592810543713,
      "loss": 0.2579,
      "num_input_tokens_seen": 12238112,
      "step": 16045
    },
    {
      "epoch": 33.36798336798337,
      "grad_norm": 0.00028216184000484645,
      "learning_rate": 0.19580319309764077,
      "loss": 0.2612,
      "num_input_tokens_seen": 12241952,
      "step": 16050
    },
    {
      "epoch": 33.37837837837838,
      "grad_norm": 0.0006488080834969878,
      "learning_rate": 0.1957470980774823,
      "loss": 0.3055,
      "num_input_tokens_seen": 12245696,
      "step": 16055
    },
    {
      "epoch": 33.38877338877339,
      "grad_norm": 0.000672056688927114,
      "learning_rate": 0.19569099600254639,
      "loss": 0.2836,
      "num_input_tokens_seen": 12249408,
      "step": 16060
    },
    {
      "epoch": 33.3991683991684,
      "grad_norm": 0.00028896002913825214,
      "learning_rate": 0.1956348868814847,
      "loss": 0.2745,
      "num_input_tokens_seen": 12253280,
      "step": 16065
    },
    {
      "epoch": 33.40956340956341,
      "grad_norm": 9.332145418738946e-05,
      "learning_rate": 0.19557877072295,
      "loss": 0.2799,
      "num_input_tokens_seen": 12257184,
      "step": 16070
    },
    {
      "epoch": 33.41995841995842,
      "grad_norm": 0.00045819731894880533,
      "learning_rate": 0.19552264753559603,
      "loss": 0.2628,
      "num_input_tokens_seen": 12260864,
      "step": 16075
    },
    {
      "epoch": 33.43035343035343,
      "grad_norm": 0.0001537467906018719,
      "learning_rate": 0.19546651732807774,
      "loss": 0.2539,
      "num_input_tokens_seen": 12264672,
      "step": 16080
    },
    {
      "epoch": 33.44074844074844,
      "grad_norm": 6.919514999026433e-05,
      "learning_rate": 0.19541038010905112,
      "loss": 0.2827,
      "num_input_tokens_seen": 12268512,
      "step": 16085
    },
    {
      "epoch": 33.45114345114345,
      "grad_norm": 0.0002537929394748062,
      "learning_rate": 0.19535423588717324,
      "loss": 0.2559,
      "num_input_tokens_seen": 12272288,
      "step": 16090
    },
    {
      "epoch": 33.46153846153846,
      "grad_norm": 0.00010120011575054377,
      "learning_rate": 0.19529808467110224,
      "loss": 0.2592,
      "num_input_tokens_seen": 12276128,
      "step": 16095
    },
    {
      "epoch": 33.471933471933475,
      "grad_norm": 0.0002596256381366402,
      "learning_rate": 0.19524192646949734,
      "loss": 0.2395,
      "num_input_tokens_seen": 12279872,
      "step": 16100
    },
    {
      "epoch": 33.482328482328484,
      "grad_norm": 0.0003985886287409812,
      "learning_rate": 0.19518576129101878,
      "loss": 0.2812,
      "num_input_tokens_seen": 12283744,
      "step": 16105
    },
    {
      "epoch": 33.49272349272349,
      "grad_norm": 0.0002144862082786858,
      "learning_rate": 0.19512958914432804,
      "loss": 0.2531,
      "num_input_tokens_seen": 12287520,
      "step": 16110
    },
    {
      "epoch": 33.5031185031185,
      "grad_norm": 6.764783756807446e-05,
      "learning_rate": 0.1950734100380875,
      "loss": 0.251,
      "num_input_tokens_seen": 12291168,
      "step": 16115
    },
    {
      "epoch": 33.513513513513516,
      "grad_norm": 0.0002308595139766112,
      "learning_rate": 0.19501722398096066,
      "loss": 0.2647,
      "num_input_tokens_seen": 12295104,
      "step": 16120
    },
    {
      "epoch": 33.523908523908524,
      "grad_norm": 0.00019253579375799745,
      "learning_rate": 0.1949610309816122,
      "loss": 0.2747,
      "num_input_tokens_seen": 12299008,
      "step": 16125
    },
    {
      "epoch": 33.53430353430353,
      "grad_norm": 0.00011218348663533106,
      "learning_rate": 0.1949048310487078,
      "loss": 0.2533,
      "num_input_tokens_seen": 12302880,
      "step": 16130
    },
    {
      "epoch": 33.54469854469855,
      "grad_norm": 0.00014726178778801113,
      "learning_rate": 0.19484862419091406,
      "loss": 0.2355,
      "num_input_tokens_seen": 12306720,
      "step": 16135
    },
    {
      "epoch": 33.555093555093556,
      "grad_norm": 7.352221291512251e-05,
      "learning_rate": 0.19479241041689893,
      "loss": 0.2741,
      "num_input_tokens_seen": 12310560,
      "step": 16140
    },
    {
      "epoch": 33.565488565488565,
      "grad_norm": 0.00012450353824533522,
      "learning_rate": 0.19473618973533116,
      "loss": 0.2783,
      "num_input_tokens_seen": 12314400,
      "step": 16145
    },
    {
      "epoch": 33.57588357588357,
      "grad_norm": 0.00011042161349905655,
      "learning_rate": 0.19467996215488076,
      "loss": 0.2704,
      "num_input_tokens_seen": 12318240,
      "step": 16150
    },
    {
      "epoch": 33.58627858627859,
      "grad_norm": 0.00021205050870776176,
      "learning_rate": 0.1946237276842187,
      "loss": 0.2718,
      "num_input_tokens_seen": 12322016,
      "step": 16155
    },
    {
      "epoch": 33.5966735966736,
      "grad_norm": 0.00013148492143955082,
      "learning_rate": 0.19456748633201712,
      "loss": 0.2787,
      "num_input_tokens_seen": 12325824,
      "step": 16160
    },
    {
      "epoch": 33.607068607068605,
      "grad_norm": 0.0007922688964754343,
      "learning_rate": 0.194511238106949,
      "loss": 0.2618,
      "num_input_tokens_seen": 12329536,
      "step": 16165
    },
    {
      "epoch": 33.61746361746362,
      "grad_norm": 4.1973460611188784e-05,
      "learning_rate": 0.19445498301768863,
      "loss": 0.2629,
      "num_input_tokens_seen": 12333504,
      "step": 16170
    },
    {
      "epoch": 33.62785862785863,
      "grad_norm": 0.00040032173274084926,
      "learning_rate": 0.19439872107291126,
      "loss": 0.2779,
      "num_input_tokens_seen": 12337280,
      "step": 16175
    },
    {
      "epoch": 33.63825363825364,
      "grad_norm": 0.00044166078441776335,
      "learning_rate": 0.1943424522812931,
      "loss": 0.2551,
      "num_input_tokens_seen": 12341248,
      "step": 16180
    },
    {
      "epoch": 33.648648648648646,
      "grad_norm": 0.0006496984860859811,
      "learning_rate": 0.19428617665151157,
      "loss": 0.2751,
      "num_input_tokens_seen": 12345088,
      "step": 16185
    },
    {
      "epoch": 33.65904365904366,
      "grad_norm": 0.0002717013703659177,
      "learning_rate": 0.19422989419224507,
      "loss": 0.2737,
      "num_input_tokens_seen": 12348832,
      "step": 16190
    },
    {
      "epoch": 33.66943866943867,
      "grad_norm": 0.0003359794500283897,
      "learning_rate": 0.19417360491217303,
      "loss": 0.2846,
      "num_input_tokens_seen": 12352512,
      "step": 16195
    },
    {
      "epoch": 33.67983367983368,
      "grad_norm": 0.00025236851070076227,
      "learning_rate": 0.19411730881997605,
      "loss": 0.2549,
      "num_input_tokens_seen": 12356224,
      "step": 16200
    },
    {
      "epoch": 33.67983367983368,
      "eval_loss": 0.24902480840682983,
      "eval_runtime": 13.407,
      "eval_samples_per_second": 63.847,
      "eval_steps_per_second": 15.962,
      "num_input_tokens_seen": 12356224,
      "step": 16200
    },
    {
      "epoch": 33.69022869022869,
      "grad_norm": 0.00020736752776429057,
      "learning_rate": 0.1940610059243356,
      "loss": 0.2754,
      "num_input_tokens_seen": 12360064,
      "step": 16205
    },
    {
      "epoch": 33.7006237006237,
      "grad_norm": 7.55802757339552e-05,
      "learning_rate": 0.19400469623393435,
      "loss": 0.2654,
      "num_input_tokens_seen": 12363552,
      "step": 16210
    },
    {
      "epoch": 33.71101871101871,
      "grad_norm": 0.00015491966041736305,
      "learning_rate": 0.1939483797574559,
      "loss": 0.2691,
      "num_input_tokens_seen": 12367424,
      "step": 16215
    },
    {
      "epoch": 33.72141372141372,
      "grad_norm": 0.0004810387035831809,
      "learning_rate": 0.19389205650358504,
      "loss": 0.2451,
      "num_input_tokens_seen": 12371008,
      "step": 16220
    },
    {
      "epoch": 33.731808731808734,
      "grad_norm": 0.0003228721907362342,
      "learning_rate": 0.19383572648100747,
      "loss": 0.2646,
      "num_input_tokens_seen": 12374816,
      "step": 16225
    },
    {
      "epoch": 33.74220374220374,
      "grad_norm": 0.00026009109569713473,
      "learning_rate": 0.19377938969841,
      "loss": 0.2606,
      "num_input_tokens_seen": 12378560,
      "step": 16230
    },
    {
      "epoch": 33.75259875259875,
      "grad_norm": 0.0005737391184084117,
      "learning_rate": 0.1937230461644805,
      "loss": 0.2765,
      "num_input_tokens_seen": 12382304,
      "step": 16235
    },
    {
      "epoch": 33.762993762993766,
      "grad_norm": 0.0009987344965338707,
      "learning_rate": 0.19366669588790777,
      "loss": 0.2743,
      "num_input_tokens_seen": 12386176,
      "step": 16240
    },
    {
      "epoch": 33.773388773388774,
      "grad_norm": 0.0007878433098085225,
      "learning_rate": 0.19361033887738185,
      "loss": 0.2658,
      "num_input_tokens_seen": 12389984,
      "step": 16245
    },
    {
      "epoch": 33.78378378378378,
      "grad_norm": 0.00023748567036818713,
      "learning_rate": 0.19355397514159361,
      "loss": 0.2733,
      "num_input_tokens_seen": 12393728,
      "step": 16250
    },
    {
      "epoch": 33.79417879417879,
      "grad_norm": 0.00047703864402137697,
      "learning_rate": 0.19349760468923508,
      "loss": 0.2902,
      "num_input_tokens_seen": 12397504,
      "step": 16255
    },
    {
      "epoch": 33.804573804573806,
      "grad_norm": 0.00029075733618810773,
      "learning_rate": 0.19344122752899925,
      "loss": 0.2684,
      "num_input_tokens_seen": 12401408,
      "step": 16260
    },
    {
      "epoch": 33.814968814968815,
      "grad_norm": 0.00013549902359955013,
      "learning_rate": 0.1933848436695802,
      "loss": 0.2684,
      "num_input_tokens_seen": 12405152,
      "step": 16265
    },
    {
      "epoch": 33.82536382536382,
      "grad_norm": 0.0005686871591024101,
      "learning_rate": 0.1933284531196731,
      "loss": 0.2695,
      "num_input_tokens_seen": 12409088,
      "step": 16270
    },
    {
      "epoch": 33.83575883575884,
      "grad_norm": 0.0006737210787832737,
      "learning_rate": 0.19327205588797403,
      "loss": 0.2731,
      "num_input_tokens_seen": 12412896,
      "step": 16275
    },
    {
      "epoch": 33.84615384615385,
      "grad_norm": 7.488125993404537e-05,
      "learning_rate": 0.19321565198318014,
      "loss": 0.2744,
      "num_input_tokens_seen": 12416704,
      "step": 16280
    },
    {
      "epoch": 33.856548856548855,
      "grad_norm": 0.0001727318303892389,
      "learning_rate": 0.1931592414139896,
      "loss": 0.2702,
      "num_input_tokens_seen": 12420608,
      "step": 16285
    },
    {
      "epoch": 33.86694386694387,
      "grad_norm": 0.00054693385027349,
      "learning_rate": 0.19310282418910169,
      "loss": 0.2769,
      "num_input_tokens_seen": 12424416,
      "step": 16290
    },
    {
      "epoch": 33.87733887733888,
      "grad_norm": 0.00027967069763690233,
      "learning_rate": 0.1930464003172166,
      "loss": 0.2691,
      "num_input_tokens_seen": 12428288,
      "step": 16295
    },
    {
      "epoch": 33.88773388773389,
      "grad_norm": 0.0005294065340422094,
      "learning_rate": 0.19298996980703567,
      "loss": 0.2305,
      "num_input_tokens_seen": 12432128,
      "step": 16300
    },
    {
      "epoch": 33.898128898128896,
      "grad_norm": 9.152175334747881e-05,
      "learning_rate": 0.19293353266726113,
      "loss": 0.2522,
      "num_input_tokens_seen": 12435872,
      "step": 16305
    },
    {
      "epoch": 33.90852390852391,
      "grad_norm": 0.0005343279335647821,
      "learning_rate": 0.19287708890659633,
      "loss": 0.3056,
      "num_input_tokens_seen": 12439744,
      "step": 16310
    },
    {
      "epoch": 33.91891891891892,
      "grad_norm": 0.0006683728424832225,
      "learning_rate": 0.19282063853374556,
      "loss": 0.2509,
      "num_input_tokens_seen": 12443584,
      "step": 16315
    },
    {
      "epoch": 33.92931392931393,
      "grad_norm": 0.0003453294048085809,
      "learning_rate": 0.19276418155741423,
      "loss": 0.2641,
      "num_input_tokens_seen": 12447456,
      "step": 16320
    },
    {
      "epoch": 33.93970893970894,
      "grad_norm": 0.0001162965054390952,
      "learning_rate": 0.19270771798630867,
      "loss": 0.272,
      "num_input_tokens_seen": 12451168,
      "step": 16325
    },
    {
      "epoch": 33.95010395010395,
      "grad_norm": 0.00010587320866761729,
      "learning_rate": 0.1926512478291363,
      "loss": 0.2643,
      "num_input_tokens_seen": 12454976,
      "step": 16330
    },
    {
      "epoch": 33.96049896049896,
      "grad_norm": 0.00019101159705314785,
      "learning_rate": 0.19259477109460557,
      "loss": 0.2562,
      "num_input_tokens_seen": 12458752,
      "step": 16335
    },
    {
      "epoch": 33.97089397089397,
      "grad_norm": 0.00017973108333535492,
      "learning_rate": 0.19253828779142584,
      "loss": 0.2439,
      "num_input_tokens_seen": 12462432,
      "step": 16340
    },
    {
      "epoch": 33.981288981288984,
      "grad_norm": 0.00026924567646346986,
      "learning_rate": 0.19248179792830755,
      "loss": 0.2518,
      "num_input_tokens_seen": 12466176,
      "step": 16345
    },
    {
      "epoch": 33.99168399168399,
      "grad_norm": 0.0005582004669122398,
      "learning_rate": 0.19242530151396217,
      "loss": 0.3083,
      "num_input_tokens_seen": 12469984,
      "step": 16350
    },
    {
      "epoch": 34.002079002079,
      "grad_norm": 0.0005495326477102935,
      "learning_rate": 0.19236879855710215,
      "loss": 0.2918,
      "num_input_tokens_seen": 12473720,
      "step": 16355
    },
    {
      "epoch": 34.012474012474016,
      "grad_norm": 0.00022900123440194875,
      "learning_rate": 0.19231228906644096,
      "loss": 0.2733,
      "num_input_tokens_seen": 12477592,
      "step": 16360
    },
    {
      "epoch": 34.022869022869024,
      "grad_norm": 0.0001909696584334597,
      "learning_rate": 0.19225577305069302,
      "loss": 0.2685,
      "num_input_tokens_seen": 12481496,
      "step": 16365
    },
    {
      "epoch": 34.03326403326403,
      "grad_norm": 0.00014395166363101453,
      "learning_rate": 0.1921992505185739,
      "loss": 0.2554,
      "num_input_tokens_seen": 12485304,
      "step": 16370
    },
    {
      "epoch": 34.04365904365904,
      "grad_norm": 0.000253660895396024,
      "learning_rate": 0.19214272147880004,
      "loss": 0.291,
      "num_input_tokens_seen": 12489144,
      "step": 16375
    },
    {
      "epoch": 34.054054054054056,
      "grad_norm": 0.0006493324763141572,
      "learning_rate": 0.19208618594008892,
      "loss": 0.2526,
      "num_input_tokens_seen": 12492824,
      "step": 16380
    },
    {
      "epoch": 34.064449064449065,
      "grad_norm": 0.00018270219152327627,
      "learning_rate": 0.19202964391115904,
      "loss": 0.2479,
      "num_input_tokens_seen": 12496696,
      "step": 16385
    },
    {
      "epoch": 34.07484407484407,
      "grad_norm": 7.166177238104865e-05,
      "learning_rate": 0.1919730954007299,
      "loss": 0.2932,
      "num_input_tokens_seen": 12500440,
      "step": 16390
    },
    {
      "epoch": 34.08523908523909,
      "grad_norm": 0.0004000802000518888,
      "learning_rate": 0.19191654041752199,
      "loss": 0.2493,
      "num_input_tokens_seen": 12504216,
      "step": 16395
    },
    {
      "epoch": 34.0956340956341,
      "grad_norm": 0.0001537659263703972,
      "learning_rate": 0.19185997897025678,
      "loss": 0.2714,
      "num_input_tokens_seen": 12507960,
      "step": 16400
    },
    {
      "epoch": 34.0956340956341,
      "eval_loss": 0.2524295747280121,
      "eval_runtime": 13.406,
      "eval_samples_per_second": 63.852,
      "eval_steps_per_second": 15.963,
      "num_input_tokens_seen": 12507960,
      "step": 16400
    },
    {
      "epoch": 34.106029106029105,
      "grad_norm": 0.00023281040193978697,
      "learning_rate": 0.19180341106765672,
      "loss": 0.2728,
      "num_input_tokens_seen": 12511768,
      "step": 16405
    },
    {
      "epoch": 34.11642411642411,
      "grad_norm": 0.00026537213125266135,
      "learning_rate": 0.19174683671844536,
      "loss": 0.2591,
      "num_input_tokens_seen": 12515608,
      "step": 16410
    },
    {
      "epoch": 34.12681912681913,
      "grad_norm": 0.00015339165111072361,
      "learning_rate": 0.19169025593134717,
      "loss": 0.2317,
      "num_input_tokens_seen": 12519416,
      "step": 16415
    },
    {
      "epoch": 34.13721413721414,
      "grad_norm": 0.0006768080638721585,
      "learning_rate": 0.19163366871508764,
      "loss": 0.268,
      "num_input_tokens_seen": 12523256,
      "step": 16420
    },
    {
      "epoch": 34.147609147609145,
      "grad_norm": 0.0002964687591884285,
      "learning_rate": 0.19157707507839317,
      "loss": 0.2544,
      "num_input_tokens_seen": 12527064,
      "step": 16425
    },
    {
      "epoch": 34.15800415800416,
      "grad_norm": 0.0001422977657057345,
      "learning_rate": 0.19152047502999123,
      "loss": 0.2576,
      "num_input_tokens_seen": 12530808,
      "step": 16430
    },
    {
      "epoch": 34.16839916839917,
      "grad_norm": 0.0005182041204534471,
      "learning_rate": 0.19146386857861025,
      "loss": 0.2613,
      "num_input_tokens_seen": 12534648,
      "step": 16435
    },
    {
      "epoch": 34.17879417879418,
      "grad_norm": 0.000290843250695616,
      "learning_rate": 0.19140725573297968,
      "loss": 0.262,
      "num_input_tokens_seen": 12538680,
      "step": 16440
    },
    {
      "epoch": 34.189189189189186,
      "grad_norm": 0.00018275891488883644,
      "learning_rate": 0.19135063650182987,
      "loss": 0.2645,
      "num_input_tokens_seen": 12542712,
      "step": 16445
    },
    {
      "epoch": 34.1995841995842,
      "grad_norm": 6.097853838582523e-05,
      "learning_rate": 0.19129401089389234,
      "loss": 0.2881,
      "num_input_tokens_seen": 12546328,
      "step": 16450
    },
    {
      "epoch": 34.20997920997921,
      "grad_norm": 6.842829316155985e-05,
      "learning_rate": 0.19123737891789938,
      "loss": 0.259,
      "num_input_tokens_seen": 12550232,
      "step": 16455
    },
    {
      "epoch": 34.22037422037422,
      "grad_norm": 6.534743442898616e-05,
      "learning_rate": 0.19118074058258439,
      "loss": 0.2771,
      "num_input_tokens_seen": 12554232,
      "step": 16460
    },
    {
      "epoch": 34.23076923076923,
      "grad_norm": 0.00034814924583770335,
      "learning_rate": 0.1911240958966816,
      "loss": 0.2845,
      "num_input_tokens_seen": 12557944,
      "step": 16465
    },
    {
      "epoch": 34.24116424116424,
      "grad_norm": 0.0008581766160205007,
      "learning_rate": 0.19106744486892652,
      "loss": 0.281,
      "num_input_tokens_seen": 12561816,
      "step": 16470
    },
    {
      "epoch": 34.25155925155925,
      "grad_norm": 0.00030134874396026134,
      "learning_rate": 0.1910107875080553,
      "loss": 0.2623,
      "num_input_tokens_seen": 12565624,
      "step": 16475
    },
    {
      "epoch": 34.26195426195426,
      "grad_norm": 4.233239087625407e-05,
      "learning_rate": 0.19095412382280533,
      "loss": 0.2928,
      "num_input_tokens_seen": 12569496,
      "step": 16480
    },
    {
      "epoch": 34.272349272349274,
      "grad_norm": 0.0003774478973355144,
      "learning_rate": 0.19089745382191473,
      "loss": 0.2662,
      "num_input_tokens_seen": 12573240,
      "step": 16485
    },
    {
      "epoch": 34.28274428274428,
      "grad_norm": 0.00011659821029752493,
      "learning_rate": 0.19084077751412284,
      "loss": 0.2777,
      "num_input_tokens_seen": 12577016,
      "step": 16490
    },
    {
      "epoch": 34.29313929313929,
      "grad_norm": 0.0007570674642920494,
      "learning_rate": 0.19078409490816986,
      "loss": 0.2725,
      "num_input_tokens_seen": 12580856,
      "step": 16495
    },
    {
      "epoch": 34.303534303534306,
      "grad_norm": 0.00016461701306980103,
      "learning_rate": 0.19072740601279686,
      "loss": 0.2654,
      "num_input_tokens_seen": 12584568,
      "step": 16500
    },
    {
      "epoch": 34.313929313929314,
      "grad_norm": 0.00023914442863315344,
      "learning_rate": 0.19067071083674605,
      "loss": 0.253,
      "num_input_tokens_seen": 12588440,
      "step": 16505
    },
    {
      "epoch": 34.32432432432432,
      "grad_norm": 0.0001310602092416957,
      "learning_rate": 0.19061400938876052,
      "loss": 0.2662,
      "num_input_tokens_seen": 12592440,
      "step": 16510
    },
    {
      "epoch": 34.33471933471934,
      "grad_norm": 0.00012074862752342597,
      "learning_rate": 0.1905573016775844,
      "loss": 0.2617,
      "num_input_tokens_seen": 12596248,
      "step": 16515
    },
    {
      "epoch": 34.34511434511435,
      "grad_norm": 0.00018509275105316192,
      "learning_rate": 0.19050058771196263,
      "loss": 0.2731,
      "num_input_tokens_seen": 12600120,
      "step": 16520
    },
    {
      "epoch": 34.355509355509355,
      "grad_norm": 0.00012660337961278856,
      "learning_rate": 0.19044386750064132,
      "loss": 0.2721,
      "num_input_tokens_seen": 12603992,
      "step": 16525
    },
    {
      "epoch": 34.36590436590436,
      "grad_norm": 0.00027666555251926184,
      "learning_rate": 0.19038714105236737,
      "loss": 0.2645,
      "num_input_tokens_seen": 12607704,
      "step": 16530
    },
    {
      "epoch": 34.37629937629938,
      "grad_norm": 0.0004880151245743036,
      "learning_rate": 0.19033040837588874,
      "loss": 0.2834,
      "num_input_tokens_seen": 12611576,
      "step": 16535
    },
    {
      "epoch": 34.38669438669439,
      "grad_norm": 0.00014753633877262473,
      "learning_rate": 0.1902736694799543,
      "loss": 0.2735,
      "num_input_tokens_seen": 12615416,
      "step": 16540
    },
    {
      "epoch": 34.397089397089395,
      "grad_norm": 0.00010072724398924038,
      "learning_rate": 0.19021692437331392,
      "loss": 0.2784,
      "num_input_tokens_seen": 12619352,
      "step": 16545
    },
    {
      "epoch": 34.40748440748441,
      "grad_norm": 0.00035135503276251256,
      "learning_rate": 0.1901601730647184,
      "loss": 0.2753,
      "num_input_tokens_seen": 12623128,
      "step": 16550
    },
    {
      "epoch": 34.41787941787942,
      "grad_norm": 3.114005448878743e-05,
      "learning_rate": 0.19010341556291954,
      "loss": 0.2651,
      "num_input_tokens_seen": 12626904,
      "step": 16555
    },
    {
      "epoch": 34.42827442827443,
      "grad_norm": 0.00038274715188890696,
      "learning_rate": 0.19004665187667,
      "loss": 0.2641,
      "num_input_tokens_seen": 12630616,
      "step": 16560
    },
    {
      "epoch": 34.438669438669436,
      "grad_norm": 0.0006404719897545874,
      "learning_rate": 0.1899898820147235,
      "loss": 0.2965,
      "num_input_tokens_seen": 12634424,
      "step": 16565
    },
    {
      "epoch": 34.44906444906445,
      "grad_norm": 0.0005553974770009518,
      "learning_rate": 0.18993310598583465,
      "loss": 0.2688,
      "num_input_tokens_seen": 12638168,
      "step": 16570
    },
    {
      "epoch": 34.45945945945946,
      "grad_norm": 0.0005463711568154395,
      "learning_rate": 0.18987632379875904,
      "loss": 0.2703,
      "num_input_tokens_seen": 12641848,
      "step": 16575
    },
    {
      "epoch": 34.46985446985447,
      "grad_norm": 3.965325959143229e-05,
      "learning_rate": 0.18981953546225314,
      "loss": 0.2487,
      "num_input_tokens_seen": 12645528,
      "step": 16580
    },
    {
      "epoch": 34.48024948024948,
      "grad_norm": 0.00010026699601439759,
      "learning_rate": 0.18976274098507445,
      "loss": 0.2767,
      "num_input_tokens_seen": 12649464,
      "step": 16585
    },
    {
      "epoch": 34.49064449064449,
      "grad_norm": 0.00048031070036813617,
      "learning_rate": 0.18970594037598146,
      "loss": 0.2474,
      "num_input_tokens_seen": 12653208,
      "step": 16590
    },
    {
      "epoch": 34.5010395010395,
      "grad_norm": 2.514071820769459e-05,
      "learning_rate": 0.1896491336437335,
      "loss": 0.2733,
      "num_input_tokens_seen": 12656984,
      "step": 16595
    },
    {
      "epoch": 34.51143451143451,
      "grad_norm": 0.00013399054296314716,
      "learning_rate": 0.18959232079709085,
      "loss": 0.2736,
      "num_input_tokens_seen": 12660760,
      "step": 16600
    },
    {
      "epoch": 34.51143451143451,
      "eval_loss": 0.2697230279445648,
      "eval_runtime": 13.4038,
      "eval_samples_per_second": 63.863,
      "eval_steps_per_second": 15.966,
      "num_input_tokens_seen": 12660760,
      "step": 16600
    },
    {
      "epoch": 34.521829521829524,
      "grad_norm": 6.736852810718119e-05,
      "learning_rate": 0.18953550184481477,
      "loss": 0.2837,
      "num_input_tokens_seen": 12664696,
      "step": 16605
    },
    {
      "epoch": 34.53222453222453,
      "grad_norm": 0.00013937293260823935,
      "learning_rate": 0.18947867679566752,
      "loss": 0.2821,
      "num_input_tokens_seen": 12668440,
      "step": 16610
    },
    {
      "epoch": 34.54261954261954,
      "grad_norm": 0.00017785966338124126,
      "learning_rate": 0.18942184565841216,
      "loss": 0.2664,
      "num_input_tokens_seen": 12672312,
      "step": 16615
    },
    {
      "epoch": 34.553014553014556,
      "grad_norm": 5.3962445235811174e-05,
      "learning_rate": 0.18936500844181278,
      "loss": 0.2685,
      "num_input_tokens_seen": 12676248,
      "step": 16620
    },
    {
      "epoch": 34.563409563409564,
      "grad_norm": 0.0004858898464590311,
      "learning_rate": 0.18930816515463436,
      "loss": 0.2561,
      "num_input_tokens_seen": 12680248,
      "step": 16625
    },
    {
      "epoch": 34.57380457380457,
      "grad_norm": 0.0002345779212191701,
      "learning_rate": 0.18925131580564297,
      "loss": 0.2598,
      "num_input_tokens_seen": 12684088,
      "step": 16630
    },
    {
      "epoch": 34.58419958419958,
      "grad_norm": 0.0001271579967578873,
      "learning_rate": 0.1891944604036054,
      "loss": 0.2515,
      "num_input_tokens_seen": 12687832,
      "step": 16635
    },
    {
      "epoch": 34.5945945945946,
      "grad_norm": 4.2805033444892615e-05,
      "learning_rate": 0.1891375989572895,
      "loss": 0.2558,
      "num_input_tokens_seen": 12691864,
      "step": 16640
    },
    {
      "epoch": 34.604989604989605,
      "grad_norm": 0.00033412614720873535,
      "learning_rate": 0.18908073147546398,
      "loss": 0.2776,
      "num_input_tokens_seen": 12695704,
      "step": 16645
    },
    {
      "epoch": 34.61538461538461,
      "grad_norm": 0.00019627295841928571,
      "learning_rate": 0.18902385796689858,
      "loss": 0.2755,
      "num_input_tokens_seen": 12699576,
      "step": 16650
    },
    {
      "epoch": 34.62577962577963,
      "grad_norm": 0.00036488103796727955,
      "learning_rate": 0.18896697844036384,
      "loss": 0.2654,
      "num_input_tokens_seen": 12703480,
      "step": 16655
    },
    {
      "epoch": 34.63617463617464,
      "grad_norm": 0.00023361347848549485,
      "learning_rate": 0.18891009290463137,
      "loss": 0.2622,
      "num_input_tokens_seen": 12707256,
      "step": 16660
    },
    {
      "epoch": 34.646569646569645,
      "grad_norm": 0.0003277365758549422,
      "learning_rate": 0.18885320136847353,
      "loss": 0.2731,
      "num_input_tokens_seen": 12711160,
      "step": 16665
    },
    {
      "epoch": 34.656964656964654,
      "grad_norm": 0.0001076085027307272,
      "learning_rate": 0.1887963038406639,
      "loss": 0.2585,
      "num_input_tokens_seen": 12714936,
      "step": 16670
    },
    {
      "epoch": 34.66735966735967,
      "grad_norm": 0.00013675668742507696,
      "learning_rate": 0.18873940032997658,
      "loss": 0.2735,
      "num_input_tokens_seen": 12718808,
      "step": 16675
    },
    {
      "epoch": 34.67775467775468,
      "grad_norm": 0.00033890947815962136,
      "learning_rate": 0.18868249084518693,
      "loss": 0.2647,
      "num_input_tokens_seen": 12722584,
      "step": 16680
    },
    {
      "epoch": 34.688149688149686,
      "grad_norm": 0.0001570408494444564,
      "learning_rate": 0.18862557539507102,
      "loss": 0.2535,
      "num_input_tokens_seen": 12726424,
      "step": 16685
    },
    {
      "epoch": 34.6985446985447,
      "grad_norm": 0.00013297348050400615,
      "learning_rate": 0.18856865398840605,
      "loss": 0.2454,
      "num_input_tokens_seen": 12730168,
      "step": 16690
    },
    {
      "epoch": 34.70893970893971,
      "grad_norm": 5.4561322031077e-05,
      "learning_rate": 0.18851172663396995,
      "loss": 0.2616,
      "num_input_tokens_seen": 12733848,
      "step": 16695
    },
    {
      "epoch": 34.71933471933472,
      "grad_norm": 0.0005349882412701845,
      "learning_rate": 0.1884547933405416,
      "loss": 0.2754,
      "num_input_tokens_seen": 12737560,
      "step": 16700
    },
    {
      "epoch": 34.729729729729726,
      "grad_norm": 0.00012064033944625407,
      "learning_rate": 0.1883978541169009,
      "loss": 0.2677,
      "num_input_tokens_seen": 12741432,
      "step": 16705
    },
    {
      "epoch": 34.74012474012474,
      "grad_norm": 7.742404704913497e-05,
      "learning_rate": 0.18834090897182854,
      "loss": 0.2625,
      "num_input_tokens_seen": 12745240,
      "step": 16710
    },
    {
      "epoch": 34.75051975051975,
      "grad_norm": 0.0003235915210098028,
      "learning_rate": 0.1882839579141062,
      "loss": 0.2692,
      "num_input_tokens_seen": 12749080,
      "step": 16715
    },
    {
      "epoch": 34.76091476091476,
      "grad_norm": 0.0003752792254090309,
      "learning_rate": 0.18822700095251646,
      "loss": 0.2434,
      "num_input_tokens_seen": 12752792,
      "step": 16720
    },
    {
      "epoch": 34.771309771309774,
      "grad_norm": 0.0003952011466026306,
      "learning_rate": 0.18817003809584273,
      "loss": 0.2678,
      "num_input_tokens_seen": 12756632,
      "step": 16725
    },
    {
      "epoch": 34.78170478170478,
      "grad_norm": 8.917592640500516e-05,
      "learning_rate": 0.1881130693528695,
      "loss": 0.2819,
      "num_input_tokens_seen": 12760536,
      "step": 16730
    },
    {
      "epoch": 34.79209979209979,
      "grad_norm": 0.0004670246271416545,
      "learning_rate": 0.18805609473238197,
      "loss": 0.2826,
      "num_input_tokens_seen": 12764248,
      "step": 16735
    },
    {
      "epoch": 34.802494802494806,
      "grad_norm": 0.0005638967268168926,
      "learning_rate": 0.18799911424316643,
      "loss": 0.2786,
      "num_input_tokens_seen": 12768056,
      "step": 16740
    },
    {
      "epoch": 34.812889812889814,
      "grad_norm": 0.00013777521962765604,
      "learning_rate": 0.18794212789400994,
      "loss": 0.2675,
      "num_input_tokens_seen": 12771896,
      "step": 16745
    },
    {
      "epoch": 34.82328482328482,
      "grad_norm": 0.00013420867617242038,
      "learning_rate": 0.18788513569370052,
      "loss": 0.2589,
      "num_input_tokens_seen": 12775576,
      "step": 16750
    },
    {
      "epoch": 34.83367983367983,
      "grad_norm": 0.0007713295053690672,
      "learning_rate": 0.1878281376510271,
      "loss": 0.2974,
      "num_input_tokens_seen": 12779288,
      "step": 16755
    },
    {
      "epoch": 34.84407484407485,
      "grad_norm": 0.0002932752249762416,
      "learning_rate": 0.18777113377477941,
      "loss": 0.2837,
      "num_input_tokens_seen": 12783128,
      "step": 16760
    },
    {
      "epoch": 34.854469854469855,
      "grad_norm": 0.0009167591924779117,
      "learning_rate": 0.1877141240737483,
      "loss": 0.3014,
      "num_input_tokens_seen": 12786904,
      "step": 16765
    },
    {
      "epoch": 34.86486486486486,
      "grad_norm": 0.00026592848007567227,
      "learning_rate": 0.18765710855672527,
      "loss": 0.2812,
      "num_input_tokens_seen": 12790712,
      "step": 16770
    },
    {
      "epoch": 34.87525987525988,
      "grad_norm": 0.00012960981985088438,
      "learning_rate": 0.18760008723250288,
      "loss": 0.2517,
      "num_input_tokens_seen": 12794328,
      "step": 16775
    },
    {
      "epoch": 34.88565488565489,
      "grad_norm": 0.0004121332603972405,
      "learning_rate": 0.18754306010987457,
      "loss": 0.2611,
      "num_input_tokens_seen": 12798200,
      "step": 16780
    },
    {
      "epoch": 34.896049896049895,
      "grad_norm": 0.0001139149462687783,
      "learning_rate": 0.18748602719763457,
      "loss": 0.2674,
      "num_input_tokens_seen": 12801976,
      "step": 16785
    },
    {
      "epoch": 34.906444906444904,
      "grad_norm": 9.393378422828391e-05,
      "learning_rate": 0.18742898850457804,
      "loss": 0.2575,
      "num_input_tokens_seen": 12805816,
      "step": 16790
    },
    {
      "epoch": 34.91683991683992,
      "grad_norm": 0.0006696488708257675,
      "learning_rate": 0.1873719440395012,
      "loss": 0.2696,
      "num_input_tokens_seen": 12809528,
      "step": 16795
    },
    {
      "epoch": 34.92723492723493,
      "grad_norm": 0.0005263136699795723,
      "learning_rate": 0.1873148938112009,
      "loss": 0.2601,
      "num_input_tokens_seen": 12813272,
      "step": 16800
    },
    {
      "epoch": 34.92723492723493,
      "eval_loss": 0.2487134486436844,
      "eval_runtime": 13.3944,
      "eval_samples_per_second": 63.907,
      "eval_steps_per_second": 15.977,
      "num_input_tokens_seen": 12813272,
      "step": 16800
    },
    {
      "epoch": 34.937629937629936,
      "grad_norm": 0.000244904775172472,
      "learning_rate": 0.18725783782847508,
      "loss": 0.2642,
      "num_input_tokens_seen": 12816920,
      "step": 16805
    },
    {
      "epoch": 34.94802494802495,
      "grad_norm": 7.562178507214412e-05,
      "learning_rate": 0.1872007761001224,
      "loss": 0.248,
      "num_input_tokens_seen": 12820664,
      "step": 16810
    },
    {
      "epoch": 34.95841995841996,
      "grad_norm": 0.00025312238722108305,
      "learning_rate": 0.1871437086349426,
      "loss": 0.2699,
      "num_input_tokens_seen": 12824536,
      "step": 16815
    },
    {
      "epoch": 34.96881496881497,
      "grad_norm": 0.00018691220611799508,
      "learning_rate": 0.18708663544173615,
      "loss": 0.2701,
      "num_input_tokens_seen": 12828312,
      "step": 16820
    },
    {
      "epoch": 34.979209979209976,
      "grad_norm": 6.183881487231702e-05,
      "learning_rate": 0.18702955652930442,
      "loss": 0.2694,
      "num_input_tokens_seen": 12832120,
      "step": 16825
    },
    {
      "epoch": 34.98960498960499,
      "grad_norm": 0.00010112955351360142,
      "learning_rate": 0.18697247190644972,
      "loss": 0.2454,
      "num_input_tokens_seen": 12835992,
      "step": 16830
    },
    {
      "epoch": 35.0,
      "grad_norm": 0.0004139370284974575,
      "learning_rate": 0.18691538158197527,
      "loss": 0.228,
      "num_input_tokens_seen": 12839656,
      "step": 16835
    },
    {
      "epoch": 35.01039501039501,
      "grad_norm": 8.809124119579792e-05,
      "learning_rate": 0.1868582855646851,
      "loss": 0.2882,
      "num_input_tokens_seen": 12843528,
      "step": 16840
    },
    {
      "epoch": 35.020790020790024,
      "grad_norm": 0.00020006607519462705,
      "learning_rate": 0.18680118386338404,
      "loss": 0.2757,
      "num_input_tokens_seen": 12847272,
      "step": 16845
    },
    {
      "epoch": 35.03118503118503,
      "grad_norm": 0.0004592253244481981,
      "learning_rate": 0.18674407648687794,
      "loss": 0.2698,
      "num_input_tokens_seen": 12850984,
      "step": 16850
    },
    {
      "epoch": 35.04158004158004,
      "grad_norm": 0.0007527959533035755,
      "learning_rate": 0.1866869634439736,
      "loss": 0.2361,
      "num_input_tokens_seen": 12855048,
      "step": 16855
    },
    {
      "epoch": 35.05197505197505,
      "grad_norm": 0.000490635575260967,
      "learning_rate": 0.18662984474347838,
      "loss": 0.2342,
      "num_input_tokens_seen": 12858824,
      "step": 16860
    },
    {
      "epoch": 35.062370062370064,
      "grad_norm": 7.042768993414938e-05,
      "learning_rate": 0.1865727203942008,
      "loss": 0.2212,
      "num_input_tokens_seen": 12862664,
      "step": 16865
    },
    {
      "epoch": 35.07276507276507,
      "grad_norm": 0.0005292988498695195,
      "learning_rate": 0.1865155904049501,
      "loss": 0.317,
      "num_input_tokens_seen": 12866664,
      "step": 16870
    },
    {
      "epoch": 35.08316008316008,
      "grad_norm": 0.00017502380069345236,
      "learning_rate": 0.1864584547845365,
      "loss": 0.2595,
      "num_input_tokens_seen": 12870600,
      "step": 16875
    },
    {
      "epoch": 35.093555093555096,
      "grad_norm": 0.00010429528629174456,
      "learning_rate": 0.186401313541771,
      "loss": 0.2744,
      "num_input_tokens_seen": 12874440,
      "step": 16880
    },
    {
      "epoch": 35.103950103950105,
      "grad_norm": 0.0002257017040392384,
      "learning_rate": 0.18634416668546552,
      "loss": 0.2791,
      "num_input_tokens_seen": 12878248,
      "step": 16885
    },
    {
      "epoch": 35.11434511434511,
      "grad_norm": 5.904418503632769e-05,
      "learning_rate": 0.1862870142244328,
      "loss": 0.2747,
      "num_input_tokens_seen": 12882120,
      "step": 16890
    },
    {
      "epoch": 35.12474012474012,
      "grad_norm": 0.00010746809130068868,
      "learning_rate": 0.1862298561674865,
      "loss": 0.2666,
      "num_input_tokens_seen": 12885992,
      "step": 16895
    },
    {
      "epoch": 35.13513513513514,
      "grad_norm": 0.000387133244657889,
      "learning_rate": 0.18617269252344104,
      "loss": 0.2563,
      "num_input_tokens_seen": 12889864,
      "step": 16900
    },
    {
      "epoch": 35.145530145530145,
      "grad_norm": 0.00039089430356398225,
      "learning_rate": 0.18611552330111186,
      "loss": 0.2775,
      "num_input_tokens_seen": 12893768,
      "step": 16905
    },
    {
      "epoch": 35.15592515592515,
      "grad_norm": 0.000233535174629651,
      "learning_rate": 0.18605834850931507,
      "loss": 0.2516,
      "num_input_tokens_seen": 12897608,
      "step": 16910
    },
    {
      "epoch": 35.16632016632017,
      "grad_norm": 0.00010665982699720189,
      "learning_rate": 0.18600116815686787,
      "loss": 0.2452,
      "num_input_tokens_seen": 12901608,
      "step": 16915
    },
    {
      "epoch": 35.17671517671518,
      "grad_norm": 0.0002713435678742826,
      "learning_rate": 0.1859439822525881,
      "loss": 0.2752,
      "num_input_tokens_seen": 12905320,
      "step": 16920
    },
    {
      "epoch": 35.187110187110186,
      "grad_norm": 0.0002450247120577842,
      "learning_rate": 0.18588679080529455,
      "loss": 0.262,
      "num_input_tokens_seen": 12909224,
      "step": 16925
    },
    {
      "epoch": 35.197505197505194,
      "grad_norm": 0.00023793573200237006,
      "learning_rate": 0.1858295938238069,
      "loss": 0.2721,
      "num_input_tokens_seen": 12912936,
      "step": 16930
    },
    {
      "epoch": 35.20790020790021,
      "grad_norm": 0.0001158507657237351,
      "learning_rate": 0.18577239131694562,
      "loss": 0.2855,
      "num_input_tokens_seen": 12916648,
      "step": 16935
    },
    {
      "epoch": 35.21829521829522,
      "grad_norm": 0.00017433053290005773,
      "learning_rate": 0.18571518329353204,
      "loss": 0.2489,
      "num_input_tokens_seen": 12920424,
      "step": 16940
    },
    {
      "epoch": 35.228690228690226,
      "grad_norm": 0.0002863065747078508,
      "learning_rate": 0.18565796976238838,
      "loss": 0.2398,
      "num_input_tokens_seen": 12924136,
      "step": 16945
    },
    {
      "epoch": 35.23908523908524,
      "grad_norm": 0.00020643460447899997,
      "learning_rate": 0.18560075073233764,
      "loss": 0.2636,
      "num_input_tokens_seen": 12927912,
      "step": 16950
    },
    {
      "epoch": 35.24948024948025,
      "grad_norm": 0.00025633437326177955,
      "learning_rate": 0.18554352621220377,
      "loss": 0.2922,
      "num_input_tokens_seen": 12931752,
      "step": 16955
    },
    {
      "epoch": 35.25987525987526,
      "grad_norm": 0.00010995643242495134,
      "learning_rate": 0.18548629621081153,
      "loss": 0.2664,
      "num_input_tokens_seen": 12935432,
      "step": 16960
    },
    {
      "epoch": 35.270270270270274,
      "grad_norm": 4.525436452240683e-05,
      "learning_rate": 0.18542906073698645,
      "loss": 0.2567,
      "num_input_tokens_seen": 12939176,
      "step": 16965
    },
    {
      "epoch": 35.28066528066528,
      "grad_norm": 0.00027432964998297393,
      "learning_rate": 0.18537181979955494,
      "loss": 0.271,
      "num_input_tokens_seen": 12943112,
      "step": 16970
    },
    {
      "epoch": 35.29106029106029,
      "grad_norm": 0.00011665987403830513,
      "learning_rate": 0.18531457340734434,
      "loss": 0.255,
      "num_input_tokens_seen": 12946888,
      "step": 16975
    },
    {
      "epoch": 35.3014553014553,
      "grad_norm": 0.0003307465522084385,
      "learning_rate": 0.1852573215691827,
      "loss": 0.2933,
      "num_input_tokens_seen": 12950696,
      "step": 16980
    },
    {
      "epoch": 35.311850311850314,
      "grad_norm": 0.0006522841867990792,
      "learning_rate": 0.18520006429389904,
      "loss": 0.2381,
      "num_input_tokens_seen": 12954408,
      "step": 16985
    },
    {
      "epoch": 35.32224532224532,
      "grad_norm": 0.0007473426521755755,
      "learning_rate": 0.1851428015903231,
      "loss": 0.278,
      "num_input_tokens_seen": 12958216,
      "step": 16990
    },
    {
      "epoch": 35.33264033264033,
      "grad_norm": 0.0001662546128500253,
      "learning_rate": 0.1850855334672855,
      "loss": 0.2428,
      "num_input_tokens_seen": 12962024,
      "step": 16995
    },
    {
      "epoch": 35.343035343035346,
      "grad_norm": 0.0002309860719833523,
      "learning_rate": 0.1850282599336178,
      "loss": 0.2638,
      "num_input_tokens_seen": 12965896,
      "step": 17000
    },
    {
      "epoch": 35.343035343035346,
      "eval_loss": 0.24889925122261047,
      "eval_runtime": 13.3956,
      "eval_samples_per_second": 63.901,
      "eval_steps_per_second": 15.975,
      "num_input_tokens_seen": 12965896,
      "step": 17000
    },
    {
      "epoch": 35.353430353430355,
      "grad_norm": 0.0004286384501028806,
      "learning_rate": 0.18497098099815215,
      "loss": 0.2508,
      "num_input_tokens_seen": 12969768,
      "step": 17005
    },
    {
      "epoch": 35.36382536382536,
      "grad_norm": 2.1405317966127768e-05,
      "learning_rate": 0.18491369666972174,
      "loss": 0.2632,
      "num_input_tokens_seen": 12973608,
      "step": 17010
    },
    {
      "epoch": 35.37422037422037,
      "grad_norm": 7.8148914326448e-05,
      "learning_rate": 0.1848564069571606,
      "loss": 0.2618,
      "num_input_tokens_seen": 12977288,
      "step": 17015
    },
    {
      "epoch": 35.38461538461539,
      "grad_norm": 0.00011393125168979168,
      "learning_rate": 0.18479911186930348,
      "loss": 0.2816,
      "num_input_tokens_seen": 12981000,
      "step": 17020
    },
    {
      "epoch": 35.395010395010395,
      "grad_norm": 0.0001927412231452763,
      "learning_rate": 0.18474181141498597,
      "loss": 0.2891,
      "num_input_tokens_seen": 12984744,
      "step": 17025
    },
    {
      "epoch": 35.4054054054054,
      "grad_norm": 0.00037378535489551723,
      "learning_rate": 0.18468450560304453,
      "loss": 0.2839,
      "num_input_tokens_seen": 12988648,
      "step": 17030
    },
    {
      "epoch": 35.41580041580042,
      "grad_norm": 4.6136443415889516e-05,
      "learning_rate": 0.1846271944423165,
      "loss": 0.2726,
      "num_input_tokens_seen": 12992488,
      "step": 17035
    },
    {
      "epoch": 35.42619542619543,
      "grad_norm": 9.900882287183776e-05,
      "learning_rate": 0.18456987794163993,
      "loss": 0.2775,
      "num_input_tokens_seen": 12996296,
      "step": 17040
    },
    {
      "epoch": 35.436590436590436,
      "grad_norm": 0.00014404222019948065,
      "learning_rate": 0.18451255610985373,
      "loss": 0.2757,
      "num_input_tokens_seen": 13000072,
      "step": 17045
    },
    {
      "epoch": 35.446985446985444,
      "grad_norm": 0.00025201408425346017,
      "learning_rate": 0.18445522895579766,
      "loss": 0.267,
      "num_input_tokens_seen": 13003976,
      "step": 17050
    },
    {
      "epoch": 35.45738045738046,
      "grad_norm": 0.0001261031866306439,
      "learning_rate": 0.1843978964883123,
      "loss": 0.2571,
      "num_input_tokens_seen": 13007848,
      "step": 17055
    },
    {
      "epoch": 35.46777546777547,
      "grad_norm": 0.00010357849532738328,
      "learning_rate": 0.18434055871623906,
      "loss": 0.2723,
      "num_input_tokens_seen": 13011624,
      "step": 17060
    },
    {
      "epoch": 35.478170478170476,
      "grad_norm": 0.00022431333491113037,
      "learning_rate": 0.18428321564842007,
      "loss": 0.2557,
      "num_input_tokens_seen": 13015432,
      "step": 17065
    },
    {
      "epoch": 35.48856548856549,
      "grad_norm": 0.0003426216426305473,
      "learning_rate": 0.18422586729369841,
      "loss": 0.2555,
      "num_input_tokens_seen": 13019144,
      "step": 17070
    },
    {
      "epoch": 35.4989604989605,
      "grad_norm": 0.0005259495228528976,
      "learning_rate": 0.1841685136609179,
      "loss": 0.2831,
      "num_input_tokens_seen": 13023176,
      "step": 17075
    },
    {
      "epoch": 35.50935550935551,
      "grad_norm": 0.00013832398690283298,
      "learning_rate": 0.18411115475892326,
      "loss": 0.2593,
      "num_input_tokens_seen": 13026952,
      "step": 17080
    },
    {
      "epoch": 35.51975051975052,
      "grad_norm": 2.0800953279831447e-05,
      "learning_rate": 0.18405379059655982,
      "loss": 0.2628,
      "num_input_tokens_seen": 13030856,
      "step": 17085
    },
    {
      "epoch": 35.53014553014553,
      "grad_norm": 8.958084072219208e-05,
      "learning_rate": 0.1839964211826739,
      "loss": 0.2829,
      "num_input_tokens_seen": 13034792,
      "step": 17090
    },
    {
      "epoch": 35.54054054054054,
      "grad_norm": 0.0004163573030382395,
      "learning_rate": 0.18393904652611265,
      "loss": 0.2686,
      "num_input_tokens_seen": 13038600,
      "step": 17095
    },
    {
      "epoch": 35.55093555093555,
      "grad_norm": 0.00042450649198144674,
      "learning_rate": 0.18388166663572392,
      "loss": 0.264,
      "num_input_tokens_seen": 13042504,
      "step": 17100
    },
    {
      "epoch": 35.561330561330564,
      "grad_norm": 0.0001936911721713841,
      "learning_rate": 0.18382428152035643,
      "loss": 0.2406,
      "num_input_tokens_seen": 13046376,
      "step": 17105
    },
    {
      "epoch": 35.57172557172557,
      "grad_norm": 9.386425517732278e-05,
      "learning_rate": 0.1837668911888596,
      "loss": 0.2355,
      "num_input_tokens_seen": 13050088,
      "step": 17110
    },
    {
      "epoch": 35.58212058212058,
      "grad_norm": 0.00011518143583089113,
      "learning_rate": 0.18370949565008388,
      "loss": 0.2776,
      "num_input_tokens_seen": 13053896,
      "step": 17115
    },
    {
      "epoch": 35.59251559251559,
      "grad_norm": 6.083751213736832e-05,
      "learning_rate": 0.1836520949128803,
      "loss": 0.2569,
      "num_input_tokens_seen": 13057736,
      "step": 17120
    },
    {
      "epoch": 35.602910602910605,
      "grad_norm": 0.00035684771137312055,
      "learning_rate": 0.18359468898610076,
      "loss": 0.2829,
      "num_input_tokens_seen": 13061672,
      "step": 17125
    },
    {
      "epoch": 35.61330561330561,
      "grad_norm": 0.0005231992108747363,
      "learning_rate": 0.18353727787859797,
      "loss": 0.2725,
      "num_input_tokens_seen": 13065448,
      "step": 17130
    },
    {
      "epoch": 35.62370062370062,
      "grad_norm": 0.00024077462148852646,
      "learning_rate": 0.18347986159922552,
      "loss": 0.2711,
      "num_input_tokens_seen": 13069224,
      "step": 17135
    },
    {
      "epoch": 35.63409563409564,
      "grad_norm": 4.889236151939258e-05,
      "learning_rate": 0.1834224401568377,
      "loss": 0.2862,
      "num_input_tokens_seen": 13073000,
      "step": 17140
    },
    {
      "epoch": 35.644490644490645,
      "grad_norm": 0.00034011725801974535,
      "learning_rate": 0.1833650135602896,
      "loss": 0.2667,
      "num_input_tokens_seen": 13076776,
      "step": 17145
    },
    {
      "epoch": 35.65488565488565,
      "grad_norm": 0.00021094617841299623,
      "learning_rate": 0.18330758181843707,
      "loss": 0.2771,
      "num_input_tokens_seen": 13080680,
      "step": 17150
    },
    {
      "epoch": 35.66528066528066,
      "grad_norm": 0.0004294009995646775,
      "learning_rate": 0.18325014494013686,
      "loss": 0.2893,
      "num_input_tokens_seen": 13084552,
      "step": 17155
    },
    {
      "epoch": 35.67567567567568,
      "grad_norm": 0.000628128822427243,
      "learning_rate": 0.18319270293424647,
      "loss": 0.2836,
      "num_input_tokens_seen": 13088360,
      "step": 17160
    },
    {
      "epoch": 35.686070686070686,
      "grad_norm": 0.000327672460116446,
      "learning_rate": 0.18313525580962417,
      "loss": 0.271,
      "num_input_tokens_seen": 13092104,
      "step": 17165
    },
    {
      "epoch": 35.696465696465694,
      "grad_norm": 0.0004409045504871756,
      "learning_rate": 0.18307780357512896,
      "loss": 0.2593,
      "num_input_tokens_seen": 13095848,
      "step": 17170
    },
    {
      "epoch": 35.70686070686071,
      "grad_norm": 0.0002291866549057886,
      "learning_rate": 0.1830203462396208,
      "loss": 0.3165,
      "num_input_tokens_seen": 13099688,
      "step": 17175
    },
    {
      "epoch": 35.71725571725572,
      "grad_norm": 0.00036499035195447505,
      "learning_rate": 0.18296288381196033,
      "loss": 0.2498,
      "num_input_tokens_seen": 13103656,
      "step": 17180
    },
    {
      "epoch": 35.727650727650726,
      "grad_norm": 0.0004183951241429895,
      "learning_rate": 0.1829054163010089,
      "loss": 0.2848,
      "num_input_tokens_seen": 13107528,
      "step": 17185
    },
    {
      "epoch": 35.73804573804574,
      "grad_norm": 8.121300197672099e-05,
      "learning_rate": 0.18284794371562874,
      "loss": 0.2782,
      "num_input_tokens_seen": 13111272,
      "step": 17190
    },
    {
      "epoch": 35.74844074844075,
      "grad_norm": 0.00026350145344622433,
      "learning_rate": 0.18279046606468288,
      "loss": 0.2737,
      "num_input_tokens_seen": 13115080,
      "step": 17195
    },
    {
      "epoch": 35.75883575883576,
      "grad_norm": 0.00017929417663253844,
      "learning_rate": 0.1827329833570351,
      "loss": 0.2766,
      "num_input_tokens_seen": 13118824,
      "step": 17200
    },
    {
      "epoch": 35.75883575883576,
      "eval_loss": 0.25238293409347534,
      "eval_runtime": 13.4257,
      "eval_samples_per_second": 63.758,
      "eval_steps_per_second": 15.94,
      "num_input_tokens_seen": 13118824,
      "step": 17200
    },
    {
      "epoch": 35.76923076923077,
      "grad_norm": 0.00010151242167921737,
      "learning_rate": 0.18267549560154991,
      "loss": 0.2499,
      "num_input_tokens_seen": 13122632,
      "step": 17205
    },
    {
      "epoch": 35.77962577962578,
      "grad_norm": 0.00030825851717963815,
      "learning_rate": 0.18261800280709267,
      "loss": 0.2719,
      "num_input_tokens_seen": 13126536,
      "step": 17210
    },
    {
      "epoch": 35.79002079002079,
      "grad_norm": 0.00035874691093340516,
      "learning_rate": 0.18256050498252957,
      "loss": 0.2659,
      "num_input_tokens_seen": 13130376,
      "step": 17215
    },
    {
      "epoch": 35.8004158004158,
      "grad_norm": 0.00014793197624385357,
      "learning_rate": 0.18250300213672735,
      "loss": 0.2887,
      "num_input_tokens_seen": 13134088,
      "step": 17220
    },
    {
      "epoch": 35.810810810810814,
      "grad_norm": 0.0006373718497343361,
      "learning_rate": 0.18244549427855378,
      "loss": 0.269,
      "num_input_tokens_seen": 13137832,
      "step": 17225
    },
    {
      "epoch": 35.82120582120582,
      "grad_norm": 0.0006245653494261205,
      "learning_rate": 0.1823879814168772,
      "loss": 0.2459,
      "num_input_tokens_seen": 13141576,
      "step": 17230
    },
    {
      "epoch": 35.83160083160083,
      "grad_norm": 0.00022500309569295496,
      "learning_rate": 0.18233046356056692,
      "loss": 0.2446,
      "num_input_tokens_seen": 13145480,
      "step": 17235
    },
    {
      "epoch": 35.84199584199584,
      "grad_norm": 0.00011551700299605727,
      "learning_rate": 0.18227294071849284,
      "loss": 0.3236,
      "num_input_tokens_seen": 13149384,
      "step": 17240
    },
    {
      "epoch": 35.852390852390855,
      "grad_norm": 0.0004671224160119891,
      "learning_rate": 0.18221541289952578,
      "loss": 0.2341,
      "num_input_tokens_seen": 13153320,
      "step": 17245
    },
    {
      "epoch": 35.86278586278586,
      "grad_norm": 0.0003103829803876579,
      "learning_rate": 0.18215788011253717,
      "loss": 0.2534,
      "num_input_tokens_seen": 13157032,
      "step": 17250
    },
    {
      "epoch": 35.87318087318087,
      "grad_norm": 0.000261551613220945,
      "learning_rate": 0.18210034236639935,
      "loss": 0.2583,
      "num_input_tokens_seen": 13160968,
      "step": 17255
    },
    {
      "epoch": 35.88357588357589,
      "grad_norm": 0.0004658151010517031,
      "learning_rate": 0.1820427996699853,
      "loss": 0.2515,
      "num_input_tokens_seen": 13164776,
      "step": 17260
    },
    {
      "epoch": 35.893970893970895,
      "grad_norm": 0.00015330652240663767,
      "learning_rate": 0.1819852520321689,
      "loss": 0.2559,
      "num_input_tokens_seen": 13168456,
      "step": 17265
    },
    {
      "epoch": 35.9043659043659,
      "grad_norm": 0.00028016703436151147,
      "learning_rate": 0.18192769946182466,
      "loss": 0.2674,
      "num_input_tokens_seen": 13172168,
      "step": 17270
    },
    {
      "epoch": 35.91476091476091,
      "grad_norm": 0.0004740901349578053,
      "learning_rate": 0.18187014196782794,
      "loss": 0.2638,
      "num_input_tokens_seen": 13176040,
      "step": 17275
    },
    {
      "epoch": 35.92515592515593,
      "grad_norm": 0.00027310673613101244,
      "learning_rate": 0.18181257955905486,
      "loss": 0.2402,
      "num_input_tokens_seen": 13179976,
      "step": 17280
    },
    {
      "epoch": 35.935550935550935,
      "grad_norm": 0.0001899885101011023,
      "learning_rate": 0.18175501224438217,
      "loss": 0.2603,
      "num_input_tokens_seen": 13183720,
      "step": 17285
    },
    {
      "epoch": 35.945945945945944,
      "grad_norm": 0.00026490914751775563,
      "learning_rate": 0.18169744003268756,
      "loss": 0.2969,
      "num_input_tokens_seen": 13187496,
      "step": 17290
    },
    {
      "epoch": 35.95634095634096,
      "grad_norm": 0.00013785452756565064,
      "learning_rate": 0.18163986293284937,
      "loss": 0.2707,
      "num_input_tokens_seen": 13191336,
      "step": 17295
    },
    {
      "epoch": 35.96673596673597,
      "grad_norm": 0.00028141759685240686,
      "learning_rate": 0.18158228095374673,
      "loss": 0.2485,
      "num_input_tokens_seen": 13195240,
      "step": 17300
    },
    {
      "epoch": 35.977130977130976,
      "grad_norm": 0.00023315436556003988,
      "learning_rate": 0.18152469410425945,
      "loss": 0.2933,
      "num_input_tokens_seen": 13198952,
      "step": 17305
    },
    {
      "epoch": 35.987525987525984,
      "grad_norm": 0.00018487342458683997,
      "learning_rate": 0.18146710239326813,
      "loss": 0.2949,
      "num_input_tokens_seen": 13202728,
      "step": 17310
    },
    {
      "epoch": 35.997920997921,
      "grad_norm": 0.0006413618102669716,
      "learning_rate": 0.18140950582965423,
      "loss": 0.2792,
      "num_input_tokens_seen": 13206440,
      "step": 17315
    },
    {
      "epoch": 36.00831600831601,
      "grad_norm": 9.09929076442495e-05,
      "learning_rate": 0.1813519044222998,
      "loss": 0.2695,
      "num_input_tokens_seen": 13210144,
      "step": 17320
    },
    {
      "epoch": 36.018711018711016,
      "grad_norm": 0.0003249425790272653,
      "learning_rate": 0.18129429818008772,
      "loss": 0.2612,
      "num_input_tokens_seen": 13213984,
      "step": 17325
    },
    {
      "epoch": 36.02910602910603,
      "grad_norm": 0.00035491675953380764,
      "learning_rate": 0.18123668711190163,
      "loss": 0.2658,
      "num_input_tokens_seen": 13217856,
      "step": 17330
    },
    {
      "epoch": 36.03950103950104,
      "grad_norm": 0.00041443223017267883,
      "learning_rate": 0.18117907122662583,
      "loss": 0.2774,
      "num_input_tokens_seen": 13221504,
      "step": 17335
    },
    {
      "epoch": 36.04989604989605,
      "grad_norm": 0.000182196352398023,
      "learning_rate": 0.1811214505331454,
      "loss": 0.2567,
      "num_input_tokens_seen": 13225312,
      "step": 17340
    },
    {
      "epoch": 36.06029106029106,
      "grad_norm": 0.00018606791854836047,
      "learning_rate": 0.1810638250403462,
      "loss": 0.2725,
      "num_input_tokens_seen": 13229248,
      "step": 17345
    },
    {
      "epoch": 36.07068607068607,
      "grad_norm": 0.000773865613155067,
      "learning_rate": 0.1810061947571148,
      "loss": 0.2267,
      "num_input_tokens_seen": 13233024,
      "step": 17350
    },
    {
      "epoch": 36.08108108108108,
      "grad_norm": 0.00011020462261512876,
      "learning_rate": 0.1809485596923385,
      "loss": 0.2727,
      "num_input_tokens_seen": 13237056,
      "step": 17355
    },
    {
      "epoch": 36.09147609147609,
      "grad_norm": 0.000755861634388566,
      "learning_rate": 0.18089091985490546,
      "loss": 0.2635,
      "num_input_tokens_seen": 13240832,
      "step": 17360
    },
    {
      "epoch": 36.101871101871104,
      "grad_norm": 0.00044536785571835935,
      "learning_rate": 0.18083327525370432,
      "loss": 0.251,
      "num_input_tokens_seen": 13244768,
      "step": 17365
    },
    {
      "epoch": 36.11226611226611,
      "grad_norm": 0.00032161345006898046,
      "learning_rate": 0.18077562589762464,
      "loss": 0.2612,
      "num_input_tokens_seen": 13248704,
      "step": 17370
    },
    {
      "epoch": 36.12266112266112,
      "grad_norm": 7.415869185933843e-05,
      "learning_rate": 0.1807179717955567,
      "loss": 0.2774,
      "num_input_tokens_seen": 13252672,
      "step": 17375
    },
    {
      "epoch": 36.13305613305613,
      "grad_norm": 7.982958777574822e-05,
      "learning_rate": 0.1806603129563915,
      "loss": 0.2487,
      "num_input_tokens_seen": 13256480,
      "step": 17380
    },
    {
      "epoch": 36.143451143451145,
      "grad_norm": 0.0001779584272298962,
      "learning_rate": 0.1806026493890208,
      "loss": 0.2872,
      "num_input_tokens_seen": 13260352,
      "step": 17385
    },
    {
      "epoch": 36.15384615384615,
      "grad_norm": 9.448125638300553e-05,
      "learning_rate": 0.18054498110233688,
      "loss": 0.251,
      "num_input_tokens_seen": 13264128,
      "step": 17390
    },
    {
      "epoch": 36.16424116424116,
      "grad_norm": 0.00026116447406820953,
      "learning_rate": 0.1804873081052331,
      "loss": 0.2471,
      "num_input_tokens_seen": 13268032,
      "step": 17395
    },
    {
      "epoch": 36.17463617463618,
      "grad_norm": 0.00038582689012400806,
      "learning_rate": 0.18042963040660326,
      "loss": 0.2599,
      "num_input_tokens_seen": 13271872,
      "step": 17400
    },
    {
      "epoch": 36.17463617463618,
      "eval_loss": 0.2490328550338745,
      "eval_runtime": 13.4571,
      "eval_samples_per_second": 63.61,
      "eval_steps_per_second": 15.902,
      "num_input_tokens_seen": 13271872,
      "step": 17400
    },
    {
      "epoch": 36.185031185031185,
      "grad_norm": 5.038334347773343e-05,
      "learning_rate": 0.180371948015342,
      "loss": 0.2377,
      "num_input_tokens_seen": 13275808,
      "step": 17405
    },
    {
      "epoch": 36.195426195426194,
      "grad_norm": 0.0003406234900467098,
      "learning_rate": 0.18031426094034472,
      "loss": 0.2603,
      "num_input_tokens_seen": 13279584,
      "step": 17410
    },
    {
      "epoch": 36.20582120582121,
      "grad_norm": 0.0002932468487415463,
      "learning_rate": 0.18025656919050737,
      "loss": 0.2597,
      "num_input_tokens_seen": 13283392,
      "step": 17415
    },
    {
      "epoch": 36.21621621621622,
      "grad_norm": 0.0002370987058384344,
      "learning_rate": 0.18019887277472688,
      "loss": 0.2543,
      "num_input_tokens_seen": 13287040,
      "step": 17420
    },
    {
      "epoch": 36.226611226611226,
      "grad_norm": 0.0003863436577375978,
      "learning_rate": 0.18014117170190067,
      "loss": 0.2766,
      "num_input_tokens_seen": 13290976,
      "step": 17425
    },
    {
      "epoch": 36.237006237006234,
      "grad_norm": 0.0001916103792609647,
      "learning_rate": 0.18008346598092703,
      "loss": 0.2679,
      "num_input_tokens_seen": 13294944,
      "step": 17430
    },
    {
      "epoch": 36.24740124740125,
      "grad_norm": 0.00019920420891139656,
      "learning_rate": 0.18002575562070489,
      "loss": 0.2588,
      "num_input_tokens_seen": 13298720,
      "step": 17435
    },
    {
      "epoch": 36.25779625779626,
      "grad_norm": 0.00015447613259311765,
      "learning_rate": 0.1799680406301339,
      "loss": 0.2726,
      "num_input_tokens_seen": 13302560,
      "step": 17440
    },
    {
      "epoch": 36.268191268191266,
      "grad_norm": 0.0005707088857889175,
      "learning_rate": 0.17991032101811447,
      "loss": 0.2667,
      "num_input_tokens_seen": 13306336,
      "step": 17445
    },
    {
      "epoch": 36.27858627858628,
      "grad_norm": 0.00012345347204245627,
      "learning_rate": 0.1798525967935476,
      "loss": 0.2776,
      "num_input_tokens_seen": 13310080,
      "step": 17450
    },
    {
      "epoch": 36.28898128898129,
      "grad_norm": 3.670494697871618e-05,
      "learning_rate": 0.17979486796533517,
      "loss": 0.2688,
      "num_input_tokens_seen": 13313856,
      "step": 17455
    },
    {
      "epoch": 36.2993762993763,
      "grad_norm": 0.00022341989097185433,
      "learning_rate": 0.1797371345423797,
      "loss": 0.261,
      "num_input_tokens_seen": 13317728,
      "step": 17460
    },
    {
      "epoch": 36.30977130977131,
      "grad_norm": 0.0005145045579411089,
      "learning_rate": 0.17967939653358436,
      "loss": 0.2612,
      "num_input_tokens_seen": 13321536,
      "step": 17465
    },
    {
      "epoch": 36.32016632016632,
      "grad_norm": 7.41965341148898e-05,
      "learning_rate": 0.17962165394785315,
      "loss": 0.2747,
      "num_input_tokens_seen": 13325216,
      "step": 17470
    },
    {
      "epoch": 36.33056133056133,
      "grad_norm": 0.0006184731610119343,
      "learning_rate": 0.17956390679409057,
      "loss": 0.2728,
      "num_input_tokens_seen": 13328960,
      "step": 17475
    },
    {
      "epoch": 36.34095634095634,
      "grad_norm": 0.00042669460526667535,
      "learning_rate": 0.1795061550812021,
      "loss": 0.2803,
      "num_input_tokens_seen": 13332768,
      "step": 17480
    },
    {
      "epoch": 36.351351351351354,
      "grad_norm": 0.00027450561174191535,
      "learning_rate": 0.1794483988180937,
      "loss": 0.2643,
      "num_input_tokens_seen": 13336704,
      "step": 17485
    },
    {
      "epoch": 36.36174636174636,
      "grad_norm": 0.00046788016334176064,
      "learning_rate": 0.17939063801367214,
      "loss": 0.2774,
      "num_input_tokens_seen": 13340736,
      "step": 17490
    },
    {
      "epoch": 36.37214137214137,
      "grad_norm": 0.0002466692531015724,
      "learning_rate": 0.17933287267684483,
      "loss": 0.2701,
      "num_input_tokens_seen": 13344448,
      "step": 17495
    },
    {
      "epoch": 36.38253638253638,
      "grad_norm": 0.00035465057590045035,
      "learning_rate": 0.17927510281651995,
      "loss": 0.2726,
      "num_input_tokens_seen": 13348384,
      "step": 17500
    },
    {
      "epoch": 36.392931392931395,
      "grad_norm": 0.0003793487267103046,
      "learning_rate": 0.17921732844160634,
      "loss": 0.2792,
      "num_input_tokens_seen": 13352224,
      "step": 17505
    },
    {
      "epoch": 36.4033264033264,
      "grad_norm": 9.518696606392041e-05,
      "learning_rate": 0.17915954956101351,
      "loss": 0.27,
      "num_input_tokens_seen": 13355936,
      "step": 17510
    },
    {
      "epoch": 36.41372141372141,
      "grad_norm": 0.0002673929266165942,
      "learning_rate": 0.17910176618365165,
      "loss": 0.2563,
      "num_input_tokens_seen": 13359936,
      "step": 17515
    },
    {
      "epoch": 36.42411642411643,
      "grad_norm": 0.0003463521134108305,
      "learning_rate": 0.17904397831843177,
      "loss": 0.2748,
      "num_input_tokens_seen": 13363712,
      "step": 17520
    },
    {
      "epoch": 36.434511434511435,
      "grad_norm": 0.00024095296976156533,
      "learning_rate": 0.17898618597426547,
      "loss": 0.2521,
      "num_input_tokens_seen": 13367488,
      "step": 17525
    },
    {
      "epoch": 36.444906444906444,
      "grad_norm": 8.075922232819721e-05,
      "learning_rate": 0.17892838916006495,
      "loss": 0.2503,
      "num_input_tokens_seen": 13371264,
      "step": 17530
    },
    {
      "epoch": 36.45530145530145,
      "grad_norm": 6.309789750957862e-05,
      "learning_rate": 0.17887058788474333,
      "loss": 0.2417,
      "num_input_tokens_seen": 13374944,
      "step": 17535
    },
    {
      "epoch": 36.46569646569647,
      "grad_norm": 0.000544725451618433,
      "learning_rate": 0.17881278215721427,
      "loss": 0.2756,
      "num_input_tokens_seen": 13378752,
      "step": 17540
    },
    {
      "epoch": 36.476091476091476,
      "grad_norm": 0.00040329451439902186,
      "learning_rate": 0.1787549719863921,
      "loss": 0.2791,
      "num_input_tokens_seen": 13382496,
      "step": 17545
    },
    {
      "epoch": 36.486486486486484,
      "grad_norm": 0.0004914184100925922,
      "learning_rate": 0.17869715738119188,
      "loss": 0.2762,
      "num_input_tokens_seen": 13386304,
      "step": 17550
    },
    {
      "epoch": 36.4968814968815,
      "grad_norm": 9.815453813644126e-05,
      "learning_rate": 0.17863933835052936,
      "loss": 0.257,
      "num_input_tokens_seen": 13390176,
      "step": 17555
    },
    {
      "epoch": 36.50727650727651,
      "grad_norm": 0.0001887131220428273,
      "learning_rate": 0.17858151490332097,
      "loss": 0.2882,
      "num_input_tokens_seen": 13394048,
      "step": 17560
    },
    {
      "epoch": 36.517671517671516,
      "grad_norm": 6.773704080842435e-05,
      "learning_rate": 0.17852368704848381,
      "loss": 0.2767,
      "num_input_tokens_seen": 13397888,
      "step": 17565
    },
    {
      "epoch": 36.528066528066525,
      "grad_norm": 0.0003230652364436537,
      "learning_rate": 0.17846585479493565,
      "loss": 0.2654,
      "num_input_tokens_seen": 13401664,
      "step": 17570
    },
    {
      "epoch": 36.53846153846154,
      "grad_norm": 0.00027159383171238005,
      "learning_rate": 0.178408018151595,
      "loss": 0.2478,
      "num_input_tokens_seen": 13405472,
      "step": 17575
    },
    {
      "epoch": 36.54885654885655,
      "grad_norm": 0.00012372738274279982,
      "learning_rate": 0.17835017712738085,
      "loss": 0.2714,
      "num_input_tokens_seen": 13409152,
      "step": 17580
    },
    {
      "epoch": 36.55925155925156,
      "grad_norm": 0.00015436962712556124,
      "learning_rate": 0.17829233173121323,
      "loss": 0.2584,
      "num_input_tokens_seen": 13412960,
      "step": 17585
    },
    {
      "epoch": 36.56964656964657,
      "grad_norm": 0.0001529431319795549,
      "learning_rate": 0.17823448197201244,
      "loss": 0.2454,
      "num_input_tokens_seen": 13416672,
      "step": 17590
    },
    {
      "epoch": 36.58004158004158,
      "grad_norm": 0.00039333649328909814,
      "learning_rate": 0.1781766278586997,
      "loss": 0.285,
      "num_input_tokens_seen": 13420384,
      "step": 17595
    },
    {
      "epoch": 36.59043659043659,
      "grad_norm": 0.00039363655378110707,
      "learning_rate": 0.1781187694001969,
      "loss": 0.2689,
      "num_input_tokens_seen": 13424128,
      "step": 17600
    },
    {
      "epoch": 36.59043659043659,
      "eval_loss": 0.2603258788585663,
      "eval_runtime": 13.4013,
      "eval_samples_per_second": 63.874,
      "eval_steps_per_second": 15.969,
      "num_input_tokens_seen": 13424128,
      "step": 17600
    },
    {
      "epoch": 36.6008316008316,
      "grad_norm": 0.00012132203846704215,
      "learning_rate": 0.1780609066054265,
      "loss": 0.2846,
      "num_input_tokens_seen": 13428064,
      "step": 17605
    },
    {
      "epoch": 36.61122661122661,
      "grad_norm": 0.0001391533005516976,
      "learning_rate": 0.17800303948331164,
      "loss": 0.2743,
      "num_input_tokens_seen": 13431968,
      "step": 17610
    },
    {
      "epoch": 36.62162162162162,
      "grad_norm": 6.1971259128768e-05,
      "learning_rate": 0.1779451680427762,
      "loss": 0.2542,
      "num_input_tokens_seen": 13435712,
      "step": 17615
    },
    {
      "epoch": 36.63201663201663,
      "grad_norm": 0.0002711884444579482,
      "learning_rate": 0.17788729229274464,
      "loss": 0.2405,
      "num_input_tokens_seen": 13439456,
      "step": 17620
    },
    {
      "epoch": 36.642411642411645,
      "grad_norm": 9.95310110738501e-05,
      "learning_rate": 0.17782941224214222,
      "loss": 0.3031,
      "num_input_tokens_seen": 13443232,
      "step": 17625
    },
    {
      "epoch": 36.65280665280665,
      "grad_norm": 0.00027481201686896384,
      "learning_rate": 0.17777152789989464,
      "loss": 0.247,
      "num_input_tokens_seen": 13447008,
      "step": 17630
    },
    {
      "epoch": 36.66320166320166,
      "grad_norm": 0.0005391245940700173,
      "learning_rate": 0.17771363927492845,
      "loss": 0.2543,
      "num_input_tokens_seen": 13450720,
      "step": 17635
    },
    {
      "epoch": 36.67359667359668,
      "grad_norm": 0.0005165913025848567,
      "learning_rate": 0.17765574637617085,
      "loss": 0.2831,
      "num_input_tokens_seen": 13454464,
      "step": 17640
    },
    {
      "epoch": 36.683991683991685,
      "grad_norm": 0.0005055902292951941,
      "learning_rate": 0.17759784921254962,
      "loss": 0.259,
      "num_input_tokens_seen": 13458272,
      "step": 17645
    },
    {
      "epoch": 36.694386694386694,
      "grad_norm": 0.0004741275042761117,
      "learning_rate": 0.1775399477929932,
      "loss": 0.2678,
      "num_input_tokens_seen": 13462016,
      "step": 17650
    },
    {
      "epoch": 36.7047817047817,
      "grad_norm": 0.0002835779741872102,
      "learning_rate": 0.17748204212643076,
      "loss": 0.2712,
      "num_input_tokens_seen": 13465728,
      "step": 17655
    },
    {
      "epoch": 36.71517671517672,
      "grad_norm": 0.00030802603578194976,
      "learning_rate": 0.17742413222179204,
      "loss": 0.2549,
      "num_input_tokens_seen": 13469664,
      "step": 17660
    },
    {
      "epoch": 36.725571725571726,
      "grad_norm": 0.00016740414139349014,
      "learning_rate": 0.17736621808800754,
      "loss": 0.2492,
      "num_input_tokens_seen": 13473376,
      "step": 17665
    },
    {
      "epoch": 36.735966735966734,
      "grad_norm": 0.0004384123894851655,
      "learning_rate": 0.17730829973400827,
      "loss": 0.2867,
      "num_input_tokens_seen": 13477248,
      "step": 17670
    },
    {
      "epoch": 36.74636174636175,
      "grad_norm": 0.00034706469159573317,
      "learning_rate": 0.17725037716872602,
      "loss": 0.2554,
      "num_input_tokens_seen": 13481216,
      "step": 17675
    },
    {
      "epoch": 36.75675675675676,
      "grad_norm": 0.00011181944137206301,
      "learning_rate": 0.17719245040109313,
      "loss": 0.2712,
      "num_input_tokens_seen": 13485024,
      "step": 17680
    },
    {
      "epoch": 36.767151767151766,
      "grad_norm": 0.00016816981951706111,
      "learning_rate": 0.17713451944004271,
      "loss": 0.2437,
      "num_input_tokens_seen": 13488800,
      "step": 17685
    },
    {
      "epoch": 36.777546777546775,
      "grad_norm": 0.00011353805166436359,
      "learning_rate": 0.17707658429450843,
      "loss": 0.2882,
      "num_input_tokens_seen": 13492480,
      "step": 17690
    },
    {
      "epoch": 36.78794178794179,
      "grad_norm": 0.00016840647731442004,
      "learning_rate": 0.1770186449734245,
      "loss": 0.2569,
      "num_input_tokens_seen": 13496288,
      "step": 17695
    },
    {
      "epoch": 36.7983367983368,
      "grad_norm": 5.462926856125705e-05,
      "learning_rate": 0.17696070148572599,
      "loss": 0.2657,
      "num_input_tokens_seen": 13500224,
      "step": 17700
    },
    {
      "epoch": 36.80873180873181,
      "grad_norm": 0.0004431198467500508,
      "learning_rate": 0.17690275384034856,
      "loss": 0.2799,
      "num_input_tokens_seen": 13504096,
      "step": 17705
    },
    {
      "epoch": 36.81912681912682,
      "grad_norm": 0.00029139991966076195,
      "learning_rate": 0.17684480204622835,
      "loss": 0.2669,
      "num_input_tokens_seen": 13507968,
      "step": 17710
    },
    {
      "epoch": 36.82952182952183,
      "grad_norm": 0.0007000116747803986,
      "learning_rate": 0.1767868461123023,
      "loss": 0.2967,
      "num_input_tokens_seen": 13511968,
      "step": 17715
    },
    {
      "epoch": 36.83991683991684,
      "grad_norm": 9.667058475315571e-05,
      "learning_rate": 0.176728886047508,
      "loss": 0.277,
      "num_input_tokens_seen": 13515808,
      "step": 17720
    },
    {
      "epoch": 36.85031185031185,
      "grad_norm": 7.989753066794947e-05,
      "learning_rate": 0.17667092186078362,
      "loss": 0.2559,
      "num_input_tokens_seen": 13519488,
      "step": 17725
    },
    {
      "epoch": 36.86070686070686,
      "grad_norm": 0.0005120862042531371,
      "learning_rate": 0.17661295356106785,
      "loss": 0.2703,
      "num_input_tokens_seen": 13523552,
      "step": 17730
    },
    {
      "epoch": 36.87110187110187,
      "grad_norm": 0.00016344754840247333,
      "learning_rate": 0.1765549811573002,
      "loss": 0.2892,
      "num_input_tokens_seen": 13527360,
      "step": 17735
    },
    {
      "epoch": 36.88149688149688,
      "grad_norm": 0.00044120170059613883,
      "learning_rate": 0.17649700465842078,
      "loss": 0.2779,
      "num_input_tokens_seen": 13531072,
      "step": 17740
    },
    {
      "epoch": 36.891891891891895,
      "grad_norm": 8.828326099319384e-05,
      "learning_rate": 0.17643902407337023,
      "loss": 0.267,
      "num_input_tokens_seen": 13534944,
      "step": 17745
    },
    {
      "epoch": 36.9022869022869,
      "grad_norm": 0.00010145201667910442,
      "learning_rate": 0.17638103941108993,
      "loss": 0.27,
      "num_input_tokens_seen": 13538720,
      "step": 17750
    },
    {
      "epoch": 36.91268191268191,
      "grad_norm": 0.00012181472993688658,
      "learning_rate": 0.1763230506805218,
      "loss": 0.2643,
      "num_input_tokens_seen": 13542560,
      "step": 17755
    },
    {
      "epoch": 36.92307692307692,
      "grad_norm": 0.00019509706180542707,
      "learning_rate": 0.1762650578906085,
      "loss": 0.2939,
      "num_input_tokens_seen": 13546400,
      "step": 17760
    },
    {
      "epoch": 36.933471933471935,
      "grad_norm": 0.0002702220226638019,
      "learning_rate": 0.1762070610502932,
      "loss": 0.2613,
      "num_input_tokens_seen": 13550176,
      "step": 17765
    },
    {
      "epoch": 36.943866943866944,
      "grad_norm": 0.00010535618639551103,
      "learning_rate": 0.17614906016851975,
      "loss": 0.2758,
      "num_input_tokens_seen": 13553856,
      "step": 17770
    },
    {
      "epoch": 36.95426195426195,
      "grad_norm": 0.0003402753791306168,
      "learning_rate": 0.17609105525423258,
      "loss": 0.2585,
      "num_input_tokens_seen": 13557632,
      "step": 17775
    },
    {
      "epoch": 36.96465696465697,
      "grad_norm": 0.0002104471204802394,
      "learning_rate": 0.1760330463163768,
      "loss": 0.26,
      "num_input_tokens_seen": 13561216,
      "step": 17780
    },
    {
      "epoch": 36.975051975051976,
      "grad_norm": 6.175143789732829e-05,
      "learning_rate": 0.17597503336389816,
      "loss": 0.2654,
      "num_input_tokens_seen": 13564832,
      "step": 17785
    },
    {
      "epoch": 36.985446985446984,
      "grad_norm": 0.0004685286548919976,
      "learning_rate": 0.17591701640574298,
      "loss": 0.2647,
      "num_input_tokens_seen": 13568640,
      "step": 17790
    },
    {
      "epoch": 36.99584199584199,
      "grad_norm": 0.0005425813724286854,
      "learning_rate": 0.17585899545085815,
      "loss": 0.2816,
      "num_input_tokens_seen": 13572448,
      "step": 17795
    },
    {
      "epoch": 37.00623700623701,
      "grad_norm": 0.0003534541465342045,
      "learning_rate": 0.17580097050819124,
      "loss": 0.2689,
      "num_input_tokens_seen": 13576056,
      "step": 17800
    },
    {
      "epoch": 37.00623700623701,
      "eval_loss": 0.26566049456596375,
      "eval_runtime": 13.4131,
      "eval_samples_per_second": 63.818,
      "eval_steps_per_second": 15.955,
      "num_input_tokens_seen": 13576056,
      "step": 17800
    },
    {
      "epoch": 37.016632016632016,
      "grad_norm": 0.0001308080245507881,
      "learning_rate": 0.17574294158669046,
      "loss": 0.261,
      "num_input_tokens_seen": 13579832,
      "step": 17805
    },
    {
      "epoch": 37.027027027027025,
      "grad_norm": 0.0005428889999166131,
      "learning_rate": 0.17568490869530456,
      "loss": 0.2906,
      "num_input_tokens_seen": 13583608,
      "step": 17810
    },
    {
      "epoch": 37.03742203742204,
      "grad_norm": 0.0002659168385434896,
      "learning_rate": 0.17562687184298295,
      "loss": 0.2611,
      "num_input_tokens_seen": 13587448,
      "step": 17815
    },
    {
      "epoch": 37.04781704781705,
      "grad_norm": 0.0005051145562902093,
      "learning_rate": 0.1755688310386757,
      "loss": 0.2487,
      "num_input_tokens_seen": 13591288,
      "step": 17820
    },
    {
      "epoch": 37.05821205821206,
      "grad_norm": 0.0001192485069623217,
      "learning_rate": 0.17551078629133335,
      "loss": 0.2655,
      "num_input_tokens_seen": 13595032,
      "step": 17825
    },
    {
      "epoch": 37.06860706860707,
      "grad_norm": 3.7393612728919834e-05,
      "learning_rate": 0.17545273760990718,
      "loss": 0.2545,
      "num_input_tokens_seen": 13598712,
      "step": 17830
    },
    {
      "epoch": 37.07900207900208,
      "grad_norm": 0.00014857096539344639,
      "learning_rate": 0.17539468500334904,
      "loss": 0.2684,
      "num_input_tokens_seen": 13602552,
      "step": 17835
    },
    {
      "epoch": 37.08939708939709,
      "grad_norm": 0.0002657573204487562,
      "learning_rate": 0.17533662848061132,
      "loss": 0.2356,
      "num_input_tokens_seen": 13606264,
      "step": 17840
    },
    {
      "epoch": 37.0997920997921,
      "grad_norm": 0.0003399352717678994,
      "learning_rate": 0.1752785680506471,
      "loss": 0.2545,
      "num_input_tokens_seen": 13610040,
      "step": 17845
    },
    {
      "epoch": 37.11018711018711,
      "grad_norm": 0.00019116760813631117,
      "learning_rate": 0.17522050372241,
      "loss": 0.224,
      "num_input_tokens_seen": 13613880,
      "step": 17850
    },
    {
      "epoch": 37.12058212058212,
      "grad_norm": 0.0004366971261333674,
      "learning_rate": 0.17516243550485425,
      "loss": 0.2543,
      "num_input_tokens_seen": 13618008,
      "step": 17855
    },
    {
      "epoch": 37.13097713097713,
      "grad_norm": 0.00046624906826764345,
      "learning_rate": 0.17510436340693478,
      "loss": 0.2406,
      "num_input_tokens_seen": 13621816,
      "step": 17860
    },
    {
      "epoch": 37.141372141372145,
      "grad_norm": 0.0006874442333355546,
      "learning_rate": 0.175046287437607,
      "loss": 0.2839,
      "num_input_tokens_seen": 13625688,
      "step": 17865
    },
    {
      "epoch": 37.15176715176715,
      "grad_norm": 9.917110583046451e-05,
      "learning_rate": 0.17498820760582695,
      "loss": 0.2619,
      "num_input_tokens_seen": 13629688,
      "step": 17870
    },
    {
      "epoch": 37.16216216216216,
      "grad_norm": 4.814877684111707e-05,
      "learning_rate": 0.1749301239205512,
      "loss": 0.2626,
      "num_input_tokens_seen": 13633592,
      "step": 17875
    },
    {
      "epoch": 37.17255717255717,
      "grad_norm": 8.301712659886107e-05,
      "learning_rate": 0.1748720363907371,
      "loss": 0.2737,
      "num_input_tokens_seen": 13637400,
      "step": 17880
    },
    {
      "epoch": 37.182952182952185,
      "grad_norm": 0.00011215340055059642,
      "learning_rate": 0.17481394502534242,
      "loss": 0.2646,
      "num_input_tokens_seen": 13641144,
      "step": 17885
    },
    {
      "epoch": 37.19334719334719,
      "grad_norm": 3.621155337896198e-05,
      "learning_rate": 0.17475584983332562,
      "loss": 0.2749,
      "num_input_tokens_seen": 13644952,
      "step": 17890
    },
    {
      "epoch": 37.2037422037422,
      "grad_norm": 0.0002319800405530259,
      "learning_rate": 0.17469775082364558,
      "loss": 0.2659,
      "num_input_tokens_seen": 13648856,
      "step": 17895
    },
    {
      "epoch": 37.21413721413722,
      "grad_norm": 0.00035933632170781493,
      "learning_rate": 0.17463964800526205,
      "loss": 0.2818,
      "num_input_tokens_seen": 13652728,
      "step": 17900
    },
    {
      "epoch": 37.224532224532226,
      "grad_norm": 0.0003076719294767827,
      "learning_rate": 0.17458154138713522,
      "loss": 0.2774,
      "num_input_tokens_seen": 13656728,
      "step": 17905
    },
    {
      "epoch": 37.234927234927234,
      "grad_norm": 0.000195908680325374,
      "learning_rate": 0.17452343097822576,
      "loss": 0.2915,
      "num_input_tokens_seen": 13660504,
      "step": 17910
    },
    {
      "epoch": 37.24532224532224,
      "grad_norm": 0.00023033608158584684,
      "learning_rate": 0.17446531678749497,
      "loss": 0.2638,
      "num_input_tokens_seen": 13664216,
      "step": 17915
    },
    {
      "epoch": 37.25571725571726,
      "grad_norm": 0.00047455556341446936,
      "learning_rate": 0.17440719882390496,
      "loss": 0.2756,
      "num_input_tokens_seen": 13668088,
      "step": 17920
    },
    {
      "epoch": 37.266112266112266,
      "grad_norm": 0.00016654242062941194,
      "learning_rate": 0.17434907709641814,
      "loss": 0.2704,
      "num_input_tokens_seen": 13671928,
      "step": 17925
    },
    {
      "epoch": 37.276507276507274,
      "grad_norm": 0.0005451777833513916,
      "learning_rate": 0.17429095161399769,
      "loss": 0.2444,
      "num_input_tokens_seen": 13675736,
      "step": 17930
    },
    {
      "epoch": 37.28690228690229,
      "grad_norm": 0.00013748749915976077,
      "learning_rate": 0.1742328223856072,
      "loss": 0.2783,
      "num_input_tokens_seen": 13679608,
      "step": 17935
    },
    {
      "epoch": 37.2972972972973,
      "grad_norm": 0.00015887801419012249,
      "learning_rate": 0.174174689420211,
      "loss": 0.2742,
      "num_input_tokens_seen": 13683480,
      "step": 17940
    },
    {
      "epoch": 37.30769230769231,
      "grad_norm": 6.824556476203725e-05,
      "learning_rate": 0.1741165527267739,
      "loss": 0.2809,
      "num_input_tokens_seen": 13687320,
      "step": 17945
    },
    {
      "epoch": 37.318087318087315,
      "grad_norm": 0.00013770848454441875,
      "learning_rate": 0.17405841231426125,
      "loss": 0.2508,
      "num_input_tokens_seen": 13691000,
      "step": 17950
    },
    {
      "epoch": 37.32848232848233,
      "grad_norm": 0.0002024432469625026,
      "learning_rate": 0.1740002681916391,
      "loss": 0.2667,
      "num_input_tokens_seen": 13694808,
      "step": 17955
    },
    {
      "epoch": 37.33887733887734,
      "grad_norm": 0.0001446391543140635,
      "learning_rate": 0.17394212036787401,
      "loss": 0.2661,
      "num_input_tokens_seen": 13698584,
      "step": 17960
    },
    {
      "epoch": 37.34927234927235,
      "grad_norm": 0.000454866501968354,
      "learning_rate": 0.1738839688519331,
      "loss": 0.2843,
      "num_input_tokens_seen": 13702360,
      "step": 17965
    },
    {
      "epoch": 37.35966735966736,
      "grad_norm": 6.412077345885336e-05,
      "learning_rate": 0.17382581365278402,
      "loss": 0.2808,
      "num_input_tokens_seen": 13706104,
      "step": 17970
    },
    {
      "epoch": 37.37006237006237,
      "grad_norm": 0.0001545573177281767,
      "learning_rate": 0.17376765477939507,
      "loss": 0.2788,
      "num_input_tokens_seen": 13709880,
      "step": 17975
    },
    {
      "epoch": 37.38045738045738,
      "grad_norm": 0.00026132227503694594,
      "learning_rate": 0.1737094922407351,
      "loss": 0.2689,
      "num_input_tokens_seen": 13713688,
      "step": 17980
    },
    {
      "epoch": 37.39085239085239,
      "grad_norm": 0.00041704546310938895,
      "learning_rate": 0.1736513260457734,
      "loss": 0.2739,
      "num_input_tokens_seen": 13717464,
      "step": 17985
    },
    {
      "epoch": 37.4012474012474,
      "grad_norm": 7.604605343658477e-05,
      "learning_rate": 0.17359315620348006,
      "loss": 0.2457,
      "num_input_tokens_seen": 13721176,
      "step": 17990
    },
    {
      "epoch": 37.41164241164241,
      "grad_norm": 0.00019248152966611087,
      "learning_rate": 0.17353498272282547,
      "loss": 0.2463,
      "num_input_tokens_seen": 13724952,
      "step": 17995
    },
    {
      "epoch": 37.42203742203742,
      "grad_norm": 0.00016061661881394684,
      "learning_rate": 0.17347680561278087,
      "loss": 0.2675,
      "num_input_tokens_seen": 13728696,
      "step": 18000
    },
    {
      "epoch": 37.42203742203742,
      "eval_loss": 0.24717116355895996,
      "eval_runtime": 13.4109,
      "eval_samples_per_second": 63.829,
      "eval_steps_per_second": 15.957,
      "num_input_tokens_seen": 13728696,
      "step": 18000
    },
    {
      "epoch": 37.432432432432435,
      "grad_norm": 0.00032401090720668435,
      "learning_rate": 0.1734186248823178,
      "loss": 0.2756,
      "num_input_tokens_seen": 13732408,
      "step": 18005
    },
    {
      "epoch": 37.44282744282744,
      "grad_norm": 3.154646037728526e-05,
      "learning_rate": 0.17336044054040844,
      "loss": 0.2817,
      "num_input_tokens_seen": 13736312,
      "step": 18010
    },
    {
      "epoch": 37.45322245322245,
      "grad_norm": 0.00012456002878025174,
      "learning_rate": 0.1733022525960256,
      "loss": 0.2742,
      "num_input_tokens_seen": 13740184,
      "step": 18015
    },
    {
      "epoch": 37.46361746361746,
      "grad_norm": 0.0002284859656356275,
      "learning_rate": 0.1732440610581426,
      "loss": 0.2725,
      "num_input_tokens_seen": 13743992,
      "step": 18020
    },
    {
      "epoch": 37.474012474012476,
      "grad_norm": 0.00026105876895599067,
      "learning_rate": 0.17318586593573326,
      "loss": 0.2755,
      "num_input_tokens_seen": 13747672,
      "step": 18025
    },
    {
      "epoch": 37.484407484407484,
      "grad_norm": 0.00011538234684849158,
      "learning_rate": 0.17312766723777204,
      "loss": 0.2685,
      "num_input_tokens_seen": 13751544,
      "step": 18030
    },
    {
      "epoch": 37.49480249480249,
      "grad_norm": 0.00030437938403338194,
      "learning_rate": 0.1730694649732339,
      "loss": 0.264,
      "num_input_tokens_seen": 13755320,
      "step": 18035
    },
    {
      "epoch": 37.50519750519751,
      "grad_norm": 3.2187650504056364e-05,
      "learning_rate": 0.17301125915109428,
      "loss": 0.2754,
      "num_input_tokens_seen": 13759288,
      "step": 18040
    },
    {
      "epoch": 37.515592515592516,
      "grad_norm": 0.0005967272445559502,
      "learning_rate": 0.17295304978032938,
      "loss": 0.2377,
      "num_input_tokens_seen": 13763128,
      "step": 18045
    },
    {
      "epoch": 37.525987525987524,
      "grad_norm": 0.0003936718567274511,
      "learning_rate": 0.17289483686991577,
      "loss": 0.2502,
      "num_input_tokens_seen": 13766904,
      "step": 18050
    },
    {
      "epoch": 37.53638253638254,
      "grad_norm": 0.00026558240642771125,
      "learning_rate": 0.1728366204288306,
      "loss": 0.2691,
      "num_input_tokens_seen": 13770776,
      "step": 18055
    },
    {
      "epoch": 37.54677754677755,
      "grad_norm": 0.00019525873358361423,
      "learning_rate": 0.17277840046605153,
      "loss": 0.2795,
      "num_input_tokens_seen": 13774712,
      "step": 18060
    },
    {
      "epoch": 37.55717255717256,
      "grad_norm": 0.00030830211471766233,
      "learning_rate": 0.17272017699055686,
      "loss": 0.266,
      "num_input_tokens_seen": 13778520,
      "step": 18065
    },
    {
      "epoch": 37.567567567567565,
      "grad_norm": 3.836326868622564e-05,
      "learning_rate": 0.17266195001132542,
      "loss": 0.2665,
      "num_input_tokens_seen": 13782200,
      "step": 18070
    },
    {
      "epoch": 37.57796257796258,
      "grad_norm": 0.00025523462682031095,
      "learning_rate": 0.17260371953733647,
      "loss": 0.2709,
      "num_input_tokens_seen": 13786104,
      "step": 18075
    },
    {
      "epoch": 37.58835758835759,
      "grad_norm": 0.00011523273860802874,
      "learning_rate": 0.1725454855775699,
      "loss": 0.2586,
      "num_input_tokens_seen": 13789784,
      "step": 18080
    },
    {
      "epoch": 37.5987525987526,
      "grad_norm": 5.492265700013377e-05,
      "learning_rate": 0.17248724814100616,
      "loss": 0.2727,
      "num_input_tokens_seen": 13793400,
      "step": 18085
    },
    {
      "epoch": 37.60914760914761,
      "grad_norm": 0.00020145783491898328,
      "learning_rate": 0.17242900723662619,
      "loss": 0.2534,
      "num_input_tokens_seen": 13797176,
      "step": 18090
    },
    {
      "epoch": 37.61954261954262,
      "grad_norm": 0.00022032538254279643,
      "learning_rate": 0.1723707628734114,
      "loss": 0.2585,
      "num_input_tokens_seen": 13800984,
      "step": 18095
    },
    {
      "epoch": 37.62993762993763,
      "grad_norm": 0.00030590148526243865,
      "learning_rate": 0.1723125150603438,
      "loss": 0.2725,
      "num_input_tokens_seen": 13804792,
      "step": 18100
    },
    {
      "epoch": 37.64033264033264,
      "grad_norm": 0.00034034260897897184,
      "learning_rate": 0.1722542638064061,
      "loss": 0.2597,
      "num_input_tokens_seen": 13808664,
      "step": 18105
    },
    {
      "epoch": 37.65072765072765,
      "grad_norm": 9.597495227353647e-05,
      "learning_rate": 0.17219600912058117,
      "loss": 0.229,
      "num_input_tokens_seen": 13812408,
      "step": 18110
    },
    {
      "epoch": 37.66112266112266,
      "grad_norm": 0.0004968281136825681,
      "learning_rate": 0.17213775101185272,
      "loss": 0.2525,
      "num_input_tokens_seen": 13816248,
      "step": 18115
    },
    {
      "epoch": 37.67151767151767,
      "grad_norm": 0.00032090532477013767,
      "learning_rate": 0.17207948948920485,
      "loss": 0.3041,
      "num_input_tokens_seen": 13820024,
      "step": 18120
    },
    {
      "epoch": 37.681912681912685,
      "grad_norm": 0.00036634900607168674,
      "learning_rate": 0.17202122456162228,
      "loss": 0.2601,
      "num_input_tokens_seen": 13823896,
      "step": 18125
    },
    {
      "epoch": 37.69230769230769,
      "grad_norm": 0.00010695265518734232,
      "learning_rate": 0.17196295623809013,
      "loss": 0.2793,
      "num_input_tokens_seen": 13827704,
      "step": 18130
    },
    {
      "epoch": 37.7027027027027,
      "grad_norm": 0.00015288103895727545,
      "learning_rate": 0.1719046845275941,
      "loss": 0.2727,
      "num_input_tokens_seen": 13831480,
      "step": 18135
    },
    {
      "epoch": 37.71309771309771,
      "grad_norm": 5.0813450798159465e-05,
      "learning_rate": 0.17184640943912044,
      "loss": 0.2649,
      "num_input_tokens_seen": 13835384,
      "step": 18140
    },
    {
      "epoch": 37.723492723492726,
      "grad_norm": 9.579612378729507e-05,
      "learning_rate": 0.1717881309816559,
      "loss": 0.2863,
      "num_input_tokens_seen": 13839128,
      "step": 18145
    },
    {
      "epoch": 37.733887733887734,
      "grad_norm": 0.0004308338393457234,
      "learning_rate": 0.1717298491641878,
      "loss": 0.2754,
      "num_input_tokens_seen": 13842968,
      "step": 18150
    },
    {
      "epoch": 37.74428274428274,
      "grad_norm": 0.0004642101703211665,
      "learning_rate": 0.17167156399570385,
      "loss": 0.2555,
      "num_input_tokens_seen": 13846776,
      "step": 18155
    },
    {
      "epoch": 37.75467775467776,
      "grad_norm": 0.0002685229410417378,
      "learning_rate": 0.17161327548519242,
      "loss": 0.27,
      "num_input_tokens_seen": 13850776,
      "step": 18160
    },
    {
      "epoch": 37.765072765072766,
      "grad_norm": 0.00011166080366820097,
      "learning_rate": 0.1715549836416423,
      "loss": 0.2842,
      "num_input_tokens_seen": 13854744,
      "step": 18165
    },
    {
      "epoch": 37.775467775467774,
      "grad_norm": 0.0006886075716465712,
      "learning_rate": 0.17149668847404279,
      "loss": 0.2805,
      "num_input_tokens_seen": 13858488,
      "step": 18170
    },
    {
      "epoch": 37.78586278586278,
      "grad_norm": 9.7030853794422e-05,
      "learning_rate": 0.1714383899913838,
      "loss": 0.251,
      "num_input_tokens_seen": 13862360,
      "step": 18175
    },
    {
      "epoch": 37.7962577962578,
      "grad_norm": 0.0003226956177968532,
      "learning_rate": 0.17138008820265563,
      "loss": 0.2512,
      "num_input_tokens_seen": 13866072,
      "step": 18180
    },
    {
      "epoch": 37.80665280665281,
      "grad_norm": 0.0001526285195723176,
      "learning_rate": 0.17132178311684917,
      "loss": 0.251,
      "num_input_tokens_seen": 13869816,
      "step": 18185
    },
    {
      "epoch": 37.817047817047815,
      "grad_norm": 0.00013543125533033162,
      "learning_rate": 0.1712634747429559,
      "loss": 0.2374,
      "num_input_tokens_seen": 13873624,
      "step": 18190
    },
    {
      "epoch": 37.82744282744283,
      "grad_norm": 0.0001464757660869509,
      "learning_rate": 0.17120516308996753,
      "loss": 0.2742,
      "num_input_tokens_seen": 13877496,
      "step": 18195
    },
    {
      "epoch": 37.83783783783784,
      "grad_norm": 0.00010562594252405688,
      "learning_rate": 0.17114684816687653,
      "loss": 0.2572,
      "num_input_tokens_seen": 13881368,
      "step": 18200
    },
    {
      "epoch": 37.83783783783784,
      "eval_loss": 0.2514818608760834,
      "eval_runtime": 13.4284,
      "eval_samples_per_second": 63.746,
      "eval_steps_per_second": 15.936,
      "num_input_tokens_seen": 13881368,
      "step": 18200
    },
    {
      "epoch": 37.84823284823285,
      "grad_norm": 9.730438614496961e-05,
      "learning_rate": 0.17108852998267585,
      "loss": 0.2931,
      "num_input_tokens_seen": 13885176,
      "step": 18205
    },
    {
      "epoch": 37.858627858627855,
      "grad_norm": 0.00015595727018080652,
      "learning_rate": 0.17103020854635878,
      "loss": 0.2795,
      "num_input_tokens_seen": 13889048,
      "step": 18210
    },
    {
      "epoch": 37.86902286902287,
      "grad_norm": 0.0006141007761470973,
      "learning_rate": 0.1709718838669193,
      "loss": 0.2826,
      "num_input_tokens_seen": 13892792,
      "step": 18215
    },
    {
      "epoch": 37.87941787941788,
      "grad_norm": 0.00013923853111919016,
      "learning_rate": 0.17091355595335173,
      "loss": 0.2505,
      "num_input_tokens_seen": 13896696,
      "step": 18220
    },
    {
      "epoch": 37.88981288981289,
      "grad_norm": 0.00017420900985598564,
      "learning_rate": 0.17085522481465107,
      "loss": 0.2526,
      "num_input_tokens_seen": 13900504,
      "step": 18225
    },
    {
      "epoch": 37.9002079002079,
      "grad_norm": 0.00016382864851038903,
      "learning_rate": 0.17079689045981264,
      "loss": 0.2481,
      "num_input_tokens_seen": 13904312,
      "step": 18230
    },
    {
      "epoch": 37.91060291060291,
      "grad_norm": 0.00036020600236952305,
      "learning_rate": 0.17073855289783238,
      "loss": 0.2947,
      "num_input_tokens_seen": 13908152,
      "step": 18235
    },
    {
      "epoch": 37.92099792099792,
      "grad_norm": 0.0004964638501405716,
      "learning_rate": 0.1706802121377066,
      "loss": 0.2735,
      "num_input_tokens_seen": 13911832,
      "step": 18240
    },
    {
      "epoch": 37.931392931392935,
      "grad_norm": 0.00011524632282089442,
      "learning_rate": 0.17062186818843225,
      "loss": 0.2824,
      "num_input_tokens_seen": 13915544,
      "step": 18245
    },
    {
      "epoch": 37.94178794178794,
      "grad_norm": 0.00013317755656316876,
      "learning_rate": 0.17056352105900668,
      "loss": 0.2726,
      "num_input_tokens_seen": 13919288,
      "step": 18250
    },
    {
      "epoch": 37.95218295218295,
      "grad_norm": 0.00017390323046129197,
      "learning_rate": 0.17050517075842772,
      "loss": 0.2711,
      "num_input_tokens_seen": 13923256,
      "step": 18255
    },
    {
      "epoch": 37.96257796257796,
      "grad_norm": 5.497239180840552e-05,
      "learning_rate": 0.17044681729569375,
      "loss": 0.2592,
      "num_input_tokens_seen": 13926936,
      "step": 18260
    },
    {
      "epoch": 37.972972972972975,
      "grad_norm": 0.00011403862299630418,
      "learning_rate": 0.17038846067980365,
      "loss": 0.2657,
      "num_input_tokens_seen": 13930776,
      "step": 18265
    },
    {
      "epoch": 37.983367983367984,
      "grad_norm": 0.0002356619806960225,
      "learning_rate": 0.17033010091975664,
      "loss": 0.2551,
      "num_input_tokens_seen": 13934424,
      "step": 18270
    },
    {
      "epoch": 37.99376299376299,
      "grad_norm": 0.00011265333159826696,
      "learning_rate": 0.17027173802455262,
      "loss": 0.2585,
      "num_input_tokens_seen": 13938200,
      "step": 18275
    },
    {
      "epoch": 38.00415800415801,
      "grad_norm": 0.00018976478895638138,
      "learning_rate": 0.1702133720031918,
      "loss": 0.2567,
      "num_input_tokens_seen": 13942000,
      "step": 18280
    },
    {
      "epoch": 38.014553014553016,
      "grad_norm": 8.800016075838357e-05,
      "learning_rate": 0.17015500286467503,
      "loss": 0.2158,
      "num_input_tokens_seen": 13945744,
      "step": 18285
    },
    {
      "epoch": 38.024948024948024,
      "grad_norm": 0.00018180751067120582,
      "learning_rate": 0.17009663061800354,
      "loss": 0.2659,
      "num_input_tokens_seen": 13949520,
      "step": 18290
    },
    {
      "epoch": 38.03534303534303,
      "grad_norm": 0.0003712528559844941,
      "learning_rate": 0.17003825527217903,
      "loss": 0.2681,
      "num_input_tokens_seen": 13953328,
      "step": 18295
    },
    {
      "epoch": 38.04573804573805,
      "grad_norm": 0.00019450522086117417,
      "learning_rate": 0.16997987683620377,
      "loss": 0.2645,
      "num_input_tokens_seen": 13957264,
      "step": 18300
    },
    {
      "epoch": 38.056133056133056,
      "grad_norm": 0.00012056219566147774,
      "learning_rate": 0.16992149531908043,
      "loss": 0.2691,
      "num_input_tokens_seen": 13961136,
      "step": 18305
    },
    {
      "epoch": 38.066528066528065,
      "grad_norm": 0.0002713422290980816,
      "learning_rate": 0.16986311072981214,
      "loss": 0.2804,
      "num_input_tokens_seen": 13964848,
      "step": 18310
    },
    {
      "epoch": 38.07692307692308,
      "grad_norm": 0.0002110823552357033,
      "learning_rate": 0.16980472307740255,
      "loss": 0.2649,
      "num_input_tokens_seen": 13968560,
      "step": 18315
    },
    {
      "epoch": 38.08731808731809,
      "grad_norm": 0.00011486118455650285,
      "learning_rate": 0.1697463323708558,
      "loss": 0.265,
      "num_input_tokens_seen": 13972336,
      "step": 18320
    },
    {
      "epoch": 38.0977130977131,
      "grad_norm": 0.0002966140164062381,
      "learning_rate": 0.16968793861917641,
      "loss": 0.2651,
      "num_input_tokens_seen": 13976144,
      "step": 18325
    },
    {
      "epoch": 38.108108108108105,
      "grad_norm": 0.0004858723550569266,
      "learning_rate": 0.16962954183136952,
      "loss": 0.282,
      "num_input_tokens_seen": 13979984,
      "step": 18330
    },
    {
      "epoch": 38.11850311850312,
      "grad_norm": 0.00017579954874236137,
      "learning_rate": 0.16957114201644058,
      "loss": 0.2501,
      "num_input_tokens_seen": 13983920,
      "step": 18335
    },
    {
      "epoch": 38.12889812889813,
      "grad_norm": 1.1696202818711754e-05,
      "learning_rate": 0.16951273918339563,
      "loss": 0.2559,
      "num_input_tokens_seen": 13987792,
      "step": 18340
    },
    {
      "epoch": 38.13929313929314,
      "grad_norm": 0.00018843628640752286,
      "learning_rate": 0.16945433334124105,
      "loss": 0.2736,
      "num_input_tokens_seen": 13991472,
      "step": 18345
    },
    {
      "epoch": 38.14968814968815,
      "grad_norm": 0.00021506579651031643,
      "learning_rate": 0.1693959244989838,
      "loss": 0.2422,
      "num_input_tokens_seen": 13995216,
      "step": 18350
    },
    {
      "epoch": 38.16008316008316,
      "grad_norm": 0.0003718194493558258,
      "learning_rate": 0.16933751266563127,
      "loss": 0.2792,
      "num_input_tokens_seen": 13998992,
      "step": 18355
    },
    {
      "epoch": 38.17047817047817,
      "grad_norm": 0.00024612279958091676,
      "learning_rate": 0.16927909785019118,
      "loss": 0.2697,
      "num_input_tokens_seen": 14002896,
      "step": 18360
    },
    {
      "epoch": 38.18087318087318,
      "grad_norm": 0.00029218298732303083,
      "learning_rate": 0.169220680061672,
      "loss": 0.2789,
      "num_input_tokens_seen": 14006672,
      "step": 18365
    },
    {
      "epoch": 38.19126819126819,
      "grad_norm": 0.0004712914233095944,
      "learning_rate": 0.16916225930908244,
      "loss": 0.247,
      "num_input_tokens_seen": 14010320,
      "step": 18370
    },
    {
      "epoch": 38.2016632016632,
      "grad_norm": 0.00022511265706270933,
      "learning_rate": 0.16910383560143163,
      "loss": 0.2564,
      "num_input_tokens_seen": 14014160,
      "step": 18375
    },
    {
      "epoch": 38.21205821205821,
      "grad_norm": 0.0003480313753243536,
      "learning_rate": 0.16904540894772935,
      "loss": 0.2743,
      "num_input_tokens_seen": 14018032,
      "step": 18380
    },
    {
      "epoch": 38.222453222453225,
      "grad_norm": 0.00028071028646081686,
      "learning_rate": 0.16898697935698562,
      "loss": 0.2459,
      "num_input_tokens_seen": 14021872,
      "step": 18385
    },
    {
      "epoch": 38.232848232848234,
      "grad_norm": 0.00021995829592924565,
      "learning_rate": 0.1689285468382111,
      "loss": 0.2616,
      "num_input_tokens_seen": 14025872,
      "step": 18390
    },
    {
      "epoch": 38.24324324324324,
      "grad_norm": 4.734143294626847e-05,
      "learning_rate": 0.16887011140041677,
      "loss": 0.2651,
      "num_input_tokens_seen": 14029776,
      "step": 18395
    },
    {
      "epoch": 38.25363825363825,
      "grad_norm": 0.0006571648991666734,
      "learning_rate": 0.1688116730526141,
      "loss": 0.2781,
      "num_input_tokens_seen": 14033616,
      "step": 18400
    },
    {
      "epoch": 38.25363825363825,
      "eval_loss": 0.26234912872314453,
      "eval_runtime": 13.4397,
      "eval_samples_per_second": 63.692,
      "eval_steps_per_second": 15.923,
      "num_input_tokens_seen": 14033616,
      "step": 18400
    },
    {
      "epoch": 38.264033264033266,
      "grad_norm": 0.0002502047864254564,
      "learning_rate": 0.1687532318038151,
      "loss": 0.2762,
      "num_input_tokens_seen": 14037392,
      "step": 18405
    },
    {
      "epoch": 38.274428274428274,
      "grad_norm": 0.00025649057351984084,
      "learning_rate": 0.16869478766303206,
      "loss": 0.2533,
      "num_input_tokens_seen": 14041264,
      "step": 18410
    },
    {
      "epoch": 38.28482328482328,
      "grad_norm": 9.528319060336798e-05,
      "learning_rate": 0.16863634063927788,
      "loss": 0.2458,
      "num_input_tokens_seen": 14045168,
      "step": 18415
    },
    {
      "epoch": 38.2952182952183,
      "grad_norm": 4.3037463910877705e-05,
      "learning_rate": 0.16857789074156568,
      "loss": 0.2758,
      "num_input_tokens_seen": 14048944,
      "step": 18420
    },
    {
      "epoch": 38.305613305613306,
      "grad_norm": 8.081365376710892e-05,
      "learning_rate": 0.16851943797890928,
      "loss": 0.2961,
      "num_input_tokens_seen": 14052784,
      "step": 18425
    },
    {
      "epoch": 38.316008316008315,
      "grad_norm": 0.00034480728209018707,
      "learning_rate": 0.16846098236032284,
      "loss": 0.2726,
      "num_input_tokens_seen": 14056592,
      "step": 18430
    },
    {
      "epoch": 38.32640332640332,
      "grad_norm": 3.981667759944685e-05,
      "learning_rate": 0.16840252389482097,
      "loss": 0.2748,
      "num_input_tokens_seen": 14060208,
      "step": 18435
    },
    {
      "epoch": 38.33679833679834,
      "grad_norm": 0.00014087762974668294,
      "learning_rate": 0.16834406259141857,
      "loss": 0.2682,
      "num_input_tokens_seen": 14063984,
      "step": 18440
    },
    {
      "epoch": 38.34719334719335,
      "grad_norm": 2.8825385015807115e-05,
      "learning_rate": 0.16828559845913124,
      "loss": 0.2487,
      "num_input_tokens_seen": 14067696,
      "step": 18445
    },
    {
      "epoch": 38.357588357588355,
      "grad_norm": 9.348901949124411e-05,
      "learning_rate": 0.16822713150697488,
      "loss": 0.2818,
      "num_input_tokens_seen": 14071504,
      "step": 18450
    },
    {
      "epoch": 38.36798336798337,
      "grad_norm": 0.00020476299687288702,
      "learning_rate": 0.16816866174396575,
      "loss": 0.2428,
      "num_input_tokens_seen": 14075344,
      "step": 18455
    },
    {
      "epoch": 38.37837837837838,
      "grad_norm": 5.3525480325333774e-05,
      "learning_rate": 0.16811018917912057,
      "loss": 0.2606,
      "num_input_tokens_seen": 14079152,
      "step": 18460
    },
    {
      "epoch": 38.38877338877339,
      "grad_norm": 0.0002666098007466644,
      "learning_rate": 0.16805171382145673,
      "loss": 0.2703,
      "num_input_tokens_seen": 14082928,
      "step": 18465
    },
    {
      "epoch": 38.3991683991684,
      "grad_norm": 0.0005928517202846706,
      "learning_rate": 0.16799323567999175,
      "loss": 0.2723,
      "num_input_tokens_seen": 14086704,
      "step": 18470
    },
    {
      "epoch": 38.40956340956341,
      "grad_norm": 0.00010158977238461375,
      "learning_rate": 0.16793475476374367,
      "loss": 0.269,
      "num_input_tokens_seen": 14090512,
      "step": 18475
    },
    {
      "epoch": 38.41995841995842,
      "grad_norm": 0.0005319610936567187,
      "learning_rate": 0.1678762710817311,
      "loss": 0.2726,
      "num_input_tokens_seen": 14094160,
      "step": 18480
    },
    {
      "epoch": 38.43035343035343,
      "grad_norm": 5.5819586123107e-05,
      "learning_rate": 0.1678177846429728,
      "loss": 0.2588,
      "num_input_tokens_seen": 14098000,
      "step": 18485
    },
    {
      "epoch": 38.44074844074844,
      "grad_norm": 0.0001422505738446489,
      "learning_rate": 0.16775929545648827,
      "loss": 0.262,
      "num_input_tokens_seen": 14101616,
      "step": 18490
    },
    {
      "epoch": 38.45114345114345,
      "grad_norm": 9.092781692743301e-05,
      "learning_rate": 0.16770080353129715,
      "loss": 0.2825,
      "num_input_tokens_seen": 14105328,
      "step": 18495
    },
    {
      "epoch": 38.46153846153846,
      "grad_norm": 0.00020319767645560205,
      "learning_rate": 0.16764230887641968,
      "loss": 0.2583,
      "num_input_tokens_seen": 14109168,
      "step": 18500
    },
    {
      "epoch": 38.471933471933475,
      "grad_norm": 7.987016579136252e-05,
      "learning_rate": 0.1675838115008765,
      "loss": 0.2711,
      "num_input_tokens_seen": 14113040,
      "step": 18505
    },
    {
      "epoch": 38.482328482328484,
      "grad_norm": 3.809152985922992e-05,
      "learning_rate": 0.1675253114136886,
      "loss": 0.2376,
      "num_input_tokens_seen": 14116944,
      "step": 18510
    },
    {
      "epoch": 38.49272349272349,
      "grad_norm": 3.1968440453056246e-05,
      "learning_rate": 0.16746680862387747,
      "loss": 0.2897,
      "num_input_tokens_seen": 14120688,
      "step": 18515
    },
    {
      "epoch": 38.5031185031185,
      "grad_norm": 0.0002977905096486211,
      "learning_rate": 0.16740830314046493,
      "loss": 0.2704,
      "num_input_tokens_seen": 14124560,
      "step": 18520
    },
    {
      "epoch": 38.513513513513516,
      "grad_norm": 4.5603363105328754e-05,
      "learning_rate": 0.1673497949724733,
      "loss": 0.2873,
      "num_input_tokens_seen": 14128400,
      "step": 18525
    },
    {
      "epoch": 38.523908523908524,
      "grad_norm": 0.00021074705000501126,
      "learning_rate": 0.16729128412892522,
      "loss": 0.2898,
      "num_input_tokens_seen": 14132208,
      "step": 18530
    },
    {
      "epoch": 38.53430353430353,
      "grad_norm": 6.72453606966883e-05,
      "learning_rate": 0.16723277061884384,
      "loss": 0.2791,
      "num_input_tokens_seen": 14136016,
      "step": 18535
    },
    {
      "epoch": 38.54469854469855,
      "grad_norm": 0.00014155970711726695,
      "learning_rate": 0.16717425445125267,
      "loss": 0.2543,
      "num_input_tokens_seen": 14139856,
      "step": 18540
    },
    {
      "epoch": 38.555093555093556,
      "grad_norm": 2.937451245088596e-05,
      "learning_rate": 0.16711573563517565,
      "loss": 0.2714,
      "num_input_tokens_seen": 14143696,
      "step": 18545
    },
    {
      "epoch": 38.565488565488565,
      "grad_norm": 4.8365214752266183e-05,
      "learning_rate": 0.1670572141796371,
      "loss": 0.2567,
      "num_input_tokens_seen": 14147600,
      "step": 18550
    },
    {
      "epoch": 38.57588357588357,
      "grad_norm": 0.00045258383033797145,
      "learning_rate": 0.16699869009366175,
      "loss": 0.2781,
      "num_input_tokens_seen": 14151440,
      "step": 18555
    },
    {
      "epoch": 38.58627858627859,
      "grad_norm": 0.00012728873116429895,
      "learning_rate": 0.1669401633862748,
      "loss": 0.2709,
      "num_input_tokens_seen": 14155312,
      "step": 18560
    },
    {
      "epoch": 38.5966735966736,
      "grad_norm": 0.00024172550183720887,
      "learning_rate": 0.16688163406650178,
      "loss": 0.2724,
      "num_input_tokens_seen": 14159152,
      "step": 18565
    },
    {
      "epoch": 38.607068607068605,
      "grad_norm": 0.0006905028130859137,
      "learning_rate": 0.1668231021433686,
      "loss": 0.2677,
      "num_input_tokens_seen": 14162992,
      "step": 18570
    },
    {
      "epoch": 38.61746361746362,
      "grad_norm": 0.0003708757576532662,
      "learning_rate": 0.1667645676259017,
      "loss": 0.278,
      "num_input_tokens_seen": 14166864,
      "step": 18575
    },
    {
      "epoch": 38.62785862785863,
      "grad_norm": 0.00017842563102021813,
      "learning_rate": 0.1667060305231277,
      "loss": 0.2631,
      "num_input_tokens_seen": 14170640,
      "step": 18580
    },
    {
      "epoch": 38.63825363825364,
      "grad_norm": 0.0001081644804798998,
      "learning_rate": 0.16664749084407396,
      "loss": 0.2539,
      "num_input_tokens_seen": 14174384,
      "step": 18585
    },
    {
      "epoch": 38.648648648648646,
      "grad_norm": 0.00019887782400473952,
      "learning_rate": 0.16658894859776788,
      "loss": 0.247,
      "num_input_tokens_seen": 14178000,
      "step": 18590
    },
    {
      "epoch": 38.65904365904366,
      "grad_norm": 0.00016245538427028805,
      "learning_rate": 0.16653040379323752,
      "loss": 0.2697,
      "num_input_tokens_seen": 14181840,
      "step": 18595
    },
    {
      "epoch": 38.66943866943867,
      "grad_norm": 0.00024234774173237383,
      "learning_rate": 0.16647185643951107,
      "loss": 0.2692,
      "num_input_tokens_seen": 14185616,
      "step": 18600
    },
    {
      "epoch": 38.66943866943867,
      "eval_loss": 0.24956050515174866,
      "eval_runtime": 13.4245,
      "eval_samples_per_second": 63.764,
      "eval_steps_per_second": 15.941,
      "num_input_tokens_seen": 14185616,
      "step": 18600
    },
    {
      "epoch": 38.67983367983368,
      "grad_norm": 0.0001595732319401577,
      "learning_rate": 0.1664133065456174,
      "loss": 0.2644,
      "num_input_tokens_seen": 14189456,
      "step": 18605
    },
    {
      "epoch": 38.69022869022869,
      "grad_norm": 0.00010833367559826002,
      "learning_rate": 0.1663547541205856,
      "loss": 0.2635,
      "num_input_tokens_seen": 14193168,
      "step": 18610
    },
    {
      "epoch": 38.7006237006237,
      "grad_norm": 0.0004917460610158741,
      "learning_rate": 0.16629619917344518,
      "loss": 0.2603,
      "num_input_tokens_seen": 14197008,
      "step": 18615
    },
    {
      "epoch": 38.71101871101871,
      "grad_norm": 0.00015672293375246227,
      "learning_rate": 0.16623764171322605,
      "loss": 0.2778,
      "num_input_tokens_seen": 14200912,
      "step": 18620
    },
    {
      "epoch": 38.72141372141372,
      "grad_norm": 0.00013964631943963468,
      "learning_rate": 0.1661790817489585,
      "loss": 0.2879,
      "num_input_tokens_seen": 14204816,
      "step": 18625
    },
    {
      "epoch": 38.731808731808734,
      "grad_norm": 0.0003688719589263201,
      "learning_rate": 0.16612051928967328,
      "loss": 0.2748,
      "num_input_tokens_seen": 14208688,
      "step": 18630
    },
    {
      "epoch": 38.74220374220374,
      "grad_norm": 0.00041695183608680964,
      "learning_rate": 0.16606195434440138,
      "loss": 0.2704,
      "num_input_tokens_seen": 14212528,
      "step": 18635
    },
    {
      "epoch": 38.75259875259875,
      "grad_norm": 0.00018519068544264883,
      "learning_rate": 0.16600338692217426,
      "loss": 0.2593,
      "num_input_tokens_seen": 14216240,
      "step": 18640
    },
    {
      "epoch": 38.762993762993766,
      "grad_norm": 0.000599623192101717,
      "learning_rate": 0.16594481703202374,
      "loss": 0.28,
      "num_input_tokens_seen": 14220144,
      "step": 18645
    },
    {
      "epoch": 38.773388773388774,
      "grad_norm": 0.0002322217624168843,
      "learning_rate": 0.1658862446829821,
      "loss": 0.2562,
      "num_input_tokens_seen": 14223984,
      "step": 18650
    },
    {
      "epoch": 38.78378378378378,
      "grad_norm": 0.0001424806978320703,
      "learning_rate": 0.16582766988408187,
      "loss": 0.2689,
      "num_input_tokens_seen": 14227856,
      "step": 18655
    },
    {
      "epoch": 38.79417879417879,
      "grad_norm": 3.8309521187329665e-05,
      "learning_rate": 0.16576909264435608,
      "loss": 0.2641,
      "num_input_tokens_seen": 14231728,
      "step": 18660
    },
    {
      "epoch": 38.804573804573806,
      "grad_norm": 6.855159881524742e-05,
      "learning_rate": 0.16571051297283798,
      "loss": 0.2592,
      "num_input_tokens_seen": 14235664,
      "step": 18665
    },
    {
      "epoch": 38.814968814968815,
      "grad_norm": 0.00028376287082210183,
      "learning_rate": 0.16565193087856137,
      "loss": 0.2727,
      "num_input_tokens_seen": 14239408,
      "step": 18670
    },
    {
      "epoch": 38.82536382536382,
      "grad_norm": 0.0004970021545886993,
      "learning_rate": 0.16559334637056033,
      "loss": 0.2695,
      "num_input_tokens_seen": 14243216,
      "step": 18675
    },
    {
      "epoch": 38.83575883575884,
      "grad_norm": 0.00023062255058903247,
      "learning_rate": 0.16553475945786933,
      "loss": 0.2775,
      "num_input_tokens_seen": 14247312,
      "step": 18680
    },
    {
      "epoch": 38.84615384615385,
      "grad_norm": 0.0006265403935685754,
      "learning_rate": 0.16547617014952318,
      "loss": 0.2647,
      "num_input_tokens_seen": 14251056,
      "step": 18685
    },
    {
      "epoch": 38.856548856548855,
      "grad_norm": 0.00014990029740147293,
      "learning_rate": 0.1654175784545571,
      "loss": 0.2425,
      "num_input_tokens_seen": 14254800,
      "step": 18690
    },
    {
      "epoch": 38.86694386694387,
      "grad_norm": 0.0006807952304370701,
      "learning_rate": 0.1653589843820067,
      "loss": 0.2853,
      "num_input_tokens_seen": 14258960,
      "step": 18695
    },
    {
      "epoch": 38.87733887733888,
      "grad_norm": 0.00014907633885741234,
      "learning_rate": 0.1653003879409079,
      "loss": 0.2678,
      "num_input_tokens_seen": 14262736,
      "step": 18700
    },
    {
      "epoch": 38.88773388773389,
      "grad_norm": 0.00019776716362684965,
      "learning_rate": 0.165241789140297,
      "loss": 0.266,
      "num_input_tokens_seen": 14266512,
      "step": 18705
    },
    {
      "epoch": 38.898128898128896,
      "grad_norm": 0.00017867582209873945,
      "learning_rate": 0.16518318798921064,
      "loss": 0.262,
      "num_input_tokens_seen": 14270320,
      "step": 18710
    },
    {
      "epoch": 38.90852390852391,
      "grad_norm": 0.00011465382704045624,
      "learning_rate": 0.16512458449668593,
      "loss": 0.2772,
      "num_input_tokens_seen": 14274064,
      "step": 18715
    },
    {
      "epoch": 38.91891891891892,
      "grad_norm": 9.577245509717613e-05,
      "learning_rate": 0.1650659786717602,
      "loss": 0.2488,
      "num_input_tokens_seen": 14277968,
      "step": 18720
    },
    {
      "epoch": 38.92931392931393,
      "grad_norm": 0.00030009375768713653,
      "learning_rate": 0.1650073705234712,
      "loss": 0.2529,
      "num_input_tokens_seen": 14281904,
      "step": 18725
    },
    {
      "epoch": 38.93970893970894,
      "grad_norm": 0.00014796375762671232,
      "learning_rate": 0.16494876006085712,
      "loss": 0.2595,
      "num_input_tokens_seen": 14285712,
      "step": 18730
    },
    {
      "epoch": 38.95010395010395,
      "grad_norm": 0.0006412433576770127,
      "learning_rate": 0.16489014729295634,
      "loss": 0.2824,
      "num_input_tokens_seen": 14289488,
      "step": 18735
    },
    {
      "epoch": 38.96049896049896,
      "grad_norm": 0.00011991824430879205,
      "learning_rate": 0.16483153222880775,
      "loss": 0.2706,
      "num_input_tokens_seen": 14293264,
      "step": 18740
    },
    {
      "epoch": 38.97089397089397,
      "grad_norm": 0.0005097811808809638,
      "learning_rate": 0.16477291487745052,
      "loss": 0.2733,
      "num_input_tokens_seen": 14297040,
      "step": 18745
    },
    {
      "epoch": 38.981288981288984,
      "grad_norm": 0.0003074663982260972,
      "learning_rate": 0.16471429524792416,
      "loss": 0.2727,
      "num_input_tokens_seen": 14300688,
      "step": 18750
    },
    {
      "epoch": 38.99168399168399,
      "grad_norm": 9.364308789372444e-05,
      "learning_rate": 0.16465567334926856,
      "loss": 0.2635,
      "num_input_tokens_seen": 14304496,
      "step": 18755
    },
    {
      "epoch": 39.002079002079,
      "grad_norm": 0.00017646928608883172,
      "learning_rate": 0.16459704919052395,
      "loss": 0.2446,
      "num_input_tokens_seen": 14308320,
      "step": 18760
    },
    {
      "epoch": 39.012474012474016,
      "grad_norm": 0.00029763428028672934,
      "learning_rate": 0.16453842278073086,
      "loss": 0.259,
      "num_input_tokens_seen": 14312096,
      "step": 18765
    },
    {
      "epoch": 39.022869022869024,
      "grad_norm": 0.00030581539613194764,
      "learning_rate": 0.16447979412893038,
      "loss": 0.2727,
      "num_input_tokens_seen": 14315936,
      "step": 18770
    },
    {
      "epoch": 39.03326403326403,
      "grad_norm": 0.00025723359431140125,
      "learning_rate": 0.16442116324416367,
      "loss": 0.2451,
      "num_input_tokens_seen": 14319648,
      "step": 18775
    },
    {
      "epoch": 39.04365904365904,
      "grad_norm": 9.510527888778597e-05,
      "learning_rate": 0.1643625301354723,
      "loss": 0.2866,
      "num_input_tokens_seen": 14323456,
      "step": 18780
    },
    {
      "epoch": 39.054054054054056,
      "grad_norm": 9.29101079236716e-05,
      "learning_rate": 0.16430389481189828,
      "loss": 0.2638,
      "num_input_tokens_seen": 14327328,
      "step": 18785
    },
    {
      "epoch": 39.064449064449065,
      "grad_norm": 0.00034269114257767797,
      "learning_rate": 0.164245257282484,
      "loss": 0.2647,
      "num_input_tokens_seen": 14331136,
      "step": 18790
    },
    {
      "epoch": 39.07484407484407,
      "grad_norm": 0.00016197112563531846,
      "learning_rate": 0.16418661755627195,
      "loss": 0.2804,
      "num_input_tokens_seen": 14334944,
      "step": 18795
    },
    {
      "epoch": 39.08523908523909,
      "grad_norm": 0.0004046610265504569,
      "learning_rate": 0.16412797564230527,
      "loss": 0.2311,
      "num_input_tokens_seen": 14338720,
      "step": 18800
    },
    {
      "epoch": 39.08523908523909,
      "eval_loss": 0.24795089662075043,
      "eval_runtime": 13.4589,
      "eval_samples_per_second": 63.601,
      "eval_steps_per_second": 15.9,
      "num_input_tokens_seen": 14338720,
      "step": 18800
    },
    {
      "epoch": 39.0956340956341,
      "grad_norm": 0.00024159486929420382,
      "learning_rate": 0.16406933154962713,
      "loss": 0.2867,
      "num_input_tokens_seen": 14342400,
      "step": 18805
    },
    {
      "epoch": 39.106029106029105,
      "grad_norm": 6.994723662501201e-05,
      "learning_rate": 0.16401068528728133,
      "loss": 0.2723,
      "num_input_tokens_seen": 14346016,
      "step": 18810
    },
    {
      "epoch": 39.11642411642411,
      "grad_norm": 5.400833106250502e-05,
      "learning_rate": 0.16395203686431173,
      "loss": 0.2734,
      "num_input_tokens_seen": 14349824,
      "step": 18815
    },
    {
      "epoch": 39.12681912681913,
      "grad_norm": 0.0003098831803072244,
      "learning_rate": 0.16389338628976277,
      "loss": 0.2724,
      "num_input_tokens_seen": 14353696,
      "step": 18820
    },
    {
      "epoch": 39.13721413721414,
      "grad_norm": 0.0002838727959897369,
      "learning_rate": 0.163834733572679,
      "loss": 0.2744,
      "num_input_tokens_seen": 14357312,
      "step": 18825
    },
    {
      "epoch": 39.147609147609145,
      "grad_norm": 0.00011237752187298611,
      "learning_rate": 0.16377607872210545,
      "loss": 0.2661,
      "num_input_tokens_seen": 14361184,
      "step": 18830
    },
    {
      "epoch": 39.15800415800416,
      "grad_norm": 0.0003228596178814769,
      "learning_rate": 0.16371742174708748,
      "loss": 0.2209,
      "num_input_tokens_seen": 14364896,
      "step": 18835
    },
    {
      "epoch": 39.16839916839917,
      "grad_norm": 0.0002744507510215044,
      "learning_rate": 0.16365876265667065,
      "loss": 0.2532,
      "num_input_tokens_seen": 14368544,
      "step": 18840
    },
    {
      "epoch": 39.17879417879418,
      "grad_norm": 0.000546666793525219,
      "learning_rate": 0.163600101459901,
      "loss": 0.3089,
      "num_input_tokens_seen": 14372288,
      "step": 18845
    },
    {
      "epoch": 39.189189189189186,
      "grad_norm": 0.00024364182900171727,
      "learning_rate": 0.16354143816582484,
      "loss": 0.2577,
      "num_input_tokens_seen": 14376096,
      "step": 18850
    },
    {
      "epoch": 39.1995841995842,
      "grad_norm": 6.36070326436311e-05,
      "learning_rate": 0.1634827727834887,
      "loss": 0.2743,
      "num_input_tokens_seen": 14379872,
      "step": 18855
    },
    {
      "epoch": 39.20997920997921,
      "grad_norm": 0.00012105060886824504,
      "learning_rate": 0.16342410532193954,
      "loss": 0.2624,
      "num_input_tokens_seen": 14383712,
      "step": 18860
    },
    {
      "epoch": 39.22037422037422,
      "grad_norm": 0.000394036847865209,
      "learning_rate": 0.16336543579022464,
      "loss": 0.2792,
      "num_input_tokens_seen": 14387648,
      "step": 18865
    },
    {
      "epoch": 39.23076923076923,
      "grad_norm": 0.00014412598102353513,
      "learning_rate": 0.16330676419739157,
      "loss": 0.2391,
      "num_input_tokens_seen": 14391488,
      "step": 18870
    },
    {
      "epoch": 39.24116424116424,
      "grad_norm": 0.00044043874368071556,
      "learning_rate": 0.1632480905524883,
      "loss": 0.2876,
      "num_input_tokens_seen": 14395168,
      "step": 18875
    },
    {
      "epoch": 39.25155925155925,
      "grad_norm": 9.197332110488787e-05,
      "learning_rate": 0.16318941486456293,
      "loss": 0.2671,
      "num_input_tokens_seen": 14398976,
      "step": 18880
    },
    {
      "epoch": 39.26195426195426,
      "grad_norm": 0.0002466531877871603,
      "learning_rate": 0.16313073714266405,
      "loss": 0.2749,
      "num_input_tokens_seen": 14402816,
      "step": 18885
    },
    {
      "epoch": 39.272349272349274,
      "grad_norm": 0.000277255050605163,
      "learning_rate": 0.16307205739584052,
      "loss": 0.2639,
      "num_input_tokens_seen": 14406496,
      "step": 18890
    },
    {
      "epoch": 39.28274428274428,
      "grad_norm": 0.0004321524756960571,
      "learning_rate": 0.16301337563314144,
      "loss": 0.2454,
      "num_input_tokens_seen": 14410400,
      "step": 18895
    },
    {
      "epoch": 39.29313929313929,
      "grad_norm": 0.0004221178824082017,
      "learning_rate": 0.1629546918636163,
      "loss": 0.3007,
      "num_input_tokens_seen": 14414144,
      "step": 18900
    },
    {
      "epoch": 39.303534303534306,
      "grad_norm": 7.016614836174995e-05,
      "learning_rate": 0.16289600609631485,
      "loss": 0.2619,
      "num_input_tokens_seen": 14417952,
      "step": 18905
    },
    {
      "epoch": 39.313929313929314,
      "grad_norm": 0.00027923999004997313,
      "learning_rate": 0.16283731834028722,
      "loss": 0.2525,
      "num_input_tokens_seen": 14421632,
      "step": 18910
    },
    {
      "epoch": 39.32432432432432,
      "grad_norm": 0.0005079591064713895,
      "learning_rate": 0.16277862860458378,
      "loss": 0.289,
      "num_input_tokens_seen": 14425504,
      "step": 18915
    },
    {
      "epoch": 39.33471933471934,
      "grad_norm": 0.0005426101852208376,
      "learning_rate": 0.16271993689825526,
      "loss": 0.2808,
      "num_input_tokens_seen": 14429472,
      "step": 18920
    },
    {
      "epoch": 39.34511434511435,
      "grad_norm": 0.00028841980383731425,
      "learning_rate": 0.1626612432303526,
      "loss": 0.2728,
      "num_input_tokens_seen": 14433120,
      "step": 18925
    },
    {
      "epoch": 39.355509355509355,
      "grad_norm": 0.0003430134674999863,
      "learning_rate": 0.1626025476099271,
      "loss": 0.2733,
      "num_input_tokens_seen": 14436896,
      "step": 18930
    },
    {
      "epoch": 39.36590436590436,
      "grad_norm": 0.0004487436672206968,
      "learning_rate": 0.1625438500460304,
      "loss": 0.2576,
      "num_input_tokens_seen": 14440672,
      "step": 18935
    },
    {
      "epoch": 39.37629937629938,
      "grad_norm": 0.00038152706110849977,
      "learning_rate": 0.16248515054771442,
      "loss": 0.2201,
      "num_input_tokens_seen": 14444448,
      "step": 18940
    },
    {
      "epoch": 39.38669438669439,
      "grad_norm": 0.000663946324493736,
      "learning_rate": 0.16242644912403123,
      "loss": 0.3338,
      "num_input_tokens_seen": 14448320,
      "step": 18945
    },
    {
      "epoch": 39.397089397089395,
      "grad_norm": 0.00042350523290224373,
      "learning_rate": 0.1623677457840335,
      "loss": 0.3054,
      "num_input_tokens_seen": 14452128,
      "step": 18950
    },
    {
      "epoch": 39.40748440748441,
      "grad_norm": 0.00016630318714305758,
      "learning_rate": 0.16230904053677397,
      "loss": 0.2759,
      "num_input_tokens_seen": 14455744,
      "step": 18955
    },
    {
      "epoch": 39.41787941787942,
      "grad_norm": 0.00023443266400136054,
      "learning_rate": 0.16225033339130568,
      "loss": 0.2796,
      "num_input_tokens_seen": 14459488,
      "step": 18960
    },
    {
      "epoch": 39.42827442827443,
      "grad_norm": 0.0002817730710376054,
      "learning_rate": 0.16219162435668197,
      "loss": 0.2573,
      "num_input_tokens_seen": 14463296,
      "step": 18965
    },
    {
      "epoch": 39.438669438669436,
      "grad_norm": 0.0005313310539349914,
      "learning_rate": 0.16213291344195666,
      "loss": 0.2766,
      "num_input_tokens_seen": 14467136,
      "step": 18970
    },
    {
      "epoch": 39.44906444906445,
      "grad_norm": 0.0002998763229697943,
      "learning_rate": 0.16207420065618358,
      "loss": 0.283,
      "num_input_tokens_seen": 14471008,
      "step": 18975
    },
    {
      "epoch": 39.45945945945946,
      "grad_norm": 0.00015673093730583787,
      "learning_rate": 0.16201548600841706,
      "loss": 0.2741,
      "num_input_tokens_seen": 14474848,
      "step": 18980
    },
    {
      "epoch": 39.46985446985447,
      "grad_norm": 0.00018048730271402746,
      "learning_rate": 0.16195676950771154,
      "loss": 0.2759,
      "num_input_tokens_seen": 14478720,
      "step": 18985
    },
    {
      "epoch": 39.48024948024948,
      "grad_norm": 0.00031342243892140687,
      "learning_rate": 0.16189805116312198,
      "loss": 0.2647,
      "num_input_tokens_seen": 14482720,
      "step": 18990
    },
    {
      "epoch": 39.49064449064449,
      "grad_norm": 0.0005804307293146849,
      "learning_rate": 0.16183933098370337,
      "loss": 0.2656,
      "num_input_tokens_seen": 14486560,
      "step": 18995
    },
    {
      "epoch": 39.5010395010395,
      "grad_norm": 5.875601709703915e-05,
      "learning_rate": 0.16178060897851115,
      "loss": 0.2502,
      "num_input_tokens_seen": 14490240,
      "step": 19000
    },
    {
      "epoch": 39.5010395010395,
      "eval_loss": 0.2497202754020691,
      "eval_runtime": 13.4861,
      "eval_samples_per_second": 63.473,
      "eval_steps_per_second": 15.868,
      "num_input_tokens_seen": 14490240,
      "step": 19000
    },
    {
      "epoch": 39.51143451143451,
      "grad_norm": 0.0005644356133416295,
      "learning_rate": 0.16172188515660096,
      "loss": 0.3014,
      "num_input_tokens_seen": 14494176,
      "step": 19005
    },
    {
      "epoch": 39.521829521829524,
      "grad_norm": 8.414949115831405e-05,
      "learning_rate": 0.16166315952702878,
      "loss": 0.2902,
      "num_input_tokens_seen": 14497984,
      "step": 19010
    },
    {
      "epoch": 39.53222453222453,
      "grad_norm": 0.00014315721637103707,
      "learning_rate": 0.16160443209885084,
      "loss": 0.28,
      "num_input_tokens_seen": 14501792,
      "step": 19015
    },
    {
      "epoch": 39.54261954261954,
      "grad_norm": 0.0004750539956148714,
      "learning_rate": 0.16154570288112363,
      "loss": 0.2801,
      "num_input_tokens_seen": 14505600,
      "step": 19020
    },
    {
      "epoch": 39.553014553014556,
      "grad_norm": 0.00038815807783976197,
      "learning_rate": 0.16148697188290395,
      "loss": 0.2677,
      "num_input_tokens_seen": 14509376,
      "step": 19025
    },
    {
      "epoch": 39.563409563409564,
      "grad_norm": 0.00015546979557257146,
      "learning_rate": 0.16142823911324888,
      "loss": 0.2516,
      "num_input_tokens_seen": 14513248,
      "step": 19030
    },
    {
      "epoch": 39.57380457380457,
      "grad_norm": 0.00020375453459564596,
      "learning_rate": 0.16136950458121568,
      "loss": 0.2514,
      "num_input_tokens_seen": 14516864,
      "step": 19035
    },
    {
      "epoch": 39.58419958419958,
      "grad_norm": 0.0001250451459782198,
      "learning_rate": 0.16131076829586205,
      "loss": 0.2462,
      "num_input_tokens_seen": 14520736,
      "step": 19040
    },
    {
      "epoch": 39.5945945945946,
      "grad_norm": 0.0005256778094917536,
      "learning_rate": 0.1612520302662457,
      "loss": 0.2896,
      "num_input_tokens_seen": 14524480,
      "step": 19045
    },
    {
      "epoch": 39.604989604989605,
      "grad_norm": 0.00018821176490746439,
      "learning_rate": 0.16119329050142497,
      "loss": 0.2544,
      "num_input_tokens_seen": 14528032,
      "step": 19050
    },
    {
      "epoch": 39.61538461538461,
      "grad_norm": 0.0003220030921511352,
      "learning_rate": 0.16113454901045818,
      "loss": 0.2605,
      "num_input_tokens_seen": 14531904,
      "step": 19055
    },
    {
      "epoch": 39.62577962577963,
      "grad_norm": 0.0002236302534583956,
      "learning_rate": 0.16107580580240397,
      "loss": 0.2516,
      "num_input_tokens_seen": 14535648,
      "step": 19060
    },
    {
      "epoch": 39.63617463617464,
      "grad_norm": 0.0001901189680211246,
      "learning_rate": 0.16101706088632134,
      "loss": 0.2267,
      "num_input_tokens_seen": 14539520,
      "step": 19065
    },
    {
      "epoch": 39.646569646569645,
      "grad_norm": 0.00013239476538728923,
      "learning_rate": 0.16095831427126947,
      "loss": 0.268,
      "num_input_tokens_seen": 14543328,
      "step": 19070
    },
    {
      "epoch": 39.656964656964654,
      "grad_norm": 0.0001788314402801916,
      "learning_rate": 0.16089956596630783,
      "loss": 0.2235,
      "num_input_tokens_seen": 14547200,
      "step": 19075
    },
    {
      "epoch": 39.66735966735967,
      "grad_norm": 0.0001133709738496691,
      "learning_rate": 0.16084081598049618,
      "loss": 0.2496,
      "num_input_tokens_seen": 14550880,
      "step": 19080
    },
    {
      "epoch": 39.67775467775468,
      "grad_norm": 0.00023295370920095593,
      "learning_rate": 0.1607820643228944,
      "loss": 0.2882,
      "num_input_tokens_seen": 14554720,
      "step": 19085
    },
    {
      "epoch": 39.688149688149686,
      "grad_norm": 0.0003464298788458109,
      "learning_rate": 0.16072331100256285,
      "loss": 0.2773,
      "num_input_tokens_seen": 14558560,
      "step": 19090
    },
    {
      "epoch": 39.6985446985447,
      "grad_norm": 0.00013704427692573518,
      "learning_rate": 0.16066455602856197,
      "loss": 0.2767,
      "num_input_tokens_seen": 14562336,
      "step": 19095
    },
    {
      "epoch": 39.70893970893971,
      "grad_norm": 0.0004995701019652188,
      "learning_rate": 0.16060579940995257,
      "loss": 0.2791,
      "num_input_tokens_seen": 14566208,
      "step": 19100
    },
    {
      "epoch": 39.71933471933472,
      "grad_norm": 0.00016199800302274525,
      "learning_rate": 0.16054704115579557,
      "loss": 0.2679,
      "num_input_tokens_seen": 14569984,
      "step": 19105
    },
    {
      "epoch": 39.729729729729726,
      "grad_norm": 0.00030695617897436023,
      "learning_rate": 0.1604882812751523,
      "loss": 0.2535,
      "num_input_tokens_seen": 14573856,
      "step": 19110
    },
    {
      "epoch": 39.74012474012474,
      "grad_norm": 0.0001370328536722809,
      "learning_rate": 0.16042951977708425,
      "loss": 0.2657,
      "num_input_tokens_seen": 14577696,
      "step": 19115
    },
    {
      "epoch": 39.75051975051975,
      "grad_norm": 8.185114711523056e-05,
      "learning_rate": 0.16037075667065318,
      "loss": 0.2793,
      "num_input_tokens_seen": 14581536,
      "step": 19120
    },
    {
      "epoch": 39.76091476091476,
      "grad_norm": 7.148168515414e-05,
      "learning_rate": 0.1603119919649211,
      "loss": 0.2797,
      "num_input_tokens_seen": 14585280,
      "step": 19125
    },
    {
      "epoch": 39.771309771309774,
      "grad_norm": 0.0004562731191981584,
      "learning_rate": 0.16025322566895028,
      "loss": 0.2664,
      "num_input_tokens_seen": 14589088,
      "step": 19130
    },
    {
      "epoch": 39.78170478170478,
      "grad_norm": 3.264590122853406e-05,
      "learning_rate": 0.16019445779180322,
      "loss": 0.253,
      "num_input_tokens_seen": 14592896,
      "step": 19135
    },
    {
      "epoch": 39.79209979209979,
      "grad_norm": 0.00019897542370017618,
      "learning_rate": 0.16013568834254271,
      "loss": 0.2898,
      "num_input_tokens_seen": 14596576,
      "step": 19140
    },
    {
      "epoch": 39.802494802494806,
      "grad_norm": 5.010885070078075e-05,
      "learning_rate": 0.1600769173302316,
      "loss": 0.2262,
      "num_input_tokens_seen": 14600608,
      "step": 19145
    },
    {
      "epoch": 39.812889812889814,
      "grad_norm": 0.00017095061775762588,
      "learning_rate": 0.16001814476393322,
      "loss": 0.2563,
      "num_input_tokens_seen": 14604576,
      "step": 19150
    },
    {
      "epoch": 39.82328482328482,
      "grad_norm": 7.988956349436194e-05,
      "learning_rate": 0.15995937065271104,
      "loss": 0.2814,
      "num_input_tokens_seen": 14608448,
      "step": 19155
    },
    {
      "epoch": 39.83367983367983,
      "grad_norm": 0.00030251420685090125,
      "learning_rate": 0.15990059500562873,
      "loss": 0.2508,
      "num_input_tokens_seen": 14612256,
      "step": 19160
    },
    {
      "epoch": 39.84407484407485,
      "grad_norm": 0.0003945770440623164,
      "learning_rate": 0.15984181783175025,
      "loss": 0.2511,
      "num_input_tokens_seen": 14616032,
      "step": 19165
    },
    {
      "epoch": 39.854469854469855,
      "grad_norm": 0.0002816040941979736,
      "learning_rate": 0.1597830391401398,
      "loss": 0.2826,
      "num_input_tokens_seen": 14619968,
      "step": 19170
    },
    {
      "epoch": 39.86486486486486,
      "grad_norm": 0.00021585570357274264,
      "learning_rate": 0.15972425893986178,
      "loss": 0.2756,
      "num_input_tokens_seen": 14623872,
      "step": 19175
    },
    {
      "epoch": 39.87525987525988,
      "grad_norm": 0.00045787045382894576,
      "learning_rate": 0.15966547723998084,
      "loss": 0.2725,
      "num_input_tokens_seen": 14627680,
      "step": 19180
    },
    {
      "epoch": 39.88565488565489,
      "grad_norm": 0.00048087825416587293,
      "learning_rate": 0.15960669404956176,
      "loss": 0.2785,
      "num_input_tokens_seen": 14631520,
      "step": 19185
    },
    {
      "epoch": 39.896049896049895,
      "grad_norm": 7.098748756106943e-05,
      "learning_rate": 0.1595479093776698,
      "loss": 0.2774,
      "num_input_tokens_seen": 14635264,
      "step": 19190
    },
    {
      "epoch": 39.906444906444904,
      "grad_norm": 0.00028822338208556175,
      "learning_rate": 0.15948912323337022,
      "loss": 0.2638,
      "num_input_tokens_seen": 14639136,
      "step": 19195
    },
    {
      "epoch": 39.91683991683992,
      "grad_norm": 0.000355742871761322,
      "learning_rate": 0.1594303356257286,
      "loss": 0.2788,
      "num_input_tokens_seen": 14643072,
      "step": 19200
    },
    {
      "epoch": 39.91683991683992,
      "eval_loss": 0.24827949702739716,
      "eval_runtime": 13.4098,
      "eval_samples_per_second": 63.834,
      "eval_steps_per_second": 15.958,
      "num_input_tokens_seen": 14643072,
      "step": 19200
    },
    {
      "epoch": 39.92723492723493,
      "grad_norm": 0.00035336625296622515,
      "learning_rate": 0.15937154656381072,
      "loss": 0.2656,
      "num_input_tokens_seen": 14646944,
      "step": 19205
    },
    {
      "epoch": 39.937629937629936,
      "grad_norm": 0.0004528326098807156,
      "learning_rate": 0.15931275605668258,
      "loss": 0.2557,
      "num_input_tokens_seen": 14650784,
      "step": 19210
    },
    {
      "epoch": 39.94802494802495,
      "grad_norm": 7.95606174506247e-05,
      "learning_rate": 0.1592539641134104,
      "loss": 0.2628,
      "num_input_tokens_seen": 14654688,
      "step": 19215
    },
    {
      "epoch": 39.95841995841996,
      "grad_norm": 0.00010925047536147758,
      "learning_rate": 0.1591951707430607,
      "loss": 0.2683,
      "num_input_tokens_seen": 14658432,
      "step": 19220
    },
    {
      "epoch": 39.96881496881497,
      "grad_norm": 0.00040998036274686456,
      "learning_rate": 0.15913637595470007,
      "loss": 0.2549,
      "num_input_tokens_seen": 14662144,
      "step": 19225
    },
    {
      "epoch": 39.979209979209976,
      "grad_norm": 0.00030653062276542187,
      "learning_rate": 0.15907757975739548,
      "loss": 0.2563,
      "num_input_tokens_seen": 14666080,
      "step": 19230
    },
    {
      "epoch": 39.98960498960499,
      "grad_norm": 0.00036579446168616414,
      "learning_rate": 0.159018782160214,
      "loss": 0.2465,
      "num_input_tokens_seen": 14669920,
      "step": 19235
    },
    {
      "epoch": 40.0,
      "grad_norm": 0.0003128277894575149,
      "learning_rate": 0.158959983172223,
      "loss": 0.2569,
      "num_input_tokens_seen": 14673744,
      "step": 19240
    },
    {
      "epoch": 40.01039501039501,
      "grad_norm": 0.0006778359529562294,
      "learning_rate": 0.15890118280249,
      "loss": 0.2748,
      "num_input_tokens_seen": 14677424,
      "step": 19245
    },
    {
      "epoch": 40.020790020790024,
      "grad_norm": 0.00021710197324864566,
      "learning_rate": 0.15884238106008275,
      "loss": 0.2487,
      "num_input_tokens_seen": 14681200,
      "step": 19250
    },
    {
      "epoch": 40.03118503118503,
      "grad_norm": 0.00010688927432056516,
      "learning_rate": 0.15878357795406922,
      "loss": 0.26,
      "num_input_tokens_seen": 14684976,
      "step": 19255
    },
    {
      "epoch": 40.04158004158004,
      "grad_norm": 0.0005350644350983202,
      "learning_rate": 0.15872477349351757,
      "loss": 0.2642,
      "num_input_tokens_seen": 14688720,
      "step": 19260
    },
    {
      "epoch": 40.05197505197505,
      "grad_norm": 0.00046306231524795294,
      "learning_rate": 0.15866596768749622,
      "loss": 0.2921,
      "num_input_tokens_seen": 14692368,
      "step": 19265
    },
    {
      "epoch": 40.062370062370064,
      "grad_norm": 0.0008180579752661288,
      "learning_rate": 0.15860716054507373,
      "loss": 0.2617,
      "num_input_tokens_seen": 14696080,
      "step": 19270
    },
    {
      "epoch": 40.07276507276507,
      "grad_norm": 0.00015625995001755655,
      "learning_rate": 0.1585483520753189,
      "loss": 0.2713,
      "num_input_tokens_seen": 14699760,
      "step": 19275
    },
    {
      "epoch": 40.08316008316008,
      "grad_norm": 0.0004341186722740531,
      "learning_rate": 0.1584895422873008,
      "loss": 0.2685,
      "num_input_tokens_seen": 14703696,
      "step": 19280
    },
    {
      "epoch": 40.093555093555096,
      "grad_norm": 0.0003843106678687036,
      "learning_rate": 0.1584307311900886,
      "loss": 0.2735,
      "num_input_tokens_seen": 14707504,
      "step": 19285
    },
    {
      "epoch": 40.103950103950105,
      "grad_norm": 0.0003282174584455788,
      "learning_rate": 0.1583719187927517,
      "loss": 0.2706,
      "num_input_tokens_seen": 14711312,
      "step": 19290
    },
    {
      "epoch": 40.11434511434511,
      "grad_norm": 0.0004849833494517952,
      "learning_rate": 0.15831310510435967,
      "loss": 0.2742,
      "num_input_tokens_seen": 14715056,
      "step": 19295
    },
    {
      "epoch": 40.12474012474012,
      "grad_norm": 0.00028795923572033644,
      "learning_rate": 0.15825429013398243,
      "loss": 0.2684,
      "num_input_tokens_seen": 14718992,
      "step": 19300
    },
    {
      "epoch": 40.13513513513514,
      "grad_norm": 0.0003188580449204892,
      "learning_rate": 0.15819547389068986,
      "loss": 0.274,
      "num_input_tokens_seen": 14722672,
      "step": 19305
    },
    {
      "epoch": 40.145530145530145,
      "grad_norm": 0.00022718128457199782,
      "learning_rate": 0.1581366563835522,
      "loss": 0.2725,
      "num_input_tokens_seen": 14726416,
      "step": 19310
    },
    {
      "epoch": 40.15592515592515,
      "grad_norm": 0.00019647617591544986,
      "learning_rate": 0.15807783762163993,
      "loss": 0.282,
      "num_input_tokens_seen": 14730256,
      "step": 19315
    },
    {
      "epoch": 40.16632016632017,
      "grad_norm": 0.0002832565223798156,
      "learning_rate": 0.15801901761402365,
      "loss": 0.2766,
      "num_input_tokens_seen": 14734096,
      "step": 19320
    },
    {
      "epoch": 40.17671517671518,
      "grad_norm": 0.00012157320452388376,
      "learning_rate": 0.157960196369774,
      "loss": 0.2716,
      "num_input_tokens_seen": 14737872,
      "step": 19325
    },
    {
      "epoch": 40.187110187110186,
      "grad_norm": 0.00032345092040486634,
      "learning_rate": 0.157901373897962,
      "loss": 0.268,
      "num_input_tokens_seen": 14741616,
      "step": 19330
    },
    {
      "epoch": 40.197505197505194,
      "grad_norm": 0.00012614870502147824,
      "learning_rate": 0.15784255020765892,
      "loss": 0.2539,
      "num_input_tokens_seen": 14745456,
      "step": 19335
    },
    {
      "epoch": 40.20790020790021,
      "grad_norm": 4.1194787627318874e-05,
      "learning_rate": 0.157783725307936,
      "loss": 0.2515,
      "num_input_tokens_seen": 14749200,
      "step": 19340
    },
    {
      "epoch": 40.21829521829522,
      "grad_norm": 0.00013640409451909363,
      "learning_rate": 0.15772489920786484,
      "loss": 0.2936,
      "num_input_tokens_seen": 14753040,
      "step": 19345
    },
    {
      "epoch": 40.228690228690226,
      "grad_norm": 0.0002855884376913309,
      "learning_rate": 0.15766607191651713,
      "loss": 0.2599,
      "num_input_tokens_seen": 14757072,
      "step": 19350
    },
    {
      "epoch": 40.23908523908524,
      "grad_norm": 0.0003141595225315541,
      "learning_rate": 0.1576072434429648,
      "loss": 0.258,
      "num_input_tokens_seen": 14760720,
      "step": 19355
    },
    {
      "epoch": 40.24948024948025,
      "grad_norm": 6.51771406410262e-05,
      "learning_rate": 0.15754841379627998,
      "loss": 0.2781,
      "num_input_tokens_seen": 14764656,
      "step": 19360
    },
    {
      "epoch": 40.25987525987526,
      "grad_norm": 0.00017045206914190203,
      "learning_rate": 0.15748958298553484,
      "loss": 0.2669,
      "num_input_tokens_seen": 14768464,
      "step": 19365
    },
    {
      "epoch": 40.270270270270274,
      "grad_norm": 4.464116500457749e-05,
      "learning_rate": 0.1574307510198019,
      "loss": 0.2639,
      "num_input_tokens_seen": 14772272,
      "step": 19370
    },
    {
      "epoch": 40.28066528066528,
      "grad_norm": 0.00023470954329241067,
      "learning_rate": 0.15737191790815375,
      "loss": 0.2554,
      "num_input_tokens_seen": 14776080,
      "step": 19375
    },
    {
      "epoch": 40.29106029106029,
      "grad_norm": 6.293384649325162e-05,
      "learning_rate": 0.15731308365966323,
      "loss": 0.2256,
      "num_input_tokens_seen": 14779760,
      "step": 19380
    },
    {
      "epoch": 40.3014553014553,
      "grad_norm": 0.00042729778215289116,
      "learning_rate": 0.15725424828340331,
      "loss": 0.2532,
      "num_input_tokens_seen": 14783760,
      "step": 19385
    },
    {
      "epoch": 40.311850311850314,
      "grad_norm": 3.17596313834656e-05,
      "learning_rate": 0.15719541178844715,
      "loss": 0.3026,
      "num_input_tokens_seen": 14787664,
      "step": 19390
    },
    {
      "epoch": 40.32224532224532,
      "grad_norm": 0.00015527821960859,
      "learning_rate": 0.15713657418386806,
      "loss": 0.286,
      "num_input_tokens_seen": 14791312,
      "step": 19395
    },
    {
      "epoch": 40.33264033264033,
      "grad_norm": 0.00028506197850219905,
      "learning_rate": 0.15707773547873957,
      "loss": 0.2716,
      "num_input_tokens_seen": 14795184,
      "step": 19400
    },
    {
      "epoch": 40.33264033264033,
      "eval_loss": 0.25817909836769104,
      "eval_runtime": 13.4178,
      "eval_samples_per_second": 63.796,
      "eval_steps_per_second": 15.949,
      "num_input_tokens_seen": 14795184,
      "step": 19400
    },
    {
      "epoch": 40.343035343035346,
      "grad_norm": 0.00026690305094234645,
      "learning_rate": 0.1570188956821353,
      "loss": 0.2664,
      "num_input_tokens_seen": 14798800,
      "step": 19405
    },
    {
      "epoch": 40.353430353430355,
      "grad_norm": 0.00019005748617928475,
      "learning_rate": 0.1569600548031291,
      "loss": 0.2697,
      "num_input_tokens_seen": 14802768,
      "step": 19410
    },
    {
      "epoch": 40.36382536382536,
      "grad_norm": 0.00023546448210254312,
      "learning_rate": 0.156901212850795,
      "loss": 0.2922,
      "num_input_tokens_seen": 14806736,
      "step": 19415
    },
    {
      "epoch": 40.37422037422037,
      "grad_norm": 0.0002392926689935848,
      "learning_rate": 0.15684236983420716,
      "loss": 0.2601,
      "num_input_tokens_seen": 14810512,
      "step": 19420
    },
    {
      "epoch": 40.38461538461539,
      "grad_norm": 0.00047618846292607486,
      "learning_rate": 0.1567835257624399,
      "loss": 0.2573,
      "num_input_tokens_seen": 14814320,
      "step": 19425
    },
    {
      "epoch": 40.395010395010395,
      "grad_norm": 0.0002257501328131184,
      "learning_rate": 0.1567246806445677,
      "loss": 0.2687,
      "num_input_tokens_seen": 14818224,
      "step": 19430
    },
    {
      "epoch": 40.4054054054054,
      "grad_norm": 0.0001277759438380599,
      "learning_rate": 0.15666583448966526,
      "loss": 0.2598,
      "num_input_tokens_seen": 14821936,
      "step": 19435
    },
    {
      "epoch": 40.41580041580042,
      "grad_norm": 9.289153967984021e-05,
      "learning_rate": 0.1566069873068074,
      "loss": 0.2615,
      "num_input_tokens_seen": 14825712,
      "step": 19440
    },
    {
      "epoch": 40.42619542619543,
      "grad_norm": 0.0002395427436567843,
      "learning_rate": 0.156548139105069,
      "loss": 0.2575,
      "num_input_tokens_seen": 14829424,
      "step": 19445
    },
    {
      "epoch": 40.436590436590436,
      "grad_norm": 0.0002534229715820402,
      "learning_rate": 0.15648928989352529,
      "loss": 0.258,
      "num_input_tokens_seen": 14833232,
      "step": 19450
    },
    {
      "epoch": 40.446985446985444,
      "grad_norm": 0.00021143726189620793,
      "learning_rate": 0.15643043968125156,
      "loss": 0.2726,
      "num_input_tokens_seen": 14837040,
      "step": 19455
    },
    {
      "epoch": 40.45738045738046,
      "grad_norm": 0.0006262267706915736,
      "learning_rate": 0.15637158847732316,
      "loss": 0.2856,
      "num_input_tokens_seen": 14840880,
      "step": 19460
    },
    {
      "epoch": 40.46777546777547,
      "grad_norm": 0.000190770675544627,
      "learning_rate": 0.15631273629081582,
      "loss": 0.2828,
      "num_input_tokens_seen": 14844688,
      "step": 19465
    },
    {
      "epoch": 40.478170478170476,
      "grad_norm": 0.00032719149021431804,
      "learning_rate": 0.15625388313080518,
      "loss": 0.286,
      "num_input_tokens_seen": 14848464,
      "step": 19470
    },
    {
      "epoch": 40.48856548856549,
      "grad_norm": 0.0003644162497948855,
      "learning_rate": 0.15619502900636714,
      "loss": 0.2747,
      "num_input_tokens_seen": 14852112,
      "step": 19475
    },
    {
      "epoch": 40.4989604989605,
      "grad_norm": 0.00034886348294094205,
      "learning_rate": 0.15613617392657783,
      "loss": 0.2548,
      "num_input_tokens_seen": 14855952,
      "step": 19480
    },
    {
      "epoch": 40.50935550935551,
      "grad_norm": 0.00010568360448814929,
      "learning_rate": 0.15607731790051335,
      "loss": 0.2699,
      "num_input_tokens_seen": 14859792,
      "step": 19485
    },
    {
      "epoch": 40.51975051975052,
      "grad_norm": 0.0002857714134734124,
      "learning_rate": 0.15601846093725008,
      "loss": 0.2586,
      "num_input_tokens_seen": 14863632,
      "step": 19490
    },
    {
      "epoch": 40.53014553014553,
      "grad_norm": 0.00010057652980322018,
      "learning_rate": 0.1559596030458645,
      "loss": 0.2436,
      "num_input_tokens_seen": 14867376,
      "step": 19495
    },
    {
      "epoch": 40.54054054054054,
      "grad_norm": 0.00026237242855131626,
      "learning_rate": 0.1559007442354333,
      "loss": 0.2637,
      "num_input_tokens_seen": 14871152,
      "step": 19500
    },
    {
      "epoch": 40.55093555093555,
      "grad_norm": 0.00014281108451541513,
      "learning_rate": 0.15584188451503314,
      "loss": 0.259,
      "num_input_tokens_seen": 14875088,
      "step": 19505
    },
    {
      "epoch": 40.561330561330564,
      "grad_norm": 0.00023837907065171748,
      "learning_rate": 0.15578302389374094,
      "loss": 0.2035,
      "num_input_tokens_seen": 14878864,
      "step": 19510
    },
    {
      "epoch": 40.57172557172557,
      "grad_norm": 0.0005240822792984545,
      "learning_rate": 0.1557241623806338,
      "loss": 0.2166,
      "num_input_tokens_seen": 14882704,
      "step": 19515
    },
    {
      "epoch": 40.58212058212058,
      "grad_norm": 8.65408728714101e-05,
      "learning_rate": 0.15566529998478887,
      "loss": 0.3236,
      "num_input_tokens_seen": 14886608,
      "step": 19520
    },
    {
      "epoch": 40.59251559251559,
      "grad_norm": 0.00020667626813519746,
      "learning_rate": 0.15560643671528354,
      "loss": 0.2797,
      "num_input_tokens_seen": 14890544,
      "step": 19525
    },
    {
      "epoch": 40.602910602910605,
      "grad_norm": 9.813201177166775e-05,
      "learning_rate": 0.15554757258119514,
      "loss": 0.2688,
      "num_input_tokens_seen": 14894448,
      "step": 19530
    },
    {
      "epoch": 40.61330561330561,
      "grad_norm": 0.00023076613433659077,
      "learning_rate": 0.1554887075916014,
      "loss": 0.2781,
      "num_input_tokens_seen": 14898128,
      "step": 19535
    },
    {
      "epoch": 40.62370062370062,
      "grad_norm": 0.00041677054832689464,
      "learning_rate": 0.15542984175558,
      "loss": 0.2725,
      "num_input_tokens_seen": 14901872,
      "step": 19540
    },
    {
      "epoch": 40.63409563409564,
      "grad_norm": 7.777262362651527e-05,
      "learning_rate": 0.1553709750822087,
      "loss": 0.2738,
      "num_input_tokens_seen": 14905520,
      "step": 19545
    },
    {
      "epoch": 40.644490644490645,
      "grad_norm": 0.0003350620099809021,
      "learning_rate": 0.15531210758056554,
      "loss": 0.2908,
      "num_input_tokens_seen": 14909424,
      "step": 19550
    },
    {
      "epoch": 40.65488565488565,
      "grad_norm": 0.00021306442795321345,
      "learning_rate": 0.15525323925972867,
      "loss": 0.2305,
      "num_input_tokens_seen": 14913232,
      "step": 19555
    },
    {
      "epoch": 40.66528066528066,
      "grad_norm": 0.0005110530182719231,
      "learning_rate": 0.15519437012877627,
      "loss": 0.276,
      "num_input_tokens_seen": 14917008,
      "step": 19560
    },
    {
      "epoch": 40.67567567567568,
      "grad_norm": 0.00038366674561984837,
      "learning_rate": 0.15513550019678676,
      "loss": 0.2922,
      "num_input_tokens_seen": 14920784,
      "step": 19565
    },
    {
      "epoch": 40.686070686070686,
      "grad_norm": 0.00013617715740110725,
      "learning_rate": 0.15507662947283854,
      "loss": 0.2634,
      "num_input_tokens_seen": 14924464,
      "step": 19570
    },
    {
      "epoch": 40.696465696465694,
      "grad_norm": 8.178168354788795e-05,
      "learning_rate": 0.15501775796601028,
      "loss": 0.2756,
      "num_input_tokens_seen": 14928176,
      "step": 19575
    },
    {
      "epoch": 40.70686070686071,
      "grad_norm": 0.00022606848506256938,
      "learning_rate": 0.15495888568538066,
      "loss": 0.2737,
      "num_input_tokens_seen": 14932048,
      "step": 19580
    },
    {
      "epoch": 40.71725571725572,
      "grad_norm": 8.381425868719816e-05,
      "learning_rate": 0.1549000126400286,
      "loss": 0.2778,
      "num_input_tokens_seen": 14935728,
      "step": 19585
    },
    {
      "epoch": 40.727650727650726,
      "grad_norm": 0.0003244522085878998,
      "learning_rate": 0.15484113883903294,
      "loss": 0.2629,
      "num_input_tokens_seen": 14939760,
      "step": 19590
    },
    {
      "epoch": 40.73804573804574,
      "grad_norm": 0.00022940777125768363,
      "learning_rate": 0.15478226429147288,
      "loss": 0.2587,
      "num_input_tokens_seen": 14943568,
      "step": 19595
    },
    {
      "epoch": 40.74844074844075,
      "grad_norm": 0.00017733285494614393,
      "learning_rate": 0.15472338900642757,
      "loss": 0.2627,
      "num_input_tokens_seen": 14947312,
      "step": 19600
    },
    {
      "epoch": 40.74844074844075,
      "eval_loss": 0.24833182990550995,
      "eval_runtime": 13.4689,
      "eval_samples_per_second": 63.554,
      "eval_steps_per_second": 15.888,
      "num_input_tokens_seen": 14947312,
      "step": 19600
    },
    {
      "epoch": 40.75883575883576,
      "grad_norm": 0.000331509712850675,
      "learning_rate": 0.15466451299297632,
      "loss": 0.2446,
      "num_input_tokens_seen": 14951152,
      "step": 19605
    },
    {
      "epoch": 40.76923076923077,
      "grad_norm": 0.00031276073423214257,
      "learning_rate": 0.15460563626019852,
      "loss": 0.2699,
      "num_input_tokens_seen": 14955152,
      "step": 19610
    },
    {
      "epoch": 40.77962577962578,
      "grad_norm": 6.963621854083613e-05,
      "learning_rate": 0.15454675881717375,
      "loss": 0.2641,
      "num_input_tokens_seen": 14959088,
      "step": 19615
    },
    {
      "epoch": 40.79002079002079,
      "grad_norm": 0.00032989951432682574,
      "learning_rate": 0.1544878806729816,
      "loss": 0.2574,
      "num_input_tokens_seen": 14963056,
      "step": 19620
    },
    {
      "epoch": 40.8004158004158,
      "grad_norm": 0.00017532636411488056,
      "learning_rate": 0.1544290018367019,
      "loss": 0.2553,
      "num_input_tokens_seen": 14966832,
      "step": 19625
    },
    {
      "epoch": 40.810810810810814,
      "grad_norm": 6.198434857651591e-05,
      "learning_rate": 0.15437012231741445,
      "loss": 0.2727,
      "num_input_tokens_seen": 14970768,
      "step": 19630
    },
    {
      "epoch": 40.82120582120582,
      "grad_norm": 0.0003781394334509969,
      "learning_rate": 0.1543112421241992,
      "loss": 0.2693,
      "num_input_tokens_seen": 14974512,
      "step": 19635
    },
    {
      "epoch": 40.83160083160083,
      "grad_norm": 0.00020204659085720778,
      "learning_rate": 0.15425236126613626,
      "loss": 0.2649,
      "num_input_tokens_seen": 14978224,
      "step": 19640
    },
    {
      "epoch": 40.84199584199584,
      "grad_norm": 0.0001514485338702798,
      "learning_rate": 0.15419347975230577,
      "loss": 0.2632,
      "num_input_tokens_seen": 14982096,
      "step": 19645
    },
    {
      "epoch": 40.852390852390855,
      "grad_norm": 8.541520946891978e-05,
      "learning_rate": 0.154134597591788,
      "loss": 0.2685,
      "num_input_tokens_seen": 14985872,
      "step": 19650
    },
    {
      "epoch": 40.86278586278586,
      "grad_norm": 0.00046932362602092326,
      "learning_rate": 0.1540757147936633,
      "loss": 0.2742,
      "num_input_tokens_seen": 14989680,
      "step": 19655
    },
    {
      "epoch": 40.87318087318087,
      "grad_norm": 0.00019683956634253263,
      "learning_rate": 0.1540168313670122,
      "loss": 0.2632,
      "num_input_tokens_seen": 14993520,
      "step": 19660
    },
    {
      "epoch": 40.88357588357589,
      "grad_norm": 0.00010083048982778564,
      "learning_rate": 0.1539579473209152,
      "loss": 0.2537,
      "num_input_tokens_seen": 14997232,
      "step": 19665
    },
    {
      "epoch": 40.893970893970895,
      "grad_norm": 0.00039551942609250546,
      "learning_rate": 0.15389906266445294,
      "loss": 0.2311,
      "num_input_tokens_seen": 15000848,
      "step": 19670
    },
    {
      "epoch": 40.9043659043659,
      "grad_norm": 0.00023763089848216623,
      "learning_rate": 0.15384017740670627,
      "loss": 0.2742,
      "num_input_tokens_seen": 15004720,
      "step": 19675
    },
    {
      "epoch": 40.91476091476091,
      "grad_norm": 0.00028581643709912896,
      "learning_rate": 0.15378129155675602,
      "loss": 0.2701,
      "num_input_tokens_seen": 15008432,
      "step": 19680
    },
    {
      "epoch": 40.92515592515593,
      "grad_norm": 0.000126958591863513,
      "learning_rate": 0.15372240512368307,
      "loss": 0.269,
      "num_input_tokens_seen": 15012144,
      "step": 19685
    },
    {
      "epoch": 40.935550935550935,
      "grad_norm": 0.0002582821762189269,
      "learning_rate": 0.1536635181165684,
      "loss": 0.2742,
      "num_input_tokens_seen": 15015920,
      "step": 19690
    },
    {
      "epoch": 40.945945945945944,
      "grad_norm": 0.0004045071837026626,
      "learning_rate": 0.15360463054449328,
      "loss": 0.2439,
      "num_input_tokens_seen": 15019824,
      "step": 19695
    },
    {
      "epoch": 40.95634095634096,
      "grad_norm": 0.00011655008711386472,
      "learning_rate": 0.1535457424165388,
      "loss": 0.258,
      "num_input_tokens_seen": 15023824,
      "step": 19700
    },
    {
      "epoch": 40.96673596673597,
      "grad_norm": 0.0004144597623962909,
      "learning_rate": 0.15348685374178628,
      "loss": 0.2972,
      "num_input_tokens_seen": 15027632,
      "step": 19705
    },
    {
      "epoch": 40.977130977130976,
      "grad_norm": 0.00020229780056979507,
      "learning_rate": 0.1534279645293171,
      "loss": 0.2745,
      "num_input_tokens_seen": 15031504,
      "step": 19710
    },
    {
      "epoch": 40.987525987525984,
      "grad_norm": 0.00015313579933717847,
      "learning_rate": 0.1533690747882127,
      "loss": 0.2766,
      "num_input_tokens_seen": 15035312,
      "step": 19715
    },
    {
      "epoch": 40.997920997921,
      "grad_norm": 0.00034366751788184047,
      "learning_rate": 0.15331018452755465,
      "loss": 0.2641,
      "num_input_tokens_seen": 15039216,
      "step": 19720
    },
    {
      "epoch": 41.00831600831601,
      "grad_norm": 0.00024178162857424468,
      "learning_rate": 0.15325129375642457,
      "loss": 0.2574,
      "num_input_tokens_seen": 15042960,
      "step": 19725
    },
    {
      "epoch": 41.018711018711016,
      "grad_norm": 0.0002120835124514997,
      "learning_rate": 0.15319240248390406,
      "loss": 0.2499,
      "num_input_tokens_seen": 15046704,
      "step": 19730
    },
    {
      "epoch": 41.02910602910603,
      "grad_norm": 0.00015668269770685583,
      "learning_rate": 0.153133510719075,
      "loss": 0.2894,
      "num_input_tokens_seen": 15050608,
      "step": 19735
    },
    {
      "epoch": 41.03950103950104,
      "grad_norm": 0.00043070336687378585,
      "learning_rate": 0.15307461847101922,
      "loss": 0.2509,
      "num_input_tokens_seen": 15054320,
      "step": 19740
    },
    {
      "epoch": 41.04989604989605,
      "grad_norm": 0.00043466646457090974,
      "learning_rate": 0.15301572574881864,
      "loss": 0.2546,
      "num_input_tokens_seen": 15058128,
      "step": 19745
    },
    {
      "epoch": 41.06029106029106,
      "grad_norm": 5.977210821583867e-05,
      "learning_rate": 0.15295683256155523,
      "loss": 0.2372,
      "num_input_tokens_seen": 15061968,
      "step": 19750
    },
    {
      "epoch": 41.07068607068607,
      "grad_norm": 0.0004895455203950405,
      "learning_rate": 0.15289793891831113,
      "loss": 0.2737,
      "num_input_tokens_seen": 15065648,
      "step": 19755
    },
    {
      "epoch": 41.08108108108108,
      "grad_norm": 0.00026473545585758984,
      "learning_rate": 0.15283904482816837,
      "loss": 0.2759,
      "num_input_tokens_seen": 15069424,
      "step": 19760
    },
    {
      "epoch": 41.09147609147609,
      "grad_norm": 0.0005848917062394321,
      "learning_rate": 0.15278015030020928,
      "loss": 0.221,
      "num_input_tokens_seen": 15073168,
      "step": 19765
    },
    {
      "epoch": 41.101871101871104,
      "grad_norm": 2.508679244783707e-05,
      "learning_rate": 0.152721255343516,
      "loss": 0.2508,
      "num_input_tokens_seen": 15076976,
      "step": 19770
    },
    {
      "epoch": 41.11226611226611,
      "grad_norm": 0.00019580198568291962,
      "learning_rate": 0.15266235996717098,
      "loss": 0.2633,
      "num_input_tokens_seen": 15080912,
      "step": 19775
    },
    {
      "epoch": 41.12266112266112,
      "grad_norm": 0.00043932415428571403,
      "learning_rate": 0.15260346418025664,
      "loss": 0.2614,
      "num_input_tokens_seen": 15084944,
      "step": 19780
    },
    {
      "epoch": 41.13305613305613,
      "grad_norm": 0.0004289295757189393,
      "learning_rate": 0.15254456799185537,
      "loss": 0.2495,
      "num_input_tokens_seen": 15088816,
      "step": 19785
    },
    {
      "epoch": 41.143451143451145,
      "grad_norm": 0.00013837787264492363,
      "learning_rate": 0.15248567141104974,
      "loss": 0.2657,
      "num_input_tokens_seen": 15092688,
      "step": 19790
    },
    {
      "epoch": 41.15384615384615,
      "grad_norm": 0.00011153219384141266,
      "learning_rate": 0.15242677444692232,
      "loss": 0.2959,
      "num_input_tokens_seen": 15096656,
      "step": 19795
    },
    {
      "epoch": 41.16424116424116,
      "grad_norm": 0.0001644265721552074,
      "learning_rate": 0.15236787710855584,
      "loss": 0.2659,
      "num_input_tokens_seen": 15100336,
      "step": 19800
    },
    {
      "epoch": 41.16424116424116,
      "eval_loss": 0.2633832097053528,
      "eval_runtime": 13.3984,
      "eval_samples_per_second": 63.888,
      "eval_steps_per_second": 15.972,
      "num_input_tokens_seen": 15100336,
      "step": 19800
    },
    {
      "epoch": 41.17463617463618,
      "grad_norm": 0.0004443080979399383,
      "learning_rate": 0.1523089794050329,
      "loss": 0.2816,
      "num_input_tokens_seen": 15104208,
      "step": 19805
    },
    {
      "epoch": 41.185031185031185,
      "grad_norm": 0.00034143132506869733,
      "learning_rate": 0.15225008134543633,
      "loss": 0.2514,
      "num_input_tokens_seen": 15107920,
      "step": 19810
    },
    {
      "epoch": 41.195426195426194,
      "grad_norm": 0.00013911582936998457,
      "learning_rate": 0.15219118293884895,
      "loss": 0.274,
      "num_input_tokens_seen": 15111760,
      "step": 19815
    },
    {
      "epoch": 41.20582120582121,
      "grad_norm": 0.0008609531214460731,
      "learning_rate": 0.15213228419435362,
      "loss": 0.2805,
      "num_input_tokens_seen": 15115440,
      "step": 19820
    },
    {
      "epoch": 41.21621621621622,
      "grad_norm": 7.07189756212756e-05,
      "learning_rate": 0.15207338512103327,
      "loss": 0.2657,
      "num_input_tokens_seen": 15119280,
      "step": 19825
    },
    {
      "epoch": 41.226611226611226,
      "grad_norm": 0.00023028695432003587,
      "learning_rate": 0.1520144857279709,
      "loss": 0.26,
      "num_input_tokens_seen": 15123088,
      "step": 19830
    },
    {
      "epoch": 41.237006237006234,
      "grad_norm": 0.00030570439412258565,
      "learning_rate": 0.1519555860242495,
      "loss": 0.2575,
      "num_input_tokens_seen": 15126832,
      "step": 19835
    },
    {
      "epoch": 41.24740124740125,
      "grad_norm": 5.598764983005822e-05,
      "learning_rate": 0.15189668601895218,
      "loss": 0.257,
      "num_input_tokens_seen": 15130704,
      "step": 19840
    },
    {
      "epoch": 41.25779625779626,
      "grad_norm": 7.639340765308589e-05,
      "learning_rate": 0.151837785721162,
      "loss": 0.2696,
      "num_input_tokens_seen": 15134640,
      "step": 19845
    },
    {
      "epoch": 41.268191268191266,
      "grad_norm": 0.00020071813196409494,
      "learning_rate": 0.15177888513996218,
      "loss": 0.2521,
      "num_input_tokens_seen": 15138384,
      "step": 19850
    },
    {
      "epoch": 41.27858627858628,
      "grad_norm": 0.00043783418368548155,
      "learning_rate": 0.15171998428443592,
      "loss": 0.2777,
      "num_input_tokens_seen": 15142160,
      "step": 19855
    },
    {
      "epoch": 41.28898128898129,
      "grad_norm": 0.0006280025700107217,
      "learning_rate": 0.1516610831636665,
      "loss": 0.2676,
      "num_input_tokens_seen": 15145968,
      "step": 19860
    },
    {
      "epoch": 41.2993762993763,
      "grad_norm": 0.00031761726131662726,
      "learning_rate": 0.15160218178673715,
      "loss": 0.2731,
      "num_input_tokens_seen": 15149648,
      "step": 19865
    },
    {
      "epoch": 41.30977130977131,
      "grad_norm": 0.00023938054800964892,
      "learning_rate": 0.15154328016273122,
      "loss": 0.247,
      "num_input_tokens_seen": 15153360,
      "step": 19870
    },
    {
      "epoch": 41.32016632016632,
      "grad_norm": 0.0005181441665627062,
      "learning_rate": 0.1514843783007321,
      "loss": 0.2572,
      "num_input_tokens_seen": 15157136,
      "step": 19875
    },
    {
      "epoch": 41.33056133056133,
      "grad_norm": 0.00014944201393518597,
      "learning_rate": 0.15142547620982322,
      "loss": 0.2472,
      "num_input_tokens_seen": 15160912,
      "step": 19880
    },
    {
      "epoch": 41.34095634095634,
      "grad_norm": 0.0005177579587325454,
      "learning_rate": 0.15136657389908797,
      "loss": 0.2575,
      "num_input_tokens_seen": 15164784,
      "step": 19885
    },
    {
      "epoch": 41.351351351351354,
      "grad_norm": 0.00048187319771386683,
      "learning_rate": 0.15130767137760986,
      "loss": 0.2663,
      "num_input_tokens_seen": 15168656,
      "step": 19890
    },
    {
      "epoch": 41.36174636174636,
      "grad_norm": 0.0007022770587354898,
      "learning_rate": 0.15124876865447243,
      "loss": 0.2551,
      "num_input_tokens_seen": 15172528,
      "step": 19895
    },
    {
      "epoch": 41.37214137214137,
      "grad_norm": 0.00039229870890267193,
      "learning_rate": 0.15118986573875912,
      "loss": 0.2728,
      "num_input_tokens_seen": 15176432,
      "step": 19900
    },
    {
      "epoch": 41.38253638253638,
      "grad_norm": 0.0005361128132790327,
      "learning_rate": 0.15113096263955358,
      "loss": 0.2422,
      "num_input_tokens_seen": 15180240,
      "step": 19905
    },
    {
      "epoch": 41.392931392931395,
      "grad_norm": 0.0004896206082776189,
      "learning_rate": 0.1510720593659394,
      "loss": 0.2695,
      "num_input_tokens_seen": 15183792,
      "step": 19910
    },
    {
      "epoch": 41.4033264033264,
      "grad_norm": 0.00017406239931005985,
      "learning_rate": 0.15101315592700015,
      "loss": 0.2689,
      "num_input_tokens_seen": 15187600,
      "step": 19915
    },
    {
      "epoch": 41.41372141372141,
      "grad_norm": 0.0004011830023955554,
      "learning_rate": 0.15095425233181956,
      "loss": 0.2704,
      "num_input_tokens_seen": 15191440,
      "step": 19920
    },
    {
      "epoch": 41.42411642411643,
      "grad_norm": 0.00024775919155217707,
      "learning_rate": 0.15089534858948128,
      "loss": 0.2817,
      "num_input_tokens_seen": 15195312,
      "step": 19925
    },
    {
      "epoch": 41.434511434511435,
      "grad_norm": 0.0001525456755189225,
      "learning_rate": 0.15083644470906898,
      "loss": 0.2696,
      "num_input_tokens_seen": 15199216,
      "step": 19930
    },
    {
      "epoch": 41.444906444906444,
      "grad_norm": 0.0005426400457508862,
      "learning_rate": 0.1507775406996664,
      "loss": 0.2578,
      "num_input_tokens_seen": 15202992,
      "step": 19935
    },
    {
      "epoch": 41.45530145530145,
      "grad_norm": 9.243025124305859e-05,
      "learning_rate": 0.15071863657035725,
      "loss": 0.266,
      "num_input_tokens_seen": 15206832,
      "step": 19940
    },
    {
      "epoch": 41.46569646569647,
      "grad_norm": 9.435565152671188e-05,
      "learning_rate": 0.15065973233022534,
      "loss": 0.2689,
      "num_input_tokens_seen": 15210640,
      "step": 19945
    },
    {
      "epoch": 41.476091476091476,
      "grad_norm": 0.0004644471046049148,
      "learning_rate": 0.15060082798835442,
      "loss": 0.2718,
      "num_input_tokens_seen": 15214448,
      "step": 19950
    },
    {
      "epoch": 41.486486486486484,
      "grad_norm": 0.0002996270777657628,
      "learning_rate": 0.15054192355382823,
      "loss": 0.2732,
      "num_input_tokens_seen": 15218288,
      "step": 19955
    },
    {
      "epoch": 41.4968814968815,
      "grad_norm": 0.0004439016920514405,
      "learning_rate": 0.15048301903573066,
      "loss": 0.2485,
      "num_input_tokens_seen": 15222064,
      "step": 19960
    },
    {
      "epoch": 41.50727650727651,
      "grad_norm": 0.00010482939978828654,
      "learning_rate": 0.15042411444314546,
      "loss": 0.2551,
      "num_input_tokens_seen": 15225936,
      "step": 19965
    },
    {
      "epoch": 41.517671517671516,
      "grad_norm": 0.00015237348270602524,
      "learning_rate": 0.1503652097851565,
      "loss": 0.2537,
      "num_input_tokens_seen": 15229648,
      "step": 19970
    },
    {
      "epoch": 41.528066528066525,
      "grad_norm": 0.00012742483522742987,
      "learning_rate": 0.15030630507084758,
      "loss": 0.2862,
      "num_input_tokens_seen": 15233424,
      "step": 19975
    },
    {
      "epoch": 41.53846153846154,
      "grad_norm": 0.0005171297816559672,
      "learning_rate": 0.1502474003093026,
      "loss": 0.2683,
      "num_input_tokens_seen": 15237200,
      "step": 19980
    },
    {
      "epoch": 41.54885654885655,
      "grad_norm": 7.528687274316326e-05,
      "learning_rate": 0.15018849550960536,
      "loss": 0.2732,
      "num_input_tokens_seen": 15240944,
      "step": 19985
    },
    {
      "epoch": 41.55925155925156,
      "grad_norm": 6.710041634505615e-05,
      "learning_rate": 0.15012959068083975,
      "loss": 0.2649,
      "num_input_tokens_seen": 15244624,
      "step": 19990
    },
    {
      "epoch": 41.56964656964657,
      "grad_norm": 0.00038850161945447326,
      "learning_rate": 0.1500706858320896,
      "loss": 0.2466,
      "num_input_tokens_seen": 15248656,
      "step": 19995
    },
    {
      "epoch": 41.58004158004158,
      "grad_norm": 0.0003976233711000532,
      "learning_rate": 0.15001178097243886,
      "loss": 0.306,
      "num_input_tokens_seen": 15252464,
      "step": 20000
    },
    {
      "epoch": 41.58004158004158,
      "eval_loss": 0.24729791283607483,
      "eval_runtime": 13.4069,
      "eval_samples_per_second": 63.848,
      "eval_steps_per_second": 15.962,
      "num_input_tokens_seen": 15252464,
      "step": 20000
    },
    {
      "epoch": 41.59043659043659,
      "grad_norm": 0.00014859740622341633,
      "learning_rate": 0.1499528761109713,
      "loss": 0.2745,
      "num_input_tokens_seen": 15256144,
      "step": 20005
    },
    {
      "epoch": 41.6008316008316,
      "grad_norm": 0.00039404688868671656,
      "learning_rate": 0.14989397125677087,
      "loss": 0.262,
      "num_input_tokens_seen": 15260144,
      "step": 20010
    },
    {
      "epoch": 41.61122661122661,
      "grad_norm": 0.00026431374135427177,
      "learning_rate": 0.14983506641892141,
      "loss": 0.278,
      "num_input_tokens_seen": 15264016,
      "step": 20015
    },
    {
      "epoch": 41.62162162162162,
      "grad_norm": 9.346404112875462e-05,
      "learning_rate": 0.14977616160650672,
      "loss": 0.2474,
      "num_input_tokens_seen": 15267824,
      "step": 20020
    },
    {
      "epoch": 41.63201663201663,
      "grad_norm": 0.00016083351511042565,
      "learning_rate": 0.14971725682861076,
      "loss": 0.2468,
      "num_input_tokens_seen": 15271632,
      "step": 20025
    },
    {
      "epoch": 41.642411642411645,
      "grad_norm": 0.00033178890589624643,
      "learning_rate": 0.14965835209431738,
      "loss": 0.2893,
      "num_input_tokens_seen": 15275440,
      "step": 20030
    },
    {
      "epoch": 41.65280665280665,
      "grad_norm": 7.432598795276135e-05,
      "learning_rate": 0.14959944741271036,
      "loss": 0.2665,
      "num_input_tokens_seen": 15279408,
      "step": 20035
    },
    {
      "epoch": 41.66320166320166,
      "grad_norm": 0.00012657100160140544,
      "learning_rate": 0.14954054279287363,
      "loss": 0.2685,
      "num_input_tokens_seen": 15283184,
      "step": 20040
    },
    {
      "epoch": 41.67359667359668,
      "grad_norm": 0.00029358078609220684,
      "learning_rate": 0.14948163824389094,
      "loss": 0.2742,
      "num_input_tokens_seen": 15286736,
      "step": 20045
    },
    {
      "epoch": 41.683991683991685,
      "grad_norm": 0.00027662538923323154,
      "learning_rate": 0.14942273377484613,
      "loss": 0.2704,
      "num_input_tokens_seen": 15290416,
      "step": 20050
    },
    {
      "epoch": 41.694386694386694,
      "grad_norm": 0.00013197209045756608,
      "learning_rate": 0.1493638293948231,
      "loss": 0.2767,
      "num_input_tokens_seen": 15294192,
      "step": 20055
    },
    {
      "epoch": 41.7047817047817,
      "grad_norm": 0.00020125800801906735,
      "learning_rate": 0.14930492511290547,
      "loss": 0.2698,
      "num_input_tokens_seen": 15298192,
      "step": 20060
    },
    {
      "epoch": 41.71517671517672,
      "grad_norm": 0.00020611828949768096,
      "learning_rate": 0.14924602093817715,
      "loss": 0.2619,
      "num_input_tokens_seen": 15301968,
      "step": 20065
    },
    {
      "epoch": 41.725571725571726,
      "grad_norm": 0.00025062181521207094,
      "learning_rate": 0.14918711687972194,
      "loss": 0.2719,
      "num_input_tokens_seen": 15305648,
      "step": 20070
    },
    {
      "epoch": 41.735966735966734,
      "grad_norm": 0.0004917155019938946,
      "learning_rate": 0.14912821294662346,
      "loss": 0.2632,
      "num_input_tokens_seen": 15309648,
      "step": 20075
    },
    {
      "epoch": 41.74636174636175,
      "grad_norm": 0.00011816350161097944,
      "learning_rate": 0.14906930914796554,
      "loss": 0.2685,
      "num_input_tokens_seen": 15313552,
      "step": 20080
    },
    {
      "epoch": 41.75675675675676,
      "grad_norm": 0.00045738540939055383,
      "learning_rate": 0.14901040549283182,
      "loss": 0.2552,
      "num_input_tokens_seen": 15317296,
      "step": 20085
    },
    {
      "epoch": 41.767151767151766,
      "grad_norm": 0.00012525790953077376,
      "learning_rate": 0.148951501990306,
      "loss": 0.2594,
      "num_input_tokens_seen": 15321072,
      "step": 20090
    },
    {
      "epoch": 41.777546777546775,
      "grad_norm": 0.00028359986026771367,
      "learning_rate": 0.14889259864947177,
      "loss": 0.2771,
      "num_input_tokens_seen": 15324880,
      "step": 20095
    },
    {
      "epoch": 41.78794178794179,
      "grad_norm": 0.0008753768051974475,
      "learning_rate": 0.14883369547941272,
      "loss": 0.2706,
      "num_input_tokens_seen": 15328720,
      "step": 20100
    },
    {
      "epoch": 41.7983367983368,
      "grad_norm": 7.559965160908177e-05,
      "learning_rate": 0.14877479248921247,
      "loss": 0.2477,
      "num_input_tokens_seen": 15332400,
      "step": 20105
    },
    {
      "epoch": 41.80873180873181,
      "grad_norm": 0.00014708726666867733,
      "learning_rate": 0.14871588968795468,
      "loss": 0.2625,
      "num_input_tokens_seen": 15336272,
      "step": 20110
    },
    {
      "epoch": 41.81912681912682,
      "grad_norm": 0.00046606335672549903,
      "learning_rate": 0.1486569870847228,
      "loss": 0.2536,
      "num_input_tokens_seen": 15340016,
      "step": 20115
    },
    {
      "epoch": 41.82952182952183,
      "grad_norm": 0.0002363435341976583,
      "learning_rate": 0.1485980846886004,
      "loss": 0.2489,
      "num_input_tokens_seen": 15343856,
      "step": 20120
    },
    {
      "epoch": 41.83991683991684,
      "grad_norm": 0.0005542057915590703,
      "learning_rate": 0.14853918250867096,
      "loss": 0.2674,
      "num_input_tokens_seen": 15347664,
      "step": 20125
    },
    {
      "epoch": 41.85031185031185,
      "grad_norm": 0.0008636984275653958,
      "learning_rate": 0.1484802805540179,
      "loss": 0.2848,
      "num_input_tokens_seen": 15351440,
      "step": 20130
    },
    {
      "epoch": 41.86070686070686,
      "grad_norm": 0.0004019705520477146,
      "learning_rate": 0.14842137883372472,
      "loss": 0.2838,
      "num_input_tokens_seen": 15355440,
      "step": 20135
    },
    {
      "epoch": 41.87110187110187,
      "grad_norm": 0.0004886981914751232,
      "learning_rate": 0.14836247735687474,
      "loss": 0.2834,
      "num_input_tokens_seen": 15359280,
      "step": 20140
    },
    {
      "epoch": 41.88149688149688,
      "grad_norm": 0.0001917227345984429,
      "learning_rate": 0.14830357613255132,
      "loss": 0.2815,
      "num_input_tokens_seen": 15362992,
      "step": 20145
    },
    {
      "epoch": 41.891891891891895,
      "grad_norm": 0.0002934641670435667,
      "learning_rate": 0.1482446751698378,
      "loss": 0.2748,
      "num_input_tokens_seen": 15366864,
      "step": 20150
    },
    {
      "epoch": 41.9022869022869,
      "grad_norm": 5.305219747242518e-05,
      "learning_rate": 0.14818577447781744,
      "loss": 0.2646,
      "num_input_tokens_seen": 15370736,
      "step": 20155
    },
    {
      "epoch": 41.91268191268191,
      "grad_norm": 0.00040848529897630215,
      "learning_rate": 0.14812687406557346,
      "loss": 0.2949,
      "num_input_tokens_seen": 15374448,
      "step": 20160
    },
    {
      "epoch": 41.92307692307692,
      "grad_norm": 0.0002495793451089412,
      "learning_rate": 0.14806797394218899,
      "loss": 0.2682,
      "num_input_tokens_seen": 15378160,
      "step": 20165
    },
    {
      "epoch": 41.933471933471935,
      "grad_norm": 0.00020230650261510164,
      "learning_rate": 0.1480090741167472,
      "loss": 0.2822,
      "num_input_tokens_seen": 15381936,
      "step": 20170
    },
    {
      "epoch": 41.943866943866944,
      "grad_norm": 0.00026740634348243475,
      "learning_rate": 0.1479501745983313,
      "loss": 0.2721,
      "num_input_tokens_seen": 15385680,
      "step": 20175
    },
    {
      "epoch": 41.95426195426195,
      "grad_norm": 0.0004154109919909388,
      "learning_rate": 0.14789127539602415,
      "loss": 0.2604,
      "num_input_tokens_seen": 15389456,
      "step": 20180
    },
    {
      "epoch": 41.96465696465697,
      "grad_norm": 0.00041083982796408236,
      "learning_rate": 0.14783237651890885,
      "loss": 0.2635,
      "num_input_tokens_seen": 15393456,
      "step": 20185
    },
    {
      "epoch": 41.975051975051976,
      "grad_norm": 0.00016418518498539925,
      "learning_rate": 0.14777347797606838,
      "loss": 0.2865,
      "num_input_tokens_seen": 15397328,
      "step": 20190
    },
    {
      "epoch": 41.985446985446984,
      "grad_norm": 0.0005811268929392099,
      "learning_rate": 0.14771457977658553,
      "loss": 0.2779,
      "num_input_tokens_seen": 15401168,
      "step": 20195
    },
    {
      "epoch": 41.99584199584199,
      "grad_norm": 0.00013299698184709996,
      "learning_rate": 0.14765568192954326,
      "loss": 0.2733,
      "num_input_tokens_seen": 15404912,
      "step": 20200
    },
    {
      "epoch": 41.99584199584199,
      "eval_loss": 0.26801010966300964,
      "eval_runtime": 13.4426,
      "eval_samples_per_second": 63.678,
      "eval_steps_per_second": 15.92,
      "num_input_tokens_seen": 15404912,
      "step": 20200
    },
    {
      "epoch": 42.00623700623701,
      "grad_norm": 0.00031006947392597795,
      "learning_rate": 0.14759678444402421,
      "loss": 0.278,
      "num_input_tokens_seen": 15408760,
      "step": 20205
    },
    {
      "epoch": 42.016632016632016,
      "grad_norm": 9.294260962633416e-05,
      "learning_rate": 0.14753788732911122,
      "loss": 0.2859,
      "num_input_tokens_seen": 15412568,
      "step": 20210
    },
    {
      "epoch": 42.027027027027025,
      "grad_norm": 4.357831130619161e-05,
      "learning_rate": 0.147478990593887,
      "loss": 0.257,
      "num_input_tokens_seen": 15416312,
      "step": 20215
    },
    {
      "epoch": 42.03742203742204,
      "grad_norm": 0.000271494995104149,
      "learning_rate": 0.14742009424743405,
      "loss": 0.2626,
      "num_input_tokens_seen": 15420056,
      "step": 20220
    },
    {
      "epoch": 42.04781704781705,
      "grad_norm": 0.0005096501554362476,
      "learning_rate": 0.14736119829883504,
      "loss": 0.2538,
      "num_input_tokens_seen": 15423992,
      "step": 20225
    },
    {
      "epoch": 42.05821205821206,
      "grad_norm": 0.0002849528391379863,
      "learning_rate": 0.14730230275717243,
      "loss": 0.2684,
      "num_input_tokens_seen": 15427992,
      "step": 20230
    },
    {
      "epoch": 42.06860706860707,
      "grad_norm": 5.8325760619482026e-05,
      "learning_rate": 0.14724340763152854,
      "loss": 0.2771,
      "num_input_tokens_seen": 15431704,
      "step": 20235
    },
    {
      "epoch": 42.07900207900208,
      "grad_norm": 0.0004823118506465107,
      "learning_rate": 0.14718451293098594,
      "loss": 0.2547,
      "num_input_tokens_seen": 15435672,
      "step": 20240
    },
    {
      "epoch": 42.08939708939709,
      "grad_norm": 0.0007917492766864598,
      "learning_rate": 0.14712561866462676,
      "loss": 0.2757,
      "num_input_tokens_seen": 15439512,
      "step": 20245
    },
    {
      "epoch": 42.0997920997921,
      "grad_norm": 0.00016996277554426342,
      "learning_rate": 0.1470667248415333,
      "loss": 0.2724,
      "num_input_tokens_seen": 15443256,
      "step": 20250
    },
    {
      "epoch": 42.11018711018711,
      "grad_norm": 0.00012539848103187978,
      "learning_rate": 0.1470078314707878,
      "loss": 0.269,
      "num_input_tokens_seen": 15447032,
      "step": 20255
    },
    {
      "epoch": 42.12058212058212,
      "grad_norm": 0.00018367639859206975,
      "learning_rate": 0.14694893856147223,
      "loss": 0.2693,
      "num_input_tokens_seen": 15450808,
      "step": 20260
    },
    {
      "epoch": 42.13097713097713,
      "grad_norm": 0.0003650474245660007,
      "learning_rate": 0.14689004612266868,
      "loss": 0.2738,
      "num_input_tokens_seen": 15454712,
      "step": 20265
    },
    {
      "epoch": 42.141372141372145,
      "grad_norm": 0.00014819738862570375,
      "learning_rate": 0.14683115416345913,
      "loss": 0.2643,
      "num_input_tokens_seen": 15458488,
      "step": 20270
    },
    {
      "epoch": 42.15176715176715,
      "grad_norm": 0.00036469101905822754,
      "learning_rate": 0.1467722626929254,
      "loss": 0.2737,
      "num_input_tokens_seen": 15462360,
      "step": 20275
    },
    {
      "epoch": 42.16216216216216,
      "grad_norm": 4.69807127956301e-05,
      "learning_rate": 0.14671337172014937,
      "loss": 0.2487,
      "num_input_tokens_seen": 15466232,
      "step": 20280
    },
    {
      "epoch": 42.17255717255717,
      "grad_norm": 0.0005535532836802304,
      "learning_rate": 0.14665448125421265,
      "loss": 0.2879,
      "num_input_tokens_seen": 15470200,
      "step": 20285
    },
    {
      "epoch": 42.182952182952185,
      "grad_norm": 0.00025059442850761116,
      "learning_rate": 0.146595591304197,
      "loss": 0.2658,
      "num_input_tokens_seen": 15474104,
      "step": 20290
    },
    {
      "epoch": 42.19334719334719,
      "grad_norm": 0.00022680446272715926,
      "learning_rate": 0.14653670187918397,
      "loss": 0.2669,
      "num_input_tokens_seen": 15477816,
      "step": 20295
    },
    {
      "epoch": 42.2037422037422,
      "grad_norm": 0.00017650314839556813,
      "learning_rate": 0.14647781298825502,
      "loss": 0.2684,
      "num_input_tokens_seen": 15481592,
      "step": 20300
    },
    {
      "epoch": 42.21413721413722,
      "grad_norm": 9.735923958942294e-05,
      "learning_rate": 0.14641892464049153,
      "loss": 0.2544,
      "num_input_tokens_seen": 15485368,
      "step": 20305
    },
    {
      "epoch": 42.224532224532226,
      "grad_norm": 6.39576610410586e-05,
      "learning_rate": 0.14636003684497495,
      "loss": 0.2842,
      "num_input_tokens_seen": 15489240,
      "step": 20310
    },
    {
      "epoch": 42.234927234927234,
      "grad_norm": 0.0002896163787227124,
      "learning_rate": 0.14630114961078636,
      "loss": 0.2651,
      "num_input_tokens_seen": 15493112,
      "step": 20315
    },
    {
      "epoch": 42.24532224532224,
      "grad_norm": 0.0006976997246965766,
      "learning_rate": 0.14624226294700704,
      "loss": 0.2497,
      "num_input_tokens_seen": 15496984,
      "step": 20320
    },
    {
      "epoch": 42.25571725571726,
      "grad_norm": 0.0006990879774093628,
      "learning_rate": 0.14618337686271793,
      "loss": 0.3222,
      "num_input_tokens_seen": 15500696,
      "step": 20325
    },
    {
      "epoch": 42.266112266112266,
      "grad_norm": 0.000442896707681939,
      "learning_rate": 0.1461244913670001,
      "loss": 0.275,
      "num_input_tokens_seen": 15504536,
      "step": 20330
    },
    {
      "epoch": 42.276507276507274,
      "grad_norm": 0.00045019612298347056,
      "learning_rate": 0.1460656064689344,
      "loss": 0.2586,
      "num_input_tokens_seen": 15508152,
      "step": 20335
    },
    {
      "epoch": 42.28690228690229,
      "grad_norm": 0.00014745046792086214,
      "learning_rate": 0.14600672217760163,
      "loss": 0.2578,
      "num_input_tokens_seen": 15512024,
      "step": 20340
    },
    {
      "epoch": 42.2972972972973,
      "grad_norm": 0.00022589118452742696,
      "learning_rate": 0.14594783850208248,
      "loss": 0.2269,
      "num_input_tokens_seen": 15515608,
      "step": 20345
    },
    {
      "epoch": 42.30769230769231,
      "grad_norm": 1.6374231563531794e-05,
      "learning_rate": 0.14588895545145758,
      "loss": 0.2678,
      "num_input_tokens_seen": 15519448,
      "step": 20350
    },
    {
      "epoch": 42.318087318087315,
      "grad_norm": 0.0001519451179774478,
      "learning_rate": 0.14583007303480738,
      "loss": 0.2742,
      "num_input_tokens_seen": 15523064,
      "step": 20355
    },
    {
      "epoch": 42.32848232848233,
      "grad_norm": 3.054229455301538e-05,
      "learning_rate": 0.14577119126121235,
      "loss": 0.2743,
      "num_input_tokens_seen": 15526840,
      "step": 20360
    },
    {
      "epoch": 42.33887733887734,
      "grad_norm": 0.00022006033395882696,
      "learning_rate": 0.14571231013975272,
      "loss": 0.2557,
      "num_input_tokens_seen": 15530616,
      "step": 20365
    },
    {
      "epoch": 42.34927234927235,
      "grad_norm": 0.0002961000718642026,
      "learning_rate": 0.1456534296795088,
      "loss": 0.2625,
      "num_input_tokens_seen": 15534520,
      "step": 20370
    },
    {
      "epoch": 42.35966735966736,
      "grad_norm": 9.562297782395035e-05,
      "learning_rate": 0.14559454988956066,
      "loss": 0.2679,
      "num_input_tokens_seen": 15538360,
      "step": 20375
    },
    {
      "epoch": 42.37006237006237,
      "grad_norm": 0.0006060078158043325,
      "learning_rate": 0.1455356707789882,
      "loss": 0.2767,
      "num_input_tokens_seen": 15541976,
      "step": 20380
    },
    {
      "epoch": 42.38045738045738,
      "grad_norm": 0.0003607240505516529,
      "learning_rate": 0.14547679235687147,
      "loss": 0.2382,
      "num_input_tokens_seen": 15545784,
      "step": 20385
    },
    {
      "epoch": 42.39085239085239,
      "grad_norm": 0.00011270622053416446,
      "learning_rate": 0.14541791463229023,
      "loss": 0.2718,
      "num_input_tokens_seen": 15549592,
      "step": 20390
    },
    {
      "epoch": 42.4012474012474,
      "grad_norm": 0.00017741376359481364,
      "learning_rate": 0.14535903761432406,
      "loss": 0.2586,
      "num_input_tokens_seen": 15553368,
      "step": 20395
    },
    {
      "epoch": 42.41164241164241,
      "grad_norm": 8.503738354193047e-05,
      "learning_rate": 0.1453001613120527,
      "loss": 0.2541,
      "num_input_tokens_seen": 15557176,
      "step": 20400
    },
    {
      "epoch": 42.41164241164241,
      "eval_loss": 0.2488873153924942,
      "eval_runtime": 13.399,
      "eval_samples_per_second": 63.886,
      "eval_steps_per_second": 15.971,
      "num_input_tokens_seen": 15557176,
      "step": 20400
    },
    {
      "epoch": 42.42203742203742,
      "grad_norm": 0.00014495462528429925,
      "learning_rate": 0.14524128573455547,
      "loss": 0.2615,
      "num_input_tokens_seen": 15561112,
      "step": 20405
    },
    {
      "epoch": 42.432432432432435,
      "grad_norm": 0.0005751390708610415,
      "learning_rate": 0.14518241089091177,
      "loss": 0.2665,
      "num_input_tokens_seen": 15564984,
      "step": 20410
    },
    {
      "epoch": 42.44282744282744,
      "grad_norm": 0.00036329164868220687,
      "learning_rate": 0.1451235367902009,
      "loss": 0.278,
      "num_input_tokens_seen": 15568952,
      "step": 20415
    },
    {
      "epoch": 42.45322245322245,
      "grad_norm": 0.0006299822707660496,
      "learning_rate": 0.1450646634415019,
      "loss": 0.2688,
      "num_input_tokens_seen": 15572696,
      "step": 20420
    },
    {
      "epoch": 42.46361746361746,
      "grad_norm": 0.0002691232948563993,
      "learning_rate": 0.1450057908538938,
      "loss": 0.2781,
      "num_input_tokens_seen": 15576568,
      "step": 20425
    },
    {
      "epoch": 42.474012474012476,
      "grad_norm": 0.0002322325890418142,
      "learning_rate": 0.14494691903645557,
      "loss": 0.2768,
      "num_input_tokens_seen": 15580536,
      "step": 20430
    },
    {
      "epoch": 42.484407484407484,
      "grad_norm": 0.00011574888048926368,
      "learning_rate": 0.14488804799826588,
      "loss": 0.27,
      "num_input_tokens_seen": 15584312,
      "step": 20435
    },
    {
      "epoch": 42.49480249480249,
      "grad_norm": 0.00023743572819512337,
      "learning_rate": 0.14482917774840348,
      "loss": 0.2532,
      "num_input_tokens_seen": 15588024,
      "step": 20440
    },
    {
      "epoch": 42.50519750519751,
      "grad_norm": 0.00011074316716985777,
      "learning_rate": 0.14477030829594684,
      "loss": 0.2726,
      "num_input_tokens_seen": 15591928,
      "step": 20445
    },
    {
      "epoch": 42.515592515592516,
      "grad_norm": 9.598352335160598e-05,
      "learning_rate": 0.14471143964997432,
      "loss": 0.2704,
      "num_input_tokens_seen": 15595800,
      "step": 20450
    },
    {
      "epoch": 42.525987525987524,
      "grad_norm": 0.00012516751303337514,
      "learning_rate": 0.14465257181956434,
      "loss": 0.2755,
      "num_input_tokens_seen": 15599704,
      "step": 20455
    },
    {
      "epoch": 42.53638253638254,
      "grad_norm": 0.0007414927240461111,
      "learning_rate": 0.1445937048137949,
      "loss": 0.266,
      "num_input_tokens_seen": 15603544,
      "step": 20460
    },
    {
      "epoch": 42.54677754677755,
      "grad_norm": 3.4345830499660224e-05,
      "learning_rate": 0.14453483864174416,
      "loss": 0.2767,
      "num_input_tokens_seen": 15607224,
      "step": 20465
    },
    {
      "epoch": 42.55717255717256,
      "grad_norm": 0.0004697073600254953,
      "learning_rate": 0.14447597331249,
      "loss": 0.2602,
      "num_input_tokens_seen": 15611064,
      "step": 20470
    },
    {
      "epoch": 42.567567567567565,
      "grad_norm": 0.00012706419511232525,
      "learning_rate": 0.1444171088351102,
      "loss": 0.2526,
      "num_input_tokens_seen": 15614968,
      "step": 20475
    },
    {
      "epoch": 42.57796257796258,
      "grad_norm": 0.000415144138969481,
      "learning_rate": 0.14435824521868235,
      "loss": 0.2632,
      "num_input_tokens_seen": 15618712,
      "step": 20480
    },
    {
      "epoch": 42.58835758835759,
      "grad_norm": 0.00030475438688881695,
      "learning_rate": 0.14429938247228397,
      "loss": 0.2924,
      "num_input_tokens_seen": 15622552,
      "step": 20485
    },
    {
      "epoch": 42.5987525987526,
      "grad_norm": 0.00022947760589886457,
      "learning_rate": 0.14424052060499243,
      "loss": 0.2628,
      "num_input_tokens_seen": 15626392,
      "step": 20490
    },
    {
      "epoch": 42.60914760914761,
      "grad_norm": 0.0003667368437163532,
      "learning_rate": 0.14418165962588506,
      "loss": 0.2653,
      "num_input_tokens_seen": 15630072,
      "step": 20495
    },
    {
      "epoch": 42.61954261954262,
      "grad_norm": 0.0002552943187765777,
      "learning_rate": 0.1441227995440388,
      "loss": 0.2567,
      "num_input_tokens_seen": 15633944,
      "step": 20500
    },
    {
      "epoch": 42.62993762993763,
      "grad_norm": 0.0002152569213649258,
      "learning_rate": 0.14406394036853082,
      "loss": 0.252,
      "num_input_tokens_seen": 15637848,
      "step": 20505
    },
    {
      "epoch": 42.64033264033264,
      "grad_norm": 0.0001386357325827703,
      "learning_rate": 0.14400508210843774,
      "loss": 0.2963,
      "num_input_tokens_seen": 15641720,
      "step": 20510
    },
    {
      "epoch": 42.65072765072765,
      "grad_norm": 0.00015914726827759296,
      "learning_rate": 0.1439462247728364,
      "loss": 0.271,
      "num_input_tokens_seen": 15645496,
      "step": 20515
    },
    {
      "epoch": 42.66112266112266,
      "grad_norm": 0.00020199711434543133,
      "learning_rate": 0.14388736837080326,
      "loss": 0.2591,
      "num_input_tokens_seen": 15649208,
      "step": 20520
    },
    {
      "epoch": 42.67151767151767,
      "grad_norm": 7.850349356886e-05,
      "learning_rate": 0.14382851291141469,
      "loss": 0.2786,
      "num_input_tokens_seen": 15653048,
      "step": 20525
    },
    {
      "epoch": 42.681912681912685,
      "grad_norm": 5.353394226403907e-05,
      "learning_rate": 0.14376965840374697,
      "loss": 0.2544,
      "num_input_tokens_seen": 15656856,
      "step": 20530
    },
    {
      "epoch": 42.69230769230769,
      "grad_norm": 0.00013941607903689146,
      "learning_rate": 0.14371080485687632,
      "loss": 0.2533,
      "num_input_tokens_seen": 15660664,
      "step": 20535
    },
    {
      "epoch": 42.7027027027027,
      "grad_norm": 6.965376087464392e-05,
      "learning_rate": 0.1436519522798785,
      "loss": 0.2421,
      "num_input_tokens_seen": 15664376,
      "step": 20540
    },
    {
      "epoch": 42.71309771309771,
      "grad_norm": 0.00013098256022203714,
      "learning_rate": 0.14359310068182948,
      "loss": 0.2805,
      "num_input_tokens_seen": 15668184,
      "step": 20545
    },
    {
      "epoch": 42.723492723492726,
      "grad_norm": 0.00017030752496793866,
      "learning_rate": 0.14353425007180484,
      "loss": 0.2301,
      "num_input_tokens_seen": 15671896,
      "step": 20550
    },
    {
      "epoch": 42.733887733887734,
      "grad_norm": 0.0005642463220283389,
      "learning_rate": 0.14347540045888005,
      "loss": 0.2759,
      "num_input_tokens_seen": 15675672,
      "step": 20555
    },
    {
      "epoch": 42.74428274428274,
      "grad_norm": 0.00025218207156285644,
      "learning_rate": 0.14341655185213056,
      "loss": 0.2658,
      "num_input_tokens_seen": 15679384,
      "step": 20560
    },
    {
      "epoch": 42.75467775467776,
      "grad_norm": 0.00020625402976293117,
      "learning_rate": 0.14335770426063144,
      "loss": 0.2579,
      "num_input_tokens_seen": 15683448,
      "step": 20565
    },
    {
      "epoch": 42.765072765072766,
      "grad_norm": 0.00019057997269555926,
      "learning_rate": 0.1432988576934578,
      "loss": 0.2433,
      "num_input_tokens_seen": 15687224,
      "step": 20570
    },
    {
      "epoch": 42.775467775467774,
      "grad_norm": 0.00020051821775268763,
      "learning_rate": 0.14324001215968457,
      "loss": 0.2732,
      "num_input_tokens_seen": 15690968,
      "step": 20575
    },
    {
      "epoch": 42.78586278586278,
      "grad_norm": 6.650371506111696e-05,
      "learning_rate": 0.14318116766838637,
      "loss": 0.2739,
      "num_input_tokens_seen": 15694712,
      "step": 20580
    },
    {
      "epoch": 42.7962577962578,
      "grad_norm": 6.460359873017296e-05,
      "learning_rate": 0.14312232422863788,
      "loss": 0.2549,
      "num_input_tokens_seen": 15698584,
      "step": 20585
    },
    {
      "epoch": 42.80665280665281,
      "grad_norm": 5.0154740165453404e-05,
      "learning_rate": 0.14306348184951334,
      "loss": 0.2665,
      "num_input_tokens_seen": 15702360,
      "step": 20590
    },
    {
      "epoch": 42.817047817047815,
      "grad_norm": 0.000191645827726461,
      "learning_rate": 0.1430046405400871,
      "loss": 0.2626,
      "num_input_tokens_seen": 15706168,
      "step": 20595
    },
    {
      "epoch": 42.82744282744283,
      "grad_norm": 0.0001799809542717412,
      "learning_rate": 0.14294580030943324,
      "loss": 0.2715,
      "num_input_tokens_seen": 15709912,
      "step": 20600
    },
    {
      "epoch": 42.82744282744283,
      "eval_loss": 0.2510470747947693,
      "eval_runtime": 13.4238,
      "eval_samples_per_second": 63.767,
      "eval_steps_per_second": 15.942,
      "num_input_tokens_seen": 15709912,
      "step": 20600
    },
    {
      "epoch": 42.83783783783784,
      "grad_norm": 0.0003508008667267859,
      "learning_rate": 0.14288696116662553,
      "loss": 0.2921,
      "num_input_tokens_seen": 15713752,
      "step": 20605
    },
    {
      "epoch": 42.84823284823285,
      "grad_norm": 0.0002212283870903775,
      "learning_rate": 0.1428281231207378,
      "loss": 0.2492,
      "num_input_tokens_seen": 15717656,
      "step": 20610
    },
    {
      "epoch": 42.858627858627855,
      "grad_norm": 0.00030565151246264577,
      "learning_rate": 0.1427692861808437,
      "loss": 0.2732,
      "num_input_tokens_seen": 15721528,
      "step": 20615
    },
    {
      "epoch": 42.86902286902287,
      "grad_norm": 0.0005401631933636963,
      "learning_rate": 0.1427104503560165,
      "loss": 0.2434,
      "num_input_tokens_seen": 15725432,
      "step": 20620
    },
    {
      "epoch": 42.87941787941788,
      "grad_norm": 0.0001575544592924416,
      "learning_rate": 0.14265161565532947,
      "loss": 0.2421,
      "num_input_tokens_seen": 15729272,
      "step": 20625
    },
    {
      "epoch": 42.88981288981289,
      "grad_norm": 0.00032919671502895653,
      "learning_rate": 0.14259278208785564,
      "loss": 0.2902,
      "num_input_tokens_seen": 15733112,
      "step": 20630
    },
    {
      "epoch": 42.9002079002079,
      "grad_norm": 0.00030925741884857416,
      "learning_rate": 0.14253394966266789,
      "loss": 0.2714,
      "num_input_tokens_seen": 15737016,
      "step": 20635
    },
    {
      "epoch": 42.91060291060291,
      "grad_norm": 0.00011206260387552902,
      "learning_rate": 0.14247511838883894,
      "loss": 0.2575,
      "num_input_tokens_seen": 15740792,
      "step": 20640
    },
    {
      "epoch": 42.92099792099792,
      "grad_norm": 0.00021110885427333415,
      "learning_rate": 0.14241628827544126,
      "loss": 0.2626,
      "num_input_tokens_seen": 15744600,
      "step": 20645
    },
    {
      "epoch": 42.931392931392935,
      "grad_norm": 6.634168676100671e-05,
      "learning_rate": 0.14235745933154723,
      "loss": 0.2642,
      "num_input_tokens_seen": 15748280,
      "step": 20650
    },
    {
      "epoch": 42.94178794178794,
      "grad_norm": 0.0002306604728801176,
      "learning_rate": 0.14229863156622907,
      "loss": 0.2787,
      "num_input_tokens_seen": 15752120,
      "step": 20655
    },
    {
      "epoch": 42.95218295218295,
      "grad_norm": 0.00016673390928190202,
      "learning_rate": 0.14223980498855868,
      "loss": 0.2656,
      "num_input_tokens_seen": 15755864,
      "step": 20660
    },
    {
      "epoch": 42.96257796257796,
      "grad_norm": 0.0002595261612441391,
      "learning_rate": 0.14218097960760792,
      "loss": 0.2578,
      "num_input_tokens_seen": 15759704,
      "step": 20665
    },
    {
      "epoch": 42.972972972972975,
      "grad_norm": 0.0002815193438436836,
      "learning_rate": 0.1421221554324483,
      "loss": 0.2898,
      "num_input_tokens_seen": 15763480,
      "step": 20670
    },
    {
      "epoch": 42.983367983367984,
      "grad_norm": 0.00032388733234256506,
      "learning_rate": 0.1420633324721513,
      "loss": 0.29,
      "num_input_tokens_seen": 15767224,
      "step": 20675
    },
    {
      "epoch": 42.99376299376299,
      "grad_norm": 0.00013253749057184905,
      "learning_rate": 0.14200451073578824,
      "loss": 0.2703,
      "num_input_tokens_seen": 15771000,
      "step": 20680
    },
    {
      "epoch": 43.00415800415801,
      "grad_norm": 0.00013155497435946018,
      "learning_rate": 0.14194569023243003,
      "loss": 0.2628,
      "num_input_tokens_seen": 15774816,
      "step": 20685
    },
    {
      "epoch": 43.014553014553016,
      "grad_norm": 7.800347520969808e-05,
      "learning_rate": 0.14188687097114766,
      "loss": 0.2649,
      "num_input_tokens_seen": 15778592,
      "step": 20690
    },
    {
      "epoch": 43.024948024948024,
      "grad_norm": 7.313511741813272e-05,
      "learning_rate": 0.14182805296101172,
      "loss": 0.2566,
      "num_input_tokens_seen": 15782496,
      "step": 20695
    },
    {
      "epoch": 43.03534303534303,
      "grad_norm": 0.0006190743879415095,
      "learning_rate": 0.14176923621109272,
      "loss": 0.264,
      "num_input_tokens_seen": 15786368,
      "step": 20700
    },
    {
      "epoch": 43.04573804573805,
      "grad_norm": 0.00037532905116677284,
      "learning_rate": 0.14171042073046097,
      "loss": 0.2925,
      "num_input_tokens_seen": 15790144,
      "step": 20705
    },
    {
      "epoch": 43.056133056133056,
      "grad_norm": 0.00034636480268090963,
      "learning_rate": 0.14165160652818642,
      "loss": 0.2753,
      "num_input_tokens_seen": 15793984,
      "step": 20710
    },
    {
      "epoch": 43.066528066528065,
      "grad_norm": 0.0009093452244997025,
      "learning_rate": 0.14159279361333907,
      "loss": 0.2687,
      "num_input_tokens_seen": 15797760,
      "step": 20715
    },
    {
      "epoch": 43.07692307692308,
      "grad_norm": 0.00019883294589817524,
      "learning_rate": 0.14153398199498868,
      "loss": 0.2828,
      "num_input_tokens_seen": 15801600,
      "step": 20720
    },
    {
      "epoch": 43.08731808731809,
      "grad_norm": 0.0004070269060321152,
      "learning_rate": 0.14147517168220458,
      "loss": 0.2499,
      "num_input_tokens_seen": 15805216,
      "step": 20725
    },
    {
      "epoch": 43.0977130977131,
      "grad_norm": 0.00012920625158585608,
      "learning_rate": 0.14141636268405616,
      "loss": 0.262,
      "num_input_tokens_seen": 15808928,
      "step": 20730
    },
    {
      "epoch": 43.108108108108105,
      "grad_norm": 5.8282905229134485e-05,
      "learning_rate": 0.14135755500961253,
      "loss": 0.2624,
      "num_input_tokens_seen": 15812800,
      "step": 20735
    },
    {
      "epoch": 43.11850311850312,
      "grad_norm": 0.00021145261416677386,
      "learning_rate": 0.14129874866794245,
      "loss": 0.2564,
      "num_input_tokens_seen": 15816576,
      "step": 20740
    },
    {
      "epoch": 43.12889812889813,
      "grad_norm": 0.00025781645672395825,
      "learning_rate": 0.14123994366811476,
      "loss": 0.2702,
      "num_input_tokens_seen": 15820320,
      "step": 20745
    },
    {
      "epoch": 43.13929313929314,
      "grad_norm": 0.0002806819975376129,
      "learning_rate": 0.14118114001919774,
      "loss": 0.2827,
      "num_input_tokens_seen": 15824096,
      "step": 20750
    },
    {
      "epoch": 43.14968814968815,
      "grad_norm": 0.00036128226201981306,
      "learning_rate": 0.14112233773025978,
      "loss": 0.2599,
      "num_input_tokens_seen": 15827840,
      "step": 20755
    },
    {
      "epoch": 43.16008316008316,
      "grad_norm": 0.00016363158647436649,
      "learning_rate": 0.14106353681036896,
      "loss": 0.2678,
      "num_input_tokens_seen": 15831584,
      "step": 20760
    },
    {
      "epoch": 43.17047817047817,
      "grad_norm": 0.00034647106076590717,
      "learning_rate": 0.14100473726859303,
      "loss": 0.2771,
      "num_input_tokens_seen": 15835200,
      "step": 20765
    },
    {
      "epoch": 43.18087318087318,
      "grad_norm": 0.00036992289824411273,
      "learning_rate": 0.14094593911399964,
      "loss": 0.2824,
      "num_input_tokens_seen": 15839104,
      "step": 20770
    },
    {
      "epoch": 43.19126819126819,
      "grad_norm": 0.00013782289170194417,
      "learning_rate": 0.14088714235565625,
      "loss": 0.2724,
      "num_input_tokens_seen": 15843008,
      "step": 20775
    },
    {
      "epoch": 43.2016632016632,
      "grad_norm": 0.00027012662030756474,
      "learning_rate": 0.14082834700263,
      "loss": 0.2624,
      "num_input_tokens_seen": 15846752,
      "step": 20780
    },
    {
      "epoch": 43.21205821205821,
      "grad_norm": 0.0004066957044415176,
      "learning_rate": 0.14076955306398795,
      "loss": 0.278,
      "num_input_tokens_seen": 15850784,
      "step": 20785
    },
    {
      "epoch": 43.222453222453225,
      "grad_norm": 0.0003724557172972709,
      "learning_rate": 0.14071076054879675,
      "loss": 0.2888,
      "num_input_tokens_seen": 15854592,
      "step": 20790
    },
    {
      "epoch": 43.232848232848234,
      "grad_norm": 0.00025916879530996084,
      "learning_rate": 0.14065196946612302,
      "loss": 0.2562,
      "num_input_tokens_seen": 15858400,
      "step": 20795
    },
    {
      "epoch": 43.24324324324324,
      "grad_norm": 9.486638009548187e-05,
      "learning_rate": 0.1405931798250331,
      "loss": 0.2508,
      "num_input_tokens_seen": 15862336,
      "step": 20800
    },
    {
      "epoch": 43.24324324324324,
      "eval_loss": 0.24934041500091553,
      "eval_runtime": 13.4151,
      "eval_samples_per_second": 63.809,
      "eval_steps_per_second": 15.952,
      "num_input_tokens_seen": 15862336,
      "step": 20800
    },
    {
      "epoch": 43.25363825363825,
      "grad_norm": 0.0001949878642335534,
      "learning_rate": 0.14053439163459308,
      "loss": 0.2596,
      "num_input_tokens_seen": 15866080,
      "step": 20805
    },
    {
      "epoch": 43.264033264033266,
      "grad_norm": 6.169541302369907e-05,
      "learning_rate": 0.14047560490386876,
      "loss": 0.2953,
      "num_input_tokens_seen": 15869952,
      "step": 20810
    },
    {
      "epoch": 43.274428274428274,
      "grad_norm": 0.00015231332508847117,
      "learning_rate": 0.14041681964192593,
      "loss": 0.2465,
      "num_input_tokens_seen": 15873792,
      "step": 20815
    },
    {
      "epoch": 43.28482328482328,
      "grad_norm": 0.00011879668454639614,
      "learning_rate": 0.14035803585782988,
      "loss": 0.2698,
      "num_input_tokens_seen": 15877728,
      "step": 20820
    },
    {
      "epoch": 43.2952182952183,
      "grad_norm": 0.0001493665622547269,
      "learning_rate": 0.14029925356064593,
      "loss": 0.2731,
      "num_input_tokens_seen": 15881504,
      "step": 20825
    },
    {
      "epoch": 43.305613305613306,
      "grad_norm": 0.0001392518461216241,
      "learning_rate": 0.1402404727594389,
      "loss": 0.2603,
      "num_input_tokens_seen": 15885280,
      "step": 20830
    },
    {
      "epoch": 43.316008316008315,
      "grad_norm": 6.962760380702093e-05,
      "learning_rate": 0.1401816934632737,
      "loss": 0.2392,
      "num_input_tokens_seen": 15889216,
      "step": 20835
    },
    {
      "epoch": 43.32640332640332,
      "grad_norm": 0.00012231871369294822,
      "learning_rate": 0.1401229156812147,
      "loss": 0.2309,
      "num_input_tokens_seen": 15892896,
      "step": 20840
    },
    {
      "epoch": 43.33679833679834,
      "grad_norm": 0.00012839803821407259,
      "learning_rate": 0.14006413942232626,
      "loss": 0.2805,
      "num_input_tokens_seen": 15896672,
      "step": 20845
    },
    {
      "epoch": 43.34719334719335,
      "grad_norm": 0.00024317018687725067,
      "learning_rate": 0.14000536469567235,
      "loss": 0.2954,
      "num_input_tokens_seen": 15900480,
      "step": 20850
    },
    {
      "epoch": 43.357588357588355,
      "grad_norm": 8.986486500361934e-05,
      "learning_rate": 0.13994659151031685,
      "loss": 0.2819,
      "num_input_tokens_seen": 15904288,
      "step": 20855
    },
    {
      "epoch": 43.36798336798337,
      "grad_norm": 0.000595803139731288,
      "learning_rate": 0.13988781987532323,
      "loss": 0.2806,
      "num_input_tokens_seen": 15908224,
      "step": 20860
    },
    {
      "epoch": 43.37837837837838,
      "grad_norm": 7.580459350720048e-05,
      "learning_rate": 0.1398290497997549,
      "loss": 0.2667,
      "num_input_tokens_seen": 15912064,
      "step": 20865
    },
    {
      "epoch": 43.38877338877339,
      "grad_norm": 0.00041864762897603214,
      "learning_rate": 0.13977028129267488,
      "loss": 0.2654,
      "num_input_tokens_seen": 15915712,
      "step": 20870
    },
    {
      "epoch": 43.3991683991684,
      "grad_norm": 7.840335456421599e-05,
      "learning_rate": 0.13971151436314605,
      "loss": 0.2692,
      "num_input_tokens_seen": 15919424,
      "step": 20875
    },
    {
      "epoch": 43.40956340956341,
      "grad_norm": 0.00011510030890349299,
      "learning_rate": 0.13965274902023103,
      "loss": 0.2738,
      "num_input_tokens_seen": 15923168,
      "step": 20880
    },
    {
      "epoch": 43.41995841995842,
      "grad_norm": 7.706279575359076e-05,
      "learning_rate": 0.13959398527299208,
      "loss": 0.2366,
      "num_input_tokens_seen": 15926976,
      "step": 20885
    },
    {
      "epoch": 43.43035343035343,
      "grad_norm": 0.00017795112216845155,
      "learning_rate": 0.13953522313049138,
      "loss": 0.2525,
      "num_input_tokens_seen": 15930656,
      "step": 20890
    },
    {
      "epoch": 43.44074844074844,
      "grad_norm": 0.00012256923946551979,
      "learning_rate": 0.13947646260179083,
      "loss": 0.2791,
      "num_input_tokens_seen": 15934368,
      "step": 20895
    },
    {
      "epoch": 43.45114345114345,
      "grad_norm": 0.00025555898901075125,
      "learning_rate": 0.13941770369595194,
      "loss": 0.2681,
      "num_input_tokens_seen": 15938144,
      "step": 20900
    },
    {
      "epoch": 43.46153846153846,
      "grad_norm": 0.0002503453870303929,
      "learning_rate": 0.1393589464220362,
      "loss": 0.2565,
      "num_input_tokens_seen": 15941824,
      "step": 20905
    },
    {
      "epoch": 43.471933471933475,
      "grad_norm": 5.845542182214558e-05,
      "learning_rate": 0.13930019078910455,
      "loss": 0.2749,
      "num_input_tokens_seen": 15945600,
      "step": 20910
    },
    {
      "epoch": 43.482328482328484,
      "grad_norm": 0.0002480710099916905,
      "learning_rate": 0.139241436806218,
      "loss": 0.2758,
      "num_input_tokens_seen": 15949440,
      "step": 20915
    },
    {
      "epoch": 43.49272349272349,
      "grad_norm": 0.000213769351830706,
      "learning_rate": 0.13918268448243712,
      "loss": 0.2725,
      "num_input_tokens_seen": 15953280,
      "step": 20920
    },
    {
      "epoch": 43.5031185031185,
      "grad_norm": 0.00019795297703240067,
      "learning_rate": 0.13912393382682217,
      "loss": 0.249,
      "num_input_tokens_seen": 15957152,
      "step": 20925
    },
    {
      "epoch": 43.513513513513516,
      "grad_norm": 0.0003817606484517455,
      "learning_rate": 0.1390651848484333,
      "loss": 0.2943,
      "num_input_tokens_seen": 15960896,
      "step": 20930
    },
    {
      "epoch": 43.523908523908524,
      "grad_norm": 0.00038954117917455733,
      "learning_rate": 0.1390064375563304,
      "loss": 0.2826,
      "num_input_tokens_seen": 15964832,
      "step": 20935
    },
    {
      "epoch": 43.53430353430353,
      "grad_norm": 0.00027874193619936705,
      "learning_rate": 0.13894769195957293,
      "loss": 0.2716,
      "num_input_tokens_seen": 15968608,
      "step": 20940
    },
    {
      "epoch": 43.54469854469855,
      "grad_norm": 9.038222196977586e-05,
      "learning_rate": 0.13888894806722032,
      "loss": 0.2722,
      "num_input_tokens_seen": 15972576,
      "step": 20945
    },
    {
      "epoch": 43.555093555093556,
      "grad_norm": 0.00016321065777447075,
      "learning_rate": 0.1388302058883315,
      "loss": 0.2678,
      "num_input_tokens_seen": 15976352,
      "step": 20950
    },
    {
      "epoch": 43.565488565488565,
      "grad_norm": 6.552860577357933e-05,
      "learning_rate": 0.13877146543196528,
      "loss": 0.2513,
      "num_input_tokens_seen": 15980256,
      "step": 20955
    },
    {
      "epoch": 43.57588357588357,
      "grad_norm": 0.00022836010612081736,
      "learning_rate": 0.13871272670718027,
      "loss": 0.2828,
      "num_input_tokens_seen": 15984096,
      "step": 20960
    },
    {
      "epoch": 43.58627858627859,
      "grad_norm": 0.00011581848957575858,
      "learning_rate": 0.13865398972303455,
      "loss": 0.2695,
      "num_input_tokens_seen": 15987840,
      "step": 20965
    },
    {
      "epoch": 43.5966735966736,
      "grad_norm": 6.513256812468171e-05,
      "learning_rate": 0.13859525448858623,
      "loss": 0.2775,
      "num_input_tokens_seen": 15991744,
      "step": 20970
    },
    {
      "epoch": 43.607068607068605,
      "grad_norm": 0.00017235387349501252,
      "learning_rate": 0.13853652101289304,
      "loss": 0.2578,
      "num_input_tokens_seen": 15995392,
      "step": 20975
    },
    {
      "epoch": 43.61746361746362,
      "grad_norm": 0.00042952195508405566,
      "learning_rate": 0.13847778930501234,
      "loss": 0.2856,
      "num_input_tokens_seen": 15999168,
      "step": 20980
    },
    {
      "epoch": 43.62785862785863,
      "grad_norm": 3.493340773275122e-05,
      "learning_rate": 0.1384190593740013,
      "loss": 0.2557,
      "num_input_tokens_seen": 16003008,
      "step": 20985
    },
    {
      "epoch": 43.63825363825364,
      "grad_norm": 0.00024644704535603523,
      "learning_rate": 0.13836033122891686,
      "loss": 0.2892,
      "num_input_tokens_seen": 16006784,
      "step": 20990
    },
    {
      "epoch": 43.648648648648646,
      "grad_norm": 0.00036609030212275684,
      "learning_rate": 0.1383016048788156,
      "loss": 0.2563,
      "num_input_tokens_seen": 16010496,
      "step": 20995
    },
    {
      "epoch": 43.65904365904366,
      "grad_norm": 7.281095895450562e-05,
      "learning_rate": 0.13824288033275392,
      "loss": 0.2563,
      "num_input_tokens_seen": 16014304,
      "step": 21000
    },
    {
      "epoch": 43.65904365904366,
      "eval_loss": 0.24751318991184235,
      "eval_runtime": 13.3865,
      "eval_samples_per_second": 63.945,
      "eval_steps_per_second": 15.986,
      "num_input_tokens_seen": 16014304,
      "step": 21000
    },
    {
      "epoch": 43.66943866943867,
      "grad_norm": 5.122081347508356e-05,
      "learning_rate": 0.1381841575997878,
      "loss": 0.2434,
      "num_input_tokens_seen": 16017952,
      "step": 21005
    },
    {
      "epoch": 43.67983367983368,
      "grad_norm": 6.415200914489105e-05,
      "learning_rate": 0.13812543668897306,
      "loss": 0.2609,
      "num_input_tokens_seen": 16021760,
      "step": 21010
    },
    {
      "epoch": 43.69022869022869,
      "grad_norm": 4.1278857679571956e-05,
      "learning_rate": 0.13806671760936526,
      "loss": 0.2855,
      "num_input_tokens_seen": 16025696,
      "step": 21015
    },
    {
      "epoch": 43.7006237006237,
      "grad_norm": 0.0004114708863198757,
      "learning_rate": 0.13800800037001956,
      "loss": 0.253,
      "num_input_tokens_seen": 16029504,
      "step": 21020
    },
    {
      "epoch": 43.71101871101871,
      "grad_norm": 0.0001820306497393176,
      "learning_rate": 0.13794928497999087,
      "loss": 0.27,
      "num_input_tokens_seen": 16033376,
      "step": 21025
    },
    {
      "epoch": 43.72141372141372,
      "grad_norm": 0.00010494939488125965,
      "learning_rate": 0.1378905714483339,
      "loss": 0.2495,
      "num_input_tokens_seen": 16037120,
      "step": 21030
    },
    {
      "epoch": 43.731808731808734,
      "grad_norm": 0.00015724134573247284,
      "learning_rate": 0.13783185978410295,
      "loss": 0.2367,
      "num_input_tokens_seen": 16040928,
      "step": 21035
    },
    {
      "epoch": 43.74220374220374,
      "grad_norm": 5.533166040549986e-05,
      "learning_rate": 0.13777314999635218,
      "loss": 0.269,
      "num_input_tokens_seen": 16044512,
      "step": 21040
    },
    {
      "epoch": 43.75259875259875,
      "grad_norm": 0.0002520230773370713,
      "learning_rate": 0.1377144420941353,
      "loss": 0.2852,
      "num_input_tokens_seen": 16048320,
      "step": 21045
    },
    {
      "epoch": 43.762993762993766,
      "grad_norm": 5.8905341575155035e-05,
      "learning_rate": 0.13765573608650586,
      "loss": 0.2635,
      "num_input_tokens_seen": 16051904,
      "step": 21050
    },
    {
      "epoch": 43.773388773388774,
      "grad_norm": 0.00038820350891910493,
      "learning_rate": 0.13759703198251702,
      "loss": 0.2594,
      "num_input_tokens_seen": 16055744,
      "step": 21055
    },
    {
      "epoch": 43.78378378378378,
      "grad_norm": 0.00010310626385034993,
      "learning_rate": 0.13753832979122174,
      "loss": 0.2633,
      "num_input_tokens_seen": 16059584,
      "step": 21060
    },
    {
      "epoch": 43.79417879417879,
      "grad_norm": 0.00019183687982149422,
      "learning_rate": 0.13747962952167264,
      "loss": 0.262,
      "num_input_tokens_seen": 16063296,
      "step": 21065
    },
    {
      "epoch": 43.804573804573806,
      "grad_norm": 0.0001501396909588948,
      "learning_rate": 0.13742093118292192,
      "loss": 0.2561,
      "num_input_tokens_seen": 16067040,
      "step": 21070
    },
    {
      "epoch": 43.814968814968815,
      "grad_norm": 0.00017984324949793518,
      "learning_rate": 0.13736223478402174,
      "loss": 0.2413,
      "num_input_tokens_seen": 16070816,
      "step": 21075
    },
    {
      "epoch": 43.82536382536382,
      "grad_norm": 8.610917575424537e-05,
      "learning_rate": 0.1373035403340238,
      "loss": 0.2857,
      "num_input_tokens_seen": 16074688,
      "step": 21080
    },
    {
      "epoch": 43.83575883575884,
      "grad_norm": 0.0003861830336973071,
      "learning_rate": 0.13724484784197943,
      "loss": 0.2629,
      "num_input_tokens_seen": 16078624,
      "step": 21085
    },
    {
      "epoch": 43.84615384615385,
      "grad_norm": 0.00041653879452496767,
      "learning_rate": 0.13718615731693987,
      "loss": 0.2679,
      "num_input_tokens_seen": 16082464,
      "step": 21090
    },
    {
      "epoch": 43.856548856548855,
      "grad_norm": 0.0006006438052281737,
      "learning_rate": 0.13712746876795587,
      "loss": 0.249,
      "num_input_tokens_seen": 16086272,
      "step": 21095
    },
    {
      "epoch": 43.86694386694387,
      "grad_norm": 0.0001967371063074097,
      "learning_rate": 0.13706878220407792,
      "loss": 0.2672,
      "num_input_tokens_seen": 16090112,
      "step": 21100
    },
    {
      "epoch": 43.87733887733888,
      "grad_norm": 0.00026333611458539963,
      "learning_rate": 0.13701009763435631,
      "loss": 0.2522,
      "num_input_tokens_seen": 16094048,
      "step": 21105
    },
    {
      "epoch": 43.88773388773389,
      "grad_norm": 8.01010537543334e-05,
      "learning_rate": 0.13695141506784084,
      "loss": 0.2892,
      "num_input_tokens_seen": 16097952,
      "step": 21110
    },
    {
      "epoch": 43.898128898128896,
      "grad_norm": 0.000379104953026399,
      "learning_rate": 0.13689273451358114,
      "loss": 0.2573,
      "num_input_tokens_seen": 16101888,
      "step": 21115
    },
    {
      "epoch": 43.90852390852391,
      "grad_norm": 4.9818871048046276e-05,
      "learning_rate": 0.13683405598062653,
      "loss": 0.2506,
      "num_input_tokens_seen": 16105600,
      "step": 21120
    },
    {
      "epoch": 43.91891891891892,
      "grad_norm": 0.0002987574553117156,
      "learning_rate": 0.1367753794780259,
      "loss": 0.281,
      "num_input_tokens_seen": 16109568,
      "step": 21125
    },
    {
      "epoch": 43.92931392931393,
      "grad_norm": 0.0002677681914065033,
      "learning_rate": 0.13671670501482802,
      "loss": 0.2504,
      "num_input_tokens_seen": 16113696,
      "step": 21130
    },
    {
      "epoch": 43.93970893970894,
      "grad_norm": 0.0002804089745040983,
      "learning_rate": 0.1366580326000811,
      "loss": 0.2548,
      "num_input_tokens_seen": 16117632,
      "step": 21135
    },
    {
      "epoch": 43.95010395010395,
      "grad_norm": 0.0002980169956572354,
      "learning_rate": 0.1365993622428332,
      "loss": 0.2743,
      "num_input_tokens_seen": 16121504,
      "step": 21140
    },
    {
      "epoch": 43.96049896049896,
      "grad_norm": 0.00032101667602546513,
      "learning_rate": 0.13654069395213211,
      "loss": 0.2769,
      "num_input_tokens_seen": 16125312,
      "step": 21145
    },
    {
      "epoch": 43.97089397089397,
      "grad_norm": 0.0002919249818660319,
      "learning_rate": 0.13648202773702509,
      "loss": 0.2369,
      "num_input_tokens_seen": 16129056,
      "step": 21150
    },
    {
      "epoch": 43.981288981288984,
      "grad_norm": 0.00021261710207909346,
      "learning_rate": 0.13642336360655927,
      "loss": 0.2511,
      "num_input_tokens_seen": 16132992,
      "step": 21155
    },
    {
      "epoch": 43.99168399168399,
      "grad_norm": 0.00012051967496518046,
      "learning_rate": 0.13636470156978145,
      "loss": 0.2674,
      "num_input_tokens_seen": 16136736,
      "step": 21160
    },
    {
      "epoch": 44.002079002079,
      "grad_norm": 7.050829299259931e-05,
      "learning_rate": 0.13630604163573798,
      "loss": 0.2782,
      "num_input_tokens_seen": 16140408,
      "step": 21165
    },
    {
      "epoch": 44.012474012474016,
      "grad_norm": 0.0005001010722480714,
      "learning_rate": 0.13624738381347495,
      "loss": 0.2609,
      "num_input_tokens_seen": 16144120,
      "step": 21170
    },
    {
      "epoch": 44.022869022869024,
      "grad_norm": 0.00011979175178566948,
      "learning_rate": 0.1361887281120382,
      "loss": 0.2605,
      "num_input_tokens_seen": 16147928,
      "step": 21175
    },
    {
      "epoch": 44.03326403326403,
      "grad_norm": 0.00019757766858674586,
      "learning_rate": 0.13613007454047307,
      "loss": 0.2593,
      "num_input_tokens_seen": 16151768,
      "step": 21180
    },
    {
      "epoch": 44.04365904365904,
      "grad_norm": 0.0002386389533057809,
      "learning_rate": 0.13607142310782486,
      "loss": 0.2648,
      "num_input_tokens_seen": 16155448,
      "step": 21185
    },
    {
      "epoch": 44.054054054054056,
      "grad_norm": 9.231379226548597e-05,
      "learning_rate": 0.13601277382313814,
      "loss": 0.2665,
      "num_input_tokens_seen": 16159128,
      "step": 21190
    },
    {
      "epoch": 44.064449064449065,
      "grad_norm": 0.00022090598940849304,
      "learning_rate": 0.1359541266954575,
      "loss": 0.2378,
      "num_input_tokens_seen": 16162872,
      "step": 21195
    },
    {
      "epoch": 44.07484407484407,
      "grad_norm": 0.00014628010103479028,
      "learning_rate": 0.13589548173382707,
      "loss": 0.2613,
      "num_input_tokens_seen": 16166680,
      "step": 21200
    },
    {
      "epoch": 44.07484407484407,
      "eval_loss": 0.24760009348392487,
      "eval_runtime": 13.3812,
      "eval_samples_per_second": 63.97,
      "eval_steps_per_second": 15.993,
      "num_input_tokens_seen": 16166680,
      "step": 21200
    },
    {
      "epoch": 44.08523908523909,
      "grad_norm": 0.00021189768449403346,
      "learning_rate": 0.1358368389472906,
      "loss": 0.2681,
      "num_input_tokens_seen": 16170616,
      "step": 21205
    },
    {
      "epoch": 44.0956340956341,
      "grad_norm": 0.0001461546344216913,
      "learning_rate": 0.13577819834489155,
      "loss": 0.2629,
      "num_input_tokens_seen": 16174680,
      "step": 21210
    },
    {
      "epoch": 44.106029106029105,
      "grad_norm": 0.0001874374720500782,
      "learning_rate": 0.135719559935673,
      "loss": 0.2433,
      "num_input_tokens_seen": 16178488,
      "step": 21215
    },
    {
      "epoch": 44.11642411642411,
      "grad_norm": 0.00021781085524708033,
      "learning_rate": 0.13566092372867775,
      "loss": 0.2846,
      "num_input_tokens_seen": 16182392,
      "step": 21220
    },
    {
      "epoch": 44.12681912681913,
      "grad_norm": 0.0004241634451318532,
      "learning_rate": 0.13560228973294833,
      "loss": 0.2877,
      "num_input_tokens_seen": 16186264,
      "step": 21225
    },
    {
      "epoch": 44.13721413721414,
      "grad_norm": 0.00023653804964851588,
      "learning_rate": 0.13554365795752668,
      "loss": 0.2609,
      "num_input_tokens_seen": 16190040,
      "step": 21230
    },
    {
      "epoch": 44.147609147609145,
      "grad_norm": 0.00027910235803574324,
      "learning_rate": 0.1354850284114547,
      "loss": 0.271,
      "num_input_tokens_seen": 16193848,
      "step": 21235
    },
    {
      "epoch": 44.15800415800416,
      "grad_norm": 0.0005078093381598592,
      "learning_rate": 0.13542640110377374,
      "loss": 0.2901,
      "num_input_tokens_seen": 16197816,
      "step": 21240
    },
    {
      "epoch": 44.16839916839917,
      "grad_norm": 0.00020517937082331628,
      "learning_rate": 0.13536777604352487,
      "loss": 0.2488,
      "num_input_tokens_seen": 16201528,
      "step": 21245
    },
    {
      "epoch": 44.17879417879418,
      "grad_norm": 0.00017930624017026275,
      "learning_rate": 0.13530915323974887,
      "loss": 0.2511,
      "num_input_tokens_seen": 16205432,
      "step": 21250
    },
    {
      "epoch": 44.189189189189186,
      "grad_norm": 0.00025428240769542754,
      "learning_rate": 0.13525053270148596,
      "loss": 0.2682,
      "num_input_tokens_seen": 16209272,
      "step": 21255
    },
    {
      "epoch": 44.1995841995842,
      "grad_norm": 0.0003311745822429657,
      "learning_rate": 0.13519191443777628,
      "loss": 0.2715,
      "num_input_tokens_seen": 16213144,
      "step": 21260
    },
    {
      "epoch": 44.20997920997921,
      "grad_norm": 0.0004858675820287317,
      "learning_rate": 0.13513329845765953,
      "loss": 0.2344,
      "num_input_tokens_seen": 16216984,
      "step": 21265
    },
    {
      "epoch": 44.22037422037422,
      "grad_norm": 0.000263742811512202,
      "learning_rate": 0.13507468477017495,
      "loss": 0.2783,
      "num_input_tokens_seen": 16220728,
      "step": 21270
    },
    {
      "epoch": 44.23076923076923,
      "grad_norm": 7.299939898075536e-05,
      "learning_rate": 0.13501607338436153,
      "loss": 0.2342,
      "num_input_tokens_seen": 16224632,
      "step": 21275
    },
    {
      "epoch": 44.24116424116424,
      "grad_norm": 0.0003884996403940022,
      "learning_rate": 0.13495746430925798,
      "loss": 0.2751,
      "num_input_tokens_seen": 16228504,
      "step": 21280
    },
    {
      "epoch": 44.25155925155925,
      "grad_norm": 0.00044079963117837906,
      "learning_rate": 0.13489885755390238,
      "loss": 0.2712,
      "num_input_tokens_seen": 16232216,
      "step": 21285
    },
    {
      "epoch": 44.26195426195426,
      "grad_norm": 4.1018534830072895e-05,
      "learning_rate": 0.13484025312733275,
      "loss": 0.2783,
      "num_input_tokens_seen": 16235928,
      "step": 21290
    },
    {
      "epoch": 44.272349272349274,
      "grad_norm": 0.00010333069803891703,
      "learning_rate": 0.13478165103858658,
      "loss": 0.268,
      "num_input_tokens_seen": 16239832,
      "step": 21295
    },
    {
      "epoch": 44.28274428274428,
      "grad_norm": 7.203936547739431e-05,
      "learning_rate": 0.13472305129670106,
      "loss": 0.2615,
      "num_input_tokens_seen": 16243736,
      "step": 21300
    },
    {
      "epoch": 44.29313929313929,
      "grad_norm": 0.00014360516797751188,
      "learning_rate": 0.13466445391071305,
      "loss": 0.2603,
      "num_input_tokens_seen": 16247512,
      "step": 21305
    },
    {
      "epoch": 44.303534303534306,
      "grad_norm": 0.00028150840080343187,
      "learning_rate": 0.13460585888965895,
      "loss": 0.2603,
      "num_input_tokens_seen": 16251352,
      "step": 21310
    },
    {
      "epoch": 44.313929313929314,
      "grad_norm": 0.0003955138963647187,
      "learning_rate": 0.13454726624257482,
      "loss": 0.2977,
      "num_input_tokens_seen": 16255096,
      "step": 21315
    },
    {
      "epoch": 44.32432432432432,
      "grad_norm": 4.251376958563924e-05,
      "learning_rate": 0.1344886759784965,
      "loss": 0.2766,
      "num_input_tokens_seen": 16258936,
      "step": 21320
    },
    {
      "epoch": 44.33471933471934,
      "grad_norm": 0.0001331439707428217,
      "learning_rate": 0.13443008810645923,
      "loss": 0.2738,
      "num_input_tokens_seen": 16262680,
      "step": 21325
    },
    {
      "epoch": 44.34511434511435,
      "grad_norm": 6.833983934484422e-05,
      "learning_rate": 0.13437150263549807,
      "loss": 0.2508,
      "num_input_tokens_seen": 16266296,
      "step": 21330
    },
    {
      "epoch": 44.355509355509355,
      "grad_norm": 0.0004538811044767499,
      "learning_rate": 0.13431291957464755,
      "loss": 0.266,
      "num_input_tokens_seen": 16270200,
      "step": 21335
    },
    {
      "epoch": 44.36590436590436,
      "grad_norm": 0.00030418054666370153,
      "learning_rate": 0.13425433893294197,
      "loss": 0.3076,
      "num_input_tokens_seen": 16274104,
      "step": 21340
    },
    {
      "epoch": 44.37629937629938,
      "grad_norm": 0.00047397197340615094,
      "learning_rate": 0.13419576071941525,
      "loss": 0.285,
      "num_input_tokens_seen": 16278008,
      "step": 21345
    },
    {
      "epoch": 44.38669438669439,
      "grad_norm": 2.938338184321765e-05,
      "learning_rate": 0.1341371849431008,
      "loss": 0.2723,
      "num_input_tokens_seen": 16281912,
      "step": 21350
    },
    {
      "epoch": 44.397089397089395,
      "grad_norm": 0.00019900860206689686,
      "learning_rate": 0.13407861161303178,
      "loss": 0.2602,
      "num_input_tokens_seen": 16285784,
      "step": 21355
    },
    {
      "epoch": 44.40748440748441,
      "grad_norm": 5.167853305465542e-05,
      "learning_rate": 0.13402004073824098,
      "loss": 0.2354,
      "num_input_tokens_seen": 16289528,
      "step": 21360
    },
    {
      "epoch": 44.41787941787942,
      "grad_norm": 0.0004037044127471745,
      "learning_rate": 0.13396147232776062,
      "loss": 0.2468,
      "num_input_tokens_seen": 16293368,
      "step": 21365
    },
    {
      "epoch": 44.42827442827443,
      "grad_norm": 0.00038667209446430206,
      "learning_rate": 0.13390290639062288,
      "loss": 0.2911,
      "num_input_tokens_seen": 16297112,
      "step": 21370
    },
    {
      "epoch": 44.438669438669436,
      "grad_norm": 3.342255877214484e-05,
      "learning_rate": 0.13384434293585917,
      "loss": 0.2607,
      "num_input_tokens_seen": 16301048,
      "step": 21375
    },
    {
      "epoch": 44.44906444906445,
      "grad_norm": 0.0001605077413842082,
      "learning_rate": 0.13378578197250088,
      "loss": 0.2571,
      "num_input_tokens_seen": 16305016,
      "step": 21380
    },
    {
      "epoch": 44.45945945945946,
      "grad_norm": 0.00023463493562303483,
      "learning_rate": 0.13372722350957872,
      "loss": 0.269,
      "num_input_tokens_seen": 16308952,
      "step": 21385
    },
    {
      "epoch": 44.46985446985447,
      "grad_norm": 7.709422061452642e-05,
      "learning_rate": 0.13366866755612322,
      "loss": 0.271,
      "num_input_tokens_seen": 16312696,
      "step": 21390
    },
    {
      "epoch": 44.48024948024948,
      "grad_norm": 9.872400551103055e-05,
      "learning_rate": 0.13361011412116436,
      "loss": 0.274,
      "num_input_tokens_seen": 16316504,
      "step": 21395
    },
    {
      "epoch": 44.49064449064449,
      "grad_norm": 4.162257755524479e-05,
      "learning_rate": 0.13355156321373196,
      "loss": 0.2594,
      "num_input_tokens_seen": 16320408,
      "step": 21400
    },
    {
      "epoch": 44.49064449064449,
      "eval_loss": 0.24985271692276,
      "eval_runtime": 13.3843,
      "eval_samples_per_second": 63.955,
      "eval_steps_per_second": 15.989,
      "num_input_tokens_seen": 16320408,
      "step": 21400
    },
    {
      "epoch": 44.5010395010395,
      "grad_norm": 9.209149720845744e-05,
      "learning_rate": 0.13349301484285514,
      "loss": 0.2616,
      "num_input_tokens_seen": 16324344,
      "step": 21405
    },
    {
      "epoch": 44.51143451143451,
      "grad_norm": 0.00010198220843449235,
      "learning_rate": 0.13343446901756295,
      "loss": 0.2652,
      "num_input_tokens_seen": 16328152,
      "step": 21410
    },
    {
      "epoch": 44.521829521829524,
      "grad_norm": 0.0003755299549084157,
      "learning_rate": 0.13337592574688376,
      "loss": 0.2542,
      "num_input_tokens_seen": 16332024,
      "step": 21415
    },
    {
      "epoch": 44.53222453222453,
      "grad_norm": 0.00013641236000694335,
      "learning_rate": 0.13331738503984572,
      "loss": 0.2693,
      "num_input_tokens_seen": 16335768,
      "step": 21420
    },
    {
      "epoch": 44.54261954261954,
      "grad_norm": 0.00010613301128614694,
      "learning_rate": 0.1332588469054766,
      "loss": 0.2777,
      "num_input_tokens_seen": 16339544,
      "step": 21425
    },
    {
      "epoch": 44.553014553014556,
      "grad_norm": 5.538835466722958e-05,
      "learning_rate": 0.1332003113528036,
      "loss": 0.263,
      "num_input_tokens_seen": 16343384,
      "step": 21430
    },
    {
      "epoch": 44.563409563409564,
      "grad_norm": 0.00014307141827885062,
      "learning_rate": 0.13314177839085373,
      "loss": 0.2728,
      "num_input_tokens_seen": 16347160,
      "step": 21435
    },
    {
      "epoch": 44.57380457380457,
      "grad_norm": 4.037943290313706e-05,
      "learning_rate": 0.13308324802865354,
      "loss": 0.2579,
      "num_input_tokens_seen": 16350872,
      "step": 21440
    },
    {
      "epoch": 44.58419958419958,
      "grad_norm": 4.118874130654149e-05,
      "learning_rate": 0.13302472027522905,
      "loss": 0.2565,
      "num_input_tokens_seen": 16354552,
      "step": 21445
    },
    {
      "epoch": 44.5945945945946,
      "grad_norm": 0.00016343964671250433,
      "learning_rate": 0.13296619513960606,
      "loss": 0.2461,
      "num_input_tokens_seen": 16358360,
      "step": 21450
    },
    {
      "epoch": 44.604989604989605,
      "grad_norm": 0.0004469746199902147,
      "learning_rate": 0.1329076726308098,
      "loss": 0.2681,
      "num_input_tokens_seen": 16362168,
      "step": 21455
    },
    {
      "epoch": 44.61538461538461,
      "grad_norm": 0.0005577249685302377,
      "learning_rate": 0.13284915275786519,
      "loss": 0.2872,
      "num_input_tokens_seen": 16365848,
      "step": 21460
    },
    {
      "epoch": 44.62577962577963,
      "grad_norm": 0.00013743994350079447,
      "learning_rate": 0.1327906355297968,
      "loss": 0.2801,
      "num_input_tokens_seen": 16369688,
      "step": 21465
    },
    {
      "epoch": 44.63617463617464,
      "grad_norm": 0.00011450327292550355,
      "learning_rate": 0.13273212095562867,
      "loss": 0.2761,
      "num_input_tokens_seen": 16373272,
      "step": 21470
    },
    {
      "epoch": 44.646569646569645,
      "grad_norm": 0.00020977872190997005,
      "learning_rate": 0.13267360904438444,
      "loss": 0.2716,
      "num_input_tokens_seen": 16376920,
      "step": 21475
    },
    {
      "epoch": 44.656964656964654,
      "grad_norm": 8.59934079926461e-05,
      "learning_rate": 0.1326150998050875,
      "loss": 0.2549,
      "num_input_tokens_seen": 16380792,
      "step": 21480
    },
    {
      "epoch": 44.66735966735967,
      "grad_norm": 0.0004057687474414706,
      "learning_rate": 0.1325565932467606,
      "loss": 0.2594,
      "num_input_tokens_seen": 16384472,
      "step": 21485
    },
    {
      "epoch": 44.67775467775468,
      "grad_norm": 0.0005032027256675065,
      "learning_rate": 0.13249808937842628,
      "loss": 0.2729,
      "num_input_tokens_seen": 16388408,
      "step": 21490
    },
    {
      "epoch": 44.688149688149686,
      "grad_norm": 0.00038425010279752314,
      "learning_rate": 0.1324395882091065,
      "loss": 0.2492,
      "num_input_tokens_seen": 16392152,
      "step": 21495
    },
    {
      "epoch": 44.6985446985447,
      "grad_norm": 0.000612241739872843,
      "learning_rate": 0.13238108974782284,
      "loss": 0.2765,
      "num_input_tokens_seen": 16395960,
      "step": 21500
    },
    {
      "epoch": 44.70893970893971,
      "grad_norm": 0.00022954182350076735,
      "learning_rate": 0.13232259400359664,
      "loss": 0.2627,
      "num_input_tokens_seen": 16399832,
      "step": 21505
    },
    {
      "epoch": 44.71933471933472,
      "grad_norm": 7.659868424525484e-05,
      "learning_rate": 0.13226410098544852,
      "loss": 0.2694,
      "num_input_tokens_seen": 16403704,
      "step": 21510
    },
    {
      "epoch": 44.729729729729726,
      "grad_norm": 0.0003492979740258306,
      "learning_rate": 0.13220561070239892,
      "loss": 0.2645,
      "num_input_tokens_seen": 16407736,
      "step": 21515
    },
    {
      "epoch": 44.74012474012474,
      "grad_norm": 0.00045646459329873323,
      "learning_rate": 0.13214712316346783,
      "loss": 0.2707,
      "num_input_tokens_seen": 16411576,
      "step": 21520
    },
    {
      "epoch": 44.75051975051975,
      "grad_norm": 0.00022545094543602318,
      "learning_rate": 0.13208863837767465,
      "loss": 0.2708,
      "num_input_tokens_seen": 16415384,
      "step": 21525
    },
    {
      "epoch": 44.76091476091476,
      "grad_norm": 0.00014819574425928295,
      "learning_rate": 0.13203015635403856,
      "loss": 0.2784,
      "num_input_tokens_seen": 16419448,
      "step": 21530
    },
    {
      "epoch": 44.771309771309774,
      "grad_norm": 0.00037124508526176214,
      "learning_rate": 0.13197167710157817,
      "loss": 0.2796,
      "num_input_tokens_seen": 16423416,
      "step": 21535
    },
    {
      "epoch": 44.78170478170478,
      "grad_norm": 0.0003178407496307045,
      "learning_rate": 0.13191320062931167,
      "loss": 0.2702,
      "num_input_tokens_seen": 16427192,
      "step": 21540
    },
    {
      "epoch": 44.79209979209979,
      "grad_norm": 0.0001378702581860125,
      "learning_rate": 0.13185472694625702,
      "loss": 0.2714,
      "num_input_tokens_seen": 16431000,
      "step": 21545
    },
    {
      "epoch": 44.802494802494806,
      "grad_norm": 2.3721042452962138e-05,
      "learning_rate": 0.13179625606143142,
      "loss": 0.2677,
      "num_input_tokens_seen": 16434904,
      "step": 21550
    },
    {
      "epoch": 44.812889812889814,
      "grad_norm": 0.0001865524100139737,
      "learning_rate": 0.13173778798385188,
      "loss": 0.2581,
      "num_input_tokens_seen": 16438776,
      "step": 21555
    },
    {
      "epoch": 44.82328482328482,
      "grad_norm": 0.00039999911678023636,
      "learning_rate": 0.13167932272253505,
      "loss": 0.2577,
      "num_input_tokens_seen": 16442584,
      "step": 21560
    },
    {
      "epoch": 44.83367983367983,
      "grad_norm": 0.00036108295898884535,
      "learning_rate": 0.1316208602864968,
      "loss": 0.2737,
      "num_input_tokens_seen": 16446488,
      "step": 21565
    },
    {
      "epoch": 44.84407484407485,
      "grad_norm": 0.00026716841966845095,
      "learning_rate": 0.13156240068475292,
      "loss": 0.2437,
      "num_input_tokens_seen": 16450200,
      "step": 21570
    },
    {
      "epoch": 44.854469854469855,
      "grad_norm": 0.00023703456099610776,
      "learning_rate": 0.1315039439263185,
      "loss": 0.2527,
      "num_input_tokens_seen": 16454008,
      "step": 21575
    },
    {
      "epoch": 44.86486486486486,
      "grad_norm": 9.033113019540906e-05,
      "learning_rate": 0.13144549002020833,
      "loss": 0.2717,
      "num_input_tokens_seen": 16457944,
      "step": 21580
    },
    {
      "epoch": 44.87525987525988,
      "grad_norm": 0.00025927714887075126,
      "learning_rate": 0.13138703897543688,
      "loss": 0.2546,
      "num_input_tokens_seen": 16461592,
      "step": 21585
    },
    {
      "epoch": 44.88565488565489,
      "grad_norm": 0.0004093958414159715,
      "learning_rate": 0.1313285908010178,
      "loss": 0.2506,
      "num_input_tokens_seen": 16465336,
      "step": 21590
    },
    {
      "epoch": 44.896049896049895,
      "grad_norm": 2.9531140171457082e-05,
      "learning_rate": 0.13127014550596475,
      "loss": 0.2629,
      "num_input_tokens_seen": 16468952,
      "step": 21595
    },
    {
      "epoch": 44.906444906444904,
      "grad_norm": 0.0003926590725313872,
      "learning_rate": 0.1312117030992906,
      "loss": 0.2615,
      "num_input_tokens_seen": 16472888,
      "step": 21600
    },
    {
      "epoch": 44.906444906444904,
      "eval_loss": 0.2488902509212494,
      "eval_runtime": 13.3818,
      "eval_samples_per_second": 63.967,
      "eval_steps_per_second": 15.992,
      "num_input_tokens_seen": 16472888,
      "step": 21600
    },
    {
      "epoch": 44.91683991683992,
      "grad_norm": 0.0001376639847876504,
      "learning_rate": 0.13115326359000795,
      "loss": 0.286,
      "num_input_tokens_seen": 16476664,
      "step": 21605
    },
    {
      "epoch": 44.92723492723493,
      "grad_norm": 7.41363619454205e-05,
      "learning_rate": 0.13109482698712896,
      "loss": 0.2699,
      "num_input_tokens_seen": 16480344,
      "step": 21610
    },
    {
      "epoch": 44.937629937629936,
      "grad_norm": 0.00011974335939157754,
      "learning_rate": 0.1310363932996651,
      "loss": 0.2781,
      "num_input_tokens_seen": 16484184,
      "step": 21615
    },
    {
      "epoch": 44.94802494802495,
      "grad_norm": 0.0003414248349145055,
      "learning_rate": 0.13097796253662775,
      "loss": 0.2584,
      "num_input_tokens_seen": 16488088,
      "step": 21620
    },
    {
      "epoch": 44.95841995841996,
      "grad_norm": 0.00020359441987238824,
      "learning_rate": 0.1309195347070277,
      "loss": 0.282,
      "num_input_tokens_seen": 16491928,
      "step": 21625
    },
    {
      "epoch": 44.96881496881497,
      "grad_norm": 0.0002986837353091687,
      "learning_rate": 0.13086110981987506,
      "loss": 0.2529,
      "num_input_tokens_seen": 16495672,
      "step": 21630
    },
    {
      "epoch": 44.979209979209976,
      "grad_norm": 0.0004336940182838589,
      "learning_rate": 0.13080268788417987,
      "loss": 0.2697,
      "num_input_tokens_seen": 16499544,
      "step": 21635
    },
    {
      "epoch": 44.98960498960499,
      "grad_norm": 0.00025523744989186525,
      "learning_rate": 0.1307442689089515,
      "loss": 0.2793,
      "num_input_tokens_seen": 16503288,
      "step": 21640
    },
    {
      "epoch": 45.0,
      "grad_norm": 7.988721336005256e-05,
      "learning_rate": 0.13068585290319873,
      "loss": 0.2803,
      "num_input_tokens_seen": 16507056,
      "step": 21645
    },
    {
      "epoch": 45.01039501039501,
      "grad_norm": 0.000308271060930565,
      "learning_rate": 0.13062743987593026,
      "loss": 0.2599,
      "num_input_tokens_seen": 16510864,
      "step": 21650
    },
    {
      "epoch": 45.020790020790024,
      "grad_norm": 0.0004340185842011124,
      "learning_rate": 0.13056902983615395,
      "loss": 0.269,
      "num_input_tokens_seen": 16514640,
      "step": 21655
    },
    {
      "epoch": 45.03118503118503,
      "grad_norm": 4.979573714081198e-05,
      "learning_rate": 0.13051062279287742,
      "loss": 0.2588,
      "num_input_tokens_seen": 16518512,
      "step": 21660
    },
    {
      "epoch": 45.04158004158004,
      "grad_norm": 0.00011068989988416433,
      "learning_rate": 0.13045221875510782,
      "loss": 0.261,
      "num_input_tokens_seen": 16522320,
      "step": 21665
    },
    {
      "epoch": 45.05197505197505,
      "grad_norm": 7.1961541834753e-05,
      "learning_rate": 0.13039381773185174,
      "loss": 0.2871,
      "num_input_tokens_seen": 16526096,
      "step": 21670
    },
    {
      "epoch": 45.062370062370064,
      "grad_norm": 0.0006801162380725145,
      "learning_rate": 0.1303354197321153,
      "loss": 0.2508,
      "num_input_tokens_seen": 16529936,
      "step": 21675
    },
    {
      "epoch": 45.07276507276507,
      "grad_norm": 0.00022337948030326515,
      "learning_rate": 0.13027702476490433,
      "loss": 0.2663,
      "num_input_tokens_seen": 16533712,
      "step": 21680
    },
    {
      "epoch": 45.08316008316008,
      "grad_norm": 8.445126150036231e-05,
      "learning_rate": 0.1302186328392239,
      "loss": 0.2756,
      "num_input_tokens_seen": 16537616,
      "step": 21685
    },
    {
      "epoch": 45.093555093555096,
      "grad_norm": 0.00040929674287326634,
      "learning_rate": 0.130160243964079,
      "loss": 0.2762,
      "num_input_tokens_seen": 16541520,
      "step": 21690
    },
    {
      "epoch": 45.103950103950105,
      "grad_norm": 0.00021836582163814455,
      "learning_rate": 0.13010185814847372,
      "loss": 0.2717,
      "num_input_tokens_seen": 16545296,
      "step": 21695
    },
    {
      "epoch": 45.11434511434511,
      "grad_norm": 0.0003535347932484001,
      "learning_rate": 0.13004347540141192,
      "loss": 0.2828,
      "num_input_tokens_seen": 16549264,
      "step": 21700
    },
    {
      "epoch": 45.12474012474012,
      "grad_norm": 0.00041933933971449733,
      "learning_rate": 0.12998509573189712,
      "loss": 0.2801,
      "num_input_tokens_seen": 16553072,
      "step": 21705
    },
    {
      "epoch": 45.13513513513514,
      "grad_norm": 0.0001975226477952674,
      "learning_rate": 0.12992671914893203,
      "loss": 0.2754,
      "num_input_tokens_seen": 16556880,
      "step": 21710
    },
    {
      "epoch": 45.145530145530145,
      "grad_norm": 0.0002361765509704128,
      "learning_rate": 0.12986834566151909,
      "loss": 0.2671,
      "num_input_tokens_seen": 16560656,
      "step": 21715
    },
    {
      "epoch": 45.15592515592515,
      "grad_norm": 7.07785293343477e-05,
      "learning_rate": 0.12980997527866028,
      "loss": 0.2707,
      "num_input_tokens_seen": 16564528,
      "step": 21720
    },
    {
      "epoch": 45.16632016632017,
      "grad_norm": 0.00020954417414031923,
      "learning_rate": 0.12975160800935692,
      "loss": 0.2843,
      "num_input_tokens_seen": 16568336,
      "step": 21725
    },
    {
      "epoch": 45.17671517671518,
      "grad_norm": 0.00036131739034317434,
      "learning_rate": 0.12969324386261016,
      "loss": 0.2696,
      "num_input_tokens_seen": 16572240,
      "step": 21730
    },
    {
      "epoch": 45.187110187110186,
      "grad_norm": 0.00024694547755643725,
      "learning_rate": 0.12963488284742034,
      "loss": 0.2597,
      "num_input_tokens_seen": 16576080,
      "step": 21735
    },
    {
      "epoch": 45.197505197505194,
      "grad_norm": 0.0002636623103171587,
      "learning_rate": 0.12957652497278752,
      "loss": 0.2259,
      "num_input_tokens_seen": 16579792,
      "step": 21740
    },
    {
      "epoch": 45.20790020790021,
      "grad_norm": 0.00033735408214852214,
      "learning_rate": 0.12951817024771117,
      "loss": 0.2546,
      "num_input_tokens_seen": 16583504,
      "step": 21745
    },
    {
      "epoch": 45.21829521829522,
      "grad_norm": 0.0004906202084384859,
      "learning_rate": 0.12945981868119041,
      "loss": 0.3106,
      "num_input_tokens_seen": 16587376,
      "step": 21750
    },
    {
      "epoch": 45.228690228690226,
      "grad_norm": 0.00024589296663179994,
      "learning_rate": 0.12940147028222376,
      "loss": 0.2801,
      "num_input_tokens_seen": 16591248,
      "step": 21755
    },
    {
      "epoch": 45.23908523908524,
      "grad_norm": 0.0003033537359442562,
      "learning_rate": 0.12934312505980916,
      "loss": 0.2627,
      "num_input_tokens_seen": 16595088,
      "step": 21760
    },
    {
      "epoch": 45.24948024948025,
      "grad_norm": 0.0005645820056088269,
      "learning_rate": 0.1292847830229443,
      "loss": 0.2442,
      "num_input_tokens_seen": 16598992,
      "step": 21765
    },
    {
      "epoch": 45.25987525987526,
      "grad_norm": 0.0002161420852644369,
      "learning_rate": 0.12922644418062626,
      "loss": 0.2772,
      "num_input_tokens_seen": 16602640,
      "step": 21770
    },
    {
      "epoch": 45.270270270270274,
      "grad_norm": 0.00021539830777328461,
      "learning_rate": 0.1291681085418515,
      "loss": 0.2748,
      "num_input_tokens_seen": 16606480,
      "step": 21775
    },
    {
      "epoch": 45.28066528066528,
      "grad_norm": 0.0001854811271186918,
      "learning_rate": 0.12910977611561628,
      "loss": 0.2613,
      "num_input_tokens_seen": 16610448,
      "step": 21780
    },
    {
      "epoch": 45.29106029106029,
      "grad_norm": 0.00014989111514296383,
      "learning_rate": 0.1290514469109161,
      "loss": 0.2737,
      "num_input_tokens_seen": 16614192,
      "step": 21785
    },
    {
      "epoch": 45.3014553014553,
      "grad_norm": 0.00013945419050287455,
      "learning_rate": 0.128993120936746,
      "loss": 0.2496,
      "num_input_tokens_seen": 16618096,
      "step": 21790
    },
    {
      "epoch": 45.311850311850314,
      "grad_norm": 0.00015988264931365848,
      "learning_rate": 0.12893479820210071,
      "loss": 0.2831,
      "num_input_tokens_seen": 16621840,
      "step": 21795
    },
    {
      "epoch": 45.32224532224532,
      "grad_norm": 0.00027094673714600503,
      "learning_rate": 0.1288764787159742,
      "loss": 0.2551,
      "num_input_tokens_seen": 16625808,
      "step": 21800
    },
    {
      "epoch": 45.32224532224532,
      "eval_loss": 0.25192537903785706,
      "eval_runtime": 13.3766,
      "eval_samples_per_second": 63.993,
      "eval_steps_per_second": 15.998,
      "num_input_tokens_seen": 16625808,
      "step": 21800
    },
    {
      "epoch": 45.33264033264033,
      "grad_norm": 0.0002906887384597212,
      "learning_rate": 0.1288181624873601,
      "loss": 0.2629,
      "num_input_tokens_seen": 16629872,
      "step": 21805
    },
    {
      "epoch": 45.343035343035346,
      "grad_norm": 8.257607987616211e-05,
      "learning_rate": 0.12875984952525163,
      "loss": 0.2681,
      "num_input_tokens_seen": 16633808,
      "step": 21810
    },
    {
      "epoch": 45.353430353430355,
      "grad_norm": 6.172782741487026e-05,
      "learning_rate": 0.12870153983864122,
      "loss": 0.2663,
      "num_input_tokens_seen": 16637424,
      "step": 21815
    },
    {
      "epoch": 45.36382536382536,
      "grad_norm": 0.0001658169785514474,
      "learning_rate": 0.12864323343652104,
      "loss": 0.2751,
      "num_input_tokens_seen": 16641232,
      "step": 21820
    },
    {
      "epoch": 45.37422037422037,
      "grad_norm": 9.83078934950754e-05,
      "learning_rate": 0.12858493032788268,
      "loss": 0.2792,
      "num_input_tokens_seen": 16644880,
      "step": 21825
    },
    {
      "epoch": 45.38461538461539,
      "grad_norm": 0.0005819424404762685,
      "learning_rate": 0.12852663052171714,
      "loss": 0.2407,
      "num_input_tokens_seen": 16648720,
      "step": 21830
    },
    {
      "epoch": 45.395010395010395,
      "grad_norm": 0.00017988719628192484,
      "learning_rate": 0.12846833402701507,
      "loss": 0.2569,
      "num_input_tokens_seen": 16652368,
      "step": 21835
    },
    {
      "epoch": 45.4054054054054,
      "grad_norm": 0.0005045937141403556,
      "learning_rate": 0.12841004085276642,
      "loss": 0.261,
      "num_input_tokens_seen": 16656272,
      "step": 21840
    },
    {
      "epoch": 45.41580041580042,
      "grad_norm": 0.00023076646903064102,
      "learning_rate": 0.12835175100796076,
      "loss": 0.2494,
      "num_input_tokens_seen": 16660176,
      "step": 21845
    },
    {
      "epoch": 45.42619542619543,
      "grad_norm": 0.00021518296853173524,
      "learning_rate": 0.12829346450158724,
      "loss": 0.2894,
      "num_input_tokens_seen": 16664016,
      "step": 21850
    },
    {
      "epoch": 45.436590436590436,
      "grad_norm": 0.00023761062766425312,
      "learning_rate": 0.12823518134263423,
      "loss": 0.2632,
      "num_input_tokens_seen": 16667888,
      "step": 21855
    },
    {
      "epoch": 45.446985446985444,
      "grad_norm": 0.0002611563540995121,
      "learning_rate": 0.12817690154008973,
      "loss": 0.2613,
      "num_input_tokens_seen": 16671728,
      "step": 21860
    },
    {
      "epoch": 45.45738045738046,
      "grad_norm": 0.0006271555321291089,
      "learning_rate": 0.12811862510294134,
      "loss": 0.2664,
      "num_input_tokens_seen": 16675472,
      "step": 21865
    },
    {
      "epoch": 45.46777546777547,
      "grad_norm": 7.920320058474317e-05,
      "learning_rate": 0.12806035204017585,
      "loss": 0.2603,
      "num_input_tokens_seen": 16679280,
      "step": 21870
    },
    {
      "epoch": 45.478170478170476,
      "grad_norm": 0.00011259713937761262,
      "learning_rate": 0.12800208236077987,
      "loss": 0.2733,
      "num_input_tokens_seen": 16683088,
      "step": 21875
    },
    {
      "epoch": 45.48856548856549,
      "grad_norm": 0.00017516405205242336,
      "learning_rate": 0.12794381607373917,
      "loss": 0.2943,
      "num_input_tokens_seen": 16687088,
      "step": 21880
    },
    {
      "epoch": 45.4989604989605,
      "grad_norm": 7.293133239727467e-05,
      "learning_rate": 0.12788555318803924,
      "loss": 0.2586,
      "num_input_tokens_seen": 16690864,
      "step": 21885
    },
    {
      "epoch": 45.50935550935551,
      "grad_norm": 0.00026224294560961425,
      "learning_rate": 0.1278272937126649,
      "loss": 0.2594,
      "num_input_tokens_seen": 16694704,
      "step": 21890
    },
    {
      "epoch": 45.51975051975052,
      "grad_norm": 0.00033755958429537714,
      "learning_rate": 0.1277690376566005,
      "loss": 0.2683,
      "num_input_tokens_seen": 16698576,
      "step": 21895
    },
    {
      "epoch": 45.53014553014553,
      "grad_norm": 7.646583981113508e-05,
      "learning_rate": 0.12771078502882985,
      "loss": 0.2612,
      "num_input_tokens_seen": 16702288,
      "step": 21900
    },
    {
      "epoch": 45.54054054054054,
      "grad_norm": 7.660726987523958e-05,
      "learning_rate": 0.12765253583833633,
      "loss": 0.2728,
      "num_input_tokens_seen": 16706160,
      "step": 21905
    },
    {
      "epoch": 45.55093555093555,
      "grad_norm": 0.00036447244929149747,
      "learning_rate": 0.12759429009410256,
      "loss": 0.2802,
      "num_input_tokens_seen": 16710064,
      "step": 21910
    },
    {
      "epoch": 45.561330561330564,
      "grad_norm": 0.000583614397328347,
      "learning_rate": 0.12753604780511085,
      "loss": 0.2658,
      "num_input_tokens_seen": 16713808,
      "step": 21915
    },
    {
      "epoch": 45.57172557172557,
      "grad_norm": 0.00017305099754594266,
      "learning_rate": 0.12747780898034283,
      "loss": 0.2614,
      "num_input_tokens_seen": 16717680,
      "step": 21920
    },
    {
      "epoch": 45.58212058212058,
      "grad_norm": 0.00039512160583399236,
      "learning_rate": 0.12741957362877973,
      "loss": 0.2824,
      "num_input_tokens_seen": 16721328,
      "step": 21925
    },
    {
      "epoch": 45.59251559251559,
      "grad_norm": 0.00017481579561717808,
      "learning_rate": 0.12736134175940214,
      "loss": 0.2762,
      "num_input_tokens_seen": 16725040,
      "step": 21930
    },
    {
      "epoch": 45.602910602910605,
      "grad_norm": 5.549371780944057e-05,
      "learning_rate": 0.12730311338119016,
      "loss": 0.2401,
      "num_input_tokens_seen": 16728816,
      "step": 21935
    },
    {
      "epoch": 45.61330561330561,
      "grad_norm": 0.0002514217921998352,
      "learning_rate": 0.12724488850312327,
      "loss": 0.2678,
      "num_input_tokens_seen": 16732688,
      "step": 21940
    },
    {
      "epoch": 45.62370062370062,
      "grad_norm": 0.0001222207793034613,
      "learning_rate": 0.1271866671341806,
      "loss": 0.2481,
      "num_input_tokens_seen": 16736368,
      "step": 21945
    },
    {
      "epoch": 45.63409563409564,
      "grad_norm": 0.0004071629955433309,
      "learning_rate": 0.12712844928334047,
      "loss": 0.2726,
      "num_input_tokens_seen": 16740240,
      "step": 21950
    },
    {
      "epoch": 45.644490644490645,
      "grad_norm": 0.0003476296260487288,
      "learning_rate": 0.12707023495958095,
      "loss": 0.2314,
      "num_input_tokens_seen": 16743920,
      "step": 21955
    },
    {
      "epoch": 45.65488565488565,
      "grad_norm": 6.0887155996169895e-05,
      "learning_rate": 0.12701202417187932,
      "loss": 0.2261,
      "num_input_tokens_seen": 16747824,
      "step": 21960
    },
    {
      "epoch": 45.66528066528066,
      "grad_norm": 0.0005414694896899164,
      "learning_rate": 0.12695381692921243,
      "loss": 0.3144,
      "num_input_tokens_seen": 16751824,
      "step": 21965
    },
    {
      "epoch": 45.67567567567568,
      "grad_norm": 0.00018057672423310578,
      "learning_rate": 0.12689561324055665,
      "loss": 0.2664,
      "num_input_tokens_seen": 16755568,
      "step": 21970
    },
    {
      "epoch": 45.686070686070686,
      "grad_norm": 0.00013447966193780303,
      "learning_rate": 0.12683741311488758,
      "loss": 0.268,
      "num_input_tokens_seen": 16759184,
      "step": 21975
    },
    {
      "epoch": 45.696465696465694,
      "grad_norm": 0.00041670765494927764,
      "learning_rate": 0.1267792165611805,
      "loss": 0.2732,
      "num_input_tokens_seen": 16762928,
      "step": 21980
    },
    {
      "epoch": 45.70686070686071,
      "grad_norm": 0.00018882739823311567,
      "learning_rate": 0.1267210235884101,
      "loss": 0.2674,
      "num_input_tokens_seen": 16766704,
      "step": 21985
    },
    {
      "epoch": 45.71725571725572,
      "grad_norm": 0.00014354908489622176,
      "learning_rate": 0.12666283420555033,
      "loss": 0.2761,
      "num_input_tokens_seen": 16770544,
      "step": 21990
    },
    {
      "epoch": 45.727650727650726,
      "grad_norm": 0.00031965802190825343,
      "learning_rate": 0.12660464842157487,
      "loss": 0.2598,
      "num_input_tokens_seen": 16774384,
      "step": 21995
    },
    {
      "epoch": 45.73804573804574,
      "grad_norm": 0.00010237919195787981,
      "learning_rate": 0.1265464662454566,
      "loss": 0.261,
      "num_input_tokens_seen": 16778288,
      "step": 22000
    },
    {
      "epoch": 45.73804573804574,
      "eval_loss": 0.2485373318195343,
      "eval_runtime": 13.3685,
      "eval_samples_per_second": 64.031,
      "eval_steps_per_second": 16.008,
      "num_input_tokens_seen": 16778288,
      "step": 22000
    },
    {
      "epoch": 45.74844074844075,
      "grad_norm": 0.00037750197225250304,
      "learning_rate": 0.12648828768616793,
      "loss": 0.2745,
      "num_input_tokens_seen": 16782032,
      "step": 22005
    },
    {
      "epoch": 45.75883575883576,
      "grad_norm": 0.0003712645557243377,
      "learning_rate": 0.12643011275268085,
      "loss": 0.2424,
      "num_input_tokens_seen": 16786064,
      "step": 22010
    },
    {
      "epoch": 45.76923076923077,
      "grad_norm": 5.295631126500666e-05,
      "learning_rate": 0.1263719414539665,
      "loss": 0.2647,
      "num_input_tokens_seen": 16789872,
      "step": 22015
    },
    {
      "epoch": 45.77962577962578,
      "grad_norm": 0.00017828235286287963,
      "learning_rate": 0.1263137737989957,
      "loss": 0.2614,
      "num_input_tokens_seen": 16793776,
      "step": 22020
    },
    {
      "epoch": 45.79002079002079,
      "grad_norm": 0.00031687956652604043,
      "learning_rate": 0.1262556097967387,
      "loss": 0.2726,
      "num_input_tokens_seen": 16797680,
      "step": 22025
    },
    {
      "epoch": 45.8004158004158,
      "grad_norm": 8.944228466134518e-05,
      "learning_rate": 0.126197449456165,
      "loss": 0.2565,
      "num_input_tokens_seen": 16801616,
      "step": 22030
    },
    {
      "epoch": 45.810810810810814,
      "grad_norm": 8.557483670301735e-05,
      "learning_rate": 0.12613929278624378,
      "loss": 0.2689,
      "num_input_tokens_seen": 16805456,
      "step": 22035
    },
    {
      "epoch": 45.82120582120582,
      "grad_norm": 0.00010479663615114987,
      "learning_rate": 0.12608113979594343,
      "loss": 0.2486,
      "num_input_tokens_seen": 16809200,
      "step": 22040
    },
    {
      "epoch": 45.83160083160083,
      "grad_norm": 7.175657810876146e-05,
      "learning_rate": 0.1260229904942319,
      "loss": 0.2912,
      "num_input_tokens_seen": 16813072,
      "step": 22045
    },
    {
      "epoch": 45.84199584199584,
      "grad_norm": 5.344607779989019e-05,
      "learning_rate": 0.12596484489007662,
      "loss": 0.2695,
      "num_input_tokens_seen": 16816944,
      "step": 22050
    },
    {
      "epoch": 45.852390852390855,
      "grad_norm": 0.00020244310144335032,
      "learning_rate": 0.1259067029924442,
      "loss": 0.2673,
      "num_input_tokens_seen": 16820688,
      "step": 22055
    },
    {
      "epoch": 45.86278586278586,
      "grad_norm": 0.00012761795369442552,
      "learning_rate": 0.12584856481030096,
      "loss": 0.253,
      "num_input_tokens_seen": 16824464,
      "step": 22060
    },
    {
      "epoch": 45.87318087318087,
      "grad_norm": 3.073545667575672e-05,
      "learning_rate": 0.12579043035261261,
      "loss": 0.2452,
      "num_input_tokens_seen": 16828240,
      "step": 22065
    },
    {
      "epoch": 45.88357588357589,
      "grad_norm": 0.000197981673409231,
      "learning_rate": 0.1257322996283441,
      "loss": 0.2538,
      "num_input_tokens_seen": 16832112,
      "step": 22070
    },
    {
      "epoch": 45.893970893970895,
      "grad_norm": 0.00019772742234636098,
      "learning_rate": 0.12567417264645994,
      "loss": 0.2208,
      "num_input_tokens_seen": 16835984,
      "step": 22075
    },
    {
      "epoch": 45.9043659043659,
      "grad_norm": 5.160536238690838e-05,
      "learning_rate": 0.12561604941592408,
      "loss": 0.3101,
      "num_input_tokens_seen": 16839888,
      "step": 22080
    },
    {
      "epoch": 45.91476091476091,
      "grad_norm": 6.732078327331692e-05,
      "learning_rate": 0.12555792994569978,
      "loss": 0.2735,
      "num_input_tokens_seen": 16843600,
      "step": 22085
    },
    {
      "epoch": 45.92515592515593,
      "grad_norm": 0.0004544588446151465,
      "learning_rate": 0.1254998142447499,
      "loss": 0.2676,
      "num_input_tokens_seen": 16847568,
      "step": 22090
    },
    {
      "epoch": 45.935550935550935,
      "grad_norm": 4.943975727655925e-05,
      "learning_rate": 0.1254417023220365,
      "loss": 0.263,
      "num_input_tokens_seen": 16851376,
      "step": 22095
    },
    {
      "epoch": 45.945945945945944,
      "grad_norm": 0.00025042553897947073,
      "learning_rate": 0.12538359418652126,
      "loss": 0.2797,
      "num_input_tokens_seen": 16855088,
      "step": 22100
    },
    {
      "epoch": 45.95634095634096,
      "grad_norm": 8.274863648694009e-05,
      "learning_rate": 0.12532548984716513,
      "loss": 0.2728,
      "num_input_tokens_seen": 16858864,
      "step": 22105
    },
    {
      "epoch": 45.96673596673597,
      "grad_norm": 0.00010645118891261518,
      "learning_rate": 0.12526738931292855,
      "loss": 0.265,
      "num_input_tokens_seen": 16862608,
      "step": 22110
    },
    {
      "epoch": 45.977130977130976,
      "grad_norm": 0.00015119793533813208,
      "learning_rate": 0.1252092925927714,
      "loss": 0.2646,
      "num_input_tokens_seen": 16866416,
      "step": 22115
    },
    {
      "epoch": 45.987525987525984,
      "grad_norm": 0.00017371904687024653,
      "learning_rate": 0.12515119969565278,
      "loss": 0.2775,
      "num_input_tokens_seen": 16870160,
      "step": 22120
    },
    {
      "epoch": 45.997920997921,
      "grad_norm": 0.00016638576926197857,
      "learning_rate": 0.12509311063053144,
      "loss": 0.2666,
      "num_input_tokens_seen": 16873872,
      "step": 22125
    },
    {
      "epoch": 46.00831600831601,
      "grad_norm": 0.0004470042185857892,
      "learning_rate": 0.1250350254063655,
      "loss": 0.2609,
      "num_input_tokens_seen": 16877672,
      "step": 22130
    },
    {
      "epoch": 46.018711018711016,
      "grad_norm": 0.0001230023626703769,
      "learning_rate": 0.1249769440321123,
      "loss": 0.2712,
      "num_input_tokens_seen": 16881512,
      "step": 22135
    },
    {
      "epoch": 46.02910602910603,
      "grad_norm": 7.218630344141275e-05,
      "learning_rate": 0.12491886651672884,
      "loss": 0.2638,
      "num_input_tokens_seen": 16885320,
      "step": 22140
    },
    {
      "epoch": 46.03950103950104,
      "grad_norm": 0.0003252634487580508,
      "learning_rate": 0.12486079286917139,
      "loss": 0.2761,
      "num_input_tokens_seen": 16889224,
      "step": 22145
    },
    {
      "epoch": 46.04989604989605,
      "grad_norm": 0.00032207617186941206,
      "learning_rate": 0.12480272309839553,
      "loss": 0.2598,
      "num_input_tokens_seen": 16893032,
      "step": 22150
    },
    {
      "epoch": 46.06029106029106,
      "grad_norm": 0.0003125071234535426,
      "learning_rate": 0.12474465721335648,
      "loss": 0.2944,
      "num_input_tokens_seen": 16896808,
      "step": 22155
    },
    {
      "epoch": 46.07068607068607,
      "grad_norm": 0.0003102972113993019,
      "learning_rate": 0.12468659522300861,
      "loss": 0.2682,
      "num_input_tokens_seen": 16900680,
      "step": 22160
    },
    {
      "epoch": 46.08108108108108,
      "grad_norm": 0.00018275667389389127,
      "learning_rate": 0.12462853713630584,
      "loss": 0.2626,
      "num_input_tokens_seen": 16904680,
      "step": 22165
    },
    {
      "epoch": 46.09147609147609,
      "grad_norm": 0.00015696596528869122,
      "learning_rate": 0.12457048296220156,
      "loss": 0.2624,
      "num_input_tokens_seen": 16908392,
      "step": 22170
    },
    {
      "epoch": 46.101871101871104,
      "grad_norm": 7.242074934765697e-05,
      "learning_rate": 0.12451243270964832,
      "loss": 0.2855,
      "num_input_tokens_seen": 16912200,
      "step": 22175
    },
    {
      "epoch": 46.11226611226611,
      "grad_norm": 0.00020489588496275246,
      "learning_rate": 0.12445438638759827,
      "loss": 0.2547,
      "num_input_tokens_seen": 16916072,
      "step": 22180
    },
    {
      "epoch": 46.12266112266112,
      "grad_norm": 0.0005089446785859764,
      "learning_rate": 0.1243963440050029,
      "loss": 0.2841,
      "num_input_tokens_seen": 16919976,
      "step": 22185
    },
    {
      "epoch": 46.13305613305613,
      "grad_norm": 0.0003243469982407987,
      "learning_rate": 0.12433830557081298,
      "loss": 0.2576,
      "num_input_tokens_seen": 16923720,
      "step": 22190
    },
    {
      "epoch": 46.143451143451145,
      "grad_norm": 0.00043446035124361515,
      "learning_rate": 0.12428027109397889,
      "loss": 0.2558,
      "num_input_tokens_seen": 16927688,
      "step": 22195
    },
    {
      "epoch": 46.15384615384615,
      "grad_norm": 0.0001216045202454552,
      "learning_rate": 0.12422224058345015,
      "loss": 0.2597,
      "num_input_tokens_seen": 16931560,
      "step": 22200
    },
    {
      "epoch": 46.15384615384615,
      "eval_loss": 0.24865058064460754,
      "eval_runtime": 13.3945,
      "eval_samples_per_second": 63.907,
      "eval_steps_per_second": 15.977,
      "num_input_tokens_seen": 16931560,
      "step": 22200
    },
    {
      "epoch": 46.16424116424116,
      "grad_norm": 0.0001540331431897357,
      "learning_rate": 0.12416421404817583,
      "loss": 0.2954,
      "num_input_tokens_seen": 16935400,
      "step": 22205
    },
    {
      "epoch": 46.17463617463618,
      "grad_norm": 0.00020360035705380142,
      "learning_rate": 0.12410619149710447,
      "loss": 0.2773,
      "num_input_tokens_seen": 16939208,
      "step": 22210
    },
    {
      "epoch": 46.185031185031185,
      "grad_norm": 0.00012955625425092876,
      "learning_rate": 0.12404817293918374,
      "loss": 0.2601,
      "num_input_tokens_seen": 16943112,
      "step": 22215
    },
    {
      "epoch": 46.195426195426194,
      "grad_norm": 0.00030278105987235904,
      "learning_rate": 0.12399015838336086,
      "loss": 0.2698,
      "num_input_tokens_seen": 16946920,
      "step": 22220
    },
    {
      "epoch": 46.20582120582121,
      "grad_norm": 0.000351201044395566,
      "learning_rate": 0.12393214783858246,
      "loss": 0.2542,
      "num_input_tokens_seen": 16950600,
      "step": 22225
    },
    {
      "epoch": 46.21621621621622,
      "grad_norm": 0.0005220185848884284,
      "learning_rate": 0.1238741413137944,
      "loss": 0.2409,
      "num_input_tokens_seen": 16954472,
      "step": 22230
    },
    {
      "epoch": 46.226611226611226,
      "grad_norm": 0.00029369135154411197,
      "learning_rate": 0.12381613881794212,
      "loss": 0.2901,
      "num_input_tokens_seen": 16958440,
      "step": 22235
    },
    {
      "epoch": 46.237006237006234,
      "grad_norm": 0.00020896198111586273,
      "learning_rate": 0.12375814035997022,
      "loss": 0.2628,
      "num_input_tokens_seen": 16962312,
      "step": 22240
    },
    {
      "epoch": 46.24740124740125,
      "grad_norm": 0.0006270611193031073,
      "learning_rate": 0.12370014594882285,
      "loss": 0.2468,
      "num_input_tokens_seen": 16966184,
      "step": 22245
    },
    {
      "epoch": 46.25779625779626,
      "grad_norm": 0.00010485291568329558,
      "learning_rate": 0.12364215559344356,
      "loss": 0.2621,
      "num_input_tokens_seen": 16970216,
      "step": 22250
    },
    {
      "epoch": 46.268191268191266,
      "grad_norm": 7.513363379985094e-05,
      "learning_rate": 0.12358416930277506,
      "loss": 0.2634,
      "num_input_tokens_seen": 16974056,
      "step": 22255
    },
    {
      "epoch": 46.27858627858628,
      "grad_norm": 0.00019662165141198784,
      "learning_rate": 0.1235261870857596,
      "loss": 0.2647,
      "num_input_tokens_seen": 16977864,
      "step": 22260
    },
    {
      "epoch": 46.28898128898129,
      "grad_norm": 0.0004189387254882604,
      "learning_rate": 0.12346820895133884,
      "loss": 0.2746,
      "num_input_tokens_seen": 16981576,
      "step": 22265
    },
    {
      "epoch": 46.2993762993763,
      "grad_norm": 0.0003355878870934248,
      "learning_rate": 0.12341023490845361,
      "loss": 0.2627,
      "num_input_tokens_seen": 16985224,
      "step": 22270
    },
    {
      "epoch": 46.30977130977131,
      "grad_norm": 0.000288547424133867,
      "learning_rate": 0.12335226496604437,
      "loss": 0.2521,
      "num_input_tokens_seen": 16988936,
      "step": 22275
    },
    {
      "epoch": 46.32016632016632,
      "grad_norm": 7.834876305423677e-05,
      "learning_rate": 0.12329429913305069,
      "loss": 0.2528,
      "num_input_tokens_seen": 16992808,
      "step": 22280
    },
    {
      "epoch": 46.33056133056133,
      "grad_norm": 0.0001973739272216335,
      "learning_rate": 0.12323633741841171,
      "loss": 0.2833,
      "num_input_tokens_seen": 16996680,
      "step": 22285
    },
    {
      "epoch": 46.34095634095634,
      "grad_norm": 0.00031665078131482005,
      "learning_rate": 0.12317837983106583,
      "loss": 0.2475,
      "num_input_tokens_seen": 17000456,
      "step": 22290
    },
    {
      "epoch": 46.351351351351354,
      "grad_norm": 0.0001569976011523977,
      "learning_rate": 0.12312042637995087,
      "loss": 0.2595,
      "num_input_tokens_seen": 17004168,
      "step": 22295
    },
    {
      "epoch": 46.36174636174636,
      "grad_norm": 5.593411697191186e-05,
      "learning_rate": 0.12306247707400389,
      "loss": 0.275,
      "num_input_tokens_seen": 17007976,
      "step": 22300
    },
    {
      "epoch": 46.37214137214137,
      "grad_norm": 0.00031737436074763536,
      "learning_rate": 0.12300453192216154,
      "loss": 0.2541,
      "num_input_tokens_seen": 17011944,
      "step": 22305
    },
    {
      "epoch": 46.38253638253638,
      "grad_norm": 0.00016067716933321208,
      "learning_rate": 0.12294659093335956,
      "loss": 0.2831,
      "num_input_tokens_seen": 17015688,
      "step": 22310
    },
    {
      "epoch": 46.392931392931395,
      "grad_norm": 0.0001555733324494213,
      "learning_rate": 0.12288865411653327,
      "loss": 0.2697,
      "num_input_tokens_seen": 17019464,
      "step": 22315
    },
    {
      "epoch": 46.4033264033264,
      "grad_norm": 0.0001924388052430004,
      "learning_rate": 0.12283072148061717,
      "loss": 0.2856,
      "num_input_tokens_seen": 17023176,
      "step": 22320
    },
    {
      "epoch": 46.41372141372141,
      "grad_norm": 0.00019046859233640134,
      "learning_rate": 0.12277279303454529,
      "loss": 0.2779,
      "num_input_tokens_seen": 17026952,
      "step": 22325
    },
    {
      "epoch": 46.42411642411643,
      "grad_norm": 0.0005711017292924225,
      "learning_rate": 0.12271486878725091,
      "loss": 0.2753,
      "num_input_tokens_seen": 17030728,
      "step": 22330
    },
    {
      "epoch": 46.434511434511435,
      "grad_norm": 0.0002052215568255633,
      "learning_rate": 0.12265694874766658,
      "loss": 0.2866,
      "num_input_tokens_seen": 17034536,
      "step": 22335
    },
    {
      "epoch": 46.444906444906444,
      "grad_norm": 0.00029944657580927014,
      "learning_rate": 0.12259903292472435,
      "loss": 0.2645,
      "num_input_tokens_seen": 17038472,
      "step": 22340
    },
    {
      "epoch": 46.45530145530145,
      "grad_norm": 0.00022672700288239866,
      "learning_rate": 0.12254112132735567,
      "loss": 0.2449,
      "num_input_tokens_seen": 17042248,
      "step": 22345
    },
    {
      "epoch": 46.46569646569647,
      "grad_norm": 0.0005383013049140573,
      "learning_rate": 0.12248321396449108,
      "loss": 0.2702,
      "num_input_tokens_seen": 17046184,
      "step": 22350
    },
    {
      "epoch": 46.476091476091476,
      "grad_norm": 0.0002797792258206755,
      "learning_rate": 0.12242531084506075,
      "loss": 0.2762,
      "num_input_tokens_seen": 17049864,
      "step": 22355
    },
    {
      "epoch": 46.486486486486484,
      "grad_norm": 0.0002097926044370979,
      "learning_rate": 0.122367411977994,
      "loss": 0.2496,
      "num_input_tokens_seen": 17053576,
      "step": 22360
    },
    {
      "epoch": 46.4968814968815,
      "grad_norm": 0.00028601373196579516,
      "learning_rate": 0.12230951737221954,
      "loss": 0.2597,
      "num_input_tokens_seen": 17057352,
      "step": 22365
    },
    {
      "epoch": 46.50727650727651,
      "grad_norm": 0.0004895106540061533,
      "learning_rate": 0.12225162703666555,
      "loss": 0.2658,
      "num_input_tokens_seen": 17061224,
      "step": 22370
    },
    {
      "epoch": 46.517671517671516,
      "grad_norm": 0.0003517533768899739,
      "learning_rate": 0.1221937409802593,
      "loss": 0.2405,
      "num_input_tokens_seen": 17064840,
      "step": 22375
    },
    {
      "epoch": 46.528066528066525,
      "grad_norm": 3.7381454603746533e-05,
      "learning_rate": 0.12213585921192768,
      "loss": 0.2638,
      "num_input_tokens_seen": 17068616,
      "step": 22380
    },
    {
      "epoch": 46.53846153846154,
      "grad_norm": 0.0004954173928126693,
      "learning_rate": 0.1220779817405967,
      "loss": 0.2754,
      "num_input_tokens_seen": 17072520,
      "step": 22385
    },
    {
      "epoch": 46.54885654885655,
      "grad_norm": 0.00013085729733575135,
      "learning_rate": 0.12202010857519181,
      "loss": 0.2824,
      "num_input_tokens_seen": 17076264,
      "step": 22390
    },
    {
      "epoch": 46.55925155925156,
      "grad_norm": 0.00023271025565918535,
      "learning_rate": 0.12196223972463785,
      "loss": 0.2693,
      "num_input_tokens_seen": 17080136,
      "step": 22395
    },
    {
      "epoch": 46.56964656964657,
      "grad_norm": 0.00025900444597937167,
      "learning_rate": 0.12190437519785885,
      "loss": 0.2781,
      "num_input_tokens_seen": 17083880,
      "step": 22400
    },
    {
      "epoch": 46.56964656964657,
      "eval_loss": 0.25912460684776306,
      "eval_runtime": 13.3799,
      "eval_samples_per_second": 63.976,
      "eval_steps_per_second": 15.994,
      "num_input_tokens_seen": 17083880,
      "step": 22400
    },
    {
      "epoch": 46.58004158004158,
      "grad_norm": 6.228961865417659e-05,
      "learning_rate": 0.12184651500377823,
      "loss": 0.2744,
      "num_input_tokens_seen": 17087880,
      "step": 22405
    },
    {
      "epoch": 46.59043659043659,
      "grad_norm": 7.690744678257033e-05,
      "learning_rate": 0.12178865915131885,
      "loss": 0.2647,
      "num_input_tokens_seen": 17091688,
      "step": 22410
    },
    {
      "epoch": 46.6008316008316,
      "grad_norm": 0.00030996662098914385,
      "learning_rate": 0.1217308076494027,
      "loss": 0.2421,
      "num_input_tokens_seen": 17095400,
      "step": 22415
    },
    {
      "epoch": 46.61122661122661,
      "grad_norm": 0.0003689787699840963,
      "learning_rate": 0.12167296050695134,
      "loss": 0.2943,
      "num_input_tokens_seen": 17099272,
      "step": 22420
    },
    {
      "epoch": 46.62162162162162,
      "grad_norm": 0.00016878554015420377,
      "learning_rate": 0.12161511773288536,
      "loss": 0.2675,
      "num_input_tokens_seen": 17103016,
      "step": 22425
    },
    {
      "epoch": 46.63201663201663,
      "grad_norm": 0.00015906919725239277,
      "learning_rate": 0.121557279336125,
      "loss": 0.2513,
      "num_input_tokens_seen": 17106728,
      "step": 22430
    },
    {
      "epoch": 46.642411642411645,
      "grad_norm": 0.0004891976714134216,
      "learning_rate": 0.12149944532558957,
      "loss": 0.2527,
      "num_input_tokens_seen": 17110568,
      "step": 22435
    },
    {
      "epoch": 46.65280665280665,
      "grad_norm": 0.00012872193474322557,
      "learning_rate": 0.12144161571019785,
      "loss": 0.261,
      "num_input_tokens_seen": 17114408,
      "step": 22440
    },
    {
      "epoch": 46.66320166320166,
      "grad_norm": 0.000386368774343282,
      "learning_rate": 0.12138379049886781,
      "loss": 0.2676,
      "num_input_tokens_seen": 17118088,
      "step": 22445
    },
    {
      "epoch": 46.67359667359668,
      "grad_norm": 0.00017513229977339506,
      "learning_rate": 0.12132596970051697,
      "loss": 0.2884,
      "num_input_tokens_seen": 17121864,
      "step": 22450
    },
    {
      "epoch": 46.683991683991685,
      "grad_norm": 0.0004549797740764916,
      "learning_rate": 0.12126815332406189,
      "loss": 0.2779,
      "num_input_tokens_seen": 17125544,
      "step": 22455
    },
    {
      "epoch": 46.694386694386694,
      "grad_norm": 0.0001555889321025461,
      "learning_rate": 0.12121034137841868,
      "loss": 0.2678,
      "num_input_tokens_seen": 17129256,
      "step": 22460
    },
    {
      "epoch": 46.7047817047817,
      "grad_norm": 0.000434901041444391,
      "learning_rate": 0.12115253387250258,
      "loss": 0.2624,
      "num_input_tokens_seen": 17133096,
      "step": 22465
    },
    {
      "epoch": 46.71517671517672,
      "grad_norm": 0.00012585431977640837,
      "learning_rate": 0.12109473081522831,
      "loss": 0.2499,
      "num_input_tokens_seen": 17136712,
      "step": 22470
    },
    {
      "epoch": 46.725571725571726,
      "grad_norm": 0.00016275845700874925,
      "learning_rate": 0.12103693221550982,
      "loss": 0.2534,
      "num_input_tokens_seen": 17140488,
      "step": 22475
    },
    {
      "epoch": 46.735966735966734,
      "grad_norm": 0.00021853778162039816,
      "learning_rate": 0.12097913808226027,
      "loss": 0.2505,
      "num_input_tokens_seen": 17144520,
      "step": 22480
    },
    {
      "epoch": 46.74636174636175,
      "grad_norm": 8.459365199087188e-05,
      "learning_rate": 0.12092134842439234,
      "loss": 0.2746,
      "num_input_tokens_seen": 17148296,
      "step": 22485
    },
    {
      "epoch": 46.75675675675676,
      "grad_norm": 9.782022971194237e-05,
      "learning_rate": 0.12086356325081798,
      "loss": 0.2927,
      "num_input_tokens_seen": 17152136,
      "step": 22490
    },
    {
      "epoch": 46.767151767151766,
      "grad_norm": 5.016908471588977e-05,
      "learning_rate": 0.12080578257044824,
      "loss": 0.2796,
      "num_input_tokens_seen": 17155848,
      "step": 22495
    },
    {
      "epoch": 46.777546777546775,
      "grad_norm": 0.00017641666636336595,
      "learning_rate": 0.12074800639219378,
      "loss": 0.2688,
      "num_input_tokens_seen": 17159720,
      "step": 22500
    },
    {
      "epoch": 46.78794178794179,
      "grad_norm": 0.0013393232366070151,
      "learning_rate": 0.12069023472496428,
      "loss": 0.2668,
      "num_input_tokens_seen": 17163560,
      "step": 22505
    },
    {
      "epoch": 46.7983367983368,
      "grad_norm": 0.00012373989738989621,
      "learning_rate": 0.12063246757766893,
      "loss": 0.2719,
      "num_input_tokens_seen": 17167304,
      "step": 22510
    },
    {
      "epoch": 46.80873180873181,
      "grad_norm": 9.288824367104098e-05,
      "learning_rate": 0.12057470495921618,
      "loss": 0.2603,
      "num_input_tokens_seen": 17171144,
      "step": 22515
    },
    {
      "epoch": 46.81912681912682,
      "grad_norm": 0.0002501018170733005,
      "learning_rate": 0.12051694687851364,
      "loss": 0.2506,
      "num_input_tokens_seen": 17175016,
      "step": 22520
    },
    {
      "epoch": 46.82952182952183,
      "grad_norm": 0.0002631780516821891,
      "learning_rate": 0.12045919334446839,
      "loss": 0.2585,
      "num_input_tokens_seen": 17178984,
      "step": 22525
    },
    {
      "epoch": 46.83991683991684,
      "grad_norm": 0.0005281229969114065,
      "learning_rate": 0.12040144436598683,
      "loss": 0.2568,
      "num_input_tokens_seen": 17182696,
      "step": 22530
    },
    {
      "epoch": 46.85031185031185,
      "grad_norm": 8.637745486339554e-05,
      "learning_rate": 0.12034369995197444,
      "loss": 0.2928,
      "num_input_tokens_seen": 17186376,
      "step": 22535
    },
    {
      "epoch": 46.86070686070686,
      "grad_norm": 4.935479228151962e-05,
      "learning_rate": 0.12028596011133627,
      "loss": 0.2581,
      "num_input_tokens_seen": 17190056,
      "step": 22540
    },
    {
      "epoch": 46.87110187110187,
      "grad_norm": 0.0001000378470052965,
      "learning_rate": 0.12022822485297643,
      "loss": 0.2628,
      "num_input_tokens_seen": 17193992,
      "step": 22545
    },
    {
      "epoch": 46.88149688149688,
      "grad_norm": 0.0003920421586371958,
      "learning_rate": 0.12017049418579843,
      "loss": 0.2618,
      "num_input_tokens_seen": 17197800,
      "step": 22550
    },
    {
      "epoch": 46.891891891891895,
      "grad_norm": 0.00031206224230118096,
      "learning_rate": 0.12011276811870514,
      "loss": 0.2529,
      "num_input_tokens_seen": 17201576,
      "step": 22555
    },
    {
      "epoch": 46.9022869022869,
      "grad_norm": 0.00015635065210517496,
      "learning_rate": 0.12005504666059852,
      "loss": 0.2822,
      "num_input_tokens_seen": 17205448,
      "step": 22560
    },
    {
      "epoch": 46.91268191268191,
      "grad_norm": 0.00044488132698461413,
      "learning_rate": 0.11999732982038003,
      "loss": 0.2573,
      "num_input_tokens_seen": 17209160,
      "step": 22565
    },
    {
      "epoch": 46.92307692307692,
      "grad_norm": 0.0001751144736772403,
      "learning_rate": 0.11993961760695038,
      "loss": 0.2656,
      "num_input_tokens_seen": 17212936,
      "step": 22570
    },
    {
      "epoch": 46.933471933471935,
      "grad_norm": 0.00012267709826119244,
      "learning_rate": 0.11988191002920942,
      "loss": 0.2515,
      "num_input_tokens_seen": 17216936,
      "step": 22575
    },
    {
      "epoch": 46.943866943866944,
      "grad_norm": 0.000148308725329116,
      "learning_rate": 0.11982420709605641,
      "loss": 0.2445,
      "num_input_tokens_seen": 17220680,
      "step": 22580
    },
    {
      "epoch": 46.95426195426195,
      "grad_norm": 6.409818888641894e-05,
      "learning_rate": 0.11976650881638991,
      "loss": 0.2813,
      "num_input_tokens_seen": 17224360,
      "step": 22585
    },
    {
      "epoch": 46.96465696465697,
      "grad_norm": 0.00019950537534896284,
      "learning_rate": 0.11970881519910764,
      "loss": 0.2372,
      "num_input_tokens_seen": 17228296,
      "step": 22590
    },
    {
      "epoch": 46.975051975051976,
      "grad_norm": 0.00038636833778582513,
      "learning_rate": 0.1196511262531068,
      "loss": 0.2836,
      "num_input_tokens_seen": 17232200,
      "step": 22595
    },
    {
      "epoch": 46.985446985446984,
      "grad_norm": 0.0002492482017260045,
      "learning_rate": 0.11959344198728361,
      "loss": 0.2659,
      "num_input_tokens_seen": 17235976,
      "step": 22600
    },
    {
      "epoch": 46.985446985446984,
      "eval_loss": 0.25825560092926025,
      "eval_runtime": 13.3845,
      "eval_samples_per_second": 63.954,
      "eval_steps_per_second": 15.989,
      "num_input_tokens_seen": 17235976,
      "step": 22600
    },
    {
      "epoch": 46.99584199584199,
      "grad_norm": 0.0002010871103266254,
      "learning_rate": 0.11953576241053378,
      "loss": 0.2625,
      "num_input_tokens_seen": 17239784,
      "step": 22605
    },
    {
      "epoch": 47.00623700623701,
      "grad_norm": 5.701424743165262e-05,
      "learning_rate": 0.11947808753175228,
      "loss": 0.2599,
      "num_input_tokens_seen": 17243352,
      "step": 22610
    },
    {
      "epoch": 47.016632016632016,
      "grad_norm": 0.00010315680265193805,
      "learning_rate": 0.1194204173598332,
      "loss": 0.2889,
      "num_input_tokens_seen": 17247224,
      "step": 22615
    },
    {
      "epoch": 47.027027027027025,
      "grad_norm": 0.00019773351959884167,
      "learning_rate": 0.11936275190367007,
      "loss": 0.2765,
      "num_input_tokens_seen": 17251128,
      "step": 22620
    },
    {
      "epoch": 47.03742203742204,
      "grad_norm": 0.00014053906488697976,
      "learning_rate": 0.11930509117215563,
      "loss": 0.249,
      "num_input_tokens_seen": 17254968,
      "step": 22625
    },
    {
      "epoch": 47.04781704781705,
      "grad_norm": 0.00018515456758905202,
      "learning_rate": 0.11924743517418179,
      "loss": 0.2724,
      "num_input_tokens_seen": 17258712,
      "step": 22630
    },
    {
      "epoch": 47.05821205821206,
      "grad_norm": 0.0003924377670045942,
      "learning_rate": 0.11918978391864,
      "loss": 0.2446,
      "num_input_tokens_seen": 17262616,
      "step": 22635
    },
    {
      "epoch": 47.06860706860707,
      "grad_norm": 0.00011191734665771946,
      "learning_rate": 0.11913213741442065,
      "loss": 0.2719,
      "num_input_tokens_seen": 17266520,
      "step": 22640
    },
    {
      "epoch": 47.07900207900208,
      "grad_norm": 8.632544631836936e-05,
      "learning_rate": 0.11907449567041364,
      "loss": 0.2707,
      "num_input_tokens_seen": 17270296,
      "step": 22645
    },
    {
      "epoch": 47.08939708939709,
      "grad_norm": 0.0003190569404978305,
      "learning_rate": 0.11901685869550803,
      "loss": 0.2564,
      "num_input_tokens_seen": 17274104,
      "step": 22650
    },
    {
      "epoch": 47.0997920997921,
      "grad_norm": 5.2312854677438736e-05,
      "learning_rate": 0.1189592264985922,
      "loss": 0.2432,
      "num_input_tokens_seen": 17277848,
      "step": 22655
    },
    {
      "epoch": 47.11018711018711,
      "grad_norm": 0.00035345632932148874,
      "learning_rate": 0.11890159908855373,
      "loss": 0.2804,
      "num_input_tokens_seen": 17281688,
      "step": 22660
    },
    {
      "epoch": 47.12058212058212,
      "grad_norm": 5.5318269005510956e-05,
      "learning_rate": 0.11884397647427941,
      "loss": 0.2714,
      "num_input_tokens_seen": 17285432,
      "step": 22665
    },
    {
      "epoch": 47.13097713097713,
      "grad_norm": 4.3209976865909994e-05,
      "learning_rate": 0.11878635866465546,
      "loss": 0.2578,
      "num_input_tokens_seen": 17289112,
      "step": 22670
    },
    {
      "epoch": 47.141372141372145,
      "grad_norm": 0.00018643047951627523,
      "learning_rate": 0.11872874566856734,
      "loss": 0.2663,
      "num_input_tokens_seen": 17292856,
      "step": 22675
    },
    {
      "epoch": 47.15176715176715,
      "grad_norm": 8.557530964026228e-05,
      "learning_rate": 0.11867113749489955,
      "loss": 0.2748,
      "num_input_tokens_seen": 17296888,
      "step": 22680
    },
    {
      "epoch": 47.16216216216216,
      "grad_norm": 0.00011160263966303319,
      "learning_rate": 0.11861353415253607,
      "loss": 0.2805,
      "num_input_tokens_seen": 17300664,
      "step": 22685
    },
    {
      "epoch": 47.17255717255717,
      "grad_norm": 0.00020285444043111056,
      "learning_rate": 0.11855593565036011,
      "loss": 0.2655,
      "num_input_tokens_seen": 17304472,
      "step": 22690
    },
    {
      "epoch": 47.182952182952185,
      "grad_norm": 0.00017236974963452667,
      "learning_rate": 0.11849834199725394,
      "loss": 0.2619,
      "num_input_tokens_seen": 17308472,
      "step": 22695
    },
    {
      "epoch": 47.19334719334719,
      "grad_norm": 0.0002156696718884632,
      "learning_rate": 0.1184407532020994,
      "loss": 0.2704,
      "num_input_tokens_seen": 17312312,
      "step": 22700
    },
    {
      "epoch": 47.2037422037422,
      "grad_norm": 7.599627133458853e-05,
      "learning_rate": 0.11838316927377723,
      "loss": 0.2385,
      "num_input_tokens_seen": 17316216,
      "step": 22705
    },
    {
      "epoch": 47.21413721413722,
      "grad_norm": 0.00021895587269682437,
      "learning_rate": 0.11832559022116766,
      "loss": 0.2873,
      "num_input_tokens_seen": 17319960,
      "step": 22710
    },
    {
      "epoch": 47.224532224532226,
      "grad_norm": 0.00022558463388122618,
      "learning_rate": 0.11826801605315022,
      "loss": 0.2727,
      "num_input_tokens_seen": 17323768,
      "step": 22715
    },
    {
      "epoch": 47.234927234927234,
      "grad_norm": 7.658988033654168e-05,
      "learning_rate": 0.1182104467786034,
      "loss": 0.2789,
      "num_input_tokens_seen": 17327704,
      "step": 22720
    },
    {
      "epoch": 47.24532224532224,
      "grad_norm": 0.000185832119314,
      "learning_rate": 0.1181528824064052,
      "loss": 0.2583,
      "num_input_tokens_seen": 17331544,
      "step": 22725
    },
    {
      "epoch": 47.25571725571726,
      "grad_norm": 6.409339403035119e-05,
      "learning_rate": 0.11809532294543279,
      "loss": 0.2793,
      "num_input_tokens_seen": 17335480,
      "step": 22730
    },
    {
      "epoch": 47.266112266112266,
      "grad_norm": 0.00033320419606752694,
      "learning_rate": 0.11803776840456245,
      "loss": 0.2652,
      "num_input_tokens_seen": 17339416,
      "step": 22735
    },
    {
      "epoch": 47.276507276507274,
      "grad_norm": 5.4971897043287754e-05,
      "learning_rate": 0.11798021879266997,
      "loss": 0.2806,
      "num_input_tokens_seen": 17343192,
      "step": 22740
    },
    {
      "epoch": 47.28690228690229,
      "grad_norm": 0.00010137514618691057,
      "learning_rate": 0.11792267411863006,
      "loss": 0.2747,
      "num_input_tokens_seen": 17347000,
      "step": 22745
    },
    {
      "epoch": 47.2972972972973,
      "grad_norm": 0.000202342082047835,
      "learning_rate": 0.1178651343913169,
      "loss": 0.2805,
      "num_input_tokens_seen": 17350776,
      "step": 22750
    },
    {
      "epoch": 47.30769230769231,
      "grad_norm": 0.00036567350616678596,
      "learning_rate": 0.11780759961960392,
      "loss": 0.2749,
      "num_input_tokens_seen": 17354488,
      "step": 22755
    },
    {
      "epoch": 47.318087318087315,
      "grad_norm": 0.0003085223725065589,
      "learning_rate": 0.1177500698123636,
      "loss": 0.2762,
      "num_input_tokens_seen": 17358200,
      "step": 22760
    },
    {
      "epoch": 47.32848232848233,
      "grad_norm": 0.00015022081788629293,
      "learning_rate": 0.11769254497846778,
      "loss": 0.2671,
      "num_input_tokens_seen": 17361848,
      "step": 22765
    },
    {
      "epoch": 47.33887733887734,
      "grad_norm": 0.0002670944668352604,
      "learning_rate": 0.11763502512678758,
      "loss": 0.2849,
      "num_input_tokens_seen": 17365528,
      "step": 22770
    },
    {
      "epoch": 47.34927234927235,
      "grad_norm": 0.00016507117834407836,
      "learning_rate": 0.11757751026619315,
      "loss": 0.2678,
      "num_input_tokens_seen": 17369464,
      "step": 22775
    },
    {
      "epoch": 47.35966735966736,
      "grad_norm": 0.0004100149089936167,
      "learning_rate": 0.11752000040555416,
      "loss": 0.2755,
      "num_input_tokens_seen": 17373176,
      "step": 22780
    },
    {
      "epoch": 47.37006237006237,
      "grad_norm": 0.0003952615079469979,
      "learning_rate": 0.11746249555373921,
      "loss": 0.266,
      "num_input_tokens_seen": 17376952,
      "step": 22785
    },
    {
      "epoch": 47.38045738045738,
      "grad_norm": 3.822296275757253e-05,
      "learning_rate": 0.11740499571961638,
      "loss": 0.2614,
      "num_input_tokens_seen": 17380600,
      "step": 22790
    },
    {
      "epoch": 47.39085239085239,
      "grad_norm": 0.0002612933167256415,
      "learning_rate": 0.11734750091205279,
      "loss": 0.2776,
      "num_input_tokens_seen": 17384248,
      "step": 22795
    },
    {
      "epoch": 47.4012474012474,
      "grad_norm": 0.00025843922048807144,
      "learning_rate": 0.11729001113991493,
      "loss": 0.2489,
      "num_input_tokens_seen": 17388152,
      "step": 22800
    },
    {
      "epoch": 47.4012474012474,
      "eval_loss": 0.24783852696418762,
      "eval_runtime": 13.386,
      "eval_samples_per_second": 63.947,
      "eval_steps_per_second": 15.987,
      "num_input_tokens_seen": 17388152,
      "step": 22800
    },
    {
      "epoch": 47.41164241164241,
      "grad_norm": 0.00036767584970220923,
      "learning_rate": 0.11723252641206837,
      "loss": 0.2573,
      "num_input_tokens_seen": 17391928,
      "step": 22805
    },
    {
      "epoch": 47.42203742203742,
      "grad_norm": 0.0002765416575130075,
      "learning_rate": 0.11717504673737808,
      "loss": 0.2895,
      "num_input_tokens_seen": 17395704,
      "step": 22810
    },
    {
      "epoch": 47.432432432432435,
      "grad_norm": 9.059199510375038e-05,
      "learning_rate": 0.11711757212470802,
      "loss": 0.2707,
      "num_input_tokens_seen": 17399416,
      "step": 22815
    },
    {
      "epoch": 47.44282744282744,
      "grad_norm": 0.0003309416933916509,
      "learning_rate": 0.11706010258292165,
      "loss": 0.2722,
      "num_input_tokens_seen": 17403192,
      "step": 22820
    },
    {
      "epoch": 47.45322245322245,
      "grad_norm": 0.0001459595951018855,
      "learning_rate": 0.11700263812088131,
      "loss": 0.2677,
      "num_input_tokens_seen": 17406904,
      "step": 22825
    },
    {
      "epoch": 47.46361746361746,
      "grad_norm": 0.00016632139158900827,
      "learning_rate": 0.11694517874744892,
      "loss": 0.2555,
      "num_input_tokens_seen": 17410744,
      "step": 22830
    },
    {
      "epoch": 47.474012474012476,
      "grad_norm": 6.510930688818917e-05,
      "learning_rate": 0.11688772447148532,
      "loss": 0.2443,
      "num_input_tokens_seen": 17414648,
      "step": 22835
    },
    {
      "epoch": 47.484407484407484,
      "grad_norm": 0.00019824094488285482,
      "learning_rate": 0.11683027530185074,
      "loss": 0.2732,
      "num_input_tokens_seen": 17418520,
      "step": 22840
    },
    {
      "epoch": 47.49480249480249,
      "grad_norm": 0.0002203934855060652,
      "learning_rate": 0.11677283124740451,
      "loss": 0.2723,
      "num_input_tokens_seen": 17422552,
      "step": 22845
    },
    {
      "epoch": 47.50519750519751,
      "grad_norm": 4.549244113150053e-05,
      "learning_rate": 0.11671539231700531,
      "loss": 0.2519,
      "num_input_tokens_seen": 17426296,
      "step": 22850
    },
    {
      "epoch": 47.515592515592516,
      "grad_norm": 0.00019838643493130803,
      "learning_rate": 0.11665795851951084,
      "loss": 0.2469,
      "num_input_tokens_seen": 17430104,
      "step": 22855
    },
    {
      "epoch": 47.525987525987524,
      "grad_norm": 0.0006660513463430107,
      "learning_rate": 0.11660052986377825,
      "loss": 0.3004,
      "num_input_tokens_seen": 17433912,
      "step": 22860
    },
    {
      "epoch": 47.53638253638254,
      "grad_norm": 8.565449388697743e-05,
      "learning_rate": 0.1165431063586636,
      "loss": 0.2468,
      "num_input_tokens_seen": 17437816,
      "step": 22865
    },
    {
      "epoch": 47.54677754677755,
      "grad_norm": 8.209392399294302e-05,
      "learning_rate": 0.11648568801302245,
      "loss": 0.2544,
      "num_input_tokens_seen": 17441656,
      "step": 22870
    },
    {
      "epoch": 47.55717255717256,
      "grad_norm": 0.0003781815175898373,
      "learning_rate": 0.11642827483570937,
      "loss": 0.2558,
      "num_input_tokens_seen": 17445400,
      "step": 22875
    },
    {
      "epoch": 47.567567567567565,
      "grad_norm": 0.00015365683066193014,
      "learning_rate": 0.11637086683557815,
      "loss": 0.2635,
      "num_input_tokens_seen": 17449272,
      "step": 22880
    },
    {
      "epoch": 47.57796257796258,
      "grad_norm": 0.00021864441805519164,
      "learning_rate": 0.11631346402148188,
      "loss": 0.2781,
      "num_input_tokens_seen": 17453016,
      "step": 22885
    },
    {
      "epoch": 47.58835758835759,
      "grad_norm": 0.00017906819994095713,
      "learning_rate": 0.11625606640227285,
      "loss": 0.2688,
      "num_input_tokens_seen": 17456984,
      "step": 22890
    },
    {
      "epoch": 47.5987525987526,
      "grad_norm": 0.0004667652247007936,
      "learning_rate": 0.11619867398680238,
      "loss": 0.2536,
      "num_input_tokens_seen": 17460856,
      "step": 22895
    },
    {
      "epoch": 47.60914760914761,
      "grad_norm": 0.0004989434382878244,
      "learning_rate": 0.11614128678392119,
      "loss": 0.2686,
      "num_input_tokens_seen": 17464536,
      "step": 22900
    },
    {
      "epoch": 47.61954261954262,
      "grad_norm": 0.0002659035089891404,
      "learning_rate": 0.11608390480247906,
      "loss": 0.2552,
      "num_input_tokens_seen": 17468376,
      "step": 22905
    },
    {
      "epoch": 47.62993762993763,
      "grad_norm": 0.000250072218477726,
      "learning_rate": 0.11602652805132499,
      "loss": 0.2761,
      "num_input_tokens_seen": 17472248,
      "step": 22910
    },
    {
      "epoch": 47.64033264033264,
      "grad_norm": 0.00027879804838448763,
      "learning_rate": 0.11596915653930731,
      "loss": 0.2389,
      "num_input_tokens_seen": 17476152,
      "step": 22915
    },
    {
      "epoch": 47.65072765072765,
      "grad_norm": 8.028848242247477e-05,
      "learning_rate": 0.11591179027527328,
      "loss": 0.2543,
      "num_input_tokens_seen": 17480056,
      "step": 22920
    },
    {
      "epoch": 47.66112266112266,
      "grad_norm": 0.00028588916757144034,
      "learning_rate": 0.11585442926806956,
      "loss": 0.2311,
      "num_input_tokens_seen": 17483768,
      "step": 22925
    },
    {
      "epoch": 47.67151767151767,
      "grad_norm": 8.197624265449122e-05,
      "learning_rate": 0.11579707352654202,
      "loss": 0.2659,
      "num_input_tokens_seen": 17487512,
      "step": 22930
    },
    {
      "epoch": 47.681912681912685,
      "grad_norm": 0.00015521433670073748,
      "learning_rate": 0.11573972305953548,
      "loss": 0.2504,
      "num_input_tokens_seen": 17491256,
      "step": 22935
    },
    {
      "epoch": 47.69230769230769,
      "grad_norm": 0.00012910990335512906,
      "learning_rate": 0.11568237787589426,
      "loss": 0.2515,
      "num_input_tokens_seen": 17495032,
      "step": 22940
    },
    {
      "epoch": 47.7027027027027,
      "grad_norm": 0.0001905964018078521,
      "learning_rate": 0.11562503798446161,
      "loss": 0.2442,
      "num_input_tokens_seen": 17498808,
      "step": 22945
    },
    {
      "epoch": 47.71309771309771,
      "grad_norm": 0.00010095864126924425,
      "learning_rate": 0.11556770339408005,
      "loss": 0.2892,
      "num_input_tokens_seen": 17502648,
      "step": 22950
    },
    {
      "epoch": 47.723492723492726,
      "grad_norm": 0.00013444237993098795,
      "learning_rate": 0.1155103741135914,
      "loss": 0.2563,
      "num_input_tokens_seen": 17506520,
      "step": 22955
    },
    {
      "epoch": 47.733887733887734,
      "grad_norm": 4.4178672396810725e-05,
      "learning_rate": 0.1154530501518364,
      "loss": 0.2645,
      "num_input_tokens_seen": 17510264,
      "step": 22960
    },
    {
      "epoch": 47.74428274428274,
      "grad_norm": 5.697973392670974e-05,
      "learning_rate": 0.11539573151765523,
      "loss": 0.2674,
      "num_input_tokens_seen": 17514232,
      "step": 22965
    },
    {
      "epoch": 47.75467775467776,
      "grad_norm": 0.00017630786169320345,
      "learning_rate": 0.11533841821988719,
      "loss": 0.2449,
      "num_input_tokens_seen": 17518104,
      "step": 22970
    },
    {
      "epoch": 47.765072765072766,
      "grad_norm": 0.00025050199474208057,
      "learning_rate": 0.11528111026737059,
      "loss": 0.2833,
      "num_input_tokens_seen": 17521848,
      "step": 22975
    },
    {
      "epoch": 47.775467775467774,
      "grad_norm": 0.0002874883357435465,
      "learning_rate": 0.11522380766894312,
      "loss": 0.2657,
      "num_input_tokens_seen": 17525624,
      "step": 22980
    },
    {
      "epoch": 47.78586278586278,
      "grad_norm": 0.00010081563959829509,
      "learning_rate": 0.11516651043344152,
      "loss": 0.2707,
      "num_input_tokens_seen": 17529496,
      "step": 22985
    },
    {
      "epoch": 47.7962577962578,
      "grad_norm": 0.00041786671499721706,
      "learning_rate": 0.11510921856970172,
      "loss": 0.2834,
      "num_input_tokens_seen": 17533240,
      "step": 22990
    },
    {
      "epoch": 47.80665280665281,
      "grad_norm": 0.00025546379038132727,
      "learning_rate": 0.11505193208655895,
      "loss": 0.2668,
      "num_input_tokens_seen": 17536984,
      "step": 22995
    },
    {
      "epoch": 47.817047817047815,
      "grad_norm": 0.0004786074860021472,
      "learning_rate": 0.11499465099284738,
      "loss": 0.2575,
      "num_input_tokens_seen": 17540824,
      "step": 23000
    },
    {
      "epoch": 47.817047817047815,
      "eval_loss": 0.251009076833725,
      "eval_runtime": 13.3981,
      "eval_samples_per_second": 63.89,
      "eval_steps_per_second": 15.972,
      "num_input_tokens_seen": 17540824,
      "step": 23000
    },
    {
      "epoch": 47.82744282744283,
      "grad_norm": 0.0003180828061886132,
      "learning_rate": 0.1149373752974006,
      "loss": 0.2523,
      "num_input_tokens_seen": 17544696,
      "step": 23005
    },
    {
      "epoch": 47.83783783783784,
      "grad_norm": 0.0008182342280633748,
      "learning_rate": 0.11488010500905109,
      "loss": 0.2746,
      "num_input_tokens_seen": 17548472,
      "step": 23010
    },
    {
      "epoch": 47.84823284823285,
      "grad_norm": 0.0004049630369991064,
      "learning_rate": 0.11482284013663077,
      "loss": 0.2793,
      "num_input_tokens_seen": 17552152,
      "step": 23015
    },
    {
      "epoch": 47.858627858627855,
      "grad_norm": 0.00036549242213368416,
      "learning_rate": 0.11476558068897061,
      "loss": 0.2777,
      "num_input_tokens_seen": 17556056,
      "step": 23020
    },
    {
      "epoch": 47.86902286902287,
      "grad_norm": 0.00013456857413984835,
      "learning_rate": 0.11470832667490061,
      "loss": 0.265,
      "num_input_tokens_seen": 17559800,
      "step": 23025
    },
    {
      "epoch": 47.87941787941788,
      "grad_norm": 0.0003066455537918955,
      "learning_rate": 0.11465107810325013,
      "loss": 0.2765,
      "num_input_tokens_seen": 17563640,
      "step": 23030
    },
    {
      "epoch": 47.88981288981289,
      "grad_norm": 0.000221941081690602,
      "learning_rate": 0.11459383498284771,
      "loss": 0.2782,
      "num_input_tokens_seen": 17567384,
      "step": 23035
    },
    {
      "epoch": 47.9002079002079,
      "grad_norm": 0.00026090172468684614,
      "learning_rate": 0.11453659732252082,
      "loss": 0.2637,
      "num_input_tokens_seen": 17571224,
      "step": 23040
    },
    {
      "epoch": 47.91060291060291,
      "grad_norm": 0.0005144834867678583,
      "learning_rate": 0.11447936513109633,
      "loss": 0.2416,
      "num_input_tokens_seen": 17575000,
      "step": 23045
    },
    {
      "epoch": 47.92099792099792,
      "grad_norm": 0.00011775966413551942,
      "learning_rate": 0.11442213841740011,
      "loss": 0.2657,
      "num_input_tokens_seen": 17578776,
      "step": 23050
    },
    {
      "epoch": 47.931392931392935,
      "grad_norm": 0.00058674969477579,
      "learning_rate": 0.1143649171902572,
      "loss": 0.287,
      "num_input_tokens_seen": 17582776,
      "step": 23055
    },
    {
      "epoch": 47.94178794178794,
      "grad_norm": 0.00017124043370131403,
      "learning_rate": 0.11430770145849194,
      "loss": 0.2658,
      "num_input_tokens_seen": 17586680,
      "step": 23060
    },
    {
      "epoch": 47.95218295218295,
      "grad_norm": 0.00039672275306656957,
      "learning_rate": 0.11425049123092756,
      "loss": 0.2675,
      "num_input_tokens_seen": 17590488,
      "step": 23065
    },
    {
      "epoch": 47.96257796257796,
      "grad_norm": 0.0003632518055383116,
      "learning_rate": 0.11419328651638674,
      "loss": 0.2606,
      "num_input_tokens_seen": 17594264,
      "step": 23070
    },
    {
      "epoch": 47.972972972972975,
      "grad_norm": 0.00014396029291674495,
      "learning_rate": 0.11413608732369115,
      "loss": 0.264,
      "num_input_tokens_seen": 17598040,
      "step": 23075
    },
    {
      "epoch": 47.983367983367984,
      "grad_norm": 0.00018414489750284702,
      "learning_rate": 0.11407889366166153,
      "loss": 0.2557,
      "num_input_tokens_seen": 17602040,
      "step": 23080
    },
    {
      "epoch": 47.99376299376299,
      "grad_norm": 0.00019989210704807192,
      "learning_rate": 0.11402170553911797,
      "loss": 0.2577,
      "num_input_tokens_seen": 17605976,
      "step": 23085
    },
    {
      "epoch": 48.00415800415801,
      "grad_norm": 0.00023254321422427893,
      "learning_rate": 0.11396452296487955,
      "loss": 0.2626,
      "num_input_tokens_seen": 17609752,
      "step": 23090
    },
    {
      "epoch": 48.014553014553016,
      "grad_norm": 0.00030729518039152026,
      "learning_rate": 0.11390734594776449,
      "loss": 0.2802,
      "num_input_tokens_seen": 17613656,
      "step": 23095
    },
    {
      "epoch": 48.024948024948024,
      "grad_norm": 5.417638749349862e-05,
      "learning_rate": 0.11385017449659031,
      "loss": 0.2817,
      "num_input_tokens_seen": 17617592,
      "step": 23100
    },
    {
      "epoch": 48.03534303534303,
      "grad_norm": 8.244667696999386e-05,
      "learning_rate": 0.11379300862017344,
      "loss": 0.2788,
      "num_input_tokens_seen": 17621464,
      "step": 23105
    },
    {
      "epoch": 48.04573804573805,
      "grad_norm": 0.0002353794261580333,
      "learning_rate": 0.11373584832732966,
      "loss": 0.2764,
      "num_input_tokens_seen": 17625240,
      "step": 23110
    },
    {
      "epoch": 48.056133056133056,
      "grad_norm": 0.00035346869844943285,
      "learning_rate": 0.11367869362687386,
      "loss": 0.2576,
      "num_input_tokens_seen": 17629048,
      "step": 23115
    },
    {
      "epoch": 48.066528066528065,
      "grad_norm": 0.00018778469529934227,
      "learning_rate": 0.11362154452761988,
      "loss": 0.2644,
      "num_input_tokens_seen": 17632920,
      "step": 23120
    },
    {
      "epoch": 48.07692307692308,
      "grad_norm": 0.00020919187227264047,
      "learning_rate": 0.11356440103838095,
      "loss": 0.2629,
      "num_input_tokens_seen": 17636664,
      "step": 23125
    },
    {
      "epoch": 48.08731808731809,
      "grad_norm": 0.00013233014033176005,
      "learning_rate": 0.11350726316796922,
      "loss": 0.318,
      "num_input_tokens_seen": 17640632,
      "step": 23130
    },
    {
      "epoch": 48.0977130977131,
      "grad_norm": 0.0001491636357968673,
      "learning_rate": 0.11345013092519607,
      "loss": 0.2836,
      "num_input_tokens_seen": 17644440,
      "step": 23135
    },
    {
      "epoch": 48.108108108108105,
      "grad_norm": 0.0004155202186666429,
      "learning_rate": 0.11339300431887213,
      "loss": 0.2812,
      "num_input_tokens_seen": 17648376,
      "step": 23140
    },
    {
      "epoch": 48.11850311850312,
      "grad_norm": 0.0001782684266800061,
      "learning_rate": 0.11333588335780687,
      "loss": 0.2826,
      "num_input_tokens_seen": 17652312,
      "step": 23145
    },
    {
      "epoch": 48.12889812889813,
      "grad_norm": 0.00014385893882717937,
      "learning_rate": 0.11327876805080916,
      "loss": 0.2722,
      "num_input_tokens_seen": 17655960,
      "step": 23150
    },
    {
      "epoch": 48.13929313929314,
      "grad_norm": 0.0001330423983745277,
      "learning_rate": 0.11322165840668696,
      "loss": 0.2744,
      "num_input_tokens_seen": 17659672,
      "step": 23155
    },
    {
      "epoch": 48.14968814968815,
      "grad_norm": 0.0007657487876713276,
      "learning_rate": 0.11316455443424717,
      "loss": 0.2575,
      "num_input_tokens_seen": 17663384,
      "step": 23160
    },
    {
      "epoch": 48.16008316008316,
      "grad_norm": 7.932808512123302e-05,
      "learning_rate": 0.11310745614229603,
      "loss": 0.2643,
      "num_input_tokens_seen": 17667320,
      "step": 23165
    },
    {
      "epoch": 48.17047817047817,
      "grad_norm": 0.00019462747150100768,
      "learning_rate": 0.1130503635396387,
      "loss": 0.2686,
      "num_input_tokens_seen": 17671128,
      "step": 23170
    },
    {
      "epoch": 48.18087318087318,
      "grad_norm": 0.00011662027100101113,
      "learning_rate": 0.11299327663507966,
      "loss": 0.263,
      "num_input_tokens_seen": 17674872,
      "step": 23175
    },
    {
      "epoch": 48.19126819126819,
      "grad_norm": 0.00014338426990434527,
      "learning_rate": 0.11293619543742246,
      "loss": 0.2695,
      "num_input_tokens_seen": 17678808,
      "step": 23180
    },
    {
      "epoch": 48.2016632016632,
      "grad_norm": 9.592835704097524e-05,
      "learning_rate": 0.11287911995546965,
      "loss": 0.2699,
      "num_input_tokens_seen": 17682616,
      "step": 23185
    },
    {
      "epoch": 48.21205821205821,
      "grad_norm": 0.0005872222245670855,
      "learning_rate": 0.11282205019802308,
      "loss": 0.2527,
      "num_input_tokens_seen": 17686488,
      "step": 23190
    },
    {
      "epoch": 48.222453222453225,
      "grad_norm": 0.0003532466071192175,
      "learning_rate": 0.11276498617388354,
      "loss": 0.2434,
      "num_input_tokens_seen": 17690264,
      "step": 23195
    },
    {
      "epoch": 48.232848232848234,
      "grad_norm": 0.00022323931625578552,
      "learning_rate": 0.11270792789185109,
      "loss": 0.2172,
      "num_input_tokens_seen": 17693912,
      "step": 23200
    },
    {
      "epoch": 48.232848232848234,
      "eval_loss": 0.25092241168022156,
      "eval_runtime": 13.3938,
      "eval_samples_per_second": 63.91,
      "eval_steps_per_second": 15.977,
      "num_input_tokens_seen": 17693912,
      "step": 23200
    },
    {
      "epoch": 48.24324324324324,
      "grad_norm": 0.0005901767290197313,
      "learning_rate": 0.11265087536072482,
      "loss": 0.2852,
      "num_input_tokens_seen": 17697752,
      "step": 23205
    },
    {
      "epoch": 48.25363825363825,
      "grad_norm": 7.44577992008999e-05,
      "learning_rate": 0.11259382858930288,
      "loss": 0.2625,
      "num_input_tokens_seen": 17701496,
      "step": 23210
    },
    {
      "epoch": 48.264033264033266,
      "grad_norm": 8.658089791424572e-05,
      "learning_rate": 0.11253678758638262,
      "loss": 0.2566,
      "num_input_tokens_seen": 17705432,
      "step": 23215
    },
    {
      "epoch": 48.274428274428274,
      "grad_norm": 0.00031267423764802516,
      "learning_rate": 0.11247975236076059,
      "loss": 0.266,
      "num_input_tokens_seen": 17709208,
      "step": 23220
    },
    {
      "epoch": 48.28482328482328,
      "grad_norm": 0.00010747231135610491,
      "learning_rate": 0.11242272292123218,
      "loss": 0.2787,
      "num_input_tokens_seen": 17713176,
      "step": 23225
    },
    {
      "epoch": 48.2952182952183,
      "grad_norm": 0.0002441554388497025,
      "learning_rate": 0.11236569927659217,
      "loss": 0.2448,
      "num_input_tokens_seen": 17716952,
      "step": 23230
    },
    {
      "epoch": 48.305613305613306,
      "grad_norm": 9.575235890224576e-05,
      "learning_rate": 0.11230868143563429,
      "loss": 0.2276,
      "num_input_tokens_seen": 17720760,
      "step": 23235
    },
    {
      "epoch": 48.316008316008315,
      "grad_norm": 0.00047440320486202836,
      "learning_rate": 0.11225166940715131,
      "loss": 0.2973,
      "num_input_tokens_seen": 17724568,
      "step": 23240
    },
    {
      "epoch": 48.32640332640332,
      "grad_norm": 0.00022464468202088028,
      "learning_rate": 0.11219466319993537,
      "loss": 0.2286,
      "num_input_tokens_seen": 17728344,
      "step": 23245
    },
    {
      "epoch": 48.33679833679834,
      "grad_norm": 0.0004839448956772685,
      "learning_rate": 0.11213766282277739,
      "loss": 0.2855,
      "num_input_tokens_seen": 17732248,
      "step": 23250
    },
    {
      "epoch": 48.34719334719335,
      "grad_norm": 5.8999121392844245e-05,
      "learning_rate": 0.11208066828446761,
      "loss": 0.2407,
      "num_input_tokens_seen": 17736248,
      "step": 23255
    },
    {
      "epoch": 48.357588357588355,
      "grad_norm": 0.00030349905136972666,
      "learning_rate": 0.11202367959379537,
      "loss": 0.26,
      "num_input_tokens_seen": 17740056,
      "step": 23260
    },
    {
      "epoch": 48.36798336798337,
      "grad_norm": 0.0002658960875123739,
      "learning_rate": 0.11196669675954894,
      "loss": 0.2766,
      "num_input_tokens_seen": 17743832,
      "step": 23265
    },
    {
      "epoch": 48.37837837837838,
      "grad_norm": 0.0005035450449213386,
      "learning_rate": 0.1119097197905158,
      "loss": 0.2682,
      "num_input_tokens_seen": 17747608,
      "step": 23270
    },
    {
      "epoch": 48.38877338877339,
      "grad_norm": 0.0001554350310470909,
      "learning_rate": 0.11185274869548259,
      "loss": 0.2652,
      "num_input_tokens_seen": 17751352,
      "step": 23275
    },
    {
      "epoch": 48.3991683991684,
      "grad_norm": 0.00022462457127403468,
      "learning_rate": 0.11179578348323486,
      "loss": 0.2539,
      "num_input_tokens_seen": 17755096,
      "step": 23280
    },
    {
      "epoch": 48.40956340956341,
      "grad_norm": 0.00021113056573085487,
      "learning_rate": 0.1117388241625575,
      "loss": 0.2685,
      "num_input_tokens_seen": 17758872,
      "step": 23285
    },
    {
      "epoch": 48.41995841995842,
      "grad_norm": 0.00027132206014357507,
      "learning_rate": 0.11168187074223421,
      "loss": 0.2534,
      "num_input_tokens_seen": 17762744,
      "step": 23290
    },
    {
      "epoch": 48.43035343035343,
      "grad_norm": 0.00011608129716478288,
      "learning_rate": 0.11162492323104796,
      "loss": 0.2681,
      "num_input_tokens_seen": 17766648,
      "step": 23295
    },
    {
      "epoch": 48.44074844074844,
      "grad_norm": 8.047799929045141e-05,
      "learning_rate": 0.11156798163778091,
      "loss": 0.2618,
      "num_input_tokens_seen": 17770392,
      "step": 23300
    },
    {
      "epoch": 48.45114345114345,
      "grad_norm": 0.00026496706414036453,
      "learning_rate": 0.11151104597121399,
      "loss": 0.26,
      "num_input_tokens_seen": 17774264,
      "step": 23305
    },
    {
      "epoch": 48.46153846153846,
      "grad_norm": 0.00016956462059170008,
      "learning_rate": 0.11145411624012742,
      "loss": 0.2732,
      "num_input_tokens_seen": 17778200,
      "step": 23310
    },
    {
      "epoch": 48.471933471933475,
      "grad_norm": 9.155752923106775e-05,
      "learning_rate": 0.11139719245330063,
      "loss": 0.2793,
      "num_input_tokens_seen": 17782104,
      "step": 23315
    },
    {
      "epoch": 48.482328482328484,
      "grad_norm": 0.0002794752363115549,
      "learning_rate": 0.11134027461951179,
      "loss": 0.2746,
      "num_input_tokens_seen": 17785944,
      "step": 23320
    },
    {
      "epoch": 48.49272349272349,
      "grad_norm": 0.0001991379540413618,
      "learning_rate": 0.11128336274753849,
      "loss": 0.2519,
      "num_input_tokens_seen": 17789720,
      "step": 23325
    },
    {
      "epoch": 48.5031185031185,
      "grad_norm": 0.0002663952764123678,
      "learning_rate": 0.11122645684615715,
      "loss": 0.2371,
      "num_input_tokens_seen": 17793528,
      "step": 23330
    },
    {
      "epoch": 48.513513513513516,
      "grad_norm": 0.00020975255756638944,
      "learning_rate": 0.11116955692414345,
      "loss": 0.261,
      "num_input_tokens_seen": 17797336,
      "step": 23335
    },
    {
      "epoch": 48.523908523908524,
      "grad_norm": 0.00032962270779535174,
      "learning_rate": 0.11111266299027203,
      "loss": 0.2886,
      "num_input_tokens_seen": 17801208,
      "step": 23340
    },
    {
      "epoch": 48.53430353430353,
      "grad_norm": 9.004686580738053e-05,
      "learning_rate": 0.11105577505331668,
      "loss": 0.2651,
      "num_input_tokens_seen": 17805016,
      "step": 23345
    },
    {
      "epoch": 48.54469854469855,
      "grad_norm": 8.469255408272147e-05,
      "learning_rate": 0.11099889312205018,
      "loss": 0.2646,
      "num_input_tokens_seen": 17808760,
      "step": 23350
    },
    {
      "epoch": 48.555093555093556,
      "grad_norm": 0.00032976423972286284,
      "learning_rate": 0.11094201720524455,
      "loss": 0.2594,
      "num_input_tokens_seen": 17812472,
      "step": 23355
    },
    {
      "epoch": 48.565488565488565,
      "grad_norm": 0.0001228516921401024,
      "learning_rate": 0.11088514731167064,
      "loss": 0.2511,
      "num_input_tokens_seen": 17816152,
      "step": 23360
    },
    {
      "epoch": 48.57588357588357,
      "grad_norm": 0.00020506713190115988,
      "learning_rate": 0.11082828345009862,
      "loss": 0.2847,
      "num_input_tokens_seen": 17819992,
      "step": 23365
    },
    {
      "epoch": 48.58627858627859,
      "grad_norm": 0.00025999650824815035,
      "learning_rate": 0.11077142562929748,
      "loss": 0.2572,
      "num_input_tokens_seen": 17823800,
      "step": 23370
    },
    {
      "epoch": 48.5966735966736,
      "grad_norm": 0.00010288695193594322,
      "learning_rate": 0.11071457385803554,
      "loss": 0.2734,
      "num_input_tokens_seen": 17827576,
      "step": 23375
    },
    {
      "epoch": 48.607068607068605,
      "grad_norm": 0.00025981085491366684,
      "learning_rate": 0.11065772814508001,
      "loss": 0.2503,
      "num_input_tokens_seen": 17831384,
      "step": 23380
    },
    {
      "epoch": 48.61746361746362,
      "grad_norm": 0.00020400645735207945,
      "learning_rate": 0.11060088849919715,
      "loss": 0.2829,
      "num_input_tokens_seen": 17835160,
      "step": 23385
    },
    {
      "epoch": 48.62785862785863,
      "grad_norm": 0.0005605733022093773,
      "learning_rate": 0.11054405492915244,
      "loss": 0.27,
      "num_input_tokens_seen": 17838840,
      "step": 23390
    },
    {
      "epoch": 48.63825363825364,
      "grad_norm": 0.00032936997013166547,
      "learning_rate": 0.11048722744371031,
      "loss": 0.2396,
      "num_input_tokens_seen": 17842584,
      "step": 23395
    },
    {
      "epoch": 48.648648648648646,
      "grad_norm": 9.78625103016384e-05,
      "learning_rate": 0.1104304060516342,
      "loss": 0.2799,
      "num_input_tokens_seen": 17846296,
      "step": 23400
    },
    {
      "epoch": 48.648648648648646,
      "eval_loss": 0.2471221536397934,
      "eval_runtime": 13.3803,
      "eval_samples_per_second": 63.975,
      "eval_steps_per_second": 15.994,
      "num_input_tokens_seen": 17846296,
      "step": 23400
    },
    {
      "epoch": 48.65904365904366,
      "grad_norm": 0.00019502625218592584,
      "learning_rate": 0.11037359076168682,
      "loss": 0.2664,
      "num_input_tokens_seen": 17850424,
      "step": 23405
    },
    {
      "epoch": 48.66943866943867,
      "grad_norm": 0.00023033835168462247,
      "learning_rate": 0.11031678158262966,
      "loss": 0.2524,
      "num_input_tokens_seen": 17854296,
      "step": 23410
    },
    {
      "epoch": 48.67983367983368,
      "grad_norm": 0.00046526658115908504,
      "learning_rate": 0.11025997852322349,
      "loss": 0.2632,
      "num_input_tokens_seen": 17858008,
      "step": 23415
    },
    {
      "epoch": 48.69022869022869,
      "grad_norm": 0.0001972712343558669,
      "learning_rate": 0.11020318159222807,
      "loss": 0.2887,
      "num_input_tokens_seen": 17861848,
      "step": 23420
    },
    {
      "epoch": 48.7006237006237,
      "grad_norm": 0.0003987428790424019,
      "learning_rate": 0.1101463907984021,
      "loss": 0.2798,
      "num_input_tokens_seen": 17865720,
      "step": 23425
    },
    {
      "epoch": 48.71101871101871,
      "grad_norm": 0.0002829173463396728,
      "learning_rate": 0.11008960615050352,
      "loss": 0.2639,
      "num_input_tokens_seen": 17869592,
      "step": 23430
    },
    {
      "epoch": 48.72141372141372,
      "grad_norm": 0.00013211939949542284,
      "learning_rate": 0.11003282765728925,
      "loss": 0.2692,
      "num_input_tokens_seen": 17873368,
      "step": 23435
    },
    {
      "epoch": 48.731808731808734,
      "grad_norm": 0.000244815309997648,
      "learning_rate": 0.10997605532751518,
      "loss": 0.2435,
      "num_input_tokens_seen": 17877176,
      "step": 23440
    },
    {
      "epoch": 48.74220374220374,
      "grad_norm": 0.00041402026545256376,
      "learning_rate": 0.1099192891699364,
      "loss": 0.2901,
      "num_input_tokens_seen": 17880824,
      "step": 23445
    },
    {
      "epoch": 48.75259875259875,
      "grad_norm": 0.0005632441025227308,
      "learning_rate": 0.10986252919330687,
      "loss": 0.3007,
      "num_input_tokens_seen": 17884632,
      "step": 23450
    },
    {
      "epoch": 48.762993762993766,
      "grad_norm": 0.00038054314791224897,
      "learning_rate": 0.10980577540637973,
      "loss": 0.2627,
      "num_input_tokens_seen": 17888568,
      "step": 23455
    },
    {
      "epoch": 48.773388773388774,
      "grad_norm": 0.00019710694323293865,
      "learning_rate": 0.10974902781790719,
      "loss": 0.256,
      "num_input_tokens_seen": 17892280,
      "step": 23460
    },
    {
      "epoch": 48.78378378378378,
      "grad_norm": 0.00022718824038747698,
      "learning_rate": 0.10969228643664032,
      "loss": 0.2644,
      "num_input_tokens_seen": 17895992,
      "step": 23465
    },
    {
      "epoch": 48.79417879417879,
      "grad_norm": 0.00016650119505356997,
      "learning_rate": 0.10963555127132942,
      "loss": 0.2628,
      "num_input_tokens_seen": 17899928,
      "step": 23470
    },
    {
      "epoch": 48.804573804573806,
      "grad_norm": 0.00012927359784953296,
      "learning_rate": 0.10957882233072382,
      "loss": 0.2915,
      "num_input_tokens_seen": 17903736,
      "step": 23475
    },
    {
      "epoch": 48.814968814968815,
      "grad_norm": 0.0006165364757180214,
      "learning_rate": 0.10952209962357176,
      "loss": 0.2735,
      "num_input_tokens_seen": 17907448,
      "step": 23480
    },
    {
      "epoch": 48.82536382536382,
      "grad_norm": 0.0004893646691925824,
      "learning_rate": 0.10946538315862062,
      "loss": 0.2459,
      "num_input_tokens_seen": 17911288,
      "step": 23485
    },
    {
      "epoch": 48.83575883575884,
      "grad_norm": 0.00021871081844437867,
      "learning_rate": 0.10940867294461679,
      "loss": 0.2382,
      "num_input_tokens_seen": 17915032,
      "step": 23490
    },
    {
      "epoch": 48.84615384615385,
      "grad_norm": 8.697227895027027e-05,
      "learning_rate": 0.10935196899030565,
      "loss": 0.2648,
      "num_input_tokens_seen": 17918904,
      "step": 23495
    },
    {
      "epoch": 48.856548856548855,
      "grad_norm": 0.0001984906557481736,
      "learning_rate": 0.10929527130443177,
      "loss": 0.2817,
      "num_input_tokens_seen": 17922680,
      "step": 23500
    },
    {
      "epoch": 48.86694386694387,
      "grad_norm": 0.00023849379795137793,
      "learning_rate": 0.1092385798957385,
      "loss": 0.2829,
      "num_input_tokens_seen": 17926488,
      "step": 23505
    },
    {
      "epoch": 48.87733887733888,
      "grad_norm": 0.00038961332757025957,
      "learning_rate": 0.10918189477296848,
      "loss": 0.2705,
      "num_input_tokens_seen": 17930296,
      "step": 23510
    },
    {
      "epoch": 48.88773388773389,
      "grad_norm": 7.555889169452712e-05,
      "learning_rate": 0.1091252159448633,
      "loss": 0.2771,
      "num_input_tokens_seen": 17933944,
      "step": 23515
    },
    {
      "epoch": 48.898128898128896,
      "grad_norm": 0.00027400473481975496,
      "learning_rate": 0.10906854342016345,
      "loss": 0.2724,
      "num_input_tokens_seen": 17937720,
      "step": 23520
    },
    {
      "epoch": 48.90852390852391,
      "grad_norm": 0.0003160927153658122,
      "learning_rate": 0.10901187720760858,
      "loss": 0.2696,
      "num_input_tokens_seen": 17941496,
      "step": 23525
    },
    {
      "epoch": 48.91891891891892,
      "grad_norm": 5.8038014685735106e-05,
      "learning_rate": 0.10895521731593734,
      "loss": 0.2712,
      "num_input_tokens_seen": 17945368,
      "step": 23530
    },
    {
      "epoch": 48.92931392931393,
      "grad_norm": 0.0006155946757644415,
      "learning_rate": 0.10889856375388733,
      "loss": 0.2341,
      "num_input_tokens_seen": 17949112,
      "step": 23535
    },
    {
      "epoch": 48.93970893970894,
      "grad_norm": 0.00010973793541779742,
      "learning_rate": 0.1088419165301954,
      "loss": 0.2541,
      "num_input_tokens_seen": 17952984,
      "step": 23540
    },
    {
      "epoch": 48.95010395010395,
      "grad_norm": 0.00017801918147597462,
      "learning_rate": 0.1087852756535971,
      "loss": 0.268,
      "num_input_tokens_seen": 17956824,
      "step": 23545
    },
    {
      "epoch": 48.96049896049896,
      "grad_norm": 9.488354407949373e-05,
      "learning_rate": 0.10872864113282725,
      "loss": 0.3091,
      "num_input_tokens_seen": 17960728,
      "step": 23550
    },
    {
      "epoch": 48.97089397089397,
      "grad_norm": 8.429075387539342e-05,
      "learning_rate": 0.10867201297661958,
      "loss": 0.2603,
      "num_input_tokens_seen": 17964536,
      "step": 23555
    },
    {
      "epoch": 48.981288981288984,
      "grad_norm": 0.0001953737810254097,
      "learning_rate": 0.10861539119370689,
      "loss": 0.2564,
      "num_input_tokens_seen": 17968440,
      "step": 23560
    },
    {
      "epoch": 48.99168399168399,
      "grad_norm": 0.00023910404706839472,
      "learning_rate": 0.10855877579282096,
      "loss": 0.2613,
      "num_input_tokens_seen": 17972152,
      "step": 23565
    },
    {
      "epoch": 49.002079002079,
      "grad_norm": 0.0002669554087333381,
      "learning_rate": 0.10850216678269252,
      "loss": 0.266,
      "num_input_tokens_seen": 17975944,
      "step": 23570
    },
    {
      "epoch": 49.012474012474016,
      "grad_norm": 0.0006303474074229598,
      "learning_rate": 0.10844556417205146,
      "loss": 0.3078,
      "num_input_tokens_seen": 17979752,
      "step": 23575
    },
    {
      "epoch": 49.022869022869024,
      "grad_norm": 0.00017910836322698742,
      "learning_rate": 0.10838896796962669,
      "loss": 0.2597,
      "num_input_tokens_seen": 17983496,
      "step": 23580
    },
    {
      "epoch": 49.03326403326403,
      "grad_norm": 0.00041906916885636747,
      "learning_rate": 0.1083323781841459,
      "loss": 0.2715,
      "num_input_tokens_seen": 17987176,
      "step": 23585
    },
    {
      "epoch": 49.04365904365904,
      "grad_norm": 0.0002511486236471683,
      "learning_rate": 0.10827579482433607,
      "loss": 0.2717,
      "num_input_tokens_seen": 17991112,
      "step": 23590
    },
    {
      "epoch": 49.054054054054056,
      "grad_norm": 5.151348523213528e-05,
      "learning_rate": 0.10821921789892304,
      "loss": 0.2707,
      "num_input_tokens_seen": 17994920,
      "step": 23595
    },
    {
      "epoch": 49.064449064449065,
      "grad_norm": 0.0002489004691597074,
      "learning_rate": 0.10816264741663158,
      "loss": 0.2467,
      "num_input_tokens_seen": 17998760,
      "step": 23600
    },
    {
      "epoch": 49.064449064449065,
      "eval_loss": 0.2500828802585602,
      "eval_runtime": 13.3946,
      "eval_samples_per_second": 63.906,
      "eval_steps_per_second": 15.977,
      "num_input_tokens_seen": 17998760,
      "step": 23600
    },
    {
      "epoch": 49.07484407484407,
      "grad_norm": 0.0001256687828572467,
      "learning_rate": 0.10810608338618573,
      "loss": 0.2684,
      "num_input_tokens_seen": 18002600,
      "step": 23605
    },
    {
      "epoch": 49.08523908523909,
      "grad_norm": 0.000493725819978863,
      "learning_rate": 0.10804952581630821,
      "loss": 0.2689,
      "num_input_tokens_seen": 18006536,
      "step": 23610
    },
    {
      "epoch": 49.0956340956341,
      "grad_norm": 0.00014630363148171455,
      "learning_rate": 0.10799297471572102,
      "loss": 0.2493,
      "num_input_tokens_seen": 18010504,
      "step": 23615
    },
    {
      "epoch": 49.106029106029105,
      "grad_norm": 4.367139263194986e-05,
      "learning_rate": 0.10793643009314507,
      "loss": 0.2658,
      "num_input_tokens_seen": 18014152,
      "step": 23620
    },
    {
      "epoch": 49.11642411642411,
      "grad_norm": 6.575254519702867e-05,
      "learning_rate": 0.10787989195730015,
      "loss": 0.2576,
      "num_input_tokens_seen": 18018120,
      "step": 23625
    },
    {
      "epoch": 49.12681912681913,
      "grad_norm": 0.0002481926931068301,
      "learning_rate": 0.10782336031690525,
      "loss": 0.2816,
      "num_input_tokens_seen": 18021928,
      "step": 23630
    },
    {
      "epoch": 49.13721413721414,
      "grad_norm": 0.00010424883657833561,
      "learning_rate": 0.10776683518067821,
      "loss": 0.2593,
      "num_input_tokens_seen": 18025864,
      "step": 23635
    },
    {
      "epoch": 49.147609147609145,
      "grad_norm": 0.00010949529678327963,
      "learning_rate": 0.10771031655733587,
      "loss": 0.2535,
      "num_input_tokens_seen": 18029672,
      "step": 23640
    },
    {
      "epoch": 49.15800415800416,
      "grad_norm": 0.00022108136909082532,
      "learning_rate": 0.10765380445559422,
      "loss": 0.2812,
      "num_input_tokens_seen": 18033672,
      "step": 23645
    },
    {
      "epoch": 49.16839916839917,
      "grad_norm": 0.00029403407825157046,
      "learning_rate": 0.10759729888416801,
      "loss": 0.2787,
      "num_input_tokens_seen": 18037576,
      "step": 23650
    },
    {
      "epoch": 49.17879417879418,
      "grad_norm": 9.798179962672293e-05,
      "learning_rate": 0.10754079985177119,
      "loss": 0.2593,
      "num_input_tokens_seen": 18041288,
      "step": 23655
    },
    {
      "epoch": 49.189189189189186,
      "grad_norm": 0.00022263005666900426,
      "learning_rate": 0.10748430736711667,
      "loss": 0.2696,
      "num_input_tokens_seen": 18045128,
      "step": 23660
    },
    {
      "epoch": 49.1995841995842,
      "grad_norm": 0.0002532023936510086,
      "learning_rate": 0.10742782143891623,
      "loss": 0.2423,
      "num_input_tokens_seen": 18048968,
      "step": 23665
    },
    {
      "epoch": 49.20997920997921,
      "grad_norm": 0.00018222650396637619,
      "learning_rate": 0.10737134207588069,
      "loss": 0.2709,
      "num_input_tokens_seen": 18052808,
      "step": 23670
    },
    {
      "epoch": 49.22037422037422,
      "grad_norm": 0.00020528025925159454,
      "learning_rate": 0.10731486928671992,
      "loss": 0.2744,
      "num_input_tokens_seen": 18056680,
      "step": 23675
    },
    {
      "epoch": 49.23076923076923,
      "grad_norm": 7.971689046826214e-05,
      "learning_rate": 0.10725840308014269,
      "loss": 0.2529,
      "num_input_tokens_seen": 18060488,
      "step": 23680
    },
    {
      "epoch": 49.24116424116424,
      "grad_norm": 8.92926545930095e-05,
      "learning_rate": 0.10720194346485688,
      "loss": 0.2646,
      "num_input_tokens_seen": 18064136,
      "step": 23685
    },
    {
      "epoch": 49.25155925155925,
      "grad_norm": 0.00013779282744508237,
      "learning_rate": 0.10714549044956918,
      "loss": 0.2721,
      "num_input_tokens_seen": 18067912,
      "step": 23690
    },
    {
      "epoch": 49.26195426195426,
      "grad_norm": 0.00032711660605855286,
      "learning_rate": 0.10708904404298542,
      "loss": 0.2601,
      "num_input_tokens_seen": 18071848,
      "step": 23695
    },
    {
      "epoch": 49.272349272349274,
      "grad_norm": 8.303031791001558e-05,
      "learning_rate": 0.1070326042538103,
      "loss": 0.258,
      "num_input_tokens_seen": 18075656,
      "step": 23700
    },
    {
      "epoch": 49.28274428274428,
      "grad_norm": 7.937040936667472e-05,
      "learning_rate": 0.10697617109074758,
      "loss": 0.2478,
      "num_input_tokens_seen": 18079464,
      "step": 23705
    },
    {
      "epoch": 49.29313929313929,
      "grad_norm": 4.8353205784223974e-05,
      "learning_rate": 0.10691974456249999,
      "loss": 0.2765,
      "num_input_tokens_seen": 18083240,
      "step": 23710
    },
    {
      "epoch": 49.303534303534306,
      "grad_norm": 0.0004748390056192875,
      "learning_rate": 0.10686332467776909,
      "loss": 0.2424,
      "num_input_tokens_seen": 18087112,
      "step": 23715
    },
    {
      "epoch": 49.313929313929314,
      "grad_norm": 0.00023989638430066407,
      "learning_rate": 0.10680691144525563,
      "loss": 0.262,
      "num_input_tokens_seen": 18090920,
      "step": 23720
    },
    {
      "epoch": 49.32432432432432,
      "grad_norm": 9.276289347326383e-05,
      "learning_rate": 0.10675050487365928,
      "loss": 0.2614,
      "num_input_tokens_seen": 18094760,
      "step": 23725
    },
    {
      "epoch": 49.33471933471934,
      "grad_norm": 0.00012683146633207798,
      "learning_rate": 0.10669410497167851,
      "loss": 0.298,
      "num_input_tokens_seen": 18098536,
      "step": 23730
    },
    {
      "epoch": 49.34511434511435,
      "grad_norm": 0.0004436123126652092,
      "learning_rate": 0.10663771174801102,
      "loss": 0.2703,
      "num_input_tokens_seen": 18102312,
      "step": 23735
    },
    {
      "epoch": 49.355509355509355,
      "grad_norm": 0.0002383785613346845,
      "learning_rate": 0.10658132521135329,
      "loss": 0.279,
      "num_input_tokens_seen": 18106056,
      "step": 23740
    },
    {
      "epoch": 49.36590436590436,
      "grad_norm": 0.0001367150543956086,
      "learning_rate": 0.10652494537040084,
      "loss": 0.2703,
      "num_input_tokens_seen": 18109928,
      "step": 23745
    },
    {
      "epoch": 49.37629937629938,
      "grad_norm": 0.00024961400777101517,
      "learning_rate": 0.1064685722338482,
      "loss": 0.263,
      "num_input_tokens_seen": 18113896,
      "step": 23750
    },
    {
      "epoch": 49.38669438669439,
      "grad_norm": 0.00012926464842166752,
      "learning_rate": 0.10641220581038871,
      "loss": 0.219,
      "num_input_tokens_seen": 18117768,
      "step": 23755
    },
    {
      "epoch": 49.397089397089395,
      "grad_norm": 0.00019978737691417336,
      "learning_rate": 0.10635584610871483,
      "loss": 0.2577,
      "num_input_tokens_seen": 18121512,
      "step": 23760
    },
    {
      "epoch": 49.40748440748441,
      "grad_norm": 0.0003638410125859082,
      "learning_rate": 0.10629949313751803,
      "loss": 0.2554,
      "num_input_tokens_seen": 18125288,
      "step": 23765
    },
    {
      "epoch": 49.41787941787942,
      "grad_norm": 0.00018110027303919196,
      "learning_rate": 0.10624314690548849,
      "loss": 0.2733,
      "num_input_tokens_seen": 18129064,
      "step": 23770
    },
    {
      "epoch": 49.42827442827443,
      "grad_norm": 9.28209483390674e-05,
      "learning_rate": 0.1061868074213156,
      "loss": 0.263,
      "num_input_tokens_seen": 18133064,
      "step": 23775
    },
    {
      "epoch": 49.438669438669436,
      "grad_norm": 2.908603892137762e-05,
      "learning_rate": 0.10613047469368765,
      "loss": 0.2746,
      "num_input_tokens_seen": 18136712,
      "step": 23780
    },
    {
      "epoch": 49.44906444906445,
      "grad_norm": 8.971554780146107e-05,
      "learning_rate": 0.10607414873129171,
      "loss": 0.2628,
      "num_input_tokens_seen": 18140552,
      "step": 23785
    },
    {
      "epoch": 49.45945945945946,
      "grad_norm": 0.0001949443540070206,
      "learning_rate": 0.10601782954281413,
      "loss": 0.2558,
      "num_input_tokens_seen": 18144328,
      "step": 23790
    },
    {
      "epoch": 49.46985446985447,
      "grad_norm": 0.0002551834040787071,
      "learning_rate": 0.1059615171369399,
      "loss": 0.2715,
      "num_input_tokens_seen": 18148200,
      "step": 23795
    },
    {
      "epoch": 49.48024948024948,
      "grad_norm": 0.00017766660312190652,
      "learning_rate": 0.10590521152235312,
      "loss": 0.2768,
      "num_input_tokens_seen": 18152072,
      "step": 23800
    },
    {
      "epoch": 49.48024948024948,
      "eval_loss": 0.25095635652542114,
      "eval_runtime": 13.3843,
      "eval_samples_per_second": 63.956,
      "eval_steps_per_second": 15.989,
      "num_input_tokens_seen": 18152072,
      "step": 23800
    },
    {
      "epoch": 49.49064449064449,
      "grad_norm": 0.0003702202229760587,
      "learning_rate": 0.1058489127077369,
      "loss": 0.2769,
      "num_input_tokens_seen": 18155880,
      "step": 23805
    },
    {
      "epoch": 49.5010395010395,
      "grad_norm": 0.00026751047698780894,
      "learning_rate": 0.1057926207017732,
      "loss": 0.2739,
      "num_input_tokens_seen": 18159848,
      "step": 23810
    },
    {
      "epoch": 49.51143451143451,
      "grad_norm": 0.00024904997553676367,
      "learning_rate": 0.10573633551314285,
      "loss": 0.2642,
      "num_input_tokens_seen": 18163752,
      "step": 23815
    },
    {
      "epoch": 49.521829521829524,
      "grad_norm": 0.00016037505702115595,
      "learning_rate": 0.1056800571505259,
      "loss": 0.2463,
      "num_input_tokens_seen": 18167528,
      "step": 23820
    },
    {
      "epoch": 49.53222453222453,
      "grad_norm": 0.00023132209025789052,
      "learning_rate": 0.10562378562260105,
      "loss": 0.2672,
      "num_input_tokens_seen": 18171400,
      "step": 23825
    },
    {
      "epoch": 49.54261954261954,
      "grad_norm": 0.0004048071859870106,
      "learning_rate": 0.10556752093804615,
      "loss": 0.275,
      "num_input_tokens_seen": 18175304,
      "step": 23830
    },
    {
      "epoch": 49.553014553014556,
      "grad_norm": 0.00011635488772299141,
      "learning_rate": 0.10551126310553786,
      "loss": 0.2309,
      "num_input_tokens_seen": 18179176,
      "step": 23835
    },
    {
      "epoch": 49.563409563409564,
      "grad_norm": 3.719219967024401e-05,
      "learning_rate": 0.10545501213375187,
      "loss": 0.2566,
      "num_input_tokens_seen": 18182984,
      "step": 23840
    },
    {
      "epoch": 49.57380457380457,
      "grad_norm": 0.0003588319232221693,
      "learning_rate": 0.10539876803136287,
      "loss": 0.2588,
      "num_input_tokens_seen": 18186536,
      "step": 23845
    },
    {
      "epoch": 49.58419958419958,
      "grad_norm": 0.00022439012536779046,
      "learning_rate": 0.10534253080704428,
      "loss": 0.2709,
      "num_input_tokens_seen": 18190152,
      "step": 23850
    },
    {
      "epoch": 49.5945945945946,
      "grad_norm": 0.0003745346039067954,
      "learning_rate": 0.10528630046946862,
      "loss": 0.2608,
      "num_input_tokens_seen": 18193864,
      "step": 23855
    },
    {
      "epoch": 49.604989604989605,
      "grad_norm": 0.0004120338417124003,
      "learning_rate": 0.1052300770273074,
      "loss": 0.2655,
      "num_input_tokens_seen": 18197672,
      "step": 23860
    },
    {
      "epoch": 49.61538461538461,
      "grad_norm": 0.00016493379371240735,
      "learning_rate": 0.10517386048923086,
      "loss": 0.2806,
      "num_input_tokens_seen": 18201480,
      "step": 23865
    },
    {
      "epoch": 49.62577962577963,
      "grad_norm": 0.00013868477253708988,
      "learning_rate": 0.10511765086390841,
      "loss": 0.2803,
      "num_input_tokens_seen": 18205352,
      "step": 23870
    },
    {
      "epoch": 49.63617463617464,
      "grad_norm": 0.00024250400019809604,
      "learning_rate": 0.10506144816000816,
      "loss": 0.2726,
      "num_input_tokens_seen": 18208968,
      "step": 23875
    },
    {
      "epoch": 49.646569646569645,
      "grad_norm": 0.00012117517326259986,
      "learning_rate": 0.10500525238619736,
      "loss": 0.2755,
      "num_input_tokens_seen": 18212840,
      "step": 23880
    },
    {
      "epoch": 49.656964656964654,
      "grad_norm": 0.0004627651069313288,
      "learning_rate": 0.10494906355114209,
      "loss": 0.2742,
      "num_input_tokens_seen": 18216744,
      "step": 23885
    },
    {
      "epoch": 49.66735966735967,
      "grad_norm": 0.0004094117321074009,
      "learning_rate": 0.10489288166350737,
      "loss": 0.2823,
      "num_input_tokens_seen": 18220680,
      "step": 23890
    },
    {
      "epoch": 49.67775467775468,
      "grad_norm": 0.0003043843898922205,
      "learning_rate": 0.10483670673195711,
      "loss": 0.267,
      "num_input_tokens_seen": 18224360,
      "step": 23895
    },
    {
      "epoch": 49.688149688149686,
      "grad_norm": 0.0003859902499243617,
      "learning_rate": 0.10478053876515431,
      "loss": 0.2857,
      "num_input_tokens_seen": 18228072,
      "step": 23900
    },
    {
      "epoch": 49.6985446985447,
      "grad_norm": 0.0004133705806452781,
      "learning_rate": 0.10472437777176061,
      "loss": 0.2688,
      "num_input_tokens_seen": 18231784,
      "step": 23905
    },
    {
      "epoch": 49.70893970893971,
      "grad_norm": 0.0002666472573764622,
      "learning_rate": 0.1046682237604369,
      "loss": 0.2527,
      "num_input_tokens_seen": 18235464,
      "step": 23910
    },
    {
      "epoch": 49.71933471933472,
      "grad_norm": 0.00042969704372808337,
      "learning_rate": 0.1046120767398427,
      "loss": 0.2693,
      "num_input_tokens_seen": 18239336,
      "step": 23915
    },
    {
      "epoch": 49.729729729729726,
      "grad_norm": 0.0004417806339915842,
      "learning_rate": 0.10455593671863667,
      "loss": 0.2628,
      "num_input_tokens_seen": 18243016,
      "step": 23920
    },
    {
      "epoch": 49.74012474012474,
      "grad_norm": 0.0004789374943356961,
      "learning_rate": 0.1044998037054763,
      "loss": 0.2797,
      "num_input_tokens_seen": 18246792,
      "step": 23925
    },
    {
      "epoch": 49.75051975051975,
      "grad_norm": 5.336150206858292e-05,
      "learning_rate": 0.10444367770901794,
      "loss": 0.2755,
      "num_input_tokens_seen": 18250600,
      "step": 23930
    },
    {
      "epoch": 49.76091476091476,
      "grad_norm": 7.816852303221822e-05,
      "learning_rate": 0.10438755873791698,
      "loss": 0.2893,
      "num_input_tokens_seen": 18254440,
      "step": 23935
    },
    {
      "epoch": 49.771309771309774,
      "grad_norm": 0.00030390999745577574,
      "learning_rate": 0.10433144680082775,
      "loss": 0.2582,
      "num_input_tokens_seen": 18258184,
      "step": 23940
    },
    {
      "epoch": 49.78170478170478,
      "grad_norm": 7.809425733285025e-05,
      "learning_rate": 0.10427534190640322,
      "loss": 0.2695,
      "num_input_tokens_seen": 18262024,
      "step": 23945
    },
    {
      "epoch": 49.79209979209979,
      "grad_norm": 0.00015017240366432816,
      "learning_rate": 0.10421924406329568,
      "loss": 0.2543,
      "num_input_tokens_seen": 18265960,
      "step": 23950
    },
    {
      "epoch": 49.802494802494806,
      "grad_norm": 0.0005150067736394703,
      "learning_rate": 0.10416315328015598,
      "loss": 0.2356,
      "num_input_tokens_seen": 18269800,
      "step": 23955
    },
    {
      "epoch": 49.812889812889814,
      "grad_norm": 0.0002945914748124778,
      "learning_rate": 0.10410706956563402,
      "loss": 0.2788,
      "num_input_tokens_seen": 18273576,
      "step": 23960
    },
    {
      "epoch": 49.82328482328482,
      "grad_norm": 0.0002123496524291113,
      "learning_rate": 0.10405099292837874,
      "loss": 0.275,
      "num_input_tokens_seen": 18277448,
      "step": 23965
    },
    {
      "epoch": 49.83367983367983,
      "grad_norm": 0.00026722936308942735,
      "learning_rate": 0.10399492337703771,
      "loss": 0.2791,
      "num_input_tokens_seen": 18281160,
      "step": 23970
    },
    {
      "epoch": 49.84407484407485,
      "grad_norm": 0.00010798217408591881,
      "learning_rate": 0.10393886092025764,
      "loss": 0.2626,
      "num_input_tokens_seen": 18285064,
      "step": 23975
    },
    {
      "epoch": 49.854469854469855,
      "grad_norm": 0.0001463006337871775,
      "learning_rate": 0.10388280556668412,
      "loss": 0.2686,
      "num_input_tokens_seen": 18288904,
      "step": 23980
    },
    {
      "epoch": 49.86486486486486,
      "grad_norm": 0.00013373060210142285,
      "learning_rate": 0.10382675732496145,
      "loss": 0.2536,
      "num_input_tokens_seen": 18292712,
      "step": 23985
    },
    {
      "epoch": 49.87525987525988,
      "grad_norm": 0.00023828244593460113,
      "learning_rate": 0.10377071620373311,
      "loss": 0.2169,
      "num_input_tokens_seen": 18296520,
      "step": 23990
    },
    {
      "epoch": 49.88565488565489,
      "grad_norm": 0.0002944665611721575,
      "learning_rate": 0.10371468221164128,
      "loss": 0.2938,
      "num_input_tokens_seen": 18300328,
      "step": 23995
    },
    {
      "epoch": 49.896049896049895,
      "grad_norm": 0.00022032791457604617,
      "learning_rate": 0.10365865535732706,
      "loss": 0.2944,
      "num_input_tokens_seen": 18304072,
      "step": 24000
    },
    {
      "epoch": 49.896049896049895,
      "eval_loss": 0.24930372834205627,
      "eval_runtime": 13.4028,
      "eval_samples_per_second": 63.867,
      "eval_steps_per_second": 15.967,
      "num_input_tokens_seen": 18304072,
      "step": 24000
    },
    {
      "epoch": 49.906444906444904,
      "grad_norm": 0.00022074466687627137,
      "learning_rate": 0.10360263564943062,
      "loss": 0.2732,
      "num_input_tokens_seen": 18307816,
      "step": 24005
    },
    {
      "epoch": 49.91683991683992,
      "grad_norm": 8.974806405603886e-05,
      "learning_rate": 0.10354662309659075,
      "loss": 0.2649,
      "num_input_tokens_seen": 18311624,
      "step": 24010
    },
    {
      "epoch": 49.92723492723493,
      "grad_norm": 0.0003212092851754278,
      "learning_rate": 0.10349061770744537,
      "loss": 0.2644,
      "num_input_tokens_seen": 18315496,
      "step": 24015
    },
    {
      "epoch": 49.937629937629936,
      "grad_norm": 0.0001214043004438281,
      "learning_rate": 0.10343461949063128,
      "loss": 0.2611,
      "num_input_tokens_seen": 18319208,
      "step": 24020
    },
    {
      "epoch": 49.94802494802495,
      "grad_norm": 0.0001060393187799491,
      "learning_rate": 0.103378628454784,
      "loss": 0.2634,
      "num_input_tokens_seen": 18322952,
      "step": 24025
    },
    {
      "epoch": 49.95841995841996,
      "grad_norm": 0.00013607065193355083,
      "learning_rate": 0.10332264460853811,
      "loss": 0.2866,
      "num_input_tokens_seen": 18326600,
      "step": 24030
    },
    {
      "epoch": 49.96881496881497,
      "grad_norm": 8.524490112904459e-05,
      "learning_rate": 0.10326666796052701,
      "loss": 0.2383,
      "num_input_tokens_seen": 18330440,
      "step": 24035
    },
    {
      "epoch": 49.979209979209976,
      "grad_norm": 0.0001402928028255701,
      "learning_rate": 0.10321069851938296,
      "loss": 0.2609,
      "num_input_tokens_seen": 18334184,
      "step": 24040
    },
    {
      "epoch": 49.98960498960499,
      "grad_norm": 0.00022541174257639796,
      "learning_rate": 0.10315473629373724,
      "loss": 0.2506,
      "num_input_tokens_seen": 18338120,
      "step": 24045
    },
    {
      "epoch": 50.0,
      "grad_norm": 0.00013395698624663055,
      "learning_rate": 0.10309878129221982,
      "loss": 0.2529,
      "num_input_tokens_seen": 18341840,
      "step": 24050
    },
    {
      "epoch": 50.01039501039501,
      "grad_norm": 4.0829749195836484e-05,
      "learning_rate": 0.10304283352345973,
      "loss": 0.2513,
      "num_input_tokens_seen": 18345680,
      "step": 24055
    },
    {
      "epoch": 50.020790020790024,
      "grad_norm": 0.0005817682249471545,
      "learning_rate": 0.10298689299608486,
      "loss": 0.2938,
      "num_input_tokens_seen": 18349520,
      "step": 24060
    },
    {
      "epoch": 50.03118503118503,
      "grad_norm": 0.0003067734360229224,
      "learning_rate": 0.10293095971872188,
      "loss": 0.2642,
      "num_input_tokens_seen": 18353264,
      "step": 24065
    },
    {
      "epoch": 50.04158004158004,
      "grad_norm": 0.0002419666707282886,
      "learning_rate": 0.10287503369999645,
      "loss": 0.2704,
      "num_input_tokens_seen": 18357040,
      "step": 24070
    },
    {
      "epoch": 50.05197505197505,
      "grad_norm": 0.0002429165760986507,
      "learning_rate": 0.10281911494853295,
      "loss": 0.2767,
      "num_input_tokens_seen": 18360976,
      "step": 24075
    },
    {
      "epoch": 50.062370062370064,
      "grad_norm": 0.00031340416171588004,
      "learning_rate": 0.10276320347295485,
      "loss": 0.268,
      "num_input_tokens_seen": 18364816,
      "step": 24080
    },
    {
      "epoch": 50.07276507276507,
      "grad_norm": 0.0001533517352072522,
      "learning_rate": 0.10270729928188446,
      "loss": 0.2824,
      "num_input_tokens_seen": 18368624,
      "step": 24085
    },
    {
      "epoch": 50.08316008316008,
      "grad_norm": 0.0001746221532812342,
      "learning_rate": 0.10265140238394276,
      "loss": 0.2634,
      "num_input_tokens_seen": 18372336,
      "step": 24090
    },
    {
      "epoch": 50.093555093555096,
      "grad_norm": 0.00046772032510489225,
      "learning_rate": 0.10259551278774988,
      "loss": 0.283,
      "num_input_tokens_seen": 18376016,
      "step": 24095
    },
    {
      "epoch": 50.103950103950105,
      "grad_norm": 0.00032474062754772604,
      "learning_rate": 0.10253963050192462,
      "loss": 0.2502,
      "num_input_tokens_seen": 18379952,
      "step": 24100
    },
    {
      "epoch": 50.11434511434511,
      "grad_norm": 0.0002057932724710554,
      "learning_rate": 0.10248375553508478,
      "loss": 0.2604,
      "num_input_tokens_seen": 18383664,
      "step": 24105
    },
    {
      "epoch": 50.12474012474012,
      "grad_norm": 0.00039557606214657426,
      "learning_rate": 0.102427887895847,
      "loss": 0.2682,
      "num_input_tokens_seen": 18387536,
      "step": 24110
    },
    {
      "epoch": 50.13513513513514,
      "grad_norm": 0.0002889862225856632,
      "learning_rate": 0.10237202759282668,
      "loss": 0.2564,
      "num_input_tokens_seen": 18391472,
      "step": 24115
    },
    {
      "epoch": 50.145530145530145,
      "grad_norm": 8.735974552109838e-05,
      "learning_rate": 0.10231617463463821,
      "loss": 0.2796,
      "num_input_tokens_seen": 18395152,
      "step": 24120
    },
    {
      "epoch": 50.15592515592515,
      "grad_norm": 0.0003275749913882464,
      "learning_rate": 0.10226032902989492,
      "loss": 0.2474,
      "num_input_tokens_seen": 18398896,
      "step": 24125
    },
    {
      "epoch": 50.16632016632017,
      "grad_norm": 0.0003949719830416143,
      "learning_rate": 0.10220449078720877,
      "loss": 0.2806,
      "num_input_tokens_seen": 18402640,
      "step": 24130
    },
    {
      "epoch": 50.17671517671518,
      "grad_norm": 0.0005629869410768151,
      "learning_rate": 0.1021486599151908,
      "loss": 0.2764,
      "num_input_tokens_seen": 18406448,
      "step": 24135
    },
    {
      "epoch": 50.187110187110186,
      "grad_norm": 5.7437307987129316e-05,
      "learning_rate": 0.10209283642245084,
      "loss": 0.272,
      "num_input_tokens_seen": 18410256,
      "step": 24140
    },
    {
      "epoch": 50.197505197505194,
      "grad_norm": 0.00025217654183506966,
      "learning_rate": 0.10203702031759748,
      "loss": 0.2737,
      "num_input_tokens_seen": 18414096,
      "step": 24145
    },
    {
      "epoch": 50.20790020790021,
      "grad_norm": 0.00018602272029966116,
      "learning_rate": 0.1019812116092384,
      "loss": 0.2707,
      "num_input_tokens_seen": 18417776,
      "step": 24150
    },
    {
      "epoch": 50.21829521829522,
      "grad_norm": 0.0003643235540948808,
      "learning_rate": 0.10192541030597986,
      "loss": 0.2472,
      "num_input_tokens_seen": 18421456,
      "step": 24155
    },
    {
      "epoch": 50.228690228690226,
      "grad_norm": 9.985128417611122e-05,
      "learning_rate": 0.1018696164164272,
      "loss": 0.2588,
      "num_input_tokens_seen": 18425296,
      "step": 24160
    },
    {
      "epoch": 50.23908523908524,
      "grad_norm": 0.0001157584338216111,
      "learning_rate": 0.10181382994918459,
      "loss": 0.2687,
      "num_input_tokens_seen": 18429200,
      "step": 24165
    },
    {
      "epoch": 50.24948024948025,
      "grad_norm": 0.0002264860668219626,
      "learning_rate": 0.10175805091285492,
      "loss": 0.2634,
      "num_input_tokens_seen": 18433136,
      "step": 24170
    },
    {
      "epoch": 50.25987525987526,
      "grad_norm": 0.0002869514573831111,
      "learning_rate": 0.10170227931603999,
      "loss": 0.2133,
      "num_input_tokens_seen": 18436944,
      "step": 24175
    },
    {
      "epoch": 50.270270270270274,
      "grad_norm": 0.00016675608640071005,
      "learning_rate": 0.10164651516734062,
      "loss": 0.2774,
      "num_input_tokens_seen": 18440880,
      "step": 24180
    },
    {
      "epoch": 50.28066528066528,
      "grad_norm": 0.00010287578334100544,
      "learning_rate": 0.1015907584753562,
      "loss": 0.261,
      "num_input_tokens_seen": 18444464,
      "step": 24185
    },
    {
      "epoch": 50.29106029106029,
      "grad_norm": 0.0006566105876117945,
      "learning_rate": 0.10153500924868523,
      "loss": 0.2259,
      "num_input_tokens_seen": 18448272,
      "step": 24190
    },
    {
      "epoch": 50.3014553014553,
      "grad_norm": 0.00014778364857193083,
      "learning_rate": 0.10147926749592483,
      "loss": 0.2436,
      "num_input_tokens_seen": 18451984,
      "step": 24195
    },
    {
      "epoch": 50.311850311850314,
      "grad_norm": 0.0002861306711565703,
      "learning_rate": 0.10142353322567112,
      "loss": 0.2994,
      "num_input_tokens_seen": 18455696,
      "step": 24200
    },
    {
      "epoch": 50.311850311850314,
      "eval_loss": 0.24659645557403564,
      "eval_runtime": 13.4121,
      "eval_samples_per_second": 63.823,
      "eval_steps_per_second": 15.956,
      "num_input_tokens_seen": 18455696,
      "step": 24200
    },
    {
      "epoch": 50.32224532224532,
      "grad_norm": 6.527492223540321e-05,
      "learning_rate": 0.1013678064465191,
      "loss": 0.2523,
      "num_input_tokens_seen": 18459568,
      "step": 24205
    },
    {
      "epoch": 50.33264033264033,
      "grad_norm": 0.0002931687922682613,
      "learning_rate": 0.10131208716706244,
      "loss": 0.2834,
      "num_input_tokens_seen": 18463504,
      "step": 24210
    },
    {
      "epoch": 50.343035343035346,
      "grad_norm": 0.0001677976833889261,
      "learning_rate": 0.10125637539589379,
      "loss": 0.2649,
      "num_input_tokens_seen": 18467408,
      "step": 24215
    },
    {
      "epoch": 50.353430353430355,
      "grad_norm": 0.00028112702420912683,
      "learning_rate": 0.10120067114160464,
      "loss": 0.2583,
      "num_input_tokens_seen": 18471120,
      "step": 24220
    },
    {
      "epoch": 50.36382536382536,
      "grad_norm": 0.00014877118519507349,
      "learning_rate": 0.10114497441278517,
      "loss": 0.2674,
      "num_input_tokens_seen": 18475024,
      "step": 24225
    },
    {
      "epoch": 50.37422037422037,
      "grad_norm": 0.00014939019456505775,
      "learning_rate": 0.10108928521802468,
      "loss": 0.2738,
      "num_input_tokens_seen": 18478768,
      "step": 24230
    },
    {
      "epoch": 50.38461538461539,
      "grad_norm": 9.070548549061641e-05,
      "learning_rate": 0.101033603565911,
      "loss": 0.2655,
      "num_input_tokens_seen": 18482672,
      "step": 24235
    },
    {
      "epoch": 50.395010395010395,
      "grad_norm": 0.0005833710310980678,
      "learning_rate": 0.10097792946503102,
      "loss": 0.2737,
      "num_input_tokens_seen": 18486480,
      "step": 24240
    },
    {
      "epoch": 50.4054054054054,
      "grad_norm": 0.00020180588762741536,
      "learning_rate": 0.10092226292397039,
      "loss": 0.2677,
      "num_input_tokens_seen": 18490384,
      "step": 24245
    },
    {
      "epoch": 50.41580041580042,
      "grad_norm": 0.00026685497141443193,
      "learning_rate": 0.10086660395131354,
      "loss": 0.257,
      "num_input_tokens_seen": 18494224,
      "step": 24250
    },
    {
      "epoch": 50.42619542619543,
      "grad_norm": 0.00023248510842677206,
      "learning_rate": 0.10081095255564385,
      "loss": 0.2524,
      "num_input_tokens_seen": 18498000,
      "step": 24255
    },
    {
      "epoch": 50.436590436590436,
      "grad_norm": 0.00020622876763809472,
      "learning_rate": 0.10075530874554335,
      "loss": 0.2937,
      "num_input_tokens_seen": 18501936,
      "step": 24260
    },
    {
      "epoch": 50.446985446985444,
      "grad_norm": 0.0003202626539859921,
      "learning_rate": 0.10069967252959311,
      "loss": 0.2566,
      "num_input_tokens_seen": 18505648,
      "step": 24265
    },
    {
      "epoch": 50.45738045738046,
      "grad_norm": 0.00010146533895749599,
      "learning_rate": 0.10064404391637297,
      "loss": 0.2592,
      "num_input_tokens_seen": 18509360,
      "step": 24270
    },
    {
      "epoch": 50.46777546777547,
      "grad_norm": 0.00012425774184521288,
      "learning_rate": 0.10058842291446145,
      "loss": 0.2716,
      "num_input_tokens_seen": 18513136,
      "step": 24275
    },
    {
      "epoch": 50.478170478170476,
      "grad_norm": 9.50094181462191e-05,
      "learning_rate": 0.10053280953243608,
      "loss": 0.2339,
      "num_input_tokens_seen": 18516944,
      "step": 24280
    },
    {
      "epoch": 50.48856548856549,
      "grad_norm": 0.00015902260201983154,
      "learning_rate": 0.10047720377887315,
      "loss": 0.2449,
      "num_input_tokens_seen": 18520816,
      "step": 24285
    },
    {
      "epoch": 50.4989604989605,
      "grad_norm": 7.598899537697434e-05,
      "learning_rate": 0.10042160566234767,
      "loss": 0.2353,
      "num_input_tokens_seen": 18524528,
      "step": 24290
    },
    {
      "epoch": 50.50935550935551,
      "grad_norm": 0.0002526042517274618,
      "learning_rate": 0.10036601519143372,
      "loss": 0.2917,
      "num_input_tokens_seen": 18528432,
      "step": 24295
    },
    {
      "epoch": 50.51975051975052,
      "grad_norm": 8.480740507366136e-05,
      "learning_rate": 0.1003104323747039,
      "loss": 0.2817,
      "num_input_tokens_seen": 18532080,
      "step": 24300
    },
    {
      "epoch": 50.53014553014553,
      "grad_norm": 0.00013210528413765132,
      "learning_rate": 0.10025485722072984,
      "loss": 0.2593,
      "num_input_tokens_seen": 18535984,
      "step": 24305
    },
    {
      "epoch": 50.54054054054054,
      "grad_norm": 4.810529571841471e-05,
      "learning_rate": 0.10019928973808201,
      "loss": 0.2702,
      "num_input_tokens_seen": 18539856,
      "step": 24310
    },
    {
      "epoch": 50.55093555093555,
      "grad_norm": 0.00028146570548415184,
      "learning_rate": 0.10014372993532945,
      "loss": 0.2543,
      "num_input_tokens_seen": 18543696,
      "step": 24315
    },
    {
      "epoch": 50.561330561330564,
      "grad_norm": 0.00033793141483329237,
      "learning_rate": 0.1000881778210403,
      "loss": 0.2446,
      "num_input_tokens_seen": 18547664,
      "step": 24320
    },
    {
      "epoch": 50.57172557172557,
      "grad_norm": 0.00026973572676070035,
      "learning_rate": 0.10003263340378142,
      "loss": 0.2638,
      "num_input_tokens_seen": 18551568,
      "step": 24325
    },
    {
      "epoch": 50.58212058212058,
      "grad_norm": 0.00018194993026554585,
      "learning_rate": 0.09997709669211834,
      "loss": 0.2591,
      "num_input_tokens_seen": 18555536,
      "step": 24330
    },
    {
      "epoch": 50.59251559251559,
      "grad_norm": 0.00017239435692317784,
      "learning_rate": 0.0999215676946156,
      "loss": 0.269,
      "num_input_tokens_seen": 18559472,
      "step": 24335
    },
    {
      "epoch": 50.602910602910605,
      "grad_norm": 0.00017497414955869317,
      "learning_rate": 0.0998660464198364,
      "loss": 0.3024,
      "num_input_tokens_seen": 18563248,
      "step": 24340
    },
    {
      "epoch": 50.61330561330561,
      "grad_norm": 0.0004112626484129578,
      "learning_rate": 0.09981053287634288,
      "loss": 0.2512,
      "num_input_tokens_seen": 18567088,
      "step": 24345
    },
    {
      "epoch": 50.62370062370062,
      "grad_norm": 9.315647912444547e-05,
      "learning_rate": 0.09975502707269596,
      "loss": 0.2677,
      "num_input_tokens_seen": 18571088,
      "step": 24350
    },
    {
      "epoch": 50.63409563409564,
      "grad_norm": 9.203373338095844e-05,
      "learning_rate": 0.09969952901745524,
      "loss": 0.2588,
      "num_input_tokens_seen": 18574704,
      "step": 24355
    },
    {
      "epoch": 50.644490644490645,
      "grad_norm": 6.929399387445301e-05,
      "learning_rate": 0.09964403871917925,
      "loss": 0.2486,
      "num_input_tokens_seen": 18578480,
      "step": 24360
    },
    {
      "epoch": 50.65488565488565,
      "grad_norm": 0.0001429767144145444,
      "learning_rate": 0.09958855618642536,
      "loss": 0.2645,
      "num_input_tokens_seen": 18582288,
      "step": 24365
    },
    {
      "epoch": 50.66528066528066,
      "grad_norm": 0.00015283816901501268,
      "learning_rate": 0.09953308142774955,
      "loss": 0.2614,
      "num_input_tokens_seen": 18586000,
      "step": 24370
    },
    {
      "epoch": 50.67567567567568,
      "grad_norm": 0.00022856822761241347,
      "learning_rate": 0.09947761445170686,
      "loss": 0.258,
      "num_input_tokens_seen": 18589904,
      "step": 24375
    },
    {
      "epoch": 50.686070686070686,
      "grad_norm": 0.00017272579134441912,
      "learning_rate": 0.09942215526685086,
      "loss": 0.2782,
      "num_input_tokens_seen": 18593712,
      "step": 24380
    },
    {
      "epoch": 50.696465696465694,
      "grad_norm": 0.0003653337771538645,
      "learning_rate": 0.09936670388173414,
      "loss": 0.2632,
      "num_input_tokens_seen": 18597456,
      "step": 24385
    },
    {
      "epoch": 50.70686070686071,
      "grad_norm": 0.0001519051002105698,
      "learning_rate": 0.09931126030490799,
      "loss": 0.2514,
      "num_input_tokens_seen": 18601360,
      "step": 24390
    },
    {
      "epoch": 50.71725571725572,
      "grad_norm": 0.00014558061957359314,
      "learning_rate": 0.0992558245449225,
      "loss": 0.2444,
      "num_input_tokens_seen": 18605136,
      "step": 24395
    },
    {
      "epoch": 50.727650727650726,
      "grad_norm": 0.00031312459032051265,
      "learning_rate": 0.09920039661032651,
      "loss": 0.2625,
      "num_input_tokens_seen": 18608976,
      "step": 24400
    },
    {
      "epoch": 50.727650727650726,
      "eval_loss": 0.24898014962673187,
      "eval_runtime": 13.3853,
      "eval_samples_per_second": 63.951,
      "eval_steps_per_second": 15.988,
      "num_input_tokens_seen": 18608976,
      "step": 24400
    },
    {
      "epoch": 50.73804573804574,
      "grad_norm": 0.00031856808345764875,
      "learning_rate": 0.09914497650966782,
      "loss": 0.2275,
      "num_input_tokens_seen": 18612848,
      "step": 24405
    },
    {
      "epoch": 50.74844074844075,
      "grad_norm": 0.0007534954929724336,
      "learning_rate": 0.09908956425149276,
      "loss": 0.2838,
      "num_input_tokens_seen": 18616688,
      "step": 24410
    },
    {
      "epoch": 50.75883575883576,
      "grad_norm": 0.00045412409235723317,
      "learning_rate": 0.09903415984434677,
      "loss": 0.2603,
      "num_input_tokens_seen": 18620496,
      "step": 24415
    },
    {
      "epoch": 50.76923076923077,
      "grad_norm": 0.00018290500156581402,
      "learning_rate": 0.09897876329677373,
      "loss": 0.2755,
      "num_input_tokens_seen": 18624400,
      "step": 24420
    },
    {
      "epoch": 50.77962577962578,
      "grad_norm": 0.0003129853284917772,
      "learning_rate": 0.09892337461731658,
      "loss": 0.2644,
      "num_input_tokens_seen": 18628336,
      "step": 24425
    },
    {
      "epoch": 50.79002079002079,
      "grad_norm": 0.0002791692386381328,
      "learning_rate": 0.09886799381451693,
      "loss": 0.2916,
      "num_input_tokens_seen": 18632208,
      "step": 24430
    },
    {
      "epoch": 50.8004158004158,
      "grad_norm": 0.0002489391481503844,
      "learning_rate": 0.09881262089691521,
      "loss": 0.2733,
      "num_input_tokens_seen": 18635984,
      "step": 24435
    },
    {
      "epoch": 50.810810810810814,
      "grad_norm": 0.0005622346652671695,
      "learning_rate": 0.09875725587305059,
      "loss": 0.2782,
      "num_input_tokens_seen": 18639952,
      "step": 24440
    },
    {
      "epoch": 50.82120582120582,
      "grad_norm": 4.364913183962926e-05,
      "learning_rate": 0.09870189875146111,
      "loss": 0.2752,
      "num_input_tokens_seen": 18643632,
      "step": 24445
    },
    {
      "epoch": 50.83160083160083,
      "grad_norm": 0.00022214869386516511,
      "learning_rate": 0.09864654954068346,
      "loss": 0.2765,
      "num_input_tokens_seen": 18647408,
      "step": 24450
    },
    {
      "epoch": 50.84199584199584,
      "grad_norm": 0.00011121843272121623,
      "learning_rate": 0.09859120824925326,
      "loss": 0.2581,
      "num_input_tokens_seen": 18651216,
      "step": 24455
    },
    {
      "epoch": 50.852390852390855,
      "grad_norm": 0.00023151130881160498,
      "learning_rate": 0.09853587488570474,
      "loss": 0.2602,
      "num_input_tokens_seen": 18654960,
      "step": 24460
    },
    {
      "epoch": 50.86278586278586,
      "grad_norm": 0.0005612490349449217,
      "learning_rate": 0.09848054945857107,
      "loss": 0.2782,
      "num_input_tokens_seen": 18658768,
      "step": 24465
    },
    {
      "epoch": 50.87318087318087,
      "grad_norm": 0.0002154849935323,
      "learning_rate": 0.09842523197638416,
      "loss": 0.2645,
      "num_input_tokens_seen": 18662480,
      "step": 24470
    },
    {
      "epoch": 50.88357588357589,
      "grad_norm": 0.0001781236642273143,
      "learning_rate": 0.09836992244767452,
      "loss": 0.2617,
      "num_input_tokens_seen": 18666256,
      "step": 24475
    },
    {
      "epoch": 50.893970893970895,
      "grad_norm": 5.580757715506479e-05,
      "learning_rate": 0.09831462088097168,
      "loss": 0.2845,
      "num_input_tokens_seen": 18669968,
      "step": 24480
    },
    {
      "epoch": 50.9043659043659,
      "grad_norm": 0.0002857065701391548,
      "learning_rate": 0.09825932728480385,
      "loss": 0.2729,
      "num_input_tokens_seen": 18673712,
      "step": 24485
    },
    {
      "epoch": 50.91476091476091,
      "grad_norm": 9.086565842153504e-05,
      "learning_rate": 0.09820404166769794,
      "loss": 0.2726,
      "num_input_tokens_seen": 18677360,
      "step": 24490
    },
    {
      "epoch": 50.92515592515593,
      "grad_norm": 7.799208106007427e-05,
      "learning_rate": 0.09814876403817978,
      "loss": 0.2623,
      "num_input_tokens_seen": 18681264,
      "step": 24495
    },
    {
      "epoch": 50.935550935550935,
      "grad_norm": 0.00014397941413335502,
      "learning_rate": 0.09809349440477376,
      "loss": 0.2507,
      "num_input_tokens_seen": 18685072,
      "step": 24500
    },
    {
      "epoch": 50.945945945945944,
      "grad_norm": 0.00013089526328258216,
      "learning_rate": 0.09803823277600317,
      "loss": 0.2749,
      "num_input_tokens_seen": 18688848,
      "step": 24505
    },
    {
      "epoch": 50.95634095634096,
      "grad_norm": 0.0005686940276063979,
      "learning_rate": 0.09798297916039014,
      "loss": 0.2742,
      "num_input_tokens_seen": 18692624,
      "step": 24510
    },
    {
      "epoch": 50.96673596673597,
      "grad_norm": 0.00027545986813493073,
      "learning_rate": 0.09792773356645534,
      "loss": 0.2579,
      "num_input_tokens_seen": 18696304,
      "step": 24515
    },
    {
      "epoch": 50.977130977130976,
      "grad_norm": 4.643796637537889e-05,
      "learning_rate": 0.09787249600271843,
      "loss": 0.2594,
      "num_input_tokens_seen": 18700080,
      "step": 24520
    },
    {
      "epoch": 50.987525987525984,
      "grad_norm": 0.00015421421267092228,
      "learning_rate": 0.09781726647769776,
      "loss": 0.2594,
      "num_input_tokens_seen": 18703920,
      "step": 24525
    },
    {
      "epoch": 50.997920997921,
      "grad_norm": 9.456792759010568e-05,
      "learning_rate": 0.0977620449999103,
      "loss": 0.2673,
      "num_input_tokens_seen": 18707696,
      "step": 24530
    },
    {
      "epoch": 51.00831600831601,
      "grad_norm": 0.00024646814563311636,
      "learning_rate": 0.09770683157787204,
      "loss": 0.2596,
      "num_input_tokens_seen": 18711488,
      "step": 24535
    },
    {
      "epoch": 51.018711018711016,
      "grad_norm": 0.00024192185082938522,
      "learning_rate": 0.09765162622009745,
      "loss": 0.2469,
      "num_input_tokens_seen": 18715296,
      "step": 24540
    },
    {
      "epoch": 51.02910602910603,
      "grad_norm": 7.681120769120753e-05,
      "learning_rate": 0.09759642893509995,
      "loss": 0.2298,
      "num_input_tokens_seen": 18719008,
      "step": 24545
    },
    {
      "epoch": 51.03950103950104,
      "grad_norm": 0.0002607563219498843,
      "learning_rate": 0.09754123973139169,
      "loss": 0.2435,
      "num_input_tokens_seen": 18722848,
      "step": 24550
    },
    {
      "epoch": 51.04989604989605,
      "grad_norm": 8.269328827736899e-05,
      "learning_rate": 0.09748605861748345,
      "loss": 0.305,
      "num_input_tokens_seen": 18726752,
      "step": 24555
    },
    {
      "epoch": 51.06029106029106,
      "grad_norm": 0.00023772777058184147,
      "learning_rate": 0.0974308856018849,
      "loss": 0.2769,
      "num_input_tokens_seen": 18730560,
      "step": 24560
    },
    {
      "epoch": 51.07068607068607,
      "grad_norm": 0.00020780455088242888,
      "learning_rate": 0.09737572069310449,
      "loss": 0.2697,
      "num_input_tokens_seen": 18734272,
      "step": 24565
    },
    {
      "epoch": 51.08108108108108,
      "grad_norm": 0.0003766508889384568,
      "learning_rate": 0.09732056389964922,
      "loss": 0.2632,
      "num_input_tokens_seen": 18738080,
      "step": 24570
    },
    {
      "epoch": 51.09147609147609,
      "grad_norm": 0.00036014526267535985,
      "learning_rate": 0.097265415230025,
      "loss": 0.2798,
      "num_input_tokens_seen": 18741888,
      "step": 24575
    },
    {
      "epoch": 51.101871101871104,
      "grad_norm": 0.0001421941997250542,
      "learning_rate": 0.09721027469273648,
      "loss": 0.2686,
      "num_input_tokens_seen": 18745664,
      "step": 24580
    },
    {
      "epoch": 51.11226611226611,
      "grad_norm": 0.00047153487685136497,
      "learning_rate": 0.09715514229628695,
      "loss": 0.2858,
      "num_input_tokens_seen": 18749472,
      "step": 24585
    },
    {
      "epoch": 51.12266112266112,
      "grad_norm": 0.00030557592981494963,
      "learning_rate": 0.09710001804917864,
      "loss": 0.2868,
      "num_input_tokens_seen": 18753312,
      "step": 24590
    },
    {
      "epoch": 51.13305613305613,
      "grad_norm": 0.00013361706805881113,
      "learning_rate": 0.09704490195991226,
      "loss": 0.2687,
      "num_input_tokens_seen": 18757152,
      "step": 24595
    },
    {
      "epoch": 51.143451143451145,
      "grad_norm": 0.00010781670425785705,
      "learning_rate": 0.09698979403698753,
      "loss": 0.2745,
      "num_input_tokens_seen": 18760928,
      "step": 24600
    },
    {
      "epoch": 51.143451143451145,
      "eval_loss": 0.26018255949020386,
      "eval_runtime": 13.3973,
      "eval_samples_per_second": 63.893,
      "eval_steps_per_second": 15.973,
      "num_input_tokens_seen": 18760928,
      "step": 24600
    },
    {
      "epoch": 51.15384615384615,
      "grad_norm": 0.0002476339286658913,
      "learning_rate": 0.0969346942889027,
      "loss": 0.2898,
      "num_input_tokens_seen": 18764864,
      "step": 24605
    },
    {
      "epoch": 51.16424116424116,
      "grad_norm": 0.000101707068097312,
      "learning_rate": 0.09687960272415487,
      "loss": 0.2651,
      "num_input_tokens_seen": 18768672,
      "step": 24610
    },
    {
      "epoch": 51.17463617463618,
      "grad_norm": 0.0004897169419564307,
      "learning_rate": 0.0968245193512399,
      "loss": 0.2623,
      "num_input_tokens_seen": 18772480,
      "step": 24615
    },
    {
      "epoch": 51.185031185031185,
      "grad_norm": 0.0001092800303013064,
      "learning_rate": 0.09676944417865221,
      "loss": 0.2812,
      "num_input_tokens_seen": 18776448,
      "step": 24620
    },
    {
      "epoch": 51.195426195426194,
      "grad_norm": 0.0002611938398331404,
      "learning_rate": 0.09671437721488517,
      "loss": 0.273,
      "num_input_tokens_seen": 18780128,
      "step": 24625
    },
    {
      "epoch": 51.20582120582121,
      "grad_norm": 0.00036381211248226464,
      "learning_rate": 0.09665931846843086,
      "loss": 0.261,
      "num_input_tokens_seen": 18783840,
      "step": 24630
    },
    {
      "epoch": 51.21621621621622,
      "grad_norm": 0.0002876933431252837,
      "learning_rate": 0.0966042679477799,
      "loss": 0.259,
      "num_input_tokens_seen": 18787680,
      "step": 24635
    },
    {
      "epoch": 51.226611226611226,
      "grad_norm": 6.897662387927994e-05,
      "learning_rate": 0.09654922566142186,
      "loss": 0.2686,
      "num_input_tokens_seen": 18791552,
      "step": 24640
    },
    {
      "epoch": 51.237006237006234,
      "grad_norm": 0.00016861112089827657,
      "learning_rate": 0.09649419161784498,
      "loss": 0.2784,
      "num_input_tokens_seen": 18795392,
      "step": 24645
    },
    {
      "epoch": 51.24740124740125,
      "grad_norm": 0.00033378094667568803,
      "learning_rate": 0.09643916582553606,
      "loss": 0.2494,
      "num_input_tokens_seen": 18799424,
      "step": 24650
    },
    {
      "epoch": 51.25779625779626,
      "grad_norm": 0.0004326729103922844,
      "learning_rate": 0.09638414829298093,
      "loss": 0.2654,
      "num_input_tokens_seen": 18803200,
      "step": 24655
    },
    {
      "epoch": 51.268191268191266,
      "grad_norm": 0.00012549009989015758,
      "learning_rate": 0.09632913902866386,
      "loss": 0.279,
      "num_input_tokens_seen": 18807040,
      "step": 24660
    },
    {
      "epoch": 51.27858627858628,
      "grad_norm": 0.00017018205835483968,
      "learning_rate": 0.096274138041068,
      "loss": 0.2668,
      "num_input_tokens_seen": 18810944,
      "step": 24665
    },
    {
      "epoch": 51.28898128898129,
      "grad_norm": 0.00022028469538781792,
      "learning_rate": 0.09621914533867527,
      "loss": 0.2204,
      "num_input_tokens_seen": 18814848,
      "step": 24670
    },
    {
      "epoch": 51.2993762993763,
      "grad_norm": 0.0002884832792915404,
      "learning_rate": 0.09616416092996616,
      "loss": 0.2897,
      "num_input_tokens_seen": 18818688,
      "step": 24675
    },
    {
      "epoch": 51.30977130977131,
      "grad_norm": 0.0003661253722384572,
      "learning_rate": 0.09610918482342,
      "loss": 0.2712,
      "num_input_tokens_seen": 18822560,
      "step": 24680
    },
    {
      "epoch": 51.32016632016632,
      "grad_norm": 9.568790119374171e-05,
      "learning_rate": 0.09605421702751478,
      "loss": 0.2677,
      "num_input_tokens_seen": 18826272,
      "step": 24685
    },
    {
      "epoch": 51.33056133056133,
      "grad_norm": 0.00018907897174358368,
      "learning_rate": 0.09599925755072718,
      "loss": 0.264,
      "num_input_tokens_seen": 18830176,
      "step": 24690
    },
    {
      "epoch": 51.34095634095634,
      "grad_norm": 0.0005880626849830151,
      "learning_rate": 0.09594430640153273,
      "loss": 0.2473,
      "num_input_tokens_seen": 18833984,
      "step": 24695
    },
    {
      "epoch": 51.351351351351354,
      "grad_norm": 0.0003065942437388003,
      "learning_rate": 0.09588936358840547,
      "loss": 0.2439,
      "num_input_tokens_seen": 18837664,
      "step": 24700
    },
    {
      "epoch": 51.36174636174636,
      "grad_norm": 0.0001648537872824818,
      "learning_rate": 0.09583442911981836,
      "loss": 0.2807,
      "num_input_tokens_seen": 18841376,
      "step": 24705
    },
    {
      "epoch": 51.37214137214137,
      "grad_norm": 3.6155055568087846e-05,
      "learning_rate": 0.09577950300424302,
      "loss": 0.2468,
      "num_input_tokens_seen": 18845088,
      "step": 24710
    },
    {
      "epoch": 51.38253638253638,
      "grad_norm": 0.0001401686604367569,
      "learning_rate": 0.09572458525014967,
      "loss": 0.2475,
      "num_input_tokens_seen": 18848864,
      "step": 24715
    },
    {
      "epoch": 51.392931392931395,
      "grad_norm": 8.26112664071843e-05,
      "learning_rate": 0.0956696758660073,
      "loss": 0.2979,
      "num_input_tokens_seen": 18852544,
      "step": 24720
    },
    {
      "epoch": 51.4033264033264,
      "grad_norm": 5.821938248118386e-05,
      "learning_rate": 0.09561477486028373,
      "loss": 0.2471,
      "num_input_tokens_seen": 18856512,
      "step": 24725
    },
    {
      "epoch": 51.41372141372141,
      "grad_norm": 0.00045093640801496804,
      "learning_rate": 0.09555988224144528,
      "loss": 0.2762,
      "num_input_tokens_seen": 18860288,
      "step": 24730
    },
    {
      "epoch": 51.42411642411643,
      "grad_norm": 0.0002846041170414537,
      "learning_rate": 0.09550499801795717,
      "loss": 0.2629,
      "num_input_tokens_seen": 18864224,
      "step": 24735
    },
    {
      "epoch": 51.434511434511435,
      "grad_norm": 0.00010604729322949424,
      "learning_rate": 0.09545012219828314,
      "loss": 0.2438,
      "num_input_tokens_seen": 18867904,
      "step": 24740
    },
    {
      "epoch": 51.444906444906444,
      "grad_norm": 0.0002491792547516525,
      "learning_rate": 0.09539525479088577,
      "loss": 0.3055,
      "num_input_tokens_seen": 18871744,
      "step": 24745
    },
    {
      "epoch": 51.45530145530145,
      "grad_norm": 0.00022967495897319168,
      "learning_rate": 0.0953403958042264,
      "loss": 0.2648,
      "num_input_tokens_seen": 18875648,
      "step": 24750
    },
    {
      "epoch": 51.46569646569647,
      "grad_norm": 9.365454752696678e-05,
      "learning_rate": 0.09528554524676484,
      "loss": 0.259,
      "num_input_tokens_seen": 18879584,
      "step": 24755
    },
    {
      "epoch": 51.476091476091476,
      "grad_norm": 7.376411667792127e-05,
      "learning_rate": 0.09523070312695978,
      "loss": 0.2777,
      "num_input_tokens_seen": 18883520,
      "step": 24760
    },
    {
      "epoch": 51.486486486486484,
      "grad_norm": 0.00016310378850903362,
      "learning_rate": 0.09517586945326863,
      "loss": 0.2785,
      "num_input_tokens_seen": 18887360,
      "step": 24765
    },
    {
      "epoch": 51.4968814968815,
      "grad_norm": 0.00025557290064170957,
      "learning_rate": 0.0951210442341473,
      "loss": 0.2668,
      "num_input_tokens_seen": 18891168,
      "step": 24770
    },
    {
      "epoch": 51.50727650727651,
      "grad_norm": 0.00027795921778306365,
      "learning_rate": 0.09506622747805066,
      "loss": 0.2609,
      "num_input_tokens_seen": 18894912,
      "step": 24775
    },
    {
      "epoch": 51.517671517671516,
      "grad_norm": 9.248357673641294e-05,
      "learning_rate": 0.09501141919343203,
      "loss": 0.2531,
      "num_input_tokens_seen": 18898816,
      "step": 24780
    },
    {
      "epoch": 51.528066528066525,
      "grad_norm": 6.760465475963429e-05,
      "learning_rate": 0.09495661938874361,
      "loss": 0.2784,
      "num_input_tokens_seen": 18902624,
      "step": 24785
    },
    {
      "epoch": 51.53846153846154,
      "grad_norm": 0.0001204007858177647,
      "learning_rate": 0.0949018280724362,
      "loss": 0.2584,
      "num_input_tokens_seen": 18906400,
      "step": 24790
    },
    {
      "epoch": 51.54885654885655,
      "grad_norm": 0.00013308181951288134,
      "learning_rate": 0.09484704525295934,
      "loss": 0.2505,
      "num_input_tokens_seen": 18910240,
      "step": 24795
    },
    {
      "epoch": 51.55925155925156,
      "grad_norm": 0.00047808833187446,
      "learning_rate": 0.09479227093876112,
      "loss": 0.2414,
      "num_input_tokens_seen": 18913856,
      "step": 24800
    },
    {
      "epoch": 51.55925155925156,
      "eval_loss": 0.24719470739364624,
      "eval_runtime": 13.3894,
      "eval_samples_per_second": 63.931,
      "eval_steps_per_second": 15.983,
      "num_input_tokens_seen": 18913856,
      "step": 24800
    },
    {
      "epoch": 51.56964656964657,
      "grad_norm": 0.00014398552593775094,
      "learning_rate": 0.0947375051382886,
      "loss": 0.2605,
      "num_input_tokens_seen": 18917600,
      "step": 24805
    },
    {
      "epoch": 51.58004158004158,
      "grad_norm": 0.00012486219929996878,
      "learning_rate": 0.09468274785998718,
      "loss": 0.2462,
      "num_input_tokens_seen": 18921344,
      "step": 24810
    },
    {
      "epoch": 51.59043659043659,
      "grad_norm": 0.00010473545989952981,
      "learning_rate": 0.09462799911230127,
      "loss": 0.2746,
      "num_input_tokens_seen": 18925248,
      "step": 24815
    },
    {
      "epoch": 51.6008316008316,
      "grad_norm": 0.0002909967442974448,
      "learning_rate": 0.0945732589036737,
      "loss": 0.2749,
      "num_input_tokens_seen": 18929056,
      "step": 24820
    },
    {
      "epoch": 51.61122661122661,
      "grad_norm": 0.00019620853709056973,
      "learning_rate": 0.09451852724254614,
      "loss": 0.2697,
      "num_input_tokens_seen": 18932864,
      "step": 24825
    },
    {
      "epoch": 51.62162162162162,
      "grad_norm": 0.00010172414476983249,
      "learning_rate": 0.09446380413735894,
      "loss": 0.2552,
      "num_input_tokens_seen": 18936736,
      "step": 24830
    },
    {
      "epoch": 51.63201663201663,
      "grad_norm": 0.00010708453191909939,
      "learning_rate": 0.09440908959655099,
      "loss": 0.249,
      "num_input_tokens_seen": 18940608,
      "step": 24835
    },
    {
      "epoch": 51.642411642411645,
      "grad_norm": 0.00036251608980819583,
      "learning_rate": 0.09435438362856004,
      "loss": 0.2804,
      "num_input_tokens_seen": 18944448,
      "step": 24840
    },
    {
      "epoch": 51.65280665280665,
      "grad_norm": 0.0007139127119444311,
      "learning_rate": 0.0942996862418225,
      "loss": 0.2722,
      "num_input_tokens_seen": 18948352,
      "step": 24845
    },
    {
      "epoch": 51.66320166320166,
      "grad_norm": 0.00022203786647878587,
      "learning_rate": 0.09424499744477322,
      "loss": 0.2721,
      "num_input_tokens_seen": 18952160,
      "step": 24850
    },
    {
      "epoch": 51.67359667359668,
      "grad_norm": 0.00028811063384637237,
      "learning_rate": 0.09419031724584608,
      "loss": 0.2486,
      "num_input_tokens_seen": 18955840,
      "step": 24855
    },
    {
      "epoch": 51.683991683991685,
      "grad_norm": 0.00017130660125985742,
      "learning_rate": 0.09413564565347331,
      "loss": 0.2615,
      "num_input_tokens_seen": 18959712,
      "step": 24860
    },
    {
      "epoch": 51.694386694386694,
      "grad_norm": 0.0005064566503278911,
      "learning_rate": 0.094080982676086,
      "loss": 0.2501,
      "num_input_tokens_seen": 18963616,
      "step": 24865
    },
    {
      "epoch": 51.7047817047817,
      "grad_norm": 3.131589983240701e-05,
      "learning_rate": 0.09402632832211395,
      "loss": 0.2653,
      "num_input_tokens_seen": 18967520,
      "step": 24870
    },
    {
      "epoch": 51.71517671517672,
      "grad_norm": 8.942348358687013e-05,
      "learning_rate": 0.09397168259998541,
      "loss": 0.2633,
      "num_input_tokens_seen": 18971168,
      "step": 24875
    },
    {
      "epoch": 51.725571725571726,
      "grad_norm": 0.0004058416816405952,
      "learning_rate": 0.09391704551812759,
      "loss": 0.2659,
      "num_input_tokens_seen": 18974944,
      "step": 24880
    },
    {
      "epoch": 51.735966735966734,
      "grad_norm": 0.00015918316785246134,
      "learning_rate": 0.09386241708496605,
      "loss": 0.2931,
      "num_input_tokens_seen": 18978880,
      "step": 24885
    },
    {
      "epoch": 51.74636174636175,
      "grad_norm": 0.0003292606270406395,
      "learning_rate": 0.09380779730892527,
      "loss": 0.2582,
      "num_input_tokens_seen": 18982656,
      "step": 24890
    },
    {
      "epoch": 51.75675675675676,
      "grad_norm": 0.00020529964240267873,
      "learning_rate": 0.09375318619842836,
      "loss": 0.2664,
      "num_input_tokens_seen": 18986432,
      "step": 24895
    },
    {
      "epoch": 51.767151767151766,
      "grad_norm": 0.000501190428622067,
      "learning_rate": 0.09369858376189696,
      "loss": 0.2668,
      "num_input_tokens_seen": 18990304,
      "step": 24900
    },
    {
      "epoch": 51.777546777546775,
      "grad_norm": 0.0001369376404909417,
      "learning_rate": 0.09364399000775143,
      "loss": 0.271,
      "num_input_tokens_seen": 18994112,
      "step": 24905
    },
    {
      "epoch": 51.78794178794179,
      "grad_norm": 9.09310911083594e-05,
      "learning_rate": 0.09358940494441093,
      "loss": 0.2555,
      "num_input_tokens_seen": 18998272,
      "step": 24910
    },
    {
      "epoch": 51.7983367983368,
      "grad_norm": 0.00047601928235962987,
      "learning_rate": 0.09353482858029301,
      "loss": 0.2285,
      "num_input_tokens_seen": 19002144,
      "step": 24915
    },
    {
      "epoch": 51.80873180873181,
      "grad_norm": 0.00019199545204173774,
      "learning_rate": 0.09348026092381419,
      "loss": 0.2494,
      "num_input_tokens_seen": 19006016,
      "step": 24920
    },
    {
      "epoch": 51.81912681912682,
      "grad_norm": 0.0001717114937491715,
      "learning_rate": 0.09342570198338931,
      "loss": 0.2846,
      "num_input_tokens_seen": 19009856,
      "step": 24925
    },
    {
      "epoch": 51.82952182952183,
      "grad_norm": 7.368923252215609e-05,
      "learning_rate": 0.0933711517674322,
      "loss": 0.2853,
      "num_input_tokens_seen": 19013696,
      "step": 24930
    },
    {
      "epoch": 51.83991683991684,
      "grad_norm": 0.00044414735748432577,
      "learning_rate": 0.09331661028435513,
      "loss": 0.2561,
      "num_input_tokens_seen": 19017504,
      "step": 24935
    },
    {
      "epoch": 51.85031185031185,
      "grad_norm": 0.0004908243427053094,
      "learning_rate": 0.09326207754256909,
      "loss": 0.2656,
      "num_input_tokens_seen": 19021280,
      "step": 24940
    },
    {
      "epoch": 51.86070686070686,
      "grad_norm": 0.00013243909052107483,
      "learning_rate": 0.09320755355048366,
      "loss": 0.263,
      "num_input_tokens_seen": 19025120,
      "step": 24945
    },
    {
      "epoch": 51.87110187110187,
      "grad_norm": 0.00020346205565147102,
      "learning_rate": 0.09315303831650722,
      "loss": 0.273,
      "num_input_tokens_seen": 19028896,
      "step": 24950
    },
    {
      "epoch": 51.88149688149688,
      "grad_norm": 0.0003335009969305247,
      "learning_rate": 0.09309853184904661,
      "loss": 0.2524,
      "num_input_tokens_seen": 19032832,
      "step": 24955
    },
    {
      "epoch": 51.891891891891895,
      "grad_norm": 0.0003491482057143003,
      "learning_rate": 0.09304403415650753,
      "loss": 0.2586,
      "num_input_tokens_seen": 19036608,
      "step": 24960
    },
    {
      "epoch": 51.9022869022869,
      "grad_norm": 0.00011952204658882692,
      "learning_rate": 0.09298954524729405,
      "loss": 0.2572,
      "num_input_tokens_seen": 19040384,
      "step": 24965
    },
    {
      "epoch": 51.91268191268191,
      "grad_norm": 0.00010351989476475865,
      "learning_rate": 0.09293506512980916,
      "loss": 0.2685,
      "num_input_tokens_seen": 19044128,
      "step": 24970
    },
    {
      "epoch": 51.92307692307692,
      "grad_norm": 0.00025333426310680807,
      "learning_rate": 0.0928805938124544,
      "loss": 0.2752,
      "num_input_tokens_seen": 19047936,
      "step": 24975
    },
    {
      "epoch": 51.933471933471935,
      "grad_norm": 0.0004497557529248297,
      "learning_rate": 0.09282613130362982,
      "loss": 0.2653,
      "num_input_tokens_seen": 19051712,
      "step": 24980
    },
    {
      "epoch": 51.943866943866944,
      "grad_norm": 0.00020699689048342407,
      "learning_rate": 0.09277167761173427,
      "loss": 0.2794,
      "num_input_tokens_seen": 19055392,
      "step": 24985
    },
    {
      "epoch": 51.95426195426195,
      "grad_norm": 0.00016919725749175996,
      "learning_rate": 0.0927172327451653,
      "loss": 0.2742,
      "num_input_tokens_seen": 19059136,
      "step": 24990
    },
    {
      "epoch": 51.96465696465697,
      "grad_norm": 0.00012547911319416016,
      "learning_rate": 0.09266279671231882,
      "loss": 0.2621,
      "num_input_tokens_seen": 19062880,
      "step": 24995
    },
    {
      "epoch": 51.975051975051976,
      "grad_norm": 8.184587204596028e-05,
      "learning_rate": 0.09260836952158967,
      "loss": 0.2668,
      "num_input_tokens_seen": 19066528,
      "step": 25000
    },
    {
      "epoch": 51.975051975051976,
      "eval_loss": 0.24882346391677856,
      "eval_runtime": 13.3923,
      "eval_samples_per_second": 63.918,
      "eval_steps_per_second": 15.979,
      "num_input_tokens_seen": 19066528,
      "step": 25000
    },
    {
      "epoch": 51.985446985446984,
      "grad_norm": 3.0351191526278853e-05,
      "learning_rate": 0.09255395118137114,
      "loss": 0.2454,
      "num_input_tokens_seen": 19070368,
      "step": 25005
    },
    {
      "epoch": 51.99584199584199,
      "grad_norm": 0.00014118723629508168,
      "learning_rate": 0.09249954170005527,
      "loss": 0.2351,
      "num_input_tokens_seen": 19074368,
      "step": 25010
    },
    {
      "epoch": 52.00623700623701,
      "grad_norm": 5.4178875871002674e-05,
      "learning_rate": 0.0924451410860327,
      "loss": 0.2854,
      "num_input_tokens_seen": 19078008,
      "step": 25015
    },
    {
      "epoch": 52.016632016632016,
      "grad_norm": 0.0001686487375991419,
      "learning_rate": 0.09239074934769258,
      "loss": 0.2552,
      "num_input_tokens_seen": 19081944,
      "step": 25020
    },
    {
      "epoch": 52.027027027027025,
      "grad_norm": 0.00012034373503411189,
      "learning_rate": 0.09233636649342288,
      "loss": 0.2626,
      "num_input_tokens_seen": 19085592,
      "step": 25025
    },
    {
      "epoch": 52.03742203742204,
      "grad_norm": 0.00013611235772259533,
      "learning_rate": 0.09228199253161017,
      "loss": 0.253,
      "num_input_tokens_seen": 19089496,
      "step": 25030
    },
    {
      "epoch": 52.04781704781705,
      "grad_norm": 0.00010463312355568632,
      "learning_rate": 0.09222762747063949,
      "loss": 0.2752,
      "num_input_tokens_seen": 19093368,
      "step": 25035
    },
    {
      "epoch": 52.05821205821206,
      "grad_norm": 0.00023305570357479155,
      "learning_rate": 0.09217327131889473,
      "loss": 0.2671,
      "num_input_tokens_seen": 19097048,
      "step": 25040
    },
    {
      "epoch": 52.06860706860707,
      "grad_norm": 0.0006485827616415918,
      "learning_rate": 0.09211892408475818,
      "loss": 0.2694,
      "num_input_tokens_seen": 19100792,
      "step": 25045
    },
    {
      "epoch": 52.07900207900208,
      "grad_norm": 0.00048761325888335705,
      "learning_rate": 0.09206458577661089,
      "loss": 0.2725,
      "num_input_tokens_seen": 19104568,
      "step": 25050
    },
    {
      "epoch": 52.08939708939709,
      "grad_norm": 0.00032891842420212924,
      "learning_rate": 0.09201025640283263,
      "loss": 0.2671,
      "num_input_tokens_seen": 19108632,
      "step": 25055
    },
    {
      "epoch": 52.0997920997921,
      "grad_norm": 0.0001725823531160131,
      "learning_rate": 0.09195593597180148,
      "loss": 0.2589,
      "num_input_tokens_seen": 19112376,
      "step": 25060
    },
    {
      "epoch": 52.11018711018711,
      "grad_norm": 8.355977479368448e-05,
      "learning_rate": 0.09190162449189444,
      "loss": 0.2598,
      "num_input_tokens_seen": 19116184,
      "step": 25065
    },
    {
      "epoch": 52.12058212058212,
      "grad_norm": 0.00017736456356942654,
      "learning_rate": 0.09184732197148705,
      "loss": 0.2458,
      "num_input_tokens_seen": 19119960,
      "step": 25070
    },
    {
      "epoch": 52.13097713097713,
      "grad_norm": 0.00014506731531582773,
      "learning_rate": 0.09179302841895343,
      "loss": 0.2838,
      "num_input_tokens_seen": 19123672,
      "step": 25075
    },
    {
      "epoch": 52.141372141372145,
      "grad_norm": 0.00021496537374332547,
      "learning_rate": 0.09173874384266625,
      "loss": 0.2552,
      "num_input_tokens_seen": 19127608,
      "step": 25080
    },
    {
      "epoch": 52.15176715176715,
      "grad_norm": 0.0004486608086153865,
      "learning_rate": 0.09168446825099695,
      "loss": 0.2744,
      "num_input_tokens_seen": 19131352,
      "step": 25085
    },
    {
      "epoch": 52.16216216216216,
      "grad_norm": 0.0001984643458854407,
      "learning_rate": 0.09163020165231545,
      "loss": 0.2679,
      "num_input_tokens_seen": 19135256,
      "step": 25090
    },
    {
      "epoch": 52.17255717255717,
      "grad_norm": 0.0003029173822142184,
      "learning_rate": 0.09157594405499044,
      "loss": 0.2552,
      "num_input_tokens_seen": 19139000,
      "step": 25095
    },
    {
      "epoch": 52.182952182952185,
      "grad_norm": 0.0006888463976792991,
      "learning_rate": 0.09152169546738899,
      "loss": 0.2671,
      "num_input_tokens_seen": 19142744,
      "step": 25100
    },
    {
      "epoch": 52.19334719334719,
      "grad_norm": 0.0003807439934462309,
      "learning_rate": 0.09146745589787698,
      "loss": 0.2717,
      "num_input_tokens_seen": 19146616,
      "step": 25105
    },
    {
      "epoch": 52.2037422037422,
      "grad_norm": 0.0003893484245054424,
      "learning_rate": 0.09141322535481891,
      "loss": 0.286,
      "num_input_tokens_seen": 19150328,
      "step": 25110
    },
    {
      "epoch": 52.21413721413722,
      "grad_norm": 8.591903315391392e-05,
      "learning_rate": 0.0913590038465777,
      "loss": 0.2721,
      "num_input_tokens_seen": 19154104,
      "step": 25115
    },
    {
      "epoch": 52.224532224532226,
      "grad_norm": 0.00023204648459795862,
      "learning_rate": 0.09130479138151505,
      "loss": 0.2782,
      "num_input_tokens_seen": 19157816,
      "step": 25120
    },
    {
      "epoch": 52.234927234927234,
      "grad_norm": 0.0004267179756425321,
      "learning_rate": 0.09125058796799114,
      "loss": 0.2735,
      "num_input_tokens_seen": 19161496,
      "step": 25125
    },
    {
      "epoch": 52.24532224532224,
      "grad_norm": 0.00018816835654433817,
      "learning_rate": 0.09119639361436485,
      "loss": 0.2618,
      "num_input_tokens_seen": 19165240,
      "step": 25130
    },
    {
      "epoch": 52.25571725571726,
      "grad_norm": 0.0002517510438337922,
      "learning_rate": 0.09114220832899368,
      "loss": 0.2521,
      "num_input_tokens_seen": 19169240,
      "step": 25135
    },
    {
      "epoch": 52.266112266112266,
      "grad_norm": 0.00038964030682109296,
      "learning_rate": 0.0910880321202336,
      "loss": 0.2596,
      "num_input_tokens_seen": 19172952,
      "step": 25140
    },
    {
      "epoch": 52.276507276507274,
      "grad_norm": 0.00012928710202686489,
      "learning_rate": 0.09103386499643933,
      "loss": 0.2559,
      "num_input_tokens_seen": 19176856,
      "step": 25145
    },
    {
      "epoch": 52.28690228690229,
      "grad_norm": 0.0004698314005509019,
      "learning_rate": 0.09097970696596407,
      "loss": 0.2959,
      "num_input_tokens_seen": 19180600,
      "step": 25150
    },
    {
      "epoch": 52.2972972972973,
      "grad_norm": 0.00024322712852153927,
      "learning_rate": 0.09092555803715971,
      "loss": 0.2792,
      "num_input_tokens_seen": 19184408,
      "step": 25155
    },
    {
      "epoch": 52.30769230769231,
      "grad_norm": 0.0001793637202354148,
      "learning_rate": 0.0908714182183767,
      "loss": 0.2761,
      "num_input_tokens_seen": 19188312,
      "step": 25160
    },
    {
      "epoch": 52.318087318087315,
      "grad_norm": 0.0005349906859919429,
      "learning_rate": 0.090817287517964,
      "loss": 0.2593,
      "num_input_tokens_seen": 19191992,
      "step": 25165
    },
    {
      "epoch": 52.32848232848233,
      "grad_norm": 0.00027267204131931067,
      "learning_rate": 0.09076316594426931,
      "loss": 0.292,
      "num_input_tokens_seen": 19195896,
      "step": 25170
    },
    {
      "epoch": 52.33887733887734,
      "grad_norm": 9.870494977803901e-05,
      "learning_rate": 0.09070905350563888,
      "loss": 0.2545,
      "num_input_tokens_seen": 19199736,
      "step": 25175
    },
    {
      "epoch": 52.34927234927235,
      "grad_norm": 0.00015784376591909677,
      "learning_rate": 0.09065495021041745,
      "loss": 0.247,
      "num_input_tokens_seen": 19203544,
      "step": 25180
    },
    {
      "epoch": 52.35966735966736,
      "grad_norm": 0.0001943733950611204,
      "learning_rate": 0.09060085606694851,
      "loss": 0.2453,
      "num_input_tokens_seen": 19207288,
      "step": 25185
    },
    {
      "epoch": 52.37006237006237,
      "grad_norm": 0.0005111717619001865,
      "learning_rate": 0.09054677108357405,
      "loss": 0.2551,
      "num_input_tokens_seen": 19211128,
      "step": 25190
    },
    {
      "epoch": 52.38045738045738,
      "grad_norm": 0.0001630577608011663,
      "learning_rate": 0.09049269526863457,
      "loss": 0.287,
      "num_input_tokens_seen": 19214904,
      "step": 25195
    },
    {
      "epoch": 52.39085239085239,
      "grad_norm": 0.00021629508410114795,
      "learning_rate": 0.09043862863046935,
      "loss": 0.2498,
      "num_input_tokens_seen": 19218616,
      "step": 25200
    },
    {
      "epoch": 52.39085239085239,
      "eval_loss": 0.251054972410202,
      "eval_runtime": 13.4132,
      "eval_samples_per_second": 63.818,
      "eval_steps_per_second": 15.954,
      "num_input_tokens_seen": 19218616,
      "step": 25200
    },
    {
      "epoch": 52.4012474012474,
      "grad_norm": 0.00021273968741297722,
      "learning_rate": 0.09038457117741602,
      "loss": 0.2796,
      "num_input_tokens_seen": 19222488,
      "step": 25205
    },
    {
      "epoch": 52.41164241164241,
      "grad_norm": 0.000733297667466104,
      "learning_rate": 0.09033052291781099,
      "loss": 0.246,
      "num_input_tokens_seen": 19226296,
      "step": 25210
    },
    {
      "epoch": 52.42203742203742,
      "grad_norm": 0.0006143064820207655,
      "learning_rate": 0.09027648385998926,
      "loss": 0.2878,
      "num_input_tokens_seen": 19230104,
      "step": 25215
    },
    {
      "epoch": 52.432432432432435,
      "grad_norm": 5.60740809305571e-05,
      "learning_rate": 0.09022245401228417,
      "loss": 0.2609,
      "num_input_tokens_seen": 19233912,
      "step": 25220
    },
    {
      "epoch": 52.44282744282744,
      "grad_norm": 0.00037995801540091634,
      "learning_rate": 0.09016843338302792,
      "loss": 0.2736,
      "num_input_tokens_seen": 19237688,
      "step": 25225
    },
    {
      "epoch": 52.45322245322245,
      "grad_norm": 0.00019657934899441898,
      "learning_rate": 0.09011442198055115,
      "loss": 0.2523,
      "num_input_tokens_seen": 19241464,
      "step": 25230
    },
    {
      "epoch": 52.46361746361746,
      "grad_norm": 0.0004387819499243051,
      "learning_rate": 0.09006041981318305,
      "loss": 0.2744,
      "num_input_tokens_seen": 19245496,
      "step": 25235
    },
    {
      "epoch": 52.474012474012476,
      "grad_norm": 0.0006431292276829481,
      "learning_rate": 0.09000642688925149,
      "loss": 0.2478,
      "num_input_tokens_seen": 19249240,
      "step": 25240
    },
    {
      "epoch": 52.484407484407484,
      "grad_norm": 0.00035885735996998847,
      "learning_rate": 0.0899524432170828,
      "loss": 0.2418,
      "num_input_tokens_seen": 19252984,
      "step": 25245
    },
    {
      "epoch": 52.49480249480249,
      "grad_norm": 0.00022899993928149343,
      "learning_rate": 0.08989846880500196,
      "loss": 0.2382,
      "num_input_tokens_seen": 19256760,
      "step": 25250
    },
    {
      "epoch": 52.50519750519751,
      "grad_norm": 0.00016251785564236343,
      "learning_rate": 0.08984450366133256,
      "loss": 0.2313,
      "num_input_tokens_seen": 19260664,
      "step": 25255
    },
    {
      "epoch": 52.515592515592516,
      "grad_norm": 0.00027021896676160395,
      "learning_rate": 0.08979054779439664,
      "loss": 0.2866,
      "num_input_tokens_seen": 19264472,
      "step": 25260
    },
    {
      "epoch": 52.525987525987524,
      "grad_norm": 0.00012837324175052345,
      "learning_rate": 0.08973660121251485,
      "loss": 0.3001,
      "num_input_tokens_seen": 19268216,
      "step": 25265
    },
    {
      "epoch": 52.53638253638254,
      "grad_norm": 0.0005488950992003083,
      "learning_rate": 0.08968266392400655,
      "loss": 0.2648,
      "num_input_tokens_seen": 19272120,
      "step": 25270
    },
    {
      "epoch": 52.54677754677755,
      "grad_norm": 0.00010601641406537965,
      "learning_rate": 0.0896287359371894,
      "loss": 0.2671,
      "num_input_tokens_seen": 19275992,
      "step": 25275
    },
    {
      "epoch": 52.55717255717256,
      "grad_norm": 0.0002984137099701911,
      "learning_rate": 0.08957481726037989,
      "loss": 0.2844,
      "num_input_tokens_seen": 19279896,
      "step": 25280
    },
    {
      "epoch": 52.567567567567565,
      "grad_norm": 7.243233267217875e-05,
      "learning_rate": 0.08952090790189286,
      "loss": 0.267,
      "num_input_tokens_seen": 19283800,
      "step": 25285
    },
    {
      "epoch": 52.57796257796258,
      "grad_norm": 0.00010885141819017008,
      "learning_rate": 0.08946700787004187,
      "loss": 0.2634,
      "num_input_tokens_seen": 19287480,
      "step": 25290
    },
    {
      "epoch": 52.58835758835759,
      "grad_norm": 0.00011178448767168447,
      "learning_rate": 0.08941311717313899,
      "loss": 0.2931,
      "num_input_tokens_seen": 19291288,
      "step": 25295
    },
    {
      "epoch": 52.5987525987526,
      "grad_norm": 0.00012702326057478786,
      "learning_rate": 0.08935923581949483,
      "loss": 0.2642,
      "num_input_tokens_seen": 19295192,
      "step": 25300
    },
    {
      "epoch": 52.60914760914761,
      "grad_norm": 6.52926682960242e-05,
      "learning_rate": 0.0893053638174185,
      "loss": 0.2841,
      "num_input_tokens_seen": 19298872,
      "step": 25305
    },
    {
      "epoch": 52.61954261954262,
      "grad_norm": 0.0003655454784166068,
      "learning_rate": 0.0892515011752179,
      "loss": 0.2569,
      "num_input_tokens_seen": 19302584,
      "step": 25310
    },
    {
      "epoch": 52.62993762993763,
      "grad_norm": 0.00011825864930870011,
      "learning_rate": 0.08919764790119918,
      "loss": 0.2391,
      "num_input_tokens_seen": 19306328,
      "step": 25315
    },
    {
      "epoch": 52.64033264033264,
      "grad_norm": 0.0002848030417226255,
      "learning_rate": 0.08914380400366727,
      "loss": 0.2526,
      "num_input_tokens_seen": 19310328,
      "step": 25320
    },
    {
      "epoch": 52.65072765072765,
      "grad_norm": 0.00033655259176157415,
      "learning_rate": 0.08908996949092551,
      "loss": 0.2618,
      "num_input_tokens_seen": 19314104,
      "step": 25325
    },
    {
      "epoch": 52.66112266112266,
      "grad_norm": 0.0003371915954630822,
      "learning_rate": 0.08903614437127592,
      "loss": 0.2577,
      "num_input_tokens_seen": 19317880,
      "step": 25330
    },
    {
      "epoch": 52.67151767151767,
      "grad_norm": 0.00027667218819260597,
      "learning_rate": 0.088982328653019,
      "loss": 0.2564,
      "num_input_tokens_seen": 19321720,
      "step": 25335
    },
    {
      "epoch": 52.681912681912685,
      "grad_norm": 8.618910942459479e-05,
      "learning_rate": 0.0889285223444538,
      "loss": 0.2631,
      "num_input_tokens_seen": 19325560,
      "step": 25340
    },
    {
      "epoch": 52.69230769230769,
      "grad_norm": 0.00023816869361326098,
      "learning_rate": 0.08887472545387787,
      "loss": 0.298,
      "num_input_tokens_seen": 19329368,
      "step": 25345
    },
    {
      "epoch": 52.7027027027027,
      "grad_norm": 0.0003238253411836922,
      "learning_rate": 0.08882093798958751,
      "loss": 0.2758,
      "num_input_tokens_seen": 19333368,
      "step": 25350
    },
    {
      "epoch": 52.71309771309771,
      "grad_norm": 0.0005789937567897141,
      "learning_rate": 0.08876715995987726,
      "loss": 0.2739,
      "num_input_tokens_seen": 19337112,
      "step": 25355
    },
    {
      "epoch": 52.723492723492726,
      "grad_norm": 0.0003477366117294878,
      "learning_rate": 0.08871339137304052,
      "loss": 0.2753,
      "num_input_tokens_seen": 19340952,
      "step": 25360
    },
    {
      "epoch": 52.733887733887734,
      "grad_norm": 0.00014633526734542102,
      "learning_rate": 0.0886596322373689,
      "loss": 0.2725,
      "num_input_tokens_seen": 19344728,
      "step": 25365
    },
    {
      "epoch": 52.74428274428274,
      "grad_norm": 0.0005999618442729115,
      "learning_rate": 0.08860588256115293,
      "loss": 0.2568,
      "num_input_tokens_seen": 19348504,
      "step": 25370
    },
    {
      "epoch": 52.75467775467776,
      "grad_norm": 0.00025870188255794346,
      "learning_rate": 0.0885521423526814,
      "loss": 0.2696,
      "num_input_tokens_seen": 19352152,
      "step": 25375
    },
    {
      "epoch": 52.765072765072766,
      "grad_norm": 0.0002299930201843381,
      "learning_rate": 0.08849841162024165,
      "loss": 0.2433,
      "num_input_tokens_seen": 19355992,
      "step": 25380
    },
    {
      "epoch": 52.775467775467774,
      "grad_norm": 0.00029891615849919617,
      "learning_rate": 0.08844469037211973,
      "loss": 0.2841,
      "num_input_tokens_seen": 19359736,
      "step": 25385
    },
    {
      "epoch": 52.78586278586278,
      "grad_norm": 0.0001081323207472451,
      "learning_rate": 0.08839097861660014,
      "loss": 0.2805,
      "num_input_tokens_seen": 19363384,
      "step": 25390
    },
    {
      "epoch": 52.7962577962578,
      "grad_norm": 8.824890392133966e-05,
      "learning_rate": 0.08833727636196585,
      "loss": 0.269,
      "num_input_tokens_seen": 19367192,
      "step": 25395
    },
    {
      "epoch": 52.80665280665281,
      "grad_norm": 0.0001952427119249478,
      "learning_rate": 0.08828358361649848,
      "loss": 0.2658,
      "num_input_tokens_seen": 19370872,
      "step": 25400
    },
    {
      "epoch": 52.80665280665281,
      "eval_loss": 0.2578355073928833,
      "eval_runtime": 13.3869,
      "eval_samples_per_second": 63.943,
      "eval_steps_per_second": 15.986,
      "num_input_tokens_seen": 19370872,
      "step": 25400
    },
    {
      "epoch": 52.817047817047815,
      "grad_norm": 6.909188232384622e-05,
      "learning_rate": 0.08822990038847807,
      "loss": 0.2646,
      "num_input_tokens_seen": 19374840,
      "step": 25405
    },
    {
      "epoch": 52.82744282744283,
      "grad_norm": 0.00022669322788715363,
      "learning_rate": 0.08817622668618325,
      "loss": 0.2593,
      "num_input_tokens_seen": 19378616,
      "step": 25410
    },
    {
      "epoch": 52.83783783783784,
      "grad_norm": 0.0005334245506674051,
      "learning_rate": 0.08812256251789125,
      "loss": 0.2794,
      "num_input_tokens_seen": 19382328,
      "step": 25415
    },
    {
      "epoch": 52.84823284823285,
      "grad_norm": 0.00016129582945723087,
      "learning_rate": 0.08806890789187766,
      "loss": 0.2406,
      "num_input_tokens_seen": 19385976,
      "step": 25420
    },
    {
      "epoch": 52.858627858627855,
      "grad_norm": 0.00036462044226936996,
      "learning_rate": 0.08801526281641672,
      "loss": 0.2655,
      "num_input_tokens_seen": 19389944,
      "step": 25425
    },
    {
      "epoch": 52.86902286902287,
      "grad_norm": 0.0006434862734749913,
      "learning_rate": 0.0879616272997813,
      "loss": 0.2562,
      "num_input_tokens_seen": 19393784,
      "step": 25430
    },
    {
      "epoch": 52.87941787941788,
      "grad_norm": 0.00046026933705434203,
      "learning_rate": 0.08790800135024247,
      "loss": 0.2453,
      "num_input_tokens_seen": 19397656,
      "step": 25435
    },
    {
      "epoch": 52.88981288981289,
      "grad_norm": 0.00041225148015655577,
      "learning_rate": 0.08785438497607023,
      "loss": 0.3011,
      "num_input_tokens_seen": 19401656,
      "step": 25440
    },
    {
      "epoch": 52.9002079002079,
      "grad_norm": 0.0007981536327861249,
      "learning_rate": 0.08780077818553277,
      "loss": 0.2659,
      "num_input_tokens_seen": 19405592,
      "step": 25445
    },
    {
      "epoch": 52.91060291060291,
      "grad_norm": 0.000723455857951194,
      "learning_rate": 0.0877471809868969,
      "loss": 0.2676,
      "num_input_tokens_seen": 19409496,
      "step": 25450
    },
    {
      "epoch": 52.92099792099792,
      "grad_norm": 4.023011933895759e-05,
      "learning_rate": 0.08769359338842811,
      "loss": 0.2633,
      "num_input_tokens_seen": 19413240,
      "step": 25455
    },
    {
      "epoch": 52.931392931392935,
      "grad_norm": 0.00022103404626250267,
      "learning_rate": 0.08764001539839016,
      "loss": 0.2448,
      "num_input_tokens_seen": 19417176,
      "step": 25460
    },
    {
      "epoch": 52.94178794178794,
      "grad_norm": 0.00010065849346574396,
      "learning_rate": 0.08758644702504548,
      "loss": 0.253,
      "num_input_tokens_seen": 19421016,
      "step": 25465
    },
    {
      "epoch": 52.95218295218295,
      "grad_norm": 0.00019730512576643378,
      "learning_rate": 0.0875328882766551,
      "loss": 0.3083,
      "num_input_tokens_seen": 19424760,
      "step": 25470
    },
    {
      "epoch": 52.96257796257796,
      "grad_norm": 0.00032311183167621493,
      "learning_rate": 0.08747933916147828,
      "loss": 0.2687,
      "num_input_tokens_seen": 19428568,
      "step": 25475
    },
    {
      "epoch": 52.972972972972975,
      "grad_norm": 0.0005062875570729375,
      "learning_rate": 0.0874257996877731,
      "loss": 0.2447,
      "num_input_tokens_seen": 19432344,
      "step": 25480
    },
    {
      "epoch": 52.983367983367984,
      "grad_norm": 0.0002024264686042443,
      "learning_rate": 0.08737226986379593,
      "loss": 0.254,
      "num_input_tokens_seen": 19436120,
      "step": 25485
    },
    {
      "epoch": 52.99376299376299,
      "grad_norm": 7.391955296043307e-05,
      "learning_rate": 0.08731874969780173,
      "loss": 0.2742,
      "num_input_tokens_seen": 19439960,
      "step": 25490
    },
    {
      "epoch": 53.00415800415801,
      "grad_norm": 0.00022842263570055366,
      "learning_rate": 0.08726523919804412,
      "loss": 0.245,
      "num_input_tokens_seen": 19443752,
      "step": 25495
    },
    {
      "epoch": 53.014553014553016,
      "grad_norm": 0.0002269794058520347,
      "learning_rate": 0.08721173837277492,
      "loss": 0.2573,
      "num_input_tokens_seen": 19447496,
      "step": 25500
    },
    {
      "epoch": 53.024948024948024,
      "grad_norm": 0.0003161179192829877,
      "learning_rate": 0.08715824723024479,
      "loss": 0.2689,
      "num_input_tokens_seen": 19451304,
      "step": 25505
    },
    {
      "epoch": 53.03534303534303,
      "grad_norm": 0.00027071021031588316,
      "learning_rate": 0.08710476577870258,
      "loss": 0.2705,
      "num_input_tokens_seen": 19455240,
      "step": 25510
    },
    {
      "epoch": 53.04573804573805,
      "grad_norm": 0.0002087950415443629,
      "learning_rate": 0.08705129402639587,
      "loss": 0.2858,
      "num_input_tokens_seen": 19458888,
      "step": 25515
    },
    {
      "epoch": 53.056133056133056,
      "grad_norm": 0.0003871911612804979,
      "learning_rate": 0.08699783198157078,
      "loss": 0.2792,
      "num_input_tokens_seen": 19462728,
      "step": 25520
    },
    {
      "epoch": 53.066528066528065,
      "grad_norm": 0.00022111559519544244,
      "learning_rate": 0.08694437965247163,
      "loss": 0.2683,
      "num_input_tokens_seen": 19466536,
      "step": 25525
    },
    {
      "epoch": 53.07692307692308,
      "grad_norm": 0.00046337186358869076,
      "learning_rate": 0.08689093704734165,
      "loss": 0.2738,
      "num_input_tokens_seen": 19470344,
      "step": 25530
    },
    {
      "epoch": 53.08731808731809,
      "grad_norm": 0.00012477106065489352,
      "learning_rate": 0.08683750417442222,
      "loss": 0.2704,
      "num_input_tokens_seen": 19474344,
      "step": 25535
    },
    {
      "epoch": 53.0977130977131,
      "grad_norm": 0.00011562240979401395,
      "learning_rate": 0.08678408104195334,
      "loss": 0.2814,
      "num_input_tokens_seen": 19478312,
      "step": 25540
    },
    {
      "epoch": 53.108108108108105,
      "grad_norm": 0.0009324764832854271,
      "learning_rate": 0.08673066765817365,
      "loss": 0.2691,
      "num_input_tokens_seen": 19482088,
      "step": 25545
    },
    {
      "epoch": 53.11850311850312,
      "grad_norm": 0.00013382262841332704,
      "learning_rate": 0.08667726403132005,
      "loss": 0.2506,
      "num_input_tokens_seen": 19485928,
      "step": 25550
    },
    {
      "epoch": 53.12889812889813,
      "grad_norm": 8.289182733278722e-05,
      "learning_rate": 0.0866238701696281,
      "loss": 0.2534,
      "num_input_tokens_seen": 19489736,
      "step": 25555
    },
    {
      "epoch": 53.13929313929314,
      "grad_norm": 0.00021687426487915218,
      "learning_rate": 0.08657048608133185,
      "loss": 0.275,
      "num_input_tokens_seen": 19493544,
      "step": 25560
    },
    {
      "epoch": 53.14968814968815,
      "grad_norm": 9.0418616309762e-05,
      "learning_rate": 0.08651711177466369,
      "loss": 0.2662,
      "num_input_tokens_seen": 19497416,
      "step": 25565
    },
    {
      "epoch": 53.16008316008316,
      "grad_norm": 0.0001370221289107576,
      "learning_rate": 0.08646374725785466,
      "loss": 0.274,
      "num_input_tokens_seen": 19501320,
      "step": 25570
    },
    {
      "epoch": 53.17047817047817,
      "grad_norm": 0.0005844387924298644,
      "learning_rate": 0.08641039253913434,
      "loss": 0.2659,
      "num_input_tokens_seen": 19505192,
      "step": 25575
    },
    {
      "epoch": 53.18087318087318,
      "grad_norm": 0.0003213598974980414,
      "learning_rate": 0.08635704762673052,
      "loss": 0.2813,
      "num_input_tokens_seen": 19508936,
      "step": 25580
    },
    {
      "epoch": 53.19126819126819,
      "grad_norm": 0.00029099799576215446,
      "learning_rate": 0.08630371252886981,
      "loss": 0.2554,
      "num_input_tokens_seen": 19512712,
      "step": 25585
    },
    {
      "epoch": 53.2016632016632,
      "grad_norm": 0.0002532747748773545,
      "learning_rate": 0.08625038725377704,
      "loss": 0.2625,
      "num_input_tokens_seen": 19516616,
      "step": 25590
    },
    {
      "epoch": 53.21205821205821,
      "grad_norm": 0.0002777600020635873,
      "learning_rate": 0.08619707180967566,
      "loss": 0.2094,
      "num_input_tokens_seen": 19520424,
      "step": 25595
    },
    {
      "epoch": 53.222453222453225,
      "grad_norm": 0.0005434795166365802,
      "learning_rate": 0.08614376620478768,
      "loss": 0.2538,
      "num_input_tokens_seen": 19524232,
      "step": 25600
    },
    {
      "epoch": 53.222453222453225,
      "eval_loss": 0.25216010212898254,
      "eval_runtime": 13.3884,
      "eval_samples_per_second": 63.936,
      "eval_steps_per_second": 15.984,
      "num_input_tokens_seen": 19524232,
      "step": 25600
    },
    {
      "epoch": 53.232848232848234,
      "grad_norm": 0.00014258103328756988,
      "learning_rate": 0.08609047044733344,
      "loss": 0.2688,
      "num_input_tokens_seen": 19527944,
      "step": 25605
    },
    {
      "epoch": 53.24324324324324,
      "grad_norm": 0.0005566255422309041,
      "learning_rate": 0.08603718454553168,
      "loss": 0.262,
      "num_input_tokens_seen": 19531816,
      "step": 25610
    },
    {
      "epoch": 53.25363825363825,
      "grad_norm": 0.0003099758177995682,
      "learning_rate": 0.08598390850759997,
      "loss": 0.2911,
      "num_input_tokens_seen": 19535752,
      "step": 25615
    },
    {
      "epoch": 53.264033264033266,
      "grad_norm": 0.00035618332913145423,
      "learning_rate": 0.08593064234175397,
      "loss": 0.2627,
      "num_input_tokens_seen": 19539624,
      "step": 25620
    },
    {
      "epoch": 53.274428274428274,
      "grad_norm": 0.00017494460917077959,
      "learning_rate": 0.08587738605620815,
      "loss": 0.2611,
      "num_input_tokens_seen": 19543496,
      "step": 25625
    },
    {
      "epoch": 53.28482328482328,
      "grad_norm": 0.0003928478399757296,
      "learning_rate": 0.08582413965917512,
      "loss": 0.2593,
      "num_input_tokens_seen": 19547208,
      "step": 25630
    },
    {
      "epoch": 53.2952182952183,
      "grad_norm": 5.883342237211764e-05,
      "learning_rate": 0.08577090315886628,
      "loss": 0.2401,
      "num_input_tokens_seen": 19550920,
      "step": 25635
    },
    {
      "epoch": 53.305613305613306,
      "grad_norm": 0.0002586435293778777,
      "learning_rate": 0.08571767656349136,
      "loss": 0.2549,
      "num_input_tokens_seen": 19554824,
      "step": 25640
    },
    {
      "epoch": 53.316008316008315,
      "grad_norm": 0.0002915515215136111,
      "learning_rate": 0.08566445988125847,
      "loss": 0.2923,
      "num_input_tokens_seen": 19558536,
      "step": 25645
    },
    {
      "epoch": 53.32640332640332,
      "grad_norm": 0.00010289955389453098,
      "learning_rate": 0.08561125312037436,
      "loss": 0.2683,
      "num_input_tokens_seen": 19562440,
      "step": 25650
    },
    {
      "epoch": 53.33679833679834,
      "grad_norm": 7.264727901201695e-05,
      "learning_rate": 0.08555805628904424,
      "loss": 0.2342,
      "num_input_tokens_seen": 19566088,
      "step": 25655
    },
    {
      "epoch": 53.34719334719335,
      "grad_norm": 0.0003991942503489554,
      "learning_rate": 0.08550486939547161,
      "loss": 0.258,
      "num_input_tokens_seen": 19569960,
      "step": 25660
    },
    {
      "epoch": 53.357588357588355,
      "grad_norm": 0.0003789020993281156,
      "learning_rate": 0.08545169244785869,
      "loss": 0.2841,
      "num_input_tokens_seen": 19573672,
      "step": 25665
    },
    {
      "epoch": 53.36798336798337,
      "grad_norm": 0.00028057151939719915,
      "learning_rate": 0.08539852545440589,
      "loss": 0.263,
      "num_input_tokens_seen": 19577576,
      "step": 25670
    },
    {
      "epoch": 53.37837837837838,
      "grad_norm": 7.342134631471708e-05,
      "learning_rate": 0.08534536842331235,
      "loss": 0.2592,
      "num_input_tokens_seen": 19581288,
      "step": 25675
    },
    {
      "epoch": 53.38877338877339,
      "grad_norm": 0.00014633193495683372,
      "learning_rate": 0.08529222136277545,
      "loss": 0.2775,
      "num_input_tokens_seen": 19585128,
      "step": 25680
    },
    {
      "epoch": 53.3991683991684,
      "grad_norm": 0.00021866269526071846,
      "learning_rate": 0.08523908428099125,
      "loss": 0.2767,
      "num_input_tokens_seen": 19588904,
      "step": 25685
    },
    {
      "epoch": 53.40956340956341,
      "grad_norm": 9.841060091275722e-05,
      "learning_rate": 0.08518595718615402,
      "loss": 0.2679,
      "num_input_tokens_seen": 19592776,
      "step": 25690
    },
    {
      "epoch": 53.41995841995842,
      "grad_norm": 0.00016996328486129642,
      "learning_rate": 0.08513284008645675,
      "loss": 0.2626,
      "num_input_tokens_seen": 19596616,
      "step": 25695
    },
    {
      "epoch": 53.43035343035343,
      "grad_norm": 0.000500997994095087,
      "learning_rate": 0.08507973299009065,
      "loss": 0.2711,
      "num_input_tokens_seen": 19600328,
      "step": 25700
    },
    {
      "epoch": 53.44074844074844,
      "grad_norm": 0.0004233003710396588,
      "learning_rate": 0.08502663590524563,
      "loss": 0.2369,
      "num_input_tokens_seen": 19604072,
      "step": 25705
    },
    {
      "epoch": 53.45114345114345,
      "grad_norm": 0.00017534276412334293,
      "learning_rate": 0.08497354884010981,
      "loss": 0.2773,
      "num_input_tokens_seen": 19607816,
      "step": 25710
    },
    {
      "epoch": 53.46153846153846,
      "grad_norm": 0.0002948506153188646,
      "learning_rate": 0.0849204718028699,
      "loss": 0.2856,
      "num_input_tokens_seen": 19611752,
      "step": 25715
    },
    {
      "epoch": 53.471933471933475,
      "grad_norm": 0.0010303632589057088,
      "learning_rate": 0.08486740480171118,
      "loss": 0.2692,
      "num_input_tokens_seen": 19615400,
      "step": 25720
    },
    {
      "epoch": 53.482328482328484,
      "grad_norm": 0.0006806841702200472,
      "learning_rate": 0.08481434784481706,
      "loss": 0.2632,
      "num_input_tokens_seen": 19619240,
      "step": 25725
    },
    {
      "epoch": 53.49272349272349,
      "grad_norm": 0.0002270428667543456,
      "learning_rate": 0.08476130094036968,
      "loss": 0.2682,
      "num_input_tokens_seen": 19622888,
      "step": 25730
    },
    {
      "epoch": 53.5031185031185,
      "grad_norm": 0.00016551902808714658,
      "learning_rate": 0.08470826409654961,
      "loss": 0.2563,
      "num_input_tokens_seen": 19626632,
      "step": 25735
    },
    {
      "epoch": 53.513513513513516,
      "grad_norm": 0.0002010891039390117,
      "learning_rate": 0.08465523732153564,
      "loss": 0.2673,
      "num_input_tokens_seen": 19630536,
      "step": 25740
    },
    {
      "epoch": 53.523908523908524,
      "grad_norm": 0.0003778178652282804,
      "learning_rate": 0.08460222062350532,
      "loss": 0.2742,
      "num_input_tokens_seen": 19634376,
      "step": 25745
    },
    {
      "epoch": 53.53430353430353,
      "grad_norm": 0.0005589865031652153,
      "learning_rate": 0.08454921401063442,
      "loss": 0.2604,
      "num_input_tokens_seen": 19638312,
      "step": 25750
    },
    {
      "epoch": 53.54469854469855,
      "grad_norm": 0.0009035980910994112,
      "learning_rate": 0.08449621749109716,
      "loss": 0.2738,
      "num_input_tokens_seen": 19642056,
      "step": 25755
    },
    {
      "epoch": 53.555093555093556,
      "grad_norm": 0.0005097401444800198,
      "learning_rate": 0.08444323107306641,
      "loss": 0.2582,
      "num_input_tokens_seen": 19645832,
      "step": 25760
    },
    {
      "epoch": 53.565488565488565,
      "grad_norm": 0.00041103025432676077,
      "learning_rate": 0.0843902547647132,
      "loss": 0.2737,
      "num_input_tokens_seen": 19649672,
      "step": 25765
    },
    {
      "epoch": 53.57588357588357,
      "grad_norm": 0.00017348640540149063,
      "learning_rate": 0.0843372885742072,
      "loss": 0.2818,
      "num_input_tokens_seen": 19653576,
      "step": 25770
    },
    {
      "epoch": 53.58627858627859,
      "grad_norm": 0.00044966634595766664,
      "learning_rate": 0.08428433250971652,
      "loss": 0.2423,
      "num_input_tokens_seen": 19657256,
      "step": 25775
    },
    {
      "epoch": 53.5966735966736,
      "grad_norm": 0.0002476478985045105,
      "learning_rate": 0.08423138657940757,
      "loss": 0.275,
      "num_input_tokens_seen": 19661224,
      "step": 25780
    },
    {
      "epoch": 53.607068607068605,
      "grad_norm": 0.00032057089265435934,
      "learning_rate": 0.08417845079144536,
      "loss": 0.279,
      "num_input_tokens_seen": 19664968,
      "step": 25785
    },
    {
      "epoch": 53.61746361746362,
      "grad_norm": 0.0005667418008670211,
      "learning_rate": 0.08412552515399314,
      "loss": 0.2553,
      "num_input_tokens_seen": 19668840,
      "step": 25790
    },
    {
      "epoch": 53.62785862785863,
      "grad_norm": 0.00021251654834486544,
      "learning_rate": 0.08407260967521278,
      "loss": 0.2203,
      "num_input_tokens_seen": 19672616,
      "step": 25795
    },
    {
      "epoch": 53.63825363825364,
      "grad_norm": 7.873615686548874e-05,
      "learning_rate": 0.08401970436326454,
      "loss": 0.2282,
      "num_input_tokens_seen": 19676456,
      "step": 25800
    },
    {
      "epoch": 53.63825363825364,
      "eval_loss": 0.24929235875606537,
      "eval_runtime": 13.389,
      "eval_samples_per_second": 63.933,
      "eval_steps_per_second": 15.983,
      "num_input_tokens_seen": 19676456,
      "step": 25800
    },
    {
      "epoch": 53.648648648648646,
      "grad_norm": 4.673874718719162e-05,
      "learning_rate": 0.08396680922630702,
      "loss": 0.2573,
      "num_input_tokens_seen": 19680200,
      "step": 25805
    },
    {
      "epoch": 53.65904365904366,
      "grad_norm": 0.00033500452991575,
      "learning_rate": 0.08391392427249732,
      "loss": 0.2324,
      "num_input_tokens_seen": 19683944,
      "step": 25810
    },
    {
      "epoch": 53.66943866943867,
      "grad_norm": 0.00012405934103298932,
      "learning_rate": 0.08386104950999107,
      "loss": 0.2596,
      "num_input_tokens_seen": 19687720,
      "step": 25815
    },
    {
      "epoch": 53.67983367983368,
      "grad_norm": 0.00015227448602672666,
      "learning_rate": 0.0838081849469421,
      "loss": 0.2871,
      "num_input_tokens_seen": 19691560,
      "step": 25820
    },
    {
      "epoch": 53.69022869022869,
      "grad_norm": 0.0002926836023107171,
      "learning_rate": 0.08375533059150281,
      "loss": 0.2692,
      "num_input_tokens_seen": 19695272,
      "step": 25825
    },
    {
      "epoch": 53.7006237006237,
      "grad_norm": 0.00046705154818482697,
      "learning_rate": 0.08370248645182406,
      "loss": 0.2599,
      "num_input_tokens_seen": 19699176,
      "step": 25830
    },
    {
      "epoch": 53.71101871101871,
      "grad_norm": 0.0003347960882820189,
      "learning_rate": 0.083649652536055,
      "loss": 0.2571,
      "num_input_tokens_seen": 19702824,
      "step": 25835
    },
    {
      "epoch": 53.72141372141372,
      "grad_norm": 0.00027674148441292346,
      "learning_rate": 0.08359682885234339,
      "loss": 0.2671,
      "num_input_tokens_seen": 19706728,
      "step": 25840
    },
    {
      "epoch": 53.731808731808734,
      "grad_norm": 0.0005954855587333441,
      "learning_rate": 0.08354401540883516,
      "loss": 0.2739,
      "num_input_tokens_seen": 19710568,
      "step": 25845
    },
    {
      "epoch": 53.74220374220374,
      "grad_norm": 9.156563464784995e-05,
      "learning_rate": 0.0834912122136749,
      "loss": 0.2569,
      "num_input_tokens_seen": 19714440,
      "step": 25850
    },
    {
      "epoch": 53.75259875259875,
      "grad_norm": 6.419842247851193e-05,
      "learning_rate": 0.0834384192750056,
      "loss": 0.2856,
      "num_input_tokens_seen": 19718312,
      "step": 25855
    },
    {
      "epoch": 53.762993762993766,
      "grad_norm": 0.000195767599507235,
      "learning_rate": 0.08338563660096844,
      "loss": 0.2562,
      "num_input_tokens_seen": 19721992,
      "step": 25860
    },
    {
      "epoch": 53.773388773388774,
      "grad_norm": 0.0002074228395940736,
      "learning_rate": 0.08333286419970329,
      "loss": 0.2688,
      "num_input_tokens_seen": 19725800,
      "step": 25865
    },
    {
      "epoch": 53.78378378378378,
      "grad_norm": 0.0003609775740187615,
      "learning_rate": 0.08328010207934824,
      "loss": 0.2678,
      "num_input_tokens_seen": 19729672,
      "step": 25870
    },
    {
      "epoch": 53.79417879417879,
      "grad_norm": 0.00026167414034716785,
      "learning_rate": 0.08322735024803989,
      "loss": 0.281,
      "num_input_tokens_seen": 19733480,
      "step": 25875
    },
    {
      "epoch": 53.804573804573806,
      "grad_norm": 0.00011447145516285673,
      "learning_rate": 0.08317460871391331,
      "loss": 0.2796,
      "num_input_tokens_seen": 19737320,
      "step": 25880
    },
    {
      "epoch": 53.814968814968815,
      "grad_norm": 0.00035589313483797014,
      "learning_rate": 0.08312187748510179,
      "loss": 0.2662,
      "num_input_tokens_seen": 19741160,
      "step": 25885
    },
    {
      "epoch": 53.82536382536382,
      "grad_norm": 0.0003252648457419127,
      "learning_rate": 0.08306915656973726,
      "loss": 0.2683,
      "num_input_tokens_seen": 19745032,
      "step": 25890
    },
    {
      "epoch": 53.83575883575884,
      "grad_norm": 0.0002471132611390203,
      "learning_rate": 0.08301644597594988,
      "loss": 0.2544,
      "num_input_tokens_seen": 19748776,
      "step": 25895
    },
    {
      "epoch": 53.84615384615385,
      "grad_norm": 0.00019129975407849997,
      "learning_rate": 0.08296374571186826,
      "loss": 0.2667,
      "num_input_tokens_seen": 19752648,
      "step": 25900
    },
    {
      "epoch": 53.856548856548855,
      "grad_norm": 7.480049680452794e-05,
      "learning_rate": 0.08291105578561955,
      "loss": 0.2548,
      "num_input_tokens_seen": 19756488,
      "step": 25905
    },
    {
      "epoch": 53.86694386694387,
      "grad_norm": 0.0005411587189882994,
      "learning_rate": 0.08285837620532904,
      "loss": 0.2989,
      "num_input_tokens_seen": 19760232,
      "step": 25910
    },
    {
      "epoch": 53.87733887733888,
      "grad_norm": 0.0002652540570124984,
      "learning_rate": 0.0828057069791207,
      "loss": 0.2668,
      "num_input_tokens_seen": 19763944,
      "step": 25915
    },
    {
      "epoch": 53.88773388773389,
      "grad_norm": 0.0003816888201981783,
      "learning_rate": 0.0827530481151168,
      "loss": 0.2575,
      "num_input_tokens_seen": 19767848,
      "step": 25920
    },
    {
      "epoch": 53.898128898128896,
      "grad_norm": 4.4268763303989545e-05,
      "learning_rate": 0.08270039962143792,
      "loss": 0.2658,
      "num_input_tokens_seen": 19771688,
      "step": 25925
    },
    {
      "epoch": 53.90852390852391,
      "grad_norm": 0.00022834762057755142,
      "learning_rate": 0.08264776150620314,
      "loss": 0.2639,
      "num_input_tokens_seen": 19775368,
      "step": 25930
    },
    {
      "epoch": 53.91891891891892,
      "grad_norm": 0.00016352102102246135,
      "learning_rate": 0.08259513377753,
      "loss": 0.2876,
      "num_input_tokens_seen": 19779080,
      "step": 25935
    },
    {
      "epoch": 53.92931392931393,
      "grad_norm": 7.525904948124662e-05,
      "learning_rate": 0.08254251644353423,
      "loss": 0.2647,
      "num_input_tokens_seen": 19782888,
      "step": 25940
    },
    {
      "epoch": 53.93970893970894,
      "grad_norm": 0.0002536469546612352,
      "learning_rate": 0.08248990951233022,
      "loss": 0.2605,
      "num_input_tokens_seen": 19786760,
      "step": 25945
    },
    {
      "epoch": 53.95010395010395,
      "grad_norm": 0.0002702054916881025,
      "learning_rate": 0.08243731299203048,
      "loss": 0.2754,
      "num_input_tokens_seen": 19790568,
      "step": 25950
    },
    {
      "epoch": 53.96049896049896,
      "grad_norm": 0.00038639953709207475,
      "learning_rate": 0.08238472689074612,
      "loss": 0.2575,
      "num_input_tokens_seen": 19794344,
      "step": 25955
    },
    {
      "epoch": 53.97089397089397,
      "grad_norm": 9.961640171241015e-05,
      "learning_rate": 0.08233215121658666,
      "loss": 0.2543,
      "num_input_tokens_seen": 19798248,
      "step": 25960
    },
    {
      "epoch": 53.981288981288984,
      "grad_norm": 0.00011329437984386459,
      "learning_rate": 0.08227958597765982,
      "loss": 0.2546,
      "num_input_tokens_seen": 19802184,
      "step": 25965
    },
    {
      "epoch": 53.99168399168399,
      "grad_norm": 0.0001849041727837175,
      "learning_rate": 0.08222703118207181,
      "loss": 0.2763,
      "num_input_tokens_seen": 19805960,
      "step": 25970
    },
    {
      "epoch": 54.002079002079,
      "grad_norm": 6.556616426678374e-05,
      "learning_rate": 0.08217448683792734,
      "loss": 0.2778,
      "num_input_tokens_seen": 19809560,
      "step": 25975
    },
    {
      "epoch": 54.012474012474016,
      "grad_norm": 0.00011560157872736454,
      "learning_rate": 0.08212195295332926,
      "loss": 0.2652,
      "num_input_tokens_seen": 19813400,
      "step": 25980
    },
    {
      "epoch": 54.022869022869024,
      "grad_norm": 0.0004670706985052675,
      "learning_rate": 0.08206942953637915,
      "loss": 0.2491,
      "num_input_tokens_seen": 19817304,
      "step": 25985
    },
    {
      "epoch": 54.03326403326403,
      "grad_norm": 0.0004275247920304537,
      "learning_rate": 0.08201691659517658,
      "loss": 0.2453,
      "num_input_tokens_seen": 19821208,
      "step": 25990
    },
    {
      "epoch": 54.04365904365904,
      "grad_norm": 0.00040047045331448317,
      "learning_rate": 0.08196441413781981,
      "loss": 0.2542,
      "num_input_tokens_seen": 19824760,
      "step": 25995
    },
    {
      "epoch": 54.054054054054056,
      "grad_norm": 0.0002530265774112195,
      "learning_rate": 0.08191192217240544,
      "loss": 0.2705,
      "num_input_tokens_seen": 19828504,
      "step": 26000
    },
    {
      "epoch": 54.054054054054056,
      "eval_loss": 0.24692478775978088,
      "eval_runtime": 13.3784,
      "eval_samples_per_second": 63.984,
      "eval_steps_per_second": 15.996,
      "num_input_tokens_seen": 19828504,
      "step": 26000
    },
    {
      "epoch": 54.064449064449065,
      "grad_norm": 0.0002602216845843941,
      "learning_rate": 0.08185944070702823,
      "loss": 0.2624,
      "num_input_tokens_seen": 19832376,
      "step": 26005
    },
    {
      "epoch": 54.07484407484407,
      "grad_norm": 8.60114159877412e-05,
      "learning_rate": 0.08180696974978159,
      "loss": 0.2496,
      "num_input_tokens_seen": 19836184,
      "step": 26010
    },
    {
      "epoch": 54.08523908523909,
      "grad_norm": 0.00011810174328275025,
      "learning_rate": 0.08175450930875724,
      "loss": 0.2486,
      "num_input_tokens_seen": 19839992,
      "step": 26015
    },
    {
      "epoch": 54.0956340956341,
      "grad_norm": 0.00021255902538541704,
      "learning_rate": 0.08170205939204513,
      "loss": 0.2459,
      "num_input_tokens_seen": 19843736,
      "step": 26020
    },
    {
      "epoch": 54.106029106029105,
      "grad_norm": 0.0003656491171568632,
      "learning_rate": 0.08164962000773379,
      "loss": 0.2664,
      "num_input_tokens_seen": 19847640,
      "step": 26025
    },
    {
      "epoch": 54.11642411642411,
      "grad_norm": 6.77032585372217e-05,
      "learning_rate": 0.08159719116390995,
      "loss": 0.2793,
      "num_input_tokens_seen": 19851416,
      "step": 26030
    },
    {
      "epoch": 54.12681912681913,
      "grad_norm": 0.0002658611338119954,
      "learning_rate": 0.08154477286865887,
      "loss": 0.2725,
      "num_input_tokens_seen": 19855320,
      "step": 26035
    },
    {
      "epoch": 54.13721413721414,
      "grad_norm": 4.501608054852113e-05,
      "learning_rate": 0.08149236513006404,
      "loss": 0.2573,
      "num_input_tokens_seen": 19859224,
      "step": 26040
    },
    {
      "epoch": 54.147609147609145,
      "grad_norm": 0.00012734455231111497,
      "learning_rate": 0.08143996795620746,
      "loss": 0.2729,
      "num_input_tokens_seen": 19863096,
      "step": 26045
    },
    {
      "epoch": 54.15800415800416,
      "grad_norm": 0.00014130730414763093,
      "learning_rate": 0.08138758135516938,
      "loss": 0.2712,
      "num_input_tokens_seen": 19866776,
      "step": 26050
    },
    {
      "epoch": 54.16839916839917,
      "grad_norm": 0.00011177197302458808,
      "learning_rate": 0.08133520533502851,
      "loss": 0.27,
      "num_input_tokens_seen": 19870584,
      "step": 26055
    },
    {
      "epoch": 54.17879417879418,
      "grad_norm": 0.00037128428812138736,
      "learning_rate": 0.08128283990386184,
      "loss": 0.2837,
      "num_input_tokens_seen": 19874616,
      "step": 26060
    },
    {
      "epoch": 54.189189189189186,
      "grad_norm": 0.0002411420427961275,
      "learning_rate": 0.08123048506974488,
      "loss": 0.2656,
      "num_input_tokens_seen": 19878488,
      "step": 26065
    },
    {
      "epoch": 54.1995841995842,
      "grad_norm": 5.852436152053997e-05,
      "learning_rate": 0.08117814084075124,
      "loss": 0.26,
      "num_input_tokens_seen": 19882232,
      "step": 26070
    },
    {
      "epoch": 54.20997920997921,
      "grad_norm": 0.0003427154733799398,
      "learning_rate": 0.08112580722495318,
      "loss": 0.272,
      "num_input_tokens_seen": 19885944,
      "step": 26075
    },
    {
      "epoch": 54.22037422037422,
      "grad_norm": 0.00025025615468621254,
      "learning_rate": 0.08107348423042122,
      "loss": 0.2559,
      "num_input_tokens_seen": 19889752,
      "step": 26080
    },
    {
      "epoch": 54.23076923076923,
      "grad_norm": 0.00014690423267893493,
      "learning_rate": 0.08102117186522413,
      "loss": 0.2683,
      "num_input_tokens_seen": 19893560,
      "step": 26085
    },
    {
      "epoch": 54.24116424116424,
      "grad_norm": 0.0005931174382567406,
      "learning_rate": 0.08096887013742916,
      "loss": 0.2614,
      "num_input_tokens_seen": 19897432,
      "step": 26090
    },
    {
      "epoch": 54.25155925155925,
      "grad_norm": 7.394458225462586e-05,
      "learning_rate": 0.08091657905510198,
      "loss": 0.252,
      "num_input_tokens_seen": 19901080,
      "step": 26095
    },
    {
      "epoch": 54.26195426195426,
      "grad_norm": 0.0002644641208462417,
      "learning_rate": 0.08086429862630642,
      "loss": 0.2758,
      "num_input_tokens_seen": 19904760,
      "step": 26100
    },
    {
      "epoch": 54.272349272349274,
      "grad_norm": 0.00016311694344040006,
      "learning_rate": 0.08081202885910488,
      "loss": 0.2758,
      "num_input_tokens_seen": 19908408,
      "step": 26105
    },
    {
      "epoch": 54.28274428274428,
      "grad_norm": 0.00023390664136968553,
      "learning_rate": 0.08075976976155795,
      "loss": 0.2643,
      "num_input_tokens_seen": 19912376,
      "step": 26110
    },
    {
      "epoch": 54.29313929313929,
      "grad_norm": 0.00015235489991027862,
      "learning_rate": 0.08070752134172461,
      "loss": 0.249,
      "num_input_tokens_seen": 19916344,
      "step": 26115
    },
    {
      "epoch": 54.303534303534306,
      "grad_norm": 0.00010802914039231837,
      "learning_rate": 0.08065528360766229,
      "loss": 0.2594,
      "num_input_tokens_seen": 19920088,
      "step": 26120
    },
    {
      "epoch": 54.313929313929314,
      "grad_norm": 0.00020670176309067756,
      "learning_rate": 0.08060305656742664,
      "loss": 0.2683,
      "num_input_tokens_seen": 19923832,
      "step": 26125
    },
    {
      "epoch": 54.32432432432432,
      "grad_norm": 0.00014302424096968025,
      "learning_rate": 0.08055084022907182,
      "loss": 0.2989,
      "num_input_tokens_seen": 19927544,
      "step": 26130
    },
    {
      "epoch": 54.33471933471934,
      "grad_norm": 0.00019747289479710162,
      "learning_rate": 0.08049863460065014,
      "loss": 0.2665,
      "num_input_tokens_seen": 19931480,
      "step": 26135
    },
    {
      "epoch": 54.34511434511435,
      "grad_norm": 0.0002862780529540032,
      "learning_rate": 0.0804464396902124,
      "loss": 0.2799,
      "num_input_tokens_seen": 19935288,
      "step": 26140
    },
    {
      "epoch": 54.355509355509355,
      "grad_norm": 0.00042460119584575295,
      "learning_rate": 0.08039425550580777,
      "loss": 0.2616,
      "num_input_tokens_seen": 19939128,
      "step": 26145
    },
    {
      "epoch": 54.36590436590436,
      "grad_norm": 0.00024318067880813032,
      "learning_rate": 0.08034208205548363,
      "loss": 0.2836,
      "num_input_tokens_seen": 19943000,
      "step": 26150
    },
    {
      "epoch": 54.37629937629938,
      "grad_norm": 0.00029690409428440034,
      "learning_rate": 0.08028991934728581,
      "loss": 0.2662,
      "num_input_tokens_seen": 19946904,
      "step": 26155
    },
    {
      "epoch": 54.38669438669439,
      "grad_norm": 0.0005558753618970513,
      "learning_rate": 0.0802377673892585,
      "loss": 0.2582,
      "num_input_tokens_seen": 19950712,
      "step": 26160
    },
    {
      "epoch": 54.397089397089395,
      "grad_norm": 0.0004594610072672367,
      "learning_rate": 0.0801856261894441,
      "loss": 0.2712,
      "num_input_tokens_seen": 19954456,
      "step": 26165
    },
    {
      "epoch": 54.40748440748441,
      "grad_norm": 0.00047014711890369654,
      "learning_rate": 0.08013349575588354,
      "loss": 0.2929,
      "num_input_tokens_seen": 19958296,
      "step": 26170
    },
    {
      "epoch": 54.41787941787942,
      "grad_norm": 7.162798283388838e-05,
      "learning_rate": 0.08008137609661586,
      "loss": 0.2642,
      "num_input_tokens_seen": 19961976,
      "step": 26175
    },
    {
      "epoch": 54.42827442827443,
      "grad_norm": 6.433594535337761e-05,
      "learning_rate": 0.08002926721967872,
      "loss": 0.2598,
      "num_input_tokens_seen": 19965720,
      "step": 26180
    },
    {
      "epoch": 54.438669438669436,
      "grad_norm": 0.00031317290267907083,
      "learning_rate": 0.07997716913310782,
      "loss": 0.2704,
      "num_input_tokens_seen": 19969560,
      "step": 26185
    },
    {
      "epoch": 54.44906444906445,
      "grad_norm": 0.00010351963283028454,
      "learning_rate": 0.07992508184493745,
      "loss": 0.2582,
      "num_input_tokens_seen": 19973336,
      "step": 26190
    },
    {
      "epoch": 54.45945945945946,
      "grad_norm": 8.793236338533461e-05,
      "learning_rate": 0.07987300536320001,
      "loss": 0.2564,
      "num_input_tokens_seen": 19977144,
      "step": 26195
    },
    {
      "epoch": 54.46985446985447,
      "grad_norm": 8.558998524677008e-05,
      "learning_rate": 0.07982093969592649,
      "loss": 0.2621,
      "num_input_tokens_seen": 19980856,
      "step": 26200
    },
    {
      "epoch": 54.46985446985447,
      "eval_loss": 0.24724189937114716,
      "eval_runtime": 13.3945,
      "eval_samples_per_second": 63.907,
      "eval_steps_per_second": 15.977,
      "num_input_tokens_seen": 19980856,
      "step": 26200
    },
    {
      "epoch": 54.48024948024948,
      "grad_norm": 0.00011386102414689958,
      "learning_rate": 0.07976888485114592,
      "loss": 0.2514,
      "num_input_tokens_seen": 19984696,
      "step": 26205
    },
    {
      "epoch": 54.49064449064449,
      "grad_norm": 0.0002468022285029292,
      "learning_rate": 0.07971684083688595,
      "loss": 0.2814,
      "num_input_tokens_seen": 19988376,
      "step": 26210
    },
    {
      "epoch": 54.5010395010395,
      "grad_norm": 0.00031022605253383517,
      "learning_rate": 0.0796648076611723,
      "loss": 0.2716,
      "num_input_tokens_seen": 19992152,
      "step": 26215
    },
    {
      "epoch": 54.51143451143451,
      "grad_norm": 0.00019124281243421137,
      "learning_rate": 0.07961278533202922,
      "loss": 0.2889,
      "num_input_tokens_seen": 19996152,
      "step": 26220
    },
    {
      "epoch": 54.521829521829524,
      "grad_norm": 0.00041830085683614016,
      "learning_rate": 0.07956077385747919,
      "loss": 0.2683,
      "num_input_tokens_seen": 19999960,
      "step": 26225
    },
    {
      "epoch": 54.53222453222453,
      "grad_norm": 0.00018451534560881555,
      "learning_rate": 0.079508773245543,
      "loss": 0.2732,
      "num_input_tokens_seen": 20003640,
      "step": 26230
    },
    {
      "epoch": 54.54261954261954,
      "grad_norm": 0.000328708701999858,
      "learning_rate": 0.07945678350423982,
      "loss": 0.233,
      "num_input_tokens_seen": 20007320,
      "step": 26235
    },
    {
      "epoch": 54.553014553014556,
      "grad_norm": 0.0002678821620065719,
      "learning_rate": 0.07940480464158717,
      "loss": 0.251,
      "num_input_tokens_seen": 20011128,
      "step": 26240
    },
    {
      "epoch": 54.563409563409564,
      "grad_norm": 0.00021444098092615604,
      "learning_rate": 0.07935283666560076,
      "loss": 0.2235,
      "num_input_tokens_seen": 20014872,
      "step": 26245
    },
    {
      "epoch": 54.57380457380457,
      "grad_norm": 0.00020633151871152222,
      "learning_rate": 0.07930087958429478,
      "loss": 0.3135,
      "num_input_tokens_seen": 20018680,
      "step": 26250
    },
    {
      "epoch": 54.58419958419958,
      "grad_norm": 0.0004138337098993361,
      "learning_rate": 0.07924893340568159,
      "loss": 0.2635,
      "num_input_tokens_seen": 20022584,
      "step": 26255
    },
    {
      "epoch": 54.5945945945946,
      "grad_norm": 0.00011721346527338028,
      "learning_rate": 0.07919699813777205,
      "loss": 0.2345,
      "num_input_tokens_seen": 20026328,
      "step": 26260
    },
    {
      "epoch": 54.604989604989605,
      "grad_norm": 0.00021726837439928204,
      "learning_rate": 0.07914507378857515,
      "loss": 0.2712,
      "num_input_tokens_seen": 20030040,
      "step": 26265
    },
    {
      "epoch": 54.61538461538461,
      "grad_norm": 0.00032105721766129136,
      "learning_rate": 0.07909316036609822,
      "loss": 0.2868,
      "num_input_tokens_seen": 20033880,
      "step": 26270
    },
    {
      "epoch": 54.62577962577963,
      "grad_norm": 0.00022261105186771601,
      "learning_rate": 0.07904125787834704,
      "loss": 0.2515,
      "num_input_tokens_seen": 20037752,
      "step": 26275
    },
    {
      "epoch": 54.63617463617464,
      "grad_norm": 0.0006170336273498833,
      "learning_rate": 0.07898936633332569,
      "loss": 0.2522,
      "num_input_tokens_seen": 20041560,
      "step": 26280
    },
    {
      "epoch": 54.646569646569645,
      "grad_norm": 0.00020674700499512255,
      "learning_rate": 0.07893748573903635,
      "loss": 0.2542,
      "num_input_tokens_seen": 20045368,
      "step": 26285
    },
    {
      "epoch": 54.656964656964654,
      "grad_norm": 0.00014733245188836008,
      "learning_rate": 0.0788856161034798,
      "loss": 0.2637,
      "num_input_tokens_seen": 20049144,
      "step": 26290
    },
    {
      "epoch": 54.66735966735967,
      "grad_norm": 0.00030363118276000023,
      "learning_rate": 0.07883375743465487,
      "loss": 0.2671,
      "num_input_tokens_seen": 20052920,
      "step": 26295
    },
    {
      "epoch": 54.67775467775468,
      "grad_norm": 0.0002180361479986459,
      "learning_rate": 0.07878190974055888,
      "loss": 0.2612,
      "num_input_tokens_seen": 20056760,
      "step": 26300
    },
    {
      "epoch": 54.688149688149686,
      "grad_norm": 0.0007007686654105783,
      "learning_rate": 0.07873007302918746,
      "loss": 0.2899,
      "num_input_tokens_seen": 20060792,
      "step": 26305
    },
    {
      "epoch": 54.6985446985447,
      "grad_norm": 0.00017218109860550612,
      "learning_rate": 0.07867824730853433,
      "loss": 0.2701,
      "num_input_tokens_seen": 20064408,
      "step": 26310
    },
    {
      "epoch": 54.70893970893971,
      "grad_norm": 0.00036265526432543993,
      "learning_rate": 0.07862643258659176,
      "loss": 0.2664,
      "num_input_tokens_seen": 20068120,
      "step": 26315
    },
    {
      "epoch": 54.71933471933472,
      "grad_norm": 0.0004432791320141405,
      "learning_rate": 0.07857462887135026,
      "loss": 0.2659,
      "num_input_tokens_seen": 20072184,
      "step": 26320
    },
    {
      "epoch": 54.729729729729726,
      "grad_norm": 0.0004398549208417535,
      "learning_rate": 0.0785228361707986,
      "loss": 0.2566,
      "num_input_tokens_seen": 20076152,
      "step": 26325
    },
    {
      "epoch": 54.74012474012474,
      "grad_norm": 0.00013934337766841054,
      "learning_rate": 0.07847105449292378,
      "loss": 0.2593,
      "num_input_tokens_seen": 20080152,
      "step": 26330
    },
    {
      "epoch": 54.75051975051975,
      "grad_norm": 0.00025779425050131977,
      "learning_rate": 0.0784192838457113,
      "loss": 0.2786,
      "num_input_tokens_seen": 20083928,
      "step": 26335
    },
    {
      "epoch": 54.76091476091476,
      "grad_norm": 0.00022268503380473703,
      "learning_rate": 0.07836752423714473,
      "loss": 0.2575,
      "num_input_tokens_seen": 20087736,
      "step": 26340
    },
    {
      "epoch": 54.771309771309774,
      "grad_norm": 0.0004874719015788287,
      "learning_rate": 0.07831577567520616,
      "loss": 0.2684,
      "num_input_tokens_seen": 20091672,
      "step": 26345
    },
    {
      "epoch": 54.78170478170478,
      "grad_norm": 0.0002922238491009921,
      "learning_rate": 0.07826403816787579,
      "loss": 0.2836,
      "num_input_tokens_seen": 20095576,
      "step": 26350
    },
    {
      "epoch": 54.79209979209979,
      "grad_norm": 0.00010126736742677167,
      "learning_rate": 0.0782123117231322,
      "loss": 0.2569,
      "num_input_tokens_seen": 20099384,
      "step": 26355
    },
    {
      "epoch": 54.802494802494806,
      "grad_norm": 0.0008457657531835139,
      "learning_rate": 0.07816059634895237,
      "loss": 0.2683,
      "num_input_tokens_seen": 20103096,
      "step": 26360
    },
    {
      "epoch": 54.812889812889814,
      "grad_norm": 0.00020537260570563376,
      "learning_rate": 0.0781088920533113,
      "loss": 0.2576,
      "num_input_tokens_seen": 20106840,
      "step": 26365
    },
    {
      "epoch": 54.82328482328482,
      "grad_norm": 0.00010219080286333337,
      "learning_rate": 0.07805719884418257,
      "loss": 0.2576,
      "num_input_tokens_seen": 20110648,
      "step": 26370
    },
    {
      "epoch": 54.83367983367983,
      "grad_norm": 0.00036513133090920746,
      "learning_rate": 0.07800551672953779,
      "loss": 0.2562,
      "num_input_tokens_seen": 20114520,
      "step": 26375
    },
    {
      "epoch": 54.84407484407485,
      "grad_norm": 0.00025904891663230956,
      "learning_rate": 0.07795384571734709,
      "loss": 0.2615,
      "num_input_tokens_seen": 20118264,
      "step": 26380
    },
    {
      "epoch": 54.854469854469855,
      "grad_norm": 0.0003789609472732991,
      "learning_rate": 0.07790218581557883,
      "loss": 0.2625,
      "num_input_tokens_seen": 20122264,
      "step": 26385
    },
    {
      "epoch": 54.86486486486486,
      "grad_norm": 0.00014257957809604704,
      "learning_rate": 0.07785053703219949,
      "loss": 0.2714,
      "num_input_tokens_seen": 20126168,
      "step": 26390
    },
    {
      "epoch": 54.87525987525988,
      "grad_norm": 0.00022269565670285374,
      "learning_rate": 0.07779889937517409,
      "loss": 0.2593,
      "num_input_tokens_seen": 20130040,
      "step": 26395
    },
    {
      "epoch": 54.88565488565489,
      "grad_norm": 0.00048069132026284933,
      "learning_rate": 0.0777472728524657,
      "loss": 0.263,
      "num_input_tokens_seen": 20133784,
      "step": 26400
    },
    {
      "epoch": 54.88565488565489,
      "eval_loss": 0.2537725865840912,
      "eval_runtime": 13.3917,
      "eval_samples_per_second": 63.92,
      "eval_steps_per_second": 15.98,
      "num_input_tokens_seen": 20133784,
      "step": 26400
    },
    {
      "epoch": 54.896049896049895,
      "grad_norm": 0.00019268585310783237,
      "learning_rate": 0.07769565747203584,
      "loss": 0.2442,
      "num_input_tokens_seen": 20137624,
      "step": 26405
    },
    {
      "epoch": 54.906444906444904,
      "grad_norm": 0.0003926725185010582,
      "learning_rate": 0.07764405324184427,
      "loss": 0.2695,
      "num_input_tokens_seen": 20141368,
      "step": 26410
    },
    {
      "epoch": 54.91683991683992,
      "grad_norm": 0.0005147213232703507,
      "learning_rate": 0.07759246016984889,
      "loss": 0.2808,
      "num_input_tokens_seen": 20145336,
      "step": 26415
    },
    {
      "epoch": 54.92723492723493,
      "grad_norm": 0.00014873924374114722,
      "learning_rate": 0.07754087826400609,
      "loss": 0.278,
      "num_input_tokens_seen": 20149304,
      "step": 26420
    },
    {
      "epoch": 54.937629937629936,
      "grad_norm": 0.00027317943749949336,
      "learning_rate": 0.0774893075322705,
      "loss": 0.2538,
      "num_input_tokens_seen": 20152888,
      "step": 26425
    },
    {
      "epoch": 54.94802494802495,
      "grad_norm": 0.00012338340457063168,
      "learning_rate": 0.07743774798259484,
      "loss": 0.2552,
      "num_input_tokens_seen": 20156568,
      "step": 26430
    },
    {
      "epoch": 54.95841995841996,
      "grad_norm": 0.000618341495282948,
      "learning_rate": 0.07738619962293032,
      "loss": 0.264,
      "num_input_tokens_seen": 20160376,
      "step": 26435
    },
    {
      "epoch": 54.96881496881497,
      "grad_norm": 0.0002620588056743145,
      "learning_rate": 0.0773346624612264,
      "loss": 0.2899,
      "num_input_tokens_seen": 20164248,
      "step": 26440
    },
    {
      "epoch": 54.979209979209976,
      "grad_norm": 0.00026451688609085977,
      "learning_rate": 0.07728313650543066,
      "loss": 0.2411,
      "num_input_tokens_seen": 20168088,
      "step": 26445
    },
    {
      "epoch": 54.98960498960499,
      "grad_norm": 0.00017170059436466545,
      "learning_rate": 0.07723162176348913,
      "loss": 0.2803,
      "num_input_tokens_seen": 20171864,
      "step": 26450
    },
    {
      "epoch": 55.0,
      "grad_norm": 9.429289639228955e-05,
      "learning_rate": 0.07718011824334593,
      "loss": 0.269,
      "num_input_tokens_seen": 20175464,
      "step": 26455
    },
    {
      "epoch": 55.01039501039501,
      "grad_norm": 0.00012367004819680005,
      "learning_rate": 0.07712862595294363,
      "loss": 0.2558,
      "num_input_tokens_seen": 20179496,
      "step": 26460
    },
    {
      "epoch": 55.020790020790024,
      "grad_norm": 0.0006052608950994909,
      "learning_rate": 0.07707714490022301,
      "loss": 0.2508,
      "num_input_tokens_seen": 20183272,
      "step": 26465
    },
    {
      "epoch": 55.03118503118503,
      "grad_norm": 0.0002582585730124265,
      "learning_rate": 0.07702567509312298,
      "loss": 0.2645,
      "num_input_tokens_seen": 20187112,
      "step": 26470
    },
    {
      "epoch": 55.04158004158004,
      "grad_norm": 6.263565592234954e-05,
      "learning_rate": 0.07697421653958098,
      "loss": 0.2412,
      "num_input_tokens_seen": 20190952,
      "step": 26475
    },
    {
      "epoch": 55.05197505197505,
      "grad_norm": 0.00016367787611670792,
      "learning_rate": 0.07692276924753247,
      "loss": 0.2632,
      "num_input_tokens_seen": 20194792,
      "step": 26480
    },
    {
      "epoch": 55.062370062370064,
      "grad_norm": 0.0009348075254820287,
      "learning_rate": 0.07687133322491124,
      "loss": 0.269,
      "num_input_tokens_seen": 20198600,
      "step": 26485
    },
    {
      "epoch": 55.07276507276507,
      "grad_norm": 0.00048672521370463073,
      "learning_rate": 0.07681990847964948,
      "loss": 0.2336,
      "num_input_tokens_seen": 20202312,
      "step": 26490
    },
    {
      "epoch": 55.08316008316008,
      "grad_norm": 0.00015556183643639088,
      "learning_rate": 0.0767684950196774,
      "loss": 0.2398,
      "num_input_tokens_seen": 20206088,
      "step": 26495
    },
    {
      "epoch": 55.093555093555096,
      "grad_norm": 0.0006498659495264292,
      "learning_rate": 0.0767170928529237,
      "loss": 0.255,
      "num_input_tokens_seen": 20210024,
      "step": 26500
    },
    {
      "epoch": 55.103950103950105,
      "grad_norm": 0.000219001027289778,
      "learning_rate": 0.07666570198731526,
      "loss": 0.2469,
      "num_input_tokens_seen": 20213704,
      "step": 26505
    },
    {
      "epoch": 55.11434511434511,
      "grad_norm": 6.57796481391415e-05,
      "learning_rate": 0.07661432243077708,
      "loss": 0.2415,
      "num_input_tokens_seen": 20217384,
      "step": 26510
    },
    {
      "epoch": 55.12474012474012,
      "grad_norm": 5.6014323490671813e-05,
      "learning_rate": 0.0765629541912326,
      "loss": 0.2599,
      "num_input_tokens_seen": 20221064,
      "step": 26515
    },
    {
      "epoch": 55.13513513513514,
      "grad_norm": 6.651139847235754e-05,
      "learning_rate": 0.07651159727660352,
      "loss": 0.2588,
      "num_input_tokens_seen": 20224712,
      "step": 26520
    },
    {
      "epoch": 55.145530145530145,
      "grad_norm": 0.0002601959276944399,
      "learning_rate": 0.07646025169480959,
      "loss": 0.2716,
      "num_input_tokens_seen": 20228488,
      "step": 26525
    },
    {
      "epoch": 55.15592515592515,
      "grad_norm": 0.00014527332677971572,
      "learning_rate": 0.07640891745376908,
      "loss": 0.2596,
      "num_input_tokens_seen": 20232456,
      "step": 26530
    },
    {
      "epoch": 55.16632016632017,
      "grad_norm": 0.0006158179603517056,
      "learning_rate": 0.07635759456139822,
      "loss": 0.2517,
      "num_input_tokens_seen": 20236232,
      "step": 26535
    },
    {
      "epoch": 55.17671517671518,
      "grad_norm": 8.398152567679062e-05,
      "learning_rate": 0.0763062830256118,
      "loss": 0.2555,
      "num_input_tokens_seen": 20240040,
      "step": 26540
    },
    {
      "epoch": 55.187110187110186,
      "grad_norm": 0.00020210351794958115,
      "learning_rate": 0.07625498285432258,
      "loss": 0.289,
      "num_input_tokens_seen": 20243944,
      "step": 26545
    },
    {
      "epoch": 55.197505197505194,
      "grad_norm": 0.0005895509384572506,
      "learning_rate": 0.07620369405544176,
      "loss": 0.2764,
      "num_input_tokens_seen": 20248040,
      "step": 26550
    },
    {
      "epoch": 55.20790020790021,
      "grad_norm": 0.0005683311610482633,
      "learning_rate": 0.07615241663687868,
      "loss": 0.2808,
      "num_input_tokens_seen": 20251976,
      "step": 26555
    },
    {
      "epoch": 55.21829521829522,
      "grad_norm": 9.496301208855584e-05,
      "learning_rate": 0.07610115060654106,
      "loss": 0.2644,
      "num_input_tokens_seen": 20255912,
      "step": 26560
    },
    {
      "epoch": 55.228690228690226,
      "grad_norm": 9.120270260609686e-05,
      "learning_rate": 0.07604989597233458,
      "loss": 0.252,
      "num_input_tokens_seen": 20259688,
      "step": 26565
    },
    {
      "epoch": 55.23908523908524,
      "grad_norm": 0.0005651332903653383,
      "learning_rate": 0.07599865274216352,
      "loss": 0.2415,
      "num_input_tokens_seen": 20263400,
      "step": 26570
    },
    {
      "epoch": 55.24948024948025,
      "grad_norm": 0.0010400756727904081,
      "learning_rate": 0.07594742092393013,
      "loss": 0.3186,
      "num_input_tokens_seen": 20267080,
      "step": 26575
    },
    {
      "epoch": 55.25987525987526,
      "grad_norm": 0.0003462302847765386,
      "learning_rate": 0.07589620052553503,
      "loss": 0.2503,
      "num_input_tokens_seen": 20270760,
      "step": 26580
    },
    {
      "epoch": 55.270270270270274,
      "grad_norm": 0.00023988127941265702,
      "learning_rate": 0.0758449915548771,
      "loss": 0.2829,
      "num_input_tokens_seen": 20274472,
      "step": 26585
    },
    {
      "epoch": 55.28066528066528,
      "grad_norm": 0.00013170663441997021,
      "learning_rate": 0.07579379401985332,
      "loss": 0.2803,
      "num_input_tokens_seen": 20278344,
      "step": 26590
    },
    {
      "epoch": 55.29106029106029,
      "grad_norm": 0.00028670215397141874,
      "learning_rate": 0.07574260792835905,
      "loss": 0.2687,
      "num_input_tokens_seen": 20282344,
      "step": 26595
    },
    {
      "epoch": 55.3014553014553,
      "grad_norm": 5.830689406138845e-05,
      "learning_rate": 0.07569143328828784,
      "loss": 0.273,
      "num_input_tokens_seen": 20286120,
      "step": 26600
    },
    {
      "epoch": 55.3014553014553,
      "eval_loss": 0.2565113306045532,
      "eval_runtime": 13.3937,
      "eval_samples_per_second": 63.91,
      "eval_steps_per_second": 15.978,
      "num_input_tokens_seen": 20286120,
      "step": 26600
    },
    {
      "epoch": 55.311850311850314,
      "grad_norm": 4.851643825531937e-05,
      "learning_rate": 0.0756402701075314,
      "loss": 0.2673,
      "num_input_tokens_seen": 20289992,
      "step": 26605
    },
    {
      "epoch": 55.32224532224532,
      "grad_norm": 0.0002664014755282551,
      "learning_rate": 0.07558911839397982,
      "loss": 0.2862,
      "num_input_tokens_seen": 20293800,
      "step": 26610
    },
    {
      "epoch": 55.33264033264033,
      "grad_norm": 0.0001437256287317723,
      "learning_rate": 0.07553797815552123,
      "loss": 0.2744,
      "num_input_tokens_seen": 20297608,
      "step": 26615
    },
    {
      "epoch": 55.343035343035346,
      "grad_norm": 0.00025565660325810313,
      "learning_rate": 0.07548684940004222,
      "loss": 0.2762,
      "num_input_tokens_seen": 20301416,
      "step": 26620
    },
    {
      "epoch": 55.353430353430355,
      "grad_norm": 0.0001771823881426826,
      "learning_rate": 0.07543573213542744,
      "loss": 0.262,
      "num_input_tokens_seen": 20305128,
      "step": 26625
    },
    {
      "epoch": 55.36382536382536,
      "grad_norm": 0.00010525508696446195,
      "learning_rate": 0.0753846263695597,
      "loss": 0.275,
      "num_input_tokens_seen": 20309000,
      "step": 26630
    },
    {
      "epoch": 55.37422037422037,
      "grad_norm": 0.0006201022188179195,
      "learning_rate": 0.07533353211032029,
      "loss": 0.2559,
      "num_input_tokens_seen": 20312776,
      "step": 26635
    },
    {
      "epoch": 55.38461538461539,
      "grad_norm": 7.772214303258806e-05,
      "learning_rate": 0.07528244936558857,
      "loss": 0.2662,
      "num_input_tokens_seen": 20316552,
      "step": 26640
    },
    {
      "epoch": 55.395010395010395,
      "grad_norm": 0.0005495261866599321,
      "learning_rate": 0.07523137814324206,
      "loss": 0.2699,
      "num_input_tokens_seen": 20320296,
      "step": 26645
    },
    {
      "epoch": 55.4054054054054,
      "grad_norm": 7.798885781085119e-05,
      "learning_rate": 0.07518031845115672,
      "loss": 0.2989,
      "num_input_tokens_seen": 20324232,
      "step": 26650
    },
    {
      "epoch": 55.41580041580042,
      "grad_norm": 8.477033406961709e-05,
      "learning_rate": 0.07512927029720647,
      "loss": 0.2782,
      "num_input_tokens_seen": 20327912,
      "step": 26655
    },
    {
      "epoch": 55.42619542619543,
      "grad_norm": 0.0002552078804001212,
      "learning_rate": 0.0750782336892636,
      "loss": 0.274,
      "num_input_tokens_seen": 20331688,
      "step": 26660
    },
    {
      "epoch": 55.436590436590436,
      "grad_norm": 0.0004291512304916978,
      "learning_rate": 0.0750272086351987,
      "loss": 0.2653,
      "num_input_tokens_seen": 20335432,
      "step": 26665
    },
    {
      "epoch": 55.446985446985444,
      "grad_norm": 0.000267739495029673,
      "learning_rate": 0.07497619514288031,
      "loss": 0.2799,
      "num_input_tokens_seen": 20339112,
      "step": 26670
    },
    {
      "epoch": 55.45738045738046,
      "grad_norm": 0.00014905829448252916,
      "learning_rate": 0.07492519322017545,
      "loss": 0.2807,
      "num_input_tokens_seen": 20342920,
      "step": 26675
    },
    {
      "epoch": 55.46777546777547,
      "grad_norm": 0.00010727380868047476,
      "learning_rate": 0.0748742028749493,
      "loss": 0.2651,
      "num_input_tokens_seen": 20346920,
      "step": 26680
    },
    {
      "epoch": 55.478170478170476,
      "grad_norm": 0.00040123899816535413,
      "learning_rate": 0.0748232241150651,
      "loss": 0.2419,
      "num_input_tokens_seen": 20350728,
      "step": 26685
    },
    {
      "epoch": 55.48856548856549,
      "grad_norm": 0.00024495512479916215,
      "learning_rate": 0.07477225694838453,
      "loss": 0.2704,
      "num_input_tokens_seen": 20354600,
      "step": 26690
    },
    {
      "epoch": 55.4989604989605,
      "grad_norm": 0.0003472850366961211,
      "learning_rate": 0.07472130138276731,
      "loss": 0.261,
      "num_input_tokens_seen": 20358536,
      "step": 26695
    },
    {
      "epoch": 55.50935550935551,
      "grad_norm": 6.669552385574207e-05,
      "learning_rate": 0.07467035742607138,
      "loss": 0.293,
      "num_input_tokens_seen": 20362376,
      "step": 26700
    },
    {
      "epoch": 55.51975051975052,
      "grad_norm": 0.00023361884814221412,
      "learning_rate": 0.07461942508615303,
      "loss": 0.2582,
      "num_input_tokens_seen": 20366280,
      "step": 26705
    },
    {
      "epoch": 55.53014553014553,
      "grad_norm": 0.00038345353095792234,
      "learning_rate": 0.07456850437086657,
      "loss": 0.2798,
      "num_input_tokens_seen": 20370184,
      "step": 26710
    },
    {
      "epoch": 55.54054054054054,
      "grad_norm": 0.0008303012582473457,
      "learning_rate": 0.07451759528806468,
      "loss": 0.2631,
      "num_input_tokens_seen": 20373960,
      "step": 26715
    },
    {
      "epoch": 55.55093555093555,
      "grad_norm": 0.00032848777482286096,
      "learning_rate": 0.0744666978455982,
      "loss": 0.2606,
      "num_input_tokens_seen": 20377800,
      "step": 26720
    },
    {
      "epoch": 55.561330561330564,
      "grad_norm": 0.0004507634730543941,
      "learning_rate": 0.07441581205131609,
      "loss": 0.2739,
      "num_input_tokens_seen": 20381608,
      "step": 26725
    },
    {
      "epoch": 55.57172557172557,
      "grad_norm": 0.0003685134288389236,
      "learning_rate": 0.07436493791306566,
      "loss": 0.2809,
      "num_input_tokens_seen": 20385352,
      "step": 26730
    },
    {
      "epoch": 55.58212058212058,
      "grad_norm": 6.376814417308196e-05,
      "learning_rate": 0.07431407543869223,
      "loss": 0.2255,
      "num_input_tokens_seen": 20389224,
      "step": 26735
    },
    {
      "epoch": 55.59251559251559,
      "grad_norm": 0.0004548740980681032,
      "learning_rate": 0.0742632246360395,
      "loss": 0.2675,
      "num_input_tokens_seen": 20393000,
      "step": 26740
    },
    {
      "epoch": 55.602910602910605,
      "grad_norm": 0.00047511718003079295,
      "learning_rate": 0.07421238551294934,
      "loss": 0.2092,
      "num_input_tokens_seen": 20396904,
      "step": 26745
    },
    {
      "epoch": 55.61330561330561,
      "grad_norm": 0.00021994025155436248,
      "learning_rate": 0.07416155807726171,
      "loss": 0.2737,
      "num_input_tokens_seen": 20400712,
      "step": 26750
    },
    {
      "epoch": 55.62370062370062,
      "grad_norm": 0.0002352344454266131,
      "learning_rate": 0.07411074233681492,
      "loss": 0.2523,
      "num_input_tokens_seen": 20404648,
      "step": 26755
    },
    {
      "epoch": 55.63409563409564,
      "grad_norm": 0.00011049176100641489,
      "learning_rate": 0.07405993829944528,
      "loss": 0.238,
      "num_input_tokens_seen": 20408424,
      "step": 26760
    },
    {
      "epoch": 55.644490644490645,
      "grad_norm": 0.0002080237027257681,
      "learning_rate": 0.07400914597298755,
      "loss": 0.2335,
      "num_input_tokens_seen": 20412200,
      "step": 26765
    },
    {
      "epoch": 55.65488565488565,
      "grad_norm": 0.00014888279838487506,
      "learning_rate": 0.07395836536527445,
      "loss": 0.2847,
      "num_input_tokens_seen": 20416104,
      "step": 26770
    },
    {
      "epoch": 55.66528066528066,
      "grad_norm": 8.534899825463071e-05,
      "learning_rate": 0.07390759648413696,
      "loss": 0.2651,
      "num_input_tokens_seen": 20419912,
      "step": 26775
    },
    {
      "epoch": 55.67567567567568,
      "grad_norm": 3.690117227961309e-05,
      "learning_rate": 0.07385683933740435,
      "loss": 0.2679,
      "num_input_tokens_seen": 20423688,
      "step": 26780
    },
    {
      "epoch": 55.686070686070686,
      "grad_norm": 0.00038660975405946374,
      "learning_rate": 0.07380609393290402,
      "loss": 0.2681,
      "num_input_tokens_seen": 20427464,
      "step": 26785
    },
    {
      "epoch": 55.696465696465694,
      "grad_norm": 0.00012417722609825432,
      "learning_rate": 0.07375536027846147,
      "loss": 0.274,
      "num_input_tokens_seen": 20431272,
      "step": 26790
    },
    {
      "epoch": 55.70686070686071,
      "grad_norm": 7.76619344833307e-05,
      "learning_rate": 0.07370463838190057,
      "loss": 0.2771,
      "num_input_tokens_seen": 20435272,
      "step": 26795
    },
    {
      "epoch": 55.71725571725572,
      "grad_norm": 0.00020915240747854114,
      "learning_rate": 0.07365392825104317,
      "loss": 0.2596,
      "num_input_tokens_seen": 20439016,
      "step": 26800
    },
    {
      "epoch": 55.71725571725572,
      "eval_loss": 0.2567959427833557,
      "eval_runtime": 13.3934,
      "eval_samples_per_second": 63.912,
      "eval_steps_per_second": 15.978,
      "num_input_tokens_seen": 20439016,
      "step": 26800
    },
    {
      "epoch": 55.727650727650726,
      "grad_norm": 9.20339734875597e-05,
      "learning_rate": 0.07360322989370945,
      "loss": 0.2705,
      "num_input_tokens_seen": 20442888,
      "step": 26805
    },
    {
      "epoch": 55.73804573804574,
      "grad_norm": 0.00031852704705670476,
      "learning_rate": 0.07355254331771781,
      "loss": 0.2741,
      "num_input_tokens_seen": 20446920,
      "step": 26810
    },
    {
      "epoch": 55.74844074844075,
      "grad_norm": 0.00017445560661144555,
      "learning_rate": 0.07350186853088461,
      "loss": 0.2739,
      "num_input_tokens_seen": 20450792,
      "step": 26815
    },
    {
      "epoch": 55.75883575883576,
      "grad_norm": 0.00011392993474146351,
      "learning_rate": 0.07345120554102462,
      "loss": 0.2754,
      "num_input_tokens_seen": 20454536,
      "step": 26820
    },
    {
      "epoch": 55.76923076923077,
      "grad_norm": 0.0002019075327552855,
      "learning_rate": 0.07340055435595079,
      "loss": 0.2588,
      "num_input_tokens_seen": 20458376,
      "step": 26825
    },
    {
      "epoch": 55.77962577962578,
      "grad_norm": 0.0003578672476578504,
      "learning_rate": 0.07334991498347401,
      "loss": 0.271,
      "num_input_tokens_seen": 20462184,
      "step": 26830
    },
    {
      "epoch": 55.79002079002079,
      "grad_norm": 0.00017872618627734482,
      "learning_rate": 0.07329928743140365,
      "loss": 0.2413,
      "num_input_tokens_seen": 20465992,
      "step": 26835
    },
    {
      "epoch": 55.8004158004158,
      "grad_norm": 0.00015051898662932217,
      "learning_rate": 0.07324867170754705,
      "loss": 0.2568,
      "num_input_tokens_seen": 20469928,
      "step": 26840
    },
    {
      "epoch": 55.810810810810814,
      "grad_norm": 0.00038806095835752785,
      "learning_rate": 0.07319806781970974,
      "loss": 0.2182,
      "num_input_tokens_seen": 20473576,
      "step": 26845
    },
    {
      "epoch": 55.82120582120582,
      "grad_norm": 0.00018154567806050181,
      "learning_rate": 0.07314747577569555,
      "loss": 0.2844,
      "num_input_tokens_seen": 20477288,
      "step": 26850
    },
    {
      "epoch": 55.83160083160083,
      "grad_norm": 0.00016651175974402577,
      "learning_rate": 0.07309689558330636,
      "loss": 0.2558,
      "num_input_tokens_seen": 20481160,
      "step": 26855
    },
    {
      "epoch": 55.84199584199584,
      "grad_norm": 0.0004189663741271943,
      "learning_rate": 0.0730463272503423,
      "loss": 0.2799,
      "num_input_tokens_seen": 20484968,
      "step": 26860
    },
    {
      "epoch": 55.852390852390855,
      "grad_norm": 2.620306440803688e-05,
      "learning_rate": 0.07299577078460168,
      "loss": 0.2827,
      "num_input_tokens_seen": 20488616,
      "step": 26865
    },
    {
      "epoch": 55.86278586278586,
      "grad_norm": 0.0001188592505059205,
      "learning_rate": 0.07294522619388083,
      "loss": 0.2655,
      "num_input_tokens_seen": 20492328,
      "step": 26870
    },
    {
      "epoch": 55.87318087318087,
      "grad_norm": 5.945091470493935e-05,
      "learning_rate": 0.07289469348597452,
      "loss": 0.2696,
      "num_input_tokens_seen": 20496104,
      "step": 26875
    },
    {
      "epoch": 55.88357588357589,
      "grad_norm": 0.0005189462681300938,
      "learning_rate": 0.07284417266867535,
      "loss": 0.2742,
      "num_input_tokens_seen": 20499912,
      "step": 26880
    },
    {
      "epoch": 55.893970893970895,
      "grad_norm": 0.0003031924716196954,
      "learning_rate": 0.07279366374977439,
      "loss": 0.2684,
      "num_input_tokens_seen": 20503656,
      "step": 26885
    },
    {
      "epoch": 55.9043659043659,
      "grad_norm": 2.6656425688997842e-05,
      "learning_rate": 0.07274316673706074,
      "loss": 0.2665,
      "num_input_tokens_seen": 20507400,
      "step": 26890
    },
    {
      "epoch": 55.91476091476091,
      "grad_norm": 0.0003867803025059402,
      "learning_rate": 0.07269268163832161,
      "loss": 0.2593,
      "num_input_tokens_seen": 20511272,
      "step": 26895
    },
    {
      "epoch": 55.92515592515593,
      "grad_norm": 0.00018587293743621558,
      "learning_rate": 0.07264220846134248,
      "loss": 0.2763,
      "num_input_tokens_seen": 20515176,
      "step": 26900
    },
    {
      "epoch": 55.935550935550935,
      "grad_norm": 5.33304555574432e-05,
      "learning_rate": 0.07259174721390699,
      "loss": 0.2449,
      "num_input_tokens_seen": 20518760,
      "step": 26905
    },
    {
      "epoch": 55.945945945945944,
      "grad_norm": 5.91612551943399e-05,
      "learning_rate": 0.07254129790379686,
      "loss": 0.253,
      "num_input_tokens_seen": 20522536,
      "step": 26910
    },
    {
      "epoch": 55.95634095634096,
      "grad_norm": 0.0001117538777180016,
      "learning_rate": 0.072490860538792,
      "loss": 0.2807,
      "num_input_tokens_seen": 20526344,
      "step": 26915
    },
    {
      "epoch": 55.96673596673597,
      "grad_norm": 0.00021780496172141284,
      "learning_rate": 0.07244043512667042,
      "loss": 0.2514,
      "num_input_tokens_seen": 20530312,
      "step": 26920
    },
    {
      "epoch": 55.977130977130976,
      "grad_norm": 0.00018675463797990233,
      "learning_rate": 0.07239002167520843,
      "loss": 0.2696,
      "num_input_tokens_seen": 20534088,
      "step": 26925
    },
    {
      "epoch": 55.987525987525984,
      "grad_norm": 0.00045130591024644673,
      "learning_rate": 0.07233962019218045,
      "loss": 0.2777,
      "num_input_tokens_seen": 20537896,
      "step": 26930
    },
    {
      "epoch": 55.997920997921,
      "grad_norm": 4.9080379540100694e-05,
      "learning_rate": 0.07228923068535892,
      "loss": 0.2503,
      "num_input_tokens_seen": 20541704,
      "step": 26935
    },
    {
      "epoch": 56.00831600831601,
      "grad_norm": 0.00019629186135716736,
      "learning_rate": 0.0722388531625146,
      "loss": 0.2684,
      "num_input_tokens_seen": 20545368,
      "step": 26940
    },
    {
      "epoch": 56.018711018711016,
      "grad_norm": 0.00039805221604183316,
      "learning_rate": 0.07218848763141639,
      "loss": 0.2751,
      "num_input_tokens_seen": 20549272,
      "step": 26945
    },
    {
      "epoch": 56.02910602910603,
      "grad_norm": 0.0009759733220562339,
      "learning_rate": 0.07213813409983118,
      "loss": 0.2769,
      "num_input_tokens_seen": 20552920,
      "step": 26950
    },
    {
      "epoch": 56.03950103950104,
      "grad_norm": 0.00021742662647739053,
      "learning_rate": 0.0720877925755242,
      "loss": 0.2701,
      "num_input_tokens_seen": 20556760,
      "step": 26955
    },
    {
      "epoch": 56.04989604989605,
      "grad_norm": 0.00035511149326339364,
      "learning_rate": 0.07203746306625866,
      "loss": 0.2614,
      "num_input_tokens_seen": 20560600,
      "step": 26960
    },
    {
      "epoch": 56.06029106029106,
      "grad_norm": 3.565665247151628e-05,
      "learning_rate": 0.07198714557979606,
      "loss": 0.2676,
      "num_input_tokens_seen": 20564344,
      "step": 26965
    },
    {
      "epoch": 56.07068607068607,
      "grad_norm": 0.0006292808102443814,
      "learning_rate": 0.07193684012389602,
      "loss": 0.2265,
      "num_input_tokens_seen": 20568312,
      "step": 26970
    },
    {
      "epoch": 56.08108108108108,
      "grad_norm": 7.516855839639902e-05,
      "learning_rate": 0.07188654670631621,
      "loss": 0.2383,
      "num_input_tokens_seen": 20572088,
      "step": 26975
    },
    {
      "epoch": 56.09147609147609,
      "grad_norm": 0.0002188281068811193,
      "learning_rate": 0.07183626533481258,
      "loss": 0.2951,
      "num_input_tokens_seen": 20575864,
      "step": 26980
    },
    {
      "epoch": 56.101871101871104,
      "grad_norm": 0.00010026013478636742,
      "learning_rate": 0.07178599601713909,
      "loss": 0.229,
      "num_input_tokens_seen": 20579704,
      "step": 26985
    },
    {
      "epoch": 56.11226611226611,
      "grad_norm": 0.0007259357953444123,
      "learning_rate": 0.07173573876104786,
      "loss": 0.2956,
      "num_input_tokens_seen": 20583608,
      "step": 26990
    },
    {
      "epoch": 56.12266112266112,
      "grad_norm": 0.00037151671131141484,
      "learning_rate": 0.0716854935742893,
      "loss": 0.2671,
      "num_input_tokens_seen": 20587480,
      "step": 26995
    },
    {
      "epoch": 56.13305613305613,
      "grad_norm": 0.0004270995850674808,
      "learning_rate": 0.07163526046461174,
      "loss": 0.2786,
      "num_input_tokens_seen": 20591320,
      "step": 27000
    },
    {
      "epoch": 56.13305613305613,
      "eval_loss": 0.2581345736980438,
      "eval_runtime": 13.3745,
      "eval_samples_per_second": 64.002,
      "eval_steps_per_second": 16.001,
      "num_input_tokens_seen": 20591320,
      "step": 27000
    },
    {
      "epoch": 56.143451143451145,
      "grad_norm": 0.0002719515177886933,
      "learning_rate": 0.07158503943976181,
      "loss": 0.2738,
      "num_input_tokens_seen": 20594904,
      "step": 27005
    },
    {
      "epoch": 56.15384615384615,
      "grad_norm": 0.00025673562777228653,
      "learning_rate": 0.07153483050748427,
      "loss": 0.2695,
      "num_input_tokens_seen": 20598680,
      "step": 27010
    },
    {
      "epoch": 56.16424116424116,
      "grad_norm": 0.0001781985629349947,
      "learning_rate": 0.07148463367552188,
      "loss": 0.2598,
      "num_input_tokens_seen": 20602456,
      "step": 27015
    },
    {
      "epoch": 56.17463617463618,
      "grad_norm": 0.0003269309236202389,
      "learning_rate": 0.07143444895161565,
      "loss": 0.2473,
      "num_input_tokens_seen": 20606104,
      "step": 27020
    },
    {
      "epoch": 56.185031185031185,
      "grad_norm": 0.0001319317234447226,
      "learning_rate": 0.07138427634350476,
      "loss": 0.2513,
      "num_input_tokens_seen": 20609848,
      "step": 27025
    },
    {
      "epoch": 56.195426195426194,
      "grad_norm": 5.980447167530656e-05,
      "learning_rate": 0.07133411585892636,
      "loss": 0.2628,
      "num_input_tokens_seen": 20613816,
      "step": 27030
    },
    {
      "epoch": 56.20582120582121,
      "grad_norm": 0.0001459171762689948,
      "learning_rate": 0.07128396750561593,
      "loss": 0.3004,
      "num_input_tokens_seen": 20617720,
      "step": 27035
    },
    {
      "epoch": 56.21621621621622,
      "grad_norm": 0.00025547415134496987,
      "learning_rate": 0.07123383129130685,
      "loss": 0.2705,
      "num_input_tokens_seen": 20621432,
      "step": 27040
    },
    {
      "epoch": 56.226611226611226,
      "grad_norm": 0.0007055096793919802,
      "learning_rate": 0.07118370722373084,
      "loss": 0.2568,
      "num_input_tokens_seen": 20625208,
      "step": 27045
    },
    {
      "epoch": 56.237006237006234,
      "grad_norm": 0.0001665985182626173,
      "learning_rate": 0.07113359531061769,
      "loss": 0.2587,
      "num_input_tokens_seen": 20629016,
      "step": 27050
    },
    {
      "epoch": 56.24740124740125,
      "grad_norm": 0.00030473803053610027,
      "learning_rate": 0.07108349555969525,
      "loss": 0.2744,
      "num_input_tokens_seen": 20632920,
      "step": 27055
    },
    {
      "epoch": 56.25779625779626,
      "grad_norm": 9.962715557776392e-05,
      "learning_rate": 0.07103340797868944,
      "loss": 0.2923,
      "num_input_tokens_seen": 20636728,
      "step": 27060
    },
    {
      "epoch": 56.268191268191266,
      "grad_norm": 0.00044138816883787513,
      "learning_rate": 0.07098333257532453,
      "loss": 0.2605,
      "num_input_tokens_seen": 20640568,
      "step": 27065
    },
    {
      "epoch": 56.27858627858628,
      "grad_norm": 0.00015792126941960305,
      "learning_rate": 0.07093326935732269,
      "loss": 0.2613,
      "num_input_tokens_seen": 20644376,
      "step": 27070
    },
    {
      "epoch": 56.28898128898129,
      "grad_norm": 0.00021895718236919492,
      "learning_rate": 0.0708832183324044,
      "loss": 0.2763,
      "num_input_tokens_seen": 20648184,
      "step": 27075
    },
    {
      "epoch": 56.2993762993763,
      "grad_norm": 0.0003913456457667053,
      "learning_rate": 0.07083317950828799,
      "loss": 0.2329,
      "num_input_tokens_seen": 20651960,
      "step": 27080
    },
    {
      "epoch": 56.30977130977131,
      "grad_norm": 0.000294084515189752,
      "learning_rate": 0.0707831528926902,
      "loss": 0.2805,
      "num_input_tokens_seen": 20655768,
      "step": 27085
    },
    {
      "epoch": 56.32016632016632,
      "grad_norm": 0.00021721870871260762,
      "learning_rate": 0.07073313849332578,
      "loss": 0.2741,
      "num_input_tokens_seen": 20659544,
      "step": 27090
    },
    {
      "epoch": 56.33056133056133,
      "grad_norm": 0.00021275108156260103,
      "learning_rate": 0.07068313631790749,
      "loss": 0.235,
      "num_input_tokens_seen": 20663352,
      "step": 27095
    },
    {
      "epoch": 56.34095634095634,
      "grad_norm": 6.678060890408233e-05,
      "learning_rate": 0.07063314637414632,
      "loss": 0.2458,
      "num_input_tokens_seen": 20667160,
      "step": 27100
    },
    {
      "epoch": 56.351351351351354,
      "grad_norm": 8.47870614961721e-05,
      "learning_rate": 0.07058316866975144,
      "loss": 0.2511,
      "num_input_tokens_seen": 20670968,
      "step": 27105
    },
    {
      "epoch": 56.36174636174636,
      "grad_norm": 0.00012517163122538477,
      "learning_rate": 0.0705332032124299,
      "loss": 0.2483,
      "num_input_tokens_seen": 20674872,
      "step": 27110
    },
    {
      "epoch": 56.37214137214137,
      "grad_norm": 0.00028426633798517287,
      "learning_rate": 0.0704832500098871,
      "loss": 0.2776,
      "num_input_tokens_seen": 20678808,
      "step": 27115
    },
    {
      "epoch": 56.38253638253638,
      "grad_norm": 8.44882961246185e-05,
      "learning_rate": 0.07043330906982641,
      "loss": 0.2561,
      "num_input_tokens_seen": 20682776,
      "step": 27120
    },
    {
      "epoch": 56.392931392931395,
      "grad_norm": 0.0004336985875852406,
      "learning_rate": 0.07038338039994936,
      "loss": 0.2751,
      "num_input_tokens_seen": 20686552,
      "step": 27125
    },
    {
      "epoch": 56.4033264033264,
      "grad_norm": 0.00011908068699995056,
      "learning_rate": 0.07033346400795562,
      "loss": 0.2467,
      "num_input_tokens_seen": 20690296,
      "step": 27130
    },
    {
      "epoch": 56.41372141372141,
      "grad_norm": 0.000276941922493279,
      "learning_rate": 0.07028355990154282,
      "loss": 0.2794,
      "num_input_tokens_seen": 20694040,
      "step": 27135
    },
    {
      "epoch": 56.42411642411643,
      "grad_norm": 0.00017509180179331452,
      "learning_rate": 0.07023366808840685,
      "loss": 0.2596,
      "num_input_tokens_seen": 20697816,
      "step": 27140
    },
    {
      "epoch": 56.434511434511435,
      "grad_norm": 0.00017745922377798706,
      "learning_rate": 0.07018378857624172,
      "loss": 0.2707,
      "num_input_tokens_seen": 20701624,
      "step": 27145
    },
    {
      "epoch": 56.444906444906444,
      "grad_norm": 0.0006116931908763945,
      "learning_rate": 0.0701339213727394,
      "loss": 0.2452,
      "num_input_tokens_seen": 20705560,
      "step": 27150
    },
    {
      "epoch": 56.45530145530145,
      "grad_norm": 0.00029468053253367543,
      "learning_rate": 0.07008406648559008,
      "loss": 0.2412,
      "num_input_tokens_seen": 20709272,
      "step": 27155
    },
    {
      "epoch": 56.46569646569647,
      "grad_norm": 0.0005130237550474703,
      "learning_rate": 0.07003422392248196,
      "loss": 0.2479,
      "num_input_tokens_seen": 20713048,
      "step": 27160
    },
    {
      "epoch": 56.476091476091476,
      "grad_norm": 0.0003463863395154476,
      "learning_rate": 0.06998439369110142,
      "loss": 0.2645,
      "num_input_tokens_seen": 20717016,
      "step": 27165
    },
    {
      "epoch": 56.486486486486484,
      "grad_norm": 0.0006893663667142391,
      "learning_rate": 0.06993457579913295,
      "loss": 0.3108,
      "num_input_tokens_seen": 20720824,
      "step": 27170
    },
    {
      "epoch": 56.4968814968815,
      "grad_norm": 0.0006446539191529155,
      "learning_rate": 0.06988477025425903,
      "loss": 0.2827,
      "num_input_tokens_seen": 20724664,
      "step": 27175
    },
    {
      "epoch": 56.50727650727651,
      "grad_norm": 0.00012618370237760246,
      "learning_rate": 0.06983497706416032,
      "loss": 0.2698,
      "num_input_tokens_seen": 20728408,
      "step": 27180
    },
    {
      "epoch": 56.517671517671516,
      "grad_norm": 0.00029114712378941476,
      "learning_rate": 0.0697851962365156,
      "loss": 0.2761,
      "num_input_tokens_seen": 20732184,
      "step": 27185
    },
    {
      "epoch": 56.528066528066525,
      "grad_norm": 0.00043628213461488485,
      "learning_rate": 0.06973542777900163,
      "loss": 0.271,
      "num_input_tokens_seen": 20735896,
      "step": 27190
    },
    {
      "epoch": 56.53846153846154,
      "grad_norm": 6.305332499323413e-05,
      "learning_rate": 0.06968567169929342,
      "loss": 0.2541,
      "num_input_tokens_seen": 20739992,
      "step": 27195
    },
    {
      "epoch": 56.54885654885655,
      "grad_norm": 0.00034865402267314494,
      "learning_rate": 0.06963592800506392,
      "loss": 0.2734,
      "num_input_tokens_seen": 20743736,
      "step": 27200
    },
    {
      "epoch": 56.54885654885655,
      "eval_loss": 0.24752461910247803,
      "eval_runtime": 13.3921,
      "eval_samples_per_second": 63.918,
      "eval_steps_per_second": 15.98,
      "num_input_tokens_seen": 20743736,
      "step": 27200
    },
    {
      "epoch": 56.55925155925156,
      "grad_norm": 5.927007441641763e-05,
      "learning_rate": 0.06958619670398417,
      "loss": 0.2702,
      "num_input_tokens_seen": 20747480,
      "step": 27205
    },
    {
      "epoch": 56.56964656964657,
      "grad_norm": 0.00034109526313841343,
      "learning_rate": 0.0695364778037235,
      "loss": 0.2529,
      "num_input_tokens_seen": 20751288,
      "step": 27210
    },
    {
      "epoch": 56.58004158004158,
      "grad_norm": 0.0003387674514669925,
      "learning_rate": 0.06948677131194907,
      "loss": 0.2726,
      "num_input_tokens_seen": 20755032,
      "step": 27215
    },
    {
      "epoch": 56.59043659043659,
      "grad_norm": 0.0001852315617725253,
      "learning_rate": 0.06943707723632629,
      "loss": 0.2594,
      "num_input_tokens_seen": 20758776,
      "step": 27220
    },
    {
      "epoch": 56.6008316008316,
      "grad_norm": 6.984312494751066e-05,
      "learning_rate": 0.06938739558451867,
      "loss": 0.2551,
      "num_input_tokens_seen": 20762776,
      "step": 27225
    },
    {
      "epoch": 56.61122661122661,
      "grad_norm": 0.00020217348355799913,
      "learning_rate": 0.06933772636418763,
      "loss": 0.2554,
      "num_input_tokens_seen": 20766520,
      "step": 27230
    },
    {
      "epoch": 56.62162162162162,
      "grad_norm": 4.147472282056697e-05,
      "learning_rate": 0.06928806958299293,
      "loss": 0.2791,
      "num_input_tokens_seen": 20770520,
      "step": 27235
    },
    {
      "epoch": 56.63201663201663,
      "grad_norm": 0.0004289264907129109,
      "learning_rate": 0.06923842524859211,
      "loss": 0.2416,
      "num_input_tokens_seen": 20774552,
      "step": 27240
    },
    {
      "epoch": 56.642411642411645,
      "grad_norm": 0.0005177919520065188,
      "learning_rate": 0.06918879336864105,
      "loss": 0.2557,
      "num_input_tokens_seen": 20778424,
      "step": 27245
    },
    {
      "epoch": 56.65280665280665,
      "grad_norm": 0.0002976930409204215,
      "learning_rate": 0.06913917395079362,
      "loss": 0.2837,
      "num_input_tokens_seen": 20782360,
      "step": 27250
    },
    {
      "epoch": 56.66320166320166,
      "grad_norm": 8.254634303739294e-05,
      "learning_rate": 0.0690895670027017,
      "loss": 0.2559,
      "num_input_tokens_seen": 20786328,
      "step": 27255
    },
    {
      "epoch": 56.67359667359668,
      "grad_norm": 0.00041633605724200606,
      "learning_rate": 0.06903997253201531,
      "loss": 0.2689,
      "num_input_tokens_seen": 20790008,
      "step": 27260
    },
    {
      "epoch": 56.683991683991685,
      "grad_norm": 0.00038434506859630346,
      "learning_rate": 0.06899039054638263,
      "loss": 0.2771,
      "num_input_tokens_seen": 20793752,
      "step": 27265
    },
    {
      "epoch": 56.694386694386694,
      "grad_norm": 3.790081973420456e-05,
      "learning_rate": 0.06894082105344976,
      "loss": 0.2694,
      "num_input_tokens_seen": 20797464,
      "step": 27270
    },
    {
      "epoch": 56.7047817047817,
      "grad_norm": 0.00043999950867146254,
      "learning_rate": 0.06889126406086087,
      "loss": 0.2674,
      "num_input_tokens_seen": 20801368,
      "step": 27275
    },
    {
      "epoch": 56.71517671517672,
      "grad_norm": 0.0001274142268812284,
      "learning_rate": 0.0688417195762584,
      "loss": 0.2643,
      "num_input_tokens_seen": 20805144,
      "step": 27280
    },
    {
      "epoch": 56.725571725571726,
      "grad_norm": 0.00030875412630848587,
      "learning_rate": 0.06879218760728262,
      "loss": 0.2626,
      "num_input_tokens_seen": 20808856,
      "step": 27285
    },
    {
      "epoch": 56.735966735966734,
      "grad_norm": 0.00036022934364154935,
      "learning_rate": 0.06874266816157207,
      "loss": 0.2586,
      "num_input_tokens_seen": 20812728,
      "step": 27290
    },
    {
      "epoch": 56.74636174636175,
      "grad_norm": 0.00023577686806675047,
      "learning_rate": 0.06869316124676321,
      "loss": 0.2913,
      "num_input_tokens_seen": 20816536,
      "step": 27295
    },
    {
      "epoch": 56.75675675675676,
      "grad_norm": 0.00047496502520516515,
      "learning_rate": 0.06864366687049062,
      "loss": 0.2761,
      "num_input_tokens_seen": 20820408,
      "step": 27300
    },
    {
      "epoch": 56.767151767151766,
      "grad_norm": 5.42862580914516e-05,
      "learning_rate": 0.06859418504038704,
      "loss": 0.2577,
      "num_input_tokens_seen": 20824216,
      "step": 27305
    },
    {
      "epoch": 56.777546777546775,
      "grad_norm": 7.574332994408906e-05,
      "learning_rate": 0.06854471576408311,
      "loss": 0.2802,
      "num_input_tokens_seen": 20827992,
      "step": 27310
    },
    {
      "epoch": 56.78794178794179,
      "grad_norm": 0.0003212700830772519,
      "learning_rate": 0.06849525904920767,
      "loss": 0.268,
      "num_input_tokens_seen": 20831832,
      "step": 27315
    },
    {
      "epoch": 56.7983367983368,
      "grad_norm": 0.00012190680718049407,
      "learning_rate": 0.06844581490338748,
      "loss": 0.2547,
      "num_input_tokens_seen": 20835672,
      "step": 27320
    },
    {
      "epoch": 56.80873180873181,
      "grad_norm": 0.00013367677456699312,
      "learning_rate": 0.06839638333424752,
      "loss": 0.284,
      "num_input_tokens_seen": 20839352,
      "step": 27325
    },
    {
      "epoch": 56.81912681912682,
      "grad_norm": 0.0003036781563423574,
      "learning_rate": 0.06834696434941082,
      "loss": 0.2783,
      "num_input_tokens_seen": 20843224,
      "step": 27330
    },
    {
      "epoch": 56.82952182952183,
      "grad_norm": 0.00011597020784392953,
      "learning_rate": 0.06829755795649824,
      "loss": 0.2645,
      "num_input_tokens_seen": 20847000,
      "step": 27335
    },
    {
      "epoch": 56.83991683991684,
      "grad_norm": 6.796064553782344e-05,
      "learning_rate": 0.06824816416312904,
      "loss": 0.2656,
      "num_input_tokens_seen": 20850872,
      "step": 27340
    },
    {
      "epoch": 56.85031185031185,
      "grad_norm": 0.0002623731561470777,
      "learning_rate": 0.06819878297692027,
      "loss": 0.265,
      "num_input_tokens_seen": 20854680,
      "step": 27345
    },
    {
      "epoch": 56.86070686070686,
      "grad_norm": 0.00013267601025290787,
      "learning_rate": 0.0681494144054871,
      "loss": 0.2727,
      "num_input_tokens_seen": 20858552,
      "step": 27350
    },
    {
      "epoch": 56.87110187110187,
      "grad_norm": 0.0003577958850655705,
      "learning_rate": 0.06810005845644286,
      "loss": 0.2602,
      "num_input_tokens_seen": 20862424,
      "step": 27355
    },
    {
      "epoch": 56.88149688149688,
      "grad_norm": 0.00036278337938711047,
      "learning_rate": 0.06805071513739878,
      "loss": 0.271,
      "num_input_tokens_seen": 20866200,
      "step": 27360
    },
    {
      "epoch": 56.891891891891895,
      "grad_norm": 0.0002547994372434914,
      "learning_rate": 0.06800138445596428,
      "loss": 0.2598,
      "num_input_tokens_seen": 20869944,
      "step": 27365
    },
    {
      "epoch": 56.9022869022869,
      "grad_norm": 0.00022513128351420164,
      "learning_rate": 0.06795206641974678,
      "loss": 0.2625,
      "num_input_tokens_seen": 20873752,
      "step": 27370
    },
    {
      "epoch": 56.91268191268191,
      "grad_norm": 0.0002479232207406312,
      "learning_rate": 0.06790276103635169,
      "loss": 0.2462,
      "num_input_tokens_seen": 20877432,
      "step": 27375
    },
    {
      "epoch": 56.92307692307692,
      "grad_norm": 3.6150941014057025e-05,
      "learning_rate": 0.0678534683133826,
      "loss": 0.2662,
      "num_input_tokens_seen": 20881080,
      "step": 27380
    },
    {
      "epoch": 56.933471933471935,
      "grad_norm": 0.00012071291712345555,
      "learning_rate": 0.06780418825844095,
      "loss": 0.2615,
      "num_input_tokens_seen": 20884856,
      "step": 27385
    },
    {
      "epoch": 56.943866943866944,
      "grad_norm": 9.35002535698004e-05,
      "learning_rate": 0.0677549208791264,
      "loss": 0.2819,
      "num_input_tokens_seen": 20888632,
      "step": 27390
    },
    {
      "epoch": 56.95426195426195,
      "grad_norm": 0.00014534560614265501,
      "learning_rate": 0.06770566618303668,
      "loss": 0.2742,
      "num_input_tokens_seen": 20892440,
      "step": 27395
    },
    {
      "epoch": 56.96465696465697,
      "grad_norm": 7.129237928893417e-05,
      "learning_rate": 0.06765642417776736,
      "loss": 0.2701,
      "num_input_tokens_seen": 20896184,
      "step": 27400
    },
    {
      "epoch": 56.96465696465697,
      "eval_loss": 0.25731244683265686,
      "eval_runtime": 13.412,
      "eval_samples_per_second": 63.823,
      "eval_steps_per_second": 15.956,
      "num_input_tokens_seen": 20896184,
      "step": 27400
    },
    {
      "epoch": 56.975051975051976,
      "grad_norm": 0.00025478401221334934,
      "learning_rate": 0.0676071948709122,
      "loss": 0.2744,
      "num_input_tokens_seen": 20900088,
      "step": 27405
    },
    {
      "epoch": 56.985446985446984,
      "grad_norm": 0.00022322667064145207,
      "learning_rate": 0.06755797827006307,
      "loss": 0.2725,
      "num_input_tokens_seen": 20903960,
      "step": 27410
    },
    {
      "epoch": 56.99584199584199,
      "grad_norm": 0.0004477265465538949,
      "learning_rate": 0.06750877438280974,
      "loss": 0.2495,
      "num_input_tokens_seen": 20907864,
      "step": 27415
    },
    {
      "epoch": 57.00623700623701,
      "grad_norm": 0.0004648904432542622,
      "learning_rate": 0.06745958321673998,
      "loss": 0.2387,
      "num_input_tokens_seen": 20911816,
      "step": 27420
    },
    {
      "epoch": 57.016632016632016,
      "grad_norm": 7.16396389179863e-05,
      "learning_rate": 0.0674104047794398,
      "loss": 0.2751,
      "num_input_tokens_seen": 20915720,
      "step": 27425
    },
    {
      "epoch": 57.027027027027025,
      "grad_norm": 0.00017544305592309684,
      "learning_rate": 0.06736123907849303,
      "loss": 0.2738,
      "num_input_tokens_seen": 20919560,
      "step": 27430
    },
    {
      "epoch": 57.03742203742204,
      "grad_norm": 3.616348112700507e-05,
      "learning_rate": 0.06731208612148178,
      "loss": 0.2253,
      "num_input_tokens_seen": 20923336,
      "step": 27435
    },
    {
      "epoch": 57.04781704781705,
      "grad_norm": 0.0003717425570357591,
      "learning_rate": 0.0672629459159859,
      "loss": 0.2718,
      "num_input_tokens_seen": 20927144,
      "step": 27440
    },
    {
      "epoch": 57.05821205821206,
      "grad_norm": 0.0003411647630855441,
      "learning_rate": 0.0672138184695835,
      "loss": 0.2692,
      "num_input_tokens_seen": 20930952,
      "step": 27445
    },
    {
      "epoch": 57.06860706860707,
      "grad_norm": 0.00027951670926995575,
      "learning_rate": 0.0671647037898507,
      "loss": 0.2835,
      "num_input_tokens_seen": 20934792,
      "step": 27450
    },
    {
      "epoch": 57.07900207900208,
      "grad_norm": 0.00012088056973880157,
      "learning_rate": 0.0671156018843615,
      "loss": 0.2633,
      "num_input_tokens_seen": 20938632,
      "step": 27455
    },
    {
      "epoch": 57.08939708939709,
      "grad_norm": 9.906120976665989e-05,
      "learning_rate": 0.06706651276068812,
      "loss": 0.2692,
      "num_input_tokens_seen": 20942568,
      "step": 27460
    },
    {
      "epoch": 57.0997920997921,
      "grad_norm": 0.00019783749303314835,
      "learning_rate": 0.06701743642640064,
      "loss": 0.2818,
      "num_input_tokens_seen": 20946472,
      "step": 27465
    },
    {
      "epoch": 57.11018711018711,
      "grad_norm": 0.00023209824576042593,
      "learning_rate": 0.06696837288906729,
      "loss": 0.2723,
      "num_input_tokens_seen": 20950376,
      "step": 27470
    },
    {
      "epoch": 57.12058212058212,
      "grad_norm": 0.00015431437350343913,
      "learning_rate": 0.06691932215625432,
      "loss": 0.2725,
      "num_input_tokens_seen": 20954088,
      "step": 27475
    },
    {
      "epoch": 57.13097713097713,
      "grad_norm": 0.0004136514035053551,
      "learning_rate": 0.06687028423552589,
      "loss": 0.2527,
      "num_input_tokens_seen": 20957864,
      "step": 27480
    },
    {
      "epoch": 57.141372141372145,
      "grad_norm": 0.0006636630860157311,
      "learning_rate": 0.06682125913444435,
      "loss": 0.282,
      "num_input_tokens_seen": 20961736,
      "step": 27485
    },
    {
      "epoch": 57.15176715176715,
      "grad_norm": 0.0002690414839889854,
      "learning_rate": 0.0667722468605699,
      "loss": 0.2491,
      "num_input_tokens_seen": 20965448,
      "step": 27490
    },
    {
      "epoch": 57.16216216216216,
      "grad_norm": 4.071547300554812e-05,
      "learning_rate": 0.06672324742146094,
      "loss": 0.2728,
      "num_input_tokens_seen": 20969256,
      "step": 27495
    },
    {
      "epoch": 57.17255717255717,
      "grad_norm": 5.4155843827174976e-05,
      "learning_rate": 0.06667426082467373,
      "loss": 0.2726,
      "num_input_tokens_seen": 20973192,
      "step": 27500
    },
    {
      "epoch": 57.182952182952185,
      "grad_norm": 0.0002416012721369043,
      "learning_rate": 0.0666252870777626,
      "loss": 0.2666,
      "num_input_tokens_seen": 20977000,
      "step": 27505
    },
    {
      "epoch": 57.19334719334719,
      "grad_norm": 0.0002885781868826598,
      "learning_rate": 0.06657632618827995,
      "loss": 0.2618,
      "num_input_tokens_seen": 20980744,
      "step": 27510
    },
    {
      "epoch": 57.2037422037422,
      "grad_norm": 0.00011953869397984818,
      "learning_rate": 0.06652737816377623,
      "loss": 0.2595,
      "num_input_tokens_seen": 20984424,
      "step": 27515
    },
    {
      "epoch": 57.21413721413722,
      "grad_norm": 0.0004012670833617449,
      "learning_rate": 0.06647844301179971,
      "loss": 0.2654,
      "num_input_tokens_seen": 20988136,
      "step": 27520
    },
    {
      "epoch": 57.224532224532226,
      "grad_norm": 0.00010758600546978414,
      "learning_rate": 0.06642952073989689,
      "loss": 0.2352,
      "num_input_tokens_seen": 20992040,
      "step": 27525
    },
    {
      "epoch": 57.234927234927234,
      "grad_norm": 0.0001168713642982766,
      "learning_rate": 0.06638061135561223,
      "loss": 0.247,
      "num_input_tokens_seen": 20995912,
      "step": 27530
    },
    {
      "epoch": 57.24532224532224,
      "grad_norm": 0.00017923676932696253,
      "learning_rate": 0.06633171486648808,
      "loss": 0.2436,
      "num_input_tokens_seen": 20999784,
      "step": 27535
    },
    {
      "epoch": 57.25571725571726,
      "grad_norm": 0.00033305975375697017,
      "learning_rate": 0.06628283128006499,
      "loss": 0.2826,
      "num_input_tokens_seen": 21003784,
      "step": 27540
    },
    {
      "epoch": 57.266112266112266,
      "grad_norm": 0.000500043504871428,
      "learning_rate": 0.0662339606038813,
      "loss": 0.2811,
      "num_input_tokens_seen": 21007400,
      "step": 27545
    },
    {
      "epoch": 57.276507276507274,
      "grad_norm": 0.00020137093088123947,
      "learning_rate": 0.06618510284547358,
      "loss": 0.2465,
      "num_input_tokens_seen": 21011176,
      "step": 27550
    },
    {
      "epoch": 57.28690228690229,
      "grad_norm": 0.0001073284656740725,
      "learning_rate": 0.06613625801237633,
      "loss": 0.2623,
      "num_input_tokens_seen": 21014952,
      "step": 27555
    },
    {
      "epoch": 57.2972972972973,
      "grad_norm": 0.00028848633519373834,
      "learning_rate": 0.066087426112122,
      "loss": 0.2781,
      "num_input_tokens_seen": 21018696,
      "step": 27560
    },
    {
      "epoch": 57.30769230769231,
      "grad_norm": 0.00014227887731976807,
      "learning_rate": 0.06603860715224101,
      "loss": 0.2664,
      "num_input_tokens_seen": 21022472,
      "step": 27565
    },
    {
      "epoch": 57.318087318087315,
      "grad_norm": 0.0007072836742736399,
      "learning_rate": 0.06598980114026198,
      "loss": 0.2625,
      "num_input_tokens_seen": 21026280,
      "step": 27570
    },
    {
      "epoch": 57.32848232848233,
      "grad_norm": 0.00016485484957229346,
      "learning_rate": 0.06594100808371128,
      "loss": 0.2643,
      "num_input_tokens_seen": 21030088,
      "step": 27575
    },
    {
      "epoch": 57.33887733887734,
      "grad_norm": 0.00019880793115589768,
      "learning_rate": 0.06589222799011357,
      "loss": 0.2513,
      "num_input_tokens_seen": 21033800,
      "step": 27580
    },
    {
      "epoch": 57.34927234927235,
      "grad_norm": 0.00030493922531604767,
      "learning_rate": 0.0658434608669912,
      "loss": 0.2898,
      "num_input_tokens_seen": 21037640,
      "step": 27585
    },
    {
      "epoch": 57.35966735966736,
      "grad_norm": 3.667144483188167e-05,
      "learning_rate": 0.06579470672186473,
      "loss": 0.2761,
      "num_input_tokens_seen": 21041512,
      "step": 27590
    },
    {
      "epoch": 57.37006237006237,
      "grad_norm": 0.0004897796316072345,
      "learning_rate": 0.06574596556225275,
      "loss": 0.2603,
      "num_input_tokens_seen": 21045384,
      "step": 27595
    },
    {
      "epoch": 57.38045738045738,
      "grad_norm": 4.648097456083633e-05,
      "learning_rate": 0.06569723739567161,
      "loss": 0.2462,
      "num_input_tokens_seen": 21049160,
      "step": 27600
    },
    {
      "epoch": 57.38045738045738,
      "eval_loss": 0.2511451244354248,
      "eval_runtime": 13.392,
      "eval_samples_per_second": 63.919,
      "eval_steps_per_second": 15.98,
      "num_input_tokens_seen": 21049160,
      "step": 27600
    },
    {
      "epoch": 57.39085239085239,
      "grad_norm": 0.0002854251943062991,
      "learning_rate": 0.06564852222963588,
      "loss": 0.2724,
      "num_input_tokens_seen": 21052904,
      "step": 27605
    },
    {
      "epoch": 57.4012474012474,
      "grad_norm": 0.00017001866945065558,
      "learning_rate": 0.06559982007165813,
      "loss": 0.2779,
      "num_input_tokens_seen": 21056648,
      "step": 27610
    },
    {
      "epoch": 57.41164241164241,
      "grad_norm": 0.00018727932183537632,
      "learning_rate": 0.06555113092924868,
      "loss": 0.2693,
      "num_input_tokens_seen": 21060424,
      "step": 27615
    },
    {
      "epoch": 57.42203742203742,
      "grad_norm": 0.00014927737356629223,
      "learning_rate": 0.06550245480991615,
      "loss": 0.2625,
      "num_input_tokens_seen": 21064200,
      "step": 27620
    },
    {
      "epoch": 57.432432432432435,
      "grad_norm": 0.00011720373731805012,
      "learning_rate": 0.0654537917211669,
      "loss": 0.2515,
      "num_input_tokens_seen": 21067944,
      "step": 27625
    },
    {
      "epoch": 57.44282744282744,
      "grad_norm": 0.00030888401670381427,
      "learning_rate": 0.0654051416705055,
      "loss": 0.2688,
      "num_input_tokens_seen": 21071944,
      "step": 27630
    },
    {
      "epoch": 57.45322245322245,
      "grad_norm": 0.0005090531194582582,
      "learning_rate": 0.06535650466543427,
      "loss": 0.2869,
      "num_input_tokens_seen": 21075752,
      "step": 27635
    },
    {
      "epoch": 57.46361746361746,
      "grad_norm": 0.00022100505884736776,
      "learning_rate": 0.0653078807134538,
      "loss": 0.2763,
      "num_input_tokens_seen": 21079560,
      "step": 27640
    },
    {
      "epoch": 57.474012474012476,
      "grad_norm": 0.00019075334421359003,
      "learning_rate": 0.06525926982206236,
      "loss": 0.2653,
      "num_input_tokens_seen": 21083464,
      "step": 27645
    },
    {
      "epoch": 57.484407484407484,
      "grad_norm": 0.0005287841777317226,
      "learning_rate": 0.06521067199875648,
      "loss": 0.2757,
      "num_input_tokens_seen": 21087432,
      "step": 27650
    },
    {
      "epoch": 57.49480249480249,
      "grad_norm": 0.0003339296963531524,
      "learning_rate": 0.06516208725103047,
      "loss": 0.274,
      "num_input_tokens_seen": 21091304,
      "step": 27655
    },
    {
      "epoch": 57.50519750519751,
      "grad_norm": 0.00017390261928085238,
      "learning_rate": 0.06511351558637678,
      "loss": 0.2716,
      "num_input_tokens_seen": 21095208,
      "step": 27660
    },
    {
      "epoch": 57.515592515592516,
      "grad_norm": 0.00043048374936915934,
      "learning_rate": 0.06506495701228569,
      "loss": 0.2626,
      "num_input_tokens_seen": 21099112,
      "step": 27665
    },
    {
      "epoch": 57.525987525987524,
      "grad_norm": 0.00015417866234201938,
      "learning_rate": 0.06501641153624559,
      "loss": 0.2808,
      "num_input_tokens_seen": 21102952,
      "step": 27670
    },
    {
      "epoch": 57.53638253638254,
      "grad_norm": 0.00010996204218827188,
      "learning_rate": 0.06496787916574286,
      "loss": 0.268,
      "num_input_tokens_seen": 21106728,
      "step": 27675
    },
    {
      "epoch": 57.54677754677755,
      "grad_norm": 0.001330801984295249,
      "learning_rate": 0.06491935990826168,
      "loss": 0.273,
      "num_input_tokens_seen": 21110536,
      "step": 27680
    },
    {
      "epoch": 57.55717255717256,
      "grad_norm": 0.00032189779449254274,
      "learning_rate": 0.0648708537712844,
      "loss": 0.2689,
      "num_input_tokens_seen": 21114504,
      "step": 27685
    },
    {
      "epoch": 57.567567567567565,
      "grad_norm": 0.0006891913944855332,
      "learning_rate": 0.06482236076229132,
      "loss": 0.2425,
      "num_input_tokens_seen": 21118248,
      "step": 27690
    },
    {
      "epoch": 57.57796257796258,
      "grad_norm": 0.00024683194351382554,
      "learning_rate": 0.06477388088876056,
      "loss": 0.2482,
      "num_input_tokens_seen": 21121992,
      "step": 27695
    },
    {
      "epoch": 57.58835758835759,
      "grad_norm": 0.00025702352286316454,
      "learning_rate": 0.06472541415816846,
      "loss": 0.2942,
      "num_input_tokens_seen": 21125736,
      "step": 27700
    },
    {
      "epoch": 57.5987525987526,
      "grad_norm": 0.0001310812949668616,
      "learning_rate": 0.06467696057798909,
      "loss": 0.2584,
      "num_input_tokens_seen": 21129704,
      "step": 27705
    },
    {
      "epoch": 57.60914760914761,
      "grad_norm": 0.00045357918133959174,
      "learning_rate": 0.0646285201556946,
      "loss": 0.2937,
      "num_input_tokens_seen": 21133512,
      "step": 27710
    },
    {
      "epoch": 57.61954261954262,
      "grad_norm": 0.00026637132395990193,
      "learning_rate": 0.06458009289875521,
      "loss": 0.2398,
      "num_input_tokens_seen": 21137224,
      "step": 27715
    },
    {
      "epoch": 57.62993762993763,
      "grad_norm": 0.00027478273841552436,
      "learning_rate": 0.0645316788146389,
      "loss": 0.2668,
      "num_input_tokens_seen": 21140936,
      "step": 27720
    },
    {
      "epoch": 57.64033264033264,
      "grad_norm": 0.0002912095806095749,
      "learning_rate": 0.06448327791081175,
      "loss": 0.2398,
      "num_input_tokens_seen": 21144680,
      "step": 27725
    },
    {
      "epoch": 57.65072765072765,
      "grad_norm": 0.00016119316569529474,
      "learning_rate": 0.0644348901947379,
      "loss": 0.2493,
      "num_input_tokens_seen": 21148520,
      "step": 27730
    },
    {
      "epoch": 57.66112266112266,
      "grad_norm": 0.00014505573199130595,
      "learning_rate": 0.06438651567387917,
      "loss": 0.2706,
      "num_input_tokens_seen": 21152488,
      "step": 27735
    },
    {
      "epoch": 57.67151767151767,
      "grad_norm": 8.437263750238344e-05,
      "learning_rate": 0.0643381543556957,
      "loss": 0.2775,
      "num_input_tokens_seen": 21156424,
      "step": 27740
    },
    {
      "epoch": 57.681912681912685,
      "grad_norm": 0.00010920163185801357,
      "learning_rate": 0.06428980624764526,
      "loss": 0.2526,
      "num_input_tokens_seen": 21160136,
      "step": 27745
    },
    {
      "epoch": 57.69230769230769,
      "grad_norm": 0.00016672284982632846,
      "learning_rate": 0.06424147135718378,
      "loss": 0.2875,
      "num_input_tokens_seen": 21163912,
      "step": 27750
    },
    {
      "epoch": 57.7027027027027,
      "grad_norm": 0.000708926236256957,
      "learning_rate": 0.06419314969176519,
      "loss": 0.2712,
      "num_input_tokens_seen": 21167848,
      "step": 27755
    },
    {
      "epoch": 57.71309771309771,
      "grad_norm": 4.3382948206271976e-05,
      "learning_rate": 0.06414484125884118,
      "loss": 0.2731,
      "num_input_tokens_seen": 21171592,
      "step": 27760
    },
    {
      "epoch": 57.723492723492726,
      "grad_norm": 0.0003938140580430627,
      "learning_rate": 0.06409654606586157,
      "loss": 0.2694,
      "num_input_tokens_seen": 21175432,
      "step": 27765
    },
    {
      "epoch": 57.733887733887734,
      "grad_norm": 6.75722403684631e-05,
      "learning_rate": 0.06404826412027415,
      "loss": 0.2676,
      "num_input_tokens_seen": 21179144,
      "step": 27770
    },
    {
      "epoch": 57.74428274428274,
      "grad_norm": 0.0005064334254711866,
      "learning_rate": 0.06399999542952453,
      "loss": 0.276,
      "num_input_tokens_seen": 21182792,
      "step": 27775
    },
    {
      "epoch": 57.75467775467776,
      "grad_norm": 5.749869524152018e-05,
      "learning_rate": 0.0639517400010563,
      "loss": 0.2531,
      "num_input_tokens_seen": 21186408,
      "step": 27780
    },
    {
      "epoch": 57.765072765072766,
      "grad_norm": 4.6296434447867796e-05,
      "learning_rate": 0.06390349784231118,
      "loss": 0.2667,
      "num_input_tokens_seen": 21190248,
      "step": 27785
    },
    {
      "epoch": 57.775467775467774,
      "grad_norm": 0.0001912676525535062,
      "learning_rate": 0.06385526896072859,
      "loss": 0.2605,
      "num_input_tokens_seen": 21194056,
      "step": 27790
    },
    {
      "epoch": 57.78586278586278,
      "grad_norm": 8.812930173007771e-05,
      "learning_rate": 0.06380705336374613,
      "loss": 0.2776,
      "num_input_tokens_seen": 21197832,
      "step": 27795
    },
    {
      "epoch": 57.7962577962578,
      "grad_norm": 0.00041584664722904563,
      "learning_rate": 0.06375885105879918,
      "loss": 0.269,
      "num_input_tokens_seen": 21201640,
      "step": 27800
    },
    {
      "epoch": 57.7962577962578,
      "eval_loss": 0.2509770393371582,
      "eval_runtime": 13.373,
      "eval_samples_per_second": 64.009,
      "eval_steps_per_second": 16.002,
      "num_input_tokens_seen": 21201640,
      "step": 27800
    },
    {
      "epoch": 57.80665280665281,
      "grad_norm": 0.00025830318918451667,
      "learning_rate": 0.06371066205332115,
      "loss": 0.2429,
      "num_input_tokens_seen": 21205352,
      "step": 27805
    },
    {
      "epoch": 57.817047817047815,
      "grad_norm": 5.51459270354826e-05,
      "learning_rate": 0.06366248635474347,
      "loss": 0.2515,
      "num_input_tokens_seen": 21209064,
      "step": 27810
    },
    {
      "epoch": 57.82744282744283,
      "grad_norm": 6.299043161561713e-05,
      "learning_rate": 0.06361432397049532,
      "loss": 0.2563,
      "num_input_tokens_seen": 21212840,
      "step": 27815
    },
    {
      "epoch": 57.83783783783784,
      "grad_norm": 0.0003149108379147947,
      "learning_rate": 0.06356617490800408,
      "loss": 0.2433,
      "num_input_tokens_seen": 21216744,
      "step": 27820
    },
    {
      "epoch": 57.84823284823285,
      "grad_norm": 0.0002543545560911298,
      "learning_rate": 0.06351803917469478,
      "loss": 0.3197,
      "num_input_tokens_seen": 21220808,
      "step": 27825
    },
    {
      "epoch": 57.858627858627855,
      "grad_norm": 0.0004248986078891903,
      "learning_rate": 0.06346991677799067,
      "loss": 0.2608,
      "num_input_tokens_seen": 21224648,
      "step": 27830
    },
    {
      "epoch": 57.86902286902287,
      "grad_norm": 0.00026921232347376645,
      "learning_rate": 0.06342180772531283,
      "loss": 0.2684,
      "num_input_tokens_seen": 21228552,
      "step": 27835
    },
    {
      "epoch": 57.87941787941788,
      "grad_norm": 0.00020852299348916858,
      "learning_rate": 0.06337371202408021,
      "loss": 0.261,
      "num_input_tokens_seen": 21232392,
      "step": 27840
    },
    {
      "epoch": 57.88981288981289,
      "grad_norm": 0.00013755708641838282,
      "learning_rate": 0.06332562968170984,
      "loss": 0.2454,
      "num_input_tokens_seen": 21236168,
      "step": 27845
    },
    {
      "epoch": 57.9002079002079,
      "grad_norm": 0.00022799032740294933,
      "learning_rate": 0.06327756070561656,
      "loss": 0.2665,
      "num_input_tokens_seen": 21240232,
      "step": 27850
    },
    {
      "epoch": 57.91060291060291,
      "grad_norm": 4.474133311305195e-05,
      "learning_rate": 0.06322950510321329,
      "loss": 0.2789,
      "num_input_tokens_seen": 21243976,
      "step": 27855
    },
    {
      "epoch": 57.92099792099792,
      "grad_norm": 0.0003047766222152859,
      "learning_rate": 0.06318146288191076,
      "loss": 0.282,
      "num_input_tokens_seen": 21247720,
      "step": 27860
    },
    {
      "epoch": 57.931392931392935,
      "grad_norm": 0.00011227883805986494,
      "learning_rate": 0.06313343404911763,
      "loss": 0.2559,
      "num_input_tokens_seen": 21251528,
      "step": 27865
    },
    {
      "epoch": 57.94178794178794,
      "grad_norm": 4.382493716548197e-05,
      "learning_rate": 0.0630854186122406,
      "loss": 0.2669,
      "num_input_tokens_seen": 21255176,
      "step": 27870
    },
    {
      "epoch": 57.95218295218295,
      "grad_norm": 0.00026669702492654324,
      "learning_rate": 0.06303741657868431,
      "loss": 0.2574,
      "num_input_tokens_seen": 21258920,
      "step": 27875
    },
    {
      "epoch": 57.96257796257796,
      "grad_norm": 7.003654900472611e-05,
      "learning_rate": 0.06298942795585115,
      "loss": 0.2672,
      "num_input_tokens_seen": 21262696,
      "step": 27880
    },
    {
      "epoch": 57.972972972972975,
      "grad_norm": 0.00035903751268051565,
      "learning_rate": 0.06294145275114167,
      "loss": 0.2514,
      "num_input_tokens_seen": 21266344,
      "step": 27885
    },
    {
      "epoch": 57.983367983367984,
      "grad_norm": 0.00019085004169028252,
      "learning_rate": 0.06289349097195428,
      "loss": 0.2372,
      "num_input_tokens_seen": 21270152,
      "step": 27890
    },
    {
      "epoch": 57.99376299376299,
      "grad_norm": 0.00032597981044091284,
      "learning_rate": 0.06284554262568516,
      "loss": 0.2674,
      "num_input_tokens_seen": 21274024,
      "step": 27895
    },
    {
      "epoch": 58.00415800415801,
      "grad_norm": 0.0001216758246300742,
      "learning_rate": 0.06279760771972868,
      "loss": 0.2545,
      "num_input_tokens_seen": 21277792,
      "step": 27900
    },
    {
      "epoch": 58.014553014553016,
      "grad_norm": 7.99032422946766e-05,
      "learning_rate": 0.06274968626147688,
      "loss": 0.2557,
      "num_input_tokens_seen": 21281664,
      "step": 27905
    },
    {
      "epoch": 58.024948024948024,
      "grad_norm": 0.00011803515371866524,
      "learning_rate": 0.06270177825831993,
      "loss": 0.2572,
      "num_input_tokens_seen": 21285504,
      "step": 27910
    },
    {
      "epoch": 58.03534303534303,
      "grad_norm": 0.00018923603056464344,
      "learning_rate": 0.06265388371764587,
      "loss": 0.2737,
      "num_input_tokens_seen": 21289312,
      "step": 27915
    },
    {
      "epoch": 58.04573804573805,
      "grad_norm": 0.0004680600541178137,
      "learning_rate": 0.0626060026468406,
      "loss": 0.2679,
      "num_input_tokens_seen": 21293120,
      "step": 27920
    },
    {
      "epoch": 58.056133056133056,
      "grad_norm": 0.0001631857012398541,
      "learning_rate": 0.06255813505328794,
      "loss": 0.2546,
      "num_input_tokens_seen": 21297152,
      "step": 27925
    },
    {
      "epoch": 58.066528066528065,
      "grad_norm": 0.0002768254198599607,
      "learning_rate": 0.06251028094436978,
      "loss": 0.2837,
      "num_input_tokens_seen": 21300896,
      "step": 27930
    },
    {
      "epoch": 58.07692307692308,
      "grad_norm": 0.0006342835840769112,
      "learning_rate": 0.06246244032746568,
      "loss": 0.2775,
      "num_input_tokens_seen": 21304704,
      "step": 27935
    },
    {
      "epoch": 58.08731808731809,
      "grad_norm": 0.00023807137040421367,
      "learning_rate": 0.06241461320995342,
      "loss": 0.2729,
      "num_input_tokens_seen": 21308512,
      "step": 27940
    },
    {
      "epoch": 58.0977130977131,
      "grad_norm": 0.0005684626521542668,
      "learning_rate": 0.062366799599208426,
      "loss": 0.2704,
      "num_input_tokens_seen": 21312448,
      "step": 27945
    },
    {
      "epoch": 58.108108108108105,
      "grad_norm": 6.058826329535805e-05,
      "learning_rate": 0.06231899950260418,
      "loss": 0.2559,
      "num_input_tokens_seen": 21316288,
      "step": 27950
    },
    {
      "epoch": 58.11850311850312,
      "grad_norm": 0.0005555640091188252,
      "learning_rate": 0.06227121292751214,
      "loss": 0.2737,
      "num_input_tokens_seen": 21320128,
      "step": 27955
    },
    {
      "epoch": 58.12889812889813,
      "grad_norm": 0.0012029018253087997,
      "learning_rate": 0.062223439881301496,
      "loss": 0.2763,
      "num_input_tokens_seen": 21323904,
      "step": 27960
    },
    {
      "epoch": 58.13929313929314,
      "grad_norm": 0.0002661730395630002,
      "learning_rate": 0.06217568037133948,
      "loss": 0.2586,
      "num_input_tokens_seen": 21327584,
      "step": 27965
    },
    {
      "epoch": 58.14968814968815,
      "grad_norm": 0.0002500272821635008,
      "learning_rate": 0.06212793440499126,
      "loss": 0.2553,
      "num_input_tokens_seen": 21331392,
      "step": 27970
    },
    {
      "epoch": 58.16008316008316,
      "grad_norm": 6.795930676162243e-05,
      "learning_rate": 0.062080201989619783,
      "loss": 0.2655,
      "num_input_tokens_seen": 21335168,
      "step": 27975
    },
    {
      "epoch": 58.17047817047817,
      "grad_norm": 9.237636550096795e-05,
      "learning_rate": 0.062032483132586094,
      "loss": 0.2771,
      "num_input_tokens_seen": 21339104,
      "step": 27980
    },
    {
      "epoch": 58.18087318087318,
      "grad_norm": 0.0004958044737577438,
      "learning_rate": 0.0619847778412489,
      "loss": 0.2817,
      "num_input_tokens_seen": 21342880,
      "step": 27985
    },
    {
      "epoch": 58.19126819126819,
      "grad_norm": 0.0006020687287673354,
      "learning_rate": 0.06193708612296509,
      "loss": 0.2499,
      "num_input_tokens_seen": 21346624,
      "step": 27990
    },
    {
      "epoch": 58.2016632016632,
      "grad_norm": 7.999097579158843e-05,
      "learning_rate": 0.06188940798508923,
      "loss": 0.2542,
      "num_input_tokens_seen": 21350400,
      "step": 27995
    },
    {
      "epoch": 58.21205821205821,
      "grad_norm": 0.00011698801245074719,
      "learning_rate": 0.06184174343497397,
      "loss": 0.2667,
      "num_input_tokens_seen": 21354208,
      "step": 28000
    },
    {
      "epoch": 58.21205821205821,
      "eval_loss": 0.2471611350774765,
      "eval_runtime": 13.4136,
      "eval_samples_per_second": 63.816,
      "eval_steps_per_second": 15.954,
      "num_input_tokens_seen": 21354208,
      "step": 28000
    },
    {
      "epoch": 58.222453222453225,
      "grad_norm": 0.0003057768917642534,
      "learning_rate": 0.061794092479969726,
      "loss": 0.274,
      "num_input_tokens_seen": 21357952,
      "step": 28005
    },
    {
      "epoch": 58.232848232848234,
      "grad_norm": 0.0005550442729145288,
      "learning_rate": 0.06174645512742485,
      "loss": 0.2374,
      "num_input_tokens_seen": 21361664,
      "step": 28010
    },
    {
      "epoch": 58.24324324324324,
      "grad_norm": 0.0001592139306012541,
      "learning_rate": 0.06169883138468565,
      "loss": 0.2572,
      "num_input_tokens_seen": 21365504,
      "step": 28015
    },
    {
      "epoch": 58.25363825363825,
      "grad_norm": 4.640250699594617e-05,
      "learning_rate": 0.06165122125909637,
      "loss": 0.2446,
      "num_input_tokens_seen": 21369120,
      "step": 28020
    },
    {
      "epoch": 58.264033264033266,
      "grad_norm": 0.00048565768520347774,
      "learning_rate": 0.061603624757998965,
      "loss": 0.2686,
      "num_input_tokens_seen": 21373024,
      "step": 28025
    },
    {
      "epoch": 58.274428274428274,
      "grad_norm": 0.0001465724635636434,
      "learning_rate": 0.0615560418887335,
      "loss": 0.2691,
      "num_input_tokens_seen": 21376928,
      "step": 28030
    },
    {
      "epoch": 58.28482328482328,
      "grad_norm": 0.0010097576305270195,
      "learning_rate": 0.06150847265863787,
      "loss": 0.2655,
      "num_input_tokens_seen": 21380640,
      "step": 28035
    },
    {
      "epoch": 58.2952182952183,
      "grad_norm": 0.0001094237231882289,
      "learning_rate": 0.061460917075047757,
      "loss": 0.2572,
      "num_input_tokens_seen": 21384544,
      "step": 28040
    },
    {
      "epoch": 58.305613305613306,
      "grad_norm": 0.0004562342946883291,
      "learning_rate": 0.06141337514529694,
      "loss": 0.2551,
      "num_input_tokens_seen": 21388480,
      "step": 28045
    },
    {
      "epoch": 58.316008316008315,
      "grad_norm": 0.00030350618180818856,
      "learning_rate": 0.06136584687671687,
      "loss": 0.2546,
      "num_input_tokens_seen": 21392256,
      "step": 28050
    },
    {
      "epoch": 58.32640332640332,
      "grad_norm": 0.00025551021099090576,
      "learning_rate": 0.061318332276637064,
      "loss": 0.253,
      "num_input_tokens_seen": 21396064,
      "step": 28055
    },
    {
      "epoch": 58.33679833679834,
      "grad_norm": 0.00012534890265669674,
      "learning_rate": 0.06127083135238491,
      "loss": 0.2613,
      "num_input_tokens_seen": 21399776,
      "step": 28060
    },
    {
      "epoch": 58.34719334719335,
      "grad_norm": 0.0003287180734332651,
      "learning_rate": 0.06122334411128555,
      "loss": 0.2569,
      "num_input_tokens_seen": 21403424,
      "step": 28065
    },
    {
      "epoch": 58.357588357588355,
      "grad_norm": 0.00011033266491722316,
      "learning_rate": 0.06117587056066223,
      "loss": 0.2773,
      "num_input_tokens_seen": 21407232,
      "step": 28070
    },
    {
      "epoch": 58.36798336798337,
      "grad_norm": 0.00011290735710645095,
      "learning_rate": 0.06112841070783589,
      "loss": 0.2593,
      "num_input_tokens_seen": 21411072,
      "step": 28075
    },
    {
      "epoch": 58.37837837837838,
      "grad_norm": 0.00029994198121130466,
      "learning_rate": 0.061080964560125406,
      "loss": 0.2812,
      "num_input_tokens_seen": 21415008,
      "step": 28080
    },
    {
      "epoch": 58.38877338877339,
      "grad_norm": 0.000195614222320728,
      "learning_rate": 0.06103353212484766,
      "loss": 0.284,
      "num_input_tokens_seen": 21418784,
      "step": 28085
    },
    {
      "epoch": 58.3991683991684,
      "grad_norm": 0.0003068026271648705,
      "learning_rate": 0.06098611340931722,
      "loss": 0.2571,
      "num_input_tokens_seen": 21422528,
      "step": 28090
    },
    {
      "epoch": 58.40956340956341,
      "grad_norm": 0.0003253926697652787,
      "learning_rate": 0.06093870842084672,
      "loss": 0.2646,
      "num_input_tokens_seen": 21426272,
      "step": 28095
    },
    {
      "epoch": 58.41995841995842,
      "grad_norm": 0.0004531065351329744,
      "learning_rate": 0.06089131716674666,
      "loss": 0.2646,
      "num_input_tokens_seen": 21430112,
      "step": 28100
    },
    {
      "epoch": 58.43035343035343,
      "grad_norm": 0.00031820303411222994,
      "learning_rate": 0.060843939654325226,
      "loss": 0.247,
      "num_input_tokens_seen": 21434048,
      "step": 28105
    },
    {
      "epoch": 58.44074844074844,
      "grad_norm": 5.92818578297738e-05,
      "learning_rate": 0.06079657589088873,
      "loss": 0.2556,
      "num_input_tokens_seen": 21437952,
      "step": 28110
    },
    {
      "epoch": 58.45114345114345,
      "grad_norm": 0.00018259974604006857,
      "learning_rate": 0.06074922588374126,
      "loss": 0.2659,
      "num_input_tokens_seen": 21441728,
      "step": 28115
    },
    {
      "epoch": 58.46153846153846,
      "grad_norm": 0.0004484377568587661,
      "learning_rate": 0.06070188964018472,
      "loss": 0.2485,
      "num_input_tokens_seen": 21445536,
      "step": 28120
    },
    {
      "epoch": 58.471933471933475,
      "grad_norm": 0.00017381951329298317,
      "learning_rate": 0.06065456716751902,
      "loss": 0.2752,
      "num_input_tokens_seen": 21449376,
      "step": 28125
    },
    {
      "epoch": 58.482328482328484,
      "grad_norm": 0.0003474741824902594,
      "learning_rate": 0.06060725847304182,
      "loss": 0.2345,
      "num_input_tokens_seen": 21453344,
      "step": 28130
    },
    {
      "epoch": 58.49272349272349,
      "grad_norm": 0.00015435223758686334,
      "learning_rate": 0.06055996356404877,
      "loss": 0.2644,
      "num_input_tokens_seen": 21457184,
      "step": 28135
    },
    {
      "epoch": 58.5031185031185,
      "grad_norm": 0.0003369873738847673,
      "learning_rate": 0.06051268244783327,
      "loss": 0.272,
      "num_input_tokens_seen": 21460992,
      "step": 28140
    },
    {
      "epoch": 58.513513513513516,
      "grad_norm": 0.00045230990508571267,
      "learning_rate": 0.06046541513168676,
      "loss": 0.245,
      "num_input_tokens_seen": 21464896,
      "step": 28145
    },
    {
      "epoch": 58.523908523908524,
      "grad_norm": 0.0002394439943600446,
      "learning_rate": 0.060418161622898356,
      "loss": 0.27,
      "num_input_tokens_seen": 21468640,
      "step": 28150
    },
    {
      "epoch": 58.53430353430353,
      "grad_norm": 8.886944851838052e-05,
      "learning_rate": 0.06037092192875521,
      "loss": 0.2782,
      "num_input_tokens_seen": 21472288,
      "step": 28155
    },
    {
      "epoch": 58.54469854469855,
      "grad_norm": 0.00027448433684185147,
      "learning_rate": 0.060323696056542225,
      "loss": 0.2716,
      "num_input_tokens_seen": 21476032,
      "step": 28160
    },
    {
      "epoch": 58.555093555093556,
      "grad_norm": 0.0004599363310262561,
      "learning_rate": 0.06027648401354229,
      "loss": 0.2592,
      "num_input_tokens_seen": 21479904,
      "step": 28165
    },
    {
      "epoch": 58.565488565488565,
      "grad_norm": 5.780800711363554e-05,
      "learning_rate": 0.06022928580703601,
      "loss": 0.2674,
      "num_input_tokens_seen": 21483776,
      "step": 28170
    },
    {
      "epoch": 58.57588357588357,
      "grad_norm": 0.0006084410124458373,
      "learning_rate": 0.060182101444301986,
      "loss": 0.2877,
      "num_input_tokens_seen": 21487744,
      "step": 28175
    },
    {
      "epoch": 58.58627858627859,
      "grad_norm": 0.0003769618342630565,
      "learning_rate": 0.06013493093261669,
      "loss": 0.2543,
      "num_input_tokens_seen": 21491456,
      "step": 28180
    },
    {
      "epoch": 58.5966735966736,
      "grad_norm": 0.00043927814112976193,
      "learning_rate": 0.06008777427925432,
      "loss": 0.2603,
      "num_input_tokens_seen": 21495520,
      "step": 28185
    },
    {
      "epoch": 58.607068607068605,
      "grad_norm": 0.00027972384123131633,
      "learning_rate": 0.06004063149148705,
      "loss": 0.2649,
      "num_input_tokens_seen": 21499296,
      "step": 28190
    },
    {
      "epoch": 58.61746361746362,
      "grad_norm": 0.0005749806296080351,
      "learning_rate": 0.05999350257658497,
      "loss": 0.2436,
      "num_input_tokens_seen": 21503104,
      "step": 28195
    },
    {
      "epoch": 58.62785862785863,
      "grad_norm": 0.0001661076967138797,
      "learning_rate": 0.05994638754181582,
      "loss": 0.2785,
      "num_input_tokens_seen": 21506752,
      "step": 28200
    },
    {
      "epoch": 58.62785862785863,
      "eval_loss": 0.24804453551769257,
      "eval_runtime": 13.4855,
      "eval_samples_per_second": 63.476,
      "eval_steps_per_second": 15.869,
      "num_input_tokens_seen": 21506752,
      "step": 28200
    },
    {
      "epoch": 58.63825363825364,
      "grad_norm": 8.766123210079968e-05,
      "learning_rate": 0.059899286394445445,
      "loss": 0.2677,
      "num_input_tokens_seen": 21510592,
      "step": 28205
    },
    {
      "epoch": 58.648648648648646,
      "grad_norm": 0.0002547898911871016,
      "learning_rate": 0.059852199141737346,
      "loss": 0.2834,
      "num_input_tokens_seen": 21514560,
      "step": 28210
    },
    {
      "epoch": 58.65904365904366,
      "grad_norm": 0.00027078064158558846,
      "learning_rate": 0.05980512579095304,
      "loss": 0.2383,
      "num_input_tokens_seen": 21518240,
      "step": 28215
    },
    {
      "epoch": 58.66943866943867,
      "grad_norm": 0.00023122024140320718,
      "learning_rate": 0.05975806634935181,
      "loss": 0.261,
      "num_input_tokens_seen": 21522048,
      "step": 28220
    },
    {
      "epoch": 58.67983367983368,
      "grad_norm": 0.00047529683797620237,
      "learning_rate": 0.05971102082419076,
      "loss": 0.2535,
      "num_input_tokens_seen": 21525728,
      "step": 28225
    },
    {
      "epoch": 58.69022869022869,
      "grad_norm": 6.772636697860435e-05,
      "learning_rate": 0.05966398922272492,
      "loss": 0.2898,
      "num_input_tokens_seen": 21529632,
      "step": 28230
    },
    {
      "epoch": 58.7006237006237,
      "grad_norm": 5.464633431984112e-05,
      "learning_rate": 0.059616971552207236,
      "loss": 0.258,
      "num_input_tokens_seen": 21533440,
      "step": 28235
    },
    {
      "epoch": 58.71101871101871,
      "grad_norm": 0.0004955774056725204,
      "learning_rate": 0.059569967819888305,
      "loss": 0.2572,
      "num_input_tokens_seen": 21537312,
      "step": 28240
    },
    {
      "epoch": 58.72141372141372,
      "grad_norm": 0.00041029872954823077,
      "learning_rate": 0.05952297803301681,
      "loss": 0.2752,
      "num_input_tokens_seen": 21541056,
      "step": 28245
    },
    {
      "epoch": 58.731808731808734,
      "grad_norm": 0.0002988222404383123,
      "learning_rate": 0.059476002198839056,
      "loss": 0.2683,
      "num_input_tokens_seen": 21544832,
      "step": 28250
    },
    {
      "epoch": 58.74220374220374,
      "grad_norm": 0.00037701110704801977,
      "learning_rate": 0.05942904032459935,
      "loss": 0.2838,
      "num_input_tokens_seen": 21548640,
      "step": 28255
    },
    {
      "epoch": 58.75259875259875,
      "grad_norm": 4.13027846661862e-05,
      "learning_rate": 0.05938209241753987,
      "loss": 0.2481,
      "num_input_tokens_seen": 21552352,
      "step": 28260
    },
    {
      "epoch": 58.762993762993766,
      "grad_norm": 0.0003292766341473907,
      "learning_rate": 0.05933515848490046,
      "loss": 0.2476,
      "num_input_tokens_seen": 21556160,
      "step": 28265
    },
    {
      "epoch": 58.773388773388774,
      "grad_norm": 0.0003277112846262753,
      "learning_rate": 0.059288238533918985,
      "loss": 0.2911,
      "num_input_tokens_seen": 21559936,
      "step": 28270
    },
    {
      "epoch": 58.78378378378378,
      "grad_norm": 0.00011082505079684779,
      "learning_rate": 0.05924133257183113,
      "loss": 0.2626,
      "num_input_tokens_seen": 21563712,
      "step": 28275
    },
    {
      "epoch": 58.79417879417879,
      "grad_norm": 0.000592010619584471,
      "learning_rate": 0.059194440605870285,
      "loss": 0.26,
      "num_input_tokens_seen": 21567776,
      "step": 28280
    },
    {
      "epoch": 58.804573804573806,
      "grad_norm": 0.00010882051719818264,
      "learning_rate": 0.059147562643267884,
      "loss": 0.2673,
      "num_input_tokens_seen": 21571680,
      "step": 28285
    },
    {
      "epoch": 58.814968814968815,
      "grad_norm": 0.0001854533766163513,
      "learning_rate": 0.059100698691253055,
      "loss": 0.2506,
      "num_input_tokens_seen": 21575520,
      "step": 28290
    },
    {
      "epoch": 58.82536382536382,
      "grad_norm": 0.00020167410548310727,
      "learning_rate": 0.05905384875705273,
      "loss": 0.2542,
      "num_input_tokens_seen": 21579296,
      "step": 28295
    },
    {
      "epoch": 58.83575883575884,
      "grad_norm": 0.0006994350114837289,
      "learning_rate": 0.05900701284789189,
      "loss": 0.2452,
      "num_input_tokens_seen": 21583136,
      "step": 28300
    },
    {
      "epoch": 58.84615384615385,
      "grad_norm": 0.00023889818112365901,
      "learning_rate": 0.058960190970993115,
      "loss": 0.2859,
      "num_input_tokens_seen": 21586944,
      "step": 28305
    },
    {
      "epoch": 58.856548856548855,
      "grad_norm": 0.0003332647029310465,
      "learning_rate": 0.058913383133576955,
      "loss": 0.2611,
      "num_input_tokens_seen": 21590752,
      "step": 28310
    },
    {
      "epoch": 58.86694386694387,
      "grad_norm": 0.0002455249195918441,
      "learning_rate": 0.05886658934286185,
      "loss": 0.2523,
      "num_input_tokens_seen": 21594496,
      "step": 28315
    },
    {
      "epoch": 58.87733887733888,
      "grad_norm": 0.0002897169906646013,
      "learning_rate": 0.058819809606063846,
      "loss": 0.2822,
      "num_input_tokens_seen": 21598304,
      "step": 28320
    },
    {
      "epoch": 58.88773388773389,
      "grad_norm": 0.00033308478305116296,
      "learning_rate": 0.05877304393039711,
      "loss": 0.2995,
      "num_input_tokens_seen": 21602176,
      "step": 28325
    },
    {
      "epoch": 58.898128898128896,
      "grad_norm": 0.00018867666949518025,
      "learning_rate": 0.05872629232307338,
      "loss": 0.2749,
      "num_input_tokens_seen": 21606016,
      "step": 28330
    },
    {
      "epoch": 58.90852390852391,
      "grad_norm": 0.0005097058019600809,
      "learning_rate": 0.05867955479130239,
      "loss": 0.282,
      "num_input_tokens_seen": 21609792,
      "step": 28335
    },
    {
      "epoch": 58.91891891891892,
      "grad_norm": 0.00013791833771392703,
      "learning_rate": 0.058632831342291705,
      "loss": 0.2692,
      "num_input_tokens_seen": 21613664,
      "step": 28340
    },
    {
      "epoch": 58.92931392931393,
      "grad_norm": 0.00035076276981271803,
      "learning_rate": 0.05858612198324655,
      "loss": 0.2796,
      "num_input_tokens_seen": 21617504,
      "step": 28345
    },
    {
      "epoch": 58.93970893970894,
      "grad_norm": 0.00025151160662062466,
      "learning_rate": 0.05853942672137025,
      "loss": 0.263,
      "num_input_tokens_seen": 21621216,
      "step": 28350
    },
    {
      "epoch": 58.95010395010395,
      "grad_norm": 0.00021297895000316203,
      "learning_rate": 0.05849274556386363,
      "loss": 0.242,
      "num_input_tokens_seen": 21625024,
      "step": 28355
    },
    {
      "epoch": 58.96049896049896,
      "grad_norm": 0.0002409277221886441,
      "learning_rate": 0.05844607851792567,
      "loss": 0.2574,
      "num_input_tokens_seen": 21628928,
      "step": 28360
    },
    {
      "epoch": 58.97089397089397,
      "grad_norm": 0.0004970143199898303,
      "learning_rate": 0.058399425590752924,
      "loss": 0.2959,
      "num_input_tokens_seen": 21632800,
      "step": 28365
    },
    {
      "epoch": 58.981288981288984,
      "grad_norm": 0.00031185749685391784,
      "learning_rate": 0.05835278678953985,
      "loss": 0.275,
      "num_input_tokens_seen": 21636672,
      "step": 28370
    },
    {
      "epoch": 58.99168399168399,
      "grad_norm": 0.0002325763343833387,
      "learning_rate": 0.05830616212147874,
      "loss": 0.2585,
      "num_input_tokens_seen": 21640352,
      "step": 28375
    },
    {
      "epoch": 59.002079002079,
      "grad_norm": 0.0002094060619128868,
      "learning_rate": 0.058259551593759784,
      "loss": 0.2815,
      "num_input_tokens_seen": 21644240,
      "step": 28380
    },
    {
      "epoch": 59.012474012474016,
      "grad_norm": 0.0004323652829043567,
      "learning_rate": 0.058212955213570804,
      "loss": 0.2699,
      "num_input_tokens_seen": 21648176,
      "step": 28385
    },
    {
      "epoch": 59.022869022869024,
      "grad_norm": 0.0009287319844588637,
      "learning_rate": 0.0581663729880976,
      "loss": 0.2752,
      "num_input_tokens_seen": 21651984,
      "step": 28390
    },
    {
      "epoch": 59.03326403326403,
      "grad_norm": 0.00040505407378077507,
      "learning_rate": 0.05811980492452379,
      "loss": 0.2769,
      "num_input_tokens_seen": 21655856,
      "step": 28395
    },
    {
      "epoch": 59.04365904365904,
      "grad_norm": 0.00010733300587162375,
      "learning_rate": 0.058073251030030644,
      "loss": 0.2669,
      "num_input_tokens_seen": 21659696,
      "step": 28400
    },
    {
      "epoch": 59.04365904365904,
      "eval_loss": 0.25756606459617615,
      "eval_runtime": 13.4052,
      "eval_samples_per_second": 63.856,
      "eval_steps_per_second": 15.964,
      "num_input_tokens_seen": 21659696,
      "step": 28400
    },
    {
      "epoch": 59.054054054054056,
      "grad_norm": 0.0002945765445474535,
      "learning_rate": 0.05802671131179747,
      "loss": 0.2573,
      "num_input_tokens_seen": 21663600,
      "step": 28405
    },
    {
      "epoch": 59.064449064449065,
      "grad_norm": 4.221084600430913e-05,
      "learning_rate": 0.057980185777001154,
      "loss": 0.2706,
      "num_input_tokens_seen": 21667376,
      "step": 28410
    },
    {
      "epoch": 59.07484407484407,
      "grad_norm": 9.901568409986794e-05,
      "learning_rate": 0.057933674432816606,
      "loss": 0.28,
      "num_input_tokens_seen": 21671312,
      "step": 28415
    },
    {
      "epoch": 59.08523908523909,
      "grad_norm": 9.723401308292523e-05,
      "learning_rate": 0.05788717728641648,
      "loss": 0.2674,
      "num_input_tokens_seen": 21675152,
      "step": 28420
    },
    {
      "epoch": 59.0956340956341,
      "grad_norm": 2.7791116735897958e-05,
      "learning_rate": 0.057840694344971126,
      "loss": 0.2833,
      "num_input_tokens_seen": 21678960,
      "step": 28425
    },
    {
      "epoch": 59.106029106029105,
      "grad_norm": 5.818685895064846e-05,
      "learning_rate": 0.0577942256156489,
      "loss": 0.2644,
      "num_input_tokens_seen": 21682864,
      "step": 28430
    },
    {
      "epoch": 59.11642411642411,
      "grad_norm": 0.00011367109982529655,
      "learning_rate": 0.057747771105615804,
      "loss": 0.2633,
      "num_input_tokens_seen": 21686640,
      "step": 28435
    },
    {
      "epoch": 59.12681912681913,
      "grad_norm": 0.0004155321803409606,
      "learning_rate": 0.05770133082203568,
      "loss": 0.2426,
      "num_input_tokens_seen": 21690416,
      "step": 28440
    },
    {
      "epoch": 59.13721413721414,
      "grad_norm": 0.0002099186385748908,
      "learning_rate": 0.0576549047720703,
      "loss": 0.2735,
      "num_input_tokens_seen": 21694128,
      "step": 28445
    },
    {
      "epoch": 59.147609147609145,
      "grad_norm": 9.252870222553611e-05,
      "learning_rate": 0.05760849296287902,
      "loss": 0.2595,
      "num_input_tokens_seen": 21697776,
      "step": 28450
    },
    {
      "epoch": 59.15800415800416,
      "grad_norm": 6.353397475322708e-05,
      "learning_rate": 0.05756209540161919,
      "loss": 0.2932,
      "num_input_tokens_seen": 21701552,
      "step": 28455
    },
    {
      "epoch": 59.16839916839917,
      "grad_norm": 0.0002833329781424254,
      "learning_rate": 0.05751571209544595,
      "loss": 0.266,
      "num_input_tokens_seen": 21705328,
      "step": 28460
    },
    {
      "epoch": 59.17879417879418,
      "grad_norm": 0.00023166960454545915,
      "learning_rate": 0.057469343051512085,
      "loss": 0.2398,
      "num_input_tokens_seen": 21708976,
      "step": 28465
    },
    {
      "epoch": 59.189189189189186,
      "grad_norm": 0.00025007323711179197,
      "learning_rate": 0.057422988276968324,
      "loss": 0.2739,
      "num_input_tokens_seen": 21712912,
      "step": 28470
    },
    {
      "epoch": 59.1995841995842,
      "grad_norm": 0.00024293549358844757,
      "learning_rate": 0.05737664777896323,
      "loss": 0.2551,
      "num_input_tokens_seen": 21716656,
      "step": 28475
    },
    {
      "epoch": 59.20997920997921,
      "grad_norm": 0.00024286335974466056,
      "learning_rate": 0.057330321564642975,
      "loss": 0.249,
      "num_input_tokens_seen": 21720432,
      "step": 28480
    },
    {
      "epoch": 59.22037422037422,
      "grad_norm": 0.00023277421132661402,
      "learning_rate": 0.05728400964115174,
      "loss": 0.2523,
      "num_input_tokens_seen": 21724240,
      "step": 28485
    },
    {
      "epoch": 59.23076923076923,
      "grad_norm": 3.65491105185356e-05,
      "learning_rate": 0.057237712015631305,
      "loss": 0.2625,
      "num_input_tokens_seen": 21728144,
      "step": 28490
    },
    {
      "epoch": 59.24116424116424,
      "grad_norm": 4.9077800213126466e-05,
      "learning_rate": 0.057191428695221425,
      "loss": 0.2609,
      "num_input_tokens_seen": 21731920,
      "step": 28495
    },
    {
      "epoch": 59.25155925155925,
      "grad_norm": 8.089486800599843e-05,
      "learning_rate": 0.05714515968705958,
      "loss": 0.27,
      "num_input_tokens_seen": 21735760,
      "step": 28500
    },
    {
      "epoch": 59.26195426195426,
      "grad_norm": 0.0001673252700129524,
      "learning_rate": 0.05709890499828099,
      "loss": 0.2628,
      "num_input_tokens_seen": 21739536,
      "step": 28505
    },
    {
      "epoch": 59.272349272349274,
      "grad_norm": 0.00032226540497504175,
      "learning_rate": 0.05705266463601868,
      "loss": 0.2445,
      "num_input_tokens_seen": 21743280,
      "step": 28510
    },
    {
      "epoch": 59.28274428274428,
      "grad_norm": 0.0003142687492072582,
      "learning_rate": 0.057006438607403565,
      "loss": 0.241,
      "num_input_tokens_seen": 21746992,
      "step": 28515
    },
    {
      "epoch": 59.29313929313929,
      "grad_norm": 0.0002858842199202627,
      "learning_rate": 0.056960226919564205,
      "loss": 0.257,
      "num_input_tokens_seen": 21750800,
      "step": 28520
    },
    {
      "epoch": 59.303534303534306,
      "grad_norm": 0.0004515312029980123,
      "learning_rate": 0.05691402957962713,
      "loss": 0.2857,
      "num_input_tokens_seen": 21754576,
      "step": 28525
    },
    {
      "epoch": 59.313929313929314,
      "grad_norm": 0.0004459033953025937,
      "learning_rate": 0.05686784659471642,
      "loss": 0.2688,
      "num_input_tokens_seen": 21758352,
      "step": 28530
    },
    {
      "epoch": 59.32432432432432,
      "grad_norm": 0.0004037866892758757,
      "learning_rate": 0.056821677971954136,
      "loss": 0.252,
      "num_input_tokens_seen": 21762224,
      "step": 28535
    },
    {
      "epoch": 59.33471933471934,
      "grad_norm": 0.0002533178194426,
      "learning_rate": 0.05677552371846012,
      "loss": 0.2607,
      "num_input_tokens_seen": 21766000,
      "step": 28540
    },
    {
      "epoch": 59.34511434511435,
      "grad_norm": 8.1624835729599e-05,
      "learning_rate": 0.05672938384135182,
      "loss": 0.2619,
      "num_input_tokens_seen": 21769936,
      "step": 28545
    },
    {
      "epoch": 59.355509355509355,
      "grad_norm": 0.00014089929754845798,
      "learning_rate": 0.05668325834774465,
      "loss": 0.2649,
      "num_input_tokens_seen": 21773776,
      "step": 28550
    },
    {
      "epoch": 59.36590436590436,
      "grad_norm": 0.0004051189753226936,
      "learning_rate": 0.05663714724475177,
      "loss": 0.2502,
      "num_input_tokens_seen": 21777616,
      "step": 28555
    },
    {
      "epoch": 59.37629937629938,
      "grad_norm": 0.00048392615281045437,
      "learning_rate": 0.05659105053948403,
      "loss": 0.2574,
      "num_input_tokens_seen": 21781360,
      "step": 28560
    },
    {
      "epoch": 59.38669438669439,
      "grad_norm": 0.00015165570948738605,
      "learning_rate": 0.056544968239050176,
      "loss": 0.277,
      "num_input_tokens_seen": 21785104,
      "step": 28565
    },
    {
      "epoch": 59.397089397089395,
      "grad_norm": 0.0005822497187182307,
      "learning_rate": 0.056498900350556616,
      "loss": 0.249,
      "num_input_tokens_seen": 21788816,
      "step": 28570
    },
    {
      "epoch": 59.40748440748441,
      "grad_norm": 0.0005350075662136078,
      "learning_rate": 0.05645284688110766,
      "loss": 0.2766,
      "num_input_tokens_seen": 21792592,
      "step": 28575
    },
    {
      "epoch": 59.41787941787942,
      "grad_norm": 0.00042768902494572103,
      "learning_rate": 0.05640680783780532,
      "loss": 0.2606,
      "num_input_tokens_seen": 21796400,
      "step": 28580
    },
    {
      "epoch": 59.42827442827443,
      "grad_norm": 0.0002751356805674732,
      "learning_rate": 0.056360783227749324,
      "loss": 0.2648,
      "num_input_tokens_seen": 21800080,
      "step": 28585
    },
    {
      "epoch": 59.438669438669436,
      "grad_norm": 0.0008365960675291717,
      "learning_rate": 0.05631477305803728,
      "loss": 0.2638,
      "num_input_tokens_seen": 21803728,
      "step": 28590
    },
    {
      "epoch": 59.44906444906445,
      "grad_norm": 0.00010132075840374455,
      "learning_rate": 0.05626877733576462,
      "loss": 0.2327,
      "num_input_tokens_seen": 21807728,
      "step": 28595
    },
    {
      "epoch": 59.45945945945946,
      "grad_norm": 0.00025405475753359497,
      "learning_rate": 0.05622279606802435,
      "loss": 0.2614,
      "num_input_tokens_seen": 21811600,
      "step": 28600
    },
    {
      "epoch": 59.45945945945946,
      "eval_loss": 0.24791602790355682,
      "eval_runtime": 13.4042,
      "eval_samples_per_second": 63.861,
      "eval_steps_per_second": 15.965,
      "num_input_tokens_seen": 21811600,
      "step": 28600
    },
    {
      "epoch": 59.46985446985447,
      "grad_norm": 0.00014457415090873837,
      "learning_rate": 0.05617682926190744,
      "loss": 0.2639,
      "num_input_tokens_seen": 21815344,
      "step": 28605
    },
    {
      "epoch": 59.48024948024948,
      "grad_norm": 0.00020768803369719535,
      "learning_rate": 0.05613087692450248,
      "loss": 0.2842,
      "num_input_tokens_seen": 21819312,
      "step": 28610
    },
    {
      "epoch": 59.49064449064449,
      "grad_norm": 0.00032393611036241055,
      "learning_rate": 0.05608493906289592,
      "loss": 0.2642,
      "num_input_tokens_seen": 21823184,
      "step": 28615
    },
    {
      "epoch": 59.5010395010395,
      "grad_norm": 0.00032815185841172934,
      "learning_rate": 0.05603901568417201,
      "loss": 0.2711,
      "num_input_tokens_seen": 21827152,
      "step": 28620
    },
    {
      "epoch": 59.51143451143451,
      "grad_norm": 0.000411802320741117,
      "learning_rate": 0.055993106795412625,
      "loss": 0.2442,
      "num_input_tokens_seen": 21831184,
      "step": 28625
    },
    {
      "epoch": 59.521829521829524,
      "grad_norm": 9.419520210940391e-05,
      "learning_rate": 0.05594721240369759,
      "loss": 0.273,
      "num_input_tokens_seen": 21834992,
      "step": 28630
    },
    {
      "epoch": 59.53222453222453,
      "grad_norm": 0.0004108079883735627,
      "learning_rate": 0.055901332516104296,
      "loss": 0.2675,
      "num_input_tokens_seen": 21838896,
      "step": 28635
    },
    {
      "epoch": 59.54261954261954,
      "grad_norm": 0.00042091155773960054,
      "learning_rate": 0.05585546713970804,
      "loss": 0.2484,
      "num_input_tokens_seen": 21842992,
      "step": 28640
    },
    {
      "epoch": 59.553014553014556,
      "grad_norm": 8.988139597931877e-05,
      "learning_rate": 0.05580961628158189,
      "loss": 0.2634,
      "num_input_tokens_seen": 21846672,
      "step": 28645
    },
    {
      "epoch": 59.563409563409564,
      "grad_norm": 9.53419366851449e-05,
      "learning_rate": 0.05576377994879659,
      "loss": 0.2711,
      "num_input_tokens_seen": 21850480,
      "step": 28650
    },
    {
      "epoch": 59.57380457380457,
      "grad_norm": 0.0003029521321877837,
      "learning_rate": 0.05571795814842063,
      "loss": 0.2816,
      "num_input_tokens_seen": 21854288,
      "step": 28655
    },
    {
      "epoch": 59.58419958419958,
      "grad_norm": 0.0004547022108454257,
      "learning_rate": 0.05567215088752037,
      "loss": 0.2727,
      "num_input_tokens_seen": 21858064,
      "step": 28660
    },
    {
      "epoch": 59.5945945945946,
      "grad_norm": 0.00023953290656208992,
      "learning_rate": 0.05562635817315981,
      "loss": 0.2757,
      "num_input_tokens_seen": 21861808,
      "step": 28665
    },
    {
      "epoch": 59.604989604989605,
      "grad_norm": 0.00022702181013301015,
      "learning_rate": 0.05558058001240083,
      "loss": 0.2668,
      "num_input_tokens_seen": 21865744,
      "step": 28670
    },
    {
      "epoch": 59.61538461538461,
      "grad_norm": 5.641702955472283e-05,
      "learning_rate": 0.055534816412302915,
      "loss": 0.2754,
      "num_input_tokens_seen": 21869360,
      "step": 28675
    },
    {
      "epoch": 59.62577962577963,
      "grad_norm": 0.00037678508670069277,
      "learning_rate": 0.055489067379923436,
      "loss": 0.2734,
      "num_input_tokens_seen": 21872976,
      "step": 28680
    },
    {
      "epoch": 59.63617463617464,
      "grad_norm": 4.154805719736032e-05,
      "learning_rate": 0.055443332922317505,
      "loss": 0.2673,
      "num_input_tokens_seen": 21876848,
      "step": 28685
    },
    {
      "epoch": 59.646569646569645,
      "grad_norm": 0.0001509753637947142,
      "learning_rate": 0.055397613046537876,
      "loss": 0.2692,
      "num_input_tokens_seen": 21880624,
      "step": 28690
    },
    {
      "epoch": 59.656964656964654,
      "grad_norm": 0.00053272774675861,
      "learning_rate": 0.055351907759635145,
      "loss": 0.2567,
      "num_input_tokens_seen": 21884464,
      "step": 28695
    },
    {
      "epoch": 59.66735966735967,
      "grad_norm": 0.00011537809041328728,
      "learning_rate": 0.05530621706865772,
      "loss": 0.2789,
      "num_input_tokens_seen": 21888336,
      "step": 28700
    },
    {
      "epoch": 59.67775467775468,
      "grad_norm": 6.822626892244443e-05,
      "learning_rate": 0.055260540980651564,
      "loss": 0.2539,
      "num_input_tokens_seen": 21892304,
      "step": 28705
    },
    {
      "epoch": 59.688149688149686,
      "grad_norm": 0.00023755435540806502,
      "learning_rate": 0.05521487950266062,
      "loss": 0.2614,
      "num_input_tokens_seen": 21896176,
      "step": 28710
    },
    {
      "epoch": 59.6985446985447,
      "grad_norm": 0.00015982553304638714,
      "learning_rate": 0.055169232641726344,
      "loss": 0.2797,
      "num_input_tokens_seen": 21900016,
      "step": 28715
    },
    {
      "epoch": 59.70893970893971,
      "grad_norm": 0.00027029550983570516,
      "learning_rate": 0.055123600404888166,
      "loss": 0.2566,
      "num_input_tokens_seen": 21903856,
      "step": 28720
    },
    {
      "epoch": 59.71933471933472,
      "grad_norm": 0.0006020541186444461,
      "learning_rate": 0.05507798279918309,
      "loss": 0.2547,
      "num_input_tokens_seen": 21907632,
      "step": 28725
    },
    {
      "epoch": 59.729729729729726,
      "grad_norm": 0.00022213983174879104,
      "learning_rate": 0.0550323798316459,
      "loss": 0.2538,
      "num_input_tokens_seen": 21911472,
      "step": 28730
    },
    {
      "epoch": 59.74012474012474,
      "grad_norm": 0.0003540258330758661,
      "learning_rate": 0.05498679150930916,
      "loss": 0.243,
      "num_input_tokens_seen": 21915088,
      "step": 28735
    },
    {
      "epoch": 59.75051975051975,
      "grad_norm": 3.7383029848570004e-05,
      "learning_rate": 0.05494121783920323,
      "loss": 0.2648,
      "num_input_tokens_seen": 21918800,
      "step": 28740
    },
    {
      "epoch": 59.76091476091476,
      "grad_norm": 9.713468898553401e-05,
      "learning_rate": 0.05489565882835605,
      "loss": 0.2845,
      "num_input_tokens_seen": 21922576,
      "step": 28745
    },
    {
      "epoch": 59.771309771309774,
      "grad_norm": 0.00011383960372768342,
      "learning_rate": 0.05485011448379348,
      "loss": 0.2829,
      "num_input_tokens_seen": 21926448,
      "step": 28750
    },
    {
      "epoch": 59.78170478170478,
      "grad_norm": 0.00040625041583552957,
      "learning_rate": 0.05480458481253893,
      "loss": 0.2548,
      "num_input_tokens_seen": 21930256,
      "step": 28755
    },
    {
      "epoch": 59.79209979209979,
      "grad_norm": 0.00011704157805070281,
      "learning_rate": 0.054759069821613715,
      "loss": 0.2737,
      "num_input_tokens_seen": 21934224,
      "step": 28760
    },
    {
      "epoch": 59.802494802494806,
      "grad_norm": 0.0003836415708065033,
      "learning_rate": 0.05471356951803683,
      "loss": 0.2801,
      "num_input_tokens_seen": 21937968,
      "step": 28765
    },
    {
      "epoch": 59.812889812889814,
      "grad_norm": 7.47228114050813e-05,
      "learning_rate": 0.054668083908824945,
      "loss": 0.2603,
      "num_input_tokens_seen": 21941872,
      "step": 28770
    },
    {
      "epoch": 59.82328482328482,
      "grad_norm": 0.00015790094039402902,
      "learning_rate": 0.054622613000992526,
      "loss": 0.2675,
      "num_input_tokens_seen": 21945584,
      "step": 28775
    },
    {
      "epoch": 59.83367983367983,
      "grad_norm": 0.0001254020753549412,
      "learning_rate": 0.05457715680155182,
      "loss": 0.2751,
      "num_input_tokens_seen": 21949360,
      "step": 28780
    },
    {
      "epoch": 59.84407484407485,
      "grad_norm": 0.00024265359388664365,
      "learning_rate": 0.05453171531751265,
      "loss": 0.2487,
      "num_input_tokens_seen": 21953296,
      "step": 28785
    },
    {
      "epoch": 59.854469854469855,
      "grad_norm": 0.0003725750430021435,
      "learning_rate": 0.05448628855588276,
      "loss": 0.2504,
      "num_input_tokens_seen": 21956912,
      "step": 28790
    },
    {
      "epoch": 59.86486486486486,
      "grad_norm": 0.00011277251178398728,
      "learning_rate": 0.05444087652366746,
      "loss": 0.2559,
      "num_input_tokens_seen": 21960656,
      "step": 28795
    },
    {
      "epoch": 59.87525987525988,
      "grad_norm": 2.916383346018847e-05,
      "learning_rate": 0.05439547922786984,
      "loss": 0.2765,
      "num_input_tokens_seen": 21964272,
      "step": 28800
    },
    {
      "epoch": 59.87525987525988,
      "eval_loss": 0.2477722465991974,
      "eval_runtime": 13.4029,
      "eval_samples_per_second": 63.867,
      "eval_steps_per_second": 15.967,
      "num_input_tokens_seen": 21964272,
      "step": 28800
    },
    {
      "epoch": 59.88565488565489,
      "grad_norm": 0.00024082105665002018,
      "learning_rate": 0.0543500966754908,
      "loss": 0.271,
      "num_input_tokens_seen": 21967952,
      "step": 28805
    },
    {
      "epoch": 59.896049896049895,
      "grad_norm": 0.00011456009087851271,
      "learning_rate": 0.05430472887352882,
      "loss": 0.2321,
      "num_input_tokens_seen": 21971792,
      "step": 28810
    },
    {
      "epoch": 59.906444906444904,
      "grad_norm": 0.00023738724121358246,
      "learning_rate": 0.05425937582898023,
      "loss": 0.2795,
      "num_input_tokens_seen": 21975440,
      "step": 28815
    },
    {
      "epoch": 59.91683991683992,
      "grad_norm": 6.13569063716568e-05,
      "learning_rate": 0.054214037548839085,
      "loss": 0.2535,
      "num_input_tokens_seen": 21979280,
      "step": 28820
    },
    {
      "epoch": 59.92723492723493,
      "grad_norm": 0.0003845040046144277,
      "learning_rate": 0.05416871404009703,
      "loss": 0.2976,
      "num_input_tokens_seen": 21983088,
      "step": 28825
    },
    {
      "epoch": 59.937629937629936,
      "grad_norm": 0.00019938072364311665,
      "learning_rate": 0.054123405309743605,
      "loss": 0.2625,
      "num_input_tokens_seen": 21986896,
      "step": 28830
    },
    {
      "epoch": 59.94802494802495,
      "grad_norm": 0.00038890400901436806,
      "learning_rate": 0.0540781113647659,
      "loss": 0.2778,
      "num_input_tokens_seen": 21990736,
      "step": 28835
    },
    {
      "epoch": 59.95841995841996,
      "grad_norm": 0.00011236857972107828,
      "learning_rate": 0.054032832212148836,
      "loss": 0.2729,
      "num_input_tokens_seen": 21994544,
      "step": 28840
    },
    {
      "epoch": 59.96881496881497,
      "grad_norm": 0.00034014161792583764,
      "learning_rate": 0.0539875678588751,
      "loss": 0.2552,
      "num_input_tokens_seen": 21998288,
      "step": 28845
    },
    {
      "epoch": 59.979209979209976,
      "grad_norm": 0.0002531029167585075,
      "learning_rate": 0.05394231831192492,
      "loss": 0.2591,
      "num_input_tokens_seen": 22002224,
      "step": 28850
    },
    {
      "epoch": 59.98960498960499,
      "grad_norm": 0.00017999463307205588,
      "learning_rate": 0.05389708357827639,
      "loss": 0.2547,
      "num_input_tokens_seen": 22005904,
      "step": 28855
    },
    {
      "epoch": 60.0,
      "grad_norm": 0.00018988997908309102,
      "learning_rate": 0.05385186366490533,
      "loss": 0.2652,
      "num_input_tokens_seen": 22009672,
      "step": 28860
    },
    {
      "epoch": 60.01039501039501,
      "grad_norm": 0.00035744041088037193,
      "learning_rate": 0.053806658578785166,
      "loss": 0.2619,
      "num_input_tokens_seen": 22013448,
      "step": 28865
    },
    {
      "epoch": 60.020790020790024,
      "grad_norm": 0.00014614209067076445,
      "learning_rate": 0.05376146832688705,
      "loss": 0.2467,
      "num_input_tokens_seen": 22017096,
      "step": 28870
    },
    {
      "epoch": 60.03118503118503,
      "grad_norm": 2.8996011678827927e-05,
      "learning_rate": 0.053716292916179964,
      "loss": 0.2277,
      "num_input_tokens_seen": 22020872,
      "step": 28875
    },
    {
      "epoch": 60.04158004158004,
      "grad_norm": 0.0001485464454162866,
      "learning_rate": 0.05367113235363045,
      "loss": 0.2563,
      "num_input_tokens_seen": 22024840,
      "step": 28880
    },
    {
      "epoch": 60.05197505197505,
      "grad_norm": 0.0001977020438062027,
      "learning_rate": 0.05362598664620289,
      "loss": 0.2673,
      "num_input_tokens_seen": 22028744,
      "step": 28885
    },
    {
      "epoch": 60.062370062370064,
      "grad_norm": 0.00017309439135715365,
      "learning_rate": 0.053580855800859285,
      "loss": 0.2624,
      "num_input_tokens_seen": 22032584,
      "step": 28890
    },
    {
      "epoch": 60.07276507276507,
      "grad_norm": 0.00010227455641143024,
      "learning_rate": 0.05353573982455938,
      "loss": 0.2797,
      "num_input_tokens_seen": 22036360,
      "step": 28895
    },
    {
      "epoch": 60.08316008316008,
      "grad_norm": 0.00014359166380017996,
      "learning_rate": 0.053490638724260686,
      "loss": 0.2649,
      "num_input_tokens_seen": 22040104,
      "step": 28900
    },
    {
      "epoch": 60.093555093555096,
      "grad_norm": 2.7528898499440402e-05,
      "learning_rate": 0.05344555250691827,
      "loss": 0.2602,
      "num_input_tokens_seen": 22043912,
      "step": 28905
    },
    {
      "epoch": 60.103950103950105,
      "grad_norm": 0.0004923850647173822,
      "learning_rate": 0.053400481179485086,
      "loss": 0.2711,
      "num_input_tokens_seen": 22047912,
      "step": 28910
    },
    {
      "epoch": 60.11434511434511,
      "grad_norm": 0.0004644485015887767,
      "learning_rate": 0.05335542474891159,
      "loss": 0.292,
      "num_input_tokens_seen": 22051720,
      "step": 28915
    },
    {
      "epoch": 60.12474012474012,
      "grad_norm": 8.596223779022694e-05,
      "learning_rate": 0.053310383222146124,
      "loss": 0.2754,
      "num_input_tokens_seen": 22055560,
      "step": 28920
    },
    {
      "epoch": 60.13513513513514,
      "grad_norm": 0.000365704414434731,
      "learning_rate": 0.053265356606134684,
      "loss": 0.2555,
      "num_input_tokens_seen": 22059464,
      "step": 28925
    },
    {
      "epoch": 60.145530145530145,
      "grad_norm": 0.00037110046832822263,
      "learning_rate": 0.053220344907820856,
      "loss": 0.2631,
      "num_input_tokens_seen": 22063272,
      "step": 28930
    },
    {
      "epoch": 60.15592515592515,
      "grad_norm": 0.00031425326596945524,
      "learning_rate": 0.05317534813414608,
      "loss": 0.2703,
      "num_input_tokens_seen": 22066920,
      "step": 28935
    },
    {
      "epoch": 60.16632016632017,
      "grad_norm": 0.000520041212439537,
      "learning_rate": 0.05313036629204942,
      "loss": 0.2535,
      "num_input_tokens_seen": 22070792,
      "step": 28940
    },
    {
      "epoch": 60.17671517671518,
      "grad_norm": 0.00011217831342946738,
      "learning_rate": 0.05308539938846756,
      "loss": 0.2654,
      "num_input_tokens_seen": 22074632,
      "step": 28945
    },
    {
      "epoch": 60.187110187110186,
      "grad_norm": 9.818014950724319e-05,
      "learning_rate": 0.05304044743033507,
      "loss": 0.2453,
      "num_input_tokens_seen": 22078440,
      "step": 28950
    },
    {
      "epoch": 60.197505197505194,
      "grad_norm": 0.0002047102025244385,
      "learning_rate": 0.05299551042458401,
      "loss": 0.2686,
      "num_input_tokens_seen": 22082216,
      "step": 28955
    },
    {
      "epoch": 60.20790020790021,
      "grad_norm": 0.0004223485302645713,
      "learning_rate": 0.052950588378144266,
      "loss": 0.2625,
      "num_input_tokens_seen": 22086024,
      "step": 28960
    },
    {
      "epoch": 60.21829521829522,
      "grad_norm": 0.0001951005106093362,
      "learning_rate": 0.052905681297943465,
      "loss": 0.2812,
      "num_input_tokens_seen": 22089992,
      "step": 28965
    },
    {
      "epoch": 60.228690228690226,
      "grad_norm": 5.571204383159056e-05,
      "learning_rate": 0.0528607891909067,
      "loss": 0.2885,
      "num_input_tokens_seen": 22093736,
      "step": 28970
    },
    {
      "epoch": 60.23908523908524,
      "grad_norm": 0.00020731370022986084,
      "learning_rate": 0.05281591206395697,
      "loss": 0.2638,
      "num_input_tokens_seen": 22097576,
      "step": 28975
    },
    {
      "epoch": 60.24948024948025,
      "grad_norm": 0.0002497128152754158,
      "learning_rate": 0.05277104992401496,
      "loss": 0.2743,
      "num_input_tokens_seen": 22101416,
      "step": 28980
    },
    {
      "epoch": 60.25987525987526,
      "grad_norm": 0.0001987690629903227,
      "learning_rate": 0.05272620277799884,
      "loss": 0.2631,
      "num_input_tokens_seen": 22105256,
      "step": 28985
    },
    {
      "epoch": 60.270270270270274,
      "grad_norm": 0.0004770920204464346,
      "learning_rate": 0.05268137063282473,
      "loss": 0.2732,
      "num_input_tokens_seen": 22109064,
      "step": 28990
    },
    {
      "epoch": 60.28066528066528,
      "grad_norm": 0.00012539415911305696,
      "learning_rate": 0.0526365534954062,
      "loss": 0.2685,
      "num_input_tokens_seen": 22112808,
      "step": 28995
    },
    {
      "epoch": 60.29106029106029,
      "grad_norm": 0.00015280838124454021,
      "learning_rate": 0.052591751372654656,
      "loss": 0.2601,
      "num_input_tokens_seen": 22116648,
      "step": 29000
    },
    {
      "epoch": 60.29106029106029,
      "eval_loss": 0.25114792585372925,
      "eval_runtime": 13.3985,
      "eval_samples_per_second": 63.888,
      "eval_steps_per_second": 15.972,
      "num_input_tokens_seen": 22116648,
      "step": 29000
    },
    {
      "epoch": 60.3014553014553,
      "grad_norm": 0.00017448792641516775,
      "learning_rate": 0.05254696427147921,
      "loss": 0.2777,
      "num_input_tokens_seen": 22120424,
      "step": 29005
    },
    {
      "epoch": 60.311850311850314,
      "grad_norm": 0.0002914345823228359,
      "learning_rate": 0.052502192198786546,
      "loss": 0.2475,
      "num_input_tokens_seen": 22124232,
      "step": 29010
    },
    {
      "epoch": 60.32224532224532,
      "grad_norm": 0.0006505593191832304,
      "learning_rate": 0.05245743516148103,
      "loss": 0.226,
      "num_input_tokens_seen": 22128040,
      "step": 29015
    },
    {
      "epoch": 60.33264033264033,
      "grad_norm": 0.00016514386516064405,
      "learning_rate": 0.05241269316646486,
      "loss": 0.275,
      "num_input_tokens_seen": 22131848,
      "step": 29020
    },
    {
      "epoch": 60.343035343035346,
      "grad_norm": 0.0004972768947482109,
      "learning_rate": 0.052367966220637725,
      "loss": 0.2525,
      "num_input_tokens_seen": 22135400,
      "step": 29025
    },
    {
      "epoch": 60.353430353430355,
      "grad_norm": 0.00012201243225717917,
      "learning_rate": 0.05232325433089716,
      "loss": 0.2541,
      "num_input_tokens_seen": 22139144,
      "step": 29030
    },
    {
      "epoch": 60.36382536382536,
      "grad_norm": 0.00014078826643526554,
      "learning_rate": 0.052278557504138214,
      "loss": 0.2643,
      "num_input_tokens_seen": 22142920,
      "step": 29035
    },
    {
      "epoch": 60.37422037422037,
      "grad_norm": 0.0003011383814737201,
      "learning_rate": 0.05223387574725372,
      "loss": 0.2563,
      "num_input_tokens_seen": 22146792,
      "step": 29040
    },
    {
      "epoch": 60.38461538461539,
      "grad_norm": 5.9105397667735815e-05,
      "learning_rate": 0.05218920906713428,
      "loss": 0.2528,
      "num_input_tokens_seen": 22150632,
      "step": 29045
    },
    {
      "epoch": 60.395010395010395,
      "grad_norm": 0.00043190570431761444,
      "learning_rate": 0.05214455747066789,
      "loss": 0.2856,
      "num_input_tokens_seen": 22154568,
      "step": 29050
    },
    {
      "epoch": 60.4054054054054,
      "grad_norm": 0.00024825692526064813,
      "learning_rate": 0.05209992096474048,
      "loss": 0.2477,
      "num_input_tokens_seen": 22158344,
      "step": 29055
    },
    {
      "epoch": 60.41580041580042,
      "grad_norm": 0.00041552301263436675,
      "learning_rate": 0.05205529955623559,
      "loss": 0.2677,
      "num_input_tokens_seen": 22162088,
      "step": 29060
    },
    {
      "epoch": 60.42619542619543,
      "grad_norm": 4.142362013226375e-05,
      "learning_rate": 0.052010693252034314,
      "loss": 0.2763,
      "num_input_tokens_seen": 22165800,
      "step": 29065
    },
    {
      "epoch": 60.436590436590436,
      "grad_norm": 0.00044333390542306006,
      "learning_rate": 0.0519661020590156,
      "loss": 0.2383,
      "num_input_tokens_seen": 22169672,
      "step": 29070
    },
    {
      "epoch": 60.446985446985444,
      "grad_norm": 0.00011101623385911807,
      "learning_rate": 0.05192152598405586,
      "loss": 0.2781,
      "num_input_tokens_seen": 22173448,
      "step": 29075
    },
    {
      "epoch": 60.45738045738046,
      "grad_norm": 0.0001243502920260653,
      "learning_rate": 0.05187696503402941,
      "loss": 0.2607,
      "num_input_tokens_seen": 22177320,
      "step": 29080
    },
    {
      "epoch": 60.46777546777547,
      "grad_norm": 0.0005695197614841163,
      "learning_rate": 0.05183241921580798,
      "loss": 0.2731,
      "num_input_tokens_seen": 22181064,
      "step": 29085
    },
    {
      "epoch": 60.478170478170476,
      "grad_norm": 0.00020992530335206538,
      "learning_rate": 0.051787888536261206,
      "loss": 0.2679,
      "num_input_tokens_seen": 22184936,
      "step": 29090
    },
    {
      "epoch": 60.48856548856549,
      "grad_norm": 0.0002746711252257228,
      "learning_rate": 0.051743373002256184,
      "loss": 0.2292,
      "num_input_tokens_seen": 22188712,
      "step": 29095
    },
    {
      "epoch": 60.4989604989605,
      "grad_norm": 0.00014733635180164129,
      "learning_rate": 0.05169887262065787,
      "loss": 0.2229,
      "num_input_tokens_seen": 22192456,
      "step": 29100
    },
    {
      "epoch": 60.50935550935551,
      "grad_norm": 0.00013522559311240911,
      "learning_rate": 0.051654387398328665,
      "loss": 0.2788,
      "num_input_tokens_seen": 22196264,
      "step": 29105
    },
    {
      "epoch": 60.51975051975052,
      "grad_norm": 6.926549394847825e-05,
      "learning_rate": 0.05160991734212888,
      "loss": 0.2841,
      "num_input_tokens_seen": 22200040,
      "step": 29110
    },
    {
      "epoch": 60.53014553014553,
      "grad_norm": 0.00014191550144460052,
      "learning_rate": 0.051565462458916224,
      "loss": 0.2492,
      "num_input_tokens_seen": 22203848,
      "step": 29115
    },
    {
      "epoch": 60.54054054054054,
      "grad_norm": 0.00022297805116977543,
      "learning_rate": 0.05152102275554627,
      "loss": 0.2489,
      "num_input_tokens_seen": 22207752,
      "step": 29120
    },
    {
      "epoch": 60.55093555093555,
      "grad_norm": 0.00021488354832399637,
      "learning_rate": 0.05147659823887222,
      "loss": 0.2953,
      "num_input_tokens_seen": 22211592,
      "step": 29125
    },
    {
      "epoch": 60.561330561330564,
      "grad_norm": 0.00020440088701434433,
      "learning_rate": 0.05143218891574479,
      "loss": 0.2447,
      "num_input_tokens_seen": 22215464,
      "step": 29130
    },
    {
      "epoch": 60.57172557172557,
      "grad_norm": 0.00023655552649870515,
      "learning_rate": 0.0513877947930125,
      "loss": 0.2432,
      "num_input_tokens_seen": 22219368,
      "step": 29135
    },
    {
      "epoch": 60.58212058212058,
      "grad_norm": 0.0003324352437630296,
      "learning_rate": 0.051343415877521566,
      "loss": 0.2377,
      "num_input_tokens_seen": 22223080,
      "step": 29140
    },
    {
      "epoch": 60.59251559251559,
      "grad_norm": 6.227132689673454e-05,
      "learning_rate": 0.051299052176115634,
      "loss": 0.2612,
      "num_input_tokens_seen": 22226952,
      "step": 29145
    },
    {
      "epoch": 60.602910602910605,
      "grad_norm": 0.0003988379903603345,
      "learning_rate": 0.051254703695636256,
      "loss": 0.2714,
      "num_input_tokens_seen": 22230824,
      "step": 29150
    },
    {
      "epoch": 60.61330561330561,
      "grad_norm": 0.0001947104901773855,
      "learning_rate": 0.05121037044292249,
      "loss": 0.2706,
      "num_input_tokens_seen": 22234632,
      "step": 29155
    },
    {
      "epoch": 60.62370062370062,
      "grad_norm": 0.0004076639888808131,
      "learning_rate": 0.05116605242481101,
      "loss": 0.2605,
      "num_input_tokens_seen": 22238504,
      "step": 29160
    },
    {
      "epoch": 60.63409563409564,
      "grad_norm": 0.0001246178726432845,
      "learning_rate": 0.05112174964813634,
      "loss": 0.2548,
      "num_input_tokens_seen": 22242376,
      "step": 29165
    },
    {
      "epoch": 60.644490644490645,
      "grad_norm": 7.189393363660201e-05,
      "learning_rate": 0.05107746211973038,
      "loss": 0.2713,
      "num_input_tokens_seen": 22246120,
      "step": 29170
    },
    {
      "epoch": 60.65488565488565,
      "grad_norm": 0.00023302558111026883,
      "learning_rate": 0.05103318984642291,
      "loss": 0.253,
      "num_input_tokens_seen": 22249832,
      "step": 29175
    },
    {
      "epoch": 60.66528066528066,
      "grad_norm": 0.00040809568599797785,
      "learning_rate": 0.05098893283504131,
      "loss": 0.2791,
      "num_input_tokens_seen": 22253576,
      "step": 29180
    },
    {
      "epoch": 60.67567567567568,
      "grad_norm": 0.00010664543515304103,
      "learning_rate": 0.050944691092410475,
      "loss": 0.2544,
      "num_input_tokens_seen": 22257288,
      "step": 29185
    },
    {
      "epoch": 60.686070686070686,
      "grad_norm": 0.00043277291115373373,
      "learning_rate": 0.05090046462535313,
      "loss": 0.2703,
      "num_input_tokens_seen": 22261192,
      "step": 29190
    },
    {
      "epoch": 60.696465696465694,
      "grad_norm": 0.0006072860560379922,
      "learning_rate": 0.050856253440689454,
      "loss": 0.2733,
      "num_input_tokens_seen": 22265160,
      "step": 29195
    },
    {
      "epoch": 60.70686070686071,
      "grad_norm": 0.00015070570225361735,
      "learning_rate": 0.050812057545237405,
      "loss": 0.2744,
      "num_input_tokens_seen": 22269032,
      "step": 29200
    },
    {
      "epoch": 60.70686070686071,
      "eval_loss": 0.2513991594314575,
      "eval_runtime": 13.386,
      "eval_samples_per_second": 63.948,
      "eval_steps_per_second": 15.987,
      "num_input_tokens_seen": 22269032,
      "step": 29200
    },
    {
      "epoch": 60.71725571725572,
      "grad_norm": 0.00026994626387022436,
      "learning_rate": 0.0507678769458126,
      "loss": 0.2702,
      "num_input_tokens_seen": 22272808,
      "step": 29205
    },
    {
      "epoch": 60.727650727650726,
      "grad_norm": 0.0002789971185848117,
      "learning_rate": 0.050723711649228155,
      "loss": 0.2669,
      "num_input_tokens_seen": 22276648,
      "step": 29210
    },
    {
      "epoch": 60.73804573804574,
      "grad_norm": 0.0003134275902993977,
      "learning_rate": 0.05067956166229496,
      "loss": 0.2711,
      "num_input_tokens_seen": 22280392,
      "step": 29215
    },
    {
      "epoch": 60.74844074844075,
      "grad_norm": 3.340885086799972e-05,
      "learning_rate": 0.05063542699182155,
      "loss": 0.2607,
      "num_input_tokens_seen": 22284104,
      "step": 29220
    },
    {
      "epoch": 60.75883575883576,
      "grad_norm": 0.0002111646463163197,
      "learning_rate": 0.050591307644613996,
      "loss": 0.2674,
      "num_input_tokens_seen": 22288168,
      "step": 29225
    },
    {
      "epoch": 60.76923076923077,
      "grad_norm": 0.00032830823329277337,
      "learning_rate": 0.05054720362747599,
      "loss": 0.2827,
      "num_input_tokens_seen": 22291976,
      "step": 29230
    },
    {
      "epoch": 60.77962577962578,
      "grad_norm": 0.0002581693115644157,
      "learning_rate": 0.050503114947209035,
      "loss": 0.2701,
      "num_input_tokens_seen": 22295816,
      "step": 29235
    },
    {
      "epoch": 60.79002079002079,
      "grad_norm": 0.00022006980725564063,
      "learning_rate": 0.05045904161061207,
      "loss": 0.2655,
      "num_input_tokens_seen": 22299752,
      "step": 29240
    },
    {
      "epoch": 60.8004158004158,
      "grad_norm": 0.00016812840476632118,
      "learning_rate": 0.05041498362448185,
      "loss": 0.2621,
      "num_input_tokens_seen": 22303688,
      "step": 29245
    },
    {
      "epoch": 60.810810810810814,
      "grad_norm": 0.0008603687747381628,
      "learning_rate": 0.05037094099561256,
      "loss": 0.2753,
      "num_input_tokens_seen": 22307560,
      "step": 29250
    },
    {
      "epoch": 60.82120582120582,
      "grad_norm": 0.00018224988889414817,
      "learning_rate": 0.05032691373079624,
      "loss": 0.283,
      "num_input_tokens_seen": 22311464,
      "step": 29255
    },
    {
      "epoch": 60.83160083160083,
      "grad_norm": 0.00019357199198566377,
      "learning_rate": 0.05028290183682234,
      "loss": 0.2703,
      "num_input_tokens_seen": 22315304,
      "step": 29260
    },
    {
      "epoch": 60.84199584199584,
      "grad_norm": 0.0003141420893371105,
      "learning_rate": 0.050238905320478096,
      "loss": 0.2632,
      "num_input_tokens_seen": 22319112,
      "step": 29265
    },
    {
      "epoch": 60.852390852390855,
      "grad_norm": 0.0001770677336025983,
      "learning_rate": 0.05019492418854838,
      "loss": 0.279,
      "num_input_tokens_seen": 22322920,
      "step": 29270
    },
    {
      "epoch": 60.86278586278586,
      "grad_norm": 0.0004777581780217588,
      "learning_rate": 0.05015095844781554,
      "loss": 0.277,
      "num_input_tokens_seen": 22326824,
      "step": 29275
    },
    {
      "epoch": 60.87318087318087,
      "grad_norm": 0.000332927011186257,
      "learning_rate": 0.05010700810505968,
      "loss": 0.285,
      "num_input_tokens_seen": 22330664,
      "step": 29280
    },
    {
      "epoch": 60.88357588357589,
      "grad_norm": 6.426603067666292e-05,
      "learning_rate": 0.05006307316705856,
      "loss": 0.2695,
      "num_input_tokens_seen": 22334376,
      "step": 29285
    },
    {
      "epoch": 60.893970893970895,
      "grad_norm": 0.000273017561994493,
      "learning_rate": 0.0500191536405874,
      "loss": 0.2716,
      "num_input_tokens_seen": 22338280,
      "step": 29290
    },
    {
      "epoch": 60.9043659043659,
      "grad_norm": 0.00046285157441161573,
      "learning_rate": 0.04997524953241922,
      "loss": 0.2684,
      "num_input_tokens_seen": 22342344,
      "step": 29295
    },
    {
      "epoch": 60.91476091476091,
      "grad_norm": 0.0001919929782161489,
      "learning_rate": 0.049931360849324556,
      "loss": 0.2505,
      "num_input_tokens_seen": 22346088,
      "step": 29300
    },
    {
      "epoch": 60.92515592515593,
      "grad_norm": 5.1948973123217e-05,
      "learning_rate": 0.04988748759807155,
      "loss": 0.2652,
      "num_input_tokens_seen": 22349736,
      "step": 29305
    },
    {
      "epoch": 60.935550935550935,
      "grad_norm": 6.86116109136492e-05,
      "learning_rate": 0.0498436297854261,
      "loss": 0.2388,
      "num_input_tokens_seen": 22353480,
      "step": 29310
    },
    {
      "epoch": 60.945945945945944,
      "grad_norm": 0.00023443944519385695,
      "learning_rate": 0.04979978741815152,
      "loss": 0.2489,
      "num_input_tokens_seen": 22357096,
      "step": 29315
    },
    {
      "epoch": 60.95634095634096,
      "grad_norm": 4.803876072401181e-05,
      "learning_rate": 0.04975596050300891,
      "loss": 0.2546,
      "num_input_tokens_seen": 22360840,
      "step": 29320
    },
    {
      "epoch": 60.96673596673597,
      "grad_norm": 0.0002800797810778022,
      "learning_rate": 0.049712149046757005,
      "loss": 0.2714,
      "num_input_tokens_seen": 22364872,
      "step": 29325
    },
    {
      "epoch": 60.977130977130976,
      "grad_norm": 6.997302261879668e-05,
      "learning_rate": 0.04966835305615194,
      "loss": 0.2531,
      "num_input_tokens_seen": 22368680,
      "step": 29330
    },
    {
      "epoch": 60.987525987525984,
      "grad_norm": 0.00043051608372479677,
      "learning_rate": 0.049624572537947755,
      "loss": 0.2848,
      "num_input_tokens_seen": 22372520,
      "step": 29335
    },
    {
      "epoch": 60.997920997921,
      "grad_norm": 0.00014155283861327916,
      "learning_rate": 0.04958080749889582,
      "loss": 0.264,
      "num_input_tokens_seen": 22376264,
      "step": 29340
    },
    {
      "epoch": 61.00831600831601,
      "grad_norm": 0.00045308491098694503,
      "learning_rate": 0.049537057945745304,
      "loss": 0.2402,
      "num_input_tokens_seen": 22380056,
      "step": 29345
    },
    {
      "epoch": 61.018711018711016,
      "grad_norm": 8.39843341964297e-05,
      "learning_rate": 0.049493323885243,
      "loss": 0.2631,
      "num_input_tokens_seen": 22383960,
      "step": 29350
    },
    {
      "epoch": 61.02910602910603,
      "grad_norm": 6.288572330959141e-05,
      "learning_rate": 0.04944960532413318,
      "loss": 0.2887,
      "num_input_tokens_seen": 22387768,
      "step": 29355
    },
    {
      "epoch": 61.03950103950104,
      "grad_norm": 0.00014656568237114698,
      "learning_rate": 0.049405902269157774,
      "loss": 0.2599,
      "num_input_tokens_seen": 22391576,
      "step": 29360
    },
    {
      "epoch": 61.04989604989605,
      "grad_norm": 0.00013488282274920493,
      "learning_rate": 0.04936221472705646,
      "loss": 0.2734,
      "num_input_tokens_seen": 22395320,
      "step": 29365
    },
    {
      "epoch": 61.06029106029106,
      "grad_norm": 4.0780701965559274e-05,
      "learning_rate": 0.04931854270456632,
      "loss": 0.2569,
      "num_input_tokens_seen": 22399192,
      "step": 29370
    },
    {
      "epoch": 61.07068607068607,
      "grad_norm": 0.00010651641787262633,
      "learning_rate": 0.049274886208422075,
      "loss": 0.2714,
      "num_input_tokens_seen": 22402968,
      "step": 29375
    },
    {
      "epoch": 61.08108108108108,
      "grad_norm": 0.00022704314324073493,
      "learning_rate": 0.049231245245356235,
      "loss": 0.2363,
      "num_input_tokens_seen": 22406712,
      "step": 29380
    },
    {
      "epoch": 61.09147609147609,
      "grad_norm": 0.00042226497316733,
      "learning_rate": 0.049187619822098655,
      "loss": 0.2573,
      "num_input_tokens_seen": 22410616,
      "step": 29385
    },
    {
      "epoch": 61.101871101871104,
      "grad_norm": 0.00012956962746102363,
      "learning_rate": 0.04914400994537705,
      "loss": 0.2667,
      "num_input_tokens_seen": 22414232,
      "step": 29390
    },
    {
      "epoch": 61.11226611226611,
      "grad_norm": 0.00024197938910219818,
      "learning_rate": 0.049100415621916485,
      "loss": 0.2783,
      "num_input_tokens_seen": 22418104,
      "step": 29395
    },
    {
      "epoch": 61.12266112266112,
      "grad_norm": 0.00019310711650177836,
      "learning_rate": 0.04905683685843981,
      "loss": 0.2519,
      "num_input_tokens_seen": 22421944,
      "step": 29400
    },
    {
      "epoch": 61.12266112266112,
      "eval_loss": 0.24778404831886292,
      "eval_runtime": 13.4094,
      "eval_samples_per_second": 63.836,
      "eval_steps_per_second": 15.959,
      "num_input_tokens_seen": 22421944,
      "step": 29400
    },
    {
      "epoch": 61.13305613305613,
      "grad_norm": 0.0006019994616508484,
      "learning_rate": 0.049013273661667495,
      "loss": 0.255,
      "num_input_tokens_seen": 22425688,
      "step": 29405
    },
    {
      "epoch": 61.143451143451145,
      "grad_norm": 0.0001796743308659643,
      "learning_rate": 0.048969726038317396,
      "loss": 0.2627,
      "num_input_tokens_seen": 22429592,
      "step": 29410
    },
    {
      "epoch": 61.15384615384615,
      "grad_norm": 7.682923751417547e-05,
      "learning_rate": 0.048926193995105206,
      "loss": 0.2624,
      "num_input_tokens_seen": 22433432,
      "step": 29415
    },
    {
      "epoch": 61.16424116424116,
      "grad_norm": 0.00025951353018172085,
      "learning_rate": 0.048882677538744035,
      "loss": 0.2767,
      "num_input_tokens_seen": 22437176,
      "step": 29420
    },
    {
      "epoch": 61.17463617463618,
      "grad_norm": 0.00020866339036729187,
      "learning_rate": 0.048839176675944715,
      "loss": 0.2593,
      "num_input_tokens_seen": 22441048,
      "step": 29425
    },
    {
      "epoch": 61.185031185031185,
      "grad_norm": 0.00026405375683680177,
      "learning_rate": 0.04879569141341566,
      "loss": 0.2817,
      "num_input_tokens_seen": 22444856,
      "step": 29430
    },
    {
      "epoch": 61.195426195426194,
      "grad_norm": 6.615067832171917e-05,
      "learning_rate": 0.04875222175786274,
      "loss": 0.2646,
      "num_input_tokens_seen": 22448632,
      "step": 29435
    },
    {
      "epoch": 61.20582120582121,
      "grad_norm": 0.00015455170068889856,
      "learning_rate": 0.04870876771598966,
      "loss": 0.2503,
      "num_input_tokens_seen": 22452440,
      "step": 29440
    },
    {
      "epoch": 61.21621621621622,
      "grad_norm": 0.0003888489445671439,
      "learning_rate": 0.04866532929449744,
      "loss": 0.2533,
      "num_input_tokens_seen": 22456248,
      "step": 29445
    },
    {
      "epoch": 61.226611226611226,
      "grad_norm": 0.0006923777400515974,
      "learning_rate": 0.048621906500084945,
      "loss": 0.2772,
      "num_input_tokens_seen": 22460024,
      "step": 29450
    },
    {
      "epoch": 61.237006237006234,
      "grad_norm": 0.0002870278840418905,
      "learning_rate": 0.04857849933944845,
      "loss": 0.2493,
      "num_input_tokens_seen": 22463928,
      "step": 29455
    },
    {
      "epoch": 61.24740124740125,
      "grad_norm": 0.00015916672418825328,
      "learning_rate": 0.048535107819281866,
      "loss": 0.2473,
      "num_input_tokens_seen": 22467800,
      "step": 29460
    },
    {
      "epoch": 61.25779625779626,
      "grad_norm": 0.00030887455795891583,
      "learning_rate": 0.04849173194627675,
      "loss": 0.2352,
      "num_input_tokens_seen": 22471672,
      "step": 29465
    },
    {
      "epoch": 61.268191268191266,
      "grad_norm": 0.00023358812904916704,
      "learning_rate": 0.04844837172712223,
      "loss": 0.2813,
      "num_input_tokens_seen": 22475544,
      "step": 29470
    },
    {
      "epoch": 61.27858627858628,
      "grad_norm": 0.0003031744563486427,
      "learning_rate": 0.04840502716850494,
      "loss": 0.2604,
      "num_input_tokens_seen": 22479576,
      "step": 29475
    },
    {
      "epoch": 61.28898128898129,
      "grad_norm": 0.0006664341781288385,
      "learning_rate": 0.04836169827710916,
      "loss": 0.2707,
      "num_input_tokens_seen": 22483480,
      "step": 29480
    },
    {
      "epoch": 61.2993762993763,
      "grad_norm": 0.0001676773972576484,
      "learning_rate": 0.04831838505961684,
      "loss": 0.2679,
      "num_input_tokens_seen": 22487320,
      "step": 29485
    },
    {
      "epoch": 61.30977130977131,
      "grad_norm": 0.0006817649118602276,
      "learning_rate": 0.048275087522707295,
      "loss": 0.2381,
      "num_input_tokens_seen": 22491032,
      "step": 29490
    },
    {
      "epoch": 61.32016632016632,
      "grad_norm": 0.0005874313064850867,
      "learning_rate": 0.04823180567305766,
      "loss": 0.2774,
      "num_input_tokens_seen": 22495032,
      "step": 29495
    },
    {
      "epoch": 61.33056133056133,
      "grad_norm": 0.0006164637161418796,
      "learning_rate": 0.04818853951734244,
      "loss": 0.2798,
      "num_input_tokens_seen": 22498808,
      "step": 29500
    },
    {
      "epoch": 61.34095634095634,
      "grad_norm": 0.00037387784686870873,
      "learning_rate": 0.04814528906223387,
      "loss": 0.2458,
      "num_input_tokens_seen": 22502552,
      "step": 29505
    },
    {
      "epoch": 61.351351351351354,
      "grad_norm": 8.99830847629346e-05,
      "learning_rate": 0.04810205431440177,
      "loss": 0.2581,
      "num_input_tokens_seen": 22506392,
      "step": 29510
    },
    {
      "epoch": 61.36174636174636,
      "grad_norm": 0.00037280903779901564,
      "learning_rate": 0.04805883528051341,
      "loss": 0.2564,
      "num_input_tokens_seen": 22510200,
      "step": 29515
    },
    {
      "epoch": 61.37214137214137,
      "grad_norm": 0.00019036447338294238,
      "learning_rate": 0.048015631967233685,
      "loss": 0.2753,
      "num_input_tokens_seen": 22514136,
      "step": 29520
    },
    {
      "epoch": 61.38253638253638,
      "grad_norm": 0.00044575455831363797,
      "learning_rate": 0.04797244438122517,
      "loss": 0.2634,
      "num_input_tokens_seen": 22517816,
      "step": 29525
    },
    {
      "epoch": 61.392931392931395,
      "grad_norm": 0.00013188677257858217,
      "learning_rate": 0.04792927252914784,
      "loss": 0.2696,
      "num_input_tokens_seen": 22521816,
      "step": 29530
    },
    {
      "epoch": 61.4033264033264,
      "grad_norm": 8.169235661625862e-05,
      "learning_rate": 0.04788611641765944,
      "loss": 0.2618,
      "num_input_tokens_seen": 22525464,
      "step": 29535
    },
    {
      "epoch": 61.41372141372141,
      "grad_norm": 0.00018782488768920302,
      "learning_rate": 0.04784297605341508,
      "loss": 0.2537,
      "num_input_tokens_seen": 22529144,
      "step": 29540
    },
    {
      "epoch": 61.42411642411643,
      "grad_norm": 0.00044537067878991365,
      "learning_rate": 0.04779985144306761,
      "loss": 0.2722,
      "num_input_tokens_seen": 22533016,
      "step": 29545
    },
    {
      "epoch": 61.434511434511435,
      "grad_norm": 0.00034549308475106955,
      "learning_rate": 0.047756742593267405,
      "loss": 0.264,
      "num_input_tokens_seen": 22536792,
      "step": 29550
    },
    {
      "epoch": 61.444906444906444,
      "grad_norm": 0.0002115590323228389,
      "learning_rate": 0.047713649510662315,
      "loss": 0.2627,
      "num_input_tokens_seen": 22540696,
      "step": 29555
    },
    {
      "epoch": 61.45530145530145,
      "grad_norm": 0.00027155931456945837,
      "learning_rate": 0.04767057220189789,
      "loss": 0.2658,
      "num_input_tokens_seen": 22544504,
      "step": 29560
    },
    {
      "epoch": 61.46569646569647,
      "grad_norm": 7.145785639295354e-05,
      "learning_rate": 0.04762751067361722,
      "loss": 0.2651,
      "num_input_tokens_seen": 22548312,
      "step": 29565
    },
    {
      "epoch": 61.476091476091476,
      "grad_norm": 0.00030012940987944603,
      "learning_rate": 0.04758446493246086,
      "loss": 0.2729,
      "num_input_tokens_seen": 22552056,
      "step": 29570
    },
    {
      "epoch": 61.486486486486484,
      "grad_norm": 0.00046009692596271634,
      "learning_rate": 0.047541434985067084,
      "loss": 0.2666,
      "num_input_tokens_seen": 22555896,
      "step": 29575
    },
    {
      "epoch": 61.4968814968815,
      "grad_norm": 4.3458341679070145e-05,
      "learning_rate": 0.047498420838071556,
      "loss": 0.2759,
      "num_input_tokens_seen": 22559896,
      "step": 29580
    },
    {
      "epoch": 61.50727650727651,
      "grad_norm": 7.798478327458724e-05,
      "learning_rate": 0.04745542249810772,
      "loss": 0.271,
      "num_input_tokens_seen": 22563704,
      "step": 29585
    },
    {
      "epoch": 61.517671517671516,
      "grad_norm": 6.683883111691102e-05,
      "learning_rate": 0.047412439971806324,
      "loss": 0.2674,
      "num_input_tokens_seen": 22567512,
      "step": 29590
    },
    {
      "epoch": 61.528066528066525,
      "grad_norm": 0.00041826171218417585,
      "learning_rate": 0.04736947326579592,
      "loss": 0.2721,
      "num_input_tokens_seen": 22571352,
      "step": 29595
    },
    {
      "epoch": 61.53846153846154,
      "grad_norm": 0.0002417449577478692,
      "learning_rate": 0.04732652238670245,
      "loss": 0.2741,
      "num_input_tokens_seen": 22574936,
      "step": 29600
    },
    {
      "epoch": 61.53846153846154,
      "eval_loss": 0.25261032581329346,
      "eval_runtime": 13.3994,
      "eval_samples_per_second": 63.884,
      "eval_steps_per_second": 15.971,
      "num_input_tokens_seen": 22574936,
      "step": 29600
    },
    {
      "epoch": 61.54885654885655,
      "grad_norm": 5.4587908380199224e-05,
      "learning_rate": 0.04728358734114952,
      "loss": 0.2627,
      "num_input_tokens_seen": 22578552,
      "step": 29605
    },
    {
      "epoch": 61.55925155925156,
      "grad_norm": 0.0004376430588308722,
      "learning_rate": 0.04724066813575821,
      "loss": 0.288,
      "num_input_tokens_seen": 22582360,
      "step": 29610
    },
    {
      "epoch": 61.56964656964657,
      "grad_norm": 0.00014969435869716108,
      "learning_rate": 0.04719776477714729,
      "loss": 0.27,
      "num_input_tokens_seen": 22586104,
      "step": 29615
    },
    {
      "epoch": 61.58004158004158,
      "grad_norm": 0.00023263019102159888,
      "learning_rate": 0.047154877271932856,
      "loss": 0.2659,
      "num_input_tokens_seen": 22589944,
      "step": 29620
    },
    {
      "epoch": 61.59043659043659,
      "grad_norm": 0.0001864308287622407,
      "learning_rate": 0.0471120056267288,
      "loss": 0.2723,
      "num_input_tokens_seen": 22593720,
      "step": 29625
    },
    {
      "epoch": 61.6008316008316,
      "grad_norm": 0.00021168560488149524,
      "learning_rate": 0.047069149848146495,
      "loss": 0.2719,
      "num_input_tokens_seen": 22597496,
      "step": 29630
    },
    {
      "epoch": 61.61122661122661,
      "grad_norm": 0.00044054165482521057,
      "learning_rate": 0.04702630994279473,
      "loss": 0.2655,
      "num_input_tokens_seen": 22601336,
      "step": 29635
    },
    {
      "epoch": 61.62162162162162,
      "grad_norm": 0.0003041196323465556,
      "learning_rate": 0.046983485917280035,
      "loss": 0.2588,
      "num_input_tokens_seen": 22605176,
      "step": 29640
    },
    {
      "epoch": 61.63201663201663,
      "grad_norm": 0.00019611175230238587,
      "learning_rate": 0.04694067777820644,
      "loss": 0.2717,
      "num_input_tokens_seen": 22608984,
      "step": 29645
    },
    {
      "epoch": 61.642411642411645,
      "grad_norm": 0.0003394393716007471,
      "learning_rate": 0.046897885532175415,
      "loss": 0.2557,
      "num_input_tokens_seen": 22612728,
      "step": 29650
    },
    {
      "epoch": 61.65280665280665,
      "grad_norm": 0.0002699804899748415,
      "learning_rate": 0.04685510918578613,
      "loss": 0.2739,
      "num_input_tokens_seen": 22616504,
      "step": 29655
    },
    {
      "epoch": 61.66320166320166,
      "grad_norm": 0.0001857308525359258,
      "learning_rate": 0.04681234874563519,
      "loss": 0.2469,
      "num_input_tokens_seen": 22620312,
      "step": 29660
    },
    {
      "epoch": 61.67359667359668,
      "grad_norm": 0.0004041350621264428,
      "learning_rate": 0.046769604218316836,
      "loss": 0.264,
      "num_input_tokens_seen": 22624216,
      "step": 29665
    },
    {
      "epoch": 61.683991683991685,
      "grad_norm": 0.0004000420158263296,
      "learning_rate": 0.04672687561042279,
      "loss": 0.2675,
      "num_input_tokens_seen": 22627896,
      "step": 29670
    },
    {
      "epoch": 61.694386694386694,
      "grad_norm": 0.0004125539562664926,
      "learning_rate": 0.046684162928542286,
      "loss": 0.251,
      "num_input_tokens_seen": 22631608,
      "step": 29675
    },
    {
      "epoch": 61.7047817047817,
      "grad_norm": 0.00015640696801710874,
      "learning_rate": 0.04664146617926222,
      "loss": 0.2665,
      "num_input_tokens_seen": 22635448,
      "step": 29680
    },
    {
      "epoch": 61.71517671517672,
      "grad_norm": 0.00010408647358417511,
      "learning_rate": 0.046598785369167,
      "loss": 0.2477,
      "num_input_tokens_seen": 22639224,
      "step": 29685
    },
    {
      "epoch": 61.725571725571726,
      "grad_norm": 0.00019091191643383354,
      "learning_rate": 0.046556120504838434,
      "loss": 0.2802,
      "num_input_tokens_seen": 22642968,
      "step": 29690
    },
    {
      "epoch": 61.735966735966734,
      "grad_norm": 0.0005145528120920062,
      "learning_rate": 0.04651347159285609,
      "loss": 0.2584,
      "num_input_tokens_seen": 22646712,
      "step": 29695
    },
    {
      "epoch": 61.74636174636175,
      "grad_norm": 0.0005875641363672912,
      "learning_rate": 0.04647083863979688,
      "loss": 0.2512,
      "num_input_tokens_seen": 22650520,
      "step": 29700
    },
    {
      "epoch": 61.75675675675676,
      "grad_norm": 9.322630648966879e-05,
      "learning_rate": 0.04642822165223538,
      "loss": 0.2665,
      "num_input_tokens_seen": 22654456,
      "step": 29705
    },
    {
      "epoch": 61.767151767151766,
      "grad_norm": 0.00034378416603431106,
      "learning_rate": 0.046385620636743716,
      "loss": 0.277,
      "num_input_tokens_seen": 22658328,
      "step": 29710
    },
    {
      "epoch": 61.777546777546775,
      "grad_norm": 0.00016907391545828432,
      "learning_rate": 0.04634303559989141,
      "loss": 0.2636,
      "num_input_tokens_seen": 22661976,
      "step": 29715
    },
    {
      "epoch": 61.78794178794179,
      "grad_norm": 7.173990161390975e-05,
      "learning_rate": 0.046300466548245635,
      "loss": 0.2509,
      "num_input_tokens_seen": 22665784,
      "step": 29720
    },
    {
      "epoch": 61.7983367983368,
      "grad_norm": 0.0004067953850608319,
      "learning_rate": 0.04625791348837114,
      "loss": 0.2654,
      "num_input_tokens_seen": 22669720,
      "step": 29725
    },
    {
      "epoch": 61.80873180873181,
      "grad_norm": 0.00038648044574074447,
      "learning_rate": 0.046215376426830095,
      "loss": 0.2493,
      "num_input_tokens_seen": 22673720,
      "step": 29730
    },
    {
      "epoch": 61.81912681912682,
      "grad_norm": 0.000145882906508632,
      "learning_rate": 0.04617285537018219,
      "loss": 0.2703,
      "num_input_tokens_seen": 22677464,
      "step": 29735
    },
    {
      "epoch": 61.82952182952183,
      "grad_norm": 0.00019403253099881113,
      "learning_rate": 0.046130350324984803,
      "loss": 0.2508,
      "num_input_tokens_seen": 22681336,
      "step": 29740
    },
    {
      "epoch": 61.83991683991684,
      "grad_norm": 7.360886229434982e-05,
      "learning_rate": 0.046087861297792666,
      "loss": 0.2627,
      "num_input_tokens_seen": 22685016,
      "step": 29745
    },
    {
      "epoch": 61.85031185031185,
      "grad_norm": 0.0003178084152750671,
      "learning_rate": 0.0460453882951582,
      "loss": 0.2592,
      "num_input_tokens_seen": 22688952,
      "step": 29750
    },
    {
      "epoch": 61.86070686070686,
      "grad_norm": 0.00014706020010635257,
      "learning_rate": 0.04600293132363119,
      "loss": 0.2763,
      "num_input_tokens_seen": 22692632,
      "step": 29755
    },
    {
      "epoch": 61.87110187110187,
      "grad_norm": 0.00011950283078476787,
      "learning_rate": 0.045960490389759086,
      "loss": 0.2807,
      "num_input_tokens_seen": 22696504,
      "step": 29760
    },
    {
      "epoch": 61.88149688149688,
      "grad_norm": 0.0005953563959337771,
      "learning_rate": 0.04591806550008685,
      "loss": 0.2666,
      "num_input_tokens_seen": 22700312,
      "step": 29765
    },
    {
      "epoch": 61.891891891891895,
      "grad_norm": 0.0002817870117723942,
      "learning_rate": 0.045875656661156825,
      "loss": 0.2598,
      "num_input_tokens_seen": 22703928,
      "step": 29770
    },
    {
      "epoch": 61.9022869022869,
      "grad_norm": 5.4918957175686955e-05,
      "learning_rate": 0.04583326387950911,
      "loss": 0.2521,
      "num_input_tokens_seen": 22707672,
      "step": 29775
    },
    {
      "epoch": 61.91268191268191,
      "grad_norm": 8.617278945166618e-05,
      "learning_rate": 0.0457908871616811,
      "loss": 0.2783,
      "num_input_tokens_seen": 22711480,
      "step": 29780
    },
    {
      "epoch": 61.92307692307692,
      "grad_norm": 0.0005102165159769356,
      "learning_rate": 0.04574852651420786,
      "loss": 0.245,
      "num_input_tokens_seen": 22715192,
      "step": 29785
    },
    {
      "epoch": 61.933471933471935,
      "grad_norm": 0.00015870384231675416,
      "learning_rate": 0.045706181943621985,
      "loss": 0.2697,
      "num_input_tokens_seen": 22719096,
      "step": 29790
    },
    {
      "epoch": 61.943866943866944,
      "grad_norm": 0.00010720246064011008,
      "learning_rate": 0.04566385345645344,
      "loss": 0.2531,
      "num_input_tokens_seen": 22723032,
      "step": 29795
    },
    {
      "epoch": 61.95426195426195,
      "grad_norm": 0.00012251581938471645,
      "learning_rate": 0.04562154105922993,
      "loss": 0.2615,
      "num_input_tokens_seen": 22727064,
      "step": 29800
    },
    {
      "epoch": 61.95426195426195,
      "eval_loss": 0.24720001220703125,
      "eval_runtime": 13.4156,
      "eval_samples_per_second": 63.807,
      "eval_steps_per_second": 15.952,
      "num_input_tokens_seen": 22727064,
      "step": 29800
    },
    {
      "epoch": 61.96465696465697,
      "grad_norm": 0.00013783949543721974,
      "learning_rate": 0.04557924475847642,
      "loss": 0.2804,
      "num_input_tokens_seen": 22730904,
      "step": 29805
    },
    {
      "epoch": 61.975051975051976,
      "grad_norm": 0.00031125059467740357,
      "learning_rate": 0.04553696456071567,
      "loss": 0.2474,
      "num_input_tokens_seen": 22734872,
      "step": 29810
    },
    {
      "epoch": 61.985446985446984,
      "grad_norm": 0.0003392488870304078,
      "learning_rate": 0.045494700472467724,
      "loss": 0.285,
      "num_input_tokens_seen": 22738584,
      "step": 29815
    },
    {
      "epoch": 61.99584199584199,
      "grad_norm": 8.49536299938336e-05,
      "learning_rate": 0.04545245250025024,
      "loss": 0.2519,
      "num_input_tokens_seen": 22742488,
      "step": 29820
    },
    {
      "epoch": 62.00623700623701,
      "grad_norm": 0.00032907864078879356,
      "learning_rate": 0.045410220650578384,
      "loss": 0.2576,
      "num_input_tokens_seen": 22746400,
      "step": 29825
    },
    {
      "epoch": 62.016632016632016,
      "grad_norm": 0.0005208990187384188,
      "learning_rate": 0.04536800492996492,
      "loss": 0.2777,
      "num_input_tokens_seen": 22750272,
      "step": 29830
    },
    {
      "epoch": 62.027027027027025,
      "grad_norm": 0.0005330211133696139,
      "learning_rate": 0.04532580534491994,
      "loss": 0.2477,
      "num_input_tokens_seen": 22754048,
      "step": 29835
    },
    {
      "epoch": 62.03742203742204,
      "grad_norm": 0.0002972113434225321,
      "learning_rate": 0.045283621901951183,
      "loss": 0.2843,
      "num_input_tokens_seen": 22758112,
      "step": 29840
    },
    {
      "epoch": 62.04781704781705,
      "grad_norm": 0.00015745799464639276,
      "learning_rate": 0.04524145460756393,
      "loss": 0.2756,
      "num_input_tokens_seen": 22761952,
      "step": 29845
    },
    {
      "epoch": 62.05821205821206,
      "grad_norm": 0.00020698031585197896,
      "learning_rate": 0.045199303468260794,
      "loss": 0.2539,
      "num_input_tokens_seen": 22765728,
      "step": 29850
    },
    {
      "epoch": 62.06860706860707,
      "grad_norm": 0.00015088576765265316,
      "learning_rate": 0.04515716849054214,
      "loss": 0.2656,
      "num_input_tokens_seen": 22769536,
      "step": 29855
    },
    {
      "epoch": 62.07900207900208,
      "grad_norm": 0.0002759074850473553,
      "learning_rate": 0.04511504968090558,
      "loss": 0.2675,
      "num_input_tokens_seen": 22773344,
      "step": 29860
    },
    {
      "epoch": 62.08939708939709,
      "grad_norm": 0.0008666822104714811,
      "learning_rate": 0.04507294704584644,
      "loss": 0.2562,
      "num_input_tokens_seen": 22777184,
      "step": 29865
    },
    {
      "epoch": 62.0997920997921,
      "grad_norm": 0.0007040095515549183,
      "learning_rate": 0.04503086059185749,
      "loss": 0.2682,
      "num_input_tokens_seen": 22781024,
      "step": 29870
    },
    {
      "epoch": 62.11018711018711,
      "grad_norm": 0.00021752313477918506,
      "learning_rate": 0.04498879032542893,
      "loss": 0.2478,
      "num_input_tokens_seen": 22784864,
      "step": 29875
    },
    {
      "epoch": 62.12058212058212,
      "grad_norm": 0.0005560569115914404,
      "learning_rate": 0.0449467362530486,
      "loss": 0.2713,
      "num_input_tokens_seen": 22788704,
      "step": 29880
    },
    {
      "epoch": 62.13097713097713,
      "grad_norm": 0.0003077409928664565,
      "learning_rate": 0.04490469838120171,
      "loss": 0.2909,
      "num_input_tokens_seen": 22792480,
      "step": 29885
    },
    {
      "epoch": 62.141372141372145,
      "grad_norm": 0.000513959676027298,
      "learning_rate": 0.04486267671637101,
      "loss": 0.2784,
      "num_input_tokens_seen": 22796256,
      "step": 29890
    },
    {
      "epoch": 62.15176715176715,
      "grad_norm": 0.0005690513062290847,
      "learning_rate": 0.04482067126503683,
      "loss": 0.2727,
      "num_input_tokens_seen": 22800128,
      "step": 29895
    },
    {
      "epoch": 62.16216216216216,
      "grad_norm": 0.00024337589275091887,
      "learning_rate": 0.04477868203367687,
      "loss": 0.2658,
      "num_input_tokens_seen": 22803776,
      "step": 29900
    },
    {
      "epoch": 62.17255717255717,
      "grad_norm": 0.00020775568555109203,
      "learning_rate": 0.044736709028766426,
      "loss": 0.2672,
      "num_input_tokens_seen": 22807680,
      "step": 29905
    },
    {
      "epoch": 62.182952182952185,
      "grad_norm": 0.0008607160416431725,
      "learning_rate": 0.04469475225677832,
      "loss": 0.2696,
      "num_input_tokens_seen": 22811456,
      "step": 29910
    },
    {
      "epoch": 62.19334719334719,
      "grad_norm": 0.00029529407038353384,
      "learning_rate": 0.04465281172418273,
      "loss": 0.2648,
      "num_input_tokens_seen": 22815296,
      "step": 29915
    },
    {
      "epoch": 62.2037422037422,
      "grad_norm": 0.0005173709359951317,
      "learning_rate": 0.044610887437447476,
      "loss": 0.2441,
      "num_input_tokens_seen": 22819168,
      "step": 29920
    },
    {
      "epoch": 62.21413721413722,
      "grad_norm": 0.00030752719612792134,
      "learning_rate": 0.044568979403037744,
      "loss": 0.2571,
      "num_input_tokens_seen": 22823136,
      "step": 29925
    },
    {
      "epoch": 62.224532224532226,
      "grad_norm": 0.0001653421059018001,
      "learning_rate": 0.04452708762741631,
      "loss": 0.2562,
      "num_input_tokens_seen": 22826784,
      "step": 29930
    },
    {
      "epoch": 62.234927234927234,
      "grad_norm": 0.0002511379134375602,
      "learning_rate": 0.044485212117043475,
      "loss": 0.2507,
      "num_input_tokens_seen": 22830592,
      "step": 29935
    },
    {
      "epoch": 62.24532224532224,
      "grad_norm": 0.00027206551749259233,
      "learning_rate": 0.04444335287837687,
      "loss": 0.2745,
      "num_input_tokens_seen": 22834464,
      "step": 29940
    },
    {
      "epoch": 62.25571725571726,
      "grad_norm": 0.0003265838313382119,
      "learning_rate": 0.04440150991787179,
      "loss": 0.2732,
      "num_input_tokens_seen": 22838048,
      "step": 29945
    },
    {
      "epoch": 62.266112266112266,
      "grad_norm": 0.0002748042461462319,
      "learning_rate": 0.04435968324198088,
      "loss": 0.2779,
      "num_input_tokens_seen": 22842016,
      "step": 29950
    },
    {
      "epoch": 62.276507276507274,
      "grad_norm": 5.0398575694998726e-05,
      "learning_rate": 0.04431787285715442,
      "loss": 0.262,
      "num_input_tokens_seen": 22845920,
      "step": 29955
    },
    {
      "epoch": 62.28690228690229,
      "grad_norm": 0.0003121078771073371,
      "learning_rate": 0.04427607876984004,
      "loss": 0.2567,
      "num_input_tokens_seen": 22849728,
      "step": 29960
    },
    {
      "epoch": 62.2972972972973,
      "grad_norm": 0.00014542232383973897,
      "learning_rate": 0.044234300986482886,
      "loss": 0.2603,
      "num_input_tokens_seen": 22853568,
      "step": 29965
    },
    {
      "epoch": 62.30769230769231,
      "grad_norm": 0.00026760200853459537,
      "learning_rate": 0.04419253951352566,
      "loss": 0.2539,
      "num_input_tokens_seen": 22857280,
      "step": 29970
    },
    {
      "epoch": 62.318087318087315,
      "grad_norm": 0.0003560234617907554,
      "learning_rate": 0.044150794357408533,
      "loss": 0.2356,
      "num_input_tokens_seen": 22861088,
      "step": 29975
    },
    {
      "epoch": 62.32848232848233,
      "grad_norm": 0.00013567270070780069,
      "learning_rate": 0.044109065524569065,
      "loss": 0.2568,
      "num_input_tokens_seen": 22864928,
      "step": 29980
    },
    {
      "epoch": 62.33887733887734,
      "grad_norm": 5.2321087423479185e-05,
      "learning_rate": 0.0440673530214424,
      "loss": 0.2464,
      "num_input_tokens_seen": 22868704,
      "step": 29985
    },
    {
      "epoch": 62.34927234927235,
      "grad_norm": 0.00024049948842730373,
      "learning_rate": 0.04402565685446117,
      "loss": 0.2572,
      "num_input_tokens_seen": 22872608,
      "step": 29990
    },
    {
      "epoch": 62.35966735966736,
      "grad_norm": 0.00029498376534320414,
      "learning_rate": 0.04398397703005536,
      "loss": 0.2626,
      "num_input_tokens_seen": 22876448,
      "step": 29995
    },
    {
      "epoch": 62.37006237006237,
      "grad_norm": 0.0004322648164816201,
      "learning_rate": 0.043942313554652626,
      "loss": 0.2514,
      "num_input_tokens_seen": 22880256,
      "step": 30000
    },
    {
      "epoch": 62.37006237006237,
      "eval_loss": 0.24741502106189728,
      "eval_runtime": 13.4102,
      "eval_samples_per_second": 63.832,
      "eval_steps_per_second": 15.958,
      "num_input_tokens_seen": 22880256,
      "step": 30000
    },
    {
      "epoch": 62.38045738045738,
      "grad_norm": 0.00032661177101545036,
      "learning_rate": 0.0439006664346779,
      "loss": 0.2654,
      "num_input_tokens_seen": 22883936,
      "step": 30005
    },
    {
      "epoch": 62.39085239085239,
      "grad_norm": 0.00023311520635616034,
      "learning_rate": 0.043859035676553755,
      "loss": 0.2537,
      "num_input_tokens_seen": 22887808,
      "step": 30010
    },
    {
      "epoch": 62.4012474012474,
      "grad_norm": 0.0003312415792606771,
      "learning_rate": 0.043817421286700194,
      "loss": 0.2627,
      "num_input_tokens_seen": 22891680,
      "step": 30015
    },
    {
      "epoch": 62.41164241164241,
      "grad_norm": 9.473994578002021e-05,
      "learning_rate": 0.043775823271534585,
      "loss": 0.2591,
      "num_input_tokens_seen": 22895552,
      "step": 30020
    },
    {
      "epoch": 62.42203742203742,
      "grad_norm": 0.00018004674348048866,
      "learning_rate": 0.04373424163747197,
      "loss": 0.2487,
      "num_input_tokens_seen": 22899360,
      "step": 30025
    },
    {
      "epoch": 62.432432432432435,
      "grad_norm": 0.0005645894561894238,
      "learning_rate": 0.04369267639092473,
      "loss": 0.2394,
      "num_input_tokens_seen": 22903136,
      "step": 30030
    },
    {
      "epoch": 62.44282744282744,
      "grad_norm": 0.00014490137982647866,
      "learning_rate": 0.04365112753830268,
      "loss": 0.2595,
      "num_input_tokens_seen": 22906752,
      "step": 30035
    },
    {
      "epoch": 62.45322245322245,
      "grad_norm": 0.0002702472556848079,
      "learning_rate": 0.04360959508601327,
      "loss": 0.2581,
      "num_input_tokens_seen": 22910688,
      "step": 30040
    },
    {
      "epoch": 62.46361746361746,
      "grad_norm": 0.0002903661224991083,
      "learning_rate": 0.04356807904046123,
      "loss": 0.2449,
      "num_input_tokens_seen": 22914368,
      "step": 30045
    },
    {
      "epoch": 62.474012474012476,
      "grad_norm": 0.000628048786893487,
      "learning_rate": 0.04352657940804892,
      "loss": 0.2634,
      "num_input_tokens_seen": 22918144,
      "step": 30050
    },
    {
      "epoch": 62.484407484407484,
      "grad_norm": 0.0006930386298336089,
      "learning_rate": 0.04348509619517613,
      "loss": 0.2279,
      "num_input_tokens_seen": 22921952,
      "step": 30055
    },
    {
      "epoch": 62.49480249480249,
      "grad_norm": 3.640597788034938e-05,
      "learning_rate": 0.04344362940824002,
      "loss": 0.2583,
      "num_input_tokens_seen": 22925824,
      "step": 30060
    },
    {
      "epoch": 62.50519750519751,
      "grad_norm": 0.00035645178286358714,
      "learning_rate": 0.04340217905363533,
      "loss": 0.2671,
      "num_input_tokens_seen": 22929504,
      "step": 30065
    },
    {
      "epoch": 62.515592515592516,
      "grad_norm": 0.000650412286631763,
      "learning_rate": 0.04336074513775425,
      "loss": 0.2441,
      "num_input_tokens_seen": 22933280,
      "step": 30070
    },
    {
      "epoch": 62.525987525987524,
      "grad_norm": 0.000225075549678877,
      "learning_rate": 0.04331932766698636,
      "loss": 0.2792,
      "num_input_tokens_seen": 22936928,
      "step": 30075
    },
    {
      "epoch": 62.53638253638254,
      "grad_norm": 0.00014629241195507348,
      "learning_rate": 0.0432779266477188,
      "loss": 0.2684,
      "num_input_tokens_seen": 22940864,
      "step": 30080
    },
    {
      "epoch": 62.54677754677755,
      "grad_norm": 6.334840873023495e-05,
      "learning_rate": 0.04323654208633607,
      "loss": 0.2367,
      "num_input_tokens_seen": 22944416,
      "step": 30085
    },
    {
      "epoch": 62.55717255717256,
      "grad_norm": 0.00044576849904842675,
      "learning_rate": 0.04319517398922024,
      "loss": 0.2719,
      "num_input_tokens_seen": 22948160,
      "step": 30090
    },
    {
      "epoch": 62.567567567567565,
      "grad_norm": 0.0011152951046824455,
      "learning_rate": 0.04315382236275079,
      "loss": 0.2533,
      "num_input_tokens_seen": 22951904,
      "step": 30095
    },
    {
      "epoch": 62.57796257796258,
      "grad_norm": 0.0003502096515148878,
      "learning_rate": 0.043112487213304664,
      "loss": 0.2833,
      "num_input_tokens_seen": 22955744,
      "step": 30100
    },
    {
      "epoch": 62.58835758835759,
      "grad_norm": 4.0146405808627605e-05,
      "learning_rate": 0.04307116854725618,
      "loss": 0.2655,
      "num_input_tokens_seen": 22959616,
      "step": 30105
    },
    {
      "epoch": 62.5987525987526,
      "grad_norm": 0.0002620772283989936,
      "learning_rate": 0.043029866370977325,
      "loss": 0.2915,
      "num_input_tokens_seen": 22963520,
      "step": 30110
    },
    {
      "epoch": 62.60914760914761,
      "grad_norm": 0.000122947632917203,
      "learning_rate": 0.04298858069083728,
      "loss": 0.264,
      "num_input_tokens_seen": 22967392,
      "step": 30115
    },
    {
      "epoch": 62.61954261954262,
      "grad_norm": 0.00012421427527442575,
      "learning_rate": 0.04294731151320295,
      "loss": 0.279,
      "num_input_tokens_seen": 22971168,
      "step": 30120
    },
    {
      "epoch": 62.62993762993763,
      "grad_norm": 0.0001404361828463152,
      "learning_rate": 0.04290605884443841,
      "loss": 0.2765,
      "num_input_tokens_seen": 22974944,
      "step": 30125
    },
    {
      "epoch": 62.64033264033264,
      "grad_norm": 0.00047713748062960804,
      "learning_rate": 0.04286482269090545,
      "loss": 0.2605,
      "num_input_tokens_seen": 22978624,
      "step": 30130
    },
    {
      "epoch": 62.65072765072765,
      "grad_norm": 0.000299125473247841,
      "learning_rate": 0.04282360305896323,
      "loss": 0.2679,
      "num_input_tokens_seen": 22982464,
      "step": 30135
    },
    {
      "epoch": 62.66112266112266,
      "grad_norm": 0.00031978421611711383,
      "learning_rate": 0.04278239995496822,
      "loss": 0.2682,
      "num_input_tokens_seen": 22986176,
      "step": 30140
    },
    {
      "epoch": 62.67151767151767,
      "grad_norm": 0.0005218606092967093,
      "learning_rate": 0.042741213385274514,
      "loss": 0.25,
      "num_input_tokens_seen": 22990016,
      "step": 30145
    },
    {
      "epoch": 62.681912681912685,
      "grad_norm": 0.00010784743790281937,
      "learning_rate": 0.04270004335623366,
      "loss": 0.258,
      "num_input_tokens_seen": 22993792,
      "step": 30150
    },
    {
      "epoch": 62.69230769230769,
      "grad_norm": 0.00027760062948800623,
      "learning_rate": 0.04265888987419448,
      "loss": 0.2438,
      "num_input_tokens_seen": 22997568,
      "step": 30155
    },
    {
      "epoch": 62.7027027027027,
      "grad_norm": 0.0004221794370096177,
      "learning_rate": 0.04261775294550346,
      "loss": 0.2593,
      "num_input_tokens_seen": 23001408,
      "step": 30160
    },
    {
      "epoch": 62.71309771309771,
      "grad_norm": 4.818370143766515e-05,
      "learning_rate": 0.042576632576504354,
      "loss": 0.2727,
      "num_input_tokens_seen": 23005312,
      "step": 30165
    },
    {
      "epoch": 62.723492723492726,
      "grad_norm": 0.0001188266251119785,
      "learning_rate": 0.0425355287735385,
      "loss": 0.2882,
      "num_input_tokens_seen": 23009312,
      "step": 30170
    },
    {
      "epoch": 62.733887733887734,
      "grad_norm": 0.000365598127245903,
      "learning_rate": 0.0424944415429446,
      "loss": 0.2876,
      "num_input_tokens_seen": 23013184,
      "step": 30175
    },
    {
      "epoch": 62.74428274428274,
      "grad_norm": 0.0004075798497069627,
      "learning_rate": 0.04245337089105877,
      "loss": 0.2704,
      "num_input_tokens_seen": 23017120,
      "step": 30180
    },
    {
      "epoch": 62.75467775467776,
      "grad_norm": 0.0002494051877874881,
      "learning_rate": 0.04241231682421467,
      "loss": 0.2642,
      "num_input_tokens_seen": 23021088,
      "step": 30185
    },
    {
      "epoch": 62.765072765072766,
      "grad_norm": 0.00021104175539221615,
      "learning_rate": 0.04237127934874337,
      "loss": 0.2617,
      "num_input_tokens_seen": 23025216,
      "step": 30190
    },
    {
      "epoch": 62.775467775467774,
      "grad_norm": 0.0001161335312644951,
      "learning_rate": 0.042330258470973305,
      "loss": 0.2754,
      "num_input_tokens_seen": 23028960,
      "step": 30195
    },
    {
      "epoch": 62.78586278586278,
      "grad_norm": 0.0003201899817213416,
      "learning_rate": 0.042289254197230515,
      "loss": 0.2572,
      "num_input_tokens_seen": 23032800,
      "step": 30200
    },
    {
      "epoch": 62.78586278586278,
      "eval_loss": 0.2501713037490845,
      "eval_runtime": 13.4296,
      "eval_samples_per_second": 63.74,
      "eval_steps_per_second": 15.935,
      "num_input_tokens_seen": 23032800,
      "step": 30200
    },
    {
      "epoch": 62.7962577962578,
      "grad_norm": 0.00010216915688943118,
      "learning_rate": 0.04224826653383823,
      "loss": 0.255,
      "num_input_tokens_seen": 23036704,
      "step": 30205
    },
    {
      "epoch": 62.80665280665281,
      "grad_norm": 8.968092879513279e-05,
      "learning_rate": 0.04220729548711735,
      "loss": 0.2734,
      "num_input_tokens_seen": 23040704,
      "step": 30210
    },
    {
      "epoch": 62.817047817047815,
      "grad_norm": 0.0002949389163404703,
      "learning_rate": 0.04216634106338616,
      "loss": 0.2539,
      "num_input_tokens_seen": 23044544,
      "step": 30215
    },
    {
      "epoch": 62.82744282744283,
      "grad_norm": 0.0002409187873126939,
      "learning_rate": 0.04212540326896025,
      "loss": 0.2474,
      "num_input_tokens_seen": 23048448,
      "step": 30220
    },
    {
      "epoch": 62.83783783783784,
      "grad_norm": 0.0002916075463872403,
      "learning_rate": 0.0420844821101528,
      "loss": 0.2975,
      "num_input_tokens_seen": 23052320,
      "step": 30225
    },
    {
      "epoch": 62.84823284823285,
      "grad_norm": 0.0003147415118291974,
      "learning_rate": 0.04204357759327441,
      "loss": 0.2718,
      "num_input_tokens_seen": 23056000,
      "step": 30230
    },
    {
      "epoch": 62.858627858627855,
      "grad_norm": 0.0001354873675154522,
      "learning_rate": 0.042002689724632954,
      "loss": 0.2607,
      "num_input_tokens_seen": 23059776,
      "step": 30235
    },
    {
      "epoch": 62.86902286902287,
      "grad_norm": 0.0003973265702370554,
      "learning_rate": 0.04196181851053398,
      "loss": 0.2799,
      "num_input_tokens_seen": 23063520,
      "step": 30240
    },
    {
      "epoch": 62.87941787941788,
      "grad_norm": 0.0003862975863739848,
      "learning_rate": 0.041920963957280295,
      "loss": 0.2807,
      "num_input_tokens_seen": 23067200,
      "step": 30245
    },
    {
      "epoch": 62.88981288981289,
      "grad_norm": 0.000357545621227473,
      "learning_rate": 0.04188012607117212,
      "loss": 0.2716,
      "num_input_tokens_seen": 23071008,
      "step": 30250
    },
    {
      "epoch": 62.9002079002079,
      "grad_norm": 0.0001907086552819237,
      "learning_rate": 0.04183930485850725,
      "loss": 0.2727,
      "num_input_tokens_seen": 23074688,
      "step": 30255
    },
    {
      "epoch": 62.91060291060291,
      "grad_norm": 0.00019051968411076814,
      "learning_rate": 0.04179850032558078,
      "loss": 0.2731,
      "num_input_tokens_seen": 23078560,
      "step": 30260
    },
    {
      "epoch": 62.92099792099792,
      "grad_norm": 0.000594455748796463,
      "learning_rate": 0.041757712478685295,
      "loss": 0.2522,
      "num_input_tokens_seen": 23082304,
      "step": 30265
    },
    {
      "epoch": 62.931392931392935,
      "grad_norm": 0.00017930276226252317,
      "learning_rate": 0.04171694132411085,
      "loss": 0.274,
      "num_input_tokens_seen": 23086112,
      "step": 30270
    },
    {
      "epoch": 62.94178794178794,
      "grad_norm": 6.014187238179147e-05,
      "learning_rate": 0.04167618686814479,
      "loss": 0.2598,
      "num_input_tokens_seen": 23089920,
      "step": 30275
    },
    {
      "epoch": 62.95218295218295,
      "grad_norm": 0.0003506779030431062,
      "learning_rate": 0.041635449117072024,
      "loss": 0.2483,
      "num_input_tokens_seen": 23093600,
      "step": 30280
    },
    {
      "epoch": 62.96257796257796,
      "grad_norm": 0.00033882210846059024,
      "learning_rate": 0.04159472807717477,
      "loss": 0.266,
      "num_input_tokens_seen": 23097504,
      "step": 30285
    },
    {
      "epoch": 62.972972972972975,
      "grad_norm": 0.00043650579755194485,
      "learning_rate": 0.041554023754732744,
      "loss": 0.3049,
      "num_input_tokens_seen": 23101344,
      "step": 30290
    },
    {
      "epoch": 62.983367983367984,
      "grad_norm": 0.00032060674857348204,
      "learning_rate": 0.04151333615602311,
      "loss": 0.2749,
      "num_input_tokens_seen": 23105120,
      "step": 30295
    },
    {
      "epoch": 62.99376299376299,
      "grad_norm": 8.948193135438487e-05,
      "learning_rate": 0.04147266528732034,
      "loss": 0.2532,
      "num_input_tokens_seen": 23108896,
      "step": 30300
    },
    {
      "epoch": 63.00415800415801,
      "grad_norm": 0.00022815649572294205,
      "learning_rate": 0.0414320111548964,
      "loss": 0.2616,
      "num_input_tokens_seen": 23112808,
      "step": 30305
    },
    {
      "epoch": 63.014553014553016,
      "grad_norm": 4.847226227866486e-05,
      "learning_rate": 0.04139137376502076,
      "loss": 0.2519,
      "num_input_tokens_seen": 23116488,
      "step": 30310
    },
    {
      "epoch": 63.024948024948024,
      "grad_norm": 0.000665325322188437,
      "learning_rate": 0.04135075312396014,
      "loss": 0.2778,
      "num_input_tokens_seen": 23120360,
      "step": 30315
    },
    {
      "epoch": 63.03534303534303,
      "grad_norm": 0.00036534195533022285,
      "learning_rate": 0.04131014923797875,
      "loss": 0.2797,
      "num_input_tokens_seen": 23124040,
      "step": 30320
    },
    {
      "epoch": 63.04573804573805,
      "grad_norm": 0.0006709165754728019,
      "learning_rate": 0.04126956211333819,
      "loss": 0.2962,
      "num_input_tokens_seen": 23127784,
      "step": 30325
    },
    {
      "epoch": 63.056133056133056,
      "grad_norm": 0.0003930821258109063,
      "learning_rate": 0.041228991756297545,
      "loss": 0.2758,
      "num_input_tokens_seen": 23131560,
      "step": 30330
    },
    {
      "epoch": 63.066528066528065,
      "grad_norm": 0.00016460947517771274,
      "learning_rate": 0.04118843817311332,
      "loss": 0.263,
      "num_input_tokens_seen": 23135336,
      "step": 30335
    },
    {
      "epoch": 63.07692307692308,
      "grad_norm": 8.551117934985086e-05,
      "learning_rate": 0.0411479013700393,
      "loss": 0.2714,
      "num_input_tokens_seen": 23139112,
      "step": 30340
    },
    {
      "epoch": 63.08731808731809,
      "grad_norm": 0.0005092486971989274,
      "learning_rate": 0.0411073813533268,
      "loss": 0.2767,
      "num_input_tokens_seen": 23143016,
      "step": 30345
    },
    {
      "epoch": 63.0977130977131,
      "grad_norm": 0.0003225219261366874,
      "learning_rate": 0.04106687812922456,
      "loss": 0.2649,
      "num_input_tokens_seen": 23146760,
      "step": 30350
    },
    {
      "epoch": 63.108108108108105,
      "grad_norm": 0.00022369003272615373,
      "learning_rate": 0.041026391703978635,
      "loss": 0.2417,
      "num_input_tokens_seen": 23150632,
      "step": 30355
    },
    {
      "epoch": 63.11850311850312,
      "grad_norm": 0.0002041378611465916,
      "learning_rate": 0.04098592208383259,
      "loss": 0.2932,
      "num_input_tokens_seen": 23154376,
      "step": 30360
    },
    {
      "epoch": 63.12889812889813,
      "grad_norm": 0.0004909907002002001,
      "learning_rate": 0.040945469275027256,
      "loss": 0.2585,
      "num_input_tokens_seen": 23158088,
      "step": 30365
    },
    {
      "epoch": 63.13929313929314,
      "grad_norm": 0.00029031949816271663,
      "learning_rate": 0.04090503328380104,
      "loss": 0.279,
      "num_input_tokens_seen": 23161864,
      "step": 30370
    },
    {
      "epoch": 63.14968814968815,
      "grad_norm": 0.00034845375921577215,
      "learning_rate": 0.04086461411638971,
      "loss": 0.2656,
      "num_input_tokens_seen": 23165640,
      "step": 30375
    },
    {
      "epoch": 63.16008316008316,
      "grad_norm": 0.00021484777971636504,
      "learning_rate": 0.04082421177902631,
      "loss": 0.2727,
      "num_input_tokens_seen": 23169480,
      "step": 30380
    },
    {
      "epoch": 63.17047817047817,
      "grad_norm": 0.0002114875242114067,
      "learning_rate": 0.04078382627794149,
      "loss": 0.2623,
      "num_input_tokens_seen": 23173256,
      "step": 30385
    },
    {
      "epoch": 63.18087318087318,
      "grad_norm": 0.0002916967496275902,
      "learning_rate": 0.04074345761936316,
      "loss": 0.2571,
      "num_input_tokens_seen": 23177000,
      "step": 30390
    },
    {
      "epoch": 63.19126819126819,
      "grad_norm": 0.000641339342109859,
      "learning_rate": 0.04070310580951663,
      "loss": 0.2441,
      "num_input_tokens_seen": 23180936,
      "step": 30395
    },
    {
      "epoch": 63.2016632016632,
      "grad_norm": 5.407605567597784e-05,
      "learning_rate": 0.040662770854624726,
      "loss": 0.2698,
      "num_input_tokens_seen": 23184744,
      "step": 30400
    },
    {
      "epoch": 63.2016632016632,
      "eval_loss": 0.24731118977069855,
      "eval_runtime": 13.3928,
      "eval_samples_per_second": 63.915,
      "eval_steps_per_second": 15.979,
      "num_input_tokens_seen": 23184744,
      "step": 30400
    },
    {
      "epoch": 63.21205821205821,
      "grad_norm": 0.0004855773877352476,
      "learning_rate": 0.040622452760907535,
      "loss": 0.2608,
      "num_input_tokens_seen": 23188680,
      "step": 30405
    },
    {
      "epoch": 63.222453222453225,
      "grad_norm": 0.00034704344579949975,
      "learning_rate": 0.04058215153458265,
      "loss": 0.2588,
      "num_input_tokens_seen": 23192424,
      "step": 30410
    },
    {
      "epoch": 63.232848232848234,
      "grad_norm": 0.0005789669812656939,
      "learning_rate": 0.04054186718186507,
      "loss": 0.2786,
      "num_input_tokens_seen": 23196264,
      "step": 30415
    },
    {
      "epoch": 63.24324324324324,
      "grad_norm": 0.0005630963132716715,
      "learning_rate": 0.04050159970896708,
      "loss": 0.2713,
      "num_input_tokens_seen": 23200264,
      "step": 30420
    },
    {
      "epoch": 63.25363825363825,
      "grad_norm": 0.00013894474250264466,
      "learning_rate": 0.04046134912209843,
      "loss": 0.2716,
      "num_input_tokens_seen": 23204040,
      "step": 30425
    },
    {
      "epoch": 63.264033264033266,
      "grad_norm": 0.0004776244459208101,
      "learning_rate": 0.040421115427466354,
      "loss": 0.2415,
      "num_input_tokens_seen": 23207816,
      "step": 30430
    },
    {
      "epoch": 63.274428274428274,
      "grad_norm": 0.0002857446379493922,
      "learning_rate": 0.04038089863127529,
      "loss": 0.2719,
      "num_input_tokens_seen": 23211784,
      "step": 30435
    },
    {
      "epoch": 63.28482328482328,
      "grad_norm": 0.00017533684149384499,
      "learning_rate": 0.04034069873972727,
      "loss": 0.251,
      "num_input_tokens_seen": 23215592,
      "step": 30440
    },
    {
      "epoch": 63.2952182952183,
      "grad_norm": 0.0001463294174754992,
      "learning_rate": 0.040300515759021514,
      "loss": 0.2394,
      "num_input_tokens_seen": 23219400,
      "step": 30445
    },
    {
      "epoch": 63.305613305613306,
      "grad_norm": 0.0003088580851908773,
      "learning_rate": 0.04026034969535478,
      "loss": 0.2636,
      "num_input_tokens_seen": 23223208,
      "step": 30450
    },
    {
      "epoch": 63.316008316008315,
      "grad_norm": 0.0003944539639633149,
      "learning_rate": 0.040220200554921266,
      "loss": 0.2533,
      "num_input_tokens_seen": 23227144,
      "step": 30455
    },
    {
      "epoch": 63.32640332640332,
      "grad_norm": 0.00015389127656817436,
      "learning_rate": 0.0401800683439124,
      "loss": 0.2632,
      "num_input_tokens_seen": 23230888,
      "step": 30460
    },
    {
      "epoch": 63.33679833679834,
      "grad_norm": 4.685593376052566e-05,
      "learning_rate": 0.04013995306851704,
      "loss": 0.2674,
      "num_input_tokens_seen": 23234600,
      "step": 30465
    },
    {
      "epoch": 63.34719334719335,
      "grad_norm": 0.0005774143501184881,
      "learning_rate": 0.040099854734921545,
      "loss": 0.2628,
      "num_input_tokens_seen": 23238376,
      "step": 30470
    },
    {
      "epoch": 63.357588357588355,
      "grad_norm": 0.0001762321189744398,
      "learning_rate": 0.0400597733493095,
      "loss": 0.2847,
      "num_input_tokens_seen": 23242120,
      "step": 30475
    },
    {
      "epoch": 63.36798336798337,
      "grad_norm": 0.00026515021454542875,
      "learning_rate": 0.04001970891786203,
      "loss": 0.2557,
      "num_input_tokens_seen": 23245832,
      "step": 30480
    },
    {
      "epoch": 63.37837837837838,
      "grad_norm": 0.0002766774268820882,
      "learning_rate": 0.03997966144675752,
      "loss": 0.2776,
      "num_input_tokens_seen": 23249576,
      "step": 30485
    },
    {
      "epoch": 63.38877338877339,
      "grad_norm": 0.0006170090055093169,
      "learning_rate": 0.039939630942171796,
      "loss": 0.2499,
      "num_input_tokens_seen": 23253256,
      "step": 30490
    },
    {
      "epoch": 63.3991683991684,
      "grad_norm": 7.146267307689413e-05,
      "learning_rate": 0.03989961741027815,
      "loss": 0.2632,
      "num_input_tokens_seen": 23257000,
      "step": 30495
    },
    {
      "epoch": 63.40956340956341,
      "grad_norm": 0.0002587572962511331,
      "learning_rate": 0.03985962085724704,
      "loss": 0.2648,
      "num_input_tokens_seen": 23260904,
      "step": 30500
    },
    {
      "epoch": 63.41995841995842,
      "grad_norm": 0.0002582099987193942,
      "learning_rate": 0.03981964128924656,
      "loss": 0.2585,
      "num_input_tokens_seen": 23264680,
      "step": 30505
    },
    {
      "epoch": 63.43035343035343,
      "grad_norm": 0.00040622326196171343,
      "learning_rate": 0.03977967871244197,
      "loss": 0.2457,
      "num_input_tokens_seen": 23268584,
      "step": 30510
    },
    {
      "epoch": 63.44074844074844,
      "grad_norm": 0.0006614467711187899,
      "learning_rate": 0.03973973313299602,
      "loss": 0.2454,
      "num_input_tokens_seen": 23272392,
      "step": 30515
    },
    {
      "epoch": 63.45114345114345,
      "grad_norm": 0.0002748437982518226,
      "learning_rate": 0.0396998045570689,
      "loss": 0.2839,
      "num_input_tokens_seen": 23276168,
      "step": 30520
    },
    {
      "epoch": 63.46153846153846,
      "grad_norm": 0.000268623378360644,
      "learning_rate": 0.03965989299081798,
      "loss": 0.262,
      "num_input_tokens_seen": 23279880,
      "step": 30525
    },
    {
      "epoch": 63.471933471933475,
      "grad_norm": 0.00028671478503383696,
      "learning_rate": 0.039619998440398235,
      "loss": 0.2697,
      "num_input_tokens_seen": 23283656,
      "step": 30530
    },
    {
      "epoch": 63.482328482328484,
      "grad_norm": 0.0003019487194251269,
      "learning_rate": 0.03958012091196184,
      "loss": 0.2601,
      "num_input_tokens_seen": 23287656,
      "step": 30535
    },
    {
      "epoch": 63.49272349272349,
      "grad_norm": 0.00023189772036857903,
      "learning_rate": 0.039540260411658396,
      "loss": 0.2404,
      "num_input_tokens_seen": 23291528,
      "step": 30540
    },
    {
      "epoch": 63.5031185031185,
      "grad_norm": 8.467365842079744e-05,
      "learning_rate": 0.03950041694563496,
      "loss": 0.2602,
      "num_input_tokens_seen": 23295272,
      "step": 30545
    },
    {
      "epoch": 63.513513513513516,
      "grad_norm": 8.710068505024537e-05,
      "learning_rate": 0.0394605905200358,
      "loss": 0.2761,
      "num_input_tokens_seen": 23299112,
      "step": 30550
    },
    {
      "epoch": 63.523908523908524,
      "grad_norm": 5.7084031141130254e-05,
      "learning_rate": 0.03942078114100272,
      "loss": 0.2811,
      "num_input_tokens_seen": 23302952,
      "step": 30555
    },
    {
      "epoch": 63.53430353430353,
      "grad_norm": 0.0003650786529760808,
      "learning_rate": 0.03938098881467485,
      "loss": 0.2568,
      "num_input_tokens_seen": 23306792,
      "step": 30560
    },
    {
      "epoch": 63.54469854469855,
      "grad_norm": 0.00011326957610435784,
      "learning_rate": 0.039341213547188586,
      "loss": 0.2575,
      "num_input_tokens_seen": 23310504,
      "step": 30565
    },
    {
      "epoch": 63.555093555093556,
      "grad_norm": 0.00041375195723958313,
      "learning_rate": 0.03930145534467782,
      "loss": 0.2562,
      "num_input_tokens_seen": 23314344,
      "step": 30570
    },
    {
      "epoch": 63.565488565488565,
      "grad_norm": 0.00056034279987216,
      "learning_rate": 0.0392617142132738,
      "loss": 0.2801,
      "num_input_tokens_seen": 23318152,
      "step": 30575
    },
    {
      "epoch": 63.57588357588357,
      "grad_norm": 0.0005134684033691883,
      "learning_rate": 0.03922199015910504,
      "loss": 0.2602,
      "num_input_tokens_seen": 23321992,
      "step": 30580
    },
    {
      "epoch": 63.58627858627859,
      "grad_norm": 0.0005481477710418403,
      "learning_rate": 0.039182283188297556,
      "loss": 0.2453,
      "num_input_tokens_seen": 23325672,
      "step": 30585
    },
    {
      "epoch": 63.5966735966736,
      "grad_norm": 0.0004044313100166619,
      "learning_rate": 0.039142593306974595,
      "loss": 0.2633,
      "num_input_tokens_seen": 23329416,
      "step": 30590
    },
    {
      "epoch": 63.607068607068605,
      "grad_norm": 0.0006246328121051192,
      "learning_rate": 0.039102920521256856,
      "loss": 0.2418,
      "num_input_tokens_seen": 23333128,
      "step": 30595
    },
    {
      "epoch": 63.61746361746362,
      "grad_norm": 0.0004957814235240221,
      "learning_rate": 0.03906326483726243,
      "loss": 0.2654,
      "num_input_tokens_seen": 23336904,
      "step": 30600
    },
    {
      "epoch": 63.61746361746362,
      "eval_loss": 0.24733735620975494,
      "eval_runtime": 13.3842,
      "eval_samples_per_second": 63.956,
      "eval_steps_per_second": 15.989,
      "num_input_tokens_seen": 23336904,
      "step": 30600
    },
    {
      "epoch": 63.62785862785863,
      "grad_norm": 0.00015600465121679008,
      "learning_rate": 0.039023626261106704,
      "loss": 0.2626,
      "num_input_tokens_seen": 23340712,
      "step": 30605
    },
    {
      "epoch": 63.63825363825364,
      "grad_norm": 0.0005576710682362318,
      "learning_rate": 0.03898400479890237,
      "loss": 0.2677,
      "num_input_tokens_seen": 23344456,
      "step": 30610
    },
    {
      "epoch": 63.648648648648646,
      "grad_norm": 0.00016586252604611218,
      "learning_rate": 0.038944400456759655,
      "loss": 0.2841,
      "num_input_tokens_seen": 23348296,
      "step": 30615
    },
    {
      "epoch": 63.65904365904366,
      "grad_norm": 0.00016067402611952275,
      "learning_rate": 0.038904813240785964,
      "loss": 0.2699,
      "num_input_tokens_seen": 23352008,
      "step": 30620
    },
    {
      "epoch": 63.66943866943867,
      "grad_norm": 0.0003051013336516917,
      "learning_rate": 0.03886524315708621,
      "loss": 0.282,
      "num_input_tokens_seen": 23355912,
      "step": 30625
    },
    {
      "epoch": 63.67983367983368,
      "grad_norm": 0.0003926317731384188,
      "learning_rate": 0.03882569021176255,
      "loss": 0.2686,
      "num_input_tokens_seen": 23359720,
      "step": 30630
    },
    {
      "epoch": 63.69022869022869,
      "grad_norm": 0.0003085850621573627,
      "learning_rate": 0.038786154410914535,
      "loss": 0.2688,
      "num_input_tokens_seen": 23363432,
      "step": 30635
    },
    {
      "epoch": 63.7006237006237,
      "grad_norm": 0.0001126511488109827,
      "learning_rate": 0.03874663576063917,
      "loss": 0.2733,
      "num_input_tokens_seen": 23367240,
      "step": 30640
    },
    {
      "epoch": 63.71101871101871,
      "grad_norm": 0.0003682049282360822,
      "learning_rate": 0.038707134267030624,
      "loss": 0.2633,
      "num_input_tokens_seen": 23371016,
      "step": 30645
    },
    {
      "epoch": 63.72141372141372,
      "grad_norm": 0.00016524641250725836,
      "learning_rate": 0.038667649936180555,
      "loss": 0.2532,
      "num_input_tokens_seen": 23374856,
      "step": 30650
    },
    {
      "epoch": 63.731808731808734,
      "grad_norm": 0.0005512066418305039,
      "learning_rate": 0.038628182774178,
      "loss": 0.2502,
      "num_input_tokens_seen": 23378696,
      "step": 30655
    },
    {
      "epoch": 63.74220374220374,
      "grad_norm": 8.064149733399972e-05,
      "learning_rate": 0.038588732787109226,
      "loss": 0.208,
      "num_input_tokens_seen": 23382600,
      "step": 30660
    },
    {
      "epoch": 63.75259875259875,
      "grad_norm": 0.000163375458214432,
      "learning_rate": 0.03854929998105795,
      "loss": 0.2827,
      "num_input_tokens_seen": 23386376,
      "step": 30665
    },
    {
      "epoch": 63.762993762993766,
      "grad_norm": 0.00027057400438934565,
      "learning_rate": 0.03850988436210518,
      "loss": 0.2534,
      "num_input_tokens_seen": 23390184,
      "step": 30670
    },
    {
      "epoch": 63.773388773388774,
      "grad_norm": 0.0002026645961450413,
      "learning_rate": 0.03847048593632933,
      "loss": 0.2715,
      "num_input_tokens_seen": 23393896,
      "step": 30675
    },
    {
      "epoch": 63.78378378378378,
      "grad_norm": 0.0003511596005409956,
      "learning_rate": 0.038431104709806096,
      "loss": 0.2964,
      "num_input_tokens_seen": 23397768,
      "step": 30680
    },
    {
      "epoch": 63.79417879417879,
      "grad_norm": 0.00010664512228686363,
      "learning_rate": 0.0383917406886086,
      "loss": 0.251,
      "num_input_tokens_seen": 23401704,
      "step": 30685
    },
    {
      "epoch": 63.804573804573806,
      "grad_norm": 0.0003200983046554029,
      "learning_rate": 0.03835239387880722,
      "loss": 0.2759,
      "num_input_tokens_seen": 23405544,
      "step": 30690
    },
    {
      "epoch": 63.814968814968815,
      "grad_norm": 0.00023513525957241654,
      "learning_rate": 0.03831306428646979,
      "loss": 0.2628,
      "num_input_tokens_seen": 23409352,
      "step": 30695
    },
    {
      "epoch": 63.82536382536382,
      "grad_norm": 0.00034564806264825165,
      "learning_rate": 0.03827375191766135,
      "loss": 0.2608,
      "num_input_tokens_seen": 23413192,
      "step": 30700
    },
    {
      "epoch": 63.83575883575884,
      "grad_norm": 0.0002446005237288773,
      "learning_rate": 0.03823445677844446,
      "loss": 0.2692,
      "num_input_tokens_seen": 23417320,
      "step": 30705
    },
    {
      "epoch": 63.84615384615385,
      "grad_norm": 0.0004821166512556374,
      "learning_rate": 0.03819517887487881,
      "loss": 0.2616,
      "num_input_tokens_seen": 23421096,
      "step": 30710
    },
    {
      "epoch": 63.856548856548855,
      "grad_norm": 0.0003033368266187608,
      "learning_rate": 0.03815591821302161,
      "loss": 0.2618,
      "num_input_tokens_seen": 23424968,
      "step": 30715
    },
    {
      "epoch": 63.86694386694387,
      "grad_norm": 0.00035398584441281855,
      "learning_rate": 0.03811667479892739,
      "loss": 0.2496,
      "num_input_tokens_seen": 23428808,
      "step": 30720
    },
    {
      "epoch": 63.87733887733888,
      "grad_norm": 0.0001340054877800867,
      "learning_rate": 0.03807744863864788,
      "loss": 0.2722,
      "num_input_tokens_seen": 23432488,
      "step": 30725
    },
    {
      "epoch": 63.88773388773389,
      "grad_norm": 0.00022208612062968314,
      "learning_rate": 0.03803823973823229,
      "loss": 0.2709,
      "num_input_tokens_seen": 23436328,
      "step": 30730
    },
    {
      "epoch": 63.898128898128896,
      "grad_norm": 0.0001387394586345181,
      "learning_rate": 0.03799904810372719,
      "loss": 0.2674,
      "num_input_tokens_seen": 23440168,
      "step": 30735
    },
    {
      "epoch": 63.90852390852391,
      "grad_norm": 0.000780729460529983,
      "learning_rate": 0.03795987374117632,
      "loss": 0.2339,
      "num_input_tokens_seen": 23443848,
      "step": 30740
    },
    {
      "epoch": 63.91891891891892,
      "grad_norm": 0.0001924822136061266,
      "learning_rate": 0.03792071665662093,
      "loss": 0.2713,
      "num_input_tokens_seen": 23447752,
      "step": 30745
    },
    {
      "epoch": 63.92931392931393,
      "grad_norm": 0.00019927808898501098,
      "learning_rate": 0.03788157685609952,
      "loss": 0.2752,
      "num_input_tokens_seen": 23451496,
      "step": 30750
    },
    {
      "epoch": 63.93970893970894,
      "grad_norm": 0.0011418135836720467,
      "learning_rate": 0.037842454345647876,
      "loss": 0.2657,
      "num_input_tokens_seen": 23455240,
      "step": 30755
    },
    {
      "epoch": 63.95010395010395,
      "grad_norm": 0.00015185806842055172,
      "learning_rate": 0.03780334913129929,
      "loss": 0.2805,
      "num_input_tokens_seen": 23459016,
      "step": 30760
    },
    {
      "epoch": 63.96049896049896,
      "grad_norm": 0.0014124070294201374,
      "learning_rate": 0.037764261219084175,
      "loss": 0.2666,
      "num_input_tokens_seen": 23462760,
      "step": 30765
    },
    {
      "epoch": 63.97089397089397,
      "grad_norm": 0.0005831573507748544,
      "learning_rate": 0.037725190615030414,
      "loss": 0.2548,
      "num_input_tokens_seen": 23466504,
      "step": 30770
    },
    {
      "epoch": 63.981288981288984,
      "grad_norm": 0.0003159177431371063,
      "learning_rate": 0.037686137325163224,
      "loss": 0.2724,
      "num_input_tokens_seen": 23470312,
      "step": 30775
    },
    {
      "epoch": 63.99168399168399,
      "grad_norm": 0.0002493836800567806,
      "learning_rate": 0.037647101355505065,
      "loss": 0.2551,
      "num_input_tokens_seen": 23474152,
      "step": 30780
    },
    {
      "epoch": 64.002079002079,
      "grad_norm": 0.0005750659620389342,
      "learning_rate": 0.03760808271207581,
      "loss": 0.2265,
      "num_input_tokens_seen": 23478072,
      "step": 30785
    },
    {
      "epoch": 64.01247401247402,
      "grad_norm": 0.00030464300652965903,
      "learning_rate": 0.03756908140089258,
      "loss": 0.258,
      "num_input_tokens_seen": 23481912,
      "step": 30790
    },
    {
      "epoch": 64.02286902286902,
      "grad_norm": 0.0003069822851102799,
      "learning_rate": 0.03753009742796989,
      "loss": 0.2443,
      "num_input_tokens_seen": 23485624,
      "step": 30795
    },
    {
      "epoch": 64.03326403326403,
      "grad_norm": 0.00019055369193665683,
      "learning_rate": 0.037491130799319615,
      "loss": 0.2354,
      "num_input_tokens_seen": 23489432,
      "step": 30800
    },
    {
      "epoch": 64.03326403326403,
      "eval_loss": 0.24824325740337372,
      "eval_runtime": 13.4147,
      "eval_samples_per_second": 63.81,
      "eval_steps_per_second": 15.953,
      "num_input_tokens_seen": 23489432,
      "step": 30800
    },
    {
      "epoch": 64.04365904365905,
      "grad_norm": 0.00020071028848178685,
      "learning_rate": 0.03745218152095079,
      "loss": 0.2389,
      "num_input_tokens_seen": 23493208,
      "step": 30805
    },
    {
      "epoch": 64.05405405405405,
      "grad_norm": 0.0004476974718272686,
      "learning_rate": 0.037413249598869935,
      "loss": 0.242,
      "num_input_tokens_seen": 23497112,
      "step": 30810
    },
    {
      "epoch": 64.06444906444906,
      "grad_norm": 0.0002993455855175853,
      "learning_rate": 0.037374335039080886,
      "loss": 0.2917,
      "num_input_tokens_seen": 23500824,
      "step": 30815
    },
    {
      "epoch": 64.07484407484408,
      "grad_norm": 0.00014793386799283326,
      "learning_rate": 0.037335437847584724,
      "loss": 0.2443,
      "num_input_tokens_seen": 23504536,
      "step": 30820
    },
    {
      "epoch": 64.08523908523908,
      "grad_norm": 0.0008046767907217145,
      "learning_rate": 0.03729655803037983,
      "loss": 0.2566,
      "num_input_tokens_seen": 23508280,
      "step": 30825
    },
    {
      "epoch": 64.0956340956341,
      "grad_norm": 0.00010719981219153851,
      "learning_rate": 0.03725769559346207,
      "loss": 0.2564,
      "num_input_tokens_seen": 23512120,
      "step": 30830
    },
    {
      "epoch": 64.10602910602911,
      "grad_norm": 0.00033316988265141845,
      "learning_rate": 0.03721885054282439,
      "loss": 0.2662,
      "num_input_tokens_seen": 23515960,
      "step": 30835
    },
    {
      "epoch": 64.11642411642411,
      "grad_norm": 8.170922956196591e-05,
      "learning_rate": 0.03718002288445731,
      "loss": 0.2375,
      "num_input_tokens_seen": 23519864,
      "step": 30840
    },
    {
      "epoch": 64.12681912681913,
      "grad_norm": 0.0003092300903517753,
      "learning_rate": 0.03714121262434844,
      "loss": 0.2636,
      "num_input_tokens_seen": 23523480,
      "step": 30845
    },
    {
      "epoch": 64.13721413721414,
      "grad_norm": 0.0001744226028677076,
      "learning_rate": 0.037102419768482844,
      "loss": 0.2501,
      "num_input_tokens_seen": 23527320,
      "step": 30850
    },
    {
      "epoch": 64.14760914760915,
      "grad_norm": 0.0006735483766533434,
      "learning_rate": 0.03706364432284293,
      "loss": 0.2805,
      "num_input_tokens_seen": 23531096,
      "step": 30855
    },
    {
      "epoch": 64.15800415800416,
      "grad_norm": 0.0006111849215812981,
      "learning_rate": 0.03702488629340828,
      "loss": 0.2558,
      "num_input_tokens_seen": 23534840,
      "step": 30860
    },
    {
      "epoch": 64.16839916839916,
      "grad_norm": 4.212605927023105e-05,
      "learning_rate": 0.036986145686155915,
      "loss": 0.2687,
      "num_input_tokens_seen": 23538840,
      "step": 30865
    },
    {
      "epoch": 64.17879417879418,
      "grad_norm": 6.692374881822616e-05,
      "learning_rate": 0.036947422507060075,
      "loss": 0.2598,
      "num_input_tokens_seen": 23542776,
      "step": 30870
    },
    {
      "epoch": 64.1891891891892,
      "grad_norm": 6.341277912724763e-05,
      "learning_rate": 0.0369087167620924,
      "loss": 0.2662,
      "num_input_tokens_seen": 23546552,
      "step": 30875
    },
    {
      "epoch": 64.1995841995842,
      "grad_norm": 0.0004200090479571372,
      "learning_rate": 0.03687002845722183,
      "loss": 0.2755,
      "num_input_tokens_seen": 23550232,
      "step": 30880
    },
    {
      "epoch": 64.20997920997921,
      "grad_norm": 0.00017318730533588678,
      "learning_rate": 0.03683135759841451,
      "loss": 0.2828,
      "num_input_tokens_seen": 23554008,
      "step": 30885
    },
    {
      "epoch": 64.22037422037423,
      "grad_norm": 0.00021113459661137313,
      "learning_rate": 0.03679270419163406,
      "loss": 0.2544,
      "num_input_tokens_seen": 23557848,
      "step": 30890
    },
    {
      "epoch": 64.23076923076923,
      "grad_norm": 0.0005112666985951364,
      "learning_rate": 0.03675406824284127,
      "loss": 0.2622,
      "num_input_tokens_seen": 23561592,
      "step": 30895
    },
    {
      "epoch": 64.24116424116424,
      "grad_norm": 0.0010388800874352455,
      "learning_rate": 0.03671544975799425,
      "loss": 0.2602,
      "num_input_tokens_seen": 23565368,
      "step": 30900
    },
    {
      "epoch": 64.25155925155926,
      "grad_norm": 0.00020329933613538742,
      "learning_rate": 0.03667684874304854,
      "loss": 0.2532,
      "num_input_tokens_seen": 23569144,
      "step": 30905
    },
    {
      "epoch": 64.26195426195426,
      "grad_norm": 0.00011772145080612972,
      "learning_rate": 0.03663826520395683,
      "loss": 0.2624,
      "num_input_tokens_seen": 23572920,
      "step": 30910
    },
    {
      "epoch": 64.27234927234927,
      "grad_norm": 6.005024624755606e-05,
      "learning_rate": 0.03659969914666922,
      "loss": 0.2896,
      "num_input_tokens_seen": 23576536,
      "step": 30915
    },
    {
      "epoch": 64.28274428274429,
      "grad_norm": 0.00010861018381547183,
      "learning_rate": 0.036561150577133106,
      "loss": 0.2777,
      "num_input_tokens_seen": 23580120,
      "step": 30920
    },
    {
      "epoch": 64.29313929313929,
      "grad_norm": 0.000434287911048159,
      "learning_rate": 0.036522619501293103,
      "loss": 0.2672,
      "num_input_tokens_seen": 23584056,
      "step": 30925
    },
    {
      "epoch": 64.3035343035343,
      "grad_norm": 0.0006371169583871961,
      "learning_rate": 0.03648410592509122,
      "loss": 0.2638,
      "num_input_tokens_seen": 23587960,
      "step": 30930
    },
    {
      "epoch": 64.31392931392931,
      "grad_norm": 0.0013753941748291254,
      "learning_rate": 0.03644560985446676,
      "loss": 0.2777,
      "num_input_tokens_seen": 23591832,
      "step": 30935
    },
    {
      "epoch": 64.32432432432432,
      "grad_norm": 0.0001845574879553169,
      "learning_rate": 0.036407131295356256,
      "loss": 0.2692,
      "num_input_tokens_seen": 23595544,
      "step": 30940
    },
    {
      "epoch": 64.33471933471934,
      "grad_norm": 0.00018170123803429306,
      "learning_rate": 0.03636867025369362,
      "loss": 0.2928,
      "num_input_tokens_seen": 23599320,
      "step": 30945
    },
    {
      "epoch": 64.34511434511434,
      "grad_norm": 0.000825926021207124,
      "learning_rate": 0.03633022673540999,
      "loss": 0.2886,
      "num_input_tokens_seen": 23603192,
      "step": 30950
    },
    {
      "epoch": 64.35550935550935,
      "grad_norm": 0.000158306400408037,
      "learning_rate": 0.03629180074643385,
      "loss": 0.2624,
      "num_input_tokens_seen": 23607032,
      "step": 30955
    },
    {
      "epoch": 64.36590436590437,
      "grad_norm": 0.0002621819730848074,
      "learning_rate": 0.03625339229269102,
      "loss": 0.2713,
      "num_input_tokens_seen": 23611000,
      "step": 30960
    },
    {
      "epoch": 64.37629937629937,
      "grad_norm": 0.0008049997850321233,
      "learning_rate": 0.036215001380104535,
      "loss": 0.2635,
      "num_input_tokens_seen": 23614808,
      "step": 30965
    },
    {
      "epoch": 64.38669438669439,
      "grad_norm": 0.0006757226074114442,
      "learning_rate": 0.03617662801459471,
      "loss": 0.2601,
      "num_input_tokens_seen": 23618616,
      "step": 30970
    },
    {
      "epoch": 64.3970893970894,
      "grad_norm": 0.0002419425145490095,
      "learning_rate": 0.036138272202079276,
      "loss": 0.2487,
      "num_input_tokens_seen": 23622616,
      "step": 30975
    },
    {
      "epoch": 64.4074844074844,
      "grad_norm": 0.00035483541432768106,
      "learning_rate": 0.036099933948473106,
      "loss": 0.2552,
      "num_input_tokens_seen": 23626456,
      "step": 30980
    },
    {
      "epoch": 64.41787941787942,
      "grad_norm": 5.000762394047342e-05,
      "learning_rate": 0.03606161325968851,
      "loss": 0.248,
      "num_input_tokens_seen": 23630232,
      "step": 30985
    },
    {
      "epoch": 64.42827442827443,
      "grad_norm": 0.0008898017113097012,
      "learning_rate": 0.03602331014163496,
      "loss": 0.2589,
      "num_input_tokens_seen": 23633944,
      "step": 30990
    },
    {
      "epoch": 64.43866943866944,
      "grad_norm": 0.00027639957261271775,
      "learning_rate": 0.035985024600219295,
      "loss": 0.2651,
      "num_input_tokens_seen": 23637784,
      "step": 30995
    },
    {
      "epoch": 64.44906444906445,
      "grad_norm": 6.263157411012799e-05,
      "learning_rate": 0.03594675664134569,
      "loss": 0.2766,
      "num_input_tokens_seen": 23641496,
      "step": 31000
    },
    {
      "epoch": 64.44906444906445,
      "eval_loss": 0.24707148969173431,
      "eval_runtime": 13.3957,
      "eval_samples_per_second": 63.901,
      "eval_steps_per_second": 15.975,
      "num_input_tokens_seen": 23641496,
      "step": 31000
    },
    {
      "epoch": 64.45945945945945,
      "grad_norm": 0.0002089206245727837,
      "learning_rate": 0.03590850627091545,
      "loss": 0.2955,
      "num_input_tokens_seen": 23645464,
      "step": 31005
    },
    {
      "epoch": 64.46985446985447,
      "grad_norm": 0.00026809112750925124,
      "learning_rate": 0.03587027349482731,
      "loss": 0.2624,
      "num_input_tokens_seen": 23649304,
      "step": 31010
    },
    {
      "epoch": 64.48024948024948,
      "grad_norm": 0.0002953018993139267,
      "learning_rate": 0.035832058318977275,
      "loss": 0.2628,
      "num_input_tokens_seen": 23653048,
      "step": 31015
    },
    {
      "epoch": 64.49064449064448,
      "grad_norm": 0.0022743260487914085,
      "learning_rate": 0.03579386074925853,
      "loss": 0.2731,
      "num_input_tokens_seen": 23656760,
      "step": 31020
    },
    {
      "epoch": 64.5010395010395,
      "grad_norm": 0.0003832119982689619,
      "learning_rate": 0.035755680791561696,
      "loss": 0.2672,
      "num_input_tokens_seen": 23660760,
      "step": 31025
    },
    {
      "epoch": 64.51143451143452,
      "grad_norm": 0.0004797054280061275,
      "learning_rate": 0.03571751845177454,
      "loss": 0.2683,
      "num_input_tokens_seen": 23664408,
      "step": 31030
    },
    {
      "epoch": 64.52182952182952,
      "grad_norm": 0.00039895178633742034,
      "learning_rate": 0.03567937373578225,
      "loss": 0.2797,
      "num_input_tokens_seen": 23668216,
      "step": 31035
    },
    {
      "epoch": 64.53222453222453,
      "grad_norm": 0.0002463925047777593,
      "learning_rate": 0.03564124664946711,
      "loss": 0.2558,
      "num_input_tokens_seen": 23672152,
      "step": 31040
    },
    {
      "epoch": 64.54261954261955,
      "grad_norm": 0.0003663892566692084,
      "learning_rate": 0.035603137198708924,
      "loss": 0.2539,
      "num_input_tokens_seen": 23675928,
      "step": 31045
    },
    {
      "epoch": 64.55301455301455,
      "grad_norm": 0.0004516581248026341,
      "learning_rate": 0.035565045389384514,
      "loss": 0.2702,
      "num_input_tokens_seen": 23679896,
      "step": 31050
    },
    {
      "epoch": 64.56340956340956,
      "grad_norm": 0.0002812107268255204,
      "learning_rate": 0.03552697122736823,
      "loss": 0.252,
      "num_input_tokens_seen": 23683768,
      "step": 31055
    },
    {
      "epoch": 64.57380457380458,
      "grad_norm": 0.00047311969683505595,
      "learning_rate": 0.03548891471853153,
      "loss": 0.2852,
      "num_input_tokens_seen": 23687480,
      "step": 31060
    },
    {
      "epoch": 64.58419958419958,
      "grad_norm": 0.00019577915372792631,
      "learning_rate": 0.03545087586874322,
      "loss": 0.2663,
      "num_input_tokens_seen": 23691416,
      "step": 31065
    },
    {
      "epoch": 64.5945945945946,
      "grad_norm": 0.00027959910221397877,
      "learning_rate": 0.03541285468386935,
      "loss": 0.2552,
      "num_input_tokens_seen": 23695192,
      "step": 31070
    },
    {
      "epoch": 64.60498960498961,
      "grad_norm": 0.00040719486423768103,
      "learning_rate": 0.03537485116977327,
      "loss": 0.2453,
      "num_input_tokens_seen": 23699128,
      "step": 31075
    },
    {
      "epoch": 64.61538461538461,
      "grad_norm": 0.00043097263551317155,
      "learning_rate": 0.03533686533231565,
      "loss": 0.2841,
      "num_input_tokens_seen": 23703000,
      "step": 31080
    },
    {
      "epoch": 64.62577962577963,
      "grad_norm": 0.00019425485515967011,
      "learning_rate": 0.0352988971773543,
      "loss": 0.2904,
      "num_input_tokens_seen": 23706808,
      "step": 31085
    },
    {
      "epoch": 64.63617463617463,
      "grad_norm": 0.0002332548756385222,
      "learning_rate": 0.03526094671074443,
      "loss": 0.2699,
      "num_input_tokens_seen": 23710552,
      "step": 31090
    },
    {
      "epoch": 64.64656964656965,
      "grad_norm": 0.00013271684292703867,
      "learning_rate": 0.03522301393833852,
      "loss": 0.2704,
      "num_input_tokens_seen": 23714360,
      "step": 31095
    },
    {
      "epoch": 64.65696465696466,
      "grad_norm": 0.0008996139513328671,
      "learning_rate": 0.035185098865986204,
      "loss": 0.2673,
      "num_input_tokens_seen": 23718328,
      "step": 31100
    },
    {
      "epoch": 64.66735966735966,
      "grad_norm": 0.00011638574505923316,
      "learning_rate": 0.03514720149953453,
      "loss": 0.2728,
      "num_input_tokens_seen": 23722104,
      "step": 31105
    },
    {
      "epoch": 64.67775467775468,
      "grad_norm": 0.00015800939581822604,
      "learning_rate": 0.03510932184482773,
      "loss": 0.2621,
      "num_input_tokens_seen": 23726040,
      "step": 31110
    },
    {
      "epoch": 64.6881496881497,
      "grad_norm": 0.0001915096363518387,
      "learning_rate": 0.03507145990770724,
      "loss": 0.2627,
      "num_input_tokens_seen": 23729880,
      "step": 31115
    },
    {
      "epoch": 64.6985446985447,
      "grad_norm": 0.0006221798830665648,
      "learning_rate": 0.035033615694011984,
      "loss": 0.2645,
      "num_input_tokens_seen": 23733624,
      "step": 31120
    },
    {
      "epoch": 64.70893970893971,
      "grad_norm": 0.00021186945377849042,
      "learning_rate": 0.03499578920957788,
      "loss": 0.2537,
      "num_input_tokens_seen": 23737464,
      "step": 31125
    },
    {
      "epoch": 64.71933471933473,
      "grad_norm": 0.0002337249752599746,
      "learning_rate": 0.034957980460238375,
      "loss": 0.2596,
      "num_input_tokens_seen": 23741400,
      "step": 31130
    },
    {
      "epoch": 64.72972972972973,
      "grad_norm": 0.00018540085875429213,
      "learning_rate": 0.03492018945182393,
      "loss": 0.2783,
      "num_input_tokens_seen": 23745208,
      "step": 31135
    },
    {
      "epoch": 64.74012474012474,
      "grad_norm": 0.0003771416377276182,
      "learning_rate": 0.03488241619016247,
      "loss": 0.2755,
      "num_input_tokens_seen": 23748952,
      "step": 31140
    },
    {
      "epoch": 64.75051975051976,
      "grad_norm": 0.0003336940426379442,
      "learning_rate": 0.03484466068107913,
      "loss": 0.2746,
      "num_input_tokens_seen": 23752632,
      "step": 31145
    },
    {
      "epoch": 64.76091476091476,
      "grad_norm": 0.0002736676251515746,
      "learning_rate": 0.034806922930396195,
      "loss": 0.2422,
      "num_input_tokens_seen": 23756376,
      "step": 31150
    },
    {
      "epoch": 64.77130977130977,
      "grad_norm": 0.00015524720947723836,
      "learning_rate": 0.03476920294393337,
      "loss": 0.2533,
      "num_input_tokens_seen": 23760152,
      "step": 31155
    },
    {
      "epoch": 64.78170478170478,
      "grad_norm": 0.0006728442385792732,
      "learning_rate": 0.03473150072750755,
      "loss": 0.2553,
      "num_input_tokens_seen": 23764120,
      "step": 31160
    },
    {
      "epoch": 64.79209979209979,
      "grad_norm": 0.00021102912432979792,
      "learning_rate": 0.03469381628693284,
      "loss": 0.2672,
      "num_input_tokens_seen": 23767928,
      "step": 31165
    },
    {
      "epoch": 64.8024948024948,
      "grad_norm": 8.775474998401478e-05,
      "learning_rate": 0.03465614962802072,
      "loss": 0.2505,
      "num_input_tokens_seen": 23771576,
      "step": 31170
    },
    {
      "epoch": 64.81288981288981,
      "grad_norm": 0.00034374051028862596,
      "learning_rate": 0.0346185007565798,
      "loss": 0.2571,
      "num_input_tokens_seen": 23775256,
      "step": 31175
    },
    {
      "epoch": 64.82328482328482,
      "grad_norm": 0.00028751016361638904,
      "learning_rate": 0.03458086967841609,
      "loss": 0.2751,
      "num_input_tokens_seen": 23779192,
      "step": 31180
    },
    {
      "epoch": 64.83367983367984,
      "grad_norm": 0.00020720680186059326,
      "learning_rate": 0.03454325639933266,
      "loss": 0.268,
      "num_input_tokens_seen": 23783064,
      "step": 31185
    },
    {
      "epoch": 64.84407484407484,
      "grad_norm": 6.728695734636858e-05,
      "learning_rate": 0.03450566092513007,
      "loss": 0.2617,
      "num_input_tokens_seen": 23786904,
      "step": 31190
    },
    {
      "epoch": 64.85446985446985,
      "grad_norm": 0.0005479721585288644,
      "learning_rate": 0.034468083261605914,
      "loss": 0.245,
      "num_input_tokens_seen": 23790808,
      "step": 31195
    },
    {
      "epoch": 64.86486486486487,
      "grad_norm": 0.0003646282711997628,
      "learning_rate": 0.03443052341455522,
      "loss": 0.2808,
      "num_input_tokens_seen": 23794744,
      "step": 31200
    },
    {
      "epoch": 64.86486486486487,
      "eval_loss": 0.2504383325576782,
      "eval_runtime": 13.3856,
      "eval_samples_per_second": 63.949,
      "eval_steps_per_second": 15.987,
      "num_input_tokens_seen": 23794744,
      "step": 31200
    },
    {
      "epoch": 64.87525987525987,
      "grad_norm": 6.249493890209123e-05,
      "learning_rate": 0.0343929813897701,
      "loss": 0.2945,
      "num_input_tokens_seen": 23798648,
      "step": 31205
    },
    {
      "epoch": 64.88565488565489,
      "grad_norm": 0.00011055656068492681,
      "learning_rate": 0.034355457193040125,
      "loss": 0.253,
      "num_input_tokens_seen": 23802456,
      "step": 31210
    },
    {
      "epoch": 64.8960498960499,
      "grad_norm": 0.0007412268896587193,
      "learning_rate": 0.03431795083015186,
      "loss": 0.2735,
      "num_input_tokens_seen": 23806360,
      "step": 31215
    },
    {
      "epoch": 64.9064449064449,
      "grad_norm": 0.0010586136486381292,
      "learning_rate": 0.03428046230688936,
      "loss": 0.2607,
      "num_input_tokens_seen": 23810168,
      "step": 31220
    },
    {
      "epoch": 64.91683991683992,
      "grad_norm": 0.00028996329638175666,
      "learning_rate": 0.034242991629033805,
      "loss": 0.2583,
      "num_input_tokens_seen": 23813880,
      "step": 31225
    },
    {
      "epoch": 64.92723492723492,
      "grad_norm": 0.00010236066009383649,
      "learning_rate": 0.03420553880236362,
      "loss": 0.2628,
      "num_input_tokens_seen": 23817624,
      "step": 31230
    },
    {
      "epoch": 64.93762993762994,
      "grad_norm": 0.00023848592536523938,
      "learning_rate": 0.03416810383265449,
      "loss": 0.2359,
      "num_input_tokens_seen": 23821528,
      "step": 31235
    },
    {
      "epoch": 64.94802494802495,
      "grad_norm": 0.0001134915801230818,
      "learning_rate": 0.03413068672567944,
      "loss": 0.2666,
      "num_input_tokens_seen": 23825304,
      "step": 31240
    },
    {
      "epoch": 64.95841995841995,
      "grad_norm": 5.729231270379387e-05,
      "learning_rate": 0.034093287487208565,
      "loss": 0.2592,
      "num_input_tokens_seen": 23829016,
      "step": 31245
    },
    {
      "epoch": 64.96881496881497,
      "grad_norm": 0.00024576118448749185,
      "learning_rate": 0.03405590612300937,
      "loss": 0.2611,
      "num_input_tokens_seen": 23832792,
      "step": 31250
    },
    {
      "epoch": 64.97920997920998,
      "grad_norm": 0.00017415838374290615,
      "learning_rate": 0.03401854263884646,
      "loss": 0.2606,
      "num_input_tokens_seen": 23836728,
      "step": 31255
    },
    {
      "epoch": 64.98960498960498,
      "grad_norm": 0.000707188795786351,
      "learning_rate": 0.033981197040481824,
      "loss": 0.2359,
      "num_input_tokens_seen": 23840664,
      "step": 31260
    },
    {
      "epoch": 65.0,
      "grad_norm": 0.00012469751527532935,
      "learning_rate": 0.03394386933367459,
      "loss": 0.2477,
      "num_input_tokens_seen": 23844456,
      "step": 31265
    },
    {
      "epoch": 65.01039501039502,
      "grad_norm": 0.0001524774415884167,
      "learning_rate": 0.033906559524181104,
      "loss": 0.244,
      "num_input_tokens_seen": 23848424,
      "step": 31270
    },
    {
      "epoch": 65.02079002079002,
      "grad_norm": 0.00013190801837481558,
      "learning_rate": 0.033869267617755085,
      "loss": 0.2729,
      "num_input_tokens_seen": 23852328,
      "step": 31275
    },
    {
      "epoch": 65.03118503118503,
      "grad_norm": 0.00011232318502152339,
      "learning_rate": 0.0338319936201474,
      "loss": 0.3129,
      "num_input_tokens_seen": 23856136,
      "step": 31280
    },
    {
      "epoch": 65.04158004158005,
      "grad_norm": 0.00018011359497904778,
      "learning_rate": 0.033794737537106136,
      "loss": 0.2651,
      "num_input_tokens_seen": 23860008,
      "step": 31285
    },
    {
      "epoch": 65.05197505197505,
      "grad_norm": 3.540679972502403e-05,
      "learning_rate": 0.03375749937437671,
      "loss": 0.2725,
      "num_input_tokens_seen": 23863816,
      "step": 31290
    },
    {
      "epoch": 65.06237006237006,
      "grad_norm": 6.644897803198546e-05,
      "learning_rate": 0.033720279137701634,
      "loss": 0.2741,
      "num_input_tokens_seen": 23867592,
      "step": 31295
    },
    {
      "epoch": 65.07276507276508,
      "grad_norm": 0.0003192275180481374,
      "learning_rate": 0.03368307683282078,
      "loss": 0.2493,
      "num_input_tokens_seen": 23871336,
      "step": 31300
    },
    {
      "epoch": 65.08316008316008,
      "grad_norm": 0.00016762847371865064,
      "learning_rate": 0.033645892465471235,
      "loss": 0.2565,
      "num_input_tokens_seen": 23875144,
      "step": 31305
    },
    {
      "epoch": 65.0935550935551,
      "grad_norm": 0.00027856751694343984,
      "learning_rate": 0.03360872604138724,
      "loss": 0.2485,
      "num_input_tokens_seen": 23878920,
      "step": 31310
    },
    {
      "epoch": 65.1039501039501,
      "grad_norm": 0.00011977836402365938,
      "learning_rate": 0.03357157756630034,
      "loss": 0.286,
      "num_input_tokens_seen": 23882728,
      "step": 31315
    },
    {
      "epoch": 65.11434511434511,
      "grad_norm": 0.0008518590475432575,
      "learning_rate": 0.033534447045939365,
      "loss": 0.2759,
      "num_input_tokens_seen": 23886632,
      "step": 31320
    },
    {
      "epoch": 65.12474012474013,
      "grad_norm": 0.00042627108632586896,
      "learning_rate": 0.03349733448603026,
      "loss": 0.2472,
      "num_input_tokens_seen": 23890504,
      "step": 31325
    },
    {
      "epoch": 65.13513513513513,
      "grad_norm": 0.00016863194468896836,
      "learning_rate": 0.03346023989229619,
      "loss": 0.266,
      "num_input_tokens_seen": 23894312,
      "step": 31330
    },
    {
      "epoch": 65.14553014553015,
      "grad_norm": 0.0003911313833668828,
      "learning_rate": 0.03342316327045769,
      "loss": 0.2524,
      "num_input_tokens_seen": 23898056,
      "step": 31335
    },
    {
      "epoch": 65.15592515592516,
      "grad_norm": 0.00034798047272488475,
      "learning_rate": 0.033386104626232385,
      "loss": 0.2778,
      "num_input_tokens_seen": 23901864,
      "step": 31340
    },
    {
      "epoch": 65.16632016632016,
      "grad_norm": 0.00010456281597726047,
      "learning_rate": 0.03334906396533525,
      "loss": 0.2606,
      "num_input_tokens_seen": 23905704,
      "step": 31345
    },
    {
      "epoch": 65.17671517671518,
      "grad_norm": 0.0003668017452582717,
      "learning_rate": 0.033312041293478326,
      "loss": 0.2268,
      "num_input_tokens_seen": 23909576,
      "step": 31350
    },
    {
      "epoch": 65.18711018711019,
      "grad_norm": 0.00019780281581915915,
      "learning_rate": 0.03327503661637103,
      "loss": 0.2475,
      "num_input_tokens_seen": 23913288,
      "step": 31355
    },
    {
      "epoch": 65.1975051975052,
      "grad_norm": 0.0003783702850341797,
      "learning_rate": 0.03323804993971998,
      "loss": 0.2727,
      "num_input_tokens_seen": 23917128,
      "step": 31360
    },
    {
      "epoch": 65.20790020790021,
      "grad_norm": 0.0013048759428784251,
      "learning_rate": 0.033201081269228924,
      "loss": 0.2404,
      "num_input_tokens_seen": 23920968,
      "step": 31365
    },
    {
      "epoch": 65.21829521829522,
      "grad_norm": 0.00026464962866157293,
      "learning_rate": 0.03316413061059895,
      "loss": 0.2331,
      "num_input_tokens_seen": 23924680,
      "step": 31370
    },
    {
      "epoch": 65.22869022869023,
      "grad_norm": 0.00011672815890051425,
      "learning_rate": 0.03312719796952827,
      "loss": 0.2622,
      "num_input_tokens_seen": 23928584,
      "step": 31375
    },
    {
      "epoch": 65.23908523908524,
      "grad_norm": 0.000176707559148781,
      "learning_rate": 0.03309028335171236,
      "loss": 0.2595,
      "num_input_tokens_seen": 23932296,
      "step": 31380
    },
    {
      "epoch": 65.24948024948024,
      "grad_norm": 0.0005229840753600001,
      "learning_rate": 0.03305338676284398,
      "loss": 0.2625,
      "num_input_tokens_seen": 23936040,
      "step": 31385
    },
    {
      "epoch": 65.25987525987526,
      "grad_norm": 0.000316180958179757,
      "learning_rate": 0.03301650820861296,
      "loss": 0.2847,
      "num_input_tokens_seen": 23939784,
      "step": 31390
    },
    {
      "epoch": 65.27027027027027,
      "grad_norm": 0.00020591021166183054,
      "learning_rate": 0.03297964769470652,
      "loss": 0.2757,
      "num_input_tokens_seen": 23943720,
      "step": 31395
    },
    {
      "epoch": 65.28066528066527,
      "grad_norm": 0.00044366164365783334,
      "learning_rate": 0.032942805226808945,
      "loss": 0.2555,
      "num_input_tokens_seen": 23947688,
      "step": 31400
    },
    {
      "epoch": 65.28066528066527,
      "eval_loss": 0.2505023181438446,
      "eval_runtime": 13.4141,
      "eval_samples_per_second": 63.814,
      "eval_steps_per_second": 15.953,
      "num_input_tokens_seen": 23947688,
      "step": 31400
    },
    {
      "epoch": 65.29106029106029,
      "grad_norm": 0.00042196191498078406,
      "learning_rate": 0.03290598081060187,
      "loss": 0.259,
      "num_input_tokens_seen": 23951464,
      "step": 31405
    },
    {
      "epoch": 65.3014553014553,
      "grad_norm": 0.00043268216541036963,
      "learning_rate": 0.03286917445176407,
      "loss": 0.2726,
      "num_input_tokens_seen": 23955144,
      "step": 31410
    },
    {
      "epoch": 65.3118503118503,
      "grad_norm": 0.00017379199562128633,
      "learning_rate": 0.032832386155971456,
      "loss": 0.2837,
      "num_input_tokens_seen": 23958888,
      "step": 31415
    },
    {
      "epoch": 65.32224532224532,
      "grad_norm": 0.00010136065247934312,
      "learning_rate": 0.032795615928897334,
      "loss": 0.2562,
      "num_input_tokens_seen": 23962536,
      "step": 31420
    },
    {
      "epoch": 65.33264033264034,
      "grad_norm": 4.883684596279636e-05,
      "learning_rate": 0.03275886377621215,
      "loss": 0.2529,
      "num_input_tokens_seen": 23966184,
      "step": 31425
    },
    {
      "epoch": 65.34303534303534,
      "grad_norm": 0.00029319716850295663,
      "learning_rate": 0.03272212970358348,
      "loss": 0.2743,
      "num_input_tokens_seen": 23969960,
      "step": 31430
    },
    {
      "epoch": 65.35343035343035,
      "grad_norm": 7.706792530370876e-05,
      "learning_rate": 0.032685413716676215,
      "loss": 0.2501,
      "num_input_tokens_seen": 23973832,
      "step": 31435
    },
    {
      "epoch": 65.36382536382537,
      "grad_norm": 7.348047074628994e-05,
      "learning_rate": 0.032648715821152474,
      "loss": 0.2925,
      "num_input_tokens_seen": 23977608,
      "step": 31440
    },
    {
      "epoch": 65.37422037422037,
      "grad_norm": 0.00039305517566390336,
      "learning_rate": 0.03261203602267143,
      "loss": 0.2632,
      "num_input_tokens_seen": 23981480,
      "step": 31445
    },
    {
      "epoch": 65.38461538461539,
      "grad_norm": 0.0002983359736390412,
      "learning_rate": 0.03257537432688966,
      "loss": 0.276,
      "num_input_tokens_seen": 23985288,
      "step": 31450
    },
    {
      "epoch": 65.39501039501039,
      "grad_norm": 0.00035976088838651776,
      "learning_rate": 0.03253873073946077,
      "loss": 0.2736,
      "num_input_tokens_seen": 23989064,
      "step": 31455
    },
    {
      "epoch": 65.4054054054054,
      "grad_norm": 0.0003426525217946619,
      "learning_rate": 0.03250210526603572,
      "loss": 0.2422,
      "num_input_tokens_seen": 23992808,
      "step": 31460
    },
    {
      "epoch": 65.41580041580042,
      "grad_norm": 0.0002808624121826142,
      "learning_rate": 0.03246549791226266,
      "loss": 0.2662,
      "num_input_tokens_seen": 23996552,
      "step": 31465
    },
    {
      "epoch": 65.42619542619542,
      "grad_norm": 0.0008905503782443702,
      "learning_rate": 0.03242890868378679,
      "loss": 0.2898,
      "num_input_tokens_seen": 24000264,
      "step": 31470
    },
    {
      "epoch": 65.43659043659044,
      "grad_norm": 0.0008628521463833749,
      "learning_rate": 0.03239233758625074,
      "loss": 0.2732,
      "num_input_tokens_seen": 24003976,
      "step": 31475
    },
    {
      "epoch": 65.44698544698545,
      "grad_norm": 0.0001316967827733606,
      "learning_rate": 0.032355784625294204,
      "loss": 0.2615,
      "num_input_tokens_seen": 24007752,
      "step": 31480
    },
    {
      "epoch": 65.45738045738045,
      "grad_norm": 6.294428021647036e-05,
      "learning_rate": 0.03231924980655402,
      "loss": 0.2633,
      "num_input_tokens_seen": 24011624,
      "step": 31485
    },
    {
      "epoch": 65.46777546777547,
      "grad_norm": 0.0001941964728757739,
      "learning_rate": 0.032282733135664446,
      "loss": 0.2525,
      "num_input_tokens_seen": 24015592,
      "step": 31490
    },
    {
      "epoch": 65.47817047817048,
      "grad_norm": 7.37763402867131e-05,
      "learning_rate": 0.03224623461825669,
      "loss": 0.267,
      "num_input_tokens_seen": 24019368,
      "step": 31495
    },
    {
      "epoch": 65.48856548856548,
      "grad_norm": 0.0001615682995179668,
      "learning_rate": 0.03220975425995937,
      "loss": 0.2567,
      "num_input_tokens_seen": 24023432,
      "step": 31500
    },
    {
      "epoch": 65.4989604989605,
      "grad_norm": 0.00030263225198723376,
      "learning_rate": 0.032173292066398206,
      "loss": 0.2823,
      "num_input_tokens_seen": 24027240,
      "step": 31505
    },
    {
      "epoch": 65.50935550935552,
      "grad_norm": 0.00017948824097402394,
      "learning_rate": 0.03213684804319606,
      "loss": 0.2415,
      "num_input_tokens_seen": 24031112,
      "step": 31510
    },
    {
      "epoch": 65.51975051975052,
      "grad_norm": 0.0005189067451283336,
      "learning_rate": 0.03210042219597312,
      "loss": 0.2519,
      "num_input_tokens_seen": 24034792,
      "step": 31515
    },
    {
      "epoch": 65.53014553014553,
      "grad_norm": 0.0011047226144000888,
      "learning_rate": 0.03206401453034675,
      "loss": 0.2607,
      "num_input_tokens_seen": 24038696,
      "step": 31520
    },
    {
      "epoch": 65.54054054054055,
      "grad_norm": 0.0005004229606129229,
      "learning_rate": 0.03202762505193136,
      "loss": 0.2587,
      "num_input_tokens_seen": 24042504,
      "step": 31525
    },
    {
      "epoch": 65.55093555093555,
      "grad_norm": 0.00023637371486984193,
      "learning_rate": 0.031991253766338754,
      "loss": 0.275,
      "num_input_tokens_seen": 24046344,
      "step": 31530
    },
    {
      "epoch": 65.56133056133056,
      "grad_norm": 0.000183104959432967,
      "learning_rate": 0.03195490067917778,
      "loss": 0.2674,
      "num_input_tokens_seen": 24050152,
      "step": 31535
    },
    {
      "epoch": 65.57172557172557,
      "grad_norm": 8.366480324184522e-05,
      "learning_rate": 0.03191856579605461,
      "loss": 0.2332,
      "num_input_tokens_seen": 24053832,
      "step": 31540
    },
    {
      "epoch": 65.58212058212058,
      "grad_norm": 9.493053221376613e-05,
      "learning_rate": 0.031882249122572454,
      "loss": 0.269,
      "num_input_tokens_seen": 24057640,
      "step": 31545
    },
    {
      "epoch": 65.5925155925156,
      "grad_norm": 0.00033328335848636925,
      "learning_rate": 0.03184595066433188,
      "loss": 0.2791,
      "num_input_tokens_seen": 24061576,
      "step": 31550
    },
    {
      "epoch": 65.6029106029106,
      "grad_norm": 0.0005416363710537553,
      "learning_rate": 0.03180967042693049,
      "loss": 0.2388,
      "num_input_tokens_seen": 24065384,
      "step": 31555
    },
    {
      "epoch": 65.61330561330561,
      "grad_norm": 0.0009904925245791674,
      "learning_rate": 0.03177340841596323,
      "loss": 0.2642,
      "num_input_tokens_seen": 24069256,
      "step": 31560
    },
    {
      "epoch": 65.62370062370063,
      "grad_norm": 8.842509123496711e-05,
      "learning_rate": 0.03173716463702209,
      "loss": 0.2589,
      "num_input_tokens_seen": 24073000,
      "step": 31565
    },
    {
      "epoch": 65.63409563409563,
      "grad_norm": 0.00023978106037247926,
      "learning_rate": 0.03170093909569638,
      "loss": 0.2735,
      "num_input_tokens_seen": 24076776,
      "step": 31570
    },
    {
      "epoch": 65.64449064449065,
      "grad_norm": 8.37829356896691e-05,
      "learning_rate": 0.03166473179757246,
      "loss": 0.2379,
      "num_input_tokens_seen": 24080520,
      "step": 31575
    },
    {
      "epoch": 65.65488565488566,
      "grad_norm": 6.53962924843654e-05,
      "learning_rate": 0.031628542748234005,
      "loss": 0.2812,
      "num_input_tokens_seen": 24084200,
      "step": 31580
    },
    {
      "epoch": 65.66528066528066,
      "grad_norm": 0.00028854614356532693,
      "learning_rate": 0.03159237195326184,
      "loss": 0.2864,
      "num_input_tokens_seen": 24088168,
      "step": 31585
    },
    {
      "epoch": 65.67567567567568,
      "grad_norm": 0.00015648029511794448,
      "learning_rate": 0.031556219418233875,
      "loss": 0.2692,
      "num_input_tokens_seen": 24092072,
      "step": 31590
    },
    {
      "epoch": 65.68607068607069,
      "grad_norm": 0.00015175908629316837,
      "learning_rate": 0.03152008514872533,
      "loss": 0.2589,
      "num_input_tokens_seen": 24095912,
      "step": 31595
    },
    {
      "epoch": 65.6964656964657,
      "grad_norm": 0.00013459622277878225,
      "learning_rate": 0.03148396915030862,
      "loss": 0.2635,
      "num_input_tokens_seen": 24099432,
      "step": 31600
    },
    {
      "epoch": 65.6964656964657,
      "eval_loss": 0.2520678639411926,
      "eval_runtime": 13.4145,
      "eval_samples_per_second": 63.811,
      "eval_steps_per_second": 15.953,
      "num_input_tokens_seen": 24099432,
      "step": 31600
    },
    {
      "epoch": 65.70686070686071,
      "grad_norm": 0.00014599690621253103,
      "learning_rate": 0.03144787142855318,
      "loss": 0.2675,
      "num_input_tokens_seen": 24103304,
      "step": 31605
    },
    {
      "epoch": 65.71725571725571,
      "grad_norm": 5.4997188271954656e-05,
      "learning_rate": 0.031411791989025835,
      "loss": 0.2709,
      "num_input_tokens_seen": 24107144,
      "step": 31610
    },
    {
      "epoch": 65.72765072765073,
      "grad_norm": 0.0001864663790911436,
      "learning_rate": 0.031375730837290394,
      "loss": 0.2567,
      "num_input_tokens_seen": 24110984,
      "step": 31615
    },
    {
      "epoch": 65.73804573804574,
      "grad_norm": 0.0001474396703997627,
      "learning_rate": 0.031339687978908015,
      "loss": 0.2641,
      "num_input_tokens_seen": 24114760,
      "step": 31620
    },
    {
      "epoch": 65.74844074844074,
      "grad_norm": 0.0005662969779223204,
      "learning_rate": 0.03130366341943694,
      "loss": 0.2626,
      "num_input_tokens_seen": 24118568,
      "step": 31625
    },
    {
      "epoch": 65.75883575883576,
      "grad_norm": 0.0003027894999831915,
      "learning_rate": 0.031267657164432555,
      "loss": 0.2869,
      "num_input_tokens_seen": 24122376,
      "step": 31630
    },
    {
      "epoch": 65.76923076923077,
      "grad_norm": 0.0007666753372177482,
      "learning_rate": 0.03123166921944752,
      "loss": 0.2793,
      "num_input_tokens_seen": 24126376,
      "step": 31635
    },
    {
      "epoch": 65.77962577962577,
      "grad_norm": 0.0009008229244500399,
      "learning_rate": 0.031195699590031666,
      "loss": 0.2781,
      "num_input_tokens_seen": 24130248,
      "step": 31640
    },
    {
      "epoch": 65.79002079002079,
      "grad_norm": 0.00015835900558158755,
      "learning_rate": 0.031159748281731885,
      "loss": 0.2746,
      "num_input_tokens_seen": 24134056,
      "step": 31645
    },
    {
      "epoch": 65.8004158004158,
      "grad_norm": 0.0001646050368435681,
      "learning_rate": 0.031123815300092394,
      "loss": 0.2703,
      "num_input_tokens_seen": 24137960,
      "step": 31650
    },
    {
      "epoch": 65.8108108108108,
      "grad_norm": 0.0004070030990988016,
      "learning_rate": 0.031087900650654424,
      "loss": 0.2657,
      "num_input_tokens_seen": 24141736,
      "step": 31655
    },
    {
      "epoch": 65.82120582120582,
      "grad_norm": 0.0005975066451355815,
      "learning_rate": 0.031052004338956534,
      "loss": 0.2381,
      "num_input_tokens_seen": 24145416,
      "step": 31660
    },
    {
      "epoch": 65.83160083160084,
      "grad_norm": 0.00011681431351462379,
      "learning_rate": 0.031016126370534407,
      "loss": 0.2582,
      "num_input_tokens_seen": 24149096,
      "step": 31665
    },
    {
      "epoch": 65.84199584199584,
      "grad_norm": 0.0004791850515175611,
      "learning_rate": 0.030980266750920804,
      "loss": 0.2788,
      "num_input_tokens_seen": 24152904,
      "step": 31670
    },
    {
      "epoch": 65.85239085239085,
      "grad_norm": 0.0003924318589270115,
      "learning_rate": 0.030944425485645747,
      "loss": 0.2599,
      "num_input_tokens_seen": 24156680,
      "step": 31675
    },
    {
      "epoch": 65.86278586278586,
      "grad_norm": 0.0002755174064077437,
      "learning_rate": 0.03090860258023647,
      "loss": 0.2404,
      "num_input_tokens_seen": 24160328,
      "step": 31680
    },
    {
      "epoch": 65.87318087318087,
      "grad_norm": 0.0006860743742436171,
      "learning_rate": 0.030872798040217236,
      "loss": 0.2804,
      "num_input_tokens_seen": 24164104,
      "step": 31685
    },
    {
      "epoch": 65.88357588357589,
      "grad_norm": 0.00036432233173400164,
      "learning_rate": 0.03083701187110964,
      "loss": 0.2637,
      "num_input_tokens_seen": 24168008,
      "step": 31690
    },
    {
      "epoch": 65.89397089397089,
      "grad_norm": 0.0005987982731312513,
      "learning_rate": 0.030801244078432294,
      "loss": 0.2408,
      "num_input_tokens_seen": 24171848,
      "step": 31695
    },
    {
      "epoch": 65.9043659043659,
      "grad_norm": 0.0009505784255452454,
      "learning_rate": 0.030765494667701024,
      "loss": 0.2704,
      "num_input_tokens_seen": 24175560,
      "step": 31700
    },
    {
      "epoch": 65.91476091476092,
      "grad_norm": 0.0006386747700162232,
      "learning_rate": 0.030729763644428913,
      "loss": 0.2357,
      "num_input_tokens_seen": 24179336,
      "step": 31705
    },
    {
      "epoch": 65.92515592515592,
      "grad_norm": 0.00023983839491847903,
      "learning_rate": 0.030694051014126048,
      "loss": 0.2712,
      "num_input_tokens_seen": 24183144,
      "step": 31710
    },
    {
      "epoch": 65.93555093555094,
      "grad_norm": 0.00028881296748295426,
      "learning_rate": 0.030658356782299792,
      "loss": 0.2641,
      "num_input_tokens_seen": 24186920,
      "step": 31715
    },
    {
      "epoch": 65.94594594594595,
      "grad_norm": 0.00025892938720062375,
      "learning_rate": 0.030622680954454726,
      "loss": 0.2747,
      "num_input_tokens_seen": 24190792,
      "step": 31720
    },
    {
      "epoch": 65.95634095634095,
      "grad_norm": 0.001398879219777882,
      "learning_rate": 0.030587023536092398,
      "loss": 0.272,
      "num_input_tokens_seen": 24194472,
      "step": 31725
    },
    {
      "epoch": 65.96673596673597,
      "grad_norm": 8.295443694805726e-05,
      "learning_rate": 0.03055138453271171,
      "loss": 0.2716,
      "num_input_tokens_seen": 24198376,
      "step": 31730
    },
    {
      "epoch": 65.97713097713098,
      "grad_norm": 0.0006298432708717883,
      "learning_rate": 0.03051576394980858,
      "loss": 0.2552,
      "num_input_tokens_seen": 24202088,
      "step": 31735
    },
    {
      "epoch": 65.98752598752598,
      "grad_norm": 0.0005156004917807877,
      "learning_rate": 0.030480161792876187,
      "loss": 0.2597,
      "num_input_tokens_seen": 24205928,
      "step": 31740
    },
    {
      "epoch": 65.997920997921,
      "grad_norm": 0.00019751819490920752,
      "learning_rate": 0.030444578067404846,
      "loss": 0.2539,
      "num_input_tokens_seen": 24209832,
      "step": 31745
    },
    {
      "epoch": 66.00831600831602,
      "grad_norm": 0.0001847275416366756,
      "learning_rate": 0.030409012778881975,
      "loss": 0.2679,
      "num_input_tokens_seen": 24213408,
      "step": 31750
    },
    {
      "epoch": 66.01871101871102,
      "grad_norm": 0.00043415476102381945,
      "learning_rate": 0.030373465932792235,
      "loss": 0.2394,
      "num_input_tokens_seen": 24217344,
      "step": 31755
    },
    {
      "epoch": 66.02910602910603,
      "grad_norm": 8.959716069512069e-05,
      "learning_rate": 0.030337937534617342,
      "loss": 0.2742,
      "num_input_tokens_seen": 24221184,
      "step": 31760
    },
    {
      "epoch": 66.03950103950103,
      "grad_norm": 0.0001079543653759174,
      "learning_rate": 0.030302427589836277,
      "loss": 0.2621,
      "num_input_tokens_seen": 24224960,
      "step": 31765
    },
    {
      "epoch": 66.04989604989605,
      "grad_norm": 0.0005546874017454684,
      "learning_rate": 0.030266936103925095,
      "loss": 0.2638,
      "num_input_tokens_seen": 24228704,
      "step": 31770
    },
    {
      "epoch": 66.06029106029106,
      "grad_norm": 0.0001257527037523687,
      "learning_rate": 0.030231463082356982,
      "loss": 0.2845,
      "num_input_tokens_seen": 24232352,
      "step": 31775
    },
    {
      "epoch": 66.07068607068607,
      "grad_norm": 0.0005177812417969108,
      "learning_rate": 0.030196008530602367,
      "loss": 0.2613,
      "num_input_tokens_seen": 24236160,
      "step": 31780
    },
    {
      "epoch": 66.08108108108108,
      "grad_norm": 0.00020069188030902296,
      "learning_rate": 0.030160572454128842,
      "loss": 0.251,
      "num_input_tokens_seen": 24239840,
      "step": 31785
    },
    {
      "epoch": 66.0914760914761,
      "grad_norm": 9.035712719196454e-05,
      "learning_rate": 0.03012515485840098,
      "loss": 0.2419,
      "num_input_tokens_seen": 24243648,
      "step": 31790
    },
    {
      "epoch": 66.1018711018711,
      "grad_norm": 0.0005617070128209889,
      "learning_rate": 0.030089755748880734,
      "loss": 0.2447,
      "num_input_tokens_seen": 24247360,
      "step": 31795
    },
    {
      "epoch": 66.11226611226611,
      "grad_norm": 0.0003881428565364331,
      "learning_rate": 0.030054375131027003,
      "loss": 0.2697,
      "num_input_tokens_seen": 24251200,
      "step": 31800
    },
    {
      "epoch": 66.11226611226611,
      "eval_loss": 0.24743887782096863,
      "eval_runtime": 13.3938,
      "eval_samples_per_second": 63.91,
      "eval_steps_per_second": 15.977,
      "num_input_tokens_seen": 24251200,
      "step": 31800
    },
    {
      "epoch": 66.12266112266113,
      "grad_norm": 0.0002553602389525622,
      "learning_rate": 0.030019013010295942,
      "loss": 0.2487,
      "num_input_tokens_seen": 24254848,
      "step": 31805
    },
    {
      "epoch": 66.13305613305613,
      "grad_norm": 0.00024426504387520254,
      "learning_rate": 0.029983669392140897,
      "loss": 0.2614,
      "num_input_tokens_seen": 24258752,
      "step": 31810
    },
    {
      "epoch": 66.14345114345114,
      "grad_norm": 0.0008571365615352988,
      "learning_rate": 0.029948344282012217,
      "loss": 0.2402,
      "num_input_tokens_seen": 24262560,
      "step": 31815
    },
    {
      "epoch": 66.15384615384616,
      "grad_norm": 0.0002636099234223366,
      "learning_rate": 0.029913037685357507,
      "loss": 0.2807,
      "num_input_tokens_seen": 24266464,
      "step": 31820
    },
    {
      "epoch": 66.16424116424116,
      "grad_norm": 0.00023669969232287258,
      "learning_rate": 0.029877749607621528,
      "loss": 0.2622,
      "num_input_tokens_seen": 24270304,
      "step": 31825
    },
    {
      "epoch": 66.17463617463618,
      "grad_norm": 0.00017646161722950637,
      "learning_rate": 0.029842480054246077,
      "loss": 0.2308,
      "num_input_tokens_seen": 24274080,
      "step": 31830
    },
    {
      "epoch": 66.18503118503118,
      "grad_norm": 0.00021775079949293286,
      "learning_rate": 0.02980722903067022,
      "loss": 0.2647,
      "num_input_tokens_seen": 24277920,
      "step": 31835
    },
    {
      "epoch": 66.1954261954262,
      "grad_norm": 0.0003477302088867873,
      "learning_rate": 0.029771996542330113,
      "loss": 0.2751,
      "num_input_tokens_seen": 24281760,
      "step": 31840
    },
    {
      "epoch": 66.20582120582121,
      "grad_norm": 0.00024201312044169754,
      "learning_rate": 0.029736782594658954,
      "loss": 0.2653,
      "num_input_tokens_seen": 24285728,
      "step": 31845
    },
    {
      "epoch": 66.21621621621621,
      "grad_norm": 0.00037786460597999394,
      "learning_rate": 0.029701587193087284,
      "loss": 0.2597,
      "num_input_tokens_seen": 24289632,
      "step": 31850
    },
    {
      "epoch": 66.22661122661123,
      "grad_norm": 0.00016534920723643154,
      "learning_rate": 0.0296664103430426,
      "loss": 0.2595,
      "num_input_tokens_seen": 24293504,
      "step": 31855
    },
    {
      "epoch": 66.23700623700624,
      "grad_norm": 0.00017315348668489605,
      "learning_rate": 0.029631252049949652,
      "loss": 0.2762,
      "num_input_tokens_seen": 24297344,
      "step": 31860
    },
    {
      "epoch": 66.24740124740124,
      "grad_norm": 0.0012824817094951868,
      "learning_rate": 0.02959611231923031,
      "loss": 0.2526,
      "num_input_tokens_seen": 24301280,
      "step": 31865
    },
    {
      "epoch": 66.25779625779626,
      "grad_norm": 0.00015988053928595036,
      "learning_rate": 0.029560991156303507,
      "loss": 0.265,
      "num_input_tokens_seen": 24305408,
      "step": 31870
    },
    {
      "epoch": 66.26819126819127,
      "grad_norm": 0.00023079382663127035,
      "learning_rate": 0.02952588856658544,
      "loss": 0.2746,
      "num_input_tokens_seen": 24309280,
      "step": 31875
    },
    {
      "epoch": 66.27858627858627,
      "grad_norm": 0.0002841517562046647,
      "learning_rate": 0.029490804555489296,
      "loss": 0.2556,
      "num_input_tokens_seen": 24312864,
      "step": 31880
    },
    {
      "epoch": 66.28898128898129,
      "grad_norm": 0.0015741126844659448,
      "learning_rate": 0.029455739128425484,
      "loss": 0.2719,
      "num_input_tokens_seen": 24316576,
      "step": 31885
    },
    {
      "epoch": 66.2993762993763,
      "grad_norm": 0.00044265022734180093,
      "learning_rate": 0.029420692290801607,
      "loss": 0.2593,
      "num_input_tokens_seen": 24320384,
      "step": 31890
    },
    {
      "epoch": 66.3097713097713,
      "grad_norm": 0.00036997589631937444,
      "learning_rate": 0.02938566404802223,
      "loss": 0.2677,
      "num_input_tokens_seen": 24324256,
      "step": 31895
    },
    {
      "epoch": 66.32016632016632,
      "grad_norm": 0.0003390981873963028,
      "learning_rate": 0.029350654405489195,
      "loss": 0.2797,
      "num_input_tokens_seen": 24328000,
      "step": 31900
    },
    {
      "epoch": 66.33056133056132,
      "grad_norm": 0.0001893117296276614,
      "learning_rate": 0.02931566336860145,
      "loss": 0.266,
      "num_input_tokens_seen": 24331936,
      "step": 31905
    },
    {
      "epoch": 66.34095634095634,
      "grad_norm": 0.000436124304542318,
      "learning_rate": 0.02928069094275505,
      "loss": 0.2683,
      "num_input_tokens_seen": 24335680,
      "step": 31910
    },
    {
      "epoch": 66.35135135135135,
      "grad_norm": 0.0002594403631519526,
      "learning_rate": 0.02924573713334314,
      "loss": 0.2758,
      "num_input_tokens_seen": 24339456,
      "step": 31915
    },
    {
      "epoch": 66.36174636174636,
      "grad_norm": 0.0004278372216504067,
      "learning_rate": 0.02921080194575603,
      "loss": 0.227,
      "num_input_tokens_seen": 24343264,
      "step": 31920
    },
    {
      "epoch": 66.37214137214137,
      "grad_norm": 0.0004695228999480605,
      "learning_rate": 0.029175885385381177,
      "loss": 0.2849,
      "num_input_tokens_seen": 24347040,
      "step": 31925
    },
    {
      "epoch": 66.38253638253639,
      "grad_norm": 0.00029203714802861214,
      "learning_rate": 0.029140987457603223,
      "loss": 0.2881,
      "num_input_tokens_seen": 24350816,
      "step": 31930
    },
    {
      "epoch": 66.39293139293139,
      "grad_norm": 0.0006393062067218125,
      "learning_rate": 0.029106108167803763,
      "loss": 0.2925,
      "num_input_tokens_seen": 24354720,
      "step": 31935
    },
    {
      "epoch": 66.4033264033264,
      "grad_norm": 0.000825194176286459,
      "learning_rate": 0.029071247521361674,
      "loss": 0.2599,
      "num_input_tokens_seen": 24358528,
      "step": 31940
    },
    {
      "epoch": 66.41372141372142,
      "grad_norm": 4.994422852178104e-05,
      "learning_rate": 0.029036405523652945,
      "loss": 0.2655,
      "num_input_tokens_seen": 24362336,
      "step": 31945
    },
    {
      "epoch": 66.42411642411642,
      "grad_norm": 0.0003835965762846172,
      "learning_rate": 0.029001582180050577,
      "loss": 0.2558,
      "num_input_tokens_seen": 24366208,
      "step": 31950
    },
    {
      "epoch": 66.43451143451144,
      "grad_norm": 0.00031185458647087216,
      "learning_rate": 0.02896677749592482,
      "loss": 0.2455,
      "num_input_tokens_seen": 24370048,
      "step": 31955
    },
    {
      "epoch": 66.44490644490645,
      "grad_norm": 0.0009959047893062234,
      "learning_rate": 0.028931991476642938,
      "loss": 0.2773,
      "num_input_tokens_seen": 24373952,
      "step": 31960
    },
    {
      "epoch": 66.45530145530145,
      "grad_norm": 0.00012873091327492148,
      "learning_rate": 0.028897224127569412,
      "loss": 0.2712,
      "num_input_tokens_seen": 24377728,
      "step": 31965
    },
    {
      "epoch": 66.46569646569647,
      "grad_norm": 0.0004577824438456446,
      "learning_rate": 0.028862475454065832,
      "loss": 0.2457,
      "num_input_tokens_seen": 24381696,
      "step": 31970
    },
    {
      "epoch": 66.47609147609148,
      "grad_norm": 0.0003507114597596228,
      "learning_rate": 0.028827745461490806,
      "loss": 0.2685,
      "num_input_tokens_seen": 24385664,
      "step": 31975
    },
    {
      "epoch": 66.48648648648648,
      "grad_norm": 0.0003083197516389191,
      "learning_rate": 0.028793034155200212,
      "loss": 0.295,
      "num_input_tokens_seen": 24389536,
      "step": 31980
    },
    {
      "epoch": 66.4968814968815,
      "grad_norm": 0.00040116204763762653,
      "learning_rate": 0.028758341540546944,
      "loss": 0.2801,
      "num_input_tokens_seen": 24393344,
      "step": 31985
    },
    {
      "epoch": 66.5072765072765,
      "grad_norm": 0.0008647387730889022,
      "learning_rate": 0.02872366762288098,
      "loss": 0.2591,
      "num_input_tokens_seen": 24397280,
      "step": 31990
    },
    {
      "epoch": 66.51767151767152,
      "grad_norm": 0.00014302368799690157,
      "learning_rate": 0.028689012407549567,
      "loss": 0.2456,
      "num_input_tokens_seen": 24401024,
      "step": 31995
    },
    {
      "epoch": 66.52806652806653,
      "grad_norm": 0.00019364868057891726,
      "learning_rate": 0.028654375899896892,
      "loss": 0.2765,
      "num_input_tokens_seen": 24404736,
      "step": 32000
    },
    {
      "epoch": 66.52806652806653,
      "eval_loss": 0.24981892108917236,
      "eval_runtime": 13.3965,
      "eval_samples_per_second": 63.897,
      "eval_steps_per_second": 15.974,
      "num_input_tokens_seen": 24404736,
      "step": 32000
    },
    {
      "epoch": 66.53846153846153,
      "grad_norm": 0.00015548663213849068,
      "learning_rate": 0.02861975810526437,
      "loss": 0.2516,
      "num_input_tokens_seen": 24408672,
      "step": 32005
    },
    {
      "epoch": 66.54885654885655,
      "grad_norm": 0.0002616024576127529,
      "learning_rate": 0.02858515902899056,
      "loss": 0.2473,
      "num_input_tokens_seen": 24412416,
      "step": 32010
    },
    {
      "epoch": 66.55925155925156,
      "grad_norm": 0.00019993921159766614,
      "learning_rate": 0.028550578676410976,
      "loss": 0.2418,
      "num_input_tokens_seen": 24416032,
      "step": 32015
    },
    {
      "epoch": 66.56964656964657,
      "grad_norm": 0.00020499563834164292,
      "learning_rate": 0.02851601705285837,
      "loss": 0.2864,
      "num_input_tokens_seen": 24419904,
      "step": 32020
    },
    {
      "epoch": 66.58004158004158,
      "grad_norm": 7.819920574547723e-05,
      "learning_rate": 0.028481474163662666,
      "loss": 0.2685,
      "num_input_tokens_seen": 24423776,
      "step": 32025
    },
    {
      "epoch": 66.5904365904366,
      "grad_norm": 0.00013305859465617687,
      "learning_rate": 0.028446950014150683,
      "loss": 0.2775,
      "num_input_tokens_seen": 24427424,
      "step": 32030
    },
    {
      "epoch": 66.6008316008316,
      "grad_norm": 0.0003858963609673083,
      "learning_rate": 0.028412444609646596,
      "loss": 0.2686,
      "num_input_tokens_seen": 24431200,
      "step": 32035
    },
    {
      "epoch": 66.61122661122661,
      "grad_norm": 0.0006701680831611156,
      "learning_rate": 0.028377957955471465,
      "loss": 0.2545,
      "num_input_tokens_seen": 24435136,
      "step": 32040
    },
    {
      "epoch": 66.62162162162163,
      "grad_norm": 0.00026531622279435396,
      "learning_rate": 0.0283434900569436,
      "loss": 0.2767,
      "num_input_tokens_seen": 24438880,
      "step": 32045
    },
    {
      "epoch": 66.63201663201663,
      "grad_norm": 0.0010543946409597993,
      "learning_rate": 0.028309040919378456,
      "loss": 0.2724,
      "num_input_tokens_seen": 24443072,
      "step": 32050
    },
    {
      "epoch": 66.64241164241164,
      "grad_norm": 0.00019822463218588382,
      "learning_rate": 0.02827461054808848,
      "loss": 0.2624,
      "num_input_tokens_seen": 24446816,
      "step": 32055
    },
    {
      "epoch": 66.65280665280665,
      "grad_norm": 0.0001548689033370465,
      "learning_rate": 0.028240198948383186,
      "loss": 0.2672,
      "num_input_tokens_seen": 24450784,
      "step": 32060
    },
    {
      "epoch": 66.66320166320166,
      "grad_norm": 0.00027787633007392287,
      "learning_rate": 0.028205806125569402,
      "loss": 0.2705,
      "num_input_tokens_seen": 24454528,
      "step": 32065
    },
    {
      "epoch": 66.67359667359668,
      "grad_norm": 0.0015721704112365842,
      "learning_rate": 0.028171432084950834,
      "loss": 0.2673,
      "num_input_tokens_seen": 24458336,
      "step": 32070
    },
    {
      "epoch": 66.68399168399168,
      "grad_norm": 0.00023947693989612162,
      "learning_rate": 0.028137076831828478,
      "loss": 0.2583,
      "num_input_tokens_seen": 24462144,
      "step": 32075
    },
    {
      "epoch": 66.6943866943867,
      "grad_norm": 4.022300709038973e-05,
      "learning_rate": 0.028102740371500238,
      "loss": 0.2547,
      "num_input_tokens_seen": 24465952,
      "step": 32080
    },
    {
      "epoch": 66.70478170478171,
      "grad_norm": 0.00010132717579836026,
      "learning_rate": 0.0280684227092613,
      "loss": 0.2825,
      "num_input_tokens_seen": 24469696,
      "step": 32085
    },
    {
      "epoch": 66.71517671517671,
      "grad_norm": 3.9046226447680965e-05,
      "learning_rate": 0.02803412385040392,
      "loss": 0.271,
      "num_input_tokens_seen": 24473440,
      "step": 32090
    },
    {
      "epoch": 66.72557172557173,
      "grad_norm": 0.0005616469425149262,
      "learning_rate": 0.027999843800217306,
      "loss": 0.2755,
      "num_input_tokens_seen": 24477248,
      "step": 32095
    },
    {
      "epoch": 66.73596673596674,
      "grad_norm": 0.00030461253481917083,
      "learning_rate": 0.027965582563987932,
      "loss": 0.26,
      "num_input_tokens_seen": 24481056,
      "step": 32100
    },
    {
      "epoch": 66.74636174636174,
      "grad_norm": 0.00039138534339144826,
      "learning_rate": 0.027931340146999346,
      "loss": 0.2375,
      "num_input_tokens_seen": 24484928,
      "step": 32105
    },
    {
      "epoch": 66.75675675675676,
      "grad_norm": 0.0006079064914956689,
      "learning_rate": 0.02789711655453208,
      "loss": 0.2644,
      "num_input_tokens_seen": 24488928,
      "step": 32110
    },
    {
      "epoch": 66.76715176715177,
      "grad_norm": 0.0001235850650118664,
      "learning_rate": 0.02786291179186392,
      "loss": 0.2711,
      "num_input_tokens_seen": 24492800,
      "step": 32115
    },
    {
      "epoch": 66.77754677754677,
      "grad_norm": 0.00036211672704666853,
      "learning_rate": 0.02782872586426961,
      "loss": 0.2593,
      "num_input_tokens_seen": 24496608,
      "step": 32120
    },
    {
      "epoch": 66.78794178794179,
      "grad_norm": 6.564981595147401e-05,
      "learning_rate": 0.027794558777021083,
      "loss": 0.2652,
      "num_input_tokens_seen": 24500352,
      "step": 32125
    },
    {
      "epoch": 66.7983367983368,
      "grad_norm": 0.0001920598588185385,
      "learning_rate": 0.02776041053538734,
      "loss": 0.2473,
      "num_input_tokens_seen": 24504032,
      "step": 32130
    },
    {
      "epoch": 66.8087318087318,
      "grad_norm": 0.00030004591098986566,
      "learning_rate": 0.027726281144634407,
      "loss": 0.2475,
      "num_input_tokens_seen": 24507872,
      "step": 32135
    },
    {
      "epoch": 66.81912681912682,
      "grad_norm": 0.00015475442341994494,
      "learning_rate": 0.02769217061002552,
      "loss": 0.252,
      "num_input_tokens_seen": 24511552,
      "step": 32140
    },
    {
      "epoch": 66.82952182952182,
      "grad_norm": 5.0014648877549917e-05,
      "learning_rate": 0.027658078936820967,
      "loss": 0.2843,
      "num_input_tokens_seen": 24515296,
      "step": 32145
    },
    {
      "epoch": 66.83991683991684,
      "grad_norm": 0.0004331866221036762,
      "learning_rate": 0.02762400613027805,
      "loss": 0.2649,
      "num_input_tokens_seen": 24519008,
      "step": 32150
    },
    {
      "epoch": 66.85031185031185,
      "grad_norm": 0.0003696487983688712,
      "learning_rate": 0.027589952195651295,
      "loss": 0.2858,
      "num_input_tokens_seen": 24522816,
      "step": 32155
    },
    {
      "epoch": 66.86070686070686,
      "grad_norm": 0.00017552592908032238,
      "learning_rate": 0.027555917138192186,
      "loss": 0.2725,
      "num_input_tokens_seen": 24526624,
      "step": 32160
    },
    {
      "epoch": 66.87110187110187,
      "grad_norm": 0.00044517702190205455,
      "learning_rate": 0.027521900963149375,
      "loss": 0.2354,
      "num_input_tokens_seen": 24530304,
      "step": 32165
    },
    {
      "epoch": 66.88149688149689,
      "grad_norm": 0.000365311570931226,
      "learning_rate": 0.027487903675768633,
      "loss": 0.2397,
      "num_input_tokens_seen": 24534144,
      "step": 32170
    },
    {
      "epoch": 66.89189189189189,
      "grad_norm": 0.000583117944188416,
      "learning_rate": 0.027453925281292677,
      "loss": 0.2682,
      "num_input_tokens_seen": 24537888,
      "step": 32175
    },
    {
      "epoch": 66.9022869022869,
      "grad_norm": 0.00010460512567078695,
      "learning_rate": 0.027419965784961475,
      "loss": 0.2603,
      "num_input_tokens_seen": 24541728,
      "step": 32180
    },
    {
      "epoch": 66.91268191268192,
      "grad_norm": 0.00010335264960303903,
      "learning_rate": 0.027386025192012015,
      "loss": 0.2621,
      "num_input_tokens_seen": 24545664,
      "step": 32185
    },
    {
      "epoch": 66.92307692307692,
      "grad_norm": 0.0001855616719694808,
      "learning_rate": 0.027352103507678277,
      "loss": 0.2715,
      "num_input_tokens_seen": 24549344,
      "step": 32190
    },
    {
      "epoch": 66.93347193347194,
      "grad_norm": 0.00020409049466252327,
      "learning_rate": 0.027318200737191527,
      "loss": 0.2524,
      "num_input_tokens_seen": 24553216,
      "step": 32195
    },
    {
      "epoch": 66.94386694386695,
      "grad_norm": 0.0019398892764002085,
      "learning_rate": 0.027284316885779935,
      "loss": 0.2743,
      "num_input_tokens_seen": 24557120,
      "step": 32200
    },
    {
      "epoch": 66.94386694386695,
      "eval_loss": 0.24870139360427856,
      "eval_runtime": 13.4109,
      "eval_samples_per_second": 63.829,
      "eval_steps_per_second": 15.957,
      "num_input_tokens_seen": 24557120,
      "step": 32200
    },
    {
      "epoch": 66.95426195426195,
      "grad_norm": 0.00022466557857114822,
      "learning_rate": 0.027250451958668785,
      "loss": 0.2894,
      "num_input_tokens_seen": 24560832,
      "step": 32205
    },
    {
      "epoch": 66.96465696465697,
      "grad_norm": 0.00018743482360150665,
      "learning_rate": 0.027216605961080536,
      "loss": 0.2712,
      "num_input_tokens_seen": 24564640,
      "step": 32210
    },
    {
      "epoch": 66.97505197505197,
      "grad_norm": 0.00018502064631320536,
      "learning_rate": 0.02718277889823461,
      "loss": 0.2615,
      "num_input_tokens_seen": 24568544,
      "step": 32215
    },
    {
      "epoch": 66.98544698544698,
      "grad_norm": 0.0003447768685873598,
      "learning_rate": 0.027148970775347604,
      "loss": 0.2551,
      "num_input_tokens_seen": 24572352,
      "step": 32220
    },
    {
      "epoch": 66.995841995842,
      "grad_norm": 0.0014508496969938278,
      "learning_rate": 0.027115181597633174,
      "loss": 0.2482,
      "num_input_tokens_seen": 24576192,
      "step": 32225
    },
    {
      "epoch": 67.006237006237,
      "grad_norm": 0.000303589622490108,
      "learning_rate": 0.027081411370301976,
      "loss": 0.2531,
      "num_input_tokens_seen": 24579952,
      "step": 32230
    },
    {
      "epoch": 67.01663201663202,
      "grad_norm": 0.00034578194026835263,
      "learning_rate": 0.027047660098561875,
      "loss": 0.2696,
      "num_input_tokens_seen": 24583696,
      "step": 32235
    },
    {
      "epoch": 67.02702702702703,
      "grad_norm": 0.00034689498716033995,
      "learning_rate": 0.02701392778761766,
      "loss": 0.238,
      "num_input_tokens_seen": 24587504,
      "step": 32240
    },
    {
      "epoch": 67.03742203742203,
      "grad_norm": 0.00010837531590368599,
      "learning_rate": 0.02698021444267133,
      "loss": 0.2666,
      "num_input_tokens_seen": 24591280,
      "step": 32245
    },
    {
      "epoch": 67.04781704781705,
      "grad_norm": 6.753968773409724e-05,
      "learning_rate": 0.026946520068921915,
      "loss": 0.2935,
      "num_input_tokens_seen": 24595216,
      "step": 32250
    },
    {
      "epoch": 67.05821205821206,
      "grad_norm": 0.00015085379709489644,
      "learning_rate": 0.02691284467156547,
      "loss": 0.2825,
      "num_input_tokens_seen": 24599184,
      "step": 32255
    },
    {
      "epoch": 67.06860706860707,
      "grad_norm": 0.0005276617011986673,
      "learning_rate": 0.026879188255795182,
      "loss": 0.2639,
      "num_input_tokens_seen": 24603056,
      "step": 32260
    },
    {
      "epoch": 67.07900207900208,
      "grad_norm": 0.0005873062764294446,
      "learning_rate": 0.026845550826801328,
      "loss": 0.2494,
      "num_input_tokens_seen": 24606864,
      "step": 32265
    },
    {
      "epoch": 67.0893970893971,
      "grad_norm": 0.00013601582031697035,
      "learning_rate": 0.02681193238977121,
      "loss": 0.2586,
      "num_input_tokens_seen": 24610768,
      "step": 32270
    },
    {
      "epoch": 67.0997920997921,
      "grad_norm": 0.00027289893478155136,
      "learning_rate": 0.026778332949889145,
      "loss": 0.2576,
      "num_input_tokens_seen": 24614672,
      "step": 32275
    },
    {
      "epoch": 67.11018711018711,
      "grad_norm": 0.0005473297787830234,
      "learning_rate": 0.026744752512336673,
      "loss": 0.2923,
      "num_input_tokens_seen": 24618480,
      "step": 32280
    },
    {
      "epoch": 67.12058212058211,
      "grad_norm": 4.2316532926633954e-05,
      "learning_rate": 0.02671119108229225,
      "loss": 0.2524,
      "num_input_tokens_seen": 24622192,
      "step": 32285
    },
    {
      "epoch": 67.13097713097713,
      "grad_norm": 0.0005457705701701343,
      "learning_rate": 0.026677648664931556,
      "loss": 0.2837,
      "num_input_tokens_seen": 24626000,
      "step": 32290
    },
    {
      "epoch": 67.14137214137214,
      "grad_norm": 9.365333971800283e-05,
      "learning_rate": 0.026644125265427154,
      "loss": 0.2595,
      "num_input_tokens_seen": 24629904,
      "step": 32295
    },
    {
      "epoch": 67.15176715176715,
      "grad_norm": 0.00046885712072253227,
      "learning_rate": 0.026610620888948822,
      "loss": 0.2512,
      "num_input_tokens_seen": 24633584,
      "step": 32300
    },
    {
      "epoch": 67.16216216216216,
      "grad_norm": 0.0002616364508867264,
      "learning_rate": 0.026577135540663408,
      "loss": 0.2613,
      "num_input_tokens_seen": 24637424,
      "step": 32305
    },
    {
      "epoch": 67.17255717255718,
      "grad_norm": 0.00028624775586649776,
      "learning_rate": 0.026543669225734673,
      "loss": 0.2809,
      "num_input_tokens_seen": 24641200,
      "step": 32310
    },
    {
      "epoch": 67.18295218295218,
      "grad_norm": 2.9039969376754016e-05,
      "learning_rate": 0.02651022194932363,
      "loss": 0.2681,
      "num_input_tokens_seen": 24645008,
      "step": 32315
    },
    {
      "epoch": 67.1933471933472,
      "grad_norm": 0.0010189771419391036,
      "learning_rate": 0.026476793716588194,
      "loss": 0.2543,
      "num_input_tokens_seen": 24648880,
      "step": 32320
    },
    {
      "epoch": 67.20374220374221,
      "grad_norm": 4.967190761817619e-05,
      "learning_rate": 0.026443384532683467,
      "loss": 0.2734,
      "num_input_tokens_seen": 24652560,
      "step": 32325
    },
    {
      "epoch": 67.21413721413721,
      "grad_norm": 0.0008077262318693101,
      "learning_rate": 0.026409994402761584,
      "loss": 0.2487,
      "num_input_tokens_seen": 24656336,
      "step": 32330
    },
    {
      "epoch": 67.22453222453223,
      "grad_norm": 0.0003014679823536426,
      "learning_rate": 0.026376623331971653,
      "loss": 0.2654,
      "num_input_tokens_seen": 24660208,
      "step": 32335
    },
    {
      "epoch": 67.23492723492724,
      "grad_norm": 0.00021866391762159765,
      "learning_rate": 0.026343271325459997,
      "loss": 0.2661,
      "num_input_tokens_seen": 24664080,
      "step": 32340
    },
    {
      "epoch": 67.24532224532224,
      "grad_norm": 0.00032542619737796485,
      "learning_rate": 0.02630993838836987,
      "loss": 0.2697,
      "num_input_tokens_seen": 24667632,
      "step": 32345
    },
    {
      "epoch": 67.25571725571726,
      "grad_norm": 0.00046098572784103453,
      "learning_rate": 0.026276624525841584,
      "loss": 0.2612,
      "num_input_tokens_seen": 24671568,
      "step": 32350
    },
    {
      "epoch": 67.26611226611226,
      "grad_norm": 0.00020881762611679733,
      "learning_rate": 0.026243329743012637,
      "loss": 0.2425,
      "num_input_tokens_seen": 24675472,
      "step": 32355
    },
    {
      "epoch": 67.27650727650727,
      "grad_norm": 0.0003716912760864943,
      "learning_rate": 0.026210054045017438,
      "loss": 0.2433,
      "num_input_tokens_seen": 24679248,
      "step": 32360
    },
    {
      "epoch": 67.28690228690229,
      "grad_norm": 5.221865649218671e-05,
      "learning_rate": 0.02617679743698755,
      "loss": 0.2802,
      "num_input_tokens_seen": 24683056,
      "step": 32365
    },
    {
      "epoch": 67.29729729729729,
      "grad_norm": 0.0004487780388444662,
      "learning_rate": 0.02614355992405158,
      "loss": 0.2839,
      "num_input_tokens_seen": 24686736,
      "step": 32370
    },
    {
      "epoch": 67.3076923076923,
      "grad_norm": 0.0006595671875402331,
      "learning_rate": 0.026110341511335115,
      "loss": 0.278,
      "num_input_tokens_seen": 24690480,
      "step": 32375
    },
    {
      "epoch": 67.31808731808732,
      "grad_norm": 0.0002938820980489254,
      "learning_rate": 0.02607714220396093,
      "loss": 0.275,
      "num_input_tokens_seen": 24694256,
      "step": 32380
    },
    {
      "epoch": 67.32848232848232,
      "grad_norm": 0.000242730719037354,
      "learning_rate": 0.02604396200704869,
      "loss": 0.2515,
      "num_input_tokens_seen": 24698096,
      "step": 32385
    },
    {
      "epoch": 67.33887733887734,
      "grad_norm": 0.00043907048529945314,
      "learning_rate": 0.02601080092571523,
      "loss": 0.2596,
      "num_input_tokens_seen": 24701904,
      "step": 32390
    },
    {
      "epoch": 67.34927234927235,
      "grad_norm": 6.275655323406681e-05,
      "learning_rate": 0.025977658965074455,
      "loss": 0.2681,
      "num_input_tokens_seen": 24705712,
      "step": 32395
    },
    {
      "epoch": 67.35966735966736,
      "grad_norm": 0.00016123673412948847,
      "learning_rate": 0.02594453613023719,
      "loss": 0.2775,
      "num_input_tokens_seen": 24709616,
      "step": 32400
    },
    {
      "epoch": 67.35966735966736,
      "eval_loss": 0.2504204511642456,
      "eval_runtime": 13.3757,
      "eval_samples_per_second": 63.997,
      "eval_steps_per_second": 15.999,
      "num_input_tokens_seen": 24709616,
      "step": 32400
    },
    {
      "epoch": 67.37006237006237,
      "grad_norm": 0.0005676941364072263,
      "learning_rate": 0.025911432426311443,
      "loss": 0.2469,
      "num_input_tokens_seen": 24713488,
      "step": 32405
    },
    {
      "epoch": 67.38045738045739,
      "grad_norm": 0.00014894260675646365,
      "learning_rate": 0.025878347858402234,
      "loss": 0.2874,
      "num_input_tokens_seen": 24717328,
      "step": 32410
    },
    {
      "epoch": 67.39085239085239,
      "grad_norm": 0.00030095753027126193,
      "learning_rate": 0.025845282431611598,
      "loss": 0.2804,
      "num_input_tokens_seen": 24721168,
      "step": 32415
    },
    {
      "epoch": 67.4012474012474,
      "grad_norm": 0.00035193029907532036,
      "learning_rate": 0.025812236151038608,
      "loss": 0.2741,
      "num_input_tokens_seen": 24724880,
      "step": 32420
    },
    {
      "epoch": 67.41164241164242,
      "grad_norm": 0.00010354749974794686,
      "learning_rate": 0.025779209021779468,
      "loss": 0.2638,
      "num_input_tokens_seen": 24728560,
      "step": 32425
    },
    {
      "epoch": 67.42203742203742,
      "grad_norm": 0.0006758659146726131,
      "learning_rate": 0.025746201048927324,
      "loss": 0.2568,
      "num_input_tokens_seen": 24732240,
      "step": 32430
    },
    {
      "epoch": 67.43243243243244,
      "grad_norm": 9.829413465922698e-05,
      "learning_rate": 0.025713212237572485,
      "loss": 0.271,
      "num_input_tokens_seen": 24736112,
      "step": 32435
    },
    {
      "epoch": 67.44282744282744,
      "grad_norm": 0.0009382665739394724,
      "learning_rate": 0.025680242592802164,
      "loss": 0.2727,
      "num_input_tokens_seen": 24739984,
      "step": 32440
    },
    {
      "epoch": 67.45322245322245,
      "grad_norm": 0.0003254359762649983,
      "learning_rate": 0.02564729211970073,
      "loss": 0.2462,
      "num_input_tokens_seen": 24743824,
      "step": 32445
    },
    {
      "epoch": 67.46361746361747,
      "grad_norm": 0.0004774636763613671,
      "learning_rate": 0.025614360823349617,
      "loss": 0.2755,
      "num_input_tokens_seen": 24747856,
      "step": 32450
    },
    {
      "epoch": 67.47401247401247,
      "grad_norm": 0.0007760768639855087,
      "learning_rate": 0.025581448708827146,
      "loss": 0.254,
      "num_input_tokens_seen": 24751600,
      "step": 32455
    },
    {
      "epoch": 67.48440748440748,
      "grad_norm": 0.0002239678578916937,
      "learning_rate": 0.025548555781208876,
      "loss": 0.2394,
      "num_input_tokens_seen": 24755280,
      "step": 32460
    },
    {
      "epoch": 67.4948024948025,
      "grad_norm": 0.0011168154887855053,
      "learning_rate": 0.02551568204556721,
      "loss": 0.2905,
      "num_input_tokens_seen": 24759120,
      "step": 32465
    },
    {
      "epoch": 67.5051975051975,
      "grad_norm": 0.0004033400327898562,
      "learning_rate": 0.02548282750697173,
      "loss": 0.2322,
      "num_input_tokens_seen": 24762960,
      "step": 32470
    },
    {
      "epoch": 67.51559251559252,
      "grad_norm": 0.0005359951173886657,
      "learning_rate": 0.02544999217048909,
      "loss": 0.2612,
      "num_input_tokens_seen": 24766800,
      "step": 32475
    },
    {
      "epoch": 67.52598752598753,
      "grad_norm": 0.00034861263702623546,
      "learning_rate": 0.025417176041182793,
      "loss": 0.2685,
      "num_input_tokens_seen": 24770832,
      "step": 32480
    },
    {
      "epoch": 67.53638253638253,
      "grad_norm": 0.0002008568262681365,
      "learning_rate": 0.025384379124113596,
      "loss": 0.2678,
      "num_input_tokens_seen": 24774576,
      "step": 32485
    },
    {
      "epoch": 67.54677754677755,
      "grad_norm": 0.000246087001869455,
      "learning_rate": 0.025351601424339124,
      "loss": 0.2325,
      "num_input_tokens_seen": 24778480,
      "step": 32490
    },
    {
      "epoch": 67.55717255717256,
      "grad_norm": 0.00024455785751342773,
      "learning_rate": 0.025318842946914184,
      "loss": 0.247,
      "num_input_tokens_seen": 24782224,
      "step": 32495
    },
    {
      "epoch": 67.56756756756756,
      "grad_norm": 0.0003561387420631945,
      "learning_rate": 0.025286103696890494,
      "loss": 0.2718,
      "num_input_tokens_seen": 24785968,
      "step": 32500
    },
    {
      "epoch": 67.57796257796258,
      "grad_norm": 0.0010462590726092458,
      "learning_rate": 0.025253383679316836,
      "loss": 0.2729,
      "num_input_tokens_seen": 24789744,
      "step": 32505
    },
    {
      "epoch": 67.58835758835758,
      "grad_norm": 0.0003633305022958666,
      "learning_rate": 0.025220682899239077,
      "loss": 0.271,
      "num_input_tokens_seen": 24793488,
      "step": 32510
    },
    {
      "epoch": 67.5987525987526,
      "grad_norm": 0.0001137854196713306,
      "learning_rate": 0.02518800136170013,
      "loss": 0.264,
      "num_input_tokens_seen": 24797424,
      "step": 32515
    },
    {
      "epoch": 67.60914760914761,
      "grad_norm": 0.0008328195544891059,
      "learning_rate": 0.02515533907173981,
      "loss": 0.2957,
      "num_input_tokens_seen": 24801424,
      "step": 32520
    },
    {
      "epoch": 67.61954261954261,
      "grad_norm": 0.0006467492785304785,
      "learning_rate": 0.025122696034395115,
      "loss": 0.2696,
      "num_input_tokens_seen": 24805328,
      "step": 32525
    },
    {
      "epoch": 67.62993762993763,
      "grad_norm": 0.0005138764390721917,
      "learning_rate": 0.025090072254700023,
      "loss": 0.2668,
      "num_input_tokens_seen": 24808944,
      "step": 32530
    },
    {
      "epoch": 67.64033264033264,
      "grad_norm": 0.00035850287531502545,
      "learning_rate": 0.025057467737685468,
      "loss": 0.2633,
      "num_input_tokens_seen": 24812848,
      "step": 32535
    },
    {
      "epoch": 67.65072765072765,
      "grad_norm": 0.0002940946724265814,
      "learning_rate": 0.025024882488379557,
      "loss": 0.2671,
      "num_input_tokens_seen": 24816496,
      "step": 32540
    },
    {
      "epoch": 67.66112266112266,
      "grad_norm": 0.00015440104471053928,
      "learning_rate": 0.02499231651180727,
      "loss": 0.2492,
      "num_input_tokens_seen": 24820304,
      "step": 32545
    },
    {
      "epoch": 67.67151767151768,
      "grad_norm": 0.00019005945068784058,
      "learning_rate": 0.024959769812990713,
      "loss": 0.263,
      "num_input_tokens_seen": 24824112,
      "step": 32550
    },
    {
      "epoch": 67.68191268191268,
      "grad_norm": 0.0002867590810637921,
      "learning_rate": 0.024927242396949045,
      "loss": 0.2684,
      "num_input_tokens_seen": 24827856,
      "step": 32555
    },
    {
      "epoch": 67.6923076923077,
      "grad_norm": 0.00021186417143326253,
      "learning_rate": 0.02489473426869836,
      "loss": 0.2481,
      "num_input_tokens_seen": 24831600,
      "step": 32560
    },
    {
      "epoch": 67.70270270270271,
      "grad_norm": 0.00025059765903279185,
      "learning_rate": 0.024862245433251776,
      "loss": 0.2273,
      "num_input_tokens_seen": 24835312,
      "step": 32565
    },
    {
      "epoch": 67.71309771309771,
      "grad_norm": 0.0003760459367185831,
      "learning_rate": 0.024829775895619577,
      "loss": 0.2765,
      "num_input_tokens_seen": 24839280,
      "step": 32570
    },
    {
      "epoch": 67.72349272349273,
      "grad_norm": 0.0001189417889690958,
      "learning_rate": 0.024797325660808882,
      "loss": 0.2667,
      "num_input_tokens_seen": 24843088,
      "step": 32575
    },
    {
      "epoch": 67.73388773388774,
      "grad_norm": 7.178210944402963e-05,
      "learning_rate": 0.02476489473382401,
      "loss": 0.2661,
      "num_input_tokens_seen": 24846800,
      "step": 32580
    },
    {
      "epoch": 67.74428274428274,
      "grad_norm": 0.0004469225532375276,
      "learning_rate": 0.024732483119666127,
      "loss": 0.2699,
      "num_input_tokens_seen": 24850736,
      "step": 32585
    },
    {
      "epoch": 67.75467775467776,
      "grad_norm": 0.00019983664969913661,
      "learning_rate": 0.024700090823333548,
      "loss": 0.2463,
      "num_input_tokens_seen": 24854480,
      "step": 32590
    },
    {
      "epoch": 67.76507276507276,
      "grad_norm": 0.00039125647163018584,
      "learning_rate": 0.02466771784982163,
      "loss": 0.282,
      "num_input_tokens_seen": 24858288,
      "step": 32595
    },
    {
      "epoch": 67.77546777546777,
      "grad_norm": 0.00014569636550731957,
      "learning_rate": 0.024635364204122594,
      "loss": 0.2678,
      "num_input_tokens_seen": 24862224,
      "step": 32600
    },
    {
      "epoch": 67.77546777546777,
      "eval_loss": 0.24993912875652313,
      "eval_runtime": 13.3855,
      "eval_samples_per_second": 63.95,
      "eval_steps_per_second": 15.987,
      "num_input_tokens_seen": 24862224,
      "step": 32600
    },
    {
      "epoch": 67.78586278586279,
      "grad_norm": 0.0004920413484796882,
      "learning_rate": 0.024603029891225852,
      "loss": 0.2744,
      "num_input_tokens_seen": 24866160,
      "step": 32605
    },
    {
      "epoch": 67.79625779625779,
      "grad_norm": 0.0010251975618302822,
      "learning_rate": 0.024570714916117748,
      "loss": 0.2826,
      "num_input_tokens_seen": 24869968,
      "step": 32610
    },
    {
      "epoch": 67.8066528066528,
      "grad_norm": 0.0005064397118985653,
      "learning_rate": 0.024538419283781625,
      "loss": 0.2782,
      "num_input_tokens_seen": 24873712,
      "step": 32615
    },
    {
      "epoch": 67.81704781704782,
      "grad_norm": 0.00044494454050436616,
      "learning_rate": 0.024506142999197938,
      "loss": 0.2618,
      "num_input_tokens_seen": 24877712,
      "step": 32620
    },
    {
      "epoch": 67.82744282744282,
      "grad_norm": 0.00012076876737410203,
      "learning_rate": 0.024473886067344002,
      "loss": 0.2752,
      "num_input_tokens_seen": 24881360,
      "step": 32625
    },
    {
      "epoch": 67.83783783783784,
      "grad_norm": 0.0001230451016454026,
      "learning_rate": 0.02444164849319434,
      "loss": 0.2518,
      "num_input_tokens_seen": 24885296,
      "step": 32630
    },
    {
      "epoch": 67.84823284823285,
      "grad_norm": 0.00011341610661474988,
      "learning_rate": 0.024409430281720306,
      "loss": 0.2569,
      "num_input_tokens_seen": 24889072,
      "step": 32635
    },
    {
      "epoch": 67.85862785862786,
      "grad_norm": 0.0005860665114596486,
      "learning_rate": 0.024377231437890428,
      "loss": 0.2652,
      "num_input_tokens_seen": 24892944,
      "step": 32640
    },
    {
      "epoch": 67.86902286902287,
      "grad_norm": 0.0006074450793676078,
      "learning_rate": 0.024345051966670115,
      "loss": 0.2492,
      "num_input_tokens_seen": 24896624,
      "step": 32645
    },
    {
      "epoch": 67.87941787941789,
      "grad_norm": 0.00017936533549800515,
      "learning_rate": 0.024312891873021884,
      "loss": 0.2419,
      "num_input_tokens_seen": 24900432,
      "step": 32650
    },
    {
      "epoch": 67.88981288981289,
      "grad_norm": 0.00037500725011341274,
      "learning_rate": 0.024280751161905183,
      "loss": 0.2381,
      "num_input_tokens_seen": 24904272,
      "step": 32655
    },
    {
      "epoch": 67.9002079002079,
      "grad_norm": 0.00024127135111484677,
      "learning_rate": 0.02424862983827658,
      "loss": 0.2435,
      "num_input_tokens_seen": 24908048,
      "step": 32660
    },
    {
      "epoch": 67.9106029106029,
      "grad_norm": 0.00033339319634251297,
      "learning_rate": 0.024216527907089495,
      "loss": 0.256,
      "num_input_tokens_seen": 24911920,
      "step": 32665
    },
    {
      "epoch": 67.92099792099792,
      "grad_norm": 0.001408817945048213,
      "learning_rate": 0.024184445373294505,
      "loss": 0.2452,
      "num_input_tokens_seen": 24915664,
      "step": 32670
    },
    {
      "epoch": 67.93139293139293,
      "grad_norm": 0.0003269794979132712,
      "learning_rate": 0.02415238224183918,
      "loss": 0.2764,
      "num_input_tokens_seen": 24919440,
      "step": 32675
    },
    {
      "epoch": 67.94178794178794,
      "grad_norm": 0.0001744201872497797,
      "learning_rate": 0.024120338517667973,
      "loss": 0.2621,
      "num_input_tokens_seen": 24923376,
      "step": 32680
    },
    {
      "epoch": 67.95218295218295,
      "grad_norm": 0.0002957224496640265,
      "learning_rate": 0.02408831420572247,
      "loss": 0.2896,
      "num_input_tokens_seen": 24927024,
      "step": 32685
    },
    {
      "epoch": 67.96257796257797,
      "grad_norm": 0.000302391272271052,
      "learning_rate": 0.024056309310941264,
      "loss": 0.2399,
      "num_input_tokens_seen": 24930832,
      "step": 32690
    },
    {
      "epoch": 67.97297297297297,
      "grad_norm": 8.811803127173334e-05,
      "learning_rate": 0.02402432383825982,
      "loss": 0.2718,
      "num_input_tokens_seen": 24934672,
      "step": 32695
    },
    {
      "epoch": 67.98336798336798,
      "grad_norm": 0.00011341377103235573,
      "learning_rate": 0.023992357792610792,
      "loss": 0.2405,
      "num_input_tokens_seen": 24938480,
      "step": 32700
    },
    {
      "epoch": 67.993762993763,
      "grad_norm": 0.0002103781735058874,
      "learning_rate": 0.0239604111789237,
      "loss": 0.2615,
      "num_input_tokens_seen": 24942256,
      "step": 32705
    },
    {
      "epoch": 68.004158004158,
      "grad_norm": 0.00018646445823833346,
      "learning_rate": 0.023928484002125095,
      "loss": 0.2501,
      "num_input_tokens_seen": 24946272,
      "step": 32710
    },
    {
      "epoch": 68.01455301455302,
      "grad_norm": 0.00011237052240176126,
      "learning_rate": 0.023896576267138595,
      "loss": 0.245,
      "num_input_tokens_seen": 24950080,
      "step": 32715
    },
    {
      "epoch": 68.02494802494803,
      "grad_norm": 8.91788222361356e-05,
      "learning_rate": 0.02386468797888471,
      "loss": 0.2589,
      "num_input_tokens_seen": 24953984,
      "step": 32720
    },
    {
      "epoch": 68.03534303534303,
      "grad_norm": 0.00012821443669963628,
      "learning_rate": 0.023832819142281057,
      "loss": 0.2761,
      "num_input_tokens_seen": 24957696,
      "step": 32725
    },
    {
      "epoch": 68.04573804573805,
      "grad_norm": 0.00021301429660525173,
      "learning_rate": 0.02380096976224225,
      "loss": 0.2588,
      "num_input_tokens_seen": 24961504,
      "step": 32730
    },
    {
      "epoch": 68.05613305613305,
      "grad_norm": 6.909112562425435e-05,
      "learning_rate": 0.023769139843679777,
      "loss": 0.2511,
      "num_input_tokens_seen": 24965248,
      "step": 32735
    },
    {
      "epoch": 68.06652806652806,
      "grad_norm": 0.0005465694703161716,
      "learning_rate": 0.023737329391502287,
      "loss": 0.2597,
      "num_input_tokens_seen": 24968960,
      "step": 32740
    },
    {
      "epoch": 68.07692307692308,
      "grad_norm": 0.00013574378681369126,
      "learning_rate": 0.023705538410615293,
      "loss": 0.253,
      "num_input_tokens_seen": 24972928,
      "step": 32745
    },
    {
      "epoch": 68.08731808731808,
      "grad_norm": 0.00020517090160865337,
      "learning_rate": 0.023673766905921396,
      "loss": 0.2598,
      "num_input_tokens_seen": 24976768,
      "step": 32750
    },
    {
      "epoch": 68.0977130977131,
      "grad_norm": 0.0009704021504148841,
      "learning_rate": 0.0236420148823202,
      "loss": 0.3023,
      "num_input_tokens_seen": 24980704,
      "step": 32755
    },
    {
      "epoch": 68.10810810810811,
      "grad_norm": 0.00021930535149294883,
      "learning_rate": 0.02361028234470816,
      "loss": 0.269,
      "num_input_tokens_seen": 24984704,
      "step": 32760
    },
    {
      "epoch": 68.11850311850311,
      "grad_norm": 0.0001617156231077388,
      "learning_rate": 0.023578569297978913,
      "loss": 0.2598,
      "num_input_tokens_seen": 24988384,
      "step": 32765
    },
    {
      "epoch": 68.12889812889813,
      "grad_norm": 6.561173358932137e-05,
      "learning_rate": 0.023546875747023025,
      "loss": 0.2815,
      "num_input_tokens_seen": 24992096,
      "step": 32770
    },
    {
      "epoch": 68.13929313929314,
      "grad_norm": 0.00018323773110751063,
      "learning_rate": 0.02351520169672801,
      "loss": 0.258,
      "num_input_tokens_seen": 24995936,
      "step": 32775
    },
    {
      "epoch": 68.14968814968815,
      "grad_norm": 0.00039092014776542783,
      "learning_rate": 0.023483547151978357,
      "loss": 0.2766,
      "num_input_tokens_seen": 24999744,
      "step": 32780
    },
    {
      "epoch": 68.16008316008316,
      "grad_norm": 0.0001044395612552762,
      "learning_rate": 0.023451912117655675,
      "loss": 0.2503,
      "num_input_tokens_seen": 25003584,
      "step": 32785
    },
    {
      "epoch": 68.17047817047818,
      "grad_norm": 0.0005508205504156649,
      "learning_rate": 0.023420296598638417,
      "loss": 0.2441,
      "num_input_tokens_seen": 25007424,
      "step": 32790
    },
    {
      "epoch": 68.18087318087318,
      "grad_norm": 0.00016276107635349035,
      "learning_rate": 0.023388700599802165,
      "loss": 0.2743,
      "num_input_tokens_seen": 25011328,
      "step": 32795
    },
    {
      "epoch": 68.1912681912682,
      "grad_norm": 0.002063712803646922,
      "learning_rate": 0.023357124126019334,
      "loss": 0.2658,
      "num_input_tokens_seen": 25015296,
      "step": 32800
    },
    {
      "epoch": 68.1912681912682,
      "eval_loss": 0.2485274374485016,
      "eval_runtime": 13.3989,
      "eval_samples_per_second": 63.886,
      "eval_steps_per_second": 15.972,
      "num_input_tokens_seen": 25015296,
      "step": 32800
    },
    {
      "epoch": 68.20166320166321,
      "grad_norm": 0.0008883222471922636,
      "learning_rate": 0.02332556718215945,
      "loss": 0.2653,
      "num_input_tokens_seen": 25019072,
      "step": 32805
    },
    {
      "epoch": 68.21205821205821,
      "grad_norm": 0.0003568447136785835,
      "learning_rate": 0.023294029773089035,
      "loss": 0.2608,
      "num_input_tokens_seen": 25023104,
      "step": 32810
    },
    {
      "epoch": 68.22245322245323,
      "grad_norm": 0.0004084985994268209,
      "learning_rate": 0.023262511903671484,
      "loss": 0.2631,
      "num_input_tokens_seen": 25026912,
      "step": 32815
    },
    {
      "epoch": 68.23284823284823,
      "grad_norm": 0.0003029360668733716,
      "learning_rate": 0.023231013578767324,
      "loss": 0.2559,
      "num_input_tokens_seen": 25030528,
      "step": 32820
    },
    {
      "epoch": 68.24324324324324,
      "grad_norm": 0.00018699109205044806,
      "learning_rate": 0.0231995348032339,
      "loss": 0.2592,
      "num_input_tokens_seen": 25034304,
      "step": 32825
    },
    {
      "epoch": 68.25363825363826,
      "grad_norm": 0.0002158731804229319,
      "learning_rate": 0.023168075581925685,
      "loss": 0.2458,
      "num_input_tokens_seen": 25038176,
      "step": 32830
    },
    {
      "epoch": 68.26403326403326,
      "grad_norm": 0.00014035528874956071,
      "learning_rate": 0.023136635919694126,
      "loss": 0.2699,
      "num_input_tokens_seen": 25042208,
      "step": 32835
    },
    {
      "epoch": 68.27442827442827,
      "grad_norm": 0.00022370957594830543,
      "learning_rate": 0.02310521582138753,
      "loss": 0.2888,
      "num_input_tokens_seen": 25045984,
      "step": 32840
    },
    {
      "epoch": 68.28482328482329,
      "grad_norm": 0.00040909109520725906,
      "learning_rate": 0.023073815291851357,
      "loss": 0.2568,
      "num_input_tokens_seen": 25049856,
      "step": 32845
    },
    {
      "epoch": 68.29521829521829,
      "grad_norm": 0.0003214692114852369,
      "learning_rate": 0.02304243433592788,
      "loss": 0.263,
      "num_input_tokens_seen": 25053536,
      "step": 32850
    },
    {
      "epoch": 68.3056133056133,
      "grad_norm": 0.00039924561860971153,
      "learning_rate": 0.023011072958456513,
      "loss": 0.2526,
      "num_input_tokens_seen": 25057248,
      "step": 32855
    },
    {
      "epoch": 68.31600831600832,
      "grad_norm": 0.00018264653044752777,
      "learning_rate": 0.022979731164273536,
      "loss": 0.2811,
      "num_input_tokens_seen": 25060960,
      "step": 32860
    },
    {
      "epoch": 68.32640332640332,
      "grad_norm": 0.000957558280788362,
      "learning_rate": 0.022948408958212218,
      "loss": 0.2623,
      "num_input_tokens_seen": 25064800,
      "step": 32865
    },
    {
      "epoch": 68.33679833679834,
      "grad_norm": 0.0003596000315155834,
      "learning_rate": 0.022917106345102876,
      "loss": 0.2737,
      "num_input_tokens_seen": 25068672,
      "step": 32870
    },
    {
      "epoch": 68.34719334719335,
      "grad_norm": 0.00018152670236304402,
      "learning_rate": 0.022885823329772785,
      "loss": 0.2812,
      "num_input_tokens_seen": 25072512,
      "step": 32875
    },
    {
      "epoch": 68.35758835758836,
      "grad_norm": 0.0002636392309796065,
      "learning_rate": 0.02285455991704612,
      "loss": 0.2594,
      "num_input_tokens_seen": 25076320,
      "step": 32880
    },
    {
      "epoch": 68.36798336798337,
      "grad_norm": 0.00023016483464743942,
      "learning_rate": 0.022823316111744117,
      "loss": 0.275,
      "num_input_tokens_seen": 25080064,
      "step": 32885
    },
    {
      "epoch": 68.37837837837837,
      "grad_norm": 0.0004171333566773683,
      "learning_rate": 0.022792091918685014,
      "loss": 0.2454,
      "num_input_tokens_seen": 25083936,
      "step": 32890
    },
    {
      "epoch": 68.38877338877339,
      "grad_norm": 0.0001815782452467829,
      "learning_rate": 0.022760887342683906,
      "loss": 0.26,
      "num_input_tokens_seen": 25087776,
      "step": 32895
    },
    {
      "epoch": 68.3991683991684,
      "grad_norm": 0.0001997713843593374,
      "learning_rate": 0.022729702388552975,
      "loss": 0.2495,
      "num_input_tokens_seen": 25091520,
      "step": 32900
    },
    {
      "epoch": 68.4095634095634,
      "grad_norm": 8.17597538116388e-05,
      "learning_rate": 0.022698537061101292,
      "loss": 0.2798,
      "num_input_tokens_seen": 25095360,
      "step": 32905
    },
    {
      "epoch": 68.41995841995842,
      "grad_norm": 0.0002609244838822633,
      "learning_rate": 0.022667391365134962,
      "loss": 0.2579,
      "num_input_tokens_seen": 25099232,
      "step": 32910
    },
    {
      "epoch": 68.43035343035343,
      "grad_norm": 0.0004053963057231158,
      "learning_rate": 0.022636265305457065,
      "loss": 0.267,
      "num_input_tokens_seen": 25103104,
      "step": 32915
    },
    {
      "epoch": 68.44074844074844,
      "grad_norm": 7.771192758809775e-05,
      "learning_rate": 0.02260515888686764,
      "loss": 0.2625,
      "num_input_tokens_seen": 25106976,
      "step": 32920
    },
    {
      "epoch": 68.45114345114345,
      "grad_norm": 0.0001673835067776963,
      "learning_rate": 0.022574072114163596,
      "loss": 0.2698,
      "num_input_tokens_seen": 25110912,
      "step": 32925
    },
    {
      "epoch": 68.46153846153847,
      "grad_norm": 0.00011636003182502463,
      "learning_rate": 0.022543004992139005,
      "loss": 0.2857,
      "num_input_tokens_seen": 25114688,
      "step": 32930
    },
    {
      "epoch": 68.47193347193347,
      "grad_norm": 0.0008483059937134385,
      "learning_rate": 0.022511957525584745,
      "loss": 0.2512,
      "num_input_tokens_seen": 25118464,
      "step": 32935
    },
    {
      "epoch": 68.48232848232848,
      "grad_norm": 5.464487549033947e-05,
      "learning_rate": 0.022480929719288778,
      "loss": 0.2619,
      "num_input_tokens_seen": 25122368,
      "step": 32940
    },
    {
      "epoch": 68.4927234927235,
      "grad_norm": 0.0001387945667374879,
      "learning_rate": 0.02244992157803592,
      "loss": 0.2483,
      "num_input_tokens_seen": 25126144,
      "step": 32945
    },
    {
      "epoch": 68.5031185031185,
      "grad_norm": 0.0006583930226042867,
      "learning_rate": 0.022418933106608047,
      "loss": 0.2554,
      "num_input_tokens_seen": 25129920,
      "step": 32950
    },
    {
      "epoch": 68.51351351351352,
      "grad_norm": 7.203652057796717e-05,
      "learning_rate": 0.022387964309784018,
      "loss": 0.2769,
      "num_input_tokens_seen": 25133792,
      "step": 32955
    },
    {
      "epoch": 68.52390852390852,
      "grad_norm": 0.0002533787046559155,
      "learning_rate": 0.022357015192339517,
      "loss": 0.291,
      "num_input_tokens_seen": 25137568,
      "step": 32960
    },
    {
      "epoch": 68.53430353430353,
      "grad_norm": 0.0006018451531417668,
      "learning_rate": 0.02232608575904734,
      "loss": 0.2641,
      "num_input_tokens_seen": 25141344,
      "step": 32965
    },
    {
      "epoch": 68.54469854469855,
      "grad_norm": 0.00037659157533198595,
      "learning_rate": 0.022295176014677225,
      "loss": 0.245,
      "num_input_tokens_seen": 25145216,
      "step": 32970
    },
    {
      "epoch": 68.55509355509355,
      "grad_norm": 0.00038553026388399303,
      "learning_rate": 0.02226428596399577,
      "loss": 0.2328,
      "num_input_tokens_seen": 25148736,
      "step": 32975
    },
    {
      "epoch": 68.56548856548856,
      "grad_norm": 0.00038493535248562694,
      "learning_rate": 0.02223341561176669,
      "loss": 0.2736,
      "num_input_tokens_seen": 25152576,
      "step": 32980
    },
    {
      "epoch": 68.57588357588358,
      "grad_norm": 0.00020661762391682714,
      "learning_rate": 0.0222025649627505,
      "loss": 0.2703,
      "num_input_tokens_seen": 25156384,
      "step": 32985
    },
    {
      "epoch": 68.58627858627858,
      "grad_norm": 0.00038398735341615975,
      "learning_rate": 0.022171734021704814,
      "loss": 0.2699,
      "num_input_tokens_seen": 25160128,
      "step": 32990
    },
    {
      "epoch": 68.5966735966736,
      "grad_norm": 0.0002705395163502544,
      "learning_rate": 0.022140922793384116,
      "loss": 0.2506,
      "num_input_tokens_seen": 25163904,
      "step": 32995
    },
    {
      "epoch": 68.60706860706861,
      "grad_norm": 0.00017358586774207652,
      "learning_rate": 0.022110131282539934,
      "loss": 0.2544,
      "num_input_tokens_seen": 25167744,
      "step": 33000
    },
    {
      "epoch": 68.60706860706861,
      "eval_loss": 0.24799275398254395,
      "eval_runtime": 13.4,
      "eval_samples_per_second": 63.881,
      "eval_steps_per_second": 15.97,
      "num_input_tokens_seen": 25167744,
      "step": 33000
    },
    {
      "epoch": 68.61746361746361,
      "grad_norm": 0.0007404489442706108,
      "learning_rate": 0.022079359493920675,
      "loss": 0.2498,
      "num_input_tokens_seen": 25171456,
      "step": 33005
    },
    {
      "epoch": 68.62785862785863,
      "grad_norm": 0.00015253355377353728,
      "learning_rate": 0.02204860743227169,
      "loss": 0.2264,
      "num_input_tokens_seen": 25175264,
      "step": 33010
    },
    {
      "epoch": 68.63825363825364,
      "grad_norm": 0.0002659687597770244,
      "learning_rate": 0.022017875102335365,
      "loss": 0.2727,
      "num_input_tokens_seen": 25179072,
      "step": 33015
    },
    {
      "epoch": 68.64864864864865,
      "grad_norm": 0.0003651087172329426,
      "learning_rate": 0.02198716250885108,
      "loss": 0.2525,
      "num_input_tokens_seen": 25182944,
      "step": 33020
    },
    {
      "epoch": 68.65904365904366,
      "grad_norm": 0.0003434156533330679,
      "learning_rate": 0.021956469656555,
      "loss": 0.2726,
      "num_input_tokens_seen": 25186816,
      "step": 33025
    },
    {
      "epoch": 68.66943866943868,
      "grad_norm": 0.0005010980530641973,
      "learning_rate": 0.0219257965501804,
      "loss": 0.2387,
      "num_input_tokens_seen": 25190560,
      "step": 33030
    },
    {
      "epoch": 68.67983367983368,
      "grad_norm": 0.0002615599369164556,
      "learning_rate": 0.021895143194457494,
      "loss": 0.2414,
      "num_input_tokens_seen": 25194432,
      "step": 33035
    },
    {
      "epoch": 68.6902286902287,
      "grad_norm": 0.00036230238038115203,
      "learning_rate": 0.021864509594113322,
      "loss": 0.2827,
      "num_input_tokens_seen": 25198304,
      "step": 33040
    },
    {
      "epoch": 68.7006237006237,
      "grad_norm": 0.0011337435571476817,
      "learning_rate": 0.02183389575387207,
      "loss": 0.2729,
      "num_input_tokens_seen": 25202080,
      "step": 33045
    },
    {
      "epoch": 68.71101871101871,
      "grad_norm": 0.0002490543120075017,
      "learning_rate": 0.021803301678454682,
      "loss": 0.265,
      "num_input_tokens_seen": 25205760,
      "step": 33050
    },
    {
      "epoch": 68.72141372141373,
      "grad_norm": 0.00026567679014988244,
      "learning_rate": 0.021772727372579213,
      "loss": 0.2789,
      "num_input_tokens_seen": 25209472,
      "step": 33055
    },
    {
      "epoch": 68.73180873180873,
      "grad_norm": 0.00033336292835883796,
      "learning_rate": 0.02174217284096061,
      "loss": 0.2713,
      "num_input_tokens_seen": 25213280,
      "step": 33060
    },
    {
      "epoch": 68.74220374220374,
      "grad_norm": 0.0003149026888422668,
      "learning_rate": 0.0217116380883107,
      "loss": 0.2696,
      "num_input_tokens_seen": 25217344,
      "step": 33065
    },
    {
      "epoch": 68.75259875259876,
      "grad_norm": 0.0007255618693307042,
      "learning_rate": 0.021681123119338425,
      "loss": 0.2921,
      "num_input_tokens_seen": 25221088,
      "step": 33070
    },
    {
      "epoch": 68.76299376299376,
      "grad_norm": 0.00048081984277814627,
      "learning_rate": 0.02165062793874951,
      "loss": 0.2862,
      "num_input_tokens_seen": 25224832,
      "step": 33075
    },
    {
      "epoch": 68.77338877338877,
      "grad_norm": 0.00022424502822104841,
      "learning_rate": 0.021620152551246666,
      "loss": 0.2738,
      "num_input_tokens_seen": 25228768,
      "step": 33080
    },
    {
      "epoch": 68.78378378378379,
      "grad_norm": 0.00021955319971311837,
      "learning_rate": 0.02158969696152967,
      "loss": 0.2548,
      "num_input_tokens_seen": 25232512,
      "step": 33085
    },
    {
      "epoch": 68.79417879417879,
      "grad_norm": 0.00022503330546896905,
      "learning_rate": 0.021559261174295057,
      "loss": 0.2496,
      "num_input_tokens_seen": 25236608,
      "step": 33090
    },
    {
      "epoch": 68.8045738045738,
      "grad_norm": 9.160657646134496e-05,
      "learning_rate": 0.02152884519423646,
      "loss": 0.2683,
      "num_input_tokens_seen": 25240480,
      "step": 33095
    },
    {
      "epoch": 68.81496881496882,
      "grad_norm": 0.00015715994231868535,
      "learning_rate": 0.021498449026044447,
      "loss": 0.2686,
      "num_input_tokens_seen": 25244448,
      "step": 33100
    },
    {
      "epoch": 68.82536382536382,
      "grad_norm": 0.00019138022616971284,
      "learning_rate": 0.021468072674406414,
      "loss": 0.2599,
      "num_input_tokens_seen": 25248288,
      "step": 33105
    },
    {
      "epoch": 68.83575883575884,
      "grad_norm": 0.00013834668789058924,
      "learning_rate": 0.021437716144006795,
      "loss": 0.2378,
      "num_input_tokens_seen": 25252064,
      "step": 33110
    },
    {
      "epoch": 68.84615384615384,
      "grad_norm": 0.0006101626786403358,
      "learning_rate": 0.021407379439527002,
      "loss": 0.304,
      "num_input_tokens_seen": 25255712,
      "step": 33115
    },
    {
      "epoch": 68.85654885654886,
      "grad_norm": 0.000161875577759929,
      "learning_rate": 0.021377062565645255,
      "loss": 0.2737,
      "num_input_tokens_seen": 25259552,
      "step": 33120
    },
    {
      "epoch": 68.86694386694387,
      "grad_norm": 0.0004340157029218972,
      "learning_rate": 0.02134676552703688,
      "loss": 0.228,
      "num_input_tokens_seen": 25263360,
      "step": 33125
    },
    {
      "epoch": 68.87733887733887,
      "grad_norm": 0.0006379140540957451,
      "learning_rate": 0.02131648832837398,
      "loss": 0.235,
      "num_input_tokens_seen": 25267168,
      "step": 33130
    },
    {
      "epoch": 68.88773388773389,
      "grad_norm": 0.0013916142052039504,
      "learning_rate": 0.02128623097432574,
      "loss": 0.2503,
      "num_input_tokens_seen": 25270976,
      "step": 33135
    },
    {
      "epoch": 68.8981288981289,
      "grad_norm": 0.0006769343162886798,
      "learning_rate": 0.021255993469558192,
      "loss": 0.2817,
      "num_input_tokens_seen": 25274848,
      "step": 33140
    },
    {
      "epoch": 68.9085239085239,
      "grad_norm": 5.576851253863424e-05,
      "learning_rate": 0.021225775818734364,
      "loss": 0.2525,
      "num_input_tokens_seen": 25278720,
      "step": 33145
    },
    {
      "epoch": 68.91891891891892,
      "grad_norm": 0.00017070425383280963,
      "learning_rate": 0.021195578026514166,
      "loss": 0.2817,
      "num_input_tokens_seen": 25282496,
      "step": 33150
    },
    {
      "epoch": 68.92931392931393,
      "grad_norm": 0.0003589412954170257,
      "learning_rate": 0.02116540009755452,
      "loss": 0.2542,
      "num_input_tokens_seen": 25286336,
      "step": 33155
    },
    {
      "epoch": 68.93970893970894,
      "grad_norm": 0.001447805087082088,
      "learning_rate": 0.021135242036509173,
      "loss": 0.2799,
      "num_input_tokens_seen": 25290208,
      "step": 33160
    },
    {
      "epoch": 68.95010395010395,
      "grad_norm": 0.00031258424860425293,
      "learning_rate": 0.021105103848028967,
      "loss": 0.2651,
      "num_input_tokens_seen": 25293888,
      "step": 33165
    },
    {
      "epoch": 68.96049896049897,
      "grad_norm": 0.0004397298616822809,
      "learning_rate": 0.021074985536761504,
      "loss": 0.2648,
      "num_input_tokens_seen": 25297632,
      "step": 33170
    },
    {
      "epoch": 68.97089397089397,
      "grad_norm": 0.00019049931142944843,
      "learning_rate": 0.021044887107351435,
      "loss": 0.26,
      "num_input_tokens_seen": 25301280,
      "step": 33175
    },
    {
      "epoch": 68.98128898128898,
      "grad_norm": 0.00013810255040880293,
      "learning_rate": 0.021014808564440362,
      "loss": 0.2806,
      "num_input_tokens_seen": 25305088,
      "step": 33180
    },
    {
      "epoch": 68.99168399168398,
      "grad_norm": 0.00013612354814540595,
      "learning_rate": 0.02098474991266671,
      "loss": 0.2687,
      "num_input_tokens_seen": 25309024,
      "step": 33185
    },
    {
      "epoch": 69.002079002079,
      "grad_norm": 0.00013730025966651738,
      "learning_rate": 0.02095471115666592,
      "loss": 0.2594,
      "num_input_tokens_seen": 25312984,
      "step": 33190
    },
    {
      "epoch": 69.01247401247402,
      "grad_norm": 0.0003259215736761689,
      "learning_rate": 0.020924692301070406,
      "loss": 0.2717,
      "num_input_tokens_seen": 25317016,
      "step": 33195
    },
    {
      "epoch": 69.02286902286902,
      "grad_norm": 0.00033645742223598063,
      "learning_rate": 0.020894693350509346,
      "loss": 0.2748,
      "num_input_tokens_seen": 25321016,
      "step": 33200
    },
    {
      "epoch": 69.02286902286902,
      "eval_loss": 0.2525767683982849,
      "eval_runtime": 13.3975,
      "eval_samples_per_second": 63.893,
      "eval_steps_per_second": 15.973,
      "num_input_tokens_seen": 25321016,
      "step": 33200
    },
    {
      "epoch": 69.03326403326403,
      "grad_norm": 0.0004571609606500715,
      "learning_rate": 0.020864714309609057,
      "loss": 0.2902,
      "num_input_tokens_seen": 25324824,
      "step": 33205
    },
    {
      "epoch": 69.04365904365905,
      "grad_norm": 0.00010474302689544857,
      "learning_rate": 0.020834755182992604,
      "loss": 0.2771,
      "num_input_tokens_seen": 25328440,
      "step": 33210
    },
    {
      "epoch": 69.05405405405405,
      "grad_norm": 0.0005784222157672048,
      "learning_rate": 0.02080481597528011,
      "loss": 0.2734,
      "num_input_tokens_seen": 25332312,
      "step": 33215
    },
    {
      "epoch": 69.06444906444906,
      "grad_norm": 7.020789053058252e-05,
      "learning_rate": 0.020774896691088583,
      "loss": 0.2653,
      "num_input_tokens_seen": 25336088,
      "step": 33220
    },
    {
      "epoch": 69.07484407484408,
      "grad_norm": 0.00028458278393372893,
      "learning_rate": 0.020744997335031882,
      "loss": 0.2752,
      "num_input_tokens_seen": 25339992,
      "step": 33225
    },
    {
      "epoch": 69.08523908523908,
      "grad_norm": 0.0006704650004394352,
      "learning_rate": 0.02071511791172092,
      "loss": 0.2509,
      "num_input_tokens_seen": 25343832,
      "step": 33230
    },
    {
      "epoch": 69.0956340956341,
      "grad_norm": 0.0001475746976211667,
      "learning_rate": 0.02068525842576351,
      "loss": 0.2581,
      "num_input_tokens_seen": 25347768,
      "step": 33235
    },
    {
      "epoch": 69.10602910602911,
      "grad_norm": 0.00018695763719733804,
      "learning_rate": 0.020655418881764264,
      "loss": 0.2527,
      "num_input_tokens_seen": 25351320,
      "step": 33240
    },
    {
      "epoch": 69.11642411642411,
      "grad_norm": 0.0001395938452333212,
      "learning_rate": 0.020625599284324923,
      "loss": 0.2865,
      "num_input_tokens_seen": 25355128,
      "step": 33245
    },
    {
      "epoch": 69.12681912681913,
      "grad_norm": 0.00024050565843936056,
      "learning_rate": 0.02059579963804396,
      "loss": 0.2845,
      "num_input_tokens_seen": 25358840,
      "step": 33250
    },
    {
      "epoch": 69.13721413721414,
      "grad_norm": 7.251149509102106e-05,
      "learning_rate": 0.02056601994751688,
      "loss": 0.2569,
      "num_input_tokens_seen": 25362616,
      "step": 33255
    },
    {
      "epoch": 69.14760914760915,
      "grad_norm": 0.00027005522861145437,
      "learning_rate": 0.02053626021733614,
      "loss": 0.2432,
      "num_input_tokens_seen": 25366296,
      "step": 33260
    },
    {
      "epoch": 69.15800415800416,
      "grad_norm": 0.00025118948542512953,
      "learning_rate": 0.02050652045209097,
      "loss": 0.2638,
      "num_input_tokens_seen": 25370392,
      "step": 33265
    },
    {
      "epoch": 69.16839916839916,
      "grad_norm": 0.00043222654494456947,
      "learning_rate": 0.020476800656367672,
      "loss": 0.2266,
      "num_input_tokens_seen": 25374104,
      "step": 33270
    },
    {
      "epoch": 69.17879417879418,
      "grad_norm": 0.00012028025957988575,
      "learning_rate": 0.020447100834749425,
      "loss": 0.2646,
      "num_input_tokens_seen": 25377912,
      "step": 33275
    },
    {
      "epoch": 69.1891891891892,
      "grad_norm": 0.00024721797672100365,
      "learning_rate": 0.02041742099181627,
      "loss": 0.2498,
      "num_input_tokens_seen": 25381848,
      "step": 33280
    },
    {
      "epoch": 69.1995841995842,
      "grad_norm": 0.0002243156632175669,
      "learning_rate": 0.02038776113214526,
      "loss": 0.2511,
      "num_input_tokens_seen": 25385752,
      "step": 33285
    },
    {
      "epoch": 69.20997920997921,
      "grad_norm": 6.635630415985361e-05,
      "learning_rate": 0.0203581212603103,
      "loss": 0.2377,
      "num_input_tokens_seen": 25389592,
      "step": 33290
    },
    {
      "epoch": 69.22037422037423,
      "grad_norm": 0.0002868381852749735,
      "learning_rate": 0.02032850138088219,
      "loss": 0.2882,
      "num_input_tokens_seen": 25393432,
      "step": 33295
    },
    {
      "epoch": 69.23076923076923,
      "grad_norm": 0.00026127431192435324,
      "learning_rate": 0.020298901498428754,
      "loss": 0.2504,
      "num_input_tokens_seen": 25397208,
      "step": 33300
    },
    {
      "epoch": 69.24116424116424,
      "grad_norm": 0.0002085700980387628,
      "learning_rate": 0.020269321617514595,
      "loss": 0.3077,
      "num_input_tokens_seen": 25401016,
      "step": 33305
    },
    {
      "epoch": 69.25155925155926,
      "grad_norm": 0.0002756815229076892,
      "learning_rate": 0.020239761742701343,
      "loss": 0.2556,
      "num_input_tokens_seen": 25404856,
      "step": 33310
    },
    {
      "epoch": 69.26195426195426,
      "grad_norm": 8.619811706012115e-05,
      "learning_rate": 0.02021022187854754,
      "loss": 0.26,
      "num_input_tokens_seen": 25408664,
      "step": 33315
    },
    {
      "epoch": 69.27234927234927,
      "grad_norm": 0.0001901060895761475,
      "learning_rate": 0.020180702029608522,
      "loss": 0.2458,
      "num_input_tokens_seen": 25412504,
      "step": 33320
    },
    {
      "epoch": 69.28274428274429,
      "grad_norm": 0.00027060817228630185,
      "learning_rate": 0.020151202200436695,
      "loss": 0.2511,
      "num_input_tokens_seen": 25416216,
      "step": 33325
    },
    {
      "epoch": 69.29313929313929,
      "grad_norm": 0.0004160564858466387,
      "learning_rate": 0.020121722395581226,
      "loss": 0.2525,
      "num_input_tokens_seen": 25420024,
      "step": 33330
    },
    {
      "epoch": 69.3035343035343,
      "grad_norm": 0.0006290625897236168,
      "learning_rate": 0.020092262619588342,
      "loss": 0.2671,
      "num_input_tokens_seen": 25423832,
      "step": 33335
    },
    {
      "epoch": 69.31392931392931,
      "grad_norm": 0.00018528304644860327,
      "learning_rate": 0.02006282287700109,
      "loss": 0.2787,
      "num_input_tokens_seen": 25427704,
      "step": 33340
    },
    {
      "epoch": 69.32432432432432,
      "grad_norm": 0.0003635364701040089,
      "learning_rate": 0.020033403172359427,
      "loss": 0.2449,
      "num_input_tokens_seen": 25431448,
      "step": 33345
    },
    {
      "epoch": 69.33471933471934,
      "grad_norm": 0.0002578072890173644,
      "learning_rate": 0.020004003510200284,
      "loss": 0.2354,
      "num_input_tokens_seen": 25435448,
      "step": 33350
    },
    {
      "epoch": 69.34511434511434,
      "grad_norm": 0.00034021472674794495,
      "learning_rate": 0.019974623895057407,
      "loss": 0.281,
      "num_input_tokens_seen": 25439256,
      "step": 33355
    },
    {
      "epoch": 69.35550935550935,
      "grad_norm": 9.66677616816014e-05,
      "learning_rate": 0.019945264331461553,
      "loss": 0.2592,
      "num_input_tokens_seen": 25442968,
      "step": 33360
    },
    {
      "epoch": 69.36590436590437,
      "grad_norm": 0.0004060154897160828,
      "learning_rate": 0.019915924823940317,
      "loss": 0.2505,
      "num_input_tokens_seen": 25446776,
      "step": 33365
    },
    {
      "epoch": 69.37629937629937,
      "grad_norm": 0.0005675734137184918,
      "learning_rate": 0.01988660537701816,
      "loss": 0.2504,
      "num_input_tokens_seen": 25450648,
      "step": 33370
    },
    {
      "epoch": 69.38669438669439,
      "grad_norm": 0.0004298902931623161,
      "learning_rate": 0.01985730599521659,
      "loss": 0.2795,
      "num_input_tokens_seen": 25454392,
      "step": 33375
    },
    {
      "epoch": 69.3970893970894,
      "grad_norm": 0.00045510605559684336,
      "learning_rate": 0.019828026683053918,
      "loss": 0.2634,
      "num_input_tokens_seen": 25458104,
      "step": 33380
    },
    {
      "epoch": 69.4074844074844,
      "grad_norm": 0.0005070194602012634,
      "learning_rate": 0.01979876744504535,
      "loss": 0.2405,
      "num_input_tokens_seen": 25461784,
      "step": 33385
    },
    {
      "epoch": 69.41787941787942,
      "grad_norm": 0.00026510213501751423,
      "learning_rate": 0.019769528285703046,
      "loss": 0.2761,
      "num_input_tokens_seen": 25465816,
      "step": 33390
    },
    {
      "epoch": 69.42827442827443,
      "grad_norm": 0.00040704559069126844,
      "learning_rate": 0.019740309209536098,
      "loss": 0.2681,
      "num_input_tokens_seen": 25469528,
      "step": 33395
    },
    {
      "epoch": 69.43866943866944,
      "grad_norm": 0.00013673979265149683,
      "learning_rate": 0.019711110221050387,
      "loss": 0.2713,
      "num_input_tokens_seen": 25473368,
      "step": 33400
    },
    {
      "epoch": 69.43866943866944,
      "eval_loss": 0.24969600141048431,
      "eval_runtime": 13.4019,
      "eval_samples_per_second": 63.872,
      "eval_steps_per_second": 15.968,
      "num_input_tokens_seen": 25473368,
      "step": 33400
    },
    {
      "epoch": 69.44906444906445,
      "grad_norm": 0.00023007711570244282,
      "learning_rate": 0.019681931324748825,
      "loss": 0.2683,
      "num_input_tokens_seen": 25477240,
      "step": 33405
    },
    {
      "epoch": 69.45945945945945,
      "grad_norm": 0.00042848187149502337,
      "learning_rate": 0.019652772525131094,
      "loss": 0.2879,
      "num_input_tokens_seen": 25481144,
      "step": 33410
    },
    {
      "epoch": 69.46985446985447,
      "grad_norm": 0.00060894968919456,
      "learning_rate": 0.019623633826693885,
      "loss": 0.2396,
      "num_input_tokens_seen": 25484952,
      "step": 33415
    },
    {
      "epoch": 69.48024948024948,
      "grad_norm": 0.0009679089416749775,
      "learning_rate": 0.019594515233930788,
      "loss": 0.2809,
      "num_input_tokens_seen": 25488728,
      "step": 33420
    },
    {
      "epoch": 69.49064449064448,
      "grad_norm": 0.00020810756541322917,
      "learning_rate": 0.019565416751332186,
      "loss": 0.2558,
      "num_input_tokens_seen": 25492440,
      "step": 33425
    },
    {
      "epoch": 69.5010395010395,
      "grad_norm": 0.00040405269828625023,
      "learning_rate": 0.019536338383385497,
      "loss": 0.2656,
      "num_input_tokens_seen": 25496056,
      "step": 33430
    },
    {
      "epoch": 69.51143451143452,
      "grad_norm": 0.00021596832084469497,
      "learning_rate": 0.019507280134574933,
      "loss": 0.2666,
      "num_input_tokens_seen": 25499992,
      "step": 33435
    },
    {
      "epoch": 69.52182952182952,
      "grad_norm": 0.00027918900013901293,
      "learning_rate": 0.019478242009381624,
      "loss": 0.2536,
      "num_input_tokens_seen": 25503832,
      "step": 33440
    },
    {
      "epoch": 69.53222453222453,
      "grad_norm": 0.000865179521497339,
      "learning_rate": 0.01944922401228367,
      "loss": 0.2857,
      "num_input_tokens_seen": 25507608,
      "step": 33445
    },
    {
      "epoch": 69.54261954261955,
      "grad_norm": 6.656336336163804e-05,
      "learning_rate": 0.01942022614775593,
      "loss": 0.2781,
      "num_input_tokens_seen": 25511416,
      "step": 33450
    },
    {
      "epoch": 69.55301455301455,
      "grad_norm": 0.0001633148203836754,
      "learning_rate": 0.01939124842027029,
      "loss": 0.2693,
      "num_input_tokens_seen": 25515224,
      "step": 33455
    },
    {
      "epoch": 69.56340956340956,
      "grad_norm": 0.00016383688489440829,
      "learning_rate": 0.01936229083429551,
      "loss": 0.2539,
      "num_input_tokens_seen": 25519032,
      "step": 33460
    },
    {
      "epoch": 69.57380457380458,
      "grad_norm": 0.00025212758919224143,
      "learning_rate": 0.019333353394297148,
      "loss": 0.2592,
      "num_input_tokens_seen": 25522936,
      "step": 33465
    },
    {
      "epoch": 69.58419958419958,
      "grad_norm": 0.00016369522199966013,
      "learning_rate": 0.019304436104737754,
      "loss": 0.2567,
      "num_input_tokens_seen": 25526552,
      "step": 33470
    },
    {
      "epoch": 69.5945945945946,
      "grad_norm": 0.00014449565787799656,
      "learning_rate": 0.019275538970076778,
      "loss": 0.2551,
      "num_input_tokens_seen": 25530488,
      "step": 33475
    },
    {
      "epoch": 69.60498960498961,
      "grad_norm": 0.0004477107140701264,
      "learning_rate": 0.019246661994770434,
      "loss": 0.2634,
      "num_input_tokens_seen": 25534232,
      "step": 33480
    },
    {
      "epoch": 69.61538461538461,
      "grad_norm": 0.00016228918684646487,
      "learning_rate": 0.019217805183271985,
      "loss": 0.2501,
      "num_input_tokens_seen": 25538136,
      "step": 33485
    },
    {
      "epoch": 69.62577962577963,
      "grad_norm": 0.0004337795835454017,
      "learning_rate": 0.019188968540031465,
      "loss": 0.269,
      "num_input_tokens_seen": 25542104,
      "step": 33490
    },
    {
      "epoch": 69.63617463617463,
      "grad_norm": 9.822798165259883e-05,
      "learning_rate": 0.019160152069495867,
      "loss": 0.2519,
      "num_input_tokens_seen": 25545880,
      "step": 33495
    },
    {
      "epoch": 69.64656964656965,
      "grad_norm": 0.00025473610730841756,
      "learning_rate": 0.019131355776109103,
      "loss": 0.2733,
      "num_input_tokens_seen": 25549752,
      "step": 33500
    },
    {
      "epoch": 69.65696465696466,
      "grad_norm": 0.0001143800254794769,
      "learning_rate": 0.019102579664311857,
      "loss": 0.2718,
      "num_input_tokens_seen": 25553656,
      "step": 33505
    },
    {
      "epoch": 69.66735966735966,
      "grad_norm": 0.00013031599519308656,
      "learning_rate": 0.019073823738541763,
      "loss": 0.2897,
      "num_input_tokens_seen": 25557368,
      "step": 33510
    },
    {
      "epoch": 69.67775467775468,
      "grad_norm": 0.00010841305629583076,
      "learning_rate": 0.0190450880032334,
      "loss": 0.2618,
      "num_input_tokens_seen": 25561176,
      "step": 33515
    },
    {
      "epoch": 69.6881496881497,
      "grad_norm": 0.00027366101858206093,
      "learning_rate": 0.019016372462818114,
      "loss": 0.287,
      "num_input_tokens_seen": 25565112,
      "step": 33520
    },
    {
      "epoch": 69.6985446985447,
      "grad_norm": 0.000278150342637673,
      "learning_rate": 0.018987677121724278,
      "loss": 0.259,
      "num_input_tokens_seen": 25568984,
      "step": 33525
    },
    {
      "epoch": 69.70893970893971,
      "grad_norm": 0.0014317523455247283,
      "learning_rate": 0.018959001984377,
      "loss": 0.2523,
      "num_input_tokens_seen": 25572856,
      "step": 33530
    },
    {
      "epoch": 69.71933471933473,
      "grad_norm": 0.0002784867538139224,
      "learning_rate": 0.018930347055198377,
      "loss": 0.2573,
      "num_input_tokens_seen": 25576728,
      "step": 33535
    },
    {
      "epoch": 69.72972972972973,
      "grad_norm": 0.00046777664101682603,
      "learning_rate": 0.01890171233860739,
      "loss": 0.2743,
      "num_input_tokens_seen": 25580664,
      "step": 33540
    },
    {
      "epoch": 69.74012474012474,
      "grad_norm": 0.00012935900304000825,
      "learning_rate": 0.018873097839019807,
      "loss": 0.2681,
      "num_input_tokens_seen": 25584280,
      "step": 33545
    },
    {
      "epoch": 69.75051975051976,
      "grad_norm": 0.00030318216886371374,
      "learning_rate": 0.0188445035608484,
      "loss": 0.2549,
      "num_input_tokens_seen": 25588344,
      "step": 33550
    },
    {
      "epoch": 69.76091476091476,
      "grad_norm": 0.0007277113618329167,
      "learning_rate": 0.018815929508502777,
      "loss": 0.23,
      "num_input_tokens_seen": 25592088,
      "step": 33555
    },
    {
      "epoch": 69.77130977130977,
      "grad_norm": 0.0006297500804066658,
      "learning_rate": 0.01878737568638934,
      "loss": 0.2896,
      "num_input_tokens_seen": 25595992,
      "step": 33560
    },
    {
      "epoch": 69.78170478170478,
      "grad_norm": 0.00021845132869202644,
      "learning_rate": 0.01875884209891152,
      "loss": 0.2534,
      "num_input_tokens_seen": 25599864,
      "step": 33565
    },
    {
      "epoch": 69.79209979209979,
      "grad_norm": 0.0002170616644434631,
      "learning_rate": 0.018730328750469514,
      "loss": 0.2761,
      "num_input_tokens_seen": 25603704,
      "step": 33570
    },
    {
      "epoch": 69.8024948024948,
      "grad_norm": 0.00029275112319737673,
      "learning_rate": 0.018701835645460473,
      "loss": 0.2182,
      "num_input_tokens_seen": 25607480,
      "step": 33575
    },
    {
      "epoch": 69.81288981288981,
      "grad_norm": 0.00010994344484061003,
      "learning_rate": 0.01867336278827838,
      "loss": 0.2717,
      "num_input_tokens_seen": 25611288,
      "step": 33580
    },
    {
      "epoch": 69.82328482328482,
      "grad_norm": 0.0003190810966771096,
      "learning_rate": 0.018644910183314056,
      "loss": 0.2645,
      "num_input_tokens_seen": 25615032,
      "step": 33585
    },
    {
      "epoch": 69.83367983367984,
      "grad_norm": 0.00012765720020979643,
      "learning_rate": 0.01861647783495531,
      "loss": 0.2697,
      "num_input_tokens_seen": 25618904,
      "step": 33590
    },
    {
      "epoch": 69.84407484407484,
      "grad_norm": 0.00042300455970689654,
      "learning_rate": 0.01858806574758676,
      "loss": 0.2662,
      "num_input_tokens_seen": 25622776,
      "step": 33595
    },
    {
      "epoch": 69.85446985446985,
      "grad_norm": 9.982017218135297e-05,
      "learning_rate": 0.01855967392558988,
      "loss": 0.2548,
      "num_input_tokens_seen": 25626520,
      "step": 33600
    },
    {
      "epoch": 69.85446985446985,
      "eval_loss": 0.24909155070781708,
      "eval_runtime": 13.4131,
      "eval_samples_per_second": 63.818,
      "eval_steps_per_second": 15.955,
      "num_input_tokens_seen": 25626520,
      "step": 33600
    },
    {
      "epoch": 69.86486486486487,
      "grad_norm": 0.00029304978670552373,
      "learning_rate": 0.018531302373343096,
      "loss": 0.2869,
      "num_input_tokens_seen": 25630200,
      "step": 33605
    },
    {
      "epoch": 69.87525987525987,
      "grad_norm": 0.0001853680150816217,
      "learning_rate": 0.018502951095221588,
      "loss": 0.2572,
      "num_input_tokens_seen": 25633784,
      "step": 33610
    },
    {
      "epoch": 69.88565488565489,
      "grad_norm": 0.00012747083383146673,
      "learning_rate": 0.01847462009559751,
      "loss": 0.2685,
      "num_input_tokens_seen": 25637592,
      "step": 33615
    },
    {
      "epoch": 69.8960498960499,
      "grad_norm": 0.00031032529659569263,
      "learning_rate": 0.01844630937883992,
      "loss": 0.2615,
      "num_input_tokens_seen": 25641592,
      "step": 33620
    },
    {
      "epoch": 69.9064449064449,
      "grad_norm": 7.44626740925014e-05,
      "learning_rate": 0.018418018949314573,
      "loss": 0.2613,
      "num_input_tokens_seen": 25645272,
      "step": 33625
    },
    {
      "epoch": 69.91683991683992,
      "grad_norm": 9.610377310309559e-05,
      "learning_rate": 0.018389748811384315,
      "loss": 0.2417,
      "num_input_tokens_seen": 25649048,
      "step": 33630
    },
    {
      "epoch": 69.92723492723492,
      "grad_norm": 0.00020023660908918828,
      "learning_rate": 0.018361498969408658,
      "loss": 0.2639,
      "num_input_tokens_seen": 25652792,
      "step": 33635
    },
    {
      "epoch": 69.93762993762994,
      "grad_norm": 6.828092591604218e-05,
      "learning_rate": 0.01833326942774415,
      "loss": 0.2472,
      "num_input_tokens_seen": 25656632,
      "step": 33640
    },
    {
      "epoch": 69.94802494802495,
      "grad_norm": 0.00037896123831160367,
      "learning_rate": 0.018305060190744155,
      "loss": 0.2611,
      "num_input_tokens_seen": 25660376,
      "step": 33645
    },
    {
      "epoch": 69.95841995841995,
      "grad_norm": 0.00042398928781040013,
      "learning_rate": 0.018276871262758846,
      "loss": 0.2925,
      "num_input_tokens_seen": 25664216,
      "step": 33650
    },
    {
      "epoch": 69.96881496881497,
      "grad_norm": 0.00022856802388560027,
      "learning_rate": 0.0182487026481353,
      "loss": 0.2711,
      "num_input_tokens_seen": 25668152,
      "step": 33655
    },
    {
      "epoch": 69.97920997920998,
      "grad_norm": 7.218356768134981e-05,
      "learning_rate": 0.018220554351217538,
      "loss": 0.2519,
      "num_input_tokens_seen": 25671896,
      "step": 33660
    },
    {
      "epoch": 69.98960498960498,
      "grad_norm": 0.00014351400022860616,
      "learning_rate": 0.01819242637634629,
      "loss": 0.272,
      "num_input_tokens_seen": 25675672,
      "step": 33665
    },
    {
      "epoch": 70.0,
      "grad_norm": 0.00023627256450708956,
      "learning_rate": 0.01816431872785933,
      "loss": 0.2559,
      "num_input_tokens_seen": 25679336,
      "step": 33670
    },
    {
      "epoch": 70.01039501039502,
      "grad_norm": 0.0003869260544888675,
      "learning_rate": 0.018136231410091148,
      "loss": 0.2525,
      "num_input_tokens_seen": 25682984,
      "step": 33675
    },
    {
      "epoch": 70.02079002079002,
      "grad_norm": 0.00021054453100077808,
      "learning_rate": 0.018108164427373175,
      "loss": 0.2761,
      "num_input_tokens_seen": 25686824,
      "step": 33680
    },
    {
      "epoch": 70.03118503118503,
      "grad_norm": 0.00043784594163298607,
      "learning_rate": 0.01808011778403375,
      "loss": 0.2821,
      "num_input_tokens_seen": 25690760,
      "step": 33685
    },
    {
      "epoch": 70.04158004158005,
      "grad_norm": 0.002799914451315999,
      "learning_rate": 0.01805209148439793,
      "loss": 0.2769,
      "num_input_tokens_seen": 25694536,
      "step": 33690
    },
    {
      "epoch": 70.05197505197505,
      "grad_norm": 0.0004847442323807627,
      "learning_rate": 0.018024085532787757,
      "loss": 0.2585,
      "num_input_tokens_seen": 25698248,
      "step": 33695
    },
    {
      "epoch": 70.06237006237006,
      "grad_norm": 0.00048818832146935165,
      "learning_rate": 0.017996099933522164,
      "loss": 0.2619,
      "num_input_tokens_seen": 25701992,
      "step": 33700
    },
    {
      "epoch": 70.07276507276508,
      "grad_norm": 0.00061624008230865,
      "learning_rate": 0.017968134690916775,
      "loss": 0.2523,
      "num_input_tokens_seen": 25705896,
      "step": 33705
    },
    {
      "epoch": 70.08316008316008,
      "grad_norm": 9.043209865922108e-05,
      "learning_rate": 0.017940189809284263,
      "loss": 0.2424,
      "num_input_tokens_seen": 25709512,
      "step": 33710
    },
    {
      "epoch": 70.0935550935551,
      "grad_norm": 0.0007165076676756144,
      "learning_rate": 0.017912265292934024,
      "loss": 0.2675,
      "num_input_tokens_seen": 25713512,
      "step": 33715
    },
    {
      "epoch": 70.1039501039501,
      "grad_norm": 0.00010592116450425237,
      "learning_rate": 0.017884361146172423,
      "loss": 0.2537,
      "num_input_tokens_seen": 25717256,
      "step": 33720
    },
    {
      "epoch": 70.11434511434511,
      "grad_norm": 0.0005999957793392241,
      "learning_rate": 0.01785647737330261,
      "loss": 0.2705,
      "num_input_tokens_seen": 25721256,
      "step": 33725
    },
    {
      "epoch": 70.12474012474013,
      "grad_norm": 0.00021102858590893447,
      "learning_rate": 0.017828613978624563,
      "loss": 0.2554,
      "num_input_tokens_seen": 25724936,
      "step": 33730
    },
    {
      "epoch": 70.13513513513513,
      "grad_norm": 0.00010630543692968786,
      "learning_rate": 0.01780077096643523,
      "loss": 0.2591,
      "num_input_tokens_seen": 25728680,
      "step": 33735
    },
    {
      "epoch": 70.14553014553015,
      "grad_norm": 0.0003579895419534296,
      "learning_rate": 0.017772948341028345,
      "loss": 0.2687,
      "num_input_tokens_seen": 25732360,
      "step": 33740
    },
    {
      "epoch": 70.15592515592516,
      "grad_norm": 0.00012928416253998876,
      "learning_rate": 0.01774514610669447,
      "loss": 0.2752,
      "num_input_tokens_seen": 25736040,
      "step": 33745
    },
    {
      "epoch": 70.16632016632016,
      "grad_norm": 0.00032673278474248946,
      "learning_rate": 0.017717364267721112,
      "loss": 0.2497,
      "num_input_tokens_seen": 25739848,
      "step": 33750
    },
    {
      "epoch": 70.17671517671518,
      "grad_norm": 0.00011024600826203823,
      "learning_rate": 0.017689602828392513,
      "loss": 0.2418,
      "num_input_tokens_seen": 25743688,
      "step": 33755
    },
    {
      "epoch": 70.18711018711019,
      "grad_norm": 0.0003032666281796992,
      "learning_rate": 0.017661861792989897,
      "loss": 0.2616,
      "num_input_tokens_seen": 25747688,
      "step": 33760
    },
    {
      "epoch": 70.1975051975052,
      "grad_norm": 0.0001534395560156554,
      "learning_rate": 0.017634141165791272,
      "loss": 0.2664,
      "num_input_tokens_seen": 25751496,
      "step": 33765
    },
    {
      "epoch": 70.20790020790021,
      "grad_norm": 0.0003258484648540616,
      "learning_rate": 0.017606440951071455,
      "loss": 0.2597,
      "num_input_tokens_seen": 25755368,
      "step": 33770
    },
    {
      "epoch": 70.21829521829522,
      "grad_norm": 0.00011250966781517491,
      "learning_rate": 0.017578761153102213,
      "loss": 0.2546,
      "num_input_tokens_seen": 25759080,
      "step": 33775
    },
    {
      "epoch": 70.22869022869023,
      "grad_norm": 0.0002449020103085786,
      "learning_rate": 0.017551101776152146,
      "loss": 0.2516,
      "num_input_tokens_seen": 25763080,
      "step": 33780
    },
    {
      "epoch": 70.23908523908524,
      "grad_norm": 0.00046255573397502303,
      "learning_rate": 0.017523462824486608,
      "loss": 0.2833,
      "num_input_tokens_seen": 25766920,
      "step": 33785
    },
    {
      "epoch": 70.24948024948024,
      "grad_norm": 0.0005931609193794429,
      "learning_rate": 0.01749584430236794,
      "loss": 0.2769,
      "num_input_tokens_seen": 25770792,
      "step": 33790
    },
    {
      "epoch": 70.25987525987526,
      "grad_norm": 0.0002591744705568999,
      "learning_rate": 0.01746824621405524,
      "loss": 0.2603,
      "num_input_tokens_seen": 25774632,
      "step": 33795
    },
    {
      "epoch": 70.27027027027027,
      "grad_norm": 0.00041593145579099655,
      "learning_rate": 0.017440668563804412,
      "loss": 0.2589,
      "num_input_tokens_seen": 25778248,
      "step": 33800
    },
    {
      "epoch": 70.27027027027027,
      "eval_loss": 0.24918237328529358,
      "eval_runtime": 13.4135,
      "eval_samples_per_second": 63.816,
      "eval_steps_per_second": 15.954,
      "num_input_tokens_seen": 25778248,
      "step": 33800
    },
    {
      "epoch": 70.28066528066527,
      "grad_norm": 0.00015579836326651275,
      "learning_rate": 0.017413111355868392,
      "loss": 0.2559,
      "num_input_tokens_seen": 25781896,
      "step": 33805
    },
    {
      "epoch": 70.29106029106029,
      "grad_norm": 0.0006690014852210879,
      "learning_rate": 0.017385574594496748,
      "loss": 0.2567,
      "num_input_tokens_seen": 25785704,
      "step": 33810
    },
    {
      "epoch": 70.3014553014553,
      "grad_norm": 0.0003354672808200121,
      "learning_rate": 0.01735805828393605,
      "loss": 0.2599,
      "num_input_tokens_seen": 25789704,
      "step": 33815
    },
    {
      "epoch": 70.3118503118503,
      "grad_norm": 0.00013841109466739,
      "learning_rate": 0.017330562428429667,
      "loss": 0.262,
      "num_input_tokens_seen": 25793736,
      "step": 33820
    },
    {
      "epoch": 70.32224532224532,
      "grad_norm": 0.0002482893178239465,
      "learning_rate": 0.01730308703221776,
      "loss": 0.2915,
      "num_input_tokens_seen": 25797640,
      "step": 33825
    },
    {
      "epoch": 70.33264033264034,
      "grad_norm": 0.00022579891083296388,
      "learning_rate": 0.01727563209953744,
      "loss": 0.2346,
      "num_input_tokens_seen": 25801384,
      "step": 33830
    },
    {
      "epoch": 70.34303534303534,
      "grad_norm": 0.0003142142086289823,
      "learning_rate": 0.017248197634622535,
      "loss": 0.2554,
      "num_input_tokens_seen": 25805160,
      "step": 33835
    },
    {
      "epoch": 70.35343035343035,
      "grad_norm": 0.00010582814138615504,
      "learning_rate": 0.01722078364170383,
      "loss": 0.2498,
      "num_input_tokens_seen": 25809032,
      "step": 33840
    },
    {
      "epoch": 70.36382536382537,
      "grad_norm": 0.0008987306500785053,
      "learning_rate": 0.017193390125008905,
      "loss": 0.261,
      "num_input_tokens_seen": 25812936,
      "step": 33845
    },
    {
      "epoch": 70.37422037422037,
      "grad_norm": 0.00034803838934749365,
      "learning_rate": 0.017166017088762153,
      "loss": 0.2736,
      "num_input_tokens_seen": 25816744,
      "step": 33850
    },
    {
      "epoch": 70.38461538461539,
      "grad_norm": 0.0008792418520897627,
      "learning_rate": 0.017138664537184878,
      "loss": 0.241,
      "num_input_tokens_seen": 25820616,
      "step": 33855
    },
    {
      "epoch": 70.39501039501039,
      "grad_norm": 0.00012005298776784912,
      "learning_rate": 0.017111332474495172,
      "loss": 0.2562,
      "num_input_tokens_seen": 25824424,
      "step": 33860
    },
    {
      "epoch": 70.4054054054054,
      "grad_norm": 0.00014786144311074167,
      "learning_rate": 0.017084020904907998,
      "loss": 0.2572,
      "num_input_tokens_seen": 25828072,
      "step": 33865
    },
    {
      "epoch": 70.41580041580042,
      "grad_norm": 0.00023796291498001665,
      "learning_rate": 0.017056729832635103,
      "loss": 0.2549,
      "num_input_tokens_seen": 25831784,
      "step": 33870
    },
    {
      "epoch": 70.42619542619542,
      "grad_norm": 0.0001081714071915485,
      "learning_rate": 0.017029459261885153,
      "loss": 0.2572,
      "num_input_tokens_seen": 25835592,
      "step": 33875
    },
    {
      "epoch": 70.43659043659044,
      "grad_norm": 0.00011292738054180518,
      "learning_rate": 0.01700220919686359,
      "loss": 0.2734,
      "num_input_tokens_seen": 25839528,
      "step": 33880
    },
    {
      "epoch": 70.44698544698545,
      "grad_norm": 0.0007623339770361781,
      "learning_rate": 0.016974979641772723,
      "loss": 0.2514,
      "num_input_tokens_seen": 25843432,
      "step": 33885
    },
    {
      "epoch": 70.45738045738045,
      "grad_norm": 0.001597981434315443,
      "learning_rate": 0.01694777060081169,
      "loss": 0.2723,
      "num_input_tokens_seen": 25847240,
      "step": 33890
    },
    {
      "epoch": 70.46777546777547,
      "grad_norm": 0.00038543209666386247,
      "learning_rate": 0.016920582078176444,
      "loss": 0.276,
      "num_input_tokens_seen": 25851016,
      "step": 33895
    },
    {
      "epoch": 70.47817047817048,
      "grad_norm": 0.0004299981810618192,
      "learning_rate": 0.016893414078059863,
      "loss": 0.2549,
      "num_input_tokens_seen": 25854792,
      "step": 33900
    },
    {
      "epoch": 70.48856548856548,
      "grad_norm": 0.0002463023702148348,
      "learning_rate": 0.016866266604651535,
      "loss": 0.2595,
      "num_input_tokens_seen": 25858440,
      "step": 33905
    },
    {
      "epoch": 70.4989604989605,
      "grad_norm": 0.0009466482442803681,
      "learning_rate": 0.016839139662137976,
      "loss": 0.2441,
      "num_input_tokens_seen": 25862280,
      "step": 33910
    },
    {
      "epoch": 70.50935550935552,
      "grad_norm": 0.00028445583302527666,
      "learning_rate": 0.01681203325470245,
      "loss": 0.2554,
      "num_input_tokens_seen": 25865960,
      "step": 33915
    },
    {
      "epoch": 70.51975051975052,
      "grad_norm": 0.0005626974161714315,
      "learning_rate": 0.016784947386525157,
      "loss": 0.2453,
      "num_input_tokens_seen": 25869672,
      "step": 33920
    },
    {
      "epoch": 70.53014553014553,
      "grad_norm": 0.0007579077500849962,
      "learning_rate": 0.01675788206178308,
      "loss": 0.2524,
      "num_input_tokens_seen": 25873448,
      "step": 33925
    },
    {
      "epoch": 70.54054054054055,
      "grad_norm": 9.533877164358273e-05,
      "learning_rate": 0.016730837284649986,
      "loss": 0.2904,
      "num_input_tokens_seen": 25877544,
      "step": 33930
    },
    {
      "epoch": 70.55093555093555,
      "grad_norm": 0.00034102812060154974,
      "learning_rate": 0.016703813059296583,
      "loss": 0.2602,
      "num_input_tokens_seen": 25881480,
      "step": 33935
    },
    {
      "epoch": 70.56133056133056,
      "grad_norm": 0.0007499619387090206,
      "learning_rate": 0.016676809389890294,
      "loss": 0.2399,
      "num_input_tokens_seen": 25885288,
      "step": 33940
    },
    {
      "epoch": 70.57172557172557,
      "grad_norm": 0.00011411232117097825,
      "learning_rate": 0.016649826280595435,
      "loss": 0.2574,
      "num_input_tokens_seen": 25889256,
      "step": 33945
    },
    {
      "epoch": 70.58212058212058,
      "grad_norm": 0.00017075546202249825,
      "learning_rate": 0.016622863735573163,
      "loss": 0.263,
      "num_input_tokens_seen": 25893096,
      "step": 33950
    },
    {
      "epoch": 70.5925155925156,
      "grad_norm": 0.0003233301977161318,
      "learning_rate": 0.016595921758981395,
      "loss": 0.2813,
      "num_input_tokens_seen": 25896936,
      "step": 33955
    },
    {
      "epoch": 70.6029106029106,
      "grad_norm": 0.00035117159131914377,
      "learning_rate": 0.01656900035497495,
      "loss": 0.2599,
      "num_input_tokens_seen": 25900776,
      "step": 33960
    },
    {
      "epoch": 70.61330561330561,
      "grad_norm": 0.0005780687788501382,
      "learning_rate": 0.016542099527705485,
      "loss": 0.2488,
      "num_input_tokens_seen": 25904424,
      "step": 33965
    },
    {
      "epoch": 70.62370062370063,
      "grad_norm": 0.00015171048289630562,
      "learning_rate": 0.01651521928132138,
      "loss": 0.2715,
      "num_input_tokens_seen": 25908360,
      "step": 33970
    },
    {
      "epoch": 70.63409563409563,
      "grad_norm": 0.0002794294850900769,
      "learning_rate": 0.01648835961996794,
      "loss": 0.2355,
      "num_input_tokens_seen": 25912072,
      "step": 33975
    },
    {
      "epoch": 70.64449064449065,
      "grad_norm": 0.00021177640883252025,
      "learning_rate": 0.016461520547787285,
      "loss": 0.2598,
      "num_input_tokens_seen": 25915784,
      "step": 33980
    },
    {
      "epoch": 70.65488565488566,
      "grad_norm": 0.0006138205644674599,
      "learning_rate": 0.016434702068918266,
      "loss": 0.2839,
      "num_input_tokens_seen": 25919592,
      "step": 33985
    },
    {
      "epoch": 70.66528066528066,
      "grad_norm": 0.0007318980642594397,
      "learning_rate": 0.01640790418749673,
      "loss": 0.2789,
      "num_input_tokens_seen": 25923400,
      "step": 33990
    },
    {
      "epoch": 70.67567567567568,
      "grad_norm": 4.5394597691483796e-05,
      "learning_rate": 0.016381126907655134,
      "loss": 0.2699,
      "num_input_tokens_seen": 25927112,
      "step": 33995
    },
    {
      "epoch": 70.68607068607069,
      "grad_norm": 0.00014760946214664727,
      "learning_rate": 0.016354370233522948,
      "loss": 0.2534,
      "num_input_tokens_seen": 25930920,
      "step": 34000
    },
    {
      "epoch": 70.68607068607069,
      "eval_loss": 0.25100159645080566,
      "eval_runtime": 13.3894,
      "eval_samples_per_second": 63.931,
      "eval_steps_per_second": 15.983,
      "num_input_tokens_seen": 25930920,
      "step": 34000
    },
    {
      "epoch": 70.6964656964657,
      "grad_norm": 0.00033858089591376483,
      "learning_rate": 0.016327634169226394,
      "loss": 0.2876,
      "num_input_tokens_seen": 25934856,
      "step": 34005
    },
    {
      "epoch": 70.70686070686071,
      "grad_norm": 0.00034338614204898477,
      "learning_rate": 0.016300918718888485,
      "loss": 0.2696,
      "num_input_tokens_seen": 25938504,
      "step": 34010
    },
    {
      "epoch": 70.71725571725571,
      "grad_norm": 0.00013957686314824969,
      "learning_rate": 0.016274223886629052,
      "loss": 0.2706,
      "num_input_tokens_seen": 25942216,
      "step": 34015
    },
    {
      "epoch": 70.72765072765073,
      "grad_norm": 0.00036131704109720886,
      "learning_rate": 0.01624754967656482,
      "loss": 0.2809,
      "num_input_tokens_seen": 25945864,
      "step": 34020
    },
    {
      "epoch": 70.73804573804574,
      "grad_norm": 0.0002578433195594698,
      "learning_rate": 0.016220896092809235,
      "loss": 0.2777,
      "num_input_tokens_seen": 25949864,
      "step": 34025
    },
    {
      "epoch": 70.74844074844074,
      "grad_norm": 0.00037835861439816654,
      "learning_rate": 0.01619426313947267,
      "loss": 0.2586,
      "num_input_tokens_seen": 25953832,
      "step": 34030
    },
    {
      "epoch": 70.75883575883576,
      "grad_norm": 0.0002139493590220809,
      "learning_rate": 0.016167650820662228,
      "loss": 0.2615,
      "num_input_tokens_seen": 25957576,
      "step": 34035
    },
    {
      "epoch": 70.76923076923077,
      "grad_norm": 0.00018161052139475942,
      "learning_rate": 0.016141059140481855,
      "loss": 0.2618,
      "num_input_tokens_seen": 25961384,
      "step": 34040
    },
    {
      "epoch": 70.77962577962577,
      "grad_norm": 0.0018129326635971665,
      "learning_rate": 0.016114488103032374,
      "loss": 0.2947,
      "num_input_tokens_seen": 25965224,
      "step": 34045
    },
    {
      "epoch": 70.79002079002079,
      "grad_norm": 0.0005014881025999784,
      "learning_rate": 0.016087937712411293,
      "loss": 0.2835,
      "num_input_tokens_seen": 25969032,
      "step": 34050
    },
    {
      "epoch": 70.8004158004158,
      "grad_norm": 0.0007075099274516106,
      "learning_rate": 0.01606140797271308,
      "loss": 0.2711,
      "num_input_tokens_seen": 25972712,
      "step": 34055
    },
    {
      "epoch": 70.8108108108108,
      "grad_norm": 0.00033514745882712305,
      "learning_rate": 0.01603489888802897,
      "loss": 0.25,
      "num_input_tokens_seen": 25976456,
      "step": 34060
    },
    {
      "epoch": 70.82120582120582,
      "grad_norm": 0.00043173637823201716,
      "learning_rate": 0.016008410462446918,
      "loss": 0.2652,
      "num_input_tokens_seen": 25980328,
      "step": 34065
    },
    {
      "epoch": 70.83160083160084,
      "grad_norm": 0.0003789405745919794,
      "learning_rate": 0.01598194270005185,
      "loss": 0.2515,
      "num_input_tokens_seen": 25984072,
      "step": 34070
    },
    {
      "epoch": 70.84199584199584,
      "grad_norm": 0.00015172957500908524,
      "learning_rate": 0.015955495604925356,
      "loss": 0.2641,
      "num_input_tokens_seen": 25987880,
      "step": 34075
    },
    {
      "epoch": 70.85239085239085,
      "grad_norm": 0.00028241268591955304,
      "learning_rate": 0.01592906918114598,
      "loss": 0.265,
      "num_input_tokens_seen": 25991656,
      "step": 34080
    },
    {
      "epoch": 70.86278586278586,
      "grad_norm": 0.00026847838307730854,
      "learning_rate": 0.015902663432788965,
      "loss": 0.26,
      "num_input_tokens_seen": 25995496,
      "step": 34085
    },
    {
      "epoch": 70.87318087318087,
      "grad_norm": 0.0004988065920770168,
      "learning_rate": 0.01587627836392643,
      "loss": 0.2561,
      "num_input_tokens_seen": 25999400,
      "step": 34090
    },
    {
      "epoch": 70.88357588357589,
      "grad_norm": 0.0006295868661254644,
      "learning_rate": 0.01584991397862726,
      "loss": 0.2868,
      "num_input_tokens_seen": 26003240,
      "step": 34095
    },
    {
      "epoch": 70.89397089397089,
      "grad_norm": 0.0002003811241593212,
      "learning_rate": 0.015823570280957214,
      "loss": 0.2516,
      "num_input_tokens_seen": 26007208,
      "step": 34100
    },
    {
      "epoch": 70.9043659043659,
      "grad_norm": 0.0008161681471392512,
      "learning_rate": 0.015797247274978766,
      "loss": 0.2542,
      "num_input_tokens_seen": 26011112,
      "step": 34105
    },
    {
      "epoch": 70.91476091476092,
      "grad_norm": 0.00030335740302689373,
      "learning_rate": 0.015770944964751326,
      "loss": 0.2779,
      "num_input_tokens_seen": 26015048,
      "step": 34110
    },
    {
      "epoch": 70.92515592515592,
      "grad_norm": 0.00048233720008283854,
      "learning_rate": 0.015744663354330956,
      "loss": 0.2561,
      "num_input_tokens_seen": 26018920,
      "step": 34115
    },
    {
      "epoch": 70.93555093555094,
      "grad_norm": 0.00033505153260193765,
      "learning_rate": 0.015718402447770664,
      "loss": 0.2637,
      "num_input_tokens_seen": 26022728,
      "step": 34120
    },
    {
      "epoch": 70.94594594594595,
      "grad_norm": 0.000287884206045419,
      "learning_rate": 0.015692162249120224,
      "loss": 0.2479,
      "num_input_tokens_seen": 26026504,
      "step": 34125
    },
    {
      "epoch": 70.95634095634095,
      "grad_norm": 0.0008753057918511331,
      "learning_rate": 0.01566594276242615,
      "loss": 0.2744,
      "num_input_tokens_seen": 26030280,
      "step": 34130
    },
    {
      "epoch": 70.96673596673597,
      "grad_norm": 0.0003068975347559899,
      "learning_rate": 0.015639743991731857,
      "loss": 0.2639,
      "num_input_tokens_seen": 26034088,
      "step": 34135
    },
    {
      "epoch": 70.97713097713098,
      "grad_norm": 0.0009717169450595975,
      "learning_rate": 0.01561356594107755,
      "loss": 0.2511,
      "num_input_tokens_seen": 26037704,
      "step": 34140
    },
    {
      "epoch": 70.98752598752598,
      "grad_norm": 0.0003354522632434964,
      "learning_rate": 0.015587408614500147,
      "loss": 0.2676,
      "num_input_tokens_seen": 26041384,
      "step": 34145
    },
    {
      "epoch": 70.997920997921,
      "grad_norm": 0.0006731243338435888,
      "learning_rate": 0.015561272016033505,
      "loss": 0.281,
      "num_input_tokens_seen": 26045256,
      "step": 34150
    },
    {
      "epoch": 71.00831600831602,
      "grad_norm": 0.00039801179082132876,
      "learning_rate": 0.015535156149708167,
      "loss": 0.2796,
      "num_input_tokens_seen": 26049024,
      "step": 34155
    },
    {
      "epoch": 71.01871101871102,
      "grad_norm": 0.0002682908670976758,
      "learning_rate": 0.015509061019551528,
      "loss": 0.2726,
      "num_input_tokens_seen": 26052832,
      "step": 34160
    },
    {
      "epoch": 71.02910602910603,
      "grad_norm": 0.00048176926793530583,
      "learning_rate": 0.015482986629587818,
      "loss": 0.2483,
      "num_input_tokens_seen": 26056640,
      "step": 34165
    },
    {
      "epoch": 71.03950103950103,
      "grad_norm": 0.0019338406855240464,
      "learning_rate": 0.01545693298383799,
      "loss": 0.2396,
      "num_input_tokens_seen": 26060448,
      "step": 34170
    },
    {
      "epoch": 71.04989604989605,
      "grad_norm": 0.0019806839991360903,
      "learning_rate": 0.015430900086319858,
      "loss": 0.2481,
      "num_input_tokens_seen": 26064160,
      "step": 34175
    },
    {
      "epoch": 71.06029106029106,
      "grad_norm": 0.00238628638908267,
      "learning_rate": 0.015404887941048084,
      "loss": 0.283,
      "num_input_tokens_seen": 26068032,
      "step": 34180
    },
    {
      "epoch": 71.07068607068607,
      "grad_norm": 0.00018494039250072092,
      "learning_rate": 0.01537889655203397,
      "loss": 0.2604,
      "num_input_tokens_seen": 26071840,
      "step": 34185
    },
    {
      "epoch": 71.08108108108108,
      "grad_norm": 0.00013956715702079237,
      "learning_rate": 0.015352925923285798,
      "loss": 0.2637,
      "num_input_tokens_seen": 26075680,
      "step": 34190
    },
    {
      "epoch": 71.0914760914761,
      "grad_norm": 0.0005342590739019215,
      "learning_rate": 0.015326976058808511,
      "loss": 0.2686,
      "num_input_tokens_seen": 26079584,
      "step": 34195
    },
    {
      "epoch": 71.1018711018711,
      "grad_norm": 0.00026903097750619054,
      "learning_rate": 0.015301046962603908,
      "loss": 0.2532,
      "num_input_tokens_seen": 26083456,
      "step": 34200
    },
    {
      "epoch": 71.1018711018711,
      "eval_loss": 0.2491416186094284,
      "eval_runtime": 13.4224,
      "eval_samples_per_second": 63.774,
      "eval_steps_per_second": 15.943,
      "num_input_tokens_seen": 26083456,
      "step": 34200
    },
    {
      "epoch": 71.11226611226611,
      "grad_norm": 0.00017995413509197533,
      "learning_rate": 0.015275138638670626,
      "loss": 0.2583,
      "num_input_tokens_seen": 26087200,
      "step": 34205
    },
    {
      "epoch": 71.12266112266113,
      "grad_norm": 0.0012385392328724265,
      "learning_rate": 0.015249251091004001,
      "loss": 0.2664,
      "num_input_tokens_seen": 26090944,
      "step": 34210
    },
    {
      "epoch": 71.13305613305613,
      "grad_norm": 0.00025588812422938645,
      "learning_rate": 0.01522338432359624,
      "loss": 0.2438,
      "num_input_tokens_seen": 26094592,
      "step": 34215
    },
    {
      "epoch": 71.14345114345114,
      "grad_norm": 4.990608795196749e-05,
      "learning_rate": 0.01519753834043635,
      "loss": 0.2677,
      "num_input_tokens_seen": 26098400,
      "step": 34220
    },
    {
      "epoch": 71.15384615384616,
      "grad_norm": 5.714313738280907e-05,
      "learning_rate": 0.015171713145510095,
      "loss": 0.2545,
      "num_input_tokens_seen": 26102368,
      "step": 34225
    },
    {
      "epoch": 71.16424116424116,
      "grad_norm": 0.0004953845636919141,
      "learning_rate": 0.01514590874279999,
      "loss": 0.265,
      "num_input_tokens_seen": 26106080,
      "step": 34230
    },
    {
      "epoch": 71.17463617463618,
      "grad_norm": 0.0009671784937381744,
      "learning_rate": 0.015120125136285467,
      "loss": 0.2871,
      "num_input_tokens_seen": 26109792,
      "step": 34235
    },
    {
      "epoch": 71.18503118503118,
      "grad_norm": 0.001206830726005137,
      "learning_rate": 0.015094362329942629,
      "loss": 0.2504,
      "num_input_tokens_seen": 26113536,
      "step": 34240
    },
    {
      "epoch": 71.1954261954262,
      "grad_norm": 0.0017787389224395156,
      "learning_rate": 0.01506862032774448,
      "loss": 0.275,
      "num_input_tokens_seen": 26117312,
      "step": 34245
    },
    {
      "epoch": 71.20582120582121,
      "grad_norm": 0.00016638028318993747,
      "learning_rate": 0.015042899133660697,
      "loss": 0.2516,
      "num_input_tokens_seen": 26121184,
      "step": 34250
    },
    {
      "epoch": 71.21621621621621,
      "grad_norm": 0.00012802726996596903,
      "learning_rate": 0.01501719875165789,
      "loss": 0.2856,
      "num_input_tokens_seen": 26124864,
      "step": 34255
    },
    {
      "epoch": 71.22661122661123,
      "grad_norm": 0.00014675762213300914,
      "learning_rate": 0.014991519185699286,
      "loss": 0.2743,
      "num_input_tokens_seen": 26128640,
      "step": 34260
    },
    {
      "epoch": 71.23700623700624,
      "grad_norm": 0.000537066429387778,
      "learning_rate": 0.014965860439745054,
      "loss": 0.2883,
      "num_input_tokens_seen": 26132544,
      "step": 34265
    },
    {
      "epoch": 71.24740124740124,
      "grad_norm": 0.0006858361884951591,
      "learning_rate": 0.01494022251775211,
      "loss": 0.2607,
      "num_input_tokens_seen": 26136288,
      "step": 34270
    },
    {
      "epoch": 71.25779625779626,
      "grad_norm": 0.000666318868752569,
      "learning_rate": 0.014914605423674109,
      "loss": 0.2555,
      "num_input_tokens_seen": 26140160,
      "step": 34275
    },
    {
      "epoch": 71.26819126819127,
      "grad_norm": 0.00018950880621559918,
      "learning_rate": 0.014889009161461525,
      "loss": 0.2612,
      "num_input_tokens_seen": 26144064,
      "step": 34280
    },
    {
      "epoch": 71.27858627858627,
      "grad_norm": 0.00047989084850996733,
      "learning_rate": 0.014863433735061665,
      "loss": 0.266,
      "num_input_tokens_seen": 26147872,
      "step": 34285
    },
    {
      "epoch": 71.28898128898129,
      "grad_norm": 0.00037910614628344774,
      "learning_rate": 0.014837879148418541,
      "loss": 0.2616,
      "num_input_tokens_seen": 26151744,
      "step": 34290
    },
    {
      "epoch": 71.2993762993763,
      "grad_norm": 0.00048052528291009367,
      "learning_rate": 0.01481234540547302,
      "loss": 0.2651,
      "num_input_tokens_seen": 26155616,
      "step": 34295
    },
    {
      "epoch": 71.3097713097713,
      "grad_norm": 0.0019880712497979403,
      "learning_rate": 0.014786832510162717,
      "loss": 0.2571,
      "num_input_tokens_seen": 26159424,
      "step": 34300
    },
    {
      "epoch": 71.32016632016632,
      "grad_norm": 0.00032300769817084074,
      "learning_rate": 0.014761340466422017,
      "loss": 0.2762,
      "num_input_tokens_seen": 26163296,
      "step": 34305
    },
    {
      "epoch": 71.33056133056132,
      "grad_norm": 0.0004088038404006511,
      "learning_rate": 0.014735869278182144,
      "loss": 0.2582,
      "num_input_tokens_seen": 26167104,
      "step": 34310
    },
    {
      "epoch": 71.34095634095634,
      "grad_norm": 0.00038851110730320215,
      "learning_rate": 0.014710418949371057,
      "loss": 0.2447,
      "num_input_tokens_seen": 26170816,
      "step": 34315
    },
    {
      "epoch": 71.35135135135135,
      "grad_norm": 0.0005060535622760653,
      "learning_rate": 0.014684989483913495,
      "loss": 0.2586,
      "num_input_tokens_seen": 26174656,
      "step": 34320
    },
    {
      "epoch": 71.36174636174636,
      "grad_norm": 0.0005553790833801031,
      "learning_rate": 0.014659580885731077,
      "loss": 0.2635,
      "num_input_tokens_seen": 26178624,
      "step": 34325
    },
    {
      "epoch": 71.37214137214137,
      "grad_norm": 0.0002516135573387146,
      "learning_rate": 0.014634193158742047,
      "loss": 0.2434,
      "num_input_tokens_seen": 26182400,
      "step": 34330
    },
    {
      "epoch": 71.38253638253639,
      "grad_norm": 0.00018362715491093695,
      "learning_rate": 0.014608826306861576,
      "loss": 0.2539,
      "num_input_tokens_seen": 26186240,
      "step": 34335
    },
    {
      "epoch": 71.39293139293139,
      "grad_norm": 0.0002084135339828208,
      "learning_rate": 0.014583480334001486,
      "loss": 0.2357,
      "num_input_tokens_seen": 26190016,
      "step": 34340
    },
    {
      "epoch": 71.4033264033264,
      "grad_norm": 0.0007845240761525929,
      "learning_rate": 0.014558155244070496,
      "loss": 0.2653,
      "num_input_tokens_seen": 26193792,
      "step": 34345
    },
    {
      "epoch": 71.41372141372142,
      "grad_norm": 0.00022965479001868516,
      "learning_rate": 0.014532851040974036,
      "loss": 0.2766,
      "num_input_tokens_seen": 26197600,
      "step": 34350
    },
    {
      "epoch": 71.42411642411642,
      "grad_norm": 0.00018276744231116027,
      "learning_rate": 0.014507567728614335,
      "loss": 0.2588,
      "num_input_tokens_seen": 26201248,
      "step": 34355
    },
    {
      "epoch": 71.43451143451144,
      "grad_norm": 0.00124224997125566,
      "learning_rate": 0.01448230531089037,
      "loss": 0.2618,
      "num_input_tokens_seen": 26205120,
      "step": 34360
    },
    {
      "epoch": 71.44490644490645,
      "grad_norm": 0.00011332490248605609,
      "learning_rate": 0.014457063791697993,
      "loss": 0.2733,
      "num_input_tokens_seen": 26209024,
      "step": 34365
    },
    {
      "epoch": 71.45530145530145,
      "grad_norm": 0.0004901313222944736,
      "learning_rate": 0.01443184317492971,
      "loss": 0.2671,
      "num_input_tokens_seen": 26212960,
      "step": 34370
    },
    {
      "epoch": 71.46569646569647,
      "grad_norm": 0.00019822089234367013,
      "learning_rate": 0.014406643464474822,
      "loss": 0.257,
      "num_input_tokens_seen": 26216576,
      "step": 34375
    },
    {
      "epoch": 71.47609147609148,
      "grad_norm": 0.0012744978303089738,
      "learning_rate": 0.014381464664219539,
      "loss": 0.2948,
      "num_input_tokens_seen": 26220448,
      "step": 34380
    },
    {
      "epoch": 71.48648648648648,
      "grad_norm": 0.0001673355873208493,
      "learning_rate": 0.014356306778046656,
      "loss": 0.2727,
      "num_input_tokens_seen": 26224256,
      "step": 34385
    },
    {
      "epoch": 71.4968814968815,
      "grad_norm": 0.00024221985950134695,
      "learning_rate": 0.014331169809835885,
      "loss": 0.2579,
      "num_input_tokens_seen": 26228064,
      "step": 34390
    },
    {
      "epoch": 71.5072765072765,
      "grad_norm": 0.000330470415065065,
      "learning_rate": 0.014306053763463644,
      "loss": 0.251,
      "num_input_tokens_seen": 26231776,
      "step": 34395
    },
    {
      "epoch": 71.51767151767152,
      "grad_norm": 0.000543308793567121,
      "learning_rate": 0.014280958642803147,
      "loss": 0.266,
      "num_input_tokens_seen": 26235552,
      "step": 34400
    },
    {
      "epoch": 71.51767151767152,
      "eval_loss": 0.2503880560398102,
      "eval_runtime": 13.3923,
      "eval_samples_per_second": 63.918,
      "eval_steps_per_second": 15.979,
      "num_input_tokens_seen": 26235552,
      "step": 34400
    },
    {
      "epoch": 71.52806652806653,
      "grad_norm": 0.00027296121697872877,
      "learning_rate": 0.014255884451724404,
      "loss": 0.2518,
      "num_input_tokens_seen": 26239424,
      "step": 34405
    },
    {
      "epoch": 71.53846153846153,
      "grad_norm": 0.00045671811676584184,
      "learning_rate": 0.014230831194094101,
      "loss": 0.2564,
      "num_input_tokens_seen": 26243200,
      "step": 34410
    },
    {
      "epoch": 71.54885654885655,
      "grad_norm": 0.00017378877964802086,
      "learning_rate": 0.014205798873775865,
      "loss": 0.262,
      "num_input_tokens_seen": 26247040,
      "step": 34415
    },
    {
      "epoch": 71.55925155925156,
      "grad_norm": 9.670834697317332e-05,
      "learning_rate": 0.014180787494629893,
      "loss": 0.282,
      "num_input_tokens_seen": 26250720,
      "step": 34420
    },
    {
      "epoch": 71.56964656964657,
      "grad_norm": 7.944850949570537e-05,
      "learning_rate": 0.014155797060513314,
      "loss": 0.2491,
      "num_input_tokens_seen": 26254432,
      "step": 34425
    },
    {
      "epoch": 71.58004158004158,
      "grad_norm": 0.0001057119297911413,
      "learning_rate": 0.014130827575279963,
      "loss": 0.2513,
      "num_input_tokens_seen": 26258304,
      "step": 34430
    },
    {
      "epoch": 71.5904365904366,
      "grad_norm": 0.0004385255742818117,
      "learning_rate": 0.014105879042780427,
      "loss": 0.2339,
      "num_input_tokens_seen": 26262144,
      "step": 34435
    },
    {
      "epoch": 71.6008316008316,
      "grad_norm": 0.0001551207824377343,
      "learning_rate": 0.014080951466862113,
      "loss": 0.2548,
      "num_input_tokens_seen": 26265888,
      "step": 34440
    },
    {
      "epoch": 71.61122661122661,
      "grad_norm": 0.000586564710829407,
      "learning_rate": 0.014056044851369126,
      "loss": 0.2613,
      "num_input_tokens_seen": 26269824,
      "step": 34445
    },
    {
      "epoch": 71.62162162162163,
      "grad_norm": 0.0006920543382875621,
      "learning_rate": 0.014031159200142428,
      "loss": 0.2652,
      "num_input_tokens_seen": 26273632,
      "step": 34450
    },
    {
      "epoch": 71.63201663201663,
      "grad_norm": 0.00020421632507350296,
      "learning_rate": 0.014006294517019667,
      "loss": 0.2707,
      "num_input_tokens_seen": 26277408,
      "step": 34455
    },
    {
      "epoch": 71.64241164241164,
      "grad_norm": 0.0002219188609160483,
      "learning_rate": 0.013981450805835276,
      "loss": 0.2787,
      "num_input_tokens_seen": 26281152,
      "step": 34460
    },
    {
      "epoch": 71.65280665280665,
      "grad_norm": 9.26225766306743e-05,
      "learning_rate": 0.01395662807042049,
      "loss": 0.2348,
      "num_input_tokens_seen": 26285088,
      "step": 34465
    },
    {
      "epoch": 71.66320166320166,
      "grad_norm": 0.00014120752166491002,
      "learning_rate": 0.013931826314603296,
      "loss": 0.2689,
      "num_input_tokens_seen": 26289024,
      "step": 34470
    },
    {
      "epoch": 71.67359667359668,
      "grad_norm": 0.0002083756262436509,
      "learning_rate": 0.013907045542208401,
      "loss": 0.2576,
      "num_input_tokens_seen": 26292896,
      "step": 34475
    },
    {
      "epoch": 71.68399168399168,
      "grad_norm": 0.00022319567506201565,
      "learning_rate": 0.013882285757057333,
      "loss": 0.289,
      "num_input_tokens_seen": 26296608,
      "step": 34480
    },
    {
      "epoch": 71.6943866943867,
      "grad_norm": 0.00013077058247290552,
      "learning_rate": 0.013857546962968403,
      "loss": 0.2941,
      "num_input_tokens_seen": 26300448,
      "step": 34485
    },
    {
      "epoch": 71.70478170478171,
      "grad_norm": 0.0001975538325496018,
      "learning_rate": 0.013832829163756577,
      "loss": 0.2585,
      "num_input_tokens_seen": 26304256,
      "step": 34490
    },
    {
      "epoch": 71.71517671517671,
      "grad_norm": 0.00021075663971714675,
      "learning_rate": 0.013808132363233689,
      "loss": 0.2763,
      "num_input_tokens_seen": 26308128,
      "step": 34495
    },
    {
      "epoch": 71.72557172557173,
      "grad_norm": 0.0003411433717701584,
      "learning_rate": 0.013783456565208256,
      "loss": 0.2513,
      "num_input_tokens_seen": 26312160,
      "step": 34500
    },
    {
      "epoch": 71.73596673596674,
      "grad_norm": 0.0011928318999707699,
      "learning_rate": 0.01375880177348564,
      "loss": 0.2809,
      "num_input_tokens_seen": 26315968,
      "step": 34505
    },
    {
      "epoch": 71.74636174636174,
      "grad_norm": 0.0001411486155120656,
      "learning_rate": 0.013734167991867928,
      "loss": 0.2684,
      "num_input_tokens_seen": 26319712,
      "step": 34510
    },
    {
      "epoch": 71.75675675675676,
      "grad_norm": 0.0002316038589924574,
      "learning_rate": 0.013709555224153935,
      "loss": 0.2587,
      "num_input_tokens_seen": 26323712,
      "step": 34515
    },
    {
      "epoch": 71.76715176715177,
      "grad_norm": 0.00013614627823699266,
      "learning_rate": 0.013684963474139222,
      "loss": 0.2526,
      "num_input_tokens_seen": 26327552,
      "step": 34520
    },
    {
      "epoch": 71.77754677754677,
      "grad_norm": 0.00013767197378911078,
      "learning_rate": 0.013660392745616224,
      "loss": 0.2763,
      "num_input_tokens_seen": 26331392,
      "step": 34525
    },
    {
      "epoch": 71.78794178794179,
      "grad_norm": 0.00042263694922439754,
      "learning_rate": 0.013635843042373974,
      "loss": 0.2382,
      "num_input_tokens_seen": 26335360,
      "step": 34530
    },
    {
      "epoch": 71.7983367983368,
      "grad_norm": 0.00031298157409764826,
      "learning_rate": 0.01361131436819843,
      "loss": 0.2607,
      "num_input_tokens_seen": 26339328,
      "step": 34535
    },
    {
      "epoch": 71.8087318087318,
      "grad_norm": 0.00030167793738655746,
      "learning_rate": 0.013586806726872147,
      "loss": 0.268,
      "num_input_tokens_seen": 26343200,
      "step": 34540
    },
    {
      "epoch": 71.81912681912682,
      "grad_norm": 0.0003853478701785207,
      "learning_rate": 0.013562320122174537,
      "loss": 0.2548,
      "num_input_tokens_seen": 26347040,
      "step": 34545
    },
    {
      "epoch": 71.82952182952182,
      "grad_norm": 0.0003635018365457654,
      "learning_rate": 0.013537854557881762,
      "loss": 0.2707,
      "num_input_tokens_seen": 26350912,
      "step": 34550
    },
    {
      "epoch": 71.83991683991684,
      "grad_norm": 0.000908699061255902,
      "learning_rate": 0.013513410037766687,
      "loss": 0.2829,
      "num_input_tokens_seen": 26354784,
      "step": 34555
    },
    {
      "epoch": 71.85031185031185,
      "grad_norm": 0.00015832159260753542,
      "learning_rate": 0.013488986565598998,
      "loss": 0.2579,
      "num_input_tokens_seen": 26358464,
      "step": 34560
    },
    {
      "epoch": 71.86070686070686,
      "grad_norm": 0.0003910884843207896,
      "learning_rate": 0.013464584145145097,
      "loss": 0.285,
      "num_input_tokens_seen": 26362208,
      "step": 34565
    },
    {
      "epoch": 71.87110187110187,
      "grad_norm": 0.0005502876010723412,
      "learning_rate": 0.013440202780168109,
      "loss": 0.2534,
      "num_input_tokens_seen": 26365888,
      "step": 34570
    },
    {
      "epoch": 71.88149688149689,
      "grad_norm": 0.00015817715029697865,
      "learning_rate": 0.01341584247442799,
      "loss": 0.2693,
      "num_input_tokens_seen": 26369696,
      "step": 34575
    },
    {
      "epoch": 71.89189189189189,
      "grad_norm": 0.00044241794967092574,
      "learning_rate": 0.013391503231681355,
      "loss": 0.2641,
      "num_input_tokens_seen": 26373472,
      "step": 34580
    },
    {
      "epoch": 71.9022869022869,
      "grad_norm": 0.0001169347160612233,
      "learning_rate": 0.013367185055681685,
      "loss": 0.2939,
      "num_input_tokens_seen": 26377248,
      "step": 34585
    },
    {
      "epoch": 71.91268191268192,
      "grad_norm": 0.00021283843670971692,
      "learning_rate": 0.013342887950179095,
      "loss": 0.2864,
      "num_input_tokens_seen": 26381088,
      "step": 34590
    },
    {
      "epoch": 71.92307692307692,
      "grad_norm": 0.0006728462176397443,
      "learning_rate": 0.013318611918920554,
      "loss": 0.2641,
      "num_input_tokens_seen": 26385088,
      "step": 34595
    },
    {
      "epoch": 71.93347193347194,
      "grad_norm": 0.0004620038962457329,
      "learning_rate": 0.01329435696564965,
      "loss": 0.256,
      "num_input_tokens_seen": 26388832,
      "step": 34600
    },
    {
      "epoch": 71.93347193347194,
      "eval_loss": 0.25349533557891846,
      "eval_runtime": 13.3973,
      "eval_samples_per_second": 63.893,
      "eval_steps_per_second": 15.973,
      "num_input_tokens_seen": 26388832,
      "step": 34600
    },
    {
      "epoch": 71.94386694386695,
      "grad_norm": 0.00020175377721898258,
      "learning_rate": 0.013270123094106894,
      "loss": 0.2508,
      "num_input_tokens_seen": 26392544,
      "step": 34605
    },
    {
      "epoch": 71.95426195426195,
      "grad_norm": 0.00018747425929177552,
      "learning_rate": 0.013245910308029395,
      "loss": 0.2795,
      "num_input_tokens_seen": 26396320,
      "step": 34610
    },
    {
      "epoch": 71.96465696465697,
      "grad_norm": 0.0013315289979800582,
      "learning_rate": 0.0132217186111511,
      "loss": 0.2578,
      "num_input_tokens_seen": 26400064,
      "step": 34615
    },
    {
      "epoch": 71.97505197505197,
      "grad_norm": 0.0004098436620552093,
      "learning_rate": 0.013197548007202626,
      "loss": 0.2609,
      "num_input_tokens_seen": 26403840,
      "step": 34620
    },
    {
      "epoch": 71.98544698544698,
      "grad_norm": 0.0003166873357258737,
      "learning_rate": 0.01317339849991142,
      "loss": 0.2604,
      "num_input_tokens_seen": 26407680,
      "step": 34625
    },
    {
      "epoch": 71.995841995842,
      "grad_norm": 0.0004101577214896679,
      "learning_rate": 0.013149270093001675,
      "loss": 0.2287,
      "num_input_tokens_seen": 26411584,
      "step": 34630
    },
    {
      "epoch": 72.006237006237,
      "grad_norm": 0.00016882021736819297,
      "learning_rate": 0.013125162790194227,
      "loss": 0.254,
      "num_input_tokens_seen": 26415344,
      "step": 34635
    },
    {
      "epoch": 72.01663201663202,
      "grad_norm": 0.00030382515978999436,
      "learning_rate": 0.01310107659520674,
      "loss": 0.2906,
      "num_input_tokens_seen": 26419120,
      "step": 34640
    },
    {
      "epoch": 72.02702702702703,
      "grad_norm": 0.00011020986130461097,
      "learning_rate": 0.013077011511753655,
      "loss": 0.2739,
      "num_input_tokens_seen": 26423056,
      "step": 34645
    },
    {
      "epoch": 72.03742203742203,
      "grad_norm": 0.0003761641273740679,
      "learning_rate": 0.013052967543546056,
      "loss": 0.2428,
      "num_input_tokens_seen": 26427056,
      "step": 34650
    },
    {
      "epoch": 72.04781704781705,
      "grad_norm": 8.697348675923422e-05,
      "learning_rate": 0.01302894469429186,
      "loss": 0.2513,
      "num_input_tokens_seen": 26430864,
      "step": 34655
    },
    {
      "epoch": 72.05821205821206,
      "grad_norm": 0.0002621742896735668,
      "learning_rate": 0.013004942967695653,
      "loss": 0.2484,
      "num_input_tokens_seen": 26434672,
      "step": 34660
    },
    {
      "epoch": 72.06860706860707,
      "grad_norm": 0.00011176583211636171,
      "learning_rate": 0.012980962367458859,
      "loss": 0.2544,
      "num_input_tokens_seen": 26438480,
      "step": 34665
    },
    {
      "epoch": 72.07900207900208,
      "grad_norm": 0.00014768738765269518,
      "learning_rate": 0.012957002897279567,
      "loss": 0.267,
      "num_input_tokens_seen": 26442320,
      "step": 34670
    },
    {
      "epoch": 72.0893970893971,
      "grad_norm": 0.001965432893484831,
      "learning_rate": 0.012933064560852576,
      "loss": 0.2506,
      "num_input_tokens_seen": 26446096,
      "step": 34675
    },
    {
      "epoch": 72.0997920997921,
      "grad_norm": 0.00010717882832977921,
      "learning_rate": 0.012909147361869527,
      "loss": 0.2678,
      "num_input_tokens_seen": 26450064,
      "step": 34680
    },
    {
      "epoch": 72.11018711018711,
      "grad_norm": 0.0012429632479324937,
      "learning_rate": 0.012885251304018774,
      "loss": 0.268,
      "num_input_tokens_seen": 26453904,
      "step": 34685
    },
    {
      "epoch": 72.12058212058211,
      "grad_norm": 0.0004679160483647138,
      "learning_rate": 0.012861376390985335,
      "loss": 0.2731,
      "num_input_tokens_seen": 26457808,
      "step": 34690
    },
    {
      "epoch": 72.13097713097713,
      "grad_norm": 0.00020628752827178687,
      "learning_rate": 0.012837522626451063,
      "loss": 0.2846,
      "num_input_tokens_seen": 26461520,
      "step": 34695
    },
    {
      "epoch": 72.14137214137214,
      "grad_norm": 0.00042619314626790583,
      "learning_rate": 0.01281369001409447,
      "loss": 0.2705,
      "num_input_tokens_seen": 26465392,
      "step": 34700
    },
    {
      "epoch": 72.15176715176715,
      "grad_norm": 0.0005593939567916095,
      "learning_rate": 0.012789878557590877,
      "loss": 0.269,
      "num_input_tokens_seen": 26469040,
      "step": 34705
    },
    {
      "epoch": 72.16216216216216,
      "grad_norm": 0.0006083393236622214,
      "learning_rate": 0.012766088260612334,
      "loss": 0.2298,
      "num_input_tokens_seen": 26472784,
      "step": 34710
    },
    {
      "epoch": 72.17255717255718,
      "grad_norm": 0.0003040891024284065,
      "learning_rate": 0.012742319126827523,
      "loss": 0.264,
      "num_input_tokens_seen": 26476592,
      "step": 34715
    },
    {
      "epoch": 72.18295218295218,
      "grad_norm": 0.00038121125544421375,
      "learning_rate": 0.012718571159902008,
      "loss": 0.2477,
      "num_input_tokens_seen": 26480336,
      "step": 34720
    },
    {
      "epoch": 72.1933471933472,
      "grad_norm": 0.0004151242901571095,
      "learning_rate": 0.01269484436349803,
      "loss": 0.2758,
      "num_input_tokens_seen": 26484048,
      "step": 34725
    },
    {
      "epoch": 72.20374220374221,
      "grad_norm": 0.00010516310430830345,
      "learning_rate": 0.012671138741274528,
      "loss": 0.2714,
      "num_input_tokens_seen": 26487952,
      "step": 34730
    },
    {
      "epoch": 72.21413721413721,
      "grad_norm": 0.00028693751664832234,
      "learning_rate": 0.012647454296887194,
      "loss": 0.264,
      "num_input_tokens_seen": 26491888,
      "step": 34735
    },
    {
      "epoch": 72.22453222453223,
      "grad_norm": 7.745195034658536e-05,
      "learning_rate": 0.012623791033988507,
      "loss": 0.2503,
      "num_input_tokens_seen": 26495664,
      "step": 34740
    },
    {
      "epoch": 72.23492723492724,
      "grad_norm": 0.0002716166200116277,
      "learning_rate": 0.012600148956227597,
      "loss": 0.2664,
      "num_input_tokens_seen": 26499600,
      "step": 34745
    },
    {
      "epoch": 72.24532224532224,
      "grad_norm": 0.00024009094340726733,
      "learning_rate": 0.012576528067250414,
      "loss": 0.2581,
      "num_input_tokens_seen": 26503440,
      "step": 34750
    },
    {
      "epoch": 72.25571725571726,
      "grad_norm": 0.0003295024507679045,
      "learning_rate": 0.012552928370699561,
      "loss": 0.2649,
      "num_input_tokens_seen": 26507152,
      "step": 34755
    },
    {
      "epoch": 72.26611226611226,
      "grad_norm": 0.0002833440084941685,
      "learning_rate": 0.012529349870214411,
      "loss": 0.2566,
      "num_input_tokens_seen": 26510768,
      "step": 34760
    },
    {
      "epoch": 72.27650727650727,
      "grad_norm": 0.00046820822171866894,
      "learning_rate": 0.012505792569431106,
      "loss": 0.272,
      "num_input_tokens_seen": 26514832,
      "step": 34765
    },
    {
      "epoch": 72.28690228690229,
      "grad_norm": 0.00015528849326074123,
      "learning_rate": 0.012482256471982422,
      "loss": 0.2613,
      "num_input_tokens_seen": 26518608,
      "step": 34770
    },
    {
      "epoch": 72.29729729729729,
      "grad_norm": 0.0003930999955628067,
      "learning_rate": 0.012458741581497956,
      "loss": 0.2159,
      "num_input_tokens_seen": 26522480,
      "step": 34775
    },
    {
      "epoch": 72.3076923076923,
      "grad_norm": 0.00011319175246171653,
      "learning_rate": 0.012435247901603974,
      "loss": 0.2578,
      "num_input_tokens_seen": 26526192,
      "step": 34780
    },
    {
      "epoch": 72.31808731808732,
      "grad_norm": 0.00017740637122187763,
      "learning_rate": 0.012411775435923528,
      "loss": 0.253,
      "num_input_tokens_seen": 26530000,
      "step": 34785
    },
    {
      "epoch": 72.32848232848232,
      "grad_norm": 0.0009421475697308779,
      "learning_rate": 0.012388324188076354,
      "loss": 0.2537,
      "num_input_tokens_seen": 26533872,
      "step": 34790
    },
    {
      "epoch": 72.33887733887734,
      "grad_norm": 0.00037671311292797327,
      "learning_rate": 0.012364894161678913,
      "loss": 0.2587,
      "num_input_tokens_seen": 26537680,
      "step": 34795
    },
    {
      "epoch": 72.34927234927235,
      "grad_norm": 8.218806760851294e-05,
      "learning_rate": 0.012341485360344445,
      "loss": 0.2666,
      "num_input_tokens_seen": 26541680,
      "step": 34800
    },
    {
      "epoch": 72.34927234927235,
      "eval_loss": 0.24757446348667145,
      "eval_runtime": 13.4016,
      "eval_samples_per_second": 63.873,
      "eval_steps_per_second": 15.968,
      "num_input_tokens_seen": 26541680,
      "step": 34800
    },
    {
      "epoch": 72.35966735966736,
      "grad_norm": 0.0005148149211890996,
      "learning_rate": 0.01231809778768283,
      "loss": 0.2582,
      "num_input_tokens_seen": 26545392,
      "step": 34805
    },
    {
      "epoch": 72.37006237006237,
      "grad_norm": 9.076271089725196e-05,
      "learning_rate": 0.012294731447300799,
      "loss": 0.2493,
      "num_input_tokens_seen": 26549232,
      "step": 34810
    },
    {
      "epoch": 72.38045738045739,
      "grad_norm": 0.00010365769412601367,
      "learning_rate": 0.012271386342801671,
      "loss": 0.2515,
      "num_input_tokens_seen": 26553008,
      "step": 34815
    },
    {
      "epoch": 72.39085239085239,
      "grad_norm": 0.0005789612769149244,
      "learning_rate": 0.012248062477785565,
      "loss": 0.2392,
      "num_input_tokens_seen": 26556880,
      "step": 34820
    },
    {
      "epoch": 72.4012474012474,
      "grad_norm": 0.00039866394945420325,
      "learning_rate": 0.012224759855849305,
      "loss": 0.2665,
      "num_input_tokens_seen": 26560752,
      "step": 34825
    },
    {
      "epoch": 72.41164241164242,
      "grad_norm": 0.0005092307110317051,
      "learning_rate": 0.012201478480586513,
      "loss": 0.2497,
      "num_input_tokens_seen": 26564496,
      "step": 34830
    },
    {
      "epoch": 72.42203742203742,
      "grad_norm": 0.0001708591153146699,
      "learning_rate": 0.012178218355587389,
      "loss": 0.2572,
      "num_input_tokens_seen": 26568304,
      "step": 34835
    },
    {
      "epoch": 72.43243243243244,
      "grad_norm": 0.00021920054859947413,
      "learning_rate": 0.01215497948443896,
      "loss": 0.2578,
      "num_input_tokens_seen": 26572016,
      "step": 34840
    },
    {
      "epoch": 72.44282744282744,
      "grad_norm": 0.0016531936125829816,
      "learning_rate": 0.012131761870724993,
      "loss": 0.2752,
      "num_input_tokens_seen": 26575952,
      "step": 34845
    },
    {
      "epoch": 72.45322245322245,
      "grad_norm": 0.0010893176076933742,
      "learning_rate": 0.012108565518025893,
      "loss": 0.2259,
      "num_input_tokens_seen": 26579984,
      "step": 34850
    },
    {
      "epoch": 72.46361746361747,
      "grad_norm": 0.0002617994323372841,
      "learning_rate": 0.012085390429918862,
      "loss": 0.2733,
      "num_input_tokens_seen": 26583888,
      "step": 34855
    },
    {
      "epoch": 72.47401247401247,
      "grad_norm": 0.000740803312510252,
      "learning_rate": 0.012062236609977744,
      "loss": 0.2754,
      "num_input_tokens_seen": 26587760,
      "step": 34860
    },
    {
      "epoch": 72.48440748440748,
      "grad_norm": 0.0004866077797487378,
      "learning_rate": 0.01203910406177318,
      "loss": 0.2821,
      "num_input_tokens_seen": 26591632,
      "step": 34865
    },
    {
      "epoch": 72.4948024948025,
      "grad_norm": 0.0005302297067828476,
      "learning_rate": 0.01201599278887252,
      "loss": 0.2916,
      "num_input_tokens_seen": 26595472,
      "step": 34870
    },
    {
      "epoch": 72.5051975051975,
      "grad_norm": 0.00019920691556762904,
      "learning_rate": 0.011992902794839744,
      "loss": 0.2615,
      "num_input_tokens_seen": 26599376,
      "step": 34875
    },
    {
      "epoch": 72.51559251559252,
      "grad_norm": 0.00010601821122691035,
      "learning_rate": 0.011969834083235703,
      "loss": 0.2591,
      "num_input_tokens_seen": 26603120,
      "step": 34880
    },
    {
      "epoch": 72.52598752598753,
      "grad_norm": 0.0002832012833096087,
      "learning_rate": 0.011946786657617836,
      "loss": 0.2795,
      "num_input_tokens_seen": 26606960,
      "step": 34885
    },
    {
      "epoch": 72.53638253638253,
      "grad_norm": 0.000433982175309211,
      "learning_rate": 0.011923760521540332,
      "loss": 0.2535,
      "num_input_tokens_seen": 26610768,
      "step": 34890
    },
    {
      "epoch": 72.54677754677755,
      "grad_norm": 0.0022351054940372705,
      "learning_rate": 0.011900755678554153,
      "loss": 0.2539,
      "num_input_tokens_seen": 26614768,
      "step": 34895
    },
    {
      "epoch": 72.55717255717256,
      "grad_norm": 0.00026285924832336605,
      "learning_rate": 0.011877772132206893,
      "loss": 0.2682,
      "num_input_tokens_seen": 26618640,
      "step": 34900
    },
    {
      "epoch": 72.56756756756756,
      "grad_norm": 0.00027138934819959104,
      "learning_rate": 0.011854809886042915,
      "loss": 0.2584,
      "num_input_tokens_seen": 26622320,
      "step": 34905
    },
    {
      "epoch": 72.57796257796258,
      "grad_norm": 0.0017051596660166979,
      "learning_rate": 0.011831868943603325,
      "loss": 0.2262,
      "num_input_tokens_seen": 26626064,
      "step": 34910
    },
    {
      "epoch": 72.58835758835758,
      "grad_norm": 0.0004306395712774247,
      "learning_rate": 0.011808949308425836,
      "loss": 0.2477,
      "num_input_tokens_seen": 26629968,
      "step": 34915
    },
    {
      "epoch": 72.5987525987526,
      "grad_norm": 0.0005320144700817764,
      "learning_rate": 0.01178605098404501,
      "loss": 0.2788,
      "num_input_tokens_seen": 26633712,
      "step": 34920
    },
    {
      "epoch": 72.60914760914761,
      "grad_norm": 0.000345159147400409,
      "learning_rate": 0.011763173973992002,
      "loss": 0.2629,
      "num_input_tokens_seen": 26637584,
      "step": 34925
    },
    {
      "epoch": 72.61954261954261,
      "grad_norm": 0.0005791736184619367,
      "learning_rate": 0.011740318281794776,
      "loss": 0.2926,
      "num_input_tokens_seen": 26641488,
      "step": 34930
    },
    {
      "epoch": 72.62993762993763,
      "grad_norm": 0.002580442000180483,
      "learning_rate": 0.01171748391097796,
      "loss": 0.2751,
      "num_input_tokens_seen": 26645200,
      "step": 34935
    },
    {
      "epoch": 72.64033264033264,
      "grad_norm": 0.0007091288571245968,
      "learning_rate": 0.011694670865062873,
      "loss": 0.2533,
      "num_input_tokens_seen": 26648880,
      "step": 34940
    },
    {
      "epoch": 72.65072765072765,
      "grad_norm": 0.000738053466193378,
      "learning_rate": 0.011671879147567616,
      "loss": 0.2386,
      "num_input_tokens_seen": 26652688,
      "step": 34945
    },
    {
      "epoch": 72.66112266112266,
      "grad_norm": 0.0003458709397818893,
      "learning_rate": 0.011649108762006893,
      "loss": 0.2655,
      "num_input_tokens_seen": 26656496,
      "step": 34950
    },
    {
      "epoch": 72.67151767151768,
      "grad_norm": 0.00039297930197790265,
      "learning_rate": 0.011626359711892265,
      "loss": 0.2773,
      "num_input_tokens_seen": 26660144,
      "step": 34955
    },
    {
      "epoch": 72.68191268191268,
      "grad_norm": 0.00016521291399840266,
      "learning_rate": 0.01160363200073189,
      "loss": 0.2693,
      "num_input_tokens_seen": 26664016,
      "step": 34960
    },
    {
      "epoch": 72.6923076923077,
      "grad_norm": 9.515542478766292e-05,
      "learning_rate": 0.011580925632030614,
      "loss": 0.2542,
      "num_input_tokens_seen": 26667856,
      "step": 34965
    },
    {
      "epoch": 72.70270270270271,
      "grad_norm": 6.108672096161172e-05,
      "learning_rate": 0.011558240609290104,
      "loss": 0.2673,
      "num_input_tokens_seen": 26671856,
      "step": 34970
    },
    {
      "epoch": 72.71309771309771,
      "grad_norm": 3.3291846193606034e-05,
      "learning_rate": 0.011535576936008679,
      "loss": 0.2869,
      "num_input_tokens_seen": 26675632,
      "step": 34975
    },
    {
      "epoch": 72.72349272349273,
      "grad_norm": 0.00020035800116602331,
      "learning_rate": 0.011512934615681309,
      "loss": 0.2706,
      "num_input_tokens_seen": 26679536,
      "step": 34980
    },
    {
      "epoch": 72.73388773388774,
      "grad_norm": 0.0003292696492280811,
      "learning_rate": 0.011490313651799765,
      "loss": 0.2666,
      "num_input_tokens_seen": 26683344,
      "step": 34985
    },
    {
      "epoch": 72.74428274428274,
      "grad_norm": 0.0002502183197066188,
      "learning_rate": 0.011467714047852512,
      "loss": 0.2622,
      "num_input_tokens_seen": 26687216,
      "step": 34990
    },
    {
      "epoch": 72.75467775467776,
      "grad_norm": 0.00014884970732964575,
      "learning_rate": 0.011445135807324624,
      "loss": 0.2596,
      "num_input_tokens_seen": 26691120,
      "step": 34995
    },
    {
      "epoch": 72.76507276507276,
      "grad_norm": 0.00017719414609018713,
      "learning_rate": 0.011422578933698002,
      "loss": 0.2528,
      "num_input_tokens_seen": 26694832,
      "step": 35000
    },
    {
      "epoch": 72.76507276507276,
      "eval_loss": 0.2502278983592987,
      "eval_runtime": 13.396,
      "eval_samples_per_second": 63.9,
      "eval_steps_per_second": 15.975,
      "num_input_tokens_seen": 26694832,
      "step": 35000
    },
    {
      "epoch": 72.77546777546777,
      "grad_norm": 0.00025865077623166144,
      "learning_rate": 0.011400043430451161,
      "loss": 0.2674,
      "num_input_tokens_seen": 26698608,
      "step": 35005
    },
    {
      "epoch": 72.78586278586279,
      "grad_norm": 0.00018091282981913537,
      "learning_rate": 0.011377529301059392,
      "loss": 0.2648,
      "num_input_tokens_seen": 26702352,
      "step": 35010
    },
    {
      "epoch": 72.79625779625779,
      "grad_norm": 3.604188532335684e-05,
      "learning_rate": 0.011355036548994646,
      "loss": 0.2652,
      "num_input_tokens_seen": 26706064,
      "step": 35015
    },
    {
      "epoch": 72.8066528066528,
      "grad_norm": 0.0003208151028957218,
      "learning_rate": 0.011332565177725584,
      "loss": 0.2929,
      "num_input_tokens_seen": 26709808,
      "step": 35020
    },
    {
      "epoch": 72.81704781704782,
      "grad_norm": 5.122231596033089e-05,
      "learning_rate": 0.011310115190717585,
      "loss": 0.2826,
      "num_input_tokens_seen": 26713552,
      "step": 35025
    },
    {
      "epoch": 72.82744282744282,
      "grad_norm": 0.000767905090469867,
      "learning_rate": 0.01128768659143271,
      "loss": 0.2594,
      "num_input_tokens_seen": 26717456,
      "step": 35030
    },
    {
      "epoch": 72.83783783783784,
      "grad_norm": 0.0005068514728918672,
      "learning_rate": 0.011265279383329713,
      "loss": 0.2711,
      "num_input_tokens_seen": 26721200,
      "step": 35035
    },
    {
      "epoch": 72.84823284823285,
      "grad_norm": 0.0001822207123041153,
      "learning_rate": 0.01124289356986411,
      "loss": 0.2526,
      "num_input_tokens_seen": 26724880,
      "step": 35040
    },
    {
      "epoch": 72.85862785862786,
      "grad_norm": 0.0009814156219363213,
      "learning_rate": 0.011220529154488023,
      "loss": 0.2644,
      "num_input_tokens_seen": 26728656,
      "step": 35045
    },
    {
      "epoch": 72.86902286902287,
      "grad_norm": 0.00026034426991827786,
      "learning_rate": 0.011198186140650346,
      "loss": 0.2699,
      "num_input_tokens_seen": 26732432,
      "step": 35050
    },
    {
      "epoch": 72.87941787941789,
      "grad_norm": 0.0008678545127622783,
      "learning_rate": 0.011175864531796685,
      "loss": 0.2621,
      "num_input_tokens_seen": 26736272,
      "step": 35055
    },
    {
      "epoch": 72.88981288981289,
      "grad_norm": 0.00023129310284275562,
      "learning_rate": 0.011153564331369258,
      "loss": 0.2609,
      "num_input_tokens_seen": 26740016,
      "step": 35060
    },
    {
      "epoch": 72.9002079002079,
      "grad_norm": 0.00033284572418779135,
      "learning_rate": 0.011131285542807078,
      "loss": 0.2636,
      "num_input_tokens_seen": 26743728,
      "step": 35065
    },
    {
      "epoch": 72.9106029106029,
      "grad_norm": 0.0004265870666131377,
      "learning_rate": 0.011109028169545815,
      "loss": 0.2488,
      "num_input_tokens_seen": 26747568,
      "step": 35070
    },
    {
      "epoch": 72.92099792099792,
      "grad_norm": 0.00025476107839494944,
      "learning_rate": 0.011086792215017804,
      "loss": 0.2733,
      "num_input_tokens_seen": 26751440,
      "step": 35075
    },
    {
      "epoch": 72.93139293139293,
      "grad_norm": 9.862656588666141e-05,
      "learning_rate": 0.011064577682652137,
      "loss": 0.2564,
      "num_input_tokens_seen": 26755312,
      "step": 35080
    },
    {
      "epoch": 72.94178794178794,
      "grad_norm": 0.0009094732231460512,
      "learning_rate": 0.011042384575874559,
      "loss": 0.2436,
      "num_input_tokens_seen": 26759152,
      "step": 35085
    },
    {
      "epoch": 72.95218295218295,
      "grad_norm": 0.00036140522570349276,
      "learning_rate": 0.011020212898107512,
      "loss": 0.2523,
      "num_input_tokens_seen": 26762992,
      "step": 35090
    },
    {
      "epoch": 72.96257796257797,
      "grad_norm": 0.00031674173078499734,
      "learning_rate": 0.010998062652770197,
      "loss": 0.2806,
      "num_input_tokens_seen": 26766736,
      "step": 35095
    },
    {
      "epoch": 72.97297297297297,
      "grad_norm": 0.00042393614421598613,
      "learning_rate": 0.010975933843278428,
      "loss": 0.2893,
      "num_input_tokens_seen": 26770480,
      "step": 35100
    },
    {
      "epoch": 72.98336798336798,
      "grad_norm": 0.0004447273095138371,
      "learning_rate": 0.010953826473044714,
      "loss": 0.256,
      "num_input_tokens_seen": 26774480,
      "step": 35105
    },
    {
      "epoch": 72.993762993763,
      "grad_norm": 8.807906851870939e-05,
      "learning_rate": 0.010931740545478357,
      "loss": 0.2752,
      "num_input_tokens_seen": 26778256,
      "step": 35110
    },
    {
      "epoch": 73.004158004158,
      "grad_norm": 0.0007353270193561912,
      "learning_rate": 0.010909676063985218,
      "loss": 0.2566,
      "num_input_tokens_seen": 26781952,
      "step": 35115
    },
    {
      "epoch": 73.01455301455302,
      "grad_norm": 0.00042906764429062605,
      "learning_rate": 0.010887633031967974,
      "loss": 0.2621,
      "num_input_tokens_seen": 26785824,
      "step": 35120
    },
    {
      "epoch": 73.02494802494803,
      "grad_norm": 0.0005121810827404261,
      "learning_rate": 0.01086561145282589,
      "loss": 0.2539,
      "num_input_tokens_seen": 26789664,
      "step": 35125
    },
    {
      "epoch": 73.03534303534303,
      "grad_norm": 0.00025208445731550455,
      "learning_rate": 0.010843611329954983,
      "loss": 0.2417,
      "num_input_tokens_seen": 26793440,
      "step": 35130
    },
    {
      "epoch": 73.04573804573805,
      "grad_norm": 0.0013416351284831762,
      "learning_rate": 0.010821632666747988,
      "loss": 0.2553,
      "num_input_tokens_seen": 26797408,
      "step": 35135
    },
    {
      "epoch": 73.05613305613305,
      "grad_norm": 0.0002883170382119715,
      "learning_rate": 0.010799675466594244,
      "loss": 0.261,
      "num_input_tokens_seen": 26801280,
      "step": 35140
    },
    {
      "epoch": 73.06652806652806,
      "grad_norm": 0.00020439941727090627,
      "learning_rate": 0.010777739732879826,
      "loss": 0.2691,
      "num_input_tokens_seen": 26805088,
      "step": 35145
    },
    {
      "epoch": 73.07692307692308,
      "grad_norm": 0.00012665581016335636,
      "learning_rate": 0.010755825468987562,
      "loss": 0.2714,
      "num_input_tokens_seen": 26808960,
      "step": 35150
    },
    {
      "epoch": 73.08731808731808,
      "grad_norm": 0.0001814990973798558,
      "learning_rate": 0.010733932678296814,
      "loss": 0.2468,
      "num_input_tokens_seen": 26812832,
      "step": 35155
    },
    {
      "epoch": 73.0977130977131,
      "grad_norm": 0.00015646299289073795,
      "learning_rate": 0.010712061364183817,
      "loss": 0.2685,
      "num_input_tokens_seen": 26816736,
      "step": 35160
    },
    {
      "epoch": 73.10810810810811,
      "grad_norm": 0.0002479937975294888,
      "learning_rate": 0.010690211530021337,
      "loss": 0.2608,
      "num_input_tokens_seen": 26820768,
      "step": 35165
    },
    {
      "epoch": 73.11850311850311,
      "grad_norm": 0.00042188348015770316,
      "learning_rate": 0.01066838317917893,
      "loss": 0.2432,
      "num_input_tokens_seen": 26824640,
      "step": 35170
    },
    {
      "epoch": 73.12889812889813,
      "grad_norm": 0.00025837679277174175,
      "learning_rate": 0.010646576315022787,
      "loss": 0.2348,
      "num_input_tokens_seen": 26828384,
      "step": 35175
    },
    {
      "epoch": 73.13929313929314,
      "grad_norm": 0.00011937220551772043,
      "learning_rate": 0.010624790940915785,
      "loss": 0.2774,
      "num_input_tokens_seen": 26832192,
      "step": 35180
    },
    {
      "epoch": 73.14968814968815,
      "grad_norm": 0.00022277713287621737,
      "learning_rate": 0.0106030270602175,
      "loss": 0.2651,
      "num_input_tokens_seen": 26835968,
      "step": 35185
    },
    {
      "epoch": 73.16008316008316,
      "grad_norm": 0.00046280043898150325,
      "learning_rate": 0.010581284676284252,
      "loss": 0.2747,
      "num_input_tokens_seen": 26839776,
      "step": 35190
    },
    {
      "epoch": 73.17047817047818,
      "grad_norm": 0.000174676941242069,
      "learning_rate": 0.010559563792468923,
      "loss": 0.2552,
      "num_input_tokens_seen": 26843456,
      "step": 35195
    },
    {
      "epoch": 73.18087318087318,
      "grad_norm": 0.00010934713645838201,
      "learning_rate": 0.010537864412121217,
      "loss": 0.263,
      "num_input_tokens_seen": 26847168,
      "step": 35200
    },
    {
      "epoch": 73.18087318087318,
      "eval_loss": 0.24831782281398773,
      "eval_runtime": 13.4014,
      "eval_samples_per_second": 63.874,
      "eval_steps_per_second": 15.969,
      "num_input_tokens_seen": 26847168,
      "step": 35200
    },
    {
      "epoch": 73.1912681912682,
      "grad_norm": 0.0002878326049540192,
      "learning_rate": 0.010516186538587357,
      "loss": 0.2569,
      "num_input_tokens_seen": 26850944,
      "step": 35205
    },
    {
      "epoch": 73.20166320166321,
      "grad_norm": 0.0001266978360945359,
      "learning_rate": 0.01049453017521042,
      "loss": 0.242,
      "num_input_tokens_seen": 26854720,
      "step": 35210
    },
    {
      "epoch": 73.21205821205821,
      "grad_norm": 0.0001323156029684469,
      "learning_rate": 0.010472895325330083,
      "loss": 0.2573,
      "num_input_tokens_seen": 26858720,
      "step": 35215
    },
    {
      "epoch": 73.22245322245323,
      "grad_norm": 7.0307040004991e-05,
      "learning_rate": 0.010451281992282662,
      "loss": 0.256,
      "num_input_tokens_seen": 26862432,
      "step": 35220
    },
    {
      "epoch": 73.23284823284823,
      "grad_norm": 8.258586603915319e-05,
      "learning_rate": 0.01042969017940124,
      "loss": 0.2544,
      "num_input_tokens_seen": 26865984,
      "step": 35225
    },
    {
      "epoch": 73.24324324324324,
      "grad_norm": 0.0002530558267608285,
      "learning_rate": 0.01040811989001557,
      "loss": 0.2541,
      "num_input_tokens_seen": 26869824,
      "step": 35230
    },
    {
      "epoch": 73.25363825363826,
      "grad_norm": 0.0003990020486526191,
      "learning_rate": 0.010386571127451992,
      "loss": 0.2758,
      "num_input_tokens_seen": 26873632,
      "step": 35235
    },
    {
      "epoch": 73.26403326403326,
      "grad_norm": 0.000595601974055171,
      "learning_rate": 0.010365043895033682,
      "loss": 0.2667,
      "num_input_tokens_seen": 26877632,
      "step": 35240
    },
    {
      "epoch": 73.27442827442827,
      "grad_norm": 0.0005734565202146769,
      "learning_rate": 0.010343538196080365,
      "loss": 0.2695,
      "num_input_tokens_seen": 26881504,
      "step": 35245
    },
    {
      "epoch": 73.28482328482329,
      "grad_norm": 0.00043768910109065473,
      "learning_rate": 0.010322054033908457,
      "loss": 0.2622,
      "num_input_tokens_seen": 26885280,
      "step": 35250
    },
    {
      "epoch": 73.29521829521829,
      "grad_norm": 0.0010707448236644268,
      "learning_rate": 0.010300591411831156,
      "loss": 0.2869,
      "num_input_tokens_seen": 26889184,
      "step": 35255
    },
    {
      "epoch": 73.3056133056133,
      "grad_norm": 0.00043533314601518214,
      "learning_rate": 0.010279150333158198,
      "loss": 0.2641,
      "num_input_tokens_seen": 26892896,
      "step": 35260
    },
    {
      "epoch": 73.31600831600832,
      "grad_norm": 0.0002873156627174467,
      "learning_rate": 0.010257730801196107,
      "loss": 0.2677,
      "num_input_tokens_seen": 26896768,
      "step": 35265
    },
    {
      "epoch": 73.32640332640332,
      "grad_norm": 0.0003685600240714848,
      "learning_rate": 0.010236332819248056,
      "loss": 0.2886,
      "num_input_tokens_seen": 26900544,
      "step": 35270
    },
    {
      "epoch": 73.33679833679834,
      "grad_norm": 0.0005905661382712424,
      "learning_rate": 0.010214956390613854,
      "loss": 0.2893,
      "num_input_tokens_seen": 26904352,
      "step": 35275
    },
    {
      "epoch": 73.34719334719335,
      "grad_norm": 0.00015305352280847728,
      "learning_rate": 0.010193601518590034,
      "loss": 0.27,
      "num_input_tokens_seen": 26908224,
      "step": 35280
    },
    {
      "epoch": 73.35758835758836,
      "grad_norm": 0.00047347971121780574,
      "learning_rate": 0.010172268206469758,
      "loss": 0.2355,
      "num_input_tokens_seen": 26912096,
      "step": 35285
    },
    {
      "epoch": 73.36798336798337,
      "grad_norm": 0.00020713932462967932,
      "learning_rate": 0.010150956457542897,
      "loss": 0.2737,
      "num_input_tokens_seen": 26916000,
      "step": 35290
    },
    {
      "epoch": 73.37837837837837,
      "grad_norm": 0.00038774916902184486,
      "learning_rate": 0.010129666275096054,
      "loss": 0.2623,
      "num_input_tokens_seen": 26920000,
      "step": 35295
    },
    {
      "epoch": 73.38877338877339,
      "grad_norm": 3.673564060591161e-05,
      "learning_rate": 0.010108397662412338,
      "loss": 0.263,
      "num_input_tokens_seen": 26923584,
      "step": 35300
    },
    {
      "epoch": 73.3991683991684,
      "grad_norm": 0.0003616712347138673,
      "learning_rate": 0.010087150622771707,
      "loss": 0.2408,
      "num_input_tokens_seen": 26927488,
      "step": 35305
    },
    {
      "epoch": 73.4095634095634,
      "grad_norm": 0.00017674409900791943,
      "learning_rate": 0.010065925159450739,
      "loss": 0.2737,
      "num_input_tokens_seen": 26931392,
      "step": 35310
    },
    {
      "epoch": 73.41995841995842,
      "grad_norm": 0.0002015016507357359,
      "learning_rate": 0.010044721275722618,
      "loss": 0.2634,
      "num_input_tokens_seen": 26935104,
      "step": 35315
    },
    {
      "epoch": 73.43035343035343,
      "grad_norm": 0.0006162866484373808,
      "learning_rate": 0.01002353897485726,
      "loss": 0.2684,
      "num_input_tokens_seen": 26938816,
      "step": 35320
    },
    {
      "epoch": 73.44074844074844,
      "grad_norm": 0.00018490733054932207,
      "learning_rate": 0.010002378260121236,
      "loss": 0.2867,
      "num_input_tokens_seen": 26942752,
      "step": 35325
    },
    {
      "epoch": 73.45114345114345,
      "grad_norm": 0.0002533454680815339,
      "learning_rate": 0.009981239134777786,
      "loss": 0.253,
      "num_input_tokens_seen": 26946624,
      "step": 35330
    },
    {
      "epoch": 73.46153846153847,
      "grad_norm": 7.157825166359544e-05,
      "learning_rate": 0.009960121602086884,
      "loss": 0.2601,
      "num_input_tokens_seen": 26950368,
      "step": 35335
    },
    {
      "epoch": 73.47193347193347,
      "grad_norm": 0.0005854950868524611,
      "learning_rate": 0.009939025665305062,
      "loss": 0.2556,
      "num_input_tokens_seen": 26954272,
      "step": 35340
    },
    {
      "epoch": 73.48232848232848,
      "grad_norm": 0.00026317095034755766,
      "learning_rate": 0.009917951327685597,
      "loss": 0.2564,
      "num_input_tokens_seen": 26958112,
      "step": 35345
    },
    {
      "epoch": 73.4927234927235,
      "grad_norm": 0.002402608748525381,
      "learning_rate": 0.009896898592478425,
      "loss": 0.272,
      "num_input_tokens_seen": 26961856,
      "step": 35350
    },
    {
      "epoch": 73.5031185031185,
      "grad_norm": 0.0004182198317721486,
      "learning_rate": 0.009875867462930132,
      "loss": 0.2624,
      "num_input_tokens_seen": 26965664,
      "step": 35355
    },
    {
      "epoch": 73.51351351351352,
      "grad_norm": 0.00024241018400061876,
      "learning_rate": 0.009854857942284006,
      "loss": 0.2739,
      "num_input_tokens_seen": 26969504,
      "step": 35360
    },
    {
      "epoch": 73.52390852390852,
      "grad_norm": 0.00012619365588761866,
      "learning_rate": 0.009833870033779923,
      "loss": 0.2547,
      "num_input_tokens_seen": 26973248,
      "step": 35365
    },
    {
      "epoch": 73.53430353430353,
      "grad_norm": 0.0004259055422153324,
      "learning_rate": 0.009812903740654527,
      "loss": 0.2428,
      "num_input_tokens_seen": 26976896,
      "step": 35370
    },
    {
      "epoch": 73.54469854469855,
      "grad_norm": 0.0002757531765382737,
      "learning_rate": 0.009791959066141097,
      "loss": 0.2725,
      "num_input_tokens_seen": 26980896,
      "step": 35375
    },
    {
      "epoch": 73.55509355509355,
      "grad_norm": 8.038845408009365e-05,
      "learning_rate": 0.009771036013469537,
      "loss": 0.2368,
      "num_input_tokens_seen": 26984736,
      "step": 35380
    },
    {
      "epoch": 73.56548856548856,
      "grad_norm": 0.00015068287029862404,
      "learning_rate": 0.00975013458586646,
      "loss": 0.2394,
      "num_input_tokens_seen": 26988448,
      "step": 35385
    },
    {
      "epoch": 73.57588357588358,
      "grad_norm": 0.00041127673466689885,
      "learning_rate": 0.009729254786555107,
      "loss": 0.2655,
      "num_input_tokens_seen": 26992192,
      "step": 35390
    },
    {
      "epoch": 73.58627858627858,
      "grad_norm": 0.0004211366467643529,
      "learning_rate": 0.009708396618755421,
      "loss": 0.2577,
      "num_input_tokens_seen": 26996128,
      "step": 35395
    },
    {
      "epoch": 73.5966735966736,
      "grad_norm": 0.00023710649111308157,
      "learning_rate": 0.009687560085683994,
      "loss": 0.2583,
      "num_input_tokens_seen": 27000096,
      "step": 35400
    },
    {
      "epoch": 73.5966735966736,
      "eval_loss": 0.24813279509544373,
      "eval_runtime": 13.4022,
      "eval_samples_per_second": 63.87,
      "eval_steps_per_second": 15.967,
      "num_input_tokens_seen": 27000096,
      "step": 35400
    },
    {
      "epoch": 73.60706860706861,
      "grad_norm": 0.0005935343215242028,
      "learning_rate": 0.009666745190554054,
      "loss": 0.2719,
      "num_input_tokens_seen": 27003936,
      "step": 35405
    },
    {
      "epoch": 73.61746361746361,
      "grad_norm": 0.0003768174210563302,
      "learning_rate": 0.009645951936575553,
      "loss": 0.2497,
      "num_input_tokens_seen": 27007552,
      "step": 35410
    },
    {
      "epoch": 73.62785862785863,
      "grad_norm": 0.0002991840592585504,
      "learning_rate": 0.00962518032695509,
      "loss": 0.2666,
      "num_input_tokens_seen": 27011232,
      "step": 35415
    },
    {
      "epoch": 73.63825363825364,
      "grad_norm": 8.902014087652788e-05,
      "learning_rate": 0.009604430364895855,
      "loss": 0.2591,
      "num_input_tokens_seen": 27014944,
      "step": 35420
    },
    {
      "epoch": 73.64864864864865,
      "grad_norm": 0.0010339658474549651,
      "learning_rate": 0.00958370205359777,
      "loss": 0.2538,
      "num_input_tokens_seen": 27018688,
      "step": 35425
    },
    {
      "epoch": 73.65904365904366,
      "grad_norm": 0.000444586796220392,
      "learning_rate": 0.009562995396257445,
      "loss": 0.2352,
      "num_input_tokens_seen": 27022304,
      "step": 35430
    },
    {
      "epoch": 73.66943866943868,
      "grad_norm": 0.0002212984545622021,
      "learning_rate": 0.009542310396068026,
      "loss": 0.2559,
      "num_input_tokens_seen": 27026176,
      "step": 35435
    },
    {
      "epoch": 73.67983367983368,
      "grad_norm": 0.00010784497862914577,
      "learning_rate": 0.009521647056219495,
      "loss": 0.2805,
      "num_input_tokens_seen": 27029952,
      "step": 35440
    },
    {
      "epoch": 73.6902286902287,
      "grad_norm": 0.000459110684460029,
      "learning_rate": 0.00950100537989832,
      "loss": 0.2795,
      "num_input_tokens_seen": 27033792,
      "step": 35445
    },
    {
      "epoch": 73.7006237006237,
      "grad_norm": 0.0003582810459192842,
      "learning_rate": 0.00948038537028772,
      "loss": 0.2777,
      "num_input_tokens_seen": 27037600,
      "step": 35450
    },
    {
      "epoch": 73.71101871101871,
      "grad_norm": 0.00031609792495146394,
      "learning_rate": 0.009459787030567617,
      "loss": 0.2465,
      "num_input_tokens_seen": 27041472,
      "step": 35455
    },
    {
      "epoch": 73.72141372141373,
      "grad_norm": 0.00017977422976400703,
      "learning_rate": 0.00943921036391449,
      "loss": 0.2599,
      "num_input_tokens_seen": 27045216,
      "step": 35460
    },
    {
      "epoch": 73.73180873180873,
      "grad_norm": 0.0003309495223220438,
      "learning_rate": 0.009418655373501483,
      "loss": 0.307,
      "num_input_tokens_seen": 27049184,
      "step": 35465
    },
    {
      "epoch": 73.74220374220374,
      "grad_norm": 0.00028912367997691035,
      "learning_rate": 0.00939812206249851,
      "loss": 0.2607,
      "num_input_tokens_seen": 27053088,
      "step": 35470
    },
    {
      "epoch": 73.75259875259876,
      "grad_norm": 0.00018243902013637125,
      "learning_rate": 0.009377610434072004,
      "loss": 0.2729,
      "num_input_tokens_seen": 27056928,
      "step": 35475
    },
    {
      "epoch": 73.76299376299376,
      "grad_norm": 0.0006715066265314817,
      "learning_rate": 0.009357120491385167,
      "loss": 0.2717,
      "num_input_tokens_seen": 27060640,
      "step": 35480
    },
    {
      "epoch": 73.77338877338877,
      "grad_norm": 0.00011176204861840233,
      "learning_rate": 0.009336652237597743,
      "loss": 0.2927,
      "num_input_tokens_seen": 27064480,
      "step": 35485
    },
    {
      "epoch": 73.78378378378379,
      "grad_norm": 0.00037878184230066836,
      "learning_rate": 0.009316205675866251,
      "loss": 0.2731,
      "num_input_tokens_seen": 27068320,
      "step": 35490
    },
    {
      "epoch": 73.79417879417879,
      "grad_norm": 0.0002593880635686219,
      "learning_rate": 0.00929578080934379,
      "loss": 0.2699,
      "num_input_tokens_seen": 27072160,
      "step": 35495
    },
    {
      "epoch": 73.8045738045738,
      "grad_norm": 0.0003337521920911968,
      "learning_rate": 0.00927537764118012,
      "loss": 0.2528,
      "num_input_tokens_seen": 27075968,
      "step": 35500
    },
    {
      "epoch": 73.81496881496882,
      "grad_norm": 0.0005197642603889108,
      "learning_rate": 0.009254996174521678,
      "loss": 0.2587,
      "num_input_tokens_seen": 27079712,
      "step": 35505
    },
    {
      "epoch": 73.82536382536382,
      "grad_norm": 0.00020184523600619286,
      "learning_rate": 0.009234636412511531,
      "loss": 0.2785,
      "num_input_tokens_seen": 27083584,
      "step": 35510
    },
    {
      "epoch": 73.83575883575884,
      "grad_norm": 0.0005622287862934172,
      "learning_rate": 0.009214298358289418,
      "loss": 0.2694,
      "num_input_tokens_seen": 27087328,
      "step": 35515
    },
    {
      "epoch": 73.84615384615384,
      "grad_norm": 0.0005193763645365834,
      "learning_rate": 0.00919398201499173,
      "loss": 0.2678,
      "num_input_tokens_seen": 27091136,
      "step": 35520
    },
    {
      "epoch": 73.85654885654886,
      "grad_norm": 0.0005127398180775344,
      "learning_rate": 0.009173687385751495,
      "loss": 0.2552,
      "num_input_tokens_seen": 27094976,
      "step": 35525
    },
    {
      "epoch": 73.86694386694387,
      "grad_norm": 0.00024458588450215757,
      "learning_rate": 0.009153414473698407,
      "loss": 0.2765,
      "num_input_tokens_seen": 27098688,
      "step": 35530
    },
    {
      "epoch": 73.87733887733887,
      "grad_norm": 0.000374671071767807,
      "learning_rate": 0.009133163281958784,
      "loss": 0.2609,
      "num_input_tokens_seen": 27102560,
      "step": 35535
    },
    {
      "epoch": 73.88773388773389,
      "grad_norm": 0.0005043321289122105,
      "learning_rate": 0.009112933813655627,
      "loss": 0.2696,
      "num_input_tokens_seen": 27106400,
      "step": 35540
    },
    {
      "epoch": 73.8981288981289,
      "grad_norm": 0.002000111388042569,
      "learning_rate": 0.009092726071908573,
      "loss": 0.2389,
      "num_input_tokens_seen": 27110240,
      "step": 35545
    },
    {
      "epoch": 73.9085239085239,
      "grad_norm": 0.00019785486801993102,
      "learning_rate": 0.0090725400598339,
      "loss": 0.2647,
      "num_input_tokens_seen": 27113888,
      "step": 35550
    },
    {
      "epoch": 73.91891891891892,
      "grad_norm": 0.00015866119065321982,
      "learning_rate": 0.009052375780544563,
      "loss": 0.2571,
      "num_input_tokens_seen": 27117600,
      "step": 35555
    },
    {
      "epoch": 73.92931392931393,
      "grad_norm": 0.00025204537087120116,
      "learning_rate": 0.009032233237150144,
      "loss": 0.2679,
      "num_input_tokens_seen": 27121280,
      "step": 35560
    },
    {
      "epoch": 73.93970893970894,
      "grad_norm": 0.0001546829444123432,
      "learning_rate": 0.009012112432756875,
      "loss": 0.2565,
      "num_input_tokens_seen": 27125184,
      "step": 35565
    },
    {
      "epoch": 73.95010395010395,
      "grad_norm": 0.00026677310233935714,
      "learning_rate": 0.008992013370467605,
      "loss": 0.2705,
      "num_input_tokens_seen": 27129088,
      "step": 35570
    },
    {
      "epoch": 73.96049896049897,
      "grad_norm": 0.0006169647094793618,
      "learning_rate": 0.008971936053381924,
      "loss": 0.2638,
      "num_input_tokens_seen": 27133024,
      "step": 35575
    },
    {
      "epoch": 73.97089397089397,
      "grad_norm": 0.0006587838288396597,
      "learning_rate": 0.008951880484595953,
      "loss": 0.277,
      "num_input_tokens_seen": 27136800,
      "step": 35580
    },
    {
      "epoch": 73.98128898128898,
      "grad_norm": 0.0009608623804524541,
      "learning_rate": 0.008931846667202552,
      "loss": 0.2309,
      "num_input_tokens_seen": 27140544,
      "step": 35585
    },
    {
      "epoch": 73.99168399168398,
      "grad_norm": 0.00029404196538962424,
      "learning_rate": 0.008911834604291152,
      "loss": 0.2647,
      "num_input_tokens_seen": 27144352,
      "step": 35590
    },
    {
      "epoch": 74.002079002079,
      "grad_norm": 0.000500818423461169,
      "learning_rate": 0.008891844298947882,
      "loss": 0.2463,
      "num_input_tokens_seen": 27147960,
      "step": 35595
    },
    {
      "epoch": 74.01247401247402,
      "grad_norm": 0.0007572741596959531,
      "learning_rate": 0.008871875754255508,
      "loss": 0.277,
      "num_input_tokens_seen": 27151800,
      "step": 35600
    },
    {
      "epoch": 74.01247401247402,
      "eval_loss": 0.24922853708267212,
      "eval_runtime": 13.3916,
      "eval_samples_per_second": 63.921,
      "eval_steps_per_second": 15.98,
      "num_input_tokens_seen": 27151800,
      "step": 35600
    },
    {
      "epoch": 74.02286902286902,
      "grad_norm": 0.0003930544189643115,
      "learning_rate": 0.008851928973293422,
      "loss": 0.2666,
      "num_input_tokens_seen": 27155736,
      "step": 35605
    },
    {
      "epoch": 74.03326403326403,
      "grad_norm": 0.00026405713288113475,
      "learning_rate": 0.00883200395913764,
      "loss": 0.2664,
      "num_input_tokens_seen": 27159672,
      "step": 35610
    },
    {
      "epoch": 74.04365904365905,
      "grad_norm": 0.000271029508439824,
      "learning_rate": 0.00881210071486091,
      "loss": 0.2648,
      "num_input_tokens_seen": 27163320,
      "step": 35615
    },
    {
      "epoch": 74.05405405405405,
      "grad_norm": 0.0003909572260454297,
      "learning_rate": 0.008792219243532505,
      "loss": 0.2635,
      "num_input_tokens_seen": 27167160,
      "step": 35620
    },
    {
      "epoch": 74.06444906444906,
      "grad_norm": 0.0005478140083141625,
      "learning_rate": 0.008772359548218428,
      "loss": 0.259,
      "num_input_tokens_seen": 27170968,
      "step": 35625
    },
    {
      "epoch": 74.07484407484408,
      "grad_norm": 0.0003560836485121399,
      "learning_rate": 0.008752521631981274,
      "loss": 0.2608,
      "num_input_tokens_seen": 27174776,
      "step": 35630
    },
    {
      "epoch": 74.08523908523908,
      "grad_norm": 0.0005666264914907515,
      "learning_rate": 0.008732705497880315,
      "loss": 0.2632,
      "num_input_tokens_seen": 27178808,
      "step": 35635
    },
    {
      "epoch": 74.0956340956341,
      "grad_norm": 0.0003599419433157891,
      "learning_rate": 0.008712911148971459,
      "loss": 0.2802,
      "num_input_tokens_seen": 27182520,
      "step": 35640
    },
    {
      "epoch": 74.10602910602911,
      "grad_norm": 0.00027648115064948797,
      "learning_rate": 0.008693138588307208,
      "loss": 0.2531,
      "num_input_tokens_seen": 27186168,
      "step": 35645
    },
    {
      "epoch": 74.11642411642411,
      "grad_norm": 0.0002439768722979352,
      "learning_rate": 0.008673387818936762,
      "loss": 0.2913,
      "num_input_tokens_seen": 27190040,
      "step": 35650
    },
    {
      "epoch": 74.12681912681913,
      "grad_norm": 0.0001416257000528276,
      "learning_rate": 0.008653658843905948,
      "loss": 0.2622,
      "num_input_tokens_seen": 27193848,
      "step": 35655
    },
    {
      "epoch": 74.13721413721414,
      "grad_norm": 0.00029875419568270445,
      "learning_rate": 0.0086339516662572,
      "loss": 0.2816,
      "num_input_tokens_seen": 27197752,
      "step": 35660
    },
    {
      "epoch": 74.14760914760915,
      "grad_norm": 0.00013812282122671604,
      "learning_rate": 0.008614266289029638,
      "loss": 0.2542,
      "num_input_tokens_seen": 27201688,
      "step": 35665
    },
    {
      "epoch": 74.15800415800416,
      "grad_norm": 0.00044145682477392256,
      "learning_rate": 0.008594602715258965,
      "loss": 0.2661,
      "num_input_tokens_seen": 27205368,
      "step": 35670
    },
    {
      "epoch": 74.16839916839916,
      "grad_norm": 0.0007312553352676332,
      "learning_rate": 0.008574960947977573,
      "loss": 0.2739,
      "num_input_tokens_seen": 27209240,
      "step": 35675
    },
    {
      "epoch": 74.17879417879418,
      "grad_norm": 0.0006273904000408947,
      "learning_rate": 0.008555340990214438,
      "loss": 0.2498,
      "num_input_tokens_seen": 27213048,
      "step": 35680
    },
    {
      "epoch": 74.1891891891892,
      "grad_norm": 0.00042457215022295713,
      "learning_rate": 0.008535742844995258,
      "loss": 0.261,
      "num_input_tokens_seen": 27216728,
      "step": 35685
    },
    {
      "epoch": 74.1995841995842,
      "grad_norm": 0.00018563997582532465,
      "learning_rate": 0.008516166515342266,
      "loss": 0.2599,
      "num_input_tokens_seen": 27220664,
      "step": 35690
    },
    {
      "epoch": 74.20997920997921,
      "grad_norm": 0.001080030226148665,
      "learning_rate": 0.008496612004274411,
      "loss": 0.2649,
      "num_input_tokens_seen": 27224600,
      "step": 35695
    },
    {
      "epoch": 74.22037422037423,
      "grad_norm": 0.0003102063783444464,
      "learning_rate": 0.008477079314807201,
      "loss": 0.2412,
      "num_input_tokens_seen": 27228312,
      "step": 35700
    },
    {
      "epoch": 74.23076923076923,
      "grad_norm": 0.00018369620374869555,
      "learning_rate": 0.008457568449952874,
      "loss": 0.2463,
      "num_input_tokens_seen": 27232088,
      "step": 35705
    },
    {
      "epoch": 74.24116424116424,
      "grad_norm": 0.000136344155180268,
      "learning_rate": 0.008438079412720189,
      "loss": 0.2591,
      "num_input_tokens_seen": 27235960,
      "step": 35710
    },
    {
      "epoch": 74.25155925155926,
      "grad_norm": 0.00017369123816024512,
      "learning_rate": 0.00841861220611466,
      "loss": 0.2742,
      "num_input_tokens_seen": 27239896,
      "step": 35715
    },
    {
      "epoch": 74.26195426195426,
      "grad_norm": 0.00031644431874156,
      "learning_rate": 0.008399166833138355,
      "loss": 0.2879,
      "num_input_tokens_seen": 27243736,
      "step": 35720
    },
    {
      "epoch": 74.27234927234927,
      "grad_norm": 0.0001117918873205781,
      "learning_rate": 0.008379743296789987,
      "loss": 0.2422,
      "num_input_tokens_seen": 27247576,
      "step": 35725
    },
    {
      "epoch": 74.28274428274429,
      "grad_norm": 0.0002494224754627794,
      "learning_rate": 0.008360341600064896,
      "loss": 0.2592,
      "num_input_tokens_seen": 27251320,
      "step": 35730
    },
    {
      "epoch": 74.29313929313929,
      "grad_norm": 0.0003176149621140212,
      "learning_rate": 0.008340961745955121,
      "loss": 0.2499,
      "num_input_tokens_seen": 27255160,
      "step": 35735
    },
    {
      "epoch": 74.3035343035343,
      "grad_norm": 0.0001895570894703269,
      "learning_rate": 0.008321603737449224,
      "loss": 0.2687,
      "num_input_tokens_seen": 27258936,
      "step": 35740
    },
    {
      "epoch": 74.31392931392931,
      "grad_norm": 0.0013831006363034248,
      "learning_rate": 0.008302267577532479,
      "loss": 0.2325,
      "num_input_tokens_seen": 27262776,
      "step": 35745
    },
    {
      "epoch": 74.32432432432432,
      "grad_norm": 0.0002702161727938801,
      "learning_rate": 0.008282953269186771,
      "loss": 0.2621,
      "num_input_tokens_seen": 27266680,
      "step": 35750
    },
    {
      "epoch": 74.33471933471934,
      "grad_norm": 0.0006112667615525424,
      "learning_rate": 0.008263660815390567,
      "loss": 0.2681,
      "num_input_tokens_seen": 27270296,
      "step": 35755
    },
    {
      "epoch": 74.34511434511434,
      "grad_norm": 0.0010118919890373945,
      "learning_rate": 0.008244390219119069,
      "loss": 0.2742,
      "num_input_tokens_seen": 27274104,
      "step": 35760
    },
    {
      "epoch": 74.35550935550935,
      "grad_norm": 0.000422335957409814,
      "learning_rate": 0.008225141483343967,
      "loss": 0.2533,
      "num_input_tokens_seen": 27277880,
      "step": 35765
    },
    {
      "epoch": 74.36590436590437,
      "grad_norm": 0.00032449059654027224,
      "learning_rate": 0.00820591461103372,
      "loss": 0.2684,
      "num_input_tokens_seen": 27281688,
      "step": 35770
    },
    {
      "epoch": 74.37629937629937,
      "grad_norm": 0.000574265664909035,
      "learning_rate": 0.008186709605153358,
      "loss": 0.2815,
      "num_input_tokens_seen": 27285496,
      "step": 35775
    },
    {
      "epoch": 74.38669438669439,
      "grad_norm": 0.00011316668678773567,
      "learning_rate": 0.008167526468664492,
      "loss": 0.258,
      "num_input_tokens_seen": 27289208,
      "step": 35780
    },
    {
      "epoch": 74.3970893970894,
      "grad_norm": 0.0003479458100628108,
      "learning_rate": 0.008148365204525443,
      "loss": 0.2571,
      "num_input_tokens_seen": 27292984,
      "step": 35785
    },
    {
      "epoch": 74.4074844074844,
      "grad_norm": 0.0009247698471881449,
      "learning_rate": 0.00812922581569106,
      "loss": 0.2665,
      "num_input_tokens_seen": 27296664,
      "step": 35790
    },
    {
      "epoch": 74.41787941787942,
      "grad_norm": 0.00012059063010383397,
      "learning_rate": 0.008110108305112934,
      "loss": 0.2482,
      "num_input_tokens_seen": 27300440,
      "step": 35795
    },
    {
      "epoch": 74.42827442827443,
      "grad_norm": 0.0003978035820182413,
      "learning_rate": 0.008091012675739223,
      "loss": 0.2486,
      "num_input_tokens_seen": 27304152,
      "step": 35800
    },
    {
      "epoch": 74.42827442827443,
      "eval_loss": 0.2478523552417755,
      "eval_runtime": 13.3972,
      "eval_samples_per_second": 63.894,
      "eval_steps_per_second": 15.973,
      "num_input_tokens_seen": 27304152,
      "step": 35800
    },
    {
      "epoch": 74.43866943866944,
      "grad_norm": 0.00023250629601534456,
      "learning_rate": 0.008071938930514671,
      "loss": 0.2332,
      "num_input_tokens_seen": 27307832,
      "step": 35805
    },
    {
      "epoch": 74.44906444906445,
      "grad_norm": 0.0001280247379327193,
      "learning_rate": 0.008052887072380726,
      "loss": 0.2913,
      "num_input_tokens_seen": 27311672,
      "step": 35810
    },
    {
      "epoch": 74.45945945945945,
      "grad_norm": 0.0006330225733108819,
      "learning_rate": 0.008033857104275437,
      "loss": 0.2752,
      "num_input_tokens_seen": 27315736,
      "step": 35815
    },
    {
      "epoch": 74.46985446985447,
      "grad_norm": 0.0016815828857943416,
      "learning_rate": 0.008014849029133424,
      "loss": 0.2735,
      "num_input_tokens_seen": 27319384,
      "step": 35820
    },
    {
      "epoch": 74.48024948024948,
      "grad_norm": 0.0002503742289263755,
      "learning_rate": 0.007995862849885975,
      "loss": 0.2281,
      "num_input_tokens_seen": 27323128,
      "step": 35825
    },
    {
      "epoch": 74.49064449064448,
      "grad_norm": 0.0003067964571528137,
      "learning_rate": 0.007976898569461032,
      "loss": 0.2824,
      "num_input_tokens_seen": 27327128,
      "step": 35830
    },
    {
      "epoch": 74.5010395010395,
      "grad_norm": 0.0007298548589460552,
      "learning_rate": 0.007957956190783088,
      "loss": 0.2702,
      "num_input_tokens_seen": 27330872,
      "step": 35835
    },
    {
      "epoch": 74.51143451143452,
      "grad_norm": 0.00034885347122326493,
      "learning_rate": 0.007939035716773324,
      "loss": 0.2668,
      "num_input_tokens_seen": 27334648,
      "step": 35840
    },
    {
      "epoch": 74.52182952182952,
      "grad_norm": 0.0004228349425829947,
      "learning_rate": 0.007920137150349487,
      "loss": 0.2624,
      "num_input_tokens_seen": 27338520,
      "step": 35845
    },
    {
      "epoch": 74.53222453222453,
      "grad_norm": 0.000176787783857435,
      "learning_rate": 0.007901260494425981,
      "loss": 0.2842,
      "num_input_tokens_seen": 27342424,
      "step": 35850
    },
    {
      "epoch": 74.54261954261955,
      "grad_norm": 0.0005410254816524684,
      "learning_rate": 0.007882405751913861,
      "loss": 0.2765,
      "num_input_tokens_seen": 27346232,
      "step": 35855
    },
    {
      "epoch": 74.55301455301455,
      "grad_norm": 0.0005940303090028465,
      "learning_rate": 0.007863572925720702,
      "loss": 0.2567,
      "num_input_tokens_seen": 27349720,
      "step": 35860
    },
    {
      "epoch": 74.56340956340956,
      "grad_norm": 0.0006077117286622524,
      "learning_rate": 0.007844762018750827,
      "loss": 0.2542,
      "num_input_tokens_seen": 27353560,
      "step": 35865
    },
    {
      "epoch": 74.57380457380458,
      "grad_norm": 0.0007424021605402231,
      "learning_rate": 0.007825973033905054,
      "loss": 0.2553,
      "num_input_tokens_seen": 27357368,
      "step": 35870
    },
    {
      "epoch": 74.58419958419958,
      "grad_norm": 0.0001844785874709487,
      "learning_rate": 0.007807205974080927,
      "loss": 0.2477,
      "num_input_tokens_seen": 27361304,
      "step": 35875
    },
    {
      "epoch": 74.5945945945946,
      "grad_norm": 0.00019676057854667306,
      "learning_rate": 0.007788460842172551,
      "loss": 0.2423,
      "num_input_tokens_seen": 27365112,
      "step": 35880
    },
    {
      "epoch": 74.60498960498961,
      "grad_norm": 0.00031554396264255047,
      "learning_rate": 0.0077697376410706285,
      "loss": 0.257,
      "num_input_tokens_seen": 27368920,
      "step": 35885
    },
    {
      "epoch": 74.61538461538461,
      "grad_norm": 0.0018940113950520754,
      "learning_rate": 0.007751036373662567,
      "loss": 0.2243,
      "num_input_tokens_seen": 27372600,
      "step": 35890
    },
    {
      "epoch": 74.62577962577963,
      "grad_norm": 0.0010034558363258839,
      "learning_rate": 0.00773235704283231,
      "loss": 0.2557,
      "num_input_tokens_seen": 27376568,
      "step": 35895
    },
    {
      "epoch": 74.63617463617463,
      "grad_norm": 0.0005155546241439879,
      "learning_rate": 0.007713699651460437,
      "loss": 0.2402,
      "num_input_tokens_seen": 27380376,
      "step": 35900
    },
    {
      "epoch": 74.64656964656965,
      "grad_norm": 0.0012954804114997387,
      "learning_rate": 0.007695064202424162,
      "loss": 0.2762,
      "num_input_tokens_seen": 27384024,
      "step": 35905
    },
    {
      "epoch": 74.65696465696466,
      "grad_norm": 0.00025558372726663947,
      "learning_rate": 0.007676450698597286,
      "loss": 0.2571,
      "num_input_tokens_seen": 27387736,
      "step": 35910
    },
    {
      "epoch": 74.66735966735966,
      "grad_norm": 0.00023000799410510808,
      "learning_rate": 0.007657859142850265,
      "loss": 0.2316,
      "num_input_tokens_seen": 27391576,
      "step": 35915
    },
    {
      "epoch": 74.67775467775468,
      "grad_norm": 0.001397228566929698,
      "learning_rate": 0.0076392895380501535,
      "loss": 0.3017,
      "num_input_tokens_seen": 27395384,
      "step": 35920
    },
    {
      "epoch": 74.6881496881497,
      "grad_norm": 0.0001082841627066955,
      "learning_rate": 0.007620741887060611,
      "loss": 0.2786,
      "num_input_tokens_seen": 27399352,
      "step": 35925
    },
    {
      "epoch": 74.6985446985447,
      "grad_norm": 0.00027650519041344523,
      "learning_rate": 0.007602216192741901,
      "loss": 0.2839,
      "num_input_tokens_seen": 27403288,
      "step": 35930
    },
    {
      "epoch": 74.70893970893971,
      "grad_norm": 0.0003521101316437125,
      "learning_rate": 0.007583712457950969,
      "loss": 0.2936,
      "num_input_tokens_seen": 27407224,
      "step": 35935
    },
    {
      "epoch": 74.71933471933473,
      "grad_norm": 0.0006344080320559442,
      "learning_rate": 0.007565230685541269,
      "loss": 0.2603,
      "num_input_tokens_seen": 27411064,
      "step": 35940
    },
    {
      "epoch": 74.72972972972973,
      "grad_norm": 0.00043430578080005944,
      "learning_rate": 0.007546770878362968,
      "loss": 0.2411,
      "num_input_tokens_seen": 27414808,
      "step": 35945
    },
    {
      "epoch": 74.74012474012474,
      "grad_norm": 0.00025257677771151066,
      "learning_rate": 0.0075283330392627405,
      "loss": 0.2412,
      "num_input_tokens_seen": 27418712,
      "step": 35950
    },
    {
      "epoch": 74.75051975051976,
      "grad_norm": 0.0006738262018188834,
      "learning_rate": 0.007509917171083979,
      "loss": 0.2301,
      "num_input_tokens_seen": 27422552,
      "step": 35955
    },
    {
      "epoch": 74.76091476091476,
      "grad_norm": 0.000822072965092957,
      "learning_rate": 0.007491523276666662,
      "loss": 0.267,
      "num_input_tokens_seen": 27426328,
      "step": 35960
    },
    {
      "epoch": 74.77130977130977,
      "grad_norm": 0.00027745121042244136,
      "learning_rate": 0.007473151358847318,
      "loss": 0.2645,
      "num_input_tokens_seen": 27430072,
      "step": 35965
    },
    {
      "epoch": 74.78170478170478,
      "grad_norm": 0.0002748681581579149,
      "learning_rate": 0.007454801420459117,
      "loss": 0.2651,
      "num_input_tokens_seen": 27433880,
      "step": 35970
    },
    {
      "epoch": 74.79209979209979,
      "grad_norm": 0.0006046561175026,
      "learning_rate": 0.0074364734643319105,
      "loss": 0.2885,
      "num_input_tokens_seen": 27437592,
      "step": 35975
    },
    {
      "epoch": 74.8024948024948,
      "grad_norm": 0.00027396451332606375,
      "learning_rate": 0.007418167493292022,
      "loss": 0.2618,
      "num_input_tokens_seen": 27441400,
      "step": 35980
    },
    {
      "epoch": 74.81288981288981,
      "grad_norm": 0.00034973659785464406,
      "learning_rate": 0.0073998835101625245,
      "loss": 0.2874,
      "num_input_tokens_seen": 27445080,
      "step": 35985
    },
    {
      "epoch": 74.82328482328482,
      "grad_norm": 0.00021457408729474992,
      "learning_rate": 0.007381621517762998,
      "loss": 0.2869,
      "num_input_tokens_seen": 27449048,
      "step": 35990
    },
    {
      "epoch": 74.83367983367984,
      "grad_norm": 0.0010342495515942574,
      "learning_rate": 0.007363381518909689,
      "loss": 0.2674,
      "num_input_tokens_seen": 27452984,
      "step": 35995
    },
    {
      "epoch": 74.84407484407484,
      "grad_norm": 9.518170554656535e-05,
      "learning_rate": 0.007345163516415448,
      "loss": 0.2598,
      "num_input_tokens_seen": 27456856,
      "step": 36000
    },
    {
      "epoch": 74.84407484407484,
      "eval_loss": 0.2507597804069519,
      "eval_runtime": 13.3904,
      "eval_samples_per_second": 63.927,
      "eval_steps_per_second": 15.982,
      "num_input_tokens_seen": 27456856,
      "step": 36000
    },
    {
      "epoch": 74.85446985446985,
      "grad_norm": 0.00029960134997963905,
      "learning_rate": 0.007326967513089693,
      "loss": 0.2433,
      "num_input_tokens_seen": 27460760,
      "step": 36005
    },
    {
      "epoch": 74.86486486486487,
      "grad_norm": 0.00023461967066396028,
      "learning_rate": 0.0073087935117384815,
      "loss": 0.2673,
      "num_input_tokens_seen": 27464728,
      "step": 36010
    },
    {
      "epoch": 74.87525987525987,
      "grad_norm": 0.00046880284207873046,
      "learning_rate": 0.007290641515164503,
      "loss": 0.2282,
      "num_input_tokens_seen": 27468600,
      "step": 36015
    },
    {
      "epoch": 74.88565488565489,
      "grad_norm": 0.0017863644752651453,
      "learning_rate": 0.007272511526166986,
      "loss": 0.2732,
      "num_input_tokens_seen": 27472440,
      "step": 36020
    },
    {
      "epoch": 74.8960498960499,
      "grad_norm": 0.00017287139780819416,
      "learning_rate": 0.0072544035475418265,
      "loss": 0.271,
      "num_input_tokens_seen": 27476248,
      "step": 36025
    },
    {
      "epoch": 74.9064449064449,
      "grad_norm": 0.0004917072947137058,
      "learning_rate": 0.007236317582081475,
      "loss": 0.265,
      "num_input_tokens_seen": 27480120,
      "step": 36030
    },
    {
      "epoch": 74.91683991683992,
      "grad_norm": 0.0001281521690543741,
      "learning_rate": 0.007218253632575066,
      "loss": 0.2537,
      "num_input_tokens_seen": 27484120,
      "step": 36035
    },
    {
      "epoch": 74.92723492723492,
      "grad_norm": 0.0003193325537722558,
      "learning_rate": 0.007200211701808223,
      "loss": 0.2564,
      "num_input_tokens_seen": 27487800,
      "step": 36040
    },
    {
      "epoch": 74.93762993762994,
      "grad_norm": 0.00043910779641009867,
      "learning_rate": 0.007182191792563286,
      "loss": 0.2801,
      "num_input_tokens_seen": 27491544,
      "step": 36045
    },
    {
      "epoch": 74.94802494802495,
      "grad_norm": 0.00034964369842782617,
      "learning_rate": 0.0071641939076191145,
      "loss": 0.2752,
      "num_input_tokens_seen": 27495352,
      "step": 36050
    },
    {
      "epoch": 74.95841995841995,
      "grad_norm": 0.00023063960543368012,
      "learning_rate": 0.007146218049751257,
      "loss": 0.2686,
      "num_input_tokens_seen": 27499416,
      "step": 36055
    },
    {
      "epoch": 74.96881496881497,
      "grad_norm": 0.0002742695214692503,
      "learning_rate": 0.0071282642217317775,
      "loss": 0.2578,
      "num_input_tokens_seen": 27503224,
      "step": 36060
    },
    {
      "epoch": 74.97920997920998,
      "grad_norm": 0.003657560097053647,
      "learning_rate": 0.007110332426329396,
      "loss": 0.2863,
      "num_input_tokens_seen": 27507224,
      "step": 36065
    },
    {
      "epoch": 74.98960498960498,
      "grad_norm": 0.0004119256045669317,
      "learning_rate": 0.007092422666309417,
      "loss": 0.2443,
      "num_input_tokens_seen": 27511160,
      "step": 36070
    },
    {
      "epoch": 75.0,
      "grad_norm": 0.00017305923392996192,
      "learning_rate": 0.0070745349444337295,
      "loss": 0.2649,
      "num_input_tokens_seen": 27514792,
      "step": 36075
    },
    {
      "epoch": 75.01039501039502,
      "grad_norm": 0.0001678390835877508,
      "learning_rate": 0.007056669263460913,
      "loss": 0.2543,
      "num_input_tokens_seen": 27518632,
      "step": 36080
    },
    {
      "epoch": 75.02079002079002,
      "grad_norm": 0.0006255152984522283,
      "learning_rate": 0.007038825626145995,
      "loss": 0.2587,
      "num_input_tokens_seen": 27522408,
      "step": 36085
    },
    {
      "epoch": 75.03118503118503,
      "grad_norm": 0.0003666336415335536,
      "learning_rate": 0.007021004035240724,
      "loss": 0.2499,
      "num_input_tokens_seen": 27526280,
      "step": 36090
    },
    {
      "epoch": 75.04158004158005,
      "grad_norm": 0.00011675449059111997,
      "learning_rate": 0.007003204493493453,
      "loss": 0.2808,
      "num_input_tokens_seen": 27530088,
      "step": 36095
    },
    {
      "epoch": 75.05197505197505,
      "grad_norm": 0.0002810081350617111,
      "learning_rate": 0.006985427003649036,
      "loss": 0.2589,
      "num_input_tokens_seen": 27534088,
      "step": 36100
    },
    {
      "epoch": 75.06237006237006,
      "grad_norm": 0.00029885617550462484,
      "learning_rate": 0.006967671568449013,
      "loss": 0.2763,
      "num_input_tokens_seen": 27537832,
      "step": 36105
    },
    {
      "epoch": 75.07276507276508,
      "grad_norm": 0.0006143073551356792,
      "learning_rate": 0.006949938190631511,
      "loss": 0.2445,
      "num_input_tokens_seen": 27541512,
      "step": 36110
    },
    {
      "epoch": 75.08316008316008,
      "grad_norm": 0.0012581292539834976,
      "learning_rate": 0.0069322268729311905,
      "loss": 0.2561,
      "num_input_tokens_seen": 27545448,
      "step": 36115
    },
    {
      "epoch": 75.0935550935551,
      "grad_norm": 0.0007887134561315179,
      "learning_rate": 0.006914537618079403,
      "loss": 0.2629,
      "num_input_tokens_seen": 27549224,
      "step": 36120
    },
    {
      "epoch": 75.1039501039501,
      "grad_norm": 0.000439716357504949,
      "learning_rate": 0.006896870428804031,
      "loss": 0.2571,
      "num_input_tokens_seen": 27553096,
      "step": 36125
    },
    {
      "epoch": 75.11434511434511,
      "grad_norm": 0.00021960939920973033,
      "learning_rate": 0.006879225307829595,
      "loss": 0.2726,
      "num_input_tokens_seen": 27556968,
      "step": 36130
    },
    {
      "epoch": 75.12474012474013,
      "grad_norm": 0.0004342082538641989,
      "learning_rate": 0.00686160225787717,
      "loss": 0.2434,
      "num_input_tokens_seen": 27560744,
      "step": 36135
    },
    {
      "epoch": 75.13513513513513,
      "grad_norm": 0.0001964966650120914,
      "learning_rate": 0.006844001281664463,
      "loss": 0.2775,
      "num_input_tokens_seen": 27564584,
      "step": 36140
    },
    {
      "epoch": 75.14553014553015,
      "grad_norm": 0.0001552668836666271,
      "learning_rate": 0.006826422381905789,
      "loss": 0.2856,
      "num_input_tokens_seen": 27568424,
      "step": 36145
    },
    {
      "epoch": 75.15592515592516,
      "grad_norm": 0.0002786397817544639,
      "learning_rate": 0.006808865561311994,
      "loss": 0.2756,
      "num_input_tokens_seen": 27572424,
      "step": 36150
    },
    {
      "epoch": 75.16632016632016,
      "grad_norm": 0.0004274522070772946,
      "learning_rate": 0.00679133082259058,
      "loss": 0.2648,
      "num_input_tokens_seen": 27576136,
      "step": 36155
    },
    {
      "epoch": 75.17671517671518,
      "grad_norm": 0.00017544864385854453,
      "learning_rate": 0.00677381816844565,
      "loss": 0.256,
      "num_input_tokens_seen": 27580008,
      "step": 36160
    },
    {
      "epoch": 75.18711018711019,
      "grad_norm": 0.0001505874388385564,
      "learning_rate": 0.0067563276015778434,
      "loss": 0.2581,
      "num_input_tokens_seen": 27583688,
      "step": 36165
    },
    {
      "epoch": 75.1975051975052,
      "grad_norm": 0.0006480725714936852,
      "learning_rate": 0.006738859124684437,
      "loss": 0.2923,
      "num_input_tokens_seen": 27587592,
      "step": 36170
    },
    {
      "epoch": 75.20790020790021,
      "grad_norm": 0.00019662445993162692,
      "learning_rate": 0.006721412740459259,
      "loss": 0.269,
      "num_input_tokens_seen": 27591464,
      "step": 36175
    },
    {
      "epoch": 75.21829521829522,
      "grad_norm": 0.0014681684551760554,
      "learning_rate": 0.006703988451592824,
      "loss": 0.271,
      "num_input_tokens_seen": 27595304,
      "step": 36180
    },
    {
      "epoch": 75.22869022869023,
      "grad_norm": 0.000167108591995202,
      "learning_rate": 0.006686586260772114,
      "loss": 0.2595,
      "num_input_tokens_seen": 27599112,
      "step": 36185
    },
    {
      "epoch": 75.23908523908524,
      "grad_norm": 6.415171083062887e-05,
      "learning_rate": 0.006669206170680819,
      "loss": 0.2561,
      "num_input_tokens_seen": 27602824,
      "step": 36190
    },
    {
      "epoch": 75.24948024948024,
      "grad_norm": 0.00022838514996692538,
      "learning_rate": 0.0066518481839991095,
      "loss": 0.2681,
      "num_input_tokens_seen": 27606632,
      "step": 36195
    },
    {
      "epoch": 75.25987525987526,
      "grad_norm": 0.00026006813277490437,
      "learning_rate": 0.006634512303403861,
      "loss": 0.2815,
      "num_input_tokens_seen": 27610376,
      "step": 36200
    },
    {
      "epoch": 75.25987525987526,
      "eval_loss": 0.25133857131004333,
      "eval_runtime": 13.4173,
      "eval_samples_per_second": 63.798,
      "eval_steps_per_second": 15.95,
      "num_input_tokens_seen": 27610376,
      "step": 36200
    },
    {
      "epoch": 75.27027027027027,
      "grad_norm": 0.00018262698722537607,
      "learning_rate": 0.0066171985315684355,
      "loss": 0.2776,
      "num_input_tokens_seen": 27613992,
      "step": 36205
    },
    {
      "epoch": 75.28066528066527,
      "grad_norm": 0.00017371807189192623,
      "learning_rate": 0.0065999068711628806,
      "loss": 0.2578,
      "num_input_tokens_seen": 27617832,
      "step": 36210
    },
    {
      "epoch": 75.29106029106029,
      "grad_norm": 0.00036904754233546555,
      "learning_rate": 0.0065826373248537295,
      "loss": 0.2678,
      "num_input_tokens_seen": 27621672,
      "step": 36215
    },
    {
      "epoch": 75.3014553014553,
      "grad_norm": 0.00015767717559356242,
      "learning_rate": 0.006565389895304218,
      "loss": 0.2624,
      "num_input_tokens_seen": 27625576,
      "step": 36220
    },
    {
      "epoch": 75.3118503118503,
      "grad_norm": 0.0011044838465750217,
      "learning_rate": 0.006548164585174104,
      "loss": 0.2579,
      "num_input_tokens_seen": 27629544,
      "step": 36225
    },
    {
      "epoch": 75.32224532224532,
      "grad_norm": 0.0007148599252104759,
      "learning_rate": 0.006530961397119728,
      "loss": 0.2892,
      "num_input_tokens_seen": 27633480,
      "step": 36230
    },
    {
      "epoch": 75.33264033264034,
      "grad_norm": 0.0002838864747900516,
      "learning_rate": 0.00651378033379405,
      "loss": 0.2627,
      "num_input_tokens_seen": 27637256,
      "step": 36235
    },
    {
      "epoch": 75.34303534303534,
      "grad_norm": 0.00021961372112855315,
      "learning_rate": 0.006496621397846619,
      "loss": 0.2529,
      "num_input_tokens_seen": 27641000,
      "step": 36240
    },
    {
      "epoch": 75.35343035343035,
      "grad_norm": 0.0006204022793099284,
      "learning_rate": 0.006479484591923518,
      "loss": 0.2718,
      "num_input_tokens_seen": 27644904,
      "step": 36245
    },
    {
      "epoch": 75.36382536382537,
      "grad_norm": 0.00027774160844273865,
      "learning_rate": 0.006462369918667515,
      "loss": 0.2632,
      "num_input_tokens_seen": 27648808,
      "step": 36250
    },
    {
      "epoch": 75.37422037422037,
      "grad_norm": 0.0003736469952855259,
      "learning_rate": 0.006445277380717851,
      "loss": 0.2736,
      "num_input_tokens_seen": 27652552,
      "step": 36255
    },
    {
      "epoch": 75.38461538461539,
      "grad_norm": 0.00017816714535001665,
      "learning_rate": 0.006428206980710466,
      "loss": 0.2652,
      "num_input_tokens_seen": 27656264,
      "step": 36260
    },
    {
      "epoch": 75.39501039501039,
      "grad_norm": 0.0005649793311022222,
      "learning_rate": 0.006411158721277788,
      "loss": 0.2754,
      "num_input_tokens_seen": 27659976,
      "step": 36265
    },
    {
      "epoch": 75.4054054054054,
      "grad_norm": 0.0002725072263274342,
      "learning_rate": 0.00639413260504888,
      "loss": 0.25,
      "num_input_tokens_seen": 27663784,
      "step": 36270
    },
    {
      "epoch": 75.41580041580042,
      "grad_norm": 0.00035435458994470537,
      "learning_rate": 0.006377128634649376,
      "loss": 0.2588,
      "num_input_tokens_seen": 27667560,
      "step": 36275
    },
    {
      "epoch": 75.42619542619542,
      "grad_norm": 0.000293812743620947,
      "learning_rate": 0.006360146812701528,
      "loss": 0.2655,
      "num_input_tokens_seen": 27671176,
      "step": 36280
    },
    {
      "epoch": 75.43659043659044,
      "grad_norm": 0.0002869760792236775,
      "learning_rate": 0.006343187141824125,
      "loss": 0.2601,
      "num_input_tokens_seen": 27675080,
      "step": 36285
    },
    {
      "epoch": 75.44698544698545,
      "grad_norm": 0.00024739388027228415,
      "learning_rate": 0.00632624962463259,
      "loss": 0.2588,
      "num_input_tokens_seen": 27678888,
      "step": 36290
    },
    {
      "epoch": 75.45738045738045,
      "grad_norm": 5.570205394178629e-05,
      "learning_rate": 0.006309334263738853,
      "loss": 0.251,
      "num_input_tokens_seen": 27682760,
      "step": 36295
    },
    {
      "epoch": 75.46777546777547,
      "grad_norm": 0.00031986256362870336,
      "learning_rate": 0.006292441061751508,
      "loss": 0.2666,
      "num_input_tokens_seen": 27686728,
      "step": 36300
    },
    {
      "epoch": 75.47817047817048,
      "grad_norm": 0.0002962035941891372,
      "learning_rate": 0.0062755700212757054,
      "loss": 0.2541,
      "num_input_tokens_seen": 27690568,
      "step": 36305
    },
    {
      "epoch": 75.48856548856548,
      "grad_norm": 0.0005598647985607386,
      "learning_rate": 0.006258721144913148,
      "loss": 0.2476,
      "num_input_tokens_seen": 27694344,
      "step": 36310
    },
    {
      "epoch": 75.4989604989605,
      "grad_norm": 0.00040435505798086524,
      "learning_rate": 0.0062418944352621575,
      "loss": 0.2528,
      "num_input_tokens_seen": 27698280,
      "step": 36315
    },
    {
      "epoch": 75.50935550935552,
      "grad_norm": 0.0006754234782420099,
      "learning_rate": 0.0062250898949176405,
      "loss": 0.2559,
      "num_input_tokens_seen": 27702152,
      "step": 36320
    },
    {
      "epoch": 75.51975051975052,
      "grad_norm": 7.313093374250457e-05,
      "learning_rate": 0.006208307526471041,
      "loss": 0.2943,
      "num_input_tokens_seen": 27706056,
      "step": 36325
    },
    {
      "epoch": 75.53014553014553,
      "grad_norm": 0.00019297453400213271,
      "learning_rate": 0.006191547332510405,
      "loss": 0.2818,
      "num_input_tokens_seen": 27709896,
      "step": 36330
    },
    {
      "epoch": 75.54054054054055,
      "grad_norm": 0.0007400016766041517,
      "learning_rate": 0.006174809315620416,
      "loss": 0.2651,
      "num_input_tokens_seen": 27713864,
      "step": 36335
    },
    {
      "epoch": 75.55093555093555,
      "grad_norm": 0.0002630755479913205,
      "learning_rate": 0.00615809347838221,
      "loss": 0.2815,
      "num_input_tokens_seen": 27717768,
      "step": 36340
    },
    {
      "epoch": 75.56133056133056,
      "grad_norm": 0.00019092790898866951,
      "learning_rate": 0.006141399823373655,
      "loss": 0.2672,
      "num_input_tokens_seen": 27721544,
      "step": 36345
    },
    {
      "epoch": 75.57172557172557,
      "grad_norm": 0.00023187852639239281,
      "learning_rate": 0.0061247283531690455,
      "loss": 0.2683,
      "num_input_tokens_seen": 27725160,
      "step": 36350
    },
    {
      "epoch": 75.58212058212058,
      "grad_norm": 0.00012423927546478808,
      "learning_rate": 0.0061080790703393895,
      "loss": 0.2723,
      "num_input_tokens_seen": 27728968,
      "step": 36355
    },
    {
      "epoch": 75.5925155925156,
      "grad_norm": 0.00019474186410661787,
      "learning_rate": 0.006091451977452217,
      "loss": 0.2723,
      "num_input_tokens_seen": 27732840,
      "step": 36360
    },
    {
      "epoch": 75.6029106029106,
      "grad_norm": 0.0007714293315075338,
      "learning_rate": 0.00607484707707161,
      "loss": 0.2618,
      "num_input_tokens_seen": 27736520,
      "step": 36365
    },
    {
      "epoch": 75.61330561330561,
      "grad_norm": 0.00013165008567739278,
      "learning_rate": 0.006058264371758254,
      "loss": 0.2649,
      "num_input_tokens_seen": 27740456,
      "step": 36370
    },
    {
      "epoch": 75.62370062370063,
      "grad_norm": 0.0002822806127369404,
      "learning_rate": 0.00604170386406942,
      "loss": 0.2546,
      "num_input_tokens_seen": 27744232,
      "step": 36375
    },
    {
      "epoch": 75.63409563409563,
      "grad_norm": 0.00011422712850617245,
      "learning_rate": 0.006025165556558931,
      "loss": 0.2567,
      "num_input_tokens_seen": 27748072,
      "step": 36380
    },
    {
      "epoch": 75.64449064449065,
      "grad_norm": 0.00016472676361445338,
      "learning_rate": 0.006008649451777248,
      "loss": 0.2568,
      "num_input_tokens_seen": 27751816,
      "step": 36385
    },
    {
      "epoch": 75.65488565488566,
      "grad_norm": 0.0005685825599357486,
      "learning_rate": 0.005992155552271283,
      "loss": 0.2592,
      "num_input_tokens_seen": 27755464,
      "step": 36390
    },
    {
      "epoch": 75.66528066528066,
      "grad_norm": 0.00030456818058155477,
      "learning_rate": 0.005975683860584685,
      "loss": 0.2653,
      "num_input_tokens_seen": 27759272,
      "step": 36395
    },
    {
      "epoch": 75.67567567567568,
      "grad_norm": 0.00042223164928145707,
      "learning_rate": 0.0059592343792575385,
      "loss": 0.2609,
      "num_input_tokens_seen": 27762984,
      "step": 36400
    },
    {
      "epoch": 75.67567567567568,
      "eval_loss": 0.24879048764705658,
      "eval_runtime": 13.3964,
      "eval_samples_per_second": 63.898,
      "eval_steps_per_second": 15.974,
      "num_input_tokens_seen": 27762984,
      "step": 36400
    },
    {
      "epoch": 75.68607068607069,
      "grad_norm": 0.0002706471714191139,
      "learning_rate": 0.0059428071108265975,
      "loss": 0.2297,
      "num_input_tokens_seen": 27766728,
      "step": 36405
    },
    {
      "epoch": 75.6964656964657,
      "grad_norm": 0.00012597993190865964,
      "learning_rate": 0.005926402057825136,
      "loss": 0.2712,
      "num_input_tokens_seen": 27770568,
      "step": 36410
    },
    {
      "epoch": 75.70686070686071,
      "grad_norm": 0.0002018495579250157,
      "learning_rate": 0.005910019222782997,
      "loss": 0.2606,
      "num_input_tokens_seen": 27774344,
      "step": 36415
    },
    {
      "epoch": 75.71725571725571,
      "grad_norm": 0.0010055502643808722,
      "learning_rate": 0.005893658608226643,
      "loss": 0.2451,
      "num_input_tokens_seen": 27778216,
      "step": 36420
    },
    {
      "epoch": 75.72765072765073,
      "grad_norm": 0.0005572371883317828,
      "learning_rate": 0.0058773202166791045,
      "loss": 0.2745,
      "num_input_tokens_seen": 27782056,
      "step": 36425
    },
    {
      "epoch": 75.73804573804574,
      "grad_norm": 0.00024897028924897313,
      "learning_rate": 0.005861004050659918,
      "loss": 0.2695,
      "num_input_tokens_seen": 27785992,
      "step": 36430
    },
    {
      "epoch": 75.74844074844074,
      "grad_norm": 0.00016537195187993348,
      "learning_rate": 0.005844710112685286,
      "loss": 0.2418,
      "num_input_tokens_seen": 27789800,
      "step": 36435
    },
    {
      "epoch": 75.75883575883576,
      "grad_norm": 0.00029151991475373507,
      "learning_rate": 0.005828438405267933,
      "loss": 0.2698,
      "num_input_tokens_seen": 27793544,
      "step": 36440
    },
    {
      "epoch": 75.76923076923077,
      "grad_norm": 0.00041944990516640246,
      "learning_rate": 0.00581218893091715,
      "loss": 0.2549,
      "num_input_tokens_seen": 27797416,
      "step": 36445
    },
    {
      "epoch": 75.77962577962577,
      "grad_norm": 0.0006753249326720834,
      "learning_rate": 0.005795961692138801,
      "loss": 0.2765,
      "num_input_tokens_seen": 27801416,
      "step": 36450
    },
    {
      "epoch": 75.79002079002079,
      "grad_norm": 9.621229401091114e-05,
      "learning_rate": 0.00577975669143535,
      "loss": 0.2506,
      "num_input_tokens_seen": 27805288,
      "step": 36455
    },
    {
      "epoch": 75.8004158004158,
      "grad_norm": 0.0003091284306719899,
      "learning_rate": 0.005763573931305782,
      "loss": 0.2711,
      "num_input_tokens_seen": 27809096,
      "step": 36460
    },
    {
      "epoch": 75.8108108108108,
      "grad_norm": 0.00020964314171578735,
      "learning_rate": 0.005747413414245733,
      "loss": 0.2481,
      "num_input_tokens_seen": 27812776,
      "step": 36465
    },
    {
      "epoch": 75.82120582120582,
      "grad_norm": 0.0003540139296092093,
      "learning_rate": 0.005731275142747294,
      "loss": 0.2791,
      "num_input_tokens_seen": 27816552,
      "step": 36470
    },
    {
      "epoch": 75.83160083160084,
      "grad_norm": 0.0004429105611052364,
      "learning_rate": 0.005715159119299256,
      "loss": 0.2436,
      "num_input_tokens_seen": 27820488,
      "step": 36475
    },
    {
      "epoch": 75.84199584199584,
      "grad_norm": 0.00015174214786384255,
      "learning_rate": 0.005699065346386867,
      "loss": 0.2398,
      "num_input_tokens_seen": 27824328,
      "step": 36480
    },
    {
      "epoch": 75.85239085239085,
      "grad_norm": 0.0002160083968192339,
      "learning_rate": 0.0056829938264919885,
      "loss": 0.2443,
      "num_input_tokens_seen": 27828008,
      "step": 36485
    },
    {
      "epoch": 75.86278586278586,
      "grad_norm": 0.00024172312987502664,
      "learning_rate": 0.005666944562093074,
      "loss": 0.252,
      "num_input_tokens_seen": 27831976,
      "step": 36490
    },
    {
      "epoch": 75.87318087318087,
      "grad_norm": 6.766623846488073e-05,
      "learning_rate": 0.005650917555665108,
      "loss": 0.2602,
      "num_input_tokens_seen": 27835656,
      "step": 36495
    },
    {
      "epoch": 75.88357588357589,
      "grad_norm": 0.00022125743271317333,
      "learning_rate": 0.005634912809679632,
      "loss": 0.2396,
      "num_input_tokens_seen": 27839304,
      "step": 36500
    },
    {
      "epoch": 75.89397089397089,
      "grad_norm": 0.00020492971816565841,
      "learning_rate": 0.005618930326604854,
      "loss": 0.2702,
      "num_input_tokens_seen": 27843112,
      "step": 36505
    },
    {
      "epoch": 75.9043659043659,
      "grad_norm": 0.0004243365256115794,
      "learning_rate": 0.005602970108905386,
      "loss": 0.2401,
      "num_input_tokens_seen": 27846760,
      "step": 36510
    },
    {
      "epoch": 75.91476091476092,
      "grad_norm": 0.00023723937920294702,
      "learning_rate": 0.005587032159042543,
      "loss": 0.2826,
      "num_input_tokens_seen": 27850568,
      "step": 36515
    },
    {
      "epoch": 75.92515592515592,
      "grad_norm": 0.00017458625370636582,
      "learning_rate": 0.005571116479474158,
      "loss": 0.2568,
      "num_input_tokens_seen": 27854280,
      "step": 36520
    },
    {
      "epoch": 75.93555093555094,
      "grad_norm": 8.346794493263587e-05,
      "learning_rate": 0.005555223072654619,
      "loss": 0.2406,
      "num_input_tokens_seen": 27858184,
      "step": 36525
    },
    {
      "epoch": 75.94594594594595,
      "grad_norm": 0.00020504332496784627,
      "learning_rate": 0.005539351941034881,
      "loss": 0.2794,
      "num_input_tokens_seen": 27861928,
      "step": 36530
    },
    {
      "epoch": 75.95634095634095,
      "grad_norm": 0.0004236344830133021,
      "learning_rate": 0.0055235030870624865,
      "loss": 0.2405,
      "num_input_tokens_seen": 27865864,
      "step": 36535
    },
    {
      "epoch": 75.96673596673597,
      "grad_norm": 0.0004484095552470535,
      "learning_rate": 0.005507676513181514,
      "loss": 0.2711,
      "num_input_tokens_seen": 27869768,
      "step": 36540
    },
    {
      "epoch": 75.97713097713098,
      "grad_norm": 0.0002650415408425033,
      "learning_rate": 0.005491872221832628,
      "loss": 0.2589,
      "num_input_tokens_seen": 27873608,
      "step": 36545
    },
    {
      "epoch": 75.98752598752598,
      "grad_norm": 0.00033283577067777514,
      "learning_rate": 0.005476090215453061,
      "loss": 0.2771,
      "num_input_tokens_seen": 27877352,
      "step": 36550
    },
    {
      "epoch": 75.997920997921,
      "grad_norm": 0.0003396625688765198,
      "learning_rate": 0.0054603304964765675,
      "loss": 0.2373,
      "num_input_tokens_seen": 27881128,
      "step": 36555
    },
    {
      "epoch": 76.00831600831602,
      "grad_norm": 0.00016951891302596778,
      "learning_rate": 0.005444593067333519,
      "loss": 0.2989,
      "num_input_tokens_seen": 27884944,
      "step": 36560
    },
    {
      "epoch": 76.01871101871102,
      "grad_norm": 0.0002527138276491314,
      "learning_rate": 0.00542887793045081,
      "loss": 0.2461,
      "num_input_tokens_seen": 27888784,
      "step": 36565
    },
    {
      "epoch": 76.02910602910603,
      "grad_norm": 0.0002157313865609467,
      "learning_rate": 0.005413185088251932,
      "loss": 0.2476,
      "num_input_tokens_seen": 27892752,
      "step": 36570
    },
    {
      "epoch": 76.03950103950103,
      "grad_norm": 0.0003957070875912905,
      "learning_rate": 0.005397514543156884,
      "loss": 0.249,
      "num_input_tokens_seen": 27896496,
      "step": 36575
    },
    {
      "epoch": 76.04989604989605,
      "grad_norm": 9.086212230613455e-05,
      "learning_rate": 0.0053818662975822825,
      "loss": 0.2527,
      "num_input_tokens_seen": 27900208,
      "step": 36580
    },
    {
      "epoch": 76.06029106029106,
      "grad_norm": 0.00037114444421604276,
      "learning_rate": 0.005366240353941315,
      "loss": 0.2681,
      "num_input_tokens_seen": 27903952,
      "step": 36585
    },
    {
      "epoch": 76.07068607068607,
      "grad_norm": 0.0007193362107500434,
      "learning_rate": 0.005350636714643636,
      "loss": 0.2842,
      "num_input_tokens_seen": 27907760,
      "step": 36590
    },
    {
      "epoch": 76.08108108108108,
      "grad_norm": 0.0010902669746428728,
      "learning_rate": 0.005335055382095555,
      "loss": 0.2543,
      "num_input_tokens_seen": 27911568,
      "step": 36595
    },
    {
      "epoch": 76.0914760914761,
      "grad_norm": 9.507292998023331e-05,
      "learning_rate": 0.005319496358699915,
      "loss": 0.2621,
      "num_input_tokens_seen": 27915504,
      "step": 36600
    },
    {
      "epoch": 76.0914760914761,
      "eval_loss": 0.2486761212348938,
      "eval_runtime": 13.3627,
      "eval_samples_per_second": 64.059,
      "eval_steps_per_second": 16.015,
      "num_input_tokens_seen": 27915504,
      "step": 36600
    },
    {
      "epoch": 76.1018711018711,
      "grad_norm": 0.000569100200664252,
      "learning_rate": 0.005303959646856099,
      "loss": 0.2455,
      "num_input_tokens_seen": 27919312,
      "step": 36605
    },
    {
      "epoch": 76.11226611226611,
      "grad_norm": 0.0001948667486431077,
      "learning_rate": 0.005288445248960089,
      "loss": 0.2289,
      "num_input_tokens_seen": 27923152,
      "step": 36610
    },
    {
      "epoch": 76.12266112266113,
      "grad_norm": 0.0006906175403855741,
      "learning_rate": 0.005272953167404354,
      "loss": 0.2763,
      "num_input_tokens_seen": 27926960,
      "step": 36615
    },
    {
      "epoch": 76.13305613305613,
      "grad_norm": 0.0004480722709558904,
      "learning_rate": 0.005257483404578017,
      "loss": 0.2694,
      "num_input_tokens_seen": 27930960,
      "step": 36620
    },
    {
      "epoch": 76.14345114345114,
      "grad_norm": 8.20136847323738e-05,
      "learning_rate": 0.0052420359628666865,
      "loss": 0.2696,
      "num_input_tokens_seen": 27934896,
      "step": 36625
    },
    {
      "epoch": 76.15384615384616,
      "grad_norm": 0.00019306126341689378,
      "learning_rate": 0.00522661084465254,
      "loss": 0.2547,
      "num_input_tokens_seen": 27938672,
      "step": 36630
    },
    {
      "epoch": 76.16424116424116,
      "grad_norm": 0.0002487867895979434,
      "learning_rate": 0.005211208052314326,
      "loss": 0.2666,
      "num_input_tokens_seen": 27942512,
      "step": 36635
    },
    {
      "epoch": 76.17463617463618,
      "grad_norm": 0.0005444425041787326,
      "learning_rate": 0.005195827588227391,
      "loss": 0.2573,
      "num_input_tokens_seen": 27946448,
      "step": 36640
    },
    {
      "epoch": 76.18503118503118,
      "grad_norm": 0.0005136593244969845,
      "learning_rate": 0.0051804694547635255,
      "loss": 0.2542,
      "num_input_tokens_seen": 27950192,
      "step": 36645
    },
    {
      "epoch": 76.1954261954262,
      "grad_norm": 0.00027354041230864823,
      "learning_rate": 0.005165133654291232,
      "loss": 0.2826,
      "num_input_tokens_seen": 27953872,
      "step": 36650
    },
    {
      "epoch": 76.20582120582121,
      "grad_norm": 0.0004595712525770068,
      "learning_rate": 0.005149820189175402,
      "loss": 0.2903,
      "num_input_tokens_seen": 27957584,
      "step": 36655
    },
    {
      "epoch": 76.21621621621621,
      "grad_norm": 0.00021985406056046486,
      "learning_rate": 0.005134529061777598,
      "loss": 0.2653,
      "num_input_tokens_seen": 27961424,
      "step": 36660
    },
    {
      "epoch": 76.22661122661123,
      "grad_norm": 0.00019717963004950434,
      "learning_rate": 0.005119260274455933,
      "loss": 0.2714,
      "num_input_tokens_seen": 27965264,
      "step": 36665
    },
    {
      "epoch": 76.23700623700624,
      "grad_norm": 0.00010775313421618193,
      "learning_rate": 0.005104013829565007,
      "loss": 0.2554,
      "num_input_tokens_seen": 27969104,
      "step": 36670
    },
    {
      "epoch": 76.24740124740124,
      "grad_norm": 0.00012310956662986428,
      "learning_rate": 0.005088789729456006,
      "loss": 0.276,
      "num_input_tokens_seen": 27973040,
      "step": 36675
    },
    {
      "epoch": 76.25779625779626,
      "grad_norm": 0.0009582063648849726,
      "learning_rate": 0.005073587976476735,
      "loss": 0.2711,
      "num_input_tokens_seen": 27976848,
      "step": 36680
    },
    {
      "epoch": 76.26819126819127,
      "grad_norm": 0.00040995379094965756,
      "learning_rate": 0.005058408572971418,
      "loss": 0.2473,
      "num_input_tokens_seen": 27980688,
      "step": 36685
    },
    {
      "epoch": 76.27858627858627,
      "grad_norm": 0.00039800367085263133,
      "learning_rate": 0.005043251521280983,
      "loss": 0.2542,
      "num_input_tokens_seen": 27984400,
      "step": 36690
    },
    {
      "epoch": 76.28898128898129,
      "grad_norm": 0.000407436047680676,
      "learning_rate": 0.005028116823742795,
      "loss": 0.2267,
      "num_input_tokens_seen": 27988080,
      "step": 36695
    },
    {
      "epoch": 76.2993762993763,
      "grad_norm": 0.00042980213765986264,
      "learning_rate": 0.005013004482690819,
      "loss": 0.2489,
      "num_input_tokens_seen": 27991856,
      "step": 36700
    },
    {
      "epoch": 76.3097713097713,
      "grad_norm": 0.00021001654386054724,
      "learning_rate": 0.0049979145004555746,
      "loss": 0.2494,
      "num_input_tokens_seen": 27995696,
      "step": 36705
    },
    {
      "epoch": 76.32016632016632,
      "grad_norm": 0.00014379557978827506,
      "learning_rate": 0.004982846879364116,
      "loss": 0.2516,
      "num_input_tokens_seen": 27999344,
      "step": 36710
    },
    {
      "epoch": 76.33056133056132,
      "grad_norm": 0.0003672557068057358,
      "learning_rate": 0.0049678016217400535,
      "loss": 0.2348,
      "num_input_tokens_seen": 28003056,
      "step": 36715
    },
    {
      "epoch": 76.34095634095634,
      "grad_norm": 0.00014944524446036667,
      "learning_rate": 0.004952778729903595,
      "loss": 0.27,
      "num_input_tokens_seen": 28006928,
      "step": 36720
    },
    {
      "epoch": 76.35135135135135,
      "grad_norm": 9.978645539376885e-05,
      "learning_rate": 0.004937778206171422,
      "loss": 0.2554,
      "num_input_tokens_seen": 28010768,
      "step": 36725
    },
    {
      "epoch": 76.36174636174636,
      "grad_norm": 0.0001249907654710114,
      "learning_rate": 0.004922800052856835,
      "loss": 0.2548,
      "num_input_tokens_seen": 28014576,
      "step": 36730
    },
    {
      "epoch": 76.37214137214137,
      "grad_norm": 0.0001946751435752958,
      "learning_rate": 0.004907844272269602,
      "loss": 0.2485,
      "num_input_tokens_seen": 28018448,
      "step": 36735
    },
    {
      "epoch": 76.38253638253639,
      "grad_norm": 0.0003651128790806979,
      "learning_rate": 0.004892910866716144,
      "loss": 0.2482,
      "num_input_tokens_seen": 28022192,
      "step": 36740
    },
    {
      "epoch": 76.39293139293139,
      "grad_norm": 0.0003809960908256471,
      "learning_rate": 0.004877999838499369,
      "loss": 0.2512,
      "num_input_tokens_seen": 28025968,
      "step": 36745
    },
    {
      "epoch": 76.4033264033264,
      "grad_norm": 0.0003875236143358052,
      "learning_rate": 0.0048631111899187065,
      "loss": 0.294,
      "num_input_tokens_seen": 28029808,
      "step": 36750
    },
    {
      "epoch": 76.41372141372142,
      "grad_norm": 0.0003525142674334347,
      "learning_rate": 0.0048482449232702335,
      "loss": 0.254,
      "num_input_tokens_seen": 28033488,
      "step": 36755
    },
    {
      "epoch": 76.42411642411642,
      "grad_norm": 0.0003585628292057663,
      "learning_rate": 0.004833401040846469,
      "loss": 0.2684,
      "num_input_tokens_seen": 28037264,
      "step": 36760
    },
    {
      "epoch": 76.43451143451144,
      "grad_norm": 0.00010031870624516159,
      "learning_rate": 0.004818579544936546,
      "loss": 0.284,
      "num_input_tokens_seen": 28041232,
      "step": 36765
    },
    {
      "epoch": 76.44490644490645,
      "grad_norm": 0.0005903986748307943,
      "learning_rate": 0.004803780437826121,
      "loss": 0.2611,
      "num_input_tokens_seen": 28045200,
      "step": 36770
    },
    {
      "epoch": 76.45530145530145,
      "grad_norm": 8.834667096380144e-05,
      "learning_rate": 0.004789003721797402,
      "loss": 0.2575,
      "num_input_tokens_seen": 28049232,
      "step": 36775
    },
    {
      "epoch": 76.46569646569647,
      "grad_norm": 0.00025207214639522135,
      "learning_rate": 0.004774249399129132,
      "loss": 0.26,
      "num_input_tokens_seen": 28053008,
      "step": 36780
    },
    {
      "epoch": 76.47609147609148,
      "grad_norm": 0.0005553350783884525,
      "learning_rate": 0.004759517472096642,
      "loss": 0.2383,
      "num_input_tokens_seen": 28056944,
      "step": 36785
    },
    {
      "epoch": 76.48648648648648,
      "grad_norm": 0.00025554600870236754,
      "learning_rate": 0.004744807942971746,
      "loss": 0.2795,
      "num_input_tokens_seen": 28060688,
      "step": 36790
    },
    {
      "epoch": 76.4968814968815,
      "grad_norm": 0.0004216076049488038,
      "learning_rate": 0.004730120814022881,
      "loss": 0.2541,
      "num_input_tokens_seen": 28064496,
      "step": 36795
    },
    {
      "epoch": 76.5072765072765,
      "grad_norm": 0.0004004836082458496,
      "learning_rate": 0.004715456087514935,
      "loss": 0.2488,
      "num_input_tokens_seen": 28068432,
      "step": 36800
    },
    {
      "epoch": 76.5072765072765,
      "eval_loss": 0.24843348562717438,
      "eval_runtime": 13.3776,
      "eval_samples_per_second": 63.987,
      "eval_steps_per_second": 15.997,
      "num_input_tokens_seen": 28068432,
      "step": 36800
    },
    {
      "epoch": 76.51767151767152,
      "grad_norm": 0.0002491026825737208,
      "learning_rate": 0.004700813765709432,
      "loss": 0.2713,
      "num_input_tokens_seen": 28072144,
      "step": 36805
    },
    {
      "epoch": 76.52806652806653,
      "grad_norm": 0.00027622279594652355,
      "learning_rate": 0.004686193850864401,
      "loss": 0.2463,
      "num_input_tokens_seen": 28075856,
      "step": 36810
    },
    {
      "epoch": 76.53846153846153,
      "grad_norm": 0.00043762658606283367,
      "learning_rate": 0.004671596345234385,
      "loss": 0.263,
      "num_input_tokens_seen": 28079600,
      "step": 36815
    },
    {
      "epoch": 76.54885654885655,
      "grad_norm": 0.00012006593169644475,
      "learning_rate": 0.00465702125107052,
      "loss": 0.2516,
      "num_input_tokens_seen": 28083472,
      "step": 36820
    },
    {
      "epoch": 76.55925155925156,
      "grad_norm": 0.00032866952824406326,
      "learning_rate": 0.004642468570620506,
      "loss": 0.262,
      "num_input_tokens_seen": 28087216,
      "step": 36825
    },
    {
      "epoch": 76.56964656964657,
      "grad_norm": 0.00034902579500339925,
      "learning_rate": 0.004627938306128482,
      "loss": 0.2713,
      "num_input_tokens_seen": 28090992,
      "step": 36830
    },
    {
      "epoch": 76.58004158004158,
      "grad_norm": 0.00010219719115411863,
      "learning_rate": 0.004613430459835255,
      "loss": 0.2569,
      "num_input_tokens_seen": 28094640,
      "step": 36835
    },
    {
      "epoch": 76.5904365904366,
      "grad_norm": 0.00034654364571906626,
      "learning_rate": 0.004598945033978085,
      "loss": 0.2772,
      "num_input_tokens_seen": 28098544,
      "step": 36840
    },
    {
      "epoch": 76.6008316008316,
      "grad_norm": 0.00037653459003195167,
      "learning_rate": 0.004584482030790804,
      "loss": 0.2788,
      "num_input_tokens_seen": 28102224,
      "step": 36845
    },
    {
      "epoch": 76.61122661122661,
      "grad_norm": 0.0005071199266240001,
      "learning_rate": 0.004570041452503826,
      "loss": 0.2719,
      "num_input_tokens_seen": 28105872,
      "step": 36850
    },
    {
      "epoch": 76.62162162162163,
      "grad_norm": 0.0001203383071697317,
      "learning_rate": 0.004555623301344003,
      "loss": 0.2493,
      "num_input_tokens_seen": 28109648,
      "step": 36855
    },
    {
      "epoch": 76.63201663201663,
      "grad_norm": 0.0002816346532199532,
      "learning_rate": 0.004541227579534857,
      "loss": 0.2732,
      "num_input_tokens_seen": 28113456,
      "step": 36860
    },
    {
      "epoch": 76.64241164241164,
      "grad_norm": 0.00053248624317348,
      "learning_rate": 0.004526854289296378,
      "loss": 0.2771,
      "num_input_tokens_seen": 28117008,
      "step": 36865
    },
    {
      "epoch": 76.65280665280665,
      "grad_norm": 0.0005299804033711553,
      "learning_rate": 0.004512503432845078,
      "loss": 0.2663,
      "num_input_tokens_seen": 28120816,
      "step": 36870
    },
    {
      "epoch": 76.66320166320166,
      "grad_norm": 4.515681939665228e-05,
      "learning_rate": 0.004498175012394068,
      "loss": 0.2712,
      "num_input_tokens_seen": 28124688,
      "step": 36875
    },
    {
      "epoch": 76.67359667359668,
      "grad_norm": 0.000797775574028492,
      "learning_rate": 0.004483869030152965,
      "loss": 0.2637,
      "num_input_tokens_seen": 28128528,
      "step": 36880
    },
    {
      "epoch": 76.68399168399168,
      "grad_norm": 0.0002325286332052201,
      "learning_rate": 0.004469585488327904,
      "loss": 0.2837,
      "num_input_tokens_seen": 28132240,
      "step": 36885
    },
    {
      "epoch": 76.6943866943867,
      "grad_norm": 0.0002579780702944845,
      "learning_rate": 0.0044553243891216395,
      "loss": 0.2657,
      "num_input_tokens_seen": 28136112,
      "step": 36890
    },
    {
      "epoch": 76.70478170478171,
      "grad_norm": 0.0003651200095191598,
      "learning_rate": 0.004441085734733363,
      "loss": 0.265,
      "num_input_tokens_seen": 28139824,
      "step": 36895
    },
    {
      "epoch": 76.71517671517671,
      "grad_norm": 0.00021340936655178666,
      "learning_rate": 0.004426869527358884,
      "loss": 0.2681,
      "num_input_tokens_seen": 28143600,
      "step": 36900
    },
    {
      "epoch": 76.72557172557173,
      "grad_norm": 0.0002148688945453614,
      "learning_rate": 0.0044126757691905156,
      "loss": 0.2748,
      "num_input_tokens_seen": 28147472,
      "step": 36905
    },
    {
      "epoch": 76.73596673596674,
      "grad_norm": 0.0008631636155769229,
      "learning_rate": 0.004398504462417107,
      "loss": 0.2535,
      "num_input_tokens_seen": 28151280,
      "step": 36910
    },
    {
      "epoch": 76.74636174636174,
      "grad_norm": 0.00011504167196108028,
      "learning_rate": 0.0043843556092240605,
      "loss": 0.2905,
      "num_input_tokens_seen": 28155280,
      "step": 36915
    },
    {
      "epoch": 76.75675675675676,
      "grad_norm": 0.0010818240698426962,
      "learning_rate": 0.004370229211793281,
      "loss": 0.2575,
      "num_input_tokens_seen": 28159088,
      "step": 36920
    },
    {
      "epoch": 76.76715176715177,
      "grad_norm": 0.0008599944994784892,
      "learning_rate": 0.0043561252723032405,
      "loss": 0.256,
      "num_input_tokens_seen": 28162736,
      "step": 36925
    },
    {
      "epoch": 76.77754677754677,
      "grad_norm": 0.0003508412337396294,
      "learning_rate": 0.004342043792929001,
      "loss": 0.2484,
      "num_input_tokens_seen": 28166640,
      "step": 36930
    },
    {
      "epoch": 76.78794178794179,
      "grad_norm": 0.0007711535436101258,
      "learning_rate": 0.004327984775842025,
      "loss": 0.2775,
      "num_input_tokens_seen": 28170576,
      "step": 36935
    },
    {
      "epoch": 76.7983367983368,
      "grad_norm": 0.00013414390559773892,
      "learning_rate": 0.004313948223210428,
      "loss": 0.2581,
      "num_input_tokens_seen": 28174544,
      "step": 36940
    },
    {
      "epoch": 76.8087318087318,
      "grad_norm": 0.00011797674233093858,
      "learning_rate": 0.004299934137198846,
      "loss": 0.2716,
      "num_input_tokens_seen": 28178384,
      "step": 36945
    },
    {
      "epoch": 76.81912681912682,
      "grad_norm": 0.00018428413022775203,
      "learning_rate": 0.004285942519968383,
      "loss": 0.2682,
      "num_input_tokens_seen": 28182192,
      "step": 36950
    },
    {
      "epoch": 76.82952182952182,
      "grad_norm": 0.0003074467822443694,
      "learning_rate": 0.004271973373676746,
      "loss": 0.2634,
      "num_input_tokens_seen": 28186096,
      "step": 36955
    },
    {
      "epoch": 76.83991683991684,
      "grad_norm": 0.00016882973432075232,
      "learning_rate": 0.004258026700478146,
      "loss": 0.2518,
      "num_input_tokens_seen": 28189904,
      "step": 36960
    },
    {
      "epoch": 76.85031185031185,
      "grad_norm": 0.00015716590860392898,
      "learning_rate": 0.004244102502523328,
      "loss": 0.274,
      "num_input_tokens_seen": 28193712,
      "step": 36965
    },
    {
      "epoch": 76.86070686070686,
      "grad_norm": 0.0004181308322586119,
      "learning_rate": 0.004230200781959592,
      "loss": 0.2696,
      "num_input_tokens_seen": 28197584,
      "step": 36970
    },
    {
      "epoch": 76.87110187110187,
      "grad_norm": 0.0003224349347874522,
      "learning_rate": 0.004216321540930756,
      "loss": 0.2684,
      "num_input_tokens_seen": 28201456,
      "step": 36975
    },
    {
      "epoch": 76.88149688149689,
      "grad_norm": 0.00027789699379354715,
      "learning_rate": 0.004202464781577175,
      "loss": 0.2581,
      "num_input_tokens_seen": 28205360,
      "step": 36980
    },
    {
      "epoch": 76.89189189189189,
      "grad_norm": 0.0003180943022016436,
      "learning_rate": 0.00418863050603574,
      "loss": 0.2567,
      "num_input_tokens_seen": 28209264,
      "step": 36985
    },
    {
      "epoch": 76.9022869022869,
      "grad_norm": 0.00042342691449448466,
      "learning_rate": 0.004174818716439843,
      "loss": 0.228,
      "num_input_tokens_seen": 28213072,
      "step": 36990
    },
    {
      "epoch": 76.91268191268192,
      "grad_norm": 0.00031388431671075523,
      "learning_rate": 0.004161029414919464,
      "loss": 0.2759,
      "num_input_tokens_seen": 28216912,
      "step": 36995
    },
    {
      "epoch": 76.92307692307692,
      "grad_norm": 0.00030096486443653703,
      "learning_rate": 0.004147262603601071,
      "loss": 0.2812,
      "num_input_tokens_seen": 28220720,
      "step": 37000
    },
    {
      "epoch": 76.92307692307692,
      "eval_loss": 0.2496126890182495,
      "eval_runtime": 13.3892,
      "eval_samples_per_second": 63.932,
      "eval_steps_per_second": 15.983,
      "num_input_tokens_seen": 28220720,
      "step": 37000
    },
    {
      "epoch": 76.93347193347194,
      "grad_norm": 0.00011489955795696005,
      "learning_rate": 0.004133518284607679,
      "loss": 0.2784,
      "num_input_tokens_seen": 28224624,
      "step": 37005
    },
    {
      "epoch": 76.94386694386695,
      "grad_norm": 0.0002891292970161885,
      "learning_rate": 0.004119796460058861,
      "loss": 0.2608,
      "num_input_tokens_seen": 28228496,
      "step": 37010
    },
    {
      "epoch": 76.95426195426195,
      "grad_norm": 0.0001657196698943153,
      "learning_rate": 0.00410609713207064,
      "loss": 0.2617,
      "num_input_tokens_seen": 28232336,
      "step": 37015
    },
    {
      "epoch": 76.96465696465697,
      "grad_norm": 0.0001507043489255011,
      "learning_rate": 0.004092420302755678,
      "loss": 0.2563,
      "num_input_tokens_seen": 28236112,
      "step": 37020
    },
    {
      "epoch": 76.97505197505197,
      "grad_norm": 0.00017236564599443227,
      "learning_rate": 0.004078765974223103,
      "loss": 0.2594,
      "num_input_tokens_seen": 28239856,
      "step": 37025
    },
    {
      "epoch": 76.98544698544698,
      "grad_norm": 0.0004071624716743827,
      "learning_rate": 0.004065134148578564,
      "loss": 0.2734,
      "num_input_tokens_seen": 28243600,
      "step": 37030
    },
    {
      "epoch": 76.995841995842,
      "grad_norm": 0.0004299864231143147,
      "learning_rate": 0.004051524827924279,
      "loss": 0.2817,
      "num_input_tokens_seen": 28247344,
      "step": 37035
    },
    {
      "epoch": 77.006237006237,
      "grad_norm": 0.003630937309935689,
      "learning_rate": 0.004037938014358955,
      "loss": 0.2382,
      "num_input_tokens_seen": 28251072,
      "step": 37040
    },
    {
      "epoch": 77.01663201663202,
      "grad_norm": 0.00022580896620638669,
      "learning_rate": 0.004024373709977863,
      "loss": 0.2633,
      "num_input_tokens_seen": 28255008,
      "step": 37045
    },
    {
      "epoch": 77.02702702702703,
      "grad_norm": 0.0004324744513723999,
      "learning_rate": 0.004010831916872814,
      "loss": 0.2635,
      "num_input_tokens_seen": 28258752,
      "step": 37050
    },
    {
      "epoch": 77.03742203742203,
      "grad_norm": 0.001715255668386817,
      "learning_rate": 0.003997312637132089,
      "loss": 0.2768,
      "num_input_tokens_seen": 28262656,
      "step": 37055
    },
    {
      "epoch": 77.04781704781705,
      "grad_norm": 0.0004185736470390111,
      "learning_rate": 0.003983815872840535,
      "loss": 0.2486,
      "num_input_tokens_seen": 28266688,
      "step": 37060
    },
    {
      "epoch": 77.05821205821206,
      "grad_norm": 0.00022444193018600345,
      "learning_rate": 0.003970341626079521,
      "loss": 0.2799,
      "num_input_tokens_seen": 28270592,
      "step": 37065
    },
    {
      "epoch": 77.06860706860707,
      "grad_norm": 0.00017643517639953643,
      "learning_rate": 0.003956889898926952,
      "loss": 0.2621,
      "num_input_tokens_seen": 28274336,
      "step": 37070
    },
    {
      "epoch": 77.07900207900208,
      "grad_norm": 0.00017763584037311375,
      "learning_rate": 0.0039434606934572675,
      "loss": 0.255,
      "num_input_tokens_seen": 28278432,
      "step": 37075
    },
    {
      "epoch": 77.0893970893971,
      "grad_norm": 0.0003319371317047626,
      "learning_rate": 0.003930054011741396,
      "loss": 0.2588,
      "num_input_tokens_seen": 28282240,
      "step": 37080
    },
    {
      "epoch": 77.0997920997921,
      "grad_norm": 0.0003477666468825191,
      "learning_rate": 0.0039166698558468155,
      "loss": 0.261,
      "num_input_tokens_seen": 28286080,
      "step": 37085
    },
    {
      "epoch": 77.11018711018711,
      "grad_norm": 0.00012088054063497111,
      "learning_rate": 0.0039033082278375594,
      "loss": 0.2749,
      "num_input_tokens_seen": 28289856,
      "step": 37090
    },
    {
      "epoch": 77.12058212058211,
      "grad_norm": 0.0003110474208369851,
      "learning_rate": 0.003889969129774112,
      "loss": 0.2634,
      "num_input_tokens_seen": 28293632,
      "step": 37095
    },
    {
      "epoch": 77.13097713097713,
      "grad_norm": 0.00029557262314483523,
      "learning_rate": 0.0038766525637135784,
      "loss": 0.2373,
      "num_input_tokens_seen": 28297408,
      "step": 37100
    },
    {
      "epoch": 77.14137214137214,
      "grad_norm": 0.00015276744670700282,
      "learning_rate": 0.0038633585317095318,
      "loss": 0.2638,
      "num_input_tokens_seen": 28301312,
      "step": 37105
    },
    {
      "epoch": 77.15176715176715,
      "grad_norm": 0.00018350358004681766,
      "learning_rate": 0.00385008703581205,
      "loss": 0.2686,
      "num_input_tokens_seen": 28305184,
      "step": 37110
    },
    {
      "epoch": 77.16216216216216,
      "grad_norm": 0.000589017232414335,
      "learning_rate": 0.0038368380780677944,
      "loss": 0.2537,
      "num_input_tokens_seen": 28309024,
      "step": 37115
    },
    {
      "epoch": 77.17255717255718,
      "grad_norm": 0.000341988867148757,
      "learning_rate": 0.003823611660519882,
      "loss": 0.2661,
      "num_input_tokens_seen": 28312896,
      "step": 37120
    },
    {
      "epoch": 77.18295218295218,
      "grad_norm": 0.0005191625095903873,
      "learning_rate": 0.0038104077852080475,
      "loss": 0.2756,
      "num_input_tokens_seen": 28316704,
      "step": 37125
    },
    {
      "epoch": 77.1933471933472,
      "grad_norm": 0.0005864623817615211,
      "learning_rate": 0.003797226454168462,
      "loss": 0.2705,
      "num_input_tokens_seen": 28320416,
      "step": 37130
    },
    {
      "epoch": 77.20374220374221,
      "grad_norm": 0.00032458262285217643,
      "learning_rate": 0.003784067669433849,
      "loss": 0.269,
      "num_input_tokens_seen": 28324288,
      "step": 37135
    },
    {
      "epoch": 77.21413721413721,
      "grad_norm": 0.0003057298017665744,
      "learning_rate": 0.0037709314330334528,
      "loss": 0.2724,
      "num_input_tokens_seen": 28327840,
      "step": 37140
    },
    {
      "epoch": 77.22453222453223,
      "grad_norm": 0.00019162641547154635,
      "learning_rate": 0.003757817746993086,
      "loss": 0.2661,
      "num_input_tokens_seen": 28331520,
      "step": 37145
    },
    {
      "epoch": 77.23492723492724,
      "grad_norm": 0.00037404592148959637,
      "learning_rate": 0.0037447266133349977,
      "loss": 0.2868,
      "num_input_tokens_seen": 28335392,
      "step": 37150
    },
    {
      "epoch": 77.24532224532224,
      "grad_norm": 0.0005784329841844738,
      "learning_rate": 0.003731658034078039,
      "loss": 0.2519,
      "num_input_tokens_seen": 28339264,
      "step": 37155
    },
    {
      "epoch": 77.25571725571726,
      "grad_norm": 0.000151947679114528,
      "learning_rate": 0.0037186120112375153,
      "loss": 0.2671,
      "num_input_tokens_seen": 28343136,
      "step": 37160
    },
    {
      "epoch": 77.26611226611226,
      "grad_norm": 5.967227480141446e-05,
      "learning_rate": 0.003705588546825317,
      "loss": 0.2763,
      "num_input_tokens_seen": 28346784,
      "step": 37165
    },
    {
      "epoch": 77.27650727650727,
      "grad_norm": 0.0003354424552526325,
      "learning_rate": 0.0036925876428498205,
      "loss": 0.2233,
      "num_input_tokens_seen": 28350592,
      "step": 37170
    },
    {
      "epoch": 77.28690228690229,
      "grad_norm": 0.00024831629707477987,
      "learning_rate": 0.0036796093013159057,
      "loss": 0.2538,
      "num_input_tokens_seen": 28354272,
      "step": 37175
    },
    {
      "epoch": 77.29729729729729,
      "grad_norm": 0.00013682476128451526,
      "learning_rate": 0.0036666535242250217,
      "loss": 0.2762,
      "num_input_tokens_seen": 28358080,
      "step": 37180
    },
    {
      "epoch": 77.3076923076923,
      "grad_norm": 6.173827569000423e-05,
      "learning_rate": 0.003653720313575104,
      "loss": 0.2519,
      "num_input_tokens_seen": 28362112,
      "step": 37185
    },
    {
      "epoch": 77.31808731808732,
      "grad_norm": 0.00044085155241191387,
      "learning_rate": 0.003640809671360623,
      "loss": 0.2484,
      "num_input_tokens_seen": 28365984,
      "step": 37190
    },
    {
      "epoch": 77.32848232848232,
      "grad_norm": 0.000660480058286339,
      "learning_rate": 0.003627921599572553,
      "loss": 0.2821,
      "num_input_tokens_seen": 28369888,
      "step": 37195
    },
    {
      "epoch": 77.33887733887734,
      "grad_norm": 0.00039032360655255616,
      "learning_rate": 0.003615056100198405,
      "loss": 0.276,
      "num_input_tokens_seen": 28373600,
      "step": 37200
    },
    {
      "epoch": 77.33887733887734,
      "eval_loss": 0.2491244226694107,
      "eval_runtime": 13.3625,
      "eval_samples_per_second": 64.06,
      "eval_steps_per_second": 16.015,
      "num_input_tokens_seen": 28373600,
      "step": 37200
    },
    {
      "epoch": 77.34927234927235,
      "grad_norm": 0.00012770255852956325,
      "learning_rate": 0.003602213175222174,
      "loss": 0.2446,
      "num_input_tokens_seen": 28377312,
      "step": 37205
    },
    {
      "epoch": 77.35966735966736,
      "grad_norm": 0.00030073378002271056,
      "learning_rate": 0.0035893928266244432,
      "loss": 0.2857,
      "num_input_tokens_seen": 28381152,
      "step": 37210
    },
    {
      "epoch": 77.37006237006237,
      "grad_norm": 0.0002108486951328814,
      "learning_rate": 0.003576595056382248,
      "loss": 0.2736,
      "num_input_tokens_seen": 28384896,
      "step": 37215
    },
    {
      "epoch": 77.38045738045739,
      "grad_norm": 4.768271537614055e-05,
      "learning_rate": 0.0035638198664691423,
      "loss": 0.2516,
      "num_input_tokens_seen": 28388544,
      "step": 37220
    },
    {
      "epoch": 77.39085239085239,
      "grad_norm": 0.00032628123881295323,
      "learning_rate": 0.003551067258855267,
      "loss": 0.222,
      "num_input_tokens_seen": 28392352,
      "step": 37225
    },
    {
      "epoch": 77.4012474012474,
      "grad_norm": 0.000332905154209584,
      "learning_rate": 0.0035383372355071996,
      "loss": 0.272,
      "num_input_tokens_seen": 28396288,
      "step": 37230
    },
    {
      "epoch": 77.41164241164242,
      "grad_norm": 0.002037876518443227,
      "learning_rate": 0.0035256297983881023,
      "loss": 0.2695,
      "num_input_tokens_seen": 28400000,
      "step": 37235
    },
    {
      "epoch": 77.42203742203742,
      "grad_norm": 0.00029906872077845037,
      "learning_rate": 0.0035129449494575747,
      "loss": 0.2663,
      "num_input_tokens_seen": 28403872,
      "step": 37240
    },
    {
      "epoch": 77.43243243243244,
      "grad_norm": 0.0013210372999310493,
      "learning_rate": 0.0035002826906718187,
      "loss": 0.2884,
      "num_input_tokens_seen": 28407648,
      "step": 37245
    },
    {
      "epoch": 77.44282744282744,
      "grad_norm": 0.000139789714012295,
      "learning_rate": 0.003487643023983522,
      "loss": 0.2665,
      "num_input_tokens_seen": 28411456,
      "step": 37250
    },
    {
      "epoch": 77.45322245322245,
      "grad_norm": 0.00016201545076910406,
      "learning_rate": 0.003475025951341842,
      "loss": 0.2514,
      "num_input_tokens_seen": 28415328,
      "step": 37255
    },
    {
      "epoch": 77.46361746361747,
      "grad_norm": 0.0002804503310471773,
      "learning_rate": 0.00346243147469249,
      "loss": 0.2702,
      "num_input_tokens_seen": 28419072,
      "step": 37260
    },
    {
      "epoch": 77.47401247401247,
      "grad_norm": 0.0002909105096478015,
      "learning_rate": 0.0034498595959777446,
      "loss": 0.2557,
      "num_input_tokens_seen": 28422912,
      "step": 37265
    },
    {
      "epoch": 77.48440748440748,
      "grad_norm": 0.000797420390881598,
      "learning_rate": 0.003437310317136305,
      "loss": 0.2684,
      "num_input_tokens_seen": 28426880,
      "step": 37270
    },
    {
      "epoch": 77.4948024948025,
      "grad_norm": 0.0002620832237880677,
      "learning_rate": 0.0034247836401034236,
      "loss": 0.2551,
      "num_input_tokens_seen": 28430816,
      "step": 37275
    },
    {
      "epoch": 77.5051975051975,
      "grad_norm": 0.0005849107983522117,
      "learning_rate": 0.003412279566810905,
      "loss": 0.2596,
      "num_input_tokens_seen": 28434688,
      "step": 37280
    },
    {
      "epoch": 77.51559251559252,
      "grad_norm": 0.0003156873572152108,
      "learning_rate": 0.00339979809918699,
      "loss": 0.2698,
      "num_input_tokens_seen": 28438688,
      "step": 37285
    },
    {
      "epoch": 77.52598752598753,
      "grad_norm": 0.00011453476327005774,
      "learning_rate": 0.0033873392391565228,
      "loss": 0.2679,
      "num_input_tokens_seen": 28442272,
      "step": 37290
    },
    {
      "epoch": 77.53638253638253,
      "grad_norm": 0.0007239239057525992,
      "learning_rate": 0.003374902988640782,
      "loss": 0.2658,
      "num_input_tokens_seen": 28446144,
      "step": 37295
    },
    {
      "epoch": 77.54677754677755,
      "grad_norm": 0.00017345913511235267,
      "learning_rate": 0.0033624893495576014,
      "loss": 0.2696,
      "num_input_tokens_seen": 28450048,
      "step": 37300
    },
    {
      "epoch": 77.55717255717256,
      "grad_norm": 0.0003316587535664439,
      "learning_rate": 0.0033500983238213323,
      "loss": 0.2803,
      "num_input_tokens_seen": 28453824,
      "step": 37305
    },
    {
      "epoch": 77.56756756756756,
      "grad_norm": 0.0008936517406255007,
      "learning_rate": 0.0033377299133428126,
      "loss": 0.2469,
      "num_input_tokens_seen": 28457536,
      "step": 37310
    },
    {
      "epoch": 77.57796257796258,
      "grad_norm": 0.0002861677494365722,
      "learning_rate": 0.003325384120029434,
      "loss": 0.2398,
      "num_input_tokens_seen": 28461408,
      "step": 37315
    },
    {
      "epoch": 77.58835758835758,
      "grad_norm": 0.0007366678328253329,
      "learning_rate": 0.0033130609457850233,
      "loss": 0.2603,
      "num_input_tokens_seen": 28465152,
      "step": 37320
    },
    {
      "epoch": 77.5987525987526,
      "grad_norm": 8.969330519903451e-05,
      "learning_rate": 0.0033007603925100104,
      "loss": 0.2632,
      "num_input_tokens_seen": 28469024,
      "step": 37325
    },
    {
      "epoch": 77.60914760914761,
      "grad_norm": 0.00018747115973383188,
      "learning_rate": 0.003288482462101294,
      "loss": 0.2662,
      "num_input_tokens_seen": 28472704,
      "step": 37330
    },
    {
      "epoch": 77.61954261954261,
      "grad_norm": 0.0002932991774287075,
      "learning_rate": 0.0032762271564522605,
      "loss": 0.2481,
      "num_input_tokens_seen": 28476576,
      "step": 37335
    },
    {
      "epoch": 77.62993762993763,
      "grad_norm": 0.00024183867208193988,
      "learning_rate": 0.003263994477452864,
      "loss": 0.2502,
      "num_input_tokens_seen": 28480352,
      "step": 37340
    },
    {
      "epoch": 77.64033264033264,
      "grad_norm": 0.00019381643505766988,
      "learning_rate": 0.0032517844269895125,
      "loss": 0.2863,
      "num_input_tokens_seen": 28484064,
      "step": 37345
    },
    {
      "epoch": 77.65072765072765,
      "grad_norm": 0.000633417977951467,
      "learning_rate": 0.0032395970069451496,
      "loss": 0.2502,
      "num_input_tokens_seen": 28487840,
      "step": 37350
    },
    {
      "epoch": 77.66112266112266,
      "grad_norm": 0.00039819898665882647,
      "learning_rate": 0.0032274322191992388,
      "loss": 0.2377,
      "num_input_tokens_seen": 28491648,
      "step": 37355
    },
    {
      "epoch": 77.67151767151768,
      "grad_norm": 0.0002766170946415514,
      "learning_rate": 0.0032152900656277294,
      "loss": 0.2651,
      "num_input_tokens_seen": 28495488,
      "step": 37360
    },
    {
      "epoch": 77.68191268191268,
      "grad_norm": 7.905794336693361e-05,
      "learning_rate": 0.0032031705481030902,
      "loss": 0.2577,
      "num_input_tokens_seen": 28499264,
      "step": 37365
    },
    {
      "epoch": 77.6923076923077,
      "grad_norm": 0.0003408543416298926,
      "learning_rate": 0.0031910736684943428,
      "loss": 0.259,
      "num_input_tokens_seen": 28503136,
      "step": 37370
    },
    {
      "epoch": 77.70270270270271,
      "grad_norm": 0.00024153085541911423,
      "learning_rate": 0.0031789994286669453,
      "loss": 0.283,
      "num_input_tokens_seen": 28506912,
      "step": 37375
    },
    {
      "epoch": 77.71309771309771,
      "grad_norm": 0.00040425642509944737,
      "learning_rate": 0.003166947830482908,
      "loss": 0.2536,
      "num_input_tokens_seen": 28510880,
      "step": 37380
    },
    {
      "epoch": 77.72349272349273,
      "grad_norm": 0.0010615846840664744,
      "learning_rate": 0.003154918875800727,
      "loss": 0.2614,
      "num_input_tokens_seen": 28514816,
      "step": 37385
    },
    {
      "epoch": 77.73388773388774,
      "grad_norm": 0.00023688064538873732,
      "learning_rate": 0.00314291256647542,
      "loss": 0.2499,
      "num_input_tokens_seen": 28518656,
      "step": 37390
    },
    {
      "epoch": 77.74428274428274,
      "grad_norm": 0.00032043945975601673,
      "learning_rate": 0.0031309289043585375,
      "loss": 0.2769,
      "num_input_tokens_seen": 28522528,
      "step": 37395
    },
    {
      "epoch": 77.75467775467776,
      "grad_norm": 0.0004774409462697804,
      "learning_rate": 0.003118967891298069,
      "loss": 0.2792,
      "num_input_tokens_seen": 28526304,
      "step": 37400
    },
    {
      "epoch": 77.75467775467776,
      "eval_loss": 0.2487938404083252,
      "eval_runtime": 13.3666,
      "eval_samples_per_second": 64.04,
      "eval_steps_per_second": 16.01,
      "num_input_tokens_seen": 28526304,
      "step": 37400
    },
    {
      "epoch": 77.76507276507276,
      "grad_norm": 0.00025252869818359613,
      "learning_rate": 0.003107029529138572,
      "loss": 0.2505,
      "num_input_tokens_seen": 28530080,
      "step": 37405
    },
    {
      "epoch": 77.77546777546777,
      "grad_norm": 0.00040116385207511485,
      "learning_rate": 0.0030951138197211235,
      "loss": 0.2462,
      "num_input_tokens_seen": 28533920,
      "step": 37410
    },
    {
      "epoch": 77.78586278586279,
      "grad_norm": 0.0005070054321549833,
      "learning_rate": 0.0030832207648832377,
      "loss": 0.2601,
      "num_input_tokens_seen": 28537760,
      "step": 37415
    },
    {
      "epoch": 77.79625779625779,
      "grad_norm": 0.0002695280418265611,
      "learning_rate": 0.0030713503664589635,
      "loss": 0.2623,
      "num_input_tokens_seen": 28541600,
      "step": 37420
    },
    {
      "epoch": 77.8066528066528,
      "grad_norm": 0.0003062169998884201,
      "learning_rate": 0.0030595026262788872,
      "loss": 0.2683,
      "num_input_tokens_seen": 28545344,
      "step": 37425
    },
    {
      "epoch": 77.81704781704782,
      "grad_norm": 0.00029004839598201215,
      "learning_rate": 0.00304767754617008,
      "loss": 0.2588,
      "num_input_tokens_seen": 28549088,
      "step": 37430
    },
    {
      "epoch": 77.82744282744282,
      "grad_norm": 0.000274400896159932,
      "learning_rate": 0.003035875127956117,
      "loss": 0.2611,
      "num_input_tokens_seen": 28552832,
      "step": 37435
    },
    {
      "epoch": 77.83783783783784,
      "grad_norm": 0.00010281155118718743,
      "learning_rate": 0.0030240953734570752,
      "loss": 0.2521,
      "num_input_tokens_seen": 28556608,
      "step": 37440
    },
    {
      "epoch": 77.84823284823285,
      "grad_norm": 9.546616638544947e-05,
      "learning_rate": 0.003012338284489535,
      "loss": 0.2651,
      "num_input_tokens_seen": 28560512,
      "step": 37445
    },
    {
      "epoch": 77.85862785862786,
      "grad_norm": 0.00038319017039611936,
      "learning_rate": 0.0030006038628665964,
      "loss": 0.2653,
      "num_input_tokens_seen": 28564288,
      "step": 37450
    },
    {
      "epoch": 77.86902286902287,
      "grad_norm": 8.163908933056518e-05,
      "learning_rate": 0.002988892110397845,
      "loss": 0.2765,
      "num_input_tokens_seen": 28568032,
      "step": 37455
    },
    {
      "epoch": 77.87941787941789,
      "grad_norm": 0.00032411047141067684,
      "learning_rate": 0.0029772030288894025,
      "loss": 0.2961,
      "num_input_tokens_seen": 28571872,
      "step": 37460
    },
    {
      "epoch": 77.88981288981289,
      "grad_norm": 0.0011404320830479264,
      "learning_rate": 0.0029655366201438438,
      "loss": 0.2441,
      "num_input_tokens_seen": 28575584,
      "step": 37465
    },
    {
      "epoch": 77.9002079002079,
      "grad_norm": 0.00013742709415964782,
      "learning_rate": 0.0029538928859602965,
      "loss": 0.2816,
      "num_input_tokens_seen": 28579296,
      "step": 37470
    },
    {
      "epoch": 77.9106029106029,
      "grad_norm": 0.0002540506247896701,
      "learning_rate": 0.002942271828134374,
      "loss": 0.2601,
      "num_input_tokens_seen": 28583072,
      "step": 37475
    },
    {
      "epoch": 77.92099792099792,
      "grad_norm": 0.00022144940157886595,
      "learning_rate": 0.00293067344845816,
      "loss": 0.2602,
      "num_input_tokens_seen": 28586784,
      "step": 37480
    },
    {
      "epoch": 77.93139293139293,
      "grad_norm": 0.0003978123713750392,
      "learning_rate": 0.0029190977487202896,
      "loss": 0.273,
      "num_input_tokens_seen": 28590592,
      "step": 37485
    },
    {
      "epoch": 77.94178794178794,
      "grad_norm": 0.00014536465459968895,
      "learning_rate": 0.0029075447307058853,
      "loss": 0.2623,
      "num_input_tokens_seen": 28594464,
      "step": 37490
    },
    {
      "epoch": 77.95218295218295,
      "grad_norm": 0.0004084158572368324,
      "learning_rate": 0.0028960143961965722,
      "loss": 0.2508,
      "num_input_tokens_seen": 28598272,
      "step": 37495
    },
    {
      "epoch": 77.96257796257797,
      "grad_norm": 0.0003114208811894059,
      "learning_rate": 0.002884506746970461,
      "loss": 0.2349,
      "num_input_tokens_seen": 28602048,
      "step": 37500
    },
    {
      "epoch": 77.97297297297297,
      "grad_norm": 0.0004258941044099629,
      "learning_rate": 0.0028730217848021654,
      "loss": 0.2858,
      "num_input_tokens_seen": 28605824,
      "step": 37505
    },
    {
      "epoch": 77.98336798336798,
      "grad_norm": 0.0004563272523228079,
      "learning_rate": 0.0028615595114628188,
      "loss": 0.2592,
      "num_input_tokens_seen": 28609664,
      "step": 37510
    },
    {
      "epoch": 77.993762993763,
      "grad_norm": 0.00025951562565751374,
      "learning_rate": 0.002850119928720074,
      "loss": 0.2627,
      "num_input_tokens_seen": 28613472,
      "step": 37515
    },
    {
      "epoch": 78.004158004158,
      "grad_norm": 0.00015031058865133673,
      "learning_rate": 0.0028387030383380195,
      "loss": 0.2443,
      "num_input_tokens_seen": 28617232,
      "step": 37520
    },
    {
      "epoch": 78.01455301455302,
      "grad_norm": 4.4965101551497355e-05,
      "learning_rate": 0.0028273088420772974,
      "loss": 0.2599,
      "num_input_tokens_seen": 28621104,
      "step": 37525
    },
    {
      "epoch": 78.02494802494803,
      "grad_norm": 0.00021094904514029622,
      "learning_rate": 0.002815937341695068,
      "loss": 0.3042,
      "num_input_tokens_seen": 28625008,
      "step": 37530
    },
    {
      "epoch": 78.03534303534303,
      "grad_norm": 0.00020221843442413956,
      "learning_rate": 0.0028045885389448963,
      "loss": 0.2747,
      "num_input_tokens_seen": 28628848,
      "step": 37535
    },
    {
      "epoch": 78.04573804573805,
      "grad_norm": 0.0005335089517757297,
      "learning_rate": 0.002793262435576965,
      "loss": 0.2728,
      "num_input_tokens_seen": 28632688,
      "step": 37540
    },
    {
      "epoch": 78.05613305613305,
      "grad_norm": 0.000485223630676046,
      "learning_rate": 0.0027819590333378772,
      "loss": 0.2665,
      "num_input_tokens_seen": 28636624,
      "step": 37545
    },
    {
      "epoch": 78.06652806652806,
      "grad_norm": 0.0006221400108188391,
      "learning_rate": 0.002770678333970755,
      "loss": 0.2671,
      "num_input_tokens_seen": 28640368,
      "step": 37550
    },
    {
      "epoch": 78.07692307692308,
      "grad_norm": 0.00023803344811312854,
      "learning_rate": 0.0027594203392152573,
      "loss": 0.2642,
      "num_input_tokens_seen": 28644304,
      "step": 37555
    },
    {
      "epoch": 78.08731808731808,
      "grad_norm": 0.00012186486856080592,
      "learning_rate": 0.002748185050807478,
      "loss": 0.2778,
      "num_input_tokens_seen": 28648016,
      "step": 37560
    },
    {
      "epoch": 78.0977130977131,
      "grad_norm": 0.0003063289914280176,
      "learning_rate": 0.002736972470480031,
      "loss": 0.2571,
      "num_input_tokens_seen": 28651824,
      "step": 37565
    },
    {
      "epoch": 78.10810810810811,
      "grad_norm": 0.0003650210564956069,
      "learning_rate": 0.002725782599962068,
      "loss": 0.2571,
      "num_input_tokens_seen": 28655696,
      "step": 37570
    },
    {
      "epoch": 78.11850311850311,
      "grad_norm": 0.0003527977969497442,
      "learning_rate": 0.0027146154409791734,
      "loss": 0.2614,
      "num_input_tokens_seen": 28659504,
      "step": 37575
    },
    {
      "epoch": 78.12889812889813,
      "grad_norm": 0.0002823907125275582,
      "learning_rate": 0.002703470995253504,
      "loss": 0.2585,
      "num_input_tokens_seen": 28663440,
      "step": 37580
    },
    {
      "epoch": 78.13929313929314,
      "grad_norm": 0.0007945362594909966,
      "learning_rate": 0.0026923492645036184,
      "loss": 0.2714,
      "num_input_tokens_seen": 28667216,
      "step": 37585
    },
    {
      "epoch": 78.14968814968815,
      "grad_norm": 0.0001828245585784316,
      "learning_rate": 0.0026812502504446776,
      "loss": 0.2472,
      "num_input_tokens_seen": 28670928,
      "step": 37590
    },
    {
      "epoch": 78.16008316008316,
      "grad_norm": 0.00011388651910237968,
      "learning_rate": 0.0026701739547882798,
      "loss": 0.2744,
      "num_input_tokens_seen": 28674736,
      "step": 37595
    },
    {
      "epoch": 78.17047817047818,
      "grad_norm": 0.00013194308849051595,
      "learning_rate": 0.0026591203792425077,
      "loss": 0.2478,
      "num_input_tokens_seen": 28678672,
      "step": 37600
    },
    {
      "epoch": 78.17047817047818,
      "eval_loss": 0.2498016059398651,
      "eval_runtime": 13.3771,
      "eval_samples_per_second": 63.99,
      "eval_steps_per_second": 15.997,
      "num_input_tokens_seen": 28678672,
      "step": 37600
    },
    {
      "epoch": 78.18087318087318,
      "grad_norm": 0.0006715295021422207,
      "learning_rate": 0.0026480895255119818,
      "loss": 0.2496,
      "num_input_tokens_seen": 28682512,
      "step": 37605
    },
    {
      "epoch": 78.1912681912682,
      "grad_norm": 0.000341462146025151,
      "learning_rate": 0.002637081395297791,
      "loss": 0.259,
      "num_input_tokens_seen": 28686352,
      "step": 37610
    },
    {
      "epoch": 78.20166320166321,
      "grad_norm": 0.00013333007518667728,
      "learning_rate": 0.0026260959902975113,
      "loss": 0.2537,
      "num_input_tokens_seen": 28690288,
      "step": 37615
    },
    {
      "epoch": 78.21205821205821,
      "grad_norm": 5.83543733227998e-05,
      "learning_rate": 0.00261513331220527,
      "loss": 0.281,
      "num_input_tokens_seen": 28694096,
      "step": 37620
    },
    {
      "epoch": 78.22245322245323,
      "grad_norm": 0.0006748784799128771,
      "learning_rate": 0.0026041933627116154,
      "loss": 0.2653,
      "num_input_tokens_seen": 28697808,
      "step": 37625
    },
    {
      "epoch": 78.23284823284823,
      "grad_norm": 0.0004393504641484469,
      "learning_rate": 0.0025932761435036476,
      "loss": 0.2495,
      "num_input_tokens_seen": 28701552,
      "step": 37630
    },
    {
      "epoch": 78.24324324324324,
      "grad_norm": 0.0006128849927335978,
      "learning_rate": 0.002582381656264904,
      "loss": 0.278,
      "num_input_tokens_seen": 28705296,
      "step": 37635
    },
    {
      "epoch": 78.25363825363826,
      "grad_norm": 0.00019382509344723076,
      "learning_rate": 0.0025715099026754895,
      "loss": 0.2672,
      "num_input_tokens_seen": 28709040,
      "step": 37640
    },
    {
      "epoch": 78.26403326403326,
      "grad_norm": 0.0005110765341669321,
      "learning_rate": 0.002560660884411947,
      "loss": 0.2689,
      "num_input_tokens_seen": 28712848,
      "step": 37645
    },
    {
      "epoch": 78.27442827442827,
      "grad_norm": 0.00012160575715824962,
      "learning_rate": 0.0025498346031473385,
      "loss": 0.2483,
      "num_input_tokens_seen": 28716720,
      "step": 37650
    },
    {
      "epoch": 78.28482328482329,
      "grad_norm": 6.866976764285937e-05,
      "learning_rate": 0.0025390310605511945,
      "loss": 0.2578,
      "num_input_tokens_seen": 28720592,
      "step": 37655
    },
    {
      "epoch": 78.29521829521829,
      "grad_norm": 0.0005914228968322277,
      "learning_rate": 0.0025282502582895995,
      "loss": 0.246,
      "num_input_tokens_seen": 28724336,
      "step": 37660
    },
    {
      "epoch": 78.3056133056133,
      "grad_norm": 0.0006792324129492044,
      "learning_rate": 0.002517492198025023,
      "loss": 0.2412,
      "num_input_tokens_seen": 28728208,
      "step": 37665
    },
    {
      "epoch": 78.31600831600832,
      "grad_norm": 0.0005745728267356753,
      "learning_rate": 0.0025067568814165554,
      "loss": 0.2939,
      "num_input_tokens_seen": 28731952,
      "step": 37670
    },
    {
      "epoch": 78.32640332640332,
      "grad_norm": 0.00017038217629306018,
      "learning_rate": 0.0024960443101196884,
      "loss": 0.2717,
      "num_input_tokens_seen": 28735792,
      "step": 37675
    },
    {
      "epoch": 78.33679833679834,
      "grad_norm": 0.0009870089124888182,
      "learning_rate": 0.002485354485786434,
      "loss": 0.2672,
      "num_input_tokens_seen": 28739824,
      "step": 37680
    },
    {
      "epoch": 78.34719334719335,
      "grad_norm": 6.939924787729979e-05,
      "learning_rate": 0.002474687410065307,
      "loss": 0.2605,
      "num_input_tokens_seen": 28743728,
      "step": 37685
    },
    {
      "epoch": 78.35758835758836,
      "grad_norm": 0.0002048071037279442,
      "learning_rate": 0.002464043084601308,
      "loss": 0.2775,
      "num_input_tokens_seen": 28747568,
      "step": 37690
    },
    {
      "epoch": 78.36798336798337,
      "grad_norm": 0.00020292414410505444,
      "learning_rate": 0.0024534215110358915,
      "loss": 0.2634,
      "num_input_tokens_seen": 28751408,
      "step": 37695
    },
    {
      "epoch": 78.37837837837837,
      "grad_norm": 0.0010969911236315966,
      "learning_rate": 0.002442822691007096,
      "loss": 0.2733,
      "num_input_tokens_seen": 28755088,
      "step": 37700
    },
    {
      "epoch": 78.38877338877339,
      "grad_norm": 0.0005535022937692702,
      "learning_rate": 0.002432246626149348,
      "loss": 0.2584,
      "num_input_tokens_seen": 28758896,
      "step": 37705
    },
    {
      "epoch": 78.3991683991684,
      "grad_norm": 0.00024046578619163483,
      "learning_rate": 0.002421693318093626,
      "loss": 0.2622,
      "num_input_tokens_seen": 28762736,
      "step": 37710
    },
    {
      "epoch": 78.4095634095634,
      "grad_norm": 0.00024482040316797793,
      "learning_rate": 0.0024111627684673784,
      "loss": 0.2783,
      "num_input_tokens_seen": 28766640,
      "step": 37715
    },
    {
      "epoch": 78.41995841995842,
      "grad_norm": 0.00029359760810621083,
      "learning_rate": 0.0024006549788945395,
      "loss": 0.2616,
      "num_input_tokens_seen": 28770480,
      "step": 37720
    },
    {
      "epoch": 78.43035343035343,
      "grad_norm": 0.0002024842397077009,
      "learning_rate": 0.0023901699509955463,
      "loss": 0.2518,
      "num_input_tokens_seen": 28774256,
      "step": 37725
    },
    {
      "epoch": 78.44074844074844,
      "grad_norm": 0.00023607561888638884,
      "learning_rate": 0.0023797076863873554,
      "loss": 0.2566,
      "num_input_tokens_seen": 28778096,
      "step": 37730
    },
    {
      "epoch": 78.45114345114345,
      "grad_norm": 0.0002189069928135723,
      "learning_rate": 0.0023692681866833262,
      "loss": 0.2687,
      "num_input_tokens_seen": 28781872,
      "step": 37735
    },
    {
      "epoch": 78.46153846153847,
      "grad_norm": 0.00016448796668555588,
      "learning_rate": 0.0023588514534934046,
      "loss": 0.2666,
      "num_input_tokens_seen": 28785648,
      "step": 37740
    },
    {
      "epoch": 78.47193347193347,
      "grad_norm": 0.00045395278721116483,
      "learning_rate": 0.002348457488423955,
      "loss": 0.2634,
      "num_input_tokens_seen": 28789520,
      "step": 37745
    },
    {
      "epoch": 78.48232848232848,
      "grad_norm": 0.00012836877431254834,
      "learning_rate": 0.0023380862930778624,
      "loss": 0.2821,
      "num_input_tokens_seen": 28793328,
      "step": 37750
    },
    {
      "epoch": 78.4927234927235,
      "grad_norm": 0.0006201587384566665,
      "learning_rate": 0.0023277378690545135,
      "loss": 0.2515,
      "num_input_tokens_seen": 28797168,
      "step": 37755
    },
    {
      "epoch": 78.5031185031185,
      "grad_norm": 0.00013783652684651315,
      "learning_rate": 0.0023174122179497325,
      "loss": 0.2508,
      "num_input_tokens_seen": 28800944,
      "step": 37760
    },
    {
      "epoch": 78.51351351351352,
      "grad_norm": 0.0002832898171618581,
      "learning_rate": 0.0023071093413558784,
      "loss": 0.256,
      "num_input_tokens_seen": 28804688,
      "step": 37765
    },
    {
      "epoch": 78.52390852390852,
      "grad_norm": 7.736971747362986e-05,
      "learning_rate": 0.002296829240861814,
      "loss": 0.2843,
      "num_input_tokens_seen": 28808528,
      "step": 37770
    },
    {
      "epoch": 78.53430353430353,
      "grad_norm": 0.00013587898865807801,
      "learning_rate": 0.002286571918052821,
      "loss": 0.2707,
      "num_input_tokens_seen": 28812528,
      "step": 37775
    },
    {
      "epoch": 78.54469854469855,
      "grad_norm": 0.00033153416006825864,
      "learning_rate": 0.0022763373745107174,
      "loss": 0.2677,
      "num_input_tokens_seen": 28816304,
      "step": 37780
    },
    {
      "epoch": 78.55509355509355,
      "grad_norm": 0.0015719928778707981,
      "learning_rate": 0.0022661256118138074,
      "loss": 0.2429,
      "num_input_tokens_seen": 28820336,
      "step": 37785
    },
    {
      "epoch": 78.56548856548856,
      "grad_norm": 0.00014601957809645683,
      "learning_rate": 0.0022559366315368645,
      "loss": 0.2792,
      "num_input_tokens_seen": 28824048,
      "step": 37790
    },
    {
      "epoch": 78.57588357588358,
      "grad_norm": 6.617201870540157e-05,
      "learning_rate": 0.002245770435251182,
      "loss": 0.257,
      "num_input_tokens_seen": 28827792,
      "step": 37795
    },
    {
      "epoch": 78.58627858627858,
      "grad_norm": 8.801106014288962e-05,
      "learning_rate": 0.002235627024524456,
      "loss": 0.2691,
      "num_input_tokens_seen": 28831632,
      "step": 37800
    },
    {
      "epoch": 78.58627858627858,
      "eval_loss": 0.2498367875814438,
      "eval_runtime": 13.3721,
      "eval_samples_per_second": 64.014,
      "eval_steps_per_second": 16.003,
      "num_input_tokens_seen": 28831632,
      "step": 37800
    },
    {
      "epoch": 78.5966735966736,
      "grad_norm": 0.00012745764979626983,
      "learning_rate": 0.0022255064009209847,
      "loss": 0.2632,
      "num_input_tokens_seen": 28835376,
      "step": 37805
    },
    {
      "epoch": 78.60706860706861,
      "grad_norm": 0.00037790200440213084,
      "learning_rate": 0.0022154085660014864,
      "loss": 0.2546,
      "num_input_tokens_seen": 28839088,
      "step": 37810
    },
    {
      "epoch": 78.61746361746361,
      "grad_norm": 0.00021950708469375968,
      "learning_rate": 0.0022053335213231494,
      "loss": 0.25,
      "num_input_tokens_seen": 28843024,
      "step": 37815
    },
    {
      "epoch": 78.62785862785863,
      "grad_norm": 0.00014048628509044647,
      "learning_rate": 0.002195281268439697,
      "loss": 0.2439,
      "num_input_tokens_seen": 28846640,
      "step": 37820
    },
    {
      "epoch": 78.63825363825364,
      "grad_norm": 0.00031565650715492666,
      "learning_rate": 0.002185251808901306,
      "loss": 0.2377,
      "num_input_tokens_seen": 28850448,
      "step": 37825
    },
    {
      "epoch": 78.64864864864865,
      "grad_norm": 0.00028254787321202457,
      "learning_rate": 0.0021752451442546227,
      "loss": 0.2537,
      "num_input_tokens_seen": 28854256,
      "step": 37830
    },
    {
      "epoch": 78.65904365904366,
      "grad_norm": 0.0005398524226620793,
      "learning_rate": 0.0021652612760428456,
      "loss": 0.2391,
      "num_input_tokens_seen": 28858032,
      "step": 37835
    },
    {
      "epoch": 78.66943866943868,
      "grad_norm": 0.0002082987775793299,
      "learning_rate": 0.0021553002058055603,
      "loss": 0.2702,
      "num_input_tokens_seen": 28861776,
      "step": 37840
    },
    {
      "epoch": 78.67983367983368,
      "grad_norm": 0.0008171290392056108,
      "learning_rate": 0.0021453619350789376,
      "loss": 0.2725,
      "num_input_tokens_seen": 28865520,
      "step": 37845
    },
    {
      "epoch": 78.6902286902287,
      "grad_norm": 0.00023345788940787315,
      "learning_rate": 0.0021354464653955516,
      "loss": 0.2352,
      "num_input_tokens_seen": 28869264,
      "step": 37850
    },
    {
      "epoch": 78.7006237006237,
      "grad_norm": 0.0004140743112657219,
      "learning_rate": 0.002125553798284513,
      "loss": 0.2612,
      "num_input_tokens_seen": 28873040,
      "step": 37855
    },
    {
      "epoch": 78.71101871101871,
      "grad_norm": 0.00019932870054617524,
      "learning_rate": 0.002115683935271384,
      "loss": 0.2661,
      "num_input_tokens_seen": 28876752,
      "step": 37860
    },
    {
      "epoch": 78.72141372141373,
      "grad_norm": 0.0005794940516352654,
      "learning_rate": 0.0021058368778782144,
      "loss": 0.2517,
      "num_input_tokens_seen": 28880400,
      "step": 37865
    },
    {
      "epoch": 78.73180873180873,
      "grad_norm": 0.00010781781747937202,
      "learning_rate": 0.002096012627623539,
      "loss": 0.2622,
      "num_input_tokens_seen": 28884176,
      "step": 37870
    },
    {
      "epoch": 78.74220374220374,
      "grad_norm": 0.00013670553744304925,
      "learning_rate": 0.00208621118602243,
      "loss": 0.2588,
      "num_input_tokens_seen": 28887888,
      "step": 37875
    },
    {
      "epoch": 78.75259875259876,
      "grad_norm": 0.00015113857807591558,
      "learning_rate": 0.002076432554586327,
      "loss": 0.2581,
      "num_input_tokens_seen": 28891696,
      "step": 37880
    },
    {
      "epoch": 78.76299376299376,
      "grad_norm": 0.00019641057588160038,
      "learning_rate": 0.002066676734823258,
      "loss": 0.268,
      "num_input_tokens_seen": 28895408,
      "step": 37885
    },
    {
      "epoch": 78.77338877338877,
      "grad_norm": 0.00032262125751003623,
      "learning_rate": 0.0020569437282376866,
      "loss": 0.286,
      "num_input_tokens_seen": 28899056,
      "step": 37890
    },
    {
      "epoch": 78.78378378378379,
      "grad_norm": 0.00017668158398009837,
      "learning_rate": 0.002047233536330545,
      "loss": 0.2706,
      "num_input_tokens_seen": 28902704,
      "step": 37895
    },
    {
      "epoch": 78.79417879417879,
      "grad_norm": 0.00016258470714092255,
      "learning_rate": 0.0020375461605993015,
      "loss": 0.2573,
      "num_input_tokens_seen": 28906352,
      "step": 37900
    },
    {
      "epoch": 78.8045738045738,
      "grad_norm": 0.00019959069322794676,
      "learning_rate": 0.002027881602537845,
      "loss": 0.2839,
      "num_input_tokens_seen": 28910160,
      "step": 37905
    },
    {
      "epoch": 78.81496881496882,
      "grad_norm": 0.0014921323163434863,
      "learning_rate": 0.002018239863636567,
      "loss": 0.2781,
      "num_input_tokens_seen": 28914064,
      "step": 37910
    },
    {
      "epoch": 78.82536382536382,
      "grad_norm": 0.0007923303637653589,
      "learning_rate": 0.002008620945382378,
      "loss": 0.264,
      "num_input_tokens_seen": 28917904,
      "step": 37915
    },
    {
      "epoch": 78.83575883575884,
      "grad_norm": 0.002255060477182269,
      "learning_rate": 0.001999024849258607,
      "loss": 0.2432,
      "num_input_tokens_seen": 28921744,
      "step": 37920
    },
    {
      "epoch": 78.84615384615384,
      "grad_norm": 0.0009984582429751754,
      "learning_rate": 0.001989451576745105,
      "loss": 0.2721,
      "num_input_tokens_seen": 28925616,
      "step": 37925
    },
    {
      "epoch": 78.85654885654886,
      "grad_norm": 0.0002737363975029439,
      "learning_rate": 0.00197990112931819,
      "loss": 0.2308,
      "num_input_tokens_seen": 28929488,
      "step": 37930
    },
    {
      "epoch": 78.86694386694387,
      "grad_norm": 0.00023218551359605044,
      "learning_rate": 0.0019703735084506345,
      "loss": 0.2473,
      "num_input_tokens_seen": 28933296,
      "step": 37935
    },
    {
      "epoch": 78.87733887733887,
      "grad_norm": 0.0006973809795454144,
      "learning_rate": 0.001960868715611763,
      "loss": 0.24,
      "num_input_tokens_seen": 28937072,
      "step": 37940
    },
    {
      "epoch": 78.88773388773389,
      "grad_norm": 0.00022784974134992808,
      "learning_rate": 0.0019513867522673034,
      "loss": 0.2559,
      "num_input_tokens_seen": 28940944,
      "step": 37945
    },
    {
      "epoch": 78.8981288981289,
      "grad_norm": 0.0004798508598469198,
      "learning_rate": 0.001941927619879502,
      "loss": 0.2755,
      "num_input_tokens_seen": 28944720,
      "step": 37950
    },
    {
      "epoch": 78.9085239085239,
      "grad_norm": 0.0010758964344859123,
      "learning_rate": 0.0019324913199070758,
      "loss": 0.2845,
      "num_input_tokens_seen": 28948816,
      "step": 37955
    },
    {
      "epoch": 78.91891891891892,
      "grad_norm": 0.00016881241754163057,
      "learning_rate": 0.0019230778538052106,
      "loss": 0.2549,
      "num_input_tokens_seen": 28952848,
      "step": 37960
    },
    {
      "epoch": 78.92931392931393,
      "grad_norm": 0.00033863424323499203,
      "learning_rate": 0.0019136872230255952,
      "loss": 0.2767,
      "num_input_tokens_seen": 28956688,
      "step": 37965
    },
    {
      "epoch": 78.93970893970894,
      "grad_norm": 0.0002336965117137879,
      "learning_rate": 0.0019043194290164045,
      "loss": 0.265,
      "num_input_tokens_seen": 28960400,
      "step": 37970
    },
    {
      "epoch": 78.95010395010395,
      "grad_norm": 0.00016260083066299558,
      "learning_rate": 0.0018949744732222162,
      "loss": 0.2651,
      "num_input_tokens_seen": 28964208,
      "step": 37975
    },
    {
      "epoch": 78.96049896049897,
      "grad_norm": 0.00023800510098226368,
      "learning_rate": 0.0018856523570841776,
      "loss": 0.2815,
      "num_input_tokens_seen": 28968176,
      "step": 37980
    },
    {
      "epoch": 78.97089397089397,
      "grad_norm": 0.000232863865676336,
      "learning_rate": 0.0018763530820398555,
      "loss": 0.2432,
      "num_input_tokens_seen": 28972016,
      "step": 37985
    },
    {
      "epoch": 78.98128898128898,
      "grad_norm": 0.0002703905920498073,
      "learning_rate": 0.0018670766495233525,
      "loss": 0.2359,
      "num_input_tokens_seen": 28975824,
      "step": 37990
    },
    {
      "epoch": 78.99168399168398,
      "grad_norm": 0.0004151446628384292,
      "learning_rate": 0.001857823060965158,
      "loss": 0.2633,
      "num_input_tokens_seen": 28979568,
      "step": 37995
    },
    {
      "epoch": 79.002079002079,
      "grad_norm": 0.00032281759195029736,
      "learning_rate": 0.0018485923177923467,
      "loss": 0.2608,
      "num_input_tokens_seen": 28983144,
      "step": 38000
    },
    {
      "epoch": 79.002079002079,
      "eval_loss": 0.2485750913619995,
      "eval_runtime": 13.3689,
      "eval_samples_per_second": 64.029,
      "eval_steps_per_second": 16.007,
      "num_input_tokens_seen": 28983144,
      "step": 38000
    },
    {
      "epoch": 79.01247401247402,
      "grad_norm": 0.00017966370796784759,
      "learning_rate": 0.001839384421428364,
      "loss": 0.262,
      "num_input_tokens_seen": 28986888,
      "step": 38005
    },
    {
      "epoch": 79.02286902286902,
      "grad_norm": 0.00016717262042220682,
      "learning_rate": 0.0018301993732932065,
      "loss": 0.2477,
      "num_input_tokens_seen": 28990696,
      "step": 38010
    },
    {
      "epoch": 79.03326403326403,
      "grad_norm": 0.0004760668089147657,
      "learning_rate": 0.0018210371748033248,
      "loss": 0.2709,
      "num_input_tokens_seen": 28994504,
      "step": 38015
    },
    {
      "epoch": 79.04365904365905,
      "grad_norm": 0.00012353732017800212,
      "learning_rate": 0.0018118978273716556,
      "loss": 0.2517,
      "num_input_tokens_seen": 28998184,
      "step": 38020
    },
    {
      "epoch": 79.05405405405405,
      "grad_norm": 0.0002884168934542686,
      "learning_rate": 0.001802781332407588,
      "loss": 0.2772,
      "num_input_tokens_seen": 29001960,
      "step": 38025
    },
    {
      "epoch": 79.06444906444906,
      "grad_norm": 0.00020809365378227085,
      "learning_rate": 0.0017936876913169806,
      "loss": 0.2683,
      "num_input_tokens_seen": 29005736,
      "step": 38030
    },
    {
      "epoch": 79.07484407484408,
      "grad_norm": 0.0002984378661494702,
      "learning_rate": 0.0017846169055022287,
      "loss": 0.2628,
      "num_input_tokens_seen": 29009576,
      "step": 38035
    },
    {
      "epoch": 79.08523908523908,
      "grad_norm": 0.0005485824658535421,
      "learning_rate": 0.0017755689763621295,
      "loss": 0.2568,
      "num_input_tokens_seen": 29013192,
      "step": 38040
    },
    {
      "epoch": 79.0956340956341,
      "grad_norm": 0.000435478868894279,
      "learning_rate": 0.0017665439052920173,
      "loss": 0.2855,
      "num_input_tokens_seen": 29016872,
      "step": 38045
    },
    {
      "epoch": 79.10602910602911,
      "grad_norm": 0.0005841461825184524,
      "learning_rate": 0.0017575416936836286,
      "loss": 0.2606,
      "num_input_tokens_seen": 29020744,
      "step": 38050
    },
    {
      "epoch": 79.11642411642411,
      "grad_norm": 0.00043549275142140687,
      "learning_rate": 0.0017485623429252528,
      "loss": 0.2534,
      "num_input_tokens_seen": 29024520,
      "step": 38055
    },
    {
      "epoch": 79.12681912681913,
      "grad_norm": 0.0008446744759567082,
      "learning_rate": 0.0017396058544016156,
      "loss": 0.2825,
      "num_input_tokens_seen": 29028456,
      "step": 38060
    },
    {
      "epoch": 79.13721413721414,
      "grad_norm": 0.00025966682005673647,
      "learning_rate": 0.0017306722294938958,
      "loss": 0.2656,
      "num_input_tokens_seen": 29032424,
      "step": 38065
    },
    {
      "epoch": 79.14760914760915,
      "grad_norm": 0.00022661415277980268,
      "learning_rate": 0.0017217614695798078,
      "loss": 0.2442,
      "num_input_tokens_seen": 29036136,
      "step": 38070
    },
    {
      "epoch": 79.15800415800416,
      "grad_norm": 0.00042740910430438817,
      "learning_rate": 0.001712873576033469,
      "loss": 0.2537,
      "num_input_tokens_seen": 29040136,
      "step": 38075
    },
    {
      "epoch": 79.16839916839916,
      "grad_norm": 0.0006336811347864568,
      "learning_rate": 0.0017040085502255163,
      "loss": 0.2533,
      "num_input_tokens_seen": 29043976,
      "step": 38080
    },
    {
      "epoch": 79.17879417879418,
      "grad_norm": 0.00026188933406956494,
      "learning_rate": 0.0016951663935230565,
      "loss": 0.261,
      "num_input_tokens_seen": 29047848,
      "step": 38085
    },
    {
      "epoch": 79.1891891891892,
      "grad_norm": 0.00043104070937260985,
      "learning_rate": 0.0016863471072896485,
      "loss": 0.2492,
      "num_input_tokens_seen": 29051656,
      "step": 38090
    },
    {
      "epoch": 79.1995841995842,
      "grad_norm": 0.0008819380309432745,
      "learning_rate": 0.0016775506928853377,
      "loss": 0.2843,
      "num_input_tokens_seen": 29055560,
      "step": 38095
    },
    {
      "epoch": 79.20997920997921,
      "grad_norm": 0.00024019851116463542,
      "learning_rate": 0.001668777151666656,
      "loss": 0.2419,
      "num_input_tokens_seen": 29059176,
      "step": 38100
    },
    {
      "epoch": 79.22037422037423,
      "grad_norm": 0.0003162742650602013,
      "learning_rate": 0.0016600264849865709,
      "loss": 0.2706,
      "num_input_tokens_seen": 29063048,
      "step": 38105
    },
    {
      "epoch": 79.23076923076923,
      "grad_norm": 0.0008727354579605162,
      "learning_rate": 0.0016512986941945695,
      "loss": 0.2565,
      "num_input_tokens_seen": 29066984,
      "step": 38110
    },
    {
      "epoch": 79.24116424116424,
      "grad_norm": 0.00014136887330096215,
      "learning_rate": 0.0016425937806365753,
      "loss": 0.2664,
      "num_input_tokens_seen": 29070792,
      "step": 38115
    },
    {
      "epoch": 79.25155925155926,
      "grad_norm": 0.00011961320706177503,
      "learning_rate": 0.0016339117456549979,
      "loss": 0.2415,
      "num_input_tokens_seen": 29074376,
      "step": 38120
    },
    {
      "epoch": 79.26195426195426,
      "grad_norm": 0.0005950881168246269,
      "learning_rate": 0.0016252525905886995,
      "loss": 0.2604,
      "num_input_tokens_seen": 29078184,
      "step": 38125
    },
    {
      "epoch": 79.27234927234927,
      "grad_norm": 0.00018996621656697243,
      "learning_rate": 0.0016166163167730617,
      "loss": 0.274,
      "num_input_tokens_seen": 29082184,
      "step": 38130
    },
    {
      "epoch": 79.28274428274429,
      "grad_norm": 0.0004959648940712214,
      "learning_rate": 0.0016080029255398864,
      "loss": 0.2775,
      "num_input_tokens_seen": 29086024,
      "step": 38135
    },
    {
      "epoch": 79.29313929313929,
      "grad_norm": 7.827136141713709e-05,
      "learning_rate": 0.0015994124182174606,
      "loss": 0.2498,
      "num_input_tokens_seen": 29089928,
      "step": 38140
    },
    {
      "epoch": 79.3035343035343,
      "grad_norm": 0.0004906129324808717,
      "learning_rate": 0.001590844796130575,
      "loss": 0.2543,
      "num_input_tokens_seen": 29093800,
      "step": 38145
    },
    {
      "epoch": 79.31392931392931,
      "grad_norm": 0.00034243756090290844,
      "learning_rate": 0.001582300060600439,
      "loss": 0.2735,
      "num_input_tokens_seen": 29097704,
      "step": 38150
    },
    {
      "epoch": 79.32432432432432,
      "grad_norm": 0.0002170712104998529,
      "learning_rate": 0.0015737782129447652,
      "loss": 0.2558,
      "num_input_tokens_seen": 29101672,
      "step": 38155
    },
    {
      "epoch": 79.33471933471934,
      "grad_norm": 0.00041325410711579025,
      "learning_rate": 0.0015652792544777361,
      "loss": 0.268,
      "num_input_tokens_seen": 29105480,
      "step": 38160
    },
    {
      "epoch": 79.34511434511434,
      "grad_norm": 0.00011577887198654935,
      "learning_rate": 0.0015568031865099863,
      "loss": 0.2682,
      "num_input_tokens_seen": 29109352,
      "step": 38165
    },
    {
      "epoch": 79.35550935550935,
      "grad_norm": 0.0023331954143941402,
      "learning_rate": 0.0015483500103486369,
      "loss": 0.2708,
      "num_input_tokens_seen": 29113256,
      "step": 38170
    },
    {
      "epoch": 79.36590436590437,
      "grad_norm": 0.00028183453832753,
      "learning_rate": 0.0015399197272972787,
      "loss": 0.2552,
      "num_input_tokens_seen": 29116968,
      "step": 38175
    },
    {
      "epoch": 79.37629937629937,
      "grad_norm": 0.00011987083416897804,
      "learning_rate": 0.0015315123386559714,
      "loss": 0.2516,
      "num_input_tokens_seen": 29120776,
      "step": 38180
    },
    {
      "epoch": 79.38669438669439,
      "grad_norm": 0.00038736755959689617,
      "learning_rate": 0.0015231278457212283,
      "loss": 0.2677,
      "num_input_tokens_seen": 29124616,
      "step": 38185
    },
    {
      "epoch": 79.3970893970894,
      "grad_norm": 0.00048630518722347915,
      "learning_rate": 0.001514766249786048,
      "loss": 0.2726,
      "num_input_tokens_seen": 29128360,
      "step": 38190
    },
    {
      "epoch": 79.4074844074844,
      "grad_norm": 0.00014829615247435868,
      "learning_rate": 0.0015064275521398994,
      "loss": 0.2455,
      "num_input_tokens_seen": 29132232,
      "step": 38195
    },
    {
      "epoch": 79.41787941787942,
      "grad_norm": 8.91961099114269e-05,
      "learning_rate": 0.0014981117540686872,
      "loss": 0.2627,
      "num_input_tokens_seen": 29136008,
      "step": 38200
    },
    {
      "epoch": 79.41787941787942,
      "eval_loss": 0.24872873723506927,
      "eval_runtime": 13.3604,
      "eval_samples_per_second": 64.07,
      "eval_steps_per_second": 16.017,
      "num_input_tokens_seen": 29136008,
      "step": 38200
    },
    {
      "epoch": 79.42827442827443,
      "grad_norm": 0.00026118342066183686,
      "learning_rate": 0.0014898188568548687,
      "loss": 0.2796,
      "num_input_tokens_seen": 29139912,
      "step": 38205
    },
    {
      "epoch": 79.43866943866944,
      "grad_norm": 0.0008638104773126543,
      "learning_rate": 0.0014815488617772542,
      "loss": 0.2728,
      "num_input_tokens_seen": 29143848,
      "step": 38210
    },
    {
      "epoch": 79.44906444906445,
      "grad_norm": 0.00042454988579265773,
      "learning_rate": 0.0014733017701112072,
      "loss": 0.2578,
      "num_input_tokens_seen": 29147656,
      "step": 38215
    },
    {
      "epoch": 79.45945945945945,
      "grad_norm": 0.00028132752049714327,
      "learning_rate": 0.0014650775831285435,
      "loss": 0.2601,
      "num_input_tokens_seen": 29151176,
      "step": 38220
    },
    {
      "epoch": 79.46985446985447,
      "grad_norm": 0.0004521938681136817,
      "learning_rate": 0.001456876302097515,
      "loss": 0.2728,
      "num_input_tokens_seen": 29154856,
      "step": 38225
    },
    {
      "epoch": 79.48024948024948,
      "grad_norm": 0.0001556698844069615,
      "learning_rate": 0.0014486979282828604,
      "loss": 0.2164,
      "num_input_tokens_seen": 29158696,
      "step": 38230
    },
    {
      "epoch": 79.49064449064448,
      "grad_norm": 9.598196629667655e-05,
      "learning_rate": 0.001440542462945804,
      "loss": 0.245,
      "num_input_tokens_seen": 29162568,
      "step": 38235
    },
    {
      "epoch": 79.5010395010395,
      "grad_norm": 0.0004440545162651688,
      "learning_rate": 0.0014324099073440232,
      "loss": 0.2664,
      "num_input_tokens_seen": 29166376,
      "step": 38240
    },
    {
      "epoch": 79.51143451143452,
      "grad_norm": 0.00024112591927405447,
      "learning_rate": 0.0014243002627316482,
      "loss": 0.2563,
      "num_input_tokens_seen": 29170184,
      "step": 38245
    },
    {
      "epoch": 79.52182952182952,
      "grad_norm": 0.0003036188136320561,
      "learning_rate": 0.0014162135303592781,
      "loss": 0.272,
      "num_input_tokens_seen": 29173928,
      "step": 38250
    },
    {
      "epoch": 79.53222453222453,
      "grad_norm": 6.247423152672127e-05,
      "learning_rate": 0.001408149711474016,
      "loss": 0.2677,
      "num_input_tokens_seen": 29177832,
      "step": 38255
    },
    {
      "epoch": 79.54261954261955,
      "grad_norm": 0.0001993149344343692,
      "learning_rate": 0.0014001088073193834,
      "loss": 0.2667,
      "num_input_tokens_seen": 29181608,
      "step": 38260
    },
    {
      "epoch": 79.55301455301455,
      "grad_norm": 0.0005638791481032968,
      "learning_rate": 0.0013920908191354052,
      "loss": 0.2849,
      "num_input_tokens_seen": 29185384,
      "step": 38265
    },
    {
      "epoch": 79.56340956340956,
      "grad_norm": 0.00023856728512328118,
      "learning_rate": 0.001384095748158526,
      "loss": 0.2511,
      "num_input_tokens_seen": 29189160,
      "step": 38270
    },
    {
      "epoch": 79.57380457380458,
      "grad_norm": 0.0006954625132493675,
      "learning_rate": 0.0013761235956217255,
      "loss": 0.2514,
      "num_input_tokens_seen": 29193000,
      "step": 38275
    },
    {
      "epoch": 79.58419958419958,
      "grad_norm": 0.0007552440511062741,
      "learning_rate": 0.0013681743627543873,
      "loss": 0.2747,
      "num_input_tokens_seen": 29196616,
      "step": 38280
    },
    {
      "epoch": 79.5945945945946,
      "grad_norm": 0.0001778706646291539,
      "learning_rate": 0.001360248050782381,
      "loss": 0.2456,
      "num_input_tokens_seen": 29200360,
      "step": 38285
    },
    {
      "epoch": 79.60498960498961,
      "grad_norm": 0.0003644452372100204,
      "learning_rate": 0.001352344660928062,
      "loss": 0.251,
      "num_input_tokens_seen": 29204328,
      "step": 38290
    },
    {
      "epoch": 79.61538461538461,
      "grad_norm": 0.0005265697836875916,
      "learning_rate": 0.0013444641944102052,
      "loss": 0.268,
      "num_input_tokens_seen": 29208136,
      "step": 38295
    },
    {
      "epoch": 79.62577962577963,
      "grad_norm": 0.00045020924881100655,
      "learning_rate": 0.0013366066524441056,
      "loss": 0.257,
      "num_input_tokens_seen": 29212008,
      "step": 38300
    },
    {
      "epoch": 79.63617463617463,
      "grad_norm": 0.00022348729544319212,
      "learning_rate": 0.0013287720362414768,
      "loss": 0.2642,
      "num_input_tokens_seen": 29215880,
      "step": 38305
    },
    {
      "epoch": 79.64656964656965,
      "grad_norm": 0.0001538680080557242,
      "learning_rate": 0.0013209603470105025,
      "loss": 0.2634,
      "num_input_tokens_seen": 29219880,
      "step": 38310
    },
    {
      "epoch": 79.65696465696466,
      "grad_norm": 8.355820318683982e-05,
      "learning_rate": 0.0013131715859558857,
      "loss": 0.2495,
      "num_input_tokens_seen": 29223752,
      "step": 38315
    },
    {
      "epoch": 79.66735966735966,
      "grad_norm": 0.0005104420124553144,
      "learning_rate": 0.001305405754278699,
      "loss": 0.2582,
      "num_input_tokens_seen": 29227592,
      "step": 38320
    },
    {
      "epoch": 79.67775467775468,
      "grad_norm": 0.005619477946311235,
      "learning_rate": 0.0012976628531765843,
      "loss": 0.2784,
      "num_input_tokens_seen": 29231336,
      "step": 38325
    },
    {
      "epoch": 79.6881496881497,
      "grad_norm": 0.0004730597138404846,
      "learning_rate": 0.0012899428838435533,
      "loss": 0.2602,
      "num_input_tokens_seen": 29234984,
      "step": 38330
    },
    {
      "epoch": 79.6985446985447,
      "grad_norm": 0.00043982380884699523,
      "learning_rate": 0.001282245847470137,
      "loss": 0.2362,
      "num_input_tokens_seen": 29238792,
      "step": 38335
    },
    {
      "epoch": 79.70893970893971,
      "grad_norm": 0.0006354756187647581,
      "learning_rate": 0.001274571745243319,
      "loss": 0.2552,
      "num_input_tokens_seen": 29242536,
      "step": 38340
    },
    {
      "epoch": 79.71933471933473,
      "grad_norm": 0.00026771367993205786,
      "learning_rate": 0.0012669205783465364,
      "loss": 0.2711,
      "num_input_tokens_seen": 29246312,
      "step": 38345
    },
    {
      "epoch": 79.72972972972973,
      "grad_norm": 0.0007430287660099566,
      "learning_rate": 0.001259292347959695,
      "loss": 0.295,
      "num_input_tokens_seen": 29250024,
      "step": 38350
    },
    {
      "epoch": 79.74012474012474,
      "grad_norm": 0.00045486559974960983,
      "learning_rate": 0.0012516870552591707,
      "loss": 0.2725,
      "num_input_tokens_seen": 29253832,
      "step": 38355
    },
    {
      "epoch": 79.75051975051976,
      "grad_norm": 0.00026794298901222646,
      "learning_rate": 0.001244104701417792,
      "loss": 0.2644,
      "num_input_tokens_seen": 29257640,
      "step": 38360
    },
    {
      "epoch": 79.76091476091476,
      "grad_norm": 0.00027484007296152413,
      "learning_rate": 0.0012365452876048565,
      "loss": 0.2727,
      "num_input_tokens_seen": 29261448,
      "step": 38365
    },
    {
      "epoch": 79.77130977130977,
      "grad_norm": 0.0002597956918179989,
      "learning_rate": 0.001229008814986099,
      "loss": 0.2586,
      "num_input_tokens_seen": 29265192,
      "step": 38370
    },
    {
      "epoch": 79.78170478170478,
      "grad_norm": 0.00032989526516757905,
      "learning_rate": 0.0012214952847237725,
      "loss": 0.2875,
      "num_input_tokens_seen": 29269032,
      "step": 38375
    },
    {
      "epoch": 79.79209979209979,
      "grad_norm": 0.0001911343861138448,
      "learning_rate": 0.0012140046979765339,
      "loss": 0.2701,
      "num_input_tokens_seen": 29272936,
      "step": 38380
    },
    {
      "epoch": 79.8024948024948,
      "grad_norm": 0.0003251050948165357,
      "learning_rate": 0.0012065370558995258,
      "loss": 0.2687,
      "num_input_tokens_seen": 29276680,
      "step": 38385
    },
    {
      "epoch": 79.81288981288981,
      "grad_norm": 0.0002685713116079569,
      "learning_rate": 0.0011990923596443602,
      "loss": 0.2483,
      "num_input_tokens_seen": 29280392,
      "step": 38390
    },
    {
      "epoch": 79.82328482328482,
      "grad_norm": 0.001317249028943479,
      "learning_rate": 0.001191670610359119,
      "loss": 0.262,
      "num_input_tokens_seen": 29284200,
      "step": 38395
    },
    {
      "epoch": 79.83367983367984,
      "grad_norm": 9.249548020306975e-05,
      "learning_rate": 0.0011842718091882865,
      "loss": 0.244,
      "num_input_tokens_seen": 29288104,
      "step": 38400
    },
    {
      "epoch": 79.83367983367984,
      "eval_loss": 0.24904431402683258,
      "eval_runtime": 13.3762,
      "eval_samples_per_second": 63.994,
      "eval_steps_per_second": 15.999,
      "num_input_tokens_seen": 29288104,
      "step": 38400
    },
    {
      "epoch": 79.84407484407484,
      "grad_norm": 0.00012733276525978,
      "learning_rate": 0.0011768959572729,
      "loss": 0.2582,
      "num_input_tokens_seen": 29291816,
      "step": 38405
    },
    {
      "epoch": 79.85446985446985,
      "grad_norm": 0.00010336189006920904,
      "learning_rate": 0.001169543055750366,
      "loss": 0.2555,
      "num_input_tokens_seen": 29295752,
      "step": 38410
    },
    {
      "epoch": 79.86486486486487,
      "grad_norm": 0.00013424715143628418,
      "learning_rate": 0.0011622131057546115,
      "loss": 0.2844,
      "num_input_tokens_seen": 29299528,
      "step": 38415
    },
    {
      "epoch": 79.87525987525987,
      "grad_norm": 0.0008734024595469236,
      "learning_rate": 0.0011549061084160316,
      "loss": 0.2686,
      "num_input_tokens_seen": 29303336,
      "step": 38420
    },
    {
      "epoch": 79.88565488565489,
      "grad_norm": 0.00042071370990015566,
      "learning_rate": 0.0011476220648614088,
      "loss": 0.2687,
      "num_input_tokens_seen": 29307048,
      "step": 38425
    },
    {
      "epoch": 79.8960498960499,
      "grad_norm": 0.00016543905076105148,
      "learning_rate": 0.0011403609762140777,
      "loss": 0.2599,
      "num_input_tokens_seen": 29310984,
      "step": 38430
    },
    {
      "epoch": 79.9064449064449,
      "grad_norm": 0.0020599847193807364,
      "learning_rate": 0.0011331228435937756,
      "loss": 0.2573,
      "num_input_tokens_seen": 29314664,
      "step": 38435
    },
    {
      "epoch": 79.91683991683992,
      "grad_norm": 0.00016036223678383976,
      "learning_rate": 0.0011259076681166935,
      "loss": 0.2626,
      "num_input_tokens_seen": 29318408,
      "step": 38440
    },
    {
      "epoch": 79.92723492723492,
      "grad_norm": 0.0003660772927105427,
      "learning_rate": 0.0011187154508955244,
      "loss": 0.2559,
      "num_input_tokens_seen": 29322280,
      "step": 38445
    },
    {
      "epoch": 79.93762993762994,
      "grad_norm": 0.000240333829424344,
      "learning_rate": 0.001111546193039381,
      "loss": 0.2542,
      "num_input_tokens_seen": 29326056,
      "step": 38450
    },
    {
      "epoch": 79.94802494802495,
      "grad_norm": 0.0001794466224964708,
      "learning_rate": 0.0011043998956538792,
      "loss": 0.2601,
      "num_input_tokens_seen": 29329800,
      "step": 38455
    },
    {
      "epoch": 79.95841995841995,
      "grad_norm": 0.00012207345571368933,
      "learning_rate": 0.0010972765598410538,
      "loss": 0.2451,
      "num_input_tokens_seen": 29333544,
      "step": 38460
    },
    {
      "epoch": 79.96881496881497,
      "grad_norm": 0.000610252667684108,
      "learning_rate": 0.0010901761866993931,
      "loss": 0.265,
      "num_input_tokens_seen": 29337384,
      "step": 38465
    },
    {
      "epoch": 79.97920997920998,
      "grad_norm": 0.0010604034177958965,
      "learning_rate": 0.0010830987773238876,
      "loss": 0.2758,
      "num_input_tokens_seen": 29341320,
      "step": 38470
    },
    {
      "epoch": 79.98960498960498,
      "grad_norm": 0.00017277008737437427,
      "learning_rate": 0.0010760443328059644,
      "loss": 0.2675,
      "num_input_tokens_seen": 29345160,
      "step": 38475
    },
    {
      "epoch": 80.0,
      "grad_norm": 0.0006391305360011756,
      "learning_rate": 0.001069012854233503,
      "loss": 0.2667,
      "num_input_tokens_seen": 29349080,
      "step": 38480
    },
    {
      "epoch": 80.01039501039502,
      "grad_norm": 0.0009808116592466831,
      "learning_rate": 0.0010620043426908365,
      "loss": 0.2641,
      "num_input_tokens_seen": 29352760,
      "step": 38485
    },
    {
      "epoch": 80.02079002079002,
      "grad_norm": 0.00039683966315351427,
      "learning_rate": 0.0010550187992587833,
      "loss": 0.2348,
      "num_input_tokens_seen": 29356696,
      "step": 38490
    },
    {
      "epoch": 80.03118503118503,
      "grad_norm": 0.000833352969493717,
      "learning_rate": 0.0010480562250145653,
      "loss": 0.2523,
      "num_input_tokens_seen": 29360504,
      "step": 38495
    },
    {
      "epoch": 80.04158004158005,
      "grad_norm": 0.00013748662604484707,
      "learning_rate": 0.0010411166210319567,
      "loss": 0.2602,
      "num_input_tokens_seen": 29364184,
      "step": 38500
    },
    {
      "epoch": 80.05197505197505,
      "grad_norm": 0.0008084644796326756,
      "learning_rate": 0.0010341999883810848,
      "loss": 0.2497,
      "num_input_tokens_seen": 29367992,
      "step": 38505
    },
    {
      "epoch": 80.06237006237006,
      "grad_norm": 0.0005259827012196183,
      "learning_rate": 0.0010273063281285965,
      "loss": 0.2807,
      "num_input_tokens_seen": 29371768,
      "step": 38510
    },
    {
      "epoch": 80.07276507276508,
      "grad_norm": 0.000703663972672075,
      "learning_rate": 0.0010204356413375747,
      "loss": 0.2881,
      "num_input_tokens_seen": 29375512,
      "step": 38515
    },
    {
      "epoch": 80.08316008316008,
      "grad_norm": 0.00019635447824839503,
      "learning_rate": 0.001013587929067572,
      "loss": 0.2596,
      "num_input_tokens_seen": 29379416,
      "step": 38520
    },
    {
      "epoch": 80.0935550935551,
      "grad_norm": 0.00042644611676223576,
      "learning_rate": 0.00100676319237461,
      "loss": 0.2754,
      "num_input_tokens_seen": 29383256,
      "step": 38525
    },
    {
      "epoch": 80.1039501039501,
      "grad_norm": 0.00028283815481700003,
      "learning_rate": 0.0009999614323110972,
      "loss": 0.2654,
      "num_input_tokens_seen": 29387032,
      "step": 38530
    },
    {
      "epoch": 80.11434511434511,
      "grad_norm": 0.000180199378519319,
      "learning_rate": 0.000993182649926011,
      "loss": 0.2602,
      "num_input_tokens_seen": 29390808,
      "step": 38535
    },
    {
      "epoch": 80.12474012474013,
      "grad_norm": 9.865043830359355e-05,
      "learning_rate": 0.000986426846264682,
      "loss": 0.2652,
      "num_input_tokens_seen": 29394616,
      "step": 38540
    },
    {
      "epoch": 80.13513513513513,
      "grad_norm": 0.0003328242455609143,
      "learning_rate": 0.00097969402236896,
      "loss": 0.2579,
      "num_input_tokens_seen": 29398424,
      "step": 38545
    },
    {
      "epoch": 80.14553014553015,
      "grad_norm": 0.00047965397243387997,
      "learning_rate": 0.0009729841792771143,
      "loss": 0.2733,
      "num_input_tokens_seen": 29402232,
      "step": 38550
    },
    {
      "epoch": 80.15592515592516,
      "grad_norm": 0.000206598051590845,
      "learning_rate": 0.0009662973180239176,
      "loss": 0.2662,
      "num_input_tokens_seen": 29406136,
      "step": 38555
    },
    {
      "epoch": 80.16632016632016,
      "grad_norm": 0.0001571306202095002,
      "learning_rate": 0.0009596334396405448,
      "loss": 0.2553,
      "num_input_tokens_seen": 29409816,
      "step": 38560
    },
    {
      "epoch": 80.17671517671518,
      "grad_norm": 0.0002151893131667748,
      "learning_rate": 0.0009529925451546406,
      "loss": 0.2659,
      "num_input_tokens_seen": 29413624,
      "step": 38565
    },
    {
      "epoch": 80.18711018711019,
      "grad_norm": 0.000471391569590196,
      "learning_rate": 0.0009463746355903357,
      "loss": 0.2453,
      "num_input_tokens_seen": 29417304,
      "step": 38570
    },
    {
      "epoch": 80.1975051975052,
      "grad_norm": 0.0002690582477953285,
      "learning_rate": 0.0009397797119681971,
      "loss": 0.2653,
      "num_input_tokens_seen": 29420984,
      "step": 38575
    },
    {
      "epoch": 80.20790020790021,
      "grad_norm": 0.00046467757783830166,
      "learning_rate": 0.0009332077753052281,
      "loss": 0.2603,
      "num_input_tokens_seen": 29424824,
      "step": 38580
    },
    {
      "epoch": 80.21829521829522,
      "grad_norm": 0.00018763489788398147,
      "learning_rate": 0.0009266588266149011,
      "loss": 0.2607,
      "num_input_tokens_seen": 29428696,
      "step": 38585
    },
    {
      "epoch": 80.22869022869023,
      "grad_norm": 0.0005945999873802066,
      "learning_rate": 0.0009201328669071584,
      "loss": 0.2273,
      "num_input_tokens_seen": 29432568,
      "step": 38590
    },
    {
      "epoch": 80.23908523908524,
      "grad_norm": 0.0007603820995427668,
      "learning_rate": 0.0009136298971883949,
      "loss": 0.2823,
      "num_input_tokens_seen": 29436472,
      "step": 38595
    },
    {
      "epoch": 80.24948024948024,
      "grad_norm": 0.00013617028889711946,
      "learning_rate": 0.0009071499184614251,
      "loss": 0.269,
      "num_input_tokens_seen": 29440312,
      "step": 38600
    },
    {
      "epoch": 80.24948024948024,
      "eval_loss": 0.2489522099494934,
      "eval_runtime": 13.3672,
      "eval_samples_per_second": 64.037,
      "eval_steps_per_second": 16.009,
      "num_input_tokens_seen": 29440312,
      "step": 38600
    },
    {
      "epoch": 80.25987525987526,
      "grad_norm": 0.0003282705438323319,
      "learning_rate": 0.0009006929317255663,
      "loss": 0.2673,
      "num_input_tokens_seen": 29444280,
      "step": 38605
    },
    {
      "epoch": 80.27027027027027,
      "grad_norm": 0.00017945414583664387,
      "learning_rate": 0.0008942589379765387,
      "loss": 0.2504,
      "num_input_tokens_seen": 29448024,
      "step": 38610
    },
    {
      "epoch": 80.28066528066527,
      "grad_norm": 0.00030554665136151016,
      "learning_rate": 0.0008878479382065817,
      "loss": 0.2689,
      "num_input_tokens_seen": 29451864,
      "step": 38615
    },
    {
      "epoch": 80.29106029106029,
      "grad_norm": 0.0008404122199863195,
      "learning_rate": 0.0008814599334043215,
      "loss": 0.2408,
      "num_input_tokens_seen": 29455672,
      "step": 38620
    },
    {
      "epoch": 80.3014553014553,
      "grad_norm": 0.00035232590744271874,
      "learning_rate": 0.0008750949245548866,
      "loss": 0.2668,
      "num_input_tokens_seen": 29459544,
      "step": 38625
    },
    {
      "epoch": 80.3118503118503,
      "grad_norm": 0.00018367475422564894,
      "learning_rate": 0.0008687529126398252,
      "loss": 0.2557,
      "num_input_tokens_seen": 29463256,
      "step": 38630
    },
    {
      "epoch": 80.32224532224532,
      "grad_norm": 8.786677062744275e-05,
      "learning_rate": 0.0008624338986371715,
      "loss": 0.2512,
      "num_input_tokens_seen": 29467064,
      "step": 38635
    },
    {
      "epoch": 80.33264033264034,
      "grad_norm": 0.00043246959103271365,
      "learning_rate": 0.0008561378835213962,
      "loss": 0.248,
      "num_input_tokens_seen": 29470808,
      "step": 38640
    },
    {
      "epoch": 80.34303534303534,
      "grad_norm": 0.002401023404672742,
      "learning_rate": 0.0008498648682634058,
      "loss": 0.2563,
      "num_input_tokens_seen": 29474616,
      "step": 38645
    },
    {
      "epoch": 80.35343035343035,
      "grad_norm": 6.830610072938725e-05,
      "learning_rate": 0.0008436148538306099,
      "loss": 0.2706,
      "num_input_tokens_seen": 29478392,
      "step": 38650
    },
    {
      "epoch": 80.36382536382537,
      "grad_norm": 0.0003837009717244655,
      "learning_rate": 0.0008373878411868041,
      "loss": 0.2487,
      "num_input_tokens_seen": 29482136,
      "step": 38655
    },
    {
      "epoch": 80.37422037422037,
      "grad_norm": 0.00035995221696794033,
      "learning_rate": 0.000831183831292287,
      "loss": 0.2851,
      "num_input_tokens_seen": 29485912,
      "step": 38660
    },
    {
      "epoch": 80.38461538461539,
      "grad_norm": 0.0004193721688352525,
      "learning_rate": 0.0008250028251037933,
      "loss": 0.2735,
      "num_input_tokens_seen": 29489784,
      "step": 38665
    },
    {
      "epoch": 80.39501039501039,
      "grad_norm": 0.00027916085673496127,
      "learning_rate": 0.0008188448235745271,
      "loss": 0.2552,
      "num_input_tokens_seen": 29493528,
      "step": 38670
    },
    {
      "epoch": 80.4054054054054,
      "grad_norm": 0.0002521621354389936,
      "learning_rate": 0.0008127098276541122,
      "loss": 0.2529,
      "num_input_tokens_seen": 29497272,
      "step": 38675
    },
    {
      "epoch": 80.41580041580042,
      "grad_norm": 0.00043592872680164874,
      "learning_rate": 0.0008065978382886418,
      "loss": 0.2674,
      "num_input_tokens_seen": 29501144,
      "step": 38680
    },
    {
      "epoch": 80.42619542619542,
      "grad_norm": 0.00041409360710531473,
      "learning_rate": 0.0008005088564206785,
      "loss": 0.2435,
      "num_input_tokens_seen": 29504888,
      "step": 38685
    },
    {
      "epoch": 80.43659043659044,
      "grad_norm": 0.0005598809220828116,
      "learning_rate": 0.0007944428829891881,
      "loss": 0.2531,
      "num_input_tokens_seen": 29508600,
      "step": 38690
    },
    {
      "epoch": 80.44698544698545,
      "grad_norm": 0.00016065462841652334,
      "learning_rate": 0.0007883999189296386,
      "loss": 0.2892,
      "num_input_tokens_seen": 29512408,
      "step": 38695
    },
    {
      "epoch": 80.45738045738045,
      "grad_norm": 0.00024615335860289633,
      "learning_rate": 0.0007823799651739515,
      "loss": 0.2691,
      "num_input_tokens_seen": 29516216,
      "step": 38700
    },
    {
      "epoch": 80.46777546777547,
      "grad_norm": 0.0003562385099940002,
      "learning_rate": 0.0007763830226504509,
      "loss": 0.2625,
      "num_input_tokens_seen": 29520024,
      "step": 38705
    },
    {
      "epoch": 80.47817047817048,
      "grad_norm": 0.0010006417287513614,
      "learning_rate": 0.0007704090922839468,
      "loss": 0.2752,
      "num_input_tokens_seen": 29523896,
      "step": 38710
    },
    {
      "epoch": 80.48856548856548,
      "grad_norm": 0.0006938848528079689,
      "learning_rate": 0.0007644581749957025,
      "loss": 0.2708,
      "num_input_tokens_seen": 29527800,
      "step": 38715
    },
    {
      "epoch": 80.4989604989605,
      "grad_norm": 0.006440659984946251,
      "learning_rate": 0.000758530271703417,
      "loss": 0.2778,
      "num_input_tokens_seen": 29531544,
      "step": 38720
    },
    {
      "epoch": 80.50935550935552,
      "grad_norm": 0.0003470568044576794,
      "learning_rate": 0.0007526253833212426,
      "loss": 0.2197,
      "num_input_tokens_seen": 29535480,
      "step": 38725
    },
    {
      "epoch": 80.51975051975052,
      "grad_norm": 0.00038155855145305395,
      "learning_rate": 0.0007467435107598008,
      "loss": 0.2622,
      "num_input_tokens_seen": 29539352,
      "step": 38730
    },
    {
      "epoch": 80.53014553014553,
      "grad_norm": 0.00013656876399181783,
      "learning_rate": 0.0007408846549261328,
      "loss": 0.2601,
      "num_input_tokens_seen": 29543000,
      "step": 38735
    },
    {
      "epoch": 80.54054054054055,
      "grad_norm": 0.0001753578835632652,
      "learning_rate": 0.0007350488167237656,
      "loss": 0.2473,
      "num_input_tokens_seen": 29546808,
      "step": 38740
    },
    {
      "epoch": 80.55093555093555,
      "grad_norm": 0.00041927240090444684,
      "learning_rate": 0.0007292359970526629,
      "loss": 0.27,
      "num_input_tokens_seen": 29550680,
      "step": 38745
    },
    {
      "epoch": 80.56133056133056,
      "grad_norm": 0.00014573319640476257,
      "learning_rate": 0.0007234461968092076,
      "loss": 0.2598,
      "num_input_tokens_seen": 29554584,
      "step": 38750
    },
    {
      "epoch": 80.57172557172557,
      "grad_norm": 0.00018213856674265116,
      "learning_rate": 0.0007176794168862854,
      "loss": 0.2791,
      "num_input_tokens_seen": 29558488,
      "step": 38755
    },
    {
      "epoch": 80.58212058212058,
      "grad_norm": 0.0014664718182757497,
      "learning_rate": 0.000711935658173185,
      "loss": 0.2509,
      "num_input_tokens_seen": 29562328,
      "step": 38760
    },
    {
      "epoch": 80.5925155925156,
      "grad_norm": 0.0008225443307310343,
      "learning_rate": 0.0007062149215556812,
      "loss": 0.282,
      "num_input_tokens_seen": 29566168,
      "step": 38765
    },
    {
      "epoch": 80.6029106029106,
      "grad_norm": 9.875069372355938e-05,
      "learning_rate": 0.0007005172079159849,
      "loss": 0.2358,
      "num_input_tokens_seen": 29570040,
      "step": 38770
    },
    {
      "epoch": 80.61330561330561,
      "grad_norm": 0.00019654349307529628,
      "learning_rate": 0.0006948425181327267,
      "loss": 0.2551,
      "num_input_tokens_seen": 29573816,
      "step": 38775
    },
    {
      "epoch": 80.62370062370063,
      "grad_norm": 0.0003359333495609462,
      "learning_rate": 0.000689190853081073,
      "loss": 0.2876,
      "num_input_tokens_seen": 29577688,
      "step": 38780
    },
    {
      "epoch": 80.63409563409563,
      "grad_norm": 0.0001950456207850948,
      "learning_rate": 0.000683562213632527,
      "loss": 0.2603,
      "num_input_tokens_seen": 29581464,
      "step": 38785
    },
    {
      "epoch": 80.64449064449065,
      "grad_norm": 0.00020145326561760157,
      "learning_rate": 0.0006779566006551108,
      "loss": 0.2641,
      "num_input_tokens_seen": 29585240,
      "step": 38790
    },
    {
      "epoch": 80.65488565488566,
      "grad_norm": 0.000315141020109877,
      "learning_rate": 0.0006723740150132995,
      "loss": 0.2534,
      "num_input_tokens_seen": 29589144,
      "step": 38795
    },
    {
      "epoch": 80.66528066528066,
      "grad_norm": 0.0003175473539158702,
      "learning_rate": 0.0006668144575679713,
      "loss": 0.2627,
      "num_input_tokens_seen": 29592888,
      "step": 38800
    },
    {
      "epoch": 80.66528066528066,
      "eval_loss": 0.24905961751937866,
      "eval_runtime": 13.37,
      "eval_samples_per_second": 64.024,
      "eval_steps_per_second": 16.006,
      "num_input_tokens_seen": 29592888,
      "step": 38800
    },
    {
      "epoch": 80.67567567567568,
      "grad_norm": 0.0004954852629452944,
      "learning_rate": 0.0006612779291765069,
      "loss": 0.2892,
      "num_input_tokens_seen": 29596696,
      "step": 38805
    },
    {
      "epoch": 80.68607068607069,
      "grad_norm": 0.0008529546903446317,
      "learning_rate": 0.0006557644306926736,
      "loss": 0.2736,
      "num_input_tokens_seen": 29600376,
      "step": 38810
    },
    {
      "epoch": 80.6964656964657,
      "grad_norm": 0.0004440688935574144,
      "learning_rate": 0.0006502739629667575,
      "loss": 0.2727,
      "num_input_tokens_seen": 29604216,
      "step": 38815
    },
    {
      "epoch": 80.70686070686071,
      "grad_norm": 0.0007419568719342351,
      "learning_rate": 0.0006448065268454317,
      "loss": 0.2757,
      "num_input_tokens_seen": 29608056,
      "step": 38820
    },
    {
      "epoch": 80.71725571725571,
      "grad_norm": 9.028362546814606e-05,
      "learning_rate": 0.0006393621231718549,
      "loss": 0.2662,
      "num_input_tokens_seen": 29611896,
      "step": 38825
    },
    {
      "epoch": 80.72765072765073,
      "grad_norm": 0.0003431664081290364,
      "learning_rate": 0.0006339407527856389,
      "loss": 0.267,
      "num_input_tokens_seen": 29615768,
      "step": 38830
    },
    {
      "epoch": 80.73804573804574,
      "grad_norm": 0.0011426162673160434,
      "learning_rate": 0.0006285424165227982,
      "loss": 0.2522,
      "num_input_tokens_seen": 29619768,
      "step": 38835
    },
    {
      "epoch": 80.74844074844074,
      "grad_norm": 0.00010065758397104219,
      "learning_rate": 0.0006231671152158169,
      "loss": 0.2621,
      "num_input_tokens_seen": 29623672,
      "step": 38840
    },
    {
      "epoch": 80.75883575883576,
      "grad_norm": 0.0007315367693081498,
      "learning_rate": 0.0006178148496936819,
      "loss": 0.2634,
      "num_input_tokens_seen": 29627544,
      "step": 38845
    },
    {
      "epoch": 80.76923076923077,
      "grad_norm": 7.239534897962585e-05,
      "learning_rate": 0.000612485620781733,
      "loss": 0.2523,
      "num_input_tokens_seen": 29631416,
      "step": 38850
    },
    {
      "epoch": 80.77962577962577,
      "grad_norm": 0.00030909874476492405,
      "learning_rate": 0.0006071794293018296,
      "loss": 0.2581,
      "num_input_tokens_seen": 29635096,
      "step": 38855
    },
    {
      "epoch": 80.79002079002079,
      "grad_norm": 0.000763608724810183,
      "learning_rate": 0.0006018962760722501,
      "loss": 0.2755,
      "num_input_tokens_seen": 29638936,
      "step": 38860
    },
    {
      "epoch": 80.8004158004158,
      "grad_norm": 0.0003426999901421368,
      "learning_rate": 0.0005966361619077098,
      "loss": 0.2917,
      "num_input_tokens_seen": 29642744,
      "step": 38865
    },
    {
      "epoch": 80.8108108108108,
      "grad_norm": 0.0001761830208124593,
      "learning_rate": 0.000591399087619393,
      "loss": 0.2714,
      "num_input_tokens_seen": 29646552,
      "step": 38870
    },
    {
      "epoch": 80.82120582120582,
      "grad_norm": 8.107997564366087e-05,
      "learning_rate": 0.0005861850540149371,
      "loss": 0.2613,
      "num_input_tokens_seen": 29650456,
      "step": 38875
    },
    {
      "epoch": 80.83160083160084,
      "grad_norm": 0.000688837026245892,
      "learning_rate": 0.0005809940618983822,
      "loss": 0.2806,
      "num_input_tokens_seen": 29654232,
      "step": 38880
    },
    {
      "epoch": 80.84199584199584,
      "grad_norm": 0.0016362975584343076,
      "learning_rate": 0.0005758261120702712,
      "loss": 0.2579,
      "num_input_tokens_seen": 29657976,
      "step": 38885
    },
    {
      "epoch": 80.85239085239085,
      "grad_norm": 6.6026346758008e-05,
      "learning_rate": 0.0005706812053275501,
      "loss": 0.262,
      "num_input_tokens_seen": 29661912,
      "step": 38890
    },
    {
      "epoch": 80.86278586278586,
      "grad_norm": 0.00039615307468920946,
      "learning_rate": 0.0005655593424636173,
      "loss": 0.2275,
      "num_input_tokens_seen": 29665656,
      "step": 38895
    },
    {
      "epoch": 80.87318087318087,
      "grad_norm": 0.000295100238872692,
      "learning_rate": 0.0005604605242683746,
      "loss": 0.2479,
      "num_input_tokens_seen": 29669528,
      "step": 38900
    },
    {
      "epoch": 80.88357588357589,
      "grad_norm": 0.0003159618645440787,
      "learning_rate": 0.0005553847515280596,
      "loss": 0.2704,
      "num_input_tokens_seen": 29673432,
      "step": 38905
    },
    {
      "epoch": 80.89397089397089,
      "grad_norm": 0.00014353074948303401,
      "learning_rate": 0.0005503320250254795,
      "loss": 0.2586,
      "num_input_tokens_seen": 29677464,
      "step": 38910
    },
    {
      "epoch": 80.9043659043659,
      "grad_norm": 0.00014917370572220534,
      "learning_rate": 0.0005453023455397943,
      "loss": 0.2588,
      "num_input_tokens_seen": 29681304,
      "step": 38915
    },
    {
      "epoch": 80.91476091476092,
      "grad_norm": 0.00016284633602481335,
      "learning_rate": 0.0005402957138466502,
      "loss": 0.2421,
      "num_input_tokens_seen": 29685016,
      "step": 38920
    },
    {
      "epoch": 80.92515592515592,
      "grad_norm": 0.0003256556810811162,
      "learning_rate": 0.0005353121307181463,
      "loss": 0.2558,
      "num_input_tokens_seen": 29688696,
      "step": 38925
    },
    {
      "epoch": 80.93555093555094,
      "grad_norm": 0.00019272591453045607,
      "learning_rate": 0.0005303515969227845,
      "loss": 0.2713,
      "num_input_tokens_seen": 29692440,
      "step": 38930
    },
    {
      "epoch": 80.94594594594595,
      "grad_norm": 0.00023034511832520366,
      "learning_rate": 0.0005254141132255862,
      "loss": 0.28,
      "num_input_tokens_seen": 29696216,
      "step": 38935
    },
    {
      "epoch": 80.95634095634095,
      "grad_norm": 0.000148714505485259,
      "learning_rate": 0.0005204996803879258,
      "loss": 0.2624,
      "num_input_tokens_seen": 29700056,
      "step": 38940
    },
    {
      "epoch": 80.96673596673597,
      "grad_norm": 0.0005001542740501463,
      "learning_rate": 0.0005156082991676969,
      "loss": 0.2642,
      "num_input_tokens_seen": 29703832,
      "step": 38945
    },
    {
      "epoch": 80.97713097713098,
      "grad_norm": 0.00046157496399246156,
      "learning_rate": 0.0005107399703192127,
      "loss": 0.2257,
      "num_input_tokens_seen": 29707608,
      "step": 38950
    },
    {
      "epoch": 80.98752598752598,
      "grad_norm": 0.000228060336667113,
      "learning_rate": 0.0005058946945932063,
      "loss": 0.2626,
      "num_input_tokens_seen": 29711352,
      "step": 38955
    },
    {
      "epoch": 80.997920997921,
      "grad_norm": 0.00011820188228739426,
      "learning_rate": 0.0005010724727369131,
      "loss": 0.2772,
      "num_input_tokens_seen": 29715096,
      "step": 38960
    },
    {
      "epoch": 81.00831600831602,
      "grad_norm": 0.0002430666791042313,
      "learning_rate": 0.000496273305493955,
      "loss": 0.2728,
      "num_input_tokens_seen": 29718824,
      "step": 38965
    },
    {
      "epoch": 81.01871101871102,
      "grad_norm": 0.0003784166183322668,
      "learning_rate": 0.0004914971936044399,
      "loss": 0.237,
      "num_input_tokens_seen": 29722632,
      "step": 38970
    },
    {
      "epoch": 81.02910602910603,
      "grad_norm": 0.0003129137621726841,
      "learning_rate": 0.00048674413780491196,
      "loss": 0.2859,
      "num_input_tokens_seen": 29726408,
      "step": 38975
    },
    {
      "epoch": 81.03950103950103,
      "grad_norm": 0.0002342287771170959,
      "learning_rate": 0.0004820141388283183,
      "loss": 0.2625,
      "num_input_tokens_seen": 29730088,
      "step": 38980
    },
    {
      "epoch": 81.04989604989605,
      "grad_norm": 0.0002921112172771245,
      "learning_rate": 0.00047730719740410874,
      "loss": 0.258,
      "num_input_tokens_seen": 29733992,
      "step": 38985
    },
    {
      "epoch": 81.06029106029106,
      "grad_norm": 0.0007997443899512291,
      "learning_rate": 0.00047262331425816927,
      "loss": 0.2657,
      "num_input_tokens_seen": 29737736,
      "step": 38990
    },
    {
      "epoch": 81.07068607068607,
      "grad_norm": 0.0005732898716814816,
      "learning_rate": 0.00046796249011277213,
      "loss": 0.251,
      "num_input_tokens_seen": 29741512,
      "step": 38995
    },
    {
      "epoch": 81.08108108108108,
      "grad_norm": 0.00032948559965007007,
      "learning_rate": 0.00046332472568669236,
      "loss": 0.2656,
      "num_input_tokens_seen": 29745320,
      "step": 39000
    },
    {
      "epoch": 81.08108108108108,
      "eval_loss": 0.24877241253852844,
      "eval_runtime": 13.3787,
      "eval_samples_per_second": 63.982,
      "eval_steps_per_second": 15.996,
      "num_input_tokens_seen": 29745320,
      "step": 39000
    },
    {
      "epoch": 81.0914760914761,
      "grad_norm": 6.946392386453226e-05,
      "learning_rate": 0.0004587100216951578,
      "loss": 0.2713,
      "num_input_tokens_seen": 29749032,
      "step": 39005
    },
    {
      "epoch": 81.1018711018711,
      "grad_norm": 0.0005109703633934259,
      "learning_rate": 0.00045411837884978265,
      "loss": 0.2346,
      "num_input_tokens_seen": 29752744,
      "step": 39010
    },
    {
      "epoch": 81.11226611226611,
      "grad_norm": 0.0003009248757734895,
      "learning_rate": 0.00044954979785865045,
      "loss": 0.2862,
      "num_input_tokens_seen": 29756680,
      "step": 39015
    },
    {
      "epoch": 81.12266112266113,
      "grad_norm": 0.00025270512560382485,
      "learning_rate": 0.00044500427942631426,
      "loss": 0.2781,
      "num_input_tokens_seen": 29760520,
      "step": 39020
    },
    {
      "epoch": 81.13305613305613,
      "grad_norm": 0.0008549754857085645,
      "learning_rate": 0.0004404818242537467,
      "loss": 0.2735,
      "num_input_tokens_seen": 29764232,
      "step": 39025
    },
    {
      "epoch": 81.14345114345114,
      "grad_norm": 0.0007490808493457735,
      "learning_rate": 0.00043598243303837324,
      "loss": 0.2438,
      "num_input_tokens_seen": 29767880,
      "step": 39030
    },
    {
      "epoch": 81.15384615384616,
      "grad_norm": 0.0004002944042440504,
      "learning_rate": 0.00043150610647403885,
      "loss": 0.2641,
      "num_input_tokens_seen": 29771720,
      "step": 39035
    },
    {
      "epoch": 81.16424116424116,
      "grad_norm": 0.00028097417089156806,
      "learning_rate": 0.00042705284525104134,
      "loss": 0.2787,
      "num_input_tokens_seen": 29775624,
      "step": 39040
    },
    {
      "epoch": 81.17463617463618,
      "grad_norm": 0.00032751832623034716,
      "learning_rate": 0.0004226226500561647,
      "loss": 0.2444,
      "num_input_tokens_seen": 29779400,
      "step": 39045
    },
    {
      "epoch": 81.18503118503118,
      "grad_norm": 0.000718900584615767,
      "learning_rate": 0.0004182155215725791,
      "loss": 0.2363,
      "num_input_tokens_seen": 29783368,
      "step": 39050
    },
    {
      "epoch": 81.1954261954262,
      "grad_norm": 0.0004228925099596381,
      "learning_rate": 0.00041383146047992424,
      "loss": 0.2594,
      "num_input_tokens_seen": 29787144,
      "step": 39055
    },
    {
      "epoch": 81.20582120582121,
      "grad_norm": 0.0006319344392977655,
      "learning_rate": 0.00040947046745427597,
      "loss": 0.2745,
      "num_input_tokens_seen": 29790984,
      "step": 39060
    },
    {
      "epoch": 81.21621621621621,
      "grad_norm": 0.0003901736927218735,
      "learning_rate": 0.00040513254316814625,
      "loss": 0.2524,
      "num_input_tokens_seen": 29794856,
      "step": 39065
    },
    {
      "epoch": 81.22661122661123,
      "grad_norm": 0.00023066597350407392,
      "learning_rate": 0.0004008176882905168,
      "loss": 0.247,
      "num_input_tokens_seen": 29798664,
      "step": 39070
    },
    {
      "epoch": 81.23700623700624,
      "grad_norm": 0.00028419942827895284,
      "learning_rate": 0.00039652590348677184,
      "loss": 0.2797,
      "num_input_tokens_seen": 29802344,
      "step": 39075
    },
    {
      "epoch": 81.24740124740124,
      "grad_norm": 0.0009782575070858002,
      "learning_rate": 0.00039225718941878206,
      "loss": 0.2681,
      "num_input_tokens_seen": 29806120,
      "step": 39080
    },
    {
      "epoch": 81.25779625779626,
      "grad_norm": 0.00029170283232815564,
      "learning_rate": 0.00038801154674480417,
      "loss": 0.2804,
      "num_input_tokens_seen": 29810024,
      "step": 39085
    },
    {
      "epoch": 81.26819126819127,
      "grad_norm": 0.00010834466229425743,
      "learning_rate": 0.00038378897611959784,
      "loss": 0.2646,
      "num_input_tokens_seen": 29813832,
      "step": 39090
    },
    {
      "epoch": 81.27858627858627,
      "grad_norm": 0.00031477221637032926,
      "learning_rate": 0.00037958947819430875,
      "loss": 0.2546,
      "num_input_tokens_seen": 29817800,
      "step": 39095
    },
    {
      "epoch": 81.28898128898129,
      "grad_norm": 0.00011645103222690523,
      "learning_rate": 0.0003754130536165856,
      "loss": 0.2602,
      "num_input_tokens_seen": 29821576,
      "step": 39100
    },
    {
      "epoch": 81.2993762993763,
      "grad_norm": 0.00046451727394014597,
      "learning_rate": 0.0003712597030304632,
      "loss": 0.2324,
      "num_input_tokens_seen": 29825512,
      "step": 39105
    },
    {
      "epoch": 81.3097713097713,
      "grad_norm": 0.00027248976402916014,
      "learning_rate": 0.00036712942707646247,
      "loss": 0.2426,
      "num_input_tokens_seen": 29829288,
      "step": 39110
    },
    {
      "epoch": 81.32016632016632,
      "grad_norm": 0.00030916763353161514,
      "learning_rate": 0.00036302222639149063,
      "loss": 0.251,
      "num_input_tokens_seen": 29833192,
      "step": 39115
    },
    {
      "epoch": 81.33056133056132,
      "grad_norm": 0.00022757383703719825,
      "learning_rate": 0.000358938101608941,
      "loss": 0.2704,
      "num_input_tokens_seen": 29836904,
      "step": 39120
    },
    {
      "epoch": 81.34095634095634,
      "grad_norm": 0.00039262103382498026,
      "learning_rate": 0.0003548770533586598,
      "loss": 0.2542,
      "num_input_tokens_seen": 29840744,
      "step": 39125
    },
    {
      "epoch": 81.35135135135135,
      "grad_norm": 0.0001572659966768697,
      "learning_rate": 0.0003508390822668961,
      "loss": 0.2876,
      "num_input_tokens_seen": 29844552,
      "step": 39130
    },
    {
      "epoch": 81.36174636174636,
      "grad_norm": 0.0003366860037203878,
      "learning_rate": 0.00034682418895633503,
      "loss": 0.2551,
      "num_input_tokens_seen": 29848424,
      "step": 39135
    },
    {
      "epoch": 81.37214137214137,
      "grad_norm": 0.00012265951954759657,
      "learning_rate": 0.0003428323740461647,
      "loss": 0.2631,
      "num_input_tokens_seen": 29852488,
      "step": 39140
    },
    {
      "epoch": 81.38253638253639,
      "grad_norm": 0.00023194070672616363,
      "learning_rate": 0.00033886363815194276,
      "loss": 0.2586,
      "num_input_tokens_seen": 29856328,
      "step": 39145
    },
    {
      "epoch": 81.39293139293139,
      "grad_norm": 0.00025766942417249084,
      "learning_rate": 0.0003349179818857129,
      "loss": 0.2609,
      "num_input_tokens_seen": 29860232,
      "step": 39150
    },
    {
      "epoch": 81.4033264033264,
      "grad_norm": 0.0005640723975375295,
      "learning_rate": 0.0003309954058559383,
      "loss": 0.2825,
      "num_input_tokens_seen": 29864200,
      "step": 39155
    },
    {
      "epoch": 81.41372141372142,
      "grad_norm": 0.00011428180005168542,
      "learning_rate": 0.0003270959106675186,
      "loss": 0.2418,
      "num_input_tokens_seen": 29868072,
      "step": 39160
    },
    {
      "epoch": 81.42411642411642,
      "grad_norm": 0.0005130151403136551,
      "learning_rate": 0.0003232194969218227,
      "loss": 0.2886,
      "num_input_tokens_seen": 29871976,
      "step": 39165
    },
    {
      "epoch": 81.43451143451144,
      "grad_norm": 0.0002611126401461661,
      "learning_rate": 0.00031936616521663905,
      "loss": 0.2711,
      "num_input_tokens_seen": 29875816,
      "step": 39170
    },
    {
      "epoch": 81.44490644490645,
      "grad_norm": 0.0007460659253410995,
      "learning_rate": 0.00031553591614619236,
      "loss": 0.2595,
      "num_input_tokens_seen": 29879560,
      "step": 39175
    },
    {
      "epoch": 81.45530145530145,
      "grad_norm": 0.00036192481638863683,
      "learning_rate": 0.00031172875030117676,
      "loss": 0.2557,
      "num_input_tokens_seen": 29883176,
      "step": 39180
    },
    {
      "epoch": 81.46569646569647,
      "grad_norm": 0.00012861716095358133,
      "learning_rate": 0.0003079446682686726,
      "loss": 0.2327,
      "num_input_tokens_seen": 29887176,
      "step": 39185
    },
    {
      "epoch": 81.47609147609148,
      "grad_norm": 0.00029900582740083337,
      "learning_rate": 0.0003041836706322465,
      "loss": 0.2636,
      "num_input_tokens_seen": 29890952,
      "step": 39190
    },
    {
      "epoch": 81.48648648648648,
      "grad_norm": 0.0001206764645758085,
      "learning_rate": 0.0003004457579719011,
      "loss": 0.2593,
      "num_input_tokens_seen": 29894792,
      "step": 39195
    },
    {
      "epoch": 81.4968814968815,
      "grad_norm": 0.0017916184151545167,
      "learning_rate": 0.00029673093086405867,
      "loss": 0.2409,
      "num_input_tokens_seen": 29898600,
      "step": 39200
    },
    {
      "epoch": 81.4968814968815,
      "eval_loss": 0.24881578981876373,
      "eval_runtime": 13.3805,
      "eval_samples_per_second": 63.974,
      "eval_steps_per_second": 15.993,
      "num_input_tokens_seen": 29898600,
      "step": 39200
    },
    {
      "epoch": 81.5072765072765,
      "grad_norm": 0.0003137363528367132,
      "learning_rate": 0.00029303918988159426,
      "loss": 0.2501,
      "num_input_tokens_seen": 29902440,
      "step": 39205
    },
    {
      "epoch": 81.51767151767152,
      "grad_norm": 0.0006595077575184405,
      "learning_rate": 0.0002893705355938192,
      "loss": 0.2673,
      "num_input_tokens_seen": 29906376,
      "step": 39210
    },
    {
      "epoch": 81.52806652806653,
      "grad_norm": 0.00019991387671325356,
      "learning_rate": 0.0002857249685664975,
      "loss": 0.263,
      "num_input_tokens_seen": 29910152,
      "step": 39215
    },
    {
      "epoch": 81.53846153846153,
      "grad_norm": 0.0005949345068074763,
      "learning_rate": 0.0002821024893618129,
      "loss": 0.2453,
      "num_input_tokens_seen": 29913896,
      "step": 39220
    },
    {
      "epoch": 81.54885654885655,
      "grad_norm": 0.0002964503364637494,
      "learning_rate": 0.0002785030985383852,
      "loss": 0.2553,
      "num_input_tokens_seen": 29917672,
      "step": 39225
    },
    {
      "epoch": 81.55925155925156,
      "grad_norm": 0.0003674213367048651,
      "learning_rate": 0.00027492679665130356,
      "loss": 0.2748,
      "num_input_tokens_seen": 29921448,
      "step": 39230
    },
    {
      "epoch": 81.56964656964657,
      "grad_norm": 0.0002499967231415212,
      "learning_rate": 0.000271373584252077,
      "loss": 0.2627,
      "num_input_tokens_seen": 29925288,
      "step": 39235
    },
    {
      "epoch": 81.58004158004158,
      "grad_norm": 0.000379693548893556,
      "learning_rate": 0.00026784346188865046,
      "loss": 0.2592,
      "num_input_tokens_seen": 29929192,
      "step": 39240
    },
    {
      "epoch": 81.5904365904366,
      "grad_norm": 0.0003246490960009396,
      "learning_rate": 0.0002643364301054218,
      "loss": 0.2814,
      "num_input_tokens_seen": 29932872,
      "step": 39245
    },
    {
      "epoch": 81.6008316008316,
      "grad_norm": 0.0006923344335518777,
      "learning_rate": 0.0002608524894431918,
      "loss": 0.2485,
      "num_input_tokens_seen": 29936872,
      "step": 39250
    },
    {
      "epoch": 81.61122661122661,
      "grad_norm": 0.00026829156558960676,
      "learning_rate": 0.000257391640439264,
      "loss": 0.2476,
      "num_input_tokens_seen": 29940552,
      "step": 39255
    },
    {
      "epoch": 81.62162162162163,
      "grad_norm": 3.881260272464715e-05,
      "learning_rate": 0.00025395388362732806,
      "loss": 0.2798,
      "num_input_tokens_seen": 29944488,
      "step": 39260
    },
    {
      "epoch": 81.63201663201663,
      "grad_norm": 0.0017335637239739299,
      "learning_rate": 0.00025053921953751,
      "loss": 0.2657,
      "num_input_tokens_seen": 29948296,
      "step": 39265
    },
    {
      "epoch": 81.64241164241164,
      "grad_norm": 0.0003677213389892131,
      "learning_rate": 0.00024714764869643855,
      "loss": 0.2709,
      "num_input_tokens_seen": 29951912,
      "step": 39270
    },
    {
      "epoch": 81.65280665280665,
      "grad_norm": 0.0005958377732895315,
      "learning_rate": 0.0002437791716270954,
      "loss": 0.2657,
      "num_input_tokens_seen": 29955592,
      "step": 39275
    },
    {
      "epoch": 81.66320166320166,
      "grad_norm": 0.00014409460709430277,
      "learning_rate": 0.00024043378884896493,
      "loss": 0.2596,
      "num_input_tokens_seen": 29959528,
      "step": 39280
    },
    {
      "epoch": 81.67359667359668,
      "grad_norm": 0.0011817305348813534,
      "learning_rate": 0.00023711150087793453,
      "loss": 0.2554,
      "num_input_tokens_seen": 29963304,
      "step": 39285
    },
    {
      "epoch": 81.68399168399168,
      "grad_norm": 0.0005550046917051077,
      "learning_rate": 0.000233812308226361,
      "loss": 0.2582,
      "num_input_tokens_seen": 29967176,
      "step": 39290
    },
    {
      "epoch": 81.6943866943867,
      "grad_norm": 0.001390982884913683,
      "learning_rate": 0.00023053621140300406,
      "loss": 0.2811,
      "num_input_tokens_seen": 29970824,
      "step": 39295
    },
    {
      "epoch": 81.70478170478171,
      "grad_norm": 0.00029024569084867835,
      "learning_rate": 0.00022728321091307623,
      "loss": 0.234,
      "num_input_tokens_seen": 29974568,
      "step": 39300
    },
    {
      "epoch": 81.71517671517671,
      "grad_norm": 0.00040963542414829135,
      "learning_rate": 0.0002240533072582429,
      "loss": 0.246,
      "num_input_tokens_seen": 29978536,
      "step": 39305
    },
    {
      "epoch": 81.72557172557173,
      "grad_norm": 0.00022900060866959393,
      "learning_rate": 0.00022084650093658897,
      "loss": 0.2893,
      "num_input_tokens_seen": 29982344,
      "step": 39310
    },
    {
      "epoch": 81.73596673596674,
      "grad_norm": 0.0003072502149734646,
      "learning_rate": 0.0002176627924426522,
      "loss": 0.2633,
      "num_input_tokens_seen": 29986184,
      "step": 39315
    },
    {
      "epoch": 81.74636174636174,
      "grad_norm": 0.0001275045215152204,
      "learning_rate": 0.0002145021822673898,
      "loss": 0.2463,
      "num_input_tokens_seen": 29990056,
      "step": 39320
    },
    {
      "epoch": 81.75675675675676,
      "grad_norm": 0.0003510946989990771,
      "learning_rate": 0.00021136467089822862,
      "loss": 0.2557,
      "num_input_tokens_seen": 29993768,
      "step": 39325
    },
    {
      "epoch": 81.76715176715177,
      "grad_norm": 0.0006697809440083802,
      "learning_rate": 0.00020825025881898162,
      "loss": 0.2446,
      "num_input_tokens_seen": 29997768,
      "step": 39330
    },
    {
      "epoch": 81.77754677754677,
      "grad_norm": 0.0002201414608862251,
      "learning_rate": 0.0002051589465099479,
      "loss": 0.2779,
      "num_input_tokens_seen": 30001416,
      "step": 39335
    },
    {
      "epoch": 81.78794178794179,
      "grad_norm": 0.0005062674172222614,
      "learning_rate": 0.0002020907344478462,
      "loss": 0.2636,
      "num_input_tokens_seen": 30005288,
      "step": 39340
    },
    {
      "epoch": 81.7983367983368,
      "grad_norm": 0.00018497595738153905,
      "learning_rate": 0.0001990456231058313,
      "loss": 0.2494,
      "num_input_tokens_seen": 30009064,
      "step": 39345
    },
    {
      "epoch": 81.8087318087318,
      "grad_norm": 0.00185938342474401,
      "learning_rate": 0.00019602361295349423,
      "loss": 0.2766,
      "num_input_tokens_seen": 30012904,
      "step": 39350
    },
    {
      "epoch": 81.81912681912682,
      "grad_norm": 0.00033537307172082365,
      "learning_rate": 0.0001930247044568789,
      "loss": 0.2776,
      "num_input_tokens_seen": 30016648,
      "step": 39355
    },
    {
      "epoch": 81.82952182952182,
      "grad_norm": 0.0001350029488094151,
      "learning_rate": 0.00019004889807843205,
      "loss": 0.2658,
      "num_input_tokens_seen": 30020456,
      "step": 39360
    },
    {
      "epoch": 81.83991683991684,
      "grad_norm": 0.00029065311537124217,
      "learning_rate": 0.00018709619427708656,
      "loss": 0.2825,
      "num_input_tokens_seen": 30024296,
      "step": 39365
    },
    {
      "epoch": 81.85031185031185,
      "grad_norm": 0.0003413350787013769,
      "learning_rate": 0.00018416659350817822,
      "loss": 0.2782,
      "num_input_tokens_seen": 30028136,
      "step": 39370
    },
    {
      "epoch": 81.86070686070686,
      "grad_norm": 8.406373672187328e-05,
      "learning_rate": 0.00018126009622346229,
      "loss": 0.2638,
      "num_input_tokens_seen": 30031944,
      "step": 39375
    },
    {
      "epoch": 81.87110187110187,
      "grad_norm": 0.0005150201031938195,
      "learning_rate": 0.00017837670287119687,
      "loss": 0.2387,
      "num_input_tokens_seen": 30035720,
      "step": 39380
    },
    {
      "epoch": 81.88149688149689,
      "grad_norm": 0.0007320127333514392,
      "learning_rate": 0.00017551641389602633,
      "loss": 0.2398,
      "num_input_tokens_seen": 30039400,
      "step": 39385
    },
    {
      "epoch": 81.89189189189189,
      "grad_norm": 0.0003280183009337634,
      "learning_rate": 0.00017267922973903115,
      "loss": 0.2665,
      "num_input_tokens_seen": 30043176,
      "step": 39390
    },
    {
      "epoch": 81.9022869022869,
      "grad_norm": 0.00015193450963124633,
      "learning_rate": 0.00016986515083774467,
      "loss": 0.2815,
      "num_input_tokens_seen": 30046824,
      "step": 39395
    },
    {
      "epoch": 81.91268191268192,
      "grad_norm": 0.0002811957092490047,
      "learning_rate": 0.00016707417762611975,
      "loss": 0.2619,
      "num_input_tokens_seen": 30050504,
      "step": 39400
    },
    {
      "epoch": 81.91268191268192,
      "eval_loss": 0.24916960299015045,
      "eval_runtime": 13.3732,
      "eval_samples_per_second": 64.008,
      "eval_steps_per_second": 16.002,
      "num_input_tokens_seen": 30050504,
      "step": 39400
    },
    {
      "epoch": 81.92307692307692,
      "grad_norm": 7.521611405536532e-05,
      "learning_rate": 0.00016430631053459543,
      "loss": 0.2895,
      "num_input_tokens_seen": 30054440,
      "step": 39405
    },
    {
      "epoch": 81.93347193347194,
      "grad_norm": 0.00027287990087643266,
      "learning_rate": 0.0001615615499899803,
      "loss": 0.2631,
      "num_input_tokens_seen": 30058312,
      "step": 39410
    },
    {
      "epoch": 81.94386694386695,
      "grad_norm": 0.002547452226281166,
      "learning_rate": 0.00015883989641556905,
      "loss": 0.28,
      "num_input_tokens_seen": 30062120,
      "step": 39415
    },
    {
      "epoch": 81.95426195426195,
      "grad_norm": 0.00022000692842993885,
      "learning_rate": 0.00015614135023105934,
      "loss": 0.289,
      "num_input_tokens_seen": 30065992,
      "step": 39420
    },
    {
      "epoch": 81.96465696465697,
      "grad_norm": 0.0005345044191926718,
      "learning_rate": 0.00015346591185261827,
      "loss": 0.2729,
      "num_input_tokens_seen": 30069896,
      "step": 39425
    },
    {
      "epoch": 81.97505197505197,
      "grad_norm": 0.0002813942264765501,
      "learning_rate": 0.00015081358169281576,
      "loss": 0.2506,
      "num_input_tokens_seen": 30073672,
      "step": 39430
    },
    {
      "epoch": 81.98544698544698,
      "grad_norm": 0.00011182101297890767,
      "learning_rate": 0.00014818436016069135,
      "loss": 0.2769,
      "num_input_tokens_seen": 30077512,
      "step": 39435
    },
    {
      "epoch": 81.995841995842,
      "grad_norm": 0.0004064147942699492,
      "learning_rate": 0.00014557824766168735,
      "loss": 0.2416,
      "num_input_tokens_seen": 30081416,
      "step": 39440
    },
    {
      "epoch": 82.006237006237,
      "grad_norm": 0.001103613874875009,
      "learning_rate": 0.00014299524459769896,
      "loss": 0.2466,
      "num_input_tokens_seen": 30085304,
      "step": 39445
    },
    {
      "epoch": 82.01663201663202,
      "grad_norm": 0.0007096627959981561,
      "learning_rate": 0.0001404353513670742,
      "loss": 0.2684,
      "num_input_tokens_seen": 30089080,
      "step": 39450
    },
    {
      "epoch": 82.02702702702703,
      "grad_norm": 0.00032989992178045213,
      "learning_rate": 0.0001378985683645806,
      "loss": 0.2713,
      "num_input_tokens_seen": 30092920,
      "step": 39455
    },
    {
      "epoch": 82.03742203742203,
      "grad_norm": 0.0001851690758485347,
      "learning_rate": 0.0001353848959813886,
      "loss": 0.2549,
      "num_input_tokens_seen": 30096728,
      "step": 39460
    },
    {
      "epoch": 82.04781704781705,
      "grad_norm": 0.0006134851137176156,
      "learning_rate": 0.00013289433460517142,
      "loss": 0.2332,
      "num_input_tokens_seen": 30100568,
      "step": 39465
    },
    {
      "epoch": 82.05821205821206,
      "grad_norm": 0.0002560316352173686,
      "learning_rate": 0.00013042688462000518,
      "loss": 0.2816,
      "num_input_tokens_seen": 30104376,
      "step": 39470
    },
    {
      "epoch": 82.06860706860707,
      "grad_norm": 0.0002895807265304029,
      "learning_rate": 0.0001279825464063855,
      "loss": 0.2663,
      "num_input_tokens_seen": 30108216,
      "step": 39475
    },
    {
      "epoch": 82.07900207900208,
      "grad_norm": 0.0010379767045378685,
      "learning_rate": 0.00012556132034126087,
      "loss": 0.239,
      "num_input_tokens_seen": 30112088,
      "step": 39480
    },
    {
      "epoch": 82.0893970893971,
      "grad_norm": 0.00022599281510338187,
      "learning_rate": 0.0001231632067980326,
      "loss": 0.2569,
      "num_input_tokens_seen": 30115896,
      "step": 39485
    },
    {
      "epoch": 82.0997920997921,
      "grad_norm": 5.704965224140324e-05,
      "learning_rate": 0.00012078820614650486,
      "loss": 0.2572,
      "num_input_tokens_seen": 30119576,
      "step": 39490
    },
    {
      "epoch": 82.11018711018711,
      "grad_norm": 0.00031116203172132373,
      "learning_rate": 0.00011843631875291804,
      "loss": 0.2626,
      "num_input_tokens_seen": 30123352,
      "step": 39495
    },
    {
      "epoch": 82.12058212058211,
      "grad_norm": 0.00028915132861584425,
      "learning_rate": 0.00011610754497999863,
      "loss": 0.2918,
      "num_input_tokens_seen": 30127192,
      "step": 39500
    },
    {
      "epoch": 82.13097713097713,
      "grad_norm": 0.00018232085858471692,
      "learning_rate": 0.0001138018851868594,
      "loss": 0.276,
      "num_input_tokens_seen": 30131096,
      "step": 39505
    },
    {
      "epoch": 82.14137214137214,
      "grad_norm": 0.0005327480612322688,
      "learning_rate": 0.0001115193397290326,
      "loss": 0.2566,
      "num_input_tokens_seen": 30134968,
      "step": 39510
    },
    {
      "epoch": 82.15176715176715,
      "grad_norm": 0.00025095490855164826,
      "learning_rate": 0.00010925990895856996,
      "loss": 0.2756,
      "num_input_tokens_seen": 30138808,
      "step": 39515
    },
    {
      "epoch": 82.16216216216216,
      "grad_norm": 0.0001093059909180738,
      "learning_rate": 0.00010702359322385946,
      "loss": 0.268,
      "num_input_tokens_seen": 30142552,
      "step": 39520
    },
    {
      "epoch": 82.17255717255718,
      "grad_norm": 0.0013536772457882762,
      "learning_rate": 0.00010481039286977523,
      "loss": 0.2702,
      "num_input_tokens_seen": 30146360,
      "step": 39525
    },
    {
      "epoch": 82.18295218295218,
      "grad_norm": 0.0005254528950899839,
      "learning_rate": 0.00010262030823764423,
      "loss": 0.2632,
      "num_input_tokens_seen": 30150008,
      "step": 39530
    },
    {
      "epoch": 82.1933471933472,
      "grad_norm": 0.0001025453457259573,
      "learning_rate": 0.00010045333966517966,
      "loss": 0.2697,
      "num_input_tokens_seen": 30153752,
      "step": 39535
    },
    {
      "epoch": 82.20374220374221,
      "grad_norm": 0.0001426387025276199,
      "learning_rate": 9.83094874865642e-05,
      "loss": 0.2635,
      "num_input_tokens_seen": 30157624,
      "step": 39540
    },
    {
      "epoch": 82.21413721413721,
      "grad_norm": 0.00043559836922213435,
      "learning_rate": 9.618875203241672e-05,
      "loss": 0.2667,
      "num_input_tokens_seen": 30161496,
      "step": 39545
    },
    {
      "epoch": 82.22453222453223,
      "grad_norm": 0.00016597587091382593,
      "learning_rate": 9.409113362977561e-05,
      "loss": 0.2545,
      "num_input_tokens_seen": 30165368,
      "step": 39550
    },
    {
      "epoch": 82.23492723492724,
      "grad_norm": 0.00024538402794860303,
      "learning_rate": 9.20166326020988e-05,
      "loss": 0.2807,
      "num_input_tokens_seen": 30169112,
      "step": 39555
    },
    {
      "epoch": 82.24532224532224,
      "grad_norm": 0.00018476434343028814,
      "learning_rate": 8.996524926933035e-05,
      "loss": 0.2729,
      "num_input_tokens_seen": 30172888,
      "step": 39560
    },
    {
      "epoch": 82.25571725571726,
      "grad_norm": 8.792390144662932e-05,
      "learning_rate": 8.793698394781723e-05,
      "loss": 0.2726,
      "num_input_tokens_seen": 30176792,
      "step": 39565
    },
    {
      "epoch": 82.26611226611226,
      "grad_norm": 0.00031042980845086277,
      "learning_rate": 8.593183695030926e-05,
      "loss": 0.2762,
      "num_input_tokens_seen": 30180664,
      "step": 39570
    },
    {
      "epoch": 82.27650727650727,
      "grad_norm": 0.0011632252717390656,
      "learning_rate": 8.39498085860757e-05,
      "loss": 0.2649,
      "num_input_tokens_seen": 30184536,
      "step": 39575
    },
    {
      "epoch": 82.28690228690229,
      "grad_norm": 0.00014371753786690533,
      "learning_rate": 8.199089916072211e-05,
      "loss": 0.2467,
      "num_input_tokens_seen": 30188504,
      "step": 39580
    },
    {
      "epoch": 82.29729729729729,
      "grad_norm": 0.0008486981969326735,
      "learning_rate": 8.005510897637346e-05,
      "loss": 0.2244,
      "num_input_tokens_seen": 30192280,
      "step": 39585
    },
    {
      "epoch": 82.3076923076923,
      "grad_norm": 0.0001485886168666184,
      "learning_rate": 7.8142438331541e-05,
      "loss": 0.2695,
      "num_input_tokens_seen": 30196056,
      "step": 39590
    },
    {
      "epoch": 82.31808731808732,
      "grad_norm": 0.00016745635366532952,
      "learning_rate": 7.625288752117209e-05,
      "loss": 0.2598,
      "num_input_tokens_seen": 30199768,
      "step": 39595
    },
    {
      "epoch": 82.32848232848232,
      "grad_norm": 0.00016985958791337907,
      "learning_rate": 7.4386456836667e-05,
      "loss": 0.2635,
      "num_input_tokens_seen": 30203576,
      "step": 39600
    },
    {
      "epoch": 82.32848232848232,
      "eval_loss": 0.2492460459470749,
      "eval_runtime": 13.3665,
      "eval_samples_per_second": 64.041,
      "eval_steps_per_second": 16.01,
      "num_input_tokens_seen": 30203576,
      "step": 39600
    },
    {
      "epoch": 82.33887733887734,
      "grad_norm": 0.0007666294113732874,
      "learning_rate": 7.254314656586214e-05,
      "loss": 0.272,
      "num_input_tokens_seen": 30207512,
      "step": 39605
    },
    {
      "epoch": 82.34927234927235,
      "grad_norm": 0.0005155249964445829,
      "learning_rate": 7.07229569929968e-05,
      "loss": 0.3037,
      "num_input_tokens_seen": 30211384,
      "step": 39610
    },
    {
      "epoch": 82.35966735966736,
      "grad_norm": 0.0007861669873818755,
      "learning_rate": 6.892588839879643e-05,
      "loss": 0.2513,
      "num_input_tokens_seen": 30215320,
      "step": 39615
    },
    {
      "epoch": 82.37006237006237,
      "grad_norm": 0.00015402685676235706,
      "learning_rate": 6.71519410603727e-05,
      "loss": 0.2784,
      "num_input_tokens_seen": 30219096,
      "step": 39620
    },
    {
      "epoch": 82.38045738045739,
      "grad_norm": 0.002041741507127881,
      "learning_rate": 6.540111525129011e-05,
      "loss": 0.2416,
      "num_input_tokens_seen": 30222872,
      "step": 39625
    },
    {
      "epoch": 82.39085239085239,
      "grad_norm": 0.0005694655701518059,
      "learning_rate": 6.367341124154934e-05,
      "loss": 0.2776,
      "num_input_tokens_seen": 30226680,
      "step": 39630
    },
    {
      "epoch": 82.4012474012474,
      "grad_norm": 0.0002953947114292532,
      "learning_rate": 6.19688292975873e-05,
      "loss": 0.2416,
      "num_input_tokens_seen": 30230616,
      "step": 39635
    },
    {
      "epoch": 82.41164241164242,
      "grad_norm": 0.0005577355623245239,
      "learning_rate": 6.0287369682260336e-05,
      "loss": 0.2419,
      "num_input_tokens_seen": 30234360,
      "step": 39640
    },
    {
      "epoch": 82.42203742203742,
      "grad_norm": 0.00024115278210956603,
      "learning_rate": 5.8629032654894384e-05,
      "loss": 0.2723,
      "num_input_tokens_seen": 30238040,
      "step": 39645
    },
    {
      "epoch": 82.43243243243244,
      "grad_norm": 0.00030726828845217824,
      "learning_rate": 5.699381847120155e-05,
      "loss": 0.2567,
      "num_input_tokens_seen": 30242008,
      "step": 39650
    },
    {
      "epoch": 82.44282744282744,
      "grad_norm": 0.00019583117682486773,
      "learning_rate": 5.5381727383380094e-05,
      "loss": 0.2591,
      "num_input_tokens_seen": 30245912,
      "step": 39655
    },
    {
      "epoch": 82.45322245322245,
      "grad_norm": 0.0010089607676491141,
      "learning_rate": 5.379275964001451e-05,
      "loss": 0.27,
      "num_input_tokens_seen": 30249752,
      "step": 39660
    },
    {
      "epoch": 82.46361746361747,
      "grad_norm": 5.344822056940757e-05,
      "learning_rate": 5.222691548614211e-05,
      "loss": 0.2403,
      "num_input_tokens_seen": 30253496,
      "step": 39665
    },
    {
      "epoch": 82.47401247401247,
      "grad_norm": 0.0002855885250028223,
      "learning_rate": 5.068419516323641e-05,
      "loss": 0.2419,
      "num_input_tokens_seen": 30257560,
      "step": 39670
    },
    {
      "epoch": 82.48440748440748,
      "grad_norm": 0.0008575962274335325,
      "learning_rate": 4.91645989092071e-05,
      "loss": 0.2806,
      "num_input_tokens_seen": 30261432,
      "step": 39675
    },
    {
      "epoch": 82.4948024948025,
      "grad_norm": 0.00026951474137604237,
      "learning_rate": 4.7668126958400056e-05,
      "loss": 0.277,
      "num_input_tokens_seen": 30265144,
      "step": 39680
    },
    {
      "epoch": 82.5051975051975,
      "grad_norm": 0.00010369114897912368,
      "learning_rate": 4.619477954159734e-05,
      "loss": 0.2432,
      "num_input_tokens_seen": 30269016,
      "step": 39685
    },
    {
      "epoch": 82.51559251559252,
      "grad_norm": 0.001255469978787005,
      "learning_rate": 4.4744556885983884e-05,
      "loss": 0.2662,
      "num_input_tokens_seen": 30272728,
      "step": 39690
    },
    {
      "epoch": 82.52598752598753,
      "grad_norm": 0.0003735451609827578,
      "learning_rate": 4.331745921523078e-05,
      "loss": 0.2748,
      "num_input_tokens_seen": 30276504,
      "step": 39695
    },
    {
      "epoch": 82.53638253638253,
      "grad_norm": 0.0009013859089463949,
      "learning_rate": 4.191348674937867e-05,
      "loss": 0.2641,
      "num_input_tokens_seen": 30280120,
      "step": 39700
    },
    {
      "epoch": 82.54677754677755,
      "grad_norm": 0.00038640134152956307,
      "learning_rate": 4.0532639704971006e-05,
      "loss": 0.2635,
      "num_input_tokens_seen": 30283832,
      "step": 39705
    },
    {
      "epoch": 82.55717255717256,
      "grad_norm": 0.0008163599413819611,
      "learning_rate": 3.917491829493747e-05,
      "loss": 0.2658,
      "num_input_tokens_seen": 30287736,
      "step": 39710
    },
    {
      "epoch": 82.56756756756756,
      "grad_norm": 0.00023090663307812065,
      "learning_rate": 3.78403227286439e-05,
      "loss": 0.2508,
      "num_input_tokens_seen": 30291576,
      "step": 39715
    },
    {
      "epoch": 82.57796257796258,
      "grad_norm": 0.0002808485005516559,
      "learning_rate": 3.652885321192567e-05,
      "loss": 0.2736,
      "num_input_tokens_seen": 30295416,
      "step": 39720
    },
    {
      "epoch": 82.58835758835758,
      "grad_norm": 0.00019189031445421278,
      "learning_rate": 3.524050994702099e-05,
      "loss": 0.2641,
      "num_input_tokens_seen": 30299352,
      "step": 39725
    },
    {
      "epoch": 82.5987525987526,
      "grad_norm": 0.0009198287734761834,
      "learning_rate": 3.3975293132604276e-05,
      "loss": 0.272,
      "num_input_tokens_seen": 30303192,
      "step": 39730
    },
    {
      "epoch": 82.60914760914761,
      "grad_norm": 0.00016505520034115762,
      "learning_rate": 3.2733202963786125e-05,
      "loss": 0.2671,
      "num_input_tokens_seen": 30306904,
      "step": 39735
    },
    {
      "epoch": 82.61954261954261,
      "grad_norm": 0.0004372768453322351,
      "learning_rate": 3.15142396321133e-05,
      "loss": 0.2594,
      "num_input_tokens_seen": 30310648,
      "step": 39740
    },
    {
      "epoch": 82.62993762993763,
      "grad_norm": 0.0002436877111904323,
      "learning_rate": 3.0318403325552132e-05,
      "loss": 0.2457,
      "num_input_tokens_seen": 30314488,
      "step": 39745
    },
    {
      "epoch": 82.64033264033264,
      "grad_norm": 0.000996051006950438,
      "learning_rate": 2.914569422855506e-05,
      "loss": 0.2755,
      "num_input_tokens_seen": 30318296,
      "step": 39750
    },
    {
      "epoch": 82.65072765072765,
      "grad_norm": 0.0002123875601682812,
      "learning_rate": 2.7996112521927462e-05,
      "loss": 0.2716,
      "num_input_tokens_seen": 30322232,
      "step": 39755
    },
    {
      "epoch": 82.66112266112266,
      "grad_norm": 0.0004449710249900818,
      "learning_rate": 2.68696583829775e-05,
      "loss": 0.2587,
      "num_input_tokens_seen": 30326040,
      "step": 39760
    },
    {
      "epoch": 82.67151767151768,
      "grad_norm": 0.0004485077515710145,
      "learning_rate": 2.576633198539957e-05,
      "loss": 0.268,
      "num_input_tokens_seen": 30329784,
      "step": 39765
    },
    {
      "epoch": 82.68191268191268,
      "grad_norm": 0.00030305111431516707,
      "learning_rate": 2.46861334993409e-05,
      "loss": 0.2656,
      "num_input_tokens_seen": 30333624,
      "step": 39770
    },
    {
      "epoch": 82.6923076923077,
      "grad_norm": 6.662662781309336e-05,
      "learning_rate": 2.3629063091384903e-05,
      "loss": 0.2624,
      "num_input_tokens_seen": 30337464,
      "step": 39775
    },
    {
      "epoch": 82.70270270270271,
      "grad_norm": 0.000767660269048065,
      "learning_rate": 2.2595120924567834e-05,
      "loss": 0.2537,
      "num_input_tokens_seen": 30341336,
      "step": 39780
    },
    {
      "epoch": 82.71309771309771,
      "grad_norm": 0.0005912585183978081,
      "learning_rate": 2.158430715829551e-05,
      "loss": 0.2268,
      "num_input_tokens_seen": 30345144,
      "step": 39785
    },
    {
      "epoch": 82.72349272349273,
      "grad_norm": 0.0004818747693207115,
      "learning_rate": 2.059662194849321e-05,
      "loss": 0.2772,
      "num_input_tokens_seen": 30348888,
      "step": 39790
    },
    {
      "epoch": 82.73388773388774,
      "grad_norm": 0.00025212852051481605,
      "learning_rate": 1.9632065447422463e-05,
      "loss": 0.2722,
      "num_input_tokens_seen": 30352568,
      "step": 39795
    },
    {
      "epoch": 82.74428274428274,
      "grad_norm": 0.0006203068769536912,
      "learning_rate": 1.8690637803880916e-05,
      "loss": 0.2449,
      "num_input_tokens_seen": 30356408,
      "step": 39800
    },
    {
      "epoch": 82.74428274428274,
      "eval_loss": 0.2485324889421463,
      "eval_runtime": 13.3747,
      "eval_samples_per_second": 64.001,
      "eval_steps_per_second": 16.0,
      "num_input_tokens_seen": 30356408,
      "step": 39800
    },
    {
      "epoch": 82.75467775467776,
      "grad_norm": 0.0005188753711991012,
      "learning_rate": 1.7772339163019123e-05,
      "loss": 0.2541,
      "num_input_tokens_seen": 30360216,
      "step": 39805
    },
    {
      "epoch": 82.76507276507276,
      "grad_norm": 0.0003956534492317587,
      "learning_rate": 1.6877169666457138e-05,
      "loss": 0.2616,
      "num_input_tokens_seen": 30363896,
      "step": 39810
    },
    {
      "epoch": 82.77546777546777,
      "grad_norm": 0.000282161810901016,
      "learning_rate": 1.6005129452234532e-05,
      "loss": 0.2592,
      "num_input_tokens_seen": 30367672,
      "step": 39815
    },
    {
      "epoch": 82.78586278586279,
      "grad_norm": 0.0009957605507224798,
      "learning_rate": 1.5156218654843733e-05,
      "loss": 0.2667,
      "num_input_tokens_seen": 30371480,
      "step": 39820
    },
    {
      "epoch": 82.79625779625779,
      "grad_norm": 0.00037115352461114526,
      "learning_rate": 1.4330437405196683e-05,
      "loss": 0.2775,
      "num_input_tokens_seen": 30375256,
      "step": 39825
    },
    {
      "epoch": 82.8066528066528,
      "grad_norm": 0.00042891796329058707,
      "learning_rate": 1.352778583062486e-05,
      "loss": 0.2972,
      "num_input_tokens_seen": 30379160,
      "step": 39830
    },
    {
      "epoch": 82.81704781704782,
      "grad_norm": 0.00018706402624957263,
      "learning_rate": 1.2748264054929237e-05,
      "loss": 0.2532,
      "num_input_tokens_seen": 30383064,
      "step": 39835
    },
    {
      "epoch": 82.82744282744282,
      "grad_norm": 0.0004231737111695111,
      "learning_rate": 1.1991872198297004e-05,
      "loss": 0.2205,
      "num_input_tokens_seen": 30386872,
      "step": 39840
    },
    {
      "epoch": 82.83783783783784,
      "grad_norm": 0.00023152027279138565,
      "learning_rate": 1.1258610377384847e-05,
      "loss": 0.2508,
      "num_input_tokens_seen": 30390808,
      "step": 39845
    },
    {
      "epoch": 82.84823284823285,
      "grad_norm": 0.00031264254357665777,
      "learning_rate": 1.0548478705268982e-05,
      "loss": 0.2721,
      "num_input_tokens_seen": 30394488,
      "step": 39850
    },
    {
      "epoch": 82.85862785862786,
      "grad_norm": 0.0008656340651214123,
      "learning_rate": 9.86147729147846e-06,
      "loss": 0.2369,
      "num_input_tokens_seen": 30398168,
      "step": 39855
    },
    {
      "epoch": 82.86902286902287,
      "grad_norm": 0.0005971190403215587,
      "learning_rate": 9.197606241928557e-06,
      "loss": 0.2426,
      "num_input_tokens_seen": 30402072,
      "step": 39860
    },
    {
      "epoch": 82.87941787941789,
      "grad_norm": 0.0005209440132603049,
      "learning_rate": 8.556865659004042e-06,
      "loss": 0.2922,
      "num_input_tokens_seen": 30405944,
      "step": 39865
    },
    {
      "epoch": 82.88981288981289,
      "grad_norm": 0.001681764144450426,
      "learning_rate": 7.939255641525867e-06,
      "loss": 0.2534,
      "num_input_tokens_seen": 30409592,
      "step": 39870
    },
    {
      "epoch": 82.9002079002079,
      "grad_norm": 0.00010485886741662398,
      "learning_rate": 7.344776284751164e-06,
      "loss": 0.2646,
      "num_input_tokens_seen": 30413272,
      "step": 39875
    },
    {
      "epoch": 82.9106029106029,
      "grad_norm": 0.00022654587519355118,
      "learning_rate": 6.773427680323296e-06,
      "loss": 0.2522,
      "num_input_tokens_seen": 30417016,
      "step": 39880
    },
    {
      "epoch": 82.92099792099792,
      "grad_norm": 0.0002863763365894556,
      "learning_rate": 6.225209916355112e-06,
      "loss": 0.2504,
      "num_input_tokens_seen": 30420824,
      "step": 39885
    },
    {
      "epoch": 82.93139293139293,
      "grad_norm": 0.0003275486524216831,
      "learning_rate": 5.7001230774123e-06,
      "loss": 0.274,
      "num_input_tokens_seen": 30424664,
      "step": 39890
    },
    {
      "epoch": 82.94178794178794,
      "grad_norm": 0.00010556641063885763,
      "learning_rate": 5.198167244446772e-06,
      "loss": 0.2699,
      "num_input_tokens_seen": 30428632,
      "step": 39895
    },
    {
      "epoch": 82.95218295218295,
      "grad_norm": 0.0003547542146407068,
      "learning_rate": 4.71934249487993e-06,
      "loss": 0.2715,
      "num_input_tokens_seen": 30432344,
      "step": 39900
    },
    {
      "epoch": 82.96257796257797,
      "grad_norm": 0.000539829139597714,
      "learning_rate": 4.2636489025527075e-06,
      "loss": 0.2503,
      "num_input_tokens_seen": 30435928,
      "step": 39905
    },
    {
      "epoch": 82.97297297297297,
      "grad_norm": 0.00030733918538317084,
      "learning_rate": 3.831086537742223e-06,
      "loss": 0.27,
      "num_input_tokens_seen": 30439832,
      "step": 39910
    },
    {
      "epoch": 82.98336798336798,
      "grad_norm": 5.470420001074672e-05,
      "learning_rate": 3.4216554671451236e-06,
      "loss": 0.2518,
      "num_input_tokens_seen": 30443736,
      "step": 39915
    },
    {
      "epoch": 82.993762993763,
      "grad_norm": 0.00018296035705134273,
      "learning_rate": 3.035355753894242e-06,
      "loss": 0.2552,
      "num_input_tokens_seen": 30447576,
      "step": 39920
    },
    {
      "epoch": 83.004158004158,
      "grad_norm": 0.00014197989366948605,
      "learning_rate": 2.6721874575752477e-06,
      "loss": 0.2584,
      "num_input_tokens_seen": 30451280,
      "step": 39925
    },
    {
      "epoch": 83.01455301455302,
      "grad_norm": 0.00011711655679391697,
      "learning_rate": 2.3321506341933418e-06,
      "loss": 0.2492,
      "num_input_tokens_seen": 30455088,
      "step": 39930
    },
    {
      "epoch": 83.02494802494803,
      "grad_norm": 0.0005199620500206947,
      "learning_rate": 2.0152453361732546e-06,
      "loss": 0.2897,
      "num_input_tokens_seen": 30458960,
      "step": 39935
    },
    {
      "epoch": 83.03534303534303,
      "grad_norm": 0.001643101335503161,
      "learning_rate": 1.7214716123925554e-06,
      "loss": 0.2632,
      "num_input_tokens_seen": 30462640,
      "step": 39940
    },
    {
      "epoch": 83.04573804573805,
      "grad_norm": 0.0007351182284764946,
      "learning_rate": 1.4508295081649968e-06,
      "loss": 0.2362,
      "num_input_tokens_seen": 30466512,
      "step": 39945
    },
    {
      "epoch": 83.05613305613305,
      "grad_norm": 0.00045212890836410224,
      "learning_rate": 1.2033190652238623e-06,
      "loss": 0.2562,
      "num_input_tokens_seen": 30470224,
      "step": 39950
    },
    {
      "epoch": 83.06652806652806,
      "grad_norm": 0.0005001293611712754,
      "learning_rate": 9.78940321721966e-07,
      "loss": 0.2591,
      "num_input_tokens_seen": 30473872,
      "step": 39955
    },
    {
      "epoch": 83.07692307692308,
      "grad_norm": 0.0013200159883126616,
      "learning_rate": 7.776933122816132e-07,
      "loss": 0.2608,
      "num_input_tokens_seen": 30477680,
      "step": 39960
    },
    {
      "epoch": 83.08731808731808,
      "grad_norm": 0.0002357655466767028,
      "learning_rate": 5.99578067927986e-07,
      "loss": 0.2651,
      "num_input_tokens_seen": 30481456,
      "step": 39965
    },
    {
      "epoch": 83.0977130977131,
      "grad_norm": 0.0002748349797911942,
      "learning_rate": 4.445946161224512e-07,
      "loss": 0.2543,
      "num_input_tokens_seen": 30485360,
      "step": 39970
    },
    {
      "epoch": 83.10810810810811,
      "grad_norm": 0.00012069835065631196,
      "learning_rate": 3.127429807792126e-07,
      "loss": 0.2502,
      "num_input_tokens_seen": 30489200,
      "step": 39975
    },
    {
      "epoch": 83.11850311850311,
      "grad_norm": 0.0007290628273040056,
      "learning_rate": 2.040231822320049e-07,
      "loss": 0.2554,
      "num_input_tokens_seen": 30493136,
      "step": 39980
    },
    {
      "epoch": 83.12889812889813,
      "grad_norm": 0.0005289536202326417,
      "learning_rate": 1.1843523723409354e-07,
      "loss": 0.2593,
      "num_input_tokens_seen": 30496880,
      "step": 39985
    },
    {
      "epoch": 83.13929313929314,
      "grad_norm": 0.000991914188489318,
      "learning_rate": 5.597915897492811e-08,
      "loss": 0.2623,
      "num_input_tokens_seen": 30500656,
      "step": 39990
    },
    {
      "epoch": 83.14968814968815,
      "grad_norm": 0.00026230866205878556,
      "learning_rate": 1.6654957113448885e-08,
      "loss": 0.2839,
      "num_input_tokens_seen": 30504496,
      "step": 39995
    },
    {
      "epoch": 83.16008316008316,
      "grad_norm": 0.00032065261621028185,
      "learning_rate": 4.626377114735902e-10,
      "loss": 0.2824,
      "num_input_tokens_seen": 30508240,
      "step": 40000
    },
    {
      "epoch": 83.16008316008316,
      "eval_loss": 0.2492801398038864,
      "eval_runtime": 13.378,
      "eval_samples_per_second": 63.986,
      "eval_steps_per_second": 15.996,
      "num_input_tokens_seen": 30508240,
      "step": 40000
    },
    {
      "epoch": 83.16008316008316,
      "num_input_tokens_seen": 30508240,
      "step": 40000,
      "total_flos": 1.3737722584517837e+18,
      "train_loss": 0.2765025964975357,
      "train_runtime": 24417.94,
      "train_samples_per_second": 26.21,
      "train_steps_per_second": 1.638
    }
  ],
  "logging_steps": 5,
  "max_steps": 40000,
  "num_input_tokens_seen": 30508240,
  "num_train_epochs": 84,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.3737722584517837e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}