{
  "best_global_step": 1967,
  "best_metric": 0.11949615180492401,
  "best_model_checkpoint": "saves/prompt-tuning/llama-3-8b-instruct/train_rte_1754502820/checkpoint-1967",
  "epoch": 10.0,
  "eval_steps": 281,
  "global_step": 5610,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008912655971479501,
      "grad_norm": 7.25,
      "learning_rate": 3.5650623885918005e-07,
      "loss": 0.3871,
      "num_input_tokens_seen": 3168,
      "step": 5
    },
    {
      "epoch": 0.017825311942959002,
      "grad_norm": 5.59375,
      "learning_rate": 8.021390374331552e-07,
      "loss": 0.3469,
      "num_input_tokens_seen": 6272,
      "step": 10
    },
    {
      "epoch": 0.026737967914438502,
      "grad_norm": 1.9921875,
      "learning_rate": 1.2477718360071302e-06,
      "loss": 0.1418,
      "num_input_tokens_seen": 10144,
      "step": 15
    },
    {
      "epoch": 0.035650623885918005,
      "grad_norm": 7.21875,
      "learning_rate": 1.6934046345811053e-06,
      "loss": 0.345,
      "num_input_tokens_seen": 13536,
      "step": 20
    },
    {
      "epoch": 0.044563279857397504,
      "grad_norm": 0.10546875,
      "learning_rate": 2.1390374331550802e-06,
      "loss": 0.3125,
      "num_input_tokens_seen": 16128,
      "step": 25
    },
    {
      "epoch": 0.053475935828877004,
      "grad_norm": 1.9296875,
      "learning_rate": 2.5846702317290554e-06,
      "loss": 0.0692,
      "num_input_tokens_seen": 18784,
      "step": 30
    },
    {
      "epoch": 0.062388591800356503,
      "grad_norm": 4.3125,
      "learning_rate": 3.0303030303030305e-06,
      "loss": 0.2449,
      "num_input_tokens_seen": 22336,
      "step": 35
    },
    {
      "epoch": 0.07130124777183601,
      "grad_norm": 0.1533203125,
      "learning_rate": 3.4759358288770056e-06,
      "loss": 0.3998,
      "num_input_tokens_seen": 25408,
      "step": 40
    },
    {
      "epoch": 0.08021390374331551,
      "grad_norm": 1.78125,
      "learning_rate": 3.92156862745098e-06,
      "loss": 0.2471,
      "num_input_tokens_seen": 27968,
      "step": 45
    },
    {
      "epoch": 0.08912655971479501,
      "grad_norm": 9.25,
      "learning_rate": 4.3672014260249555e-06,
      "loss": 0.4143,
      "num_input_tokens_seen": 30752,
      "step": 50
    },
    {
      "epoch": 0.09803921568627451,
      "grad_norm": 4.84375,
      "learning_rate": 4.812834224598931e-06,
      "loss": 0.2401,
      "num_input_tokens_seen": 33376,
      "step": 55
    },
    {
      "epoch": 0.10695187165775401,
      "grad_norm": 8.125,
      "learning_rate": 5.258467023172906e-06,
      "loss": 0.1603,
      "num_input_tokens_seen": 37280,
      "step": 60
    },
    {
      "epoch": 0.11586452762923351,
      "grad_norm": 4.96875,
      "learning_rate": 5.704099821746881e-06,
      "loss": 0.2917,
      "num_input_tokens_seen": 40640,
      "step": 65
    },
    {
      "epoch": 0.12477718360071301,
      "grad_norm": 7.4375,
      "learning_rate": 6.149732620320856e-06,
      "loss": 0.4022,
      "num_input_tokens_seen": 44128,
      "step": 70
    },
    {
      "epoch": 0.13368983957219252,
      "grad_norm": 1.25,
      "learning_rate": 6.59536541889483e-06,
      "loss": 0.2311,
      "num_input_tokens_seen": 47648,
      "step": 75
    },
    {
      "epoch": 0.14260249554367202,
      "grad_norm": 11.8125,
      "learning_rate": 7.040998217468805e-06,
      "loss": 0.134,
      "num_input_tokens_seen": 50816,
      "step": 80
    },
    {
      "epoch": 0.15151515151515152,
      "grad_norm": 12.1875,
      "learning_rate": 7.4866310160427806e-06,
      "loss": 0.3235,
      "num_input_tokens_seen": 53728,
      "step": 85
    },
    {
      "epoch": 0.16042780748663102,
      "grad_norm": 1.203125,
      "learning_rate": 7.932263814616755e-06,
      "loss": 0.1043,
      "num_input_tokens_seen": 57056,
      "step": 90
    },
    {
      "epoch": 0.16934046345811052,
      "grad_norm": 1.4453125,
      "learning_rate": 8.377896613190733e-06,
      "loss": 0.128,
      "num_input_tokens_seen": 59808,
      "step": 95
    },
    {
      "epoch": 0.17825311942959002,
      "grad_norm": 0.625,
      "learning_rate": 8.823529411764707e-06,
      "loss": 0.0951,
      "num_input_tokens_seen": 62848,
      "step": 100
    },
    {
      "epoch": 0.18716577540106952,
      "grad_norm": 0.11865234375,
      "learning_rate": 9.269162210338681e-06,
      "loss": 0.2309,
      "num_input_tokens_seen": 65856,
      "step": 105
    },
    {
      "epoch": 0.19607843137254902,
      "grad_norm": 0.92578125,
      "learning_rate": 9.714795008912657e-06,
      "loss": 0.1133,
      "num_input_tokens_seen": 68672,
      "step": 110
    },
    {
      "epoch": 0.20499108734402852,
      "grad_norm": 2.578125,
      "learning_rate": 1.0160427807486631e-05,
      "loss": 0.0673,
      "num_input_tokens_seen": 71840,
      "step": 115
    },
    {
      "epoch": 0.21390374331550802,
      "grad_norm": 3.5,
      "learning_rate": 1.0606060606060607e-05,
      "loss": 0.2123,
      "num_input_tokens_seen": 74624,
      "step": 120
    },
    {
      "epoch": 0.22281639928698752,
      "grad_norm": 1.21875,
      "learning_rate": 1.1051693404634582e-05,
      "loss": 0.2054,
      "num_input_tokens_seen": 78080,
      "step": 125
    },
    {
      "epoch": 0.23172905525846701,
      "grad_norm": 2.21875,
      "learning_rate": 1.1497326203208558e-05,
      "loss": 0.2639,
      "num_input_tokens_seen": 81408,
      "step": 130
    },
    {
      "epoch": 0.24064171122994651,
      "grad_norm": 1.3671875,
      "learning_rate": 1.1942959001782532e-05,
      "loss": 0.0858,
      "num_input_tokens_seen": 84192,
      "step": 135
    },
    {
      "epoch": 0.24955436720142601,
      "grad_norm": 2.5,
      "learning_rate": 1.2388591800356506e-05,
      "loss": 0.2855,
      "num_input_tokens_seen": 87264,
      "step": 140
    },
    {
      "epoch": 0.25846702317290554,
      "grad_norm": 0.0150146484375,
      "learning_rate": 1.2834224598930484e-05,
      "loss": 0.034,
      "num_input_tokens_seen": 90336,
      "step": 145
    },
    {
      "epoch": 0.26737967914438504,
      "grad_norm": 9.8125,
      "learning_rate": 1.3279857397504458e-05,
      "loss": 0.2305,
      "num_input_tokens_seen": 93760,
      "step": 150
    },
    {
      "epoch": 0.27629233511586454,
      "grad_norm": 3.875,
      "learning_rate": 1.3725490196078432e-05,
      "loss": 0.1568,
      "num_input_tokens_seen": 97120,
      "step": 155
    },
    {
      "epoch": 0.28520499108734404,
      "grad_norm": 7.34375,
      "learning_rate": 1.4171122994652408e-05,
      "loss": 0.5335,
      "num_input_tokens_seen": 100160,
      "step": 160
    },
    {
      "epoch": 0.29411764705882354,
      "grad_norm": 2.984375,
      "learning_rate": 1.4616755793226383e-05,
      "loss": 0.2975,
      "num_input_tokens_seen": 103136,
      "step": 165
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 0.2314453125,
      "learning_rate": 1.5062388591800359e-05,
      "loss": 0.1237,
      "num_input_tokens_seen": 105696,
      "step": 170
    },
    {
      "epoch": 0.31194295900178254,
      "grad_norm": 2.671875,
      "learning_rate": 1.5508021390374333e-05,
      "loss": 0.3044,
      "num_input_tokens_seen": 108800,
      "step": 175
    },
    {
      "epoch": 0.32085561497326204,
      "grad_norm": 1.578125,
      "learning_rate": 1.5953654188948307e-05,
      "loss": 0.1746,
      "num_input_tokens_seen": 111808,
      "step": 180
    },
    {
      "epoch": 0.32976827094474154,
      "grad_norm": 6.625,
      "learning_rate": 1.639928698752228e-05,
      "loss": 0.2148,
      "num_input_tokens_seen": 114944,
      "step": 185
    },
    {
      "epoch": 0.33868092691622104,
      "grad_norm": 9.3125,
      "learning_rate": 1.684491978609626e-05,
      "loss": 0.1727,
      "num_input_tokens_seen": 118112,
      "step": 190
    },
    {
      "epoch": 0.34759358288770054,
      "grad_norm": 0.23046875,
      "learning_rate": 1.7290552584670233e-05,
      "loss": 0.1062,
      "num_input_tokens_seen": 120896,
      "step": 195
    },
    {
      "epoch": 0.35650623885918004,
      "grad_norm": 2.953125,
      "learning_rate": 1.7736185383244208e-05,
      "loss": 0.1255,
      "num_input_tokens_seen": 123904,
      "step": 200
    },
    {
      "epoch": 0.36541889483065954,
      "grad_norm": 0.470703125,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 0.1552,
      "num_input_tokens_seen": 127008,
      "step": 205
    },
    {
      "epoch": 0.37433155080213903,
      "grad_norm": 3.96875,
      "learning_rate": 1.862745098039216e-05,
      "loss": 0.2139,
      "num_input_tokens_seen": 129984,
      "step": 210
    },
    {
      "epoch": 0.38324420677361853,
      "grad_norm": 1.890625,
      "learning_rate": 1.9073083778966134e-05,
      "loss": 0.1131,
      "num_input_tokens_seen": 133152,
      "step": 215
    },
    {
      "epoch": 0.39215686274509803,
      "grad_norm": 0.0888671875,
      "learning_rate": 1.951871657754011e-05,
      "loss": 0.1154,
      "num_input_tokens_seen": 136096,
      "step": 220
    },
    {
      "epoch": 0.40106951871657753,
      "grad_norm": 0.423828125,
      "learning_rate": 1.9964349376114083e-05,
      "loss": 0.2696,
      "num_input_tokens_seen": 139136,
      "step": 225
    },
    {
      "epoch": 0.40998217468805703,
      "grad_norm": 8.375,
      "learning_rate": 2.0409982174688057e-05,
      "loss": 0.2524,
      "num_input_tokens_seen": 142080,
      "step": 230
    },
    {
      "epoch": 0.41889483065953653,
      "grad_norm": 6.71875,
      "learning_rate": 2.0855614973262035e-05,
      "loss": 0.221,
      "num_input_tokens_seen": 145824,
      "step": 235
    },
    {
      "epoch": 0.42780748663101603,
      "grad_norm": 0.333984375,
      "learning_rate": 2.130124777183601e-05,
      "loss": 0.118,
      "num_input_tokens_seen": 149280,
      "step": 240
    },
    {
      "epoch": 0.43672014260249553,
      "grad_norm": 8.5,
      "learning_rate": 2.1746880570409983e-05,
      "loss": 0.2365,
      "num_input_tokens_seen": 152544,
      "step": 245
    },
    {
      "epoch": 0.44563279857397503,
      "grad_norm": 6.03125,
      "learning_rate": 2.2192513368983957e-05,
      "loss": 0.3776,
      "num_input_tokens_seen": 156416,
      "step": 250
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 0.984375,
      "learning_rate": 2.2638146167557932e-05,
      "loss": 0.1431,
      "num_input_tokens_seen": 159712,
      "step": 255
    },
    {
      "epoch": 0.46345811051693403,
      "grad_norm": 2.546875,
      "learning_rate": 2.308377896613191e-05,
      "loss": 0.0348,
      "num_input_tokens_seen": 162400,
      "step": 260
    },
    {
      "epoch": 0.47237076648841353,
      "grad_norm": 4.0625,
      "learning_rate": 2.3529411764705884e-05,
      "loss": 0.1799,
      "num_input_tokens_seen": 166048,
      "step": 265
    },
    {
      "epoch": 0.48128342245989303,
      "grad_norm": 8.5625,
      "learning_rate": 2.3975044563279858e-05,
      "loss": 0.1163,
      "num_input_tokens_seen": 168576,
      "step": 270
    },
    {
      "epoch": 0.49019607843137253,
      "grad_norm": 3.953125,
      "learning_rate": 2.4420677361853832e-05,
      "loss": 0.155,
      "num_input_tokens_seen": 172320,
      "step": 275
    },
    {
      "epoch": 0.49910873440285203,
      "grad_norm": 3.25,
      "learning_rate": 2.4866310160427807e-05,
      "loss": 0.2025,
      "num_input_tokens_seen": 175424,
      "step": 280
    },
    {
      "epoch": 0.5008912655971479,
      "eval_loss": 0.1983698159456253,
      "eval_runtime": 6.2808,
      "eval_samples_per_second": 39.644,
      "eval_steps_per_second": 10.03,
      "num_input_tokens_seen": 176032,
      "step": 281
    },
    {
      "epoch": 0.5080213903743316,
      "grad_norm": 3.078125,
      "learning_rate": 2.5311942959001784e-05,
      "loss": 0.1648,
      "num_input_tokens_seen": 178016,
      "step": 285
    },
    {
      "epoch": 0.5169340463458111,
      "grad_norm": 0.50390625,
      "learning_rate": 2.575757575757576e-05,
      "loss": 0.0759,
      "num_input_tokens_seen": 181888,
      "step": 290
    },
    {
      "epoch": 0.5258467023172906,
      "grad_norm": 2.9375,
      "learning_rate": 2.6203208556149733e-05,
      "loss": 0.1507,
      "num_input_tokens_seen": 184960,
      "step": 295
    },
    {
      "epoch": 0.5347593582887701,
      "grad_norm": 2.671875,
      "learning_rate": 2.6648841354723707e-05,
      "loss": 0.0901,
      "num_input_tokens_seen": 187488,
      "step": 300
    },
    {
      "epoch": 0.5436720142602496,
      "grad_norm": 0.375,
      "learning_rate": 2.7094474153297685e-05,
      "loss": 0.0708,
      "num_input_tokens_seen": 191232,
      "step": 305
    },
    {
      "epoch": 0.5525846702317291,
      "grad_norm": 6.03125,
      "learning_rate": 2.754010695187166e-05,
      "loss": 0.1351,
      "num_input_tokens_seen": 194272,
      "step": 310
    },
    {
      "epoch": 0.5614973262032086,
      "grad_norm": 0.48046875,
      "learning_rate": 2.7985739750445633e-05,
      "loss": 0.0961,
      "num_input_tokens_seen": 197184,
      "step": 315
    },
    {
      "epoch": 0.5704099821746881,
      "grad_norm": 0.357421875,
      "learning_rate": 2.8431372549019608e-05,
      "loss": 0.1903,
      "num_input_tokens_seen": 199840,
      "step": 320
    },
    {
      "epoch": 0.5793226381461676,
      "grad_norm": 0.53125,
      "learning_rate": 2.8877005347593582e-05,
      "loss": 0.0495,
      "num_input_tokens_seen": 203008,
      "step": 325
    },
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 6.4375,
      "learning_rate": 2.932263814616756e-05,
      "loss": 0.1168,
      "num_input_tokens_seen": 206400,
      "step": 330
    },
    {
      "epoch": 0.5971479500891266,
      "grad_norm": 1.1796875,
      "learning_rate": 2.9768270944741534e-05,
      "loss": 0.1421,
      "num_input_tokens_seen": 209440,
      "step": 335
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 2.59375,
      "learning_rate": 3.0213903743315508e-05,
      "loss": 0.2025,
      "num_input_tokens_seen": 212736,
      "step": 340
    },
    {
      "epoch": 0.6149732620320856,
      "grad_norm": 1.8359375,
      "learning_rate": 3.065953654188948e-05,
      "loss": 0.2427,
      "num_input_tokens_seen": 216096,
      "step": 345
    },
    {
      "epoch": 0.6238859180035651,
      "grad_norm": 2.65625,
      "learning_rate": 3.110516934046346e-05,
      "loss": 0.153,
      "num_input_tokens_seen": 219200,
      "step": 350
    },
    {
      "epoch": 0.6327985739750446,
      "grad_norm": 0.58203125,
      "learning_rate": 3.155080213903743e-05,
      "loss": 0.1216,
      "num_input_tokens_seen": 221952,
      "step": 355
    },
    {
      "epoch": 0.6417112299465241,
      "grad_norm": 1.6328125,
      "learning_rate": 3.199643493761141e-05,
      "loss": 0.1366,
      "num_input_tokens_seen": 225376,
      "step": 360
    },
    {
      "epoch": 0.6506238859180036,
      "grad_norm": 1.0625,
      "learning_rate": 3.2442067736185386e-05,
      "loss": 0.0615,
      "num_input_tokens_seen": 228736,
      "step": 365
    },
    {
      "epoch": 0.6595365418894831,
      "grad_norm": 0.09521484375,
      "learning_rate": 3.288770053475936e-05,
      "loss": 0.0691,
      "num_input_tokens_seen": 231648,
      "step": 370
    },
    {
      "epoch": 0.6684491978609626,
      "grad_norm": 4.59375,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.0868,
      "num_input_tokens_seen": 234976,
      "step": 375
    },
    {
      "epoch": 0.6773618538324421,
      "grad_norm": 2.078125,
      "learning_rate": 3.3778966131907306e-05,
      "loss": 0.1143,
      "num_input_tokens_seen": 238368,
      "step": 380
    },
    {
      "epoch": 0.6862745098039216,
      "grad_norm": 6.28125,
      "learning_rate": 3.4224598930481284e-05,
      "loss": 0.0611,
      "num_input_tokens_seen": 241440,
      "step": 385
    },
    {
      "epoch": 0.6951871657754011,
      "grad_norm": 9.4375,
      "learning_rate": 3.467023172905526e-05,
      "loss": 0.2413,
      "num_input_tokens_seen": 244448,
      "step": 390
    },
    {
      "epoch": 0.7040998217468806,
      "grad_norm": 1.0859375,
      "learning_rate": 3.511586452762923e-05,
      "loss": 0.0962,
      "num_input_tokens_seen": 246880,
      "step": 395
    },
    {
      "epoch": 0.7130124777183601,
      "grad_norm": 4.96875,
      "learning_rate": 3.556149732620321e-05,
      "loss": 0.289,
      "num_input_tokens_seen": 250240,
      "step": 400
    },
    {
      "epoch": 0.7219251336898396,
      "grad_norm": 1.296875,
      "learning_rate": 3.600713012477718e-05,
      "loss": 0.1467,
      "num_input_tokens_seen": 253184,
      "step": 405
    },
    {
      "epoch": 0.7308377896613191,
      "grad_norm": 3.34375,
      "learning_rate": 3.645276292335116e-05,
      "loss": 0.1721,
      "num_input_tokens_seen": 255968,
      "step": 410
    },
    {
      "epoch": 0.7397504456327986,
      "grad_norm": 3.359375,
      "learning_rate": 3.6898395721925136e-05,
      "loss": 0.1705,
      "num_input_tokens_seen": 258688,
      "step": 415
    },
    {
      "epoch": 0.7486631016042781,
      "grad_norm": 5.21875,
      "learning_rate": 3.734402852049911e-05,
      "loss": 0.1499,
      "num_input_tokens_seen": 262240,
      "step": 420
    },
    {
      "epoch": 0.7575757575757576,
      "grad_norm": 1.921875,
      "learning_rate": 3.7789661319073085e-05,
      "loss": 0.102,
      "num_input_tokens_seen": 265952,
      "step": 425
    },
    {
      "epoch": 0.7664884135472371,
      "grad_norm": 2.109375,
      "learning_rate": 3.8235294117647055e-05,
      "loss": 0.0561,
      "num_input_tokens_seen": 269312,
      "step": 430
    },
    {
      "epoch": 0.7754010695187166,
      "grad_norm": 3.125,
      "learning_rate": 3.868092691622103e-05,
      "loss": 0.1028,
      "num_input_tokens_seen": 272128,
      "step": 435
    },
    {
      "epoch": 0.7843137254901961,
      "grad_norm": 3.203125,
      "learning_rate": 3.912655971479501e-05,
      "loss": 0.1513,
      "num_input_tokens_seen": 275552,
      "step": 440
    },
    {
      "epoch": 0.7932263814616756,
      "grad_norm": 5.3125,
      "learning_rate": 3.957219251336899e-05,
      "loss": 0.1387,
      "num_input_tokens_seen": 278720,
      "step": 445
    },
    {
      "epoch": 0.8021390374331551,
      "grad_norm": 1.5546875,
      "learning_rate": 4.0017825311942966e-05,
      "loss": 0.1824,
      "num_input_tokens_seen": 281536,
      "step": 450
    },
    {
      "epoch": 0.8110516934046346,
      "grad_norm": 2.15625,
      "learning_rate": 4.046345811051694e-05,
      "loss": 0.2449,
      "num_input_tokens_seen": 284672,
      "step": 455
    },
    {
      "epoch": 0.8199643493761141,
      "grad_norm": 0.81640625,
      "learning_rate": 4.0909090909090915e-05,
      "loss": 0.1029,
      "num_input_tokens_seen": 288416,
      "step": 460
    },
    {
      "epoch": 0.8288770053475936,
      "grad_norm": 1.2890625,
      "learning_rate": 4.1354723707664886e-05,
      "loss": 0.0824,
      "num_input_tokens_seen": 291232,
      "step": 465
    },
    {
      "epoch": 0.8377896613190731,
      "grad_norm": 3.046875,
      "learning_rate": 4.180035650623886e-05,
      "loss": 0.1585,
      "num_input_tokens_seen": 294784,
      "step": 470
    },
    {
      "epoch": 0.8467023172905526,
      "grad_norm": 3.140625,
      "learning_rate": 4.224598930481284e-05,
      "loss": 0.1035,
      "num_input_tokens_seen": 297632,
      "step": 475
    },
    {
      "epoch": 0.8556149732620321,
      "grad_norm": 5.375,
      "learning_rate": 4.269162210338681e-05,
      "loss": 0.1197,
      "num_input_tokens_seen": 300416,
      "step": 480
    },
    {
      "epoch": 0.8645276292335116,
      "grad_norm": 0.55078125,
      "learning_rate": 4.313725490196079e-05,
      "loss": 0.1252,
      "num_input_tokens_seen": 303232,
      "step": 485
    },
    {
      "epoch": 0.8734402852049911,
      "grad_norm": 4.4375,
      "learning_rate": 4.358288770053476e-05,
      "loss": 0.1306,
      "num_input_tokens_seen": 306144,
      "step": 490
    },
    {
      "epoch": 0.8823529411764706,
      "grad_norm": 2.484375,
      "learning_rate": 4.402852049910874e-05,
      "loss": 0.0748,
      "num_input_tokens_seen": 308576,
      "step": 495
    },
    {
      "epoch": 0.8912655971479501,
      "grad_norm": 0.427734375,
      "learning_rate": 4.4474153297682716e-05,
      "loss": 0.1202,
      "num_input_tokens_seen": 312000,
      "step": 500
    },
    {
      "epoch": 0.9001782531194296,
      "grad_norm": 13.125,
      "learning_rate": 4.491978609625669e-05,
      "loss": 0.1321,
      "num_input_tokens_seen": 314848,
      "step": 505
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 3.953125,
      "learning_rate": 4.5365418894830664e-05,
      "loss": 0.0629,
      "num_input_tokens_seen": 318112,
      "step": 510
    },
    {
      "epoch": 0.9180035650623886,
      "grad_norm": 0.25390625,
      "learning_rate": 4.5811051693404635e-05,
      "loss": 0.0765,
      "num_input_tokens_seen": 321152,
      "step": 515
    },
    {
      "epoch": 0.9269162210338681,
      "grad_norm": 4.5,
      "learning_rate": 4.625668449197861e-05,
      "loss": 0.0566,
      "num_input_tokens_seen": 323552,
      "step": 520
    },
    {
      "epoch": 0.9358288770053476,
      "grad_norm": 3.296875,
      "learning_rate": 4.670231729055259e-05,
      "loss": 0.058,
      "num_input_tokens_seen": 326112,
      "step": 525
    },
    {
      "epoch": 0.9447415329768271,
      "grad_norm": 11.4375,
      "learning_rate": 4.714795008912656e-05,
      "loss": 0.1965,
      "num_input_tokens_seen": 328800,
      "step": 530
    },
    {
      "epoch": 0.9536541889483066,
      "grad_norm": 5.625,
      "learning_rate": 4.759358288770054e-05,
      "loss": 0.1333,
      "num_input_tokens_seen": 332512,
      "step": 535
    },
    {
      "epoch": 0.9625668449197861,
      "grad_norm": 4.59375,
      "learning_rate": 4.803921568627452e-05,
      "loss": 0.1171,
      "num_input_tokens_seen": 335360,
      "step": 540
    },
    {
      "epoch": 0.9714795008912656,
      "grad_norm": 5.5,
      "learning_rate": 4.848484848484849e-05,
      "loss": 0.0733,
      "num_input_tokens_seen": 339488,
      "step": 545
    },
    {
      "epoch": 0.9803921568627451,
      "grad_norm": 0.72265625,
      "learning_rate": 4.8930481283422465e-05,
      "loss": 0.0912,
      "num_input_tokens_seen": 342176,
      "step": 550
    },
    {
      "epoch": 0.9893048128342246,
      "grad_norm": 3.640625,
      "learning_rate": 4.9376114081996436e-05,
      "loss": 0.109,
      "num_input_tokens_seen": 345568,
      "step": 555
    },
    {
      "epoch": 0.9982174688057041,
      "grad_norm": 15.9375,
      "learning_rate": 4.9821746880570414e-05,
      "loss": 0.1763,
      "num_input_tokens_seen": 348000,
      "step": 560
    },
    {
      "epoch": 1.0017825311942958,
      "eval_loss": 0.13491526246070862,
      "eval_runtime": 6.252,
      "eval_samples_per_second": 39.827,
      "eval_steps_per_second": 10.077,
      "num_input_tokens_seen": 349200,
      "step": 562
    },
    {
      "epoch": 1.0071301247771836,
      "grad_norm": 4.53125,
      "learning_rate": 4.99999564446608e-05,
      "loss": 0.182,
      "num_input_tokens_seen": 350960,
      "step": 565
    },
    {
      "epoch": 1.0160427807486632,
      "grad_norm": 7.28125,
      "learning_rate": 4.9999690273693036e-05,
      "loss": 0.0825,
      "num_input_tokens_seen": 354288,
      "step": 570
    },
    {
      "epoch": 1.0249554367201426,
      "grad_norm": 1.609375,
      "learning_rate": 4.999918213174131e-05,
      "loss": 0.0908,
      "num_input_tokens_seen": 357648,
      "step": 575
    },
    {
      "epoch": 1.0338680926916222,
      "grad_norm": 1.9296875,
      "learning_rate": 4.9998432023723915e-05,
      "loss": 0.1092,
      "num_input_tokens_seen": 360496,
      "step": 580
    },
    {
      "epoch": 1.0427807486631016,
      "grad_norm": 2.1875,
      "learning_rate": 4.9997439956901106e-05,
      "loss": 0.0588,
      "num_input_tokens_seen": 363376,
      "step": 585
    },
    {
      "epoch": 1.0516934046345812,
      "grad_norm": 31.625,
      "learning_rate": 4.999620594087507e-05,
      "loss": 0.1617,
      "num_input_tokens_seen": 366320,
      "step": 590
    },
    {
      "epoch": 1.0606060606060606,
      "grad_norm": 4.875,
      "learning_rate": 4.999472998758978e-05,
      "loss": 0.1548,
      "num_input_tokens_seen": 369488,
      "step": 595
    },
    {
      "epoch": 1.0695187165775402,
      "grad_norm": 3.34375,
      "learning_rate": 4.999301211133095e-05,
      "loss": 0.0569,
      "num_input_tokens_seen": 372656,
      "step": 600
    },
    {
      "epoch": 1.0784313725490196,
      "grad_norm": 1.5859375,
      "learning_rate": 4.999105232872582e-05,
      "loss": 0.1505,
      "num_input_tokens_seen": 376048,
      "step": 605
    },
    {
      "epoch": 1.0873440285204992,
      "grad_norm": 10.0,
      "learning_rate": 4.998885065874305e-05,
      "loss": 0.2991,
      "num_input_tokens_seen": 379472,
      "step": 610
    },
    {
      "epoch": 1.0962566844919786,
      "grad_norm": 6.34375,
      "learning_rate": 4.9986407122692504e-05,
      "loss": 0.2261,
      "num_input_tokens_seen": 382288,
      "step": 615
    },
    {
      "epoch": 1.1051693404634582,
      "grad_norm": 1.6171875,
      "learning_rate": 4.998372174422507e-05,
      "loss": 0.2298,
      "num_input_tokens_seen": 385392,
      "step": 620
    },
    {
      "epoch": 1.1140819964349375,
      "grad_norm": 12.125,
      "learning_rate": 4.998079454933244e-05,
      "loss": 0.0593,
      "num_input_tokens_seen": 389200,
      "step": 625
    },
    {
      "epoch": 1.1229946524064172,
      "grad_norm": 7.53125,
      "learning_rate": 4.99776255663468e-05,
      "loss": 0.1117,
      "num_input_tokens_seen": 391664,
      "step": 630
    },
    {
      "epoch": 1.1319073083778965,
      "grad_norm": 4.75,
      "learning_rate": 4.997421482594059e-05,
      "loss": 0.0618,
      "num_input_tokens_seen": 394416,
      "step": 635
    },
    {
      "epoch": 1.1408199643493762,
      "grad_norm": 2.3125,
      "learning_rate": 4.997056236112625e-05,
      "loss": 0.0573,
      "num_input_tokens_seen": 399248,
      "step": 640
    },
    {
      "epoch": 1.1497326203208555,
      "grad_norm": 0.8203125,
      "learning_rate": 4.9966668207255826e-05,
      "loss": 0.0672,
      "num_input_tokens_seen": 402032,
      "step": 645
    },
    {
      "epoch": 1.1586452762923352,
      "grad_norm": 4.9375,
      "learning_rate": 4.996253240202069e-05,
      "loss": 0.081,
      "num_input_tokens_seen": 405296,
      "step": 650
    },
    {
      "epoch": 1.1675579322638145,
      "grad_norm": 11.875,
      "learning_rate": 4.9958154985451114e-05,
      "loss": 0.0902,
      "num_input_tokens_seen": 408400,
      "step": 655
    },
    {
      "epoch": 1.1764705882352942,
      "grad_norm": 3.515625,
      "learning_rate": 4.995353599991595e-05,
      "loss": 0.062,
      "num_input_tokens_seen": 412016,
      "step": 660
    },
    {
      "epoch": 1.1853832442067735,
      "grad_norm": 24.625,
      "learning_rate": 4.994867549012215e-05,
      "loss": 0.2733,
      "num_input_tokens_seen": 415504,
      "step": 665
    },
    {
      "epoch": 1.1942959001782532,
      "grad_norm": 11.375,
      "learning_rate": 4.99435735031144e-05,
      "loss": 0.1307,
      "num_input_tokens_seen": 418448,
      "step": 670
    },
    {
      "epoch": 1.2032085561497325,
      "grad_norm": 2.6875,
      "learning_rate": 4.993823008827465e-05,
      "loss": 0.0424,
      "num_input_tokens_seen": 421168,
      "step": 675
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 5.25,
      "learning_rate": 4.9932645297321555e-05,
      "loss": 0.1269,
      "num_input_tokens_seen": 423632,
      "step": 680
    },
    {
      "epoch": 1.2210338680926915,
      "grad_norm": 4.71875,
      "learning_rate": 4.9926819184310103e-05,
      "loss": 0.0861,
      "num_input_tokens_seen": 426640,
      "step": 685
    },
    {
      "epoch": 1.2299465240641712,
      "grad_norm": 0.3046875,
      "learning_rate": 4.9920751805631e-05,
      "loss": 0.0286,
      "num_input_tokens_seen": 430032,
      "step": 690
    },
    {
      "epoch": 1.2388591800356505,
      "grad_norm": 0.314453125,
      "learning_rate": 4.991444322001014e-05,
      "loss": 0.0113,
      "num_input_tokens_seen": 433008,
      "step": 695
    },
    {
      "epoch": 1.2477718360071302,
      "grad_norm": 3.1875,
      "learning_rate": 4.99078934885081e-05,
      "loss": 0.1576,
      "num_input_tokens_seen": 436400,
      "step": 700
    },
    {
      "epoch": 1.2566844919786098,
      "grad_norm": 0.050537109375,
      "learning_rate": 4.990110267451944e-05,
      "loss": 0.1207,
      "num_input_tokens_seen": 439248,
      "step": 705
    },
    {
      "epoch": 1.2655971479500892,
      "grad_norm": 8.8125,
      "learning_rate": 4.989407084377218e-05,
      "loss": 0.0847,
      "num_input_tokens_seen": 442416,
      "step": 710
    },
    {
      "epoch": 1.2745098039215685,
      "grad_norm": 0.04345703125,
      "learning_rate": 4.988679806432712e-05,
      "loss": 0.003,
      "num_input_tokens_seen": 445616,
      "step": 715
    },
    {
      "epoch": 1.2834224598930482,
      "grad_norm": 9.25,
      "learning_rate": 4.9879284406577195e-05,
      "loss": 0.1655,
      "num_input_tokens_seen": 448528,
      "step": 720
    },
    {
      "epoch": 1.2923351158645278,
      "grad_norm": 0.0595703125,
      "learning_rate": 4.98715299432468e-05,
      "loss": 0.0605,
      "num_input_tokens_seen": 451664,
      "step": 725
    },
    {
      "epoch": 1.3012477718360071,
      "grad_norm": 6.90625,
      "learning_rate": 4.986353474939106e-05,
      "loss": 0.0255,
      "num_input_tokens_seen": 455120,
      "step": 730
    },
    {
      "epoch": 1.3101604278074865,
      "grad_norm": 3.515625,
      "learning_rate": 4.9855298902395134e-05,
      "loss": 0.0086,
      "num_input_tokens_seen": 458352,
      "step": 735
    },
    {
      "epoch": 1.3190730837789661,
      "grad_norm": 21.75,
      "learning_rate": 4.9846822481973455e-05,
      "loss": 0.1697,
      "num_input_tokens_seen": 461488,
      "step": 740
    },
    {
      "epoch": 1.3279857397504458,
      "grad_norm": 0.28515625,
      "learning_rate": 4.9838105570168946e-05,
      "loss": 0.0289,
      "num_input_tokens_seen": 464848,
      "step": 745
    },
    {
      "epoch": 1.3368983957219251,
      "grad_norm": 2.078125,
      "learning_rate": 4.982914825135224e-05,
      "loss": 0.3793,
      "num_input_tokens_seen": 468944,
      "step": 750
    },
    {
      "epoch": 1.3458110516934045,
      "grad_norm": 18.0,
      "learning_rate": 4.981995061222087e-05,
      "loss": 0.2285,
      "num_input_tokens_seen": 471312,
      "step": 755
    },
    {
      "epoch": 1.3547237076648841,
      "grad_norm": 2.84375,
      "learning_rate": 4.98105127417984e-05,
      "loss": 0.0773,
      "num_input_tokens_seen": 474128,
      "step": 760
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 4.71875,
      "learning_rate": 4.9800834731433596e-05,
      "loss": 0.1334,
      "num_input_tokens_seen": 476592,
      "step": 765
    },
    {
      "epoch": 1.3725490196078431,
      "grad_norm": 9.75,
      "learning_rate": 4.9790916674799526e-05,
      "loss": 0.0512,
      "num_input_tokens_seen": 480240,
      "step": 770
    },
    {
      "epoch": 1.3814616755793225,
      "grad_norm": 14.5625,
      "learning_rate": 4.9780758667892656e-05,
      "loss": 0.0634,
      "num_input_tokens_seen": 483472,
      "step": 775
    },
    {
      "epoch": 1.3903743315508021,
      "grad_norm": 0.06396484375,
      "learning_rate": 4.977036080903193e-05,
      "loss": 0.2462,
      "num_input_tokens_seen": 486768,
      "step": 780
    },
    {
      "epoch": 1.3992869875222818,
      "grad_norm": 13.25,
      "learning_rate": 4.975972319885779e-05,
      "loss": 0.0367,
      "num_input_tokens_seen": 489392,
      "step": 785
    },
    {
      "epoch": 1.4081996434937611,
      "grad_norm": 2.390625,
      "learning_rate": 4.974884594033123e-05,
      "loss": 0.0759,
      "num_input_tokens_seen": 492560,
      "step": 790
    },
    {
      "epoch": 1.4171122994652405,
      "grad_norm": 10.125,
      "learning_rate": 4.9737729138732805e-05,
      "loss": 0.1909,
      "num_input_tokens_seen": 495344,
      "step": 795
    },
    {
      "epoch": 1.4260249554367201,
      "grad_norm": 13.1875,
      "learning_rate": 4.972637290166158e-05,
      "loss": 0.1073,
      "num_input_tokens_seen": 498128,
      "step": 800
    },
    {
      "epoch": 1.4349376114081998,
      "grad_norm": 0.43359375,
      "learning_rate": 4.97147773390341e-05,
      "loss": 0.0423,
      "num_input_tokens_seen": 501488,
      "step": 805
    },
    {
      "epoch": 1.4438502673796791,
      "grad_norm": 11.5625,
      "learning_rate": 4.9702942563083356e-05,
      "loss": 0.1348,
      "num_input_tokens_seen": 504272,
      "step": 810
    },
    {
      "epoch": 1.4527629233511585,
      "grad_norm": 27.0,
      "learning_rate": 4.969086868835765e-05,
      "loss": 0.2365,
      "num_input_tokens_seen": 506672,
      "step": 815
    },
    {
      "epoch": 1.4616755793226381,
      "grad_norm": 4.71875,
      "learning_rate": 4.967855583171954e-05,
      "loss": 0.1498,
      "num_input_tokens_seen": 509232,
      "step": 820
    },
    {
      "epoch": 1.4705882352941178,
      "grad_norm": 4.75,
      "learning_rate": 4.9666004112344656e-05,
      "loss": 0.1283,
      "num_input_tokens_seen": 512528,
      "step": 825
    },
    {
      "epoch": 1.4795008912655971,
      "grad_norm": 2.625,
      "learning_rate": 4.965321365172057e-05,
      "loss": 0.0449,
      "num_input_tokens_seen": 514896,
      "step": 830
    },
    {
      "epoch": 1.4884135472370765,
      "grad_norm": 10.3125,
      "learning_rate": 4.9640184573645646e-05,
      "loss": 0.0548,
      "num_input_tokens_seen": 518384,
      "step": 835
    },
    {
      "epoch": 1.4973262032085561,
      "grad_norm": 20.875,
      "learning_rate": 4.962691700422778e-05,
      "loss": 0.0961,
      "num_input_tokens_seen": 522448,
      "step": 840
    },
    {
      "epoch": 1.5026737967914439,
      "eval_loss": 0.12453080713748932,
      "eval_runtime": 6.2756,
      "eval_samples_per_second": 39.678,
      "eval_steps_per_second": 10.039,
      "num_input_tokens_seen": 524208,
      "step": 843
    },
    {
      "epoch": 1.5062388591800357,
      "grad_norm": 18.5,
      "learning_rate": 4.9613411071883267e-05,
      "loss": 0.0782,
      "num_input_tokens_seen": 525264,
      "step": 845
    },
    {
      "epoch": 1.5151515151515151,
      "grad_norm": 29.125,
      "learning_rate": 4.959966690733544e-05,
      "loss": 0.1161,
      "num_input_tokens_seen": 528528,
      "step": 850
    },
    {
      "epoch": 1.5240641711229945,
      "grad_norm": 2.015625,
      "learning_rate": 4.958568464361353e-05,
      "loss": 0.0688,
      "num_input_tokens_seen": 531536,
      "step": 855
    },
    {
      "epoch": 1.5329768270944741,
      "grad_norm": 21.125,
      "learning_rate": 4.9571464416051294e-05,
      "loss": 0.0848,
      "num_input_tokens_seen": 534704,
      "step": 860
    },
    {
      "epoch": 1.5418894830659537,
      "grad_norm": 12.125,
      "learning_rate": 4.955700636228573e-05,
      "loss": 0.0262,
      "num_input_tokens_seen": 537264,
      "step": 865
    },
    {
      "epoch": 1.5508021390374331,
      "grad_norm": 2.578125,
      "learning_rate": 4.954231062225576e-05,
      "loss": 0.0095,
      "num_input_tokens_seen": 541328,
      "step": 870
    },
    {
      "epoch": 1.5597147950089125,
      "grad_norm": 7.90625,
      "learning_rate": 4.9527377338200855e-05,
      "loss": 0.0443,
      "num_input_tokens_seen": 544496,
      "step": 875
    },
    {
      "epoch": 1.5686274509803921,
      "grad_norm": 15.9375,
      "learning_rate": 4.951220665465964e-05,
      "loss": 0.0583,
      "num_input_tokens_seen": 547696,
      "step": 880
    },
    {
      "epoch": 1.5775401069518717,
      "grad_norm": 11.4375,
      "learning_rate": 4.949679871846857e-05,
      "loss": 0.3259,
      "num_input_tokens_seen": 550416,
      "step": 885
    },
    {
      "epoch": 1.5864527629233511,
      "grad_norm": 10.0625,
      "learning_rate": 4.948115367876043e-05,
      "loss": 0.155,
      "num_input_tokens_seen": 553968,
      "step": 890
    },
    {
      "epoch": 1.5953654188948305,
      "grad_norm": 2.421875,
      "learning_rate": 4.94652716869629e-05,
      "loss": 0.0245,
      "num_input_tokens_seen": 556656,
      "step": 895
    },
    {
      "epoch": 1.6042780748663101,
      "grad_norm": 0.7421875,
      "learning_rate": 4.944915289679716e-05,
      "loss": 0.1417,
      "num_input_tokens_seen": 559536,
      "step": 900
    },
    {
      "epoch": 1.6131907308377897,
      "grad_norm": 1.0234375,
      "learning_rate": 4.94327974642763e-05,
      "loss": 0.2026,
      "num_input_tokens_seen": 562704,
      "step": 905
    },
    {
      "epoch": 1.6221033868092691,
      "grad_norm": 21.75,
      "learning_rate": 4.94162055477039e-05,
      "loss": 0.0659,
      "num_input_tokens_seen": 566352,
      "step": 910
    },
    {
      "epoch": 1.6310160427807485,
      "grad_norm": 33.75,
      "learning_rate": 4.939937730767243e-05,
      "loss": 0.0742,
      "num_input_tokens_seen": 569584,
      "step": 915
    },
    {
      "epoch": 1.6399286987522281,
      "grad_norm": 24.625,
      "learning_rate": 4.9382312907061755e-05,
      "loss": 0.1525,
      "num_input_tokens_seen": 571824,
      "step": 920
    },
    {
      "epoch": 1.6488413547237077,
      "grad_norm": 18.875,
      "learning_rate": 4.9365012511037514e-05,
      "loss": 0.0786,
      "num_input_tokens_seen": 575248,
      "step": 925
    },
    {
      "epoch": 1.6577540106951871,
      "grad_norm": 17.875,
      "learning_rate": 4.934747628704952e-05,
      "loss": 0.1018,
      "num_input_tokens_seen": 578032,
      "step": 930
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 17.75,
      "learning_rate": 4.932970440483018e-05,
      "loss": 0.2083,
      "num_input_tokens_seen": 581744,
      "step": 935
    },
    {
      "epoch": 1.6755793226381461,
      "grad_norm": 5.625,
      "learning_rate": 4.931169703639282e-05,
      "loss": 0.2742,
      "num_input_tokens_seen": 584880,
      "step": 940
    },
    {
      "epoch": 1.6844919786096257,
      "grad_norm": 4.15625,
      "learning_rate": 4.929345435603003e-05,
      "loss": 0.1039,
      "num_input_tokens_seen": 587856,
      "step": 945
    },
    {
      "epoch": 1.6934046345811051,
      "grad_norm": 0.1962890625,
      "learning_rate": 4.9274976540311956e-05,
      "loss": 0.0851,
      "num_input_tokens_seen": 590928,
      "step": 950
    },
    {
      "epoch": 1.7023172905525845,
      "grad_norm": 1.2890625,
      "learning_rate": 4.9256263768084635e-05,
      "loss": 0.0727,
      "num_input_tokens_seen": 594096,
      "step": 955
    },
    {
      "epoch": 1.7112299465240641,
      "grad_norm": 0.19140625,
      "learning_rate": 4.923731622046823e-05,
      "loss": 0.0307,
      "num_input_tokens_seen": 597136,
      "step": 960
    },
    {
      "epoch": 1.7201426024955437,
      "grad_norm": 10.1875,
      "learning_rate": 4.9218134080855273e-05,
      "loss": 0.1659,
      "num_input_tokens_seen": 600912,
      "step": 965
    },
    {
      "epoch": 1.7290552584670231,
      "grad_norm": 0.060791015625,
      "learning_rate": 4.919871753490891e-05,
      "loss": 0.1413,
      "num_input_tokens_seen": 604240,
      "step": 970
    },
    {
      "epoch": 1.7379679144385025,
      "grad_norm": 0.045166015625,
      "learning_rate": 4.917906677056111e-05,
      "loss": 0.0924,
      "num_input_tokens_seen": 607248,
      "step": 975
    },
    {
      "epoch": 1.7468805704099821,
      "grad_norm": 15.625,
      "learning_rate": 4.9159181978010814e-05,
      "loss": 0.1579,
      "num_input_tokens_seen": 610736,
      "step": 980
    },
    {
      "epoch": 1.7557932263814617,
      "grad_norm": 0.09423828125,
      "learning_rate": 4.9139063349722113e-05,
      "loss": 0.0559,
      "num_input_tokens_seen": 614128,
      "step": 985
    },
    {
      "epoch": 1.7647058823529411,
      "grad_norm": 19.0,
      "learning_rate": 4.911871108042241e-05,
      "loss": 0.0736,
      "num_input_tokens_seen": 617232,
      "step": 990
    },
    {
      "epoch": 1.7736185383244205,
      "grad_norm": 36.25,
      "learning_rate": 4.909812536710048e-05,
      "loss": 0.064,
      "num_input_tokens_seen": 620880,
      "step": 995
    },
    {
      "epoch": 1.7825311942959001,
      "grad_norm": 10.5,
      "learning_rate": 4.9077306409004585e-05,
      "loss": 0.0652,
      "num_input_tokens_seen": 624368,
      "step": 1000
    },
    {
      "epoch": 1.7914438502673797,
      "grad_norm": 8.4375,
      "learning_rate": 4.9056254407640604e-05,
      "loss": 0.0429,
      "num_input_tokens_seen": 627152,
      "step": 1005
    },
    {
      "epoch": 1.8003565062388591,
      "grad_norm": 4.3125,
      "learning_rate": 4.903496956676998e-05,
      "loss": 0.0436,
      "num_input_tokens_seen": 629680,
      "step": 1010
    },
    {
      "epoch": 1.8092691622103387,
      "grad_norm": 27.625,
      "learning_rate": 4.901345209240784e-05,
      "loss": 0.1416,
      "num_input_tokens_seen": 632848,
      "step": 1015
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.2255859375,
      "learning_rate": 4.8991702192820924e-05,
      "loss": 0.0298,
      "num_input_tokens_seen": 635920,
      "step": 1020
    },
    {
      "epoch": 1.8270944741532977,
      "grad_norm": 4.4375,
      "learning_rate": 4.896972007852563e-05,
      "loss": 0.0748,
      "num_input_tokens_seen": 639056,
      "step": 1025
    },
    {
      "epoch": 1.8360071301247771,
      "grad_norm": 41.0,
      "learning_rate": 4.894750596228594e-05,
      "loss": 0.0602,
      "num_input_tokens_seen": 642192,
      "step": 1030
    },
    {
      "epoch": 1.8449197860962567,
      "grad_norm": 1.8828125,
      "learning_rate": 4.8925060059111394e-05,
      "loss": 0.0054,
      "num_input_tokens_seen": 645488,
      "step": 1035
    },
    {
      "epoch": 1.8538324420677363,
      "grad_norm": 20.5,
      "learning_rate": 4.890238258625496e-05,
      "loss": 0.1834,
      "num_input_tokens_seen": 648336,
      "step": 1040
    },
    {
      "epoch": 1.8627450980392157,
      "grad_norm": 1.671875,
      "learning_rate": 4.887947376321099e-05,
      "loss": 0.0974,
      "num_input_tokens_seen": 651696,
      "step": 1045
    },
    {
      "epoch": 1.8716577540106951,
      "grad_norm": 0.10107421875,
      "learning_rate": 4.885633381171304e-05,
      "loss": 0.0829,
      "num_input_tokens_seen": 654640,
      "step": 1050
    },
    {
      "epoch": 1.8805704099821747,
      "grad_norm": 0.061279296875,
      "learning_rate": 4.883296295573176e-05,
      "loss": 0.157,
      "num_input_tokens_seen": 658128,
      "step": 1055
    },
    {
      "epoch": 1.8894830659536543,
      "grad_norm": 13.25,
      "learning_rate": 4.880936142147271e-05,
      "loss": 0.1809,
      "num_input_tokens_seen": 660848,
      "step": 1060
    },
    {
      "epoch": 1.8983957219251337,
      "grad_norm": 0.75,
      "learning_rate": 4.878552943737418e-05,
      "loss": 0.1404,
      "num_input_tokens_seen": 663120,
      "step": 1065
    },
    {
      "epoch": 1.9073083778966131,
      "grad_norm": 2.9375,
      "learning_rate": 4.876146723410498e-05,
      "loss": 0.05,
      "num_input_tokens_seen": 666288,
      "step": 1070
    },
    {
      "epoch": 1.9162210338680927,
      "grad_norm": 3.859375,
      "learning_rate": 4.873717504456219e-05,
      "loss": 0.0693,
      "num_input_tokens_seen": 669360,
      "step": 1075
    },
    {
      "epoch": 1.9251336898395723,
      "grad_norm": 14.4375,
      "learning_rate": 4.8712653103868916e-05,
      "loss": 0.1736,
      "num_input_tokens_seen": 671344,
      "step": 1080
    },
    {
      "epoch": 1.9340463458110517,
      "grad_norm": 0.66015625,
      "learning_rate": 4.868790164937204e-05,
      "loss": 0.0195,
      "num_input_tokens_seen": 674672,
      "step": 1085
    },
    {
      "epoch": 1.9429590017825311,
      "grad_norm": 4.25,
      "learning_rate": 4.8662920920639866e-05,
      "loss": 0.1089,
      "num_input_tokens_seen": 677968,
      "step": 1090
    },
    {
      "epoch": 1.9518716577540107,
      "grad_norm": 5.375,
      "learning_rate": 4.8637711159459855e-05,
      "loss": 0.0145,
      "num_input_tokens_seen": 680560,
      "step": 1095
    },
    {
      "epoch": 1.9607843137254903,
      "grad_norm": 8.625,
      "learning_rate": 4.8612272609836263e-05,
      "loss": 0.0347,
      "num_input_tokens_seen": 683824,
      "step": 1100
    },
    {
      "epoch": 1.9696969696969697,
      "grad_norm": 0.73828125,
      "learning_rate": 4.858660551798778e-05,
      "loss": 0.0616,
      "num_input_tokens_seen": 687216,
      "step": 1105
    },
    {
      "epoch": 1.9786096256684491,
      "grad_norm": 28.125,
      "learning_rate": 4.856071013234513e-05,
      "loss": 0.1747,
      "num_input_tokens_seen": 690128,
      "step": 1110
    },
    {
      "epoch": 1.9875222816399287,
      "grad_norm": 18.75,
      "learning_rate": 4.85345867035487e-05,
      "loss": 0.1298,
      "num_input_tokens_seen": 693232,
      "step": 1115
    },
    {
      "epoch": 1.9964349376114083,
      "grad_norm": 0.43359375,
      "learning_rate": 4.8508235484446095e-05,
      "loss": 0.1243,
      "num_input_tokens_seen": 696880,
      "step": 1120
    },
    {
      "epoch": 2.0035650623885917,
      "eval_loss": 0.15137933194637299,
      "eval_runtime": 6.2699,
      "eval_samples_per_second": 39.714,
      "eval_steps_per_second": 10.048,
      "num_input_tokens_seen": 699264,
      "step": 1124
    },
    {
      "epoch": 2.0053475935828877,
      "grad_norm": 27.625,
      "learning_rate": 4.8481656730089695e-05,
      "loss": 0.0738,
      "num_input_tokens_seen": 700096,
      "step": 1125
    },
    {
      "epoch": 2.014260249554367,
      "grad_norm": 15.5,
      "learning_rate": 4.8454850697734174e-05,
      "loss": 0.1989,
      "num_input_tokens_seen": 703360,
      "step": 1130
    },
    {
      "epoch": 2.0231729055258465,
      "grad_norm": 0.7109375,
      "learning_rate": 4.842781764683403e-05,
      "loss": 0.0507,
      "num_input_tokens_seen": 706624,
      "step": 1135
    },
    {
      "epoch": 2.0320855614973263,
      "grad_norm": 17.625,
      "learning_rate": 4.8400557839041064e-05,
      "loss": 0.1906,
      "num_input_tokens_seen": 709472,
      "step": 1140
    },
    {
      "epoch": 2.0409982174688057,
      "grad_norm": 12.4375,
      "learning_rate": 4.837307153820184e-05,
      "loss": 0.0311,
      "num_input_tokens_seen": 713152,
      "step": 1145
    },
    {
      "epoch": 2.049910873440285,
      "grad_norm": 7.3125,
      "learning_rate": 4.8345359010355155e-05,
      "loss": 0.0409,
      "num_input_tokens_seen": 716480,
      "step": 1150
    },
    {
      "epoch": 2.0588235294117645,
      "grad_norm": 0.59375,
      "learning_rate": 4.831742052372943e-05,
      "loss": 0.0037,
      "num_input_tokens_seen": 719104,
      "step": 1155
    },
    {
      "epoch": 2.0677361853832443,
      "grad_norm": 17.75,
      "learning_rate": 4.828925634874014e-05,
      "loss": 0.0205,
      "num_input_tokens_seen": 722016,
      "step": 1160
    },
    {
      "epoch": 2.0766488413547237,
      "grad_norm": 16.875,
      "learning_rate": 4.8260866757987177e-05,
      "loss": 0.1525,
      "num_input_tokens_seen": 725184,
      "step": 1165
    },
    {
      "epoch": 2.085561497326203,
      "grad_norm": 25.0,
      "learning_rate": 4.823225202625226e-05,
      "loss": 0.1205,
      "num_input_tokens_seen": 728352,
      "step": 1170
    },
    {
      "epoch": 2.0944741532976825,
      "grad_norm": 0.0108642578125,
      "learning_rate": 4.820341243049618e-05,
      "loss": 0.1138,
      "num_input_tokens_seen": 731712,
      "step": 1175
    },
    {
      "epoch": 2.1033868092691623,
      "grad_norm": 0.1123046875,
      "learning_rate": 4.8174348249856236e-05,
      "loss": 0.0186,
      "num_input_tokens_seen": 734880,
      "step": 1180
    },
    {
      "epoch": 2.1122994652406417,
      "grad_norm": 3.359375,
      "learning_rate": 4.814505976564343e-05,
      "loss": 0.066,
      "num_input_tokens_seen": 737728,
      "step": 1185
    },
    {
      "epoch": 2.121212121212121,
      "grad_norm": 25.0,
      "learning_rate": 4.8115547261339824e-05,
      "loss": 0.0528,
      "num_input_tokens_seen": 741376,
      "step": 1190
    },
    {
      "epoch": 2.1301247771836005,
      "grad_norm": 18.0,
      "learning_rate": 4.808581102259573e-05,
      "loss": 0.1943,
      "num_input_tokens_seen": 744256,
      "step": 1195
    },
    {
      "epoch": 2.1390374331550803,
      "grad_norm": 25.375,
      "learning_rate": 4.8055851337227006e-05,
      "loss": 0.1003,
      "num_input_tokens_seen": 746944,
      "step": 1200
    },
    {
      "epoch": 2.1479500891265597,
      "grad_norm": 14.4375,
      "learning_rate": 4.802566849521222e-05,
      "loss": 0.1651,
      "num_input_tokens_seen": 750272,
      "step": 1205
    },
    {
      "epoch": 2.156862745098039,
      "grad_norm": 2.453125,
      "learning_rate": 4.799526278868987e-05,
      "loss": 0.2084,
      "num_input_tokens_seen": 753024,
      "step": 1210
    },
    {
      "epoch": 2.165775401069519,
      "grad_norm": 0.03125,
      "learning_rate": 4.796463451195554e-05,
      "loss": 0.0164,
      "num_input_tokens_seen": 756576,
      "step": 1215
    },
    {
      "epoch": 2.1746880570409983,
      "grad_norm": 6.90625,
      "learning_rate": 4.7933783961459094e-05,
      "loss": 0.0881,
      "num_input_tokens_seen": 759680,
      "step": 1220
    },
    {
      "epoch": 2.1836007130124777,
      "grad_norm": 0.140625,
      "learning_rate": 4.790271143580174e-05,
      "loss": 0.0008,
      "num_input_tokens_seen": 762880,
      "step": 1225
    },
    {
      "epoch": 2.192513368983957,
      "grad_norm": 0.01544189453125,
      "learning_rate": 4.7871417235733196e-05,
      "loss": 0.0389,
      "num_input_tokens_seen": 765920,
      "step": 1230
    },
    {
      "epoch": 2.2014260249554365,
      "grad_norm": 0.486328125,
      "learning_rate": 4.783990166414875e-05,
      "loss": 0.0392,
      "num_input_tokens_seen": 769728,
      "step": 1235
    },
    {
      "epoch": 2.2103386809269163,
      "grad_norm": 0.07421875,
      "learning_rate": 4.780816502608632e-05,
      "loss": 0.109,
      "num_input_tokens_seen": 772832,
      "step": 1240
    },
    {
      "epoch": 2.2192513368983957,
      "grad_norm": 17.875,
      "learning_rate": 4.777620762872355e-05,
      "loss": 0.1271,
      "num_input_tokens_seen": 776352,
      "step": 1245
    },
    {
      "epoch": 2.228163992869875,
      "grad_norm": 0.0177001953125,
      "learning_rate": 4.774402978137479e-05,
      "loss": 0.0143,
      "num_input_tokens_seen": 779456,
      "step": 1250
    },
    {
      "epoch": 2.237076648841355,
      "grad_norm": 1.25,
      "learning_rate": 4.7711631795488096e-05,
      "loss": 0.0065,
      "num_input_tokens_seen": 782112,
      "step": 1255
    },
    {
      "epoch": 2.2459893048128343,
      "grad_norm": 0.1875,
      "learning_rate": 4.767901398464227e-05,
      "loss": 0.0725,
      "num_input_tokens_seen": 784864,
      "step": 1260
    },
    {
      "epoch": 2.2549019607843137,
      "grad_norm": 0.171875,
      "learning_rate": 4.7646176664543763e-05,
      "loss": 0.0415,
      "num_input_tokens_seen": 787936,
      "step": 1265
    },
    {
      "epoch": 2.263814616755793,
      "grad_norm": 0.052490234375,
      "learning_rate": 4.761312015302367e-05,
      "loss": 0.0875,
      "num_input_tokens_seen": 790976,
      "step": 1270
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 2.84375,
      "learning_rate": 4.757984477003462e-05,
      "loss": 0.0067,
      "num_input_tokens_seen": 794016,
      "step": 1275
    },
    {
      "epoch": 2.2816399286987523,
      "grad_norm": 2.140625,
      "learning_rate": 4.7546350837647666e-05,
      "loss": 0.0988,
      "num_input_tokens_seen": 796864,
      "step": 1280
    },
    {
      "epoch": 2.2905525846702317,
      "grad_norm": 0.376953125,
      "learning_rate": 4.7512638680049245e-05,
      "loss": 0.0008,
      "num_input_tokens_seen": 800096,
      "step": 1285
    },
    {
      "epoch": 2.299465240641711,
      "grad_norm": 25.875,
      "learning_rate": 4.7478708623537956e-05,
      "loss": 0.141,
      "num_input_tokens_seen": 803392,
      "step": 1290
    },
    {
      "epoch": 2.308377896613191,
      "grad_norm": 0.02490234375,
      "learning_rate": 4.7444560996521415e-05,
      "loss": 0.0787,
      "num_input_tokens_seen": 806400,
      "step": 1295
    },
    {
      "epoch": 2.3172905525846703,
      "grad_norm": 35.75,
      "learning_rate": 4.741019612951312e-05,
      "loss": 0.3434,
      "num_input_tokens_seen": 809568,
      "step": 1300
    },
    {
      "epoch": 2.3262032085561497,
      "grad_norm": 4.3125,
      "learning_rate": 4.737561435512923e-05,
      "loss": 0.0448,
      "num_input_tokens_seen": 812768,
      "step": 1305
    },
    {
      "epoch": 2.335115864527629,
      "grad_norm": 0.625,
      "learning_rate": 4.734081600808531e-05,
      "loss": 0.0257,
      "num_input_tokens_seen": 815968,
      "step": 1310
    },
    {
      "epoch": 2.344028520499109,
      "grad_norm": 0.322265625,
      "learning_rate": 4.7305801425193165e-05,
      "loss": 0.0109,
      "num_input_tokens_seen": 818976,
      "step": 1315
    },
    {
      "epoch": 2.3529411764705883,
      "grad_norm": 0.388671875,
      "learning_rate": 4.727057094535749e-05,
      "loss": 0.1287,
      "num_input_tokens_seen": 821760,
      "step": 1320
    },
    {
      "epoch": 2.3618538324420677,
      "grad_norm": 11.0625,
      "learning_rate": 4.72351249095727e-05,
      "loss": 0.1399,
      "num_input_tokens_seen": 824288,
      "step": 1325
    },
    {
      "epoch": 2.370766488413547,
      "grad_norm": 0.01220703125,
      "learning_rate": 4.7199463660919514e-05,
      "loss": 0.0546,
      "num_input_tokens_seen": 827424,
      "step": 1330
    },
    {
      "epoch": 2.379679144385027,
      "grad_norm": 15.1875,
      "learning_rate": 4.7163587544561705e-05,
      "loss": 0.1861,
      "num_input_tokens_seen": 830176,
      "step": 1335
    },
    {
      "epoch": 2.3885918003565063,
      "grad_norm": 0.08154296875,
      "learning_rate": 4.7127496907742734e-05,
      "loss": 0.0112,
      "num_input_tokens_seen": 833664,
      "step": 1340
    },
    {
      "epoch": 2.3975044563279857,
      "grad_norm": 12.4375,
      "learning_rate": 4.709119209978242e-05,
      "loss": 0.108,
      "num_input_tokens_seen": 836736,
      "step": 1345
    },
    {
      "epoch": 2.406417112299465,
      "grad_norm": 0.0240478515625,
      "learning_rate": 4.7054673472073506e-05,
      "loss": 0.0531,
      "num_input_tokens_seen": 840160,
      "step": 1350
    },
    {
      "epoch": 2.415329768270945,
      "grad_norm": 19.625,
      "learning_rate": 4.7017941378078314e-05,
      "loss": 0.1849,
      "num_input_tokens_seen": 843168,
      "step": 1355
    },
    {
      "epoch": 2.4242424242424243,
      "grad_norm": 0.1416015625,
      "learning_rate": 4.698099617332528e-05,
      "loss": 0.0017,
      "num_input_tokens_seen": 845952,
      "step": 1360
    },
    {
      "epoch": 2.4331550802139037,
      "grad_norm": 2.75,
      "learning_rate": 4.694383821540555e-05,
      "loss": 0.005,
      "num_input_tokens_seen": 848448,
      "step": 1365
    },
    {
      "epoch": 2.442067736185383,
      "grad_norm": 0.044921875,
      "learning_rate": 4.690646786396945e-05,
      "loss": 0.0923,
      "num_input_tokens_seen": 851552,
      "step": 1370
    },
    {
      "epoch": 2.450980392156863,
      "grad_norm": 15.75,
      "learning_rate": 4.686888548072312e-05,
      "loss": 0.0954,
      "num_input_tokens_seen": 854752,
      "step": 1375
    },
    {
      "epoch": 2.4598930481283423,
      "grad_norm": 0.02294921875,
      "learning_rate": 4.683109142942492e-05,
      "loss": 0.0676,
      "num_input_tokens_seen": 857600,
      "step": 1380
    },
    {
      "epoch": 2.4688057040998217,
      "grad_norm": 1.484375,
      "learning_rate": 4.679308607588192e-05,
      "loss": 0.0957,
      "num_input_tokens_seen": 861248,
      "step": 1385
    },
    {
      "epoch": 2.477718360071301,
      "grad_norm": 24.875,
      "learning_rate": 4.6754869787946386e-05,
      "loss": 0.1243,
      "num_input_tokens_seen": 865056,
      "step": 1390
    },
    {
      "epoch": 2.486631016042781,
      "grad_norm": 19.25,
      "learning_rate": 4.6716442935512214e-05,
      "loss": 0.1454,
      "num_input_tokens_seen": 867936,
      "step": 1395
    },
    {
      "epoch": 2.4955436720142603,
      "grad_norm": 1.2890625,
      "learning_rate": 4.6677805890511354e-05,
      "loss": 0.1073,
      "num_input_tokens_seen": 871136,
      "step": 1400
    },
    {
      "epoch": 2.5044563279857397,
      "grad_norm": 2.046875,
      "learning_rate": 4.663895902691018e-05,
      "loss": 0.0632,
      "num_input_tokens_seen": 873600,
      "step": 1405
    },
    {
      "epoch": 2.5044563279857397,
      "eval_loss": 0.1448797881603241,
      "eval_runtime": 6.3011,
      "eval_samples_per_second": 39.517,
      "eval_steps_per_second": 9.998,
      "num_input_tokens_seen": 873600,
      "step": 1405
    },
    {
      "epoch": 2.5133689839572195,
      "grad_norm": 14.0,
      "learning_rate": 4.659990272070591e-05,
      "loss": 0.0324,
      "num_input_tokens_seen": 877152,
      "step": 1410
    },
    {
      "epoch": 2.522281639928699,
      "grad_norm": 0.053466796875,
      "learning_rate": 4.656063734992294e-05,
      "loss": 0.0354,
      "num_input_tokens_seen": 880096,
      "step": 1415
    },
    {
      "epoch": 2.5311942959001783,
      "grad_norm": 0.103515625,
      "learning_rate": 4.6521163294609196e-05,
      "loss": 0.0523,
      "num_input_tokens_seen": 882944,
      "step": 1420
    },
    {
      "epoch": 2.5401069518716577,
      "grad_norm": 22.0,
      "learning_rate": 4.6481480936832444e-05,
      "loss": 0.0934,
      "num_input_tokens_seen": 886848,
      "step": 1425
    },
    {
      "epoch": 2.549019607843137,
      "grad_norm": 24.5,
      "learning_rate": 4.644159066067662e-05,
      "loss": 0.0791,
      "num_input_tokens_seen": 890272,
      "step": 1430
    },
    {
      "epoch": 2.557932263814617,
      "grad_norm": 22.375,
      "learning_rate": 4.640149285223806e-05,
      "loss": 0.0603,
      "num_input_tokens_seen": 893600,
      "step": 1435
    },
    {
      "epoch": 2.5668449197860963,
      "grad_norm": 2.421875,
      "learning_rate": 4.636118789962184e-05,
      "loss": 0.0062,
      "num_input_tokens_seen": 896448,
      "step": 1440
    },
    {
      "epoch": 2.5757575757575757,
      "grad_norm": 32.75,
      "learning_rate": 4.632067619293795e-05,
      "loss": 0.2594,
      "num_input_tokens_seen": 899424,
      "step": 1445
    },
    {
      "epoch": 2.5846702317290555,
      "grad_norm": 0.451171875,
      "learning_rate": 4.6279958124297554e-05,
      "loss": 0.0197,
      "num_input_tokens_seen": 902624,
      "step": 1450
    },
    {
      "epoch": 2.593582887700535,
      "grad_norm": 7.375,
      "learning_rate": 4.623903408780916e-05,
      "loss": 0.0287,
      "num_input_tokens_seen": 905568,
      "step": 1455
    },
    {
      "epoch": 2.6024955436720143,
      "grad_norm": 0.86328125,
      "learning_rate": 4.619790447957488e-05,
      "loss": 0.0068,
      "num_input_tokens_seen": 908960,
      "step": 1460
    },
    {
      "epoch": 2.6114081996434937,
      "grad_norm": 0.17578125,
      "learning_rate": 4.615656969768649e-05,
      "loss": 0.0148,
      "num_input_tokens_seen": 912640,
      "step": 1465
    },
    {
      "epoch": 2.620320855614973,
      "grad_norm": 0.07666015625,
      "learning_rate": 4.611503014222168e-05,
      "loss": 0.0079,
      "num_input_tokens_seen": 915328,
      "step": 1470
    },
    {
      "epoch": 2.629233511586453,
      "grad_norm": 31.125,
      "learning_rate": 4.6073286215240105e-05,
      "loss": 0.0921,
      "num_input_tokens_seen": 918656,
      "step": 1475
    },
    {
      "epoch": 2.6381461675579323,
      "grad_norm": 0.75,
      "learning_rate": 4.6031338320779534e-05,
      "loss": 0.1445,
      "num_input_tokens_seen": 921344,
      "step": 1480
    },
    {
      "epoch": 2.6470588235294117,
      "grad_norm": 0.16015625,
      "learning_rate": 4.598918686485193e-05,
      "loss": 0.1085,
      "num_input_tokens_seen": 924192,
      "step": 1485
    },
    {
      "epoch": 2.6559714795008915,
      "grad_norm": 0.828125,
      "learning_rate": 4.594683225543952e-05,
      "loss": 0.0044,
      "num_input_tokens_seen": 927424,
      "step": 1490
    },
    {
      "epoch": 2.664884135472371,
      "grad_norm": 0.59375,
      "learning_rate": 4.590427490249084e-05,
      "loss": 0.088,
      "num_input_tokens_seen": 930080,
      "step": 1495
    },
    {
      "epoch": 2.6737967914438503,
      "grad_norm": 7.3125,
      "learning_rate": 4.5861515217916785e-05,
      "loss": 0.227,
      "num_input_tokens_seen": 932768,
      "step": 1500
    },
    {
      "epoch": 2.6827094474153297,
      "grad_norm": 0.017822265625,
      "learning_rate": 4.581855361558659e-05,
      "loss": 0.095,
      "num_input_tokens_seen": 935904,
      "step": 1505
    },
    {
      "epoch": 2.691622103386809,
      "grad_norm": 19.5,
      "learning_rate": 4.577539051132386e-05,
      "loss": 0.2288,
      "num_input_tokens_seen": 938784,
      "step": 1510
    },
    {
      "epoch": 2.700534759358289,
      "grad_norm": 0.041748046875,
      "learning_rate": 4.573202632290252e-05,
      "loss": 0.0119,
      "num_input_tokens_seen": 941280,
      "step": 1515
    },
    {
      "epoch": 2.7094474153297683,
      "grad_norm": 6.5625,
      "learning_rate": 4.568846147004279e-05,
      "loss": 0.0493,
      "num_input_tokens_seen": 944672,
      "step": 1520
    },
    {
      "epoch": 2.7183600713012477,
      "grad_norm": 16.375,
      "learning_rate": 4.5644696374407105e-05,
      "loss": 0.0657,
      "num_input_tokens_seen": 948032,
      "step": 1525
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 4.03125,
      "learning_rate": 4.560073145959602e-05,
      "loss": 0.0595,
      "num_input_tokens_seen": 952000,
      "step": 1530
    },
    {
      "epoch": 2.736185383244207,
      "grad_norm": 21.125,
      "learning_rate": 4.555656715114419e-05,
      "loss": 0.1502,
      "num_input_tokens_seen": 955456,
      "step": 1535
    },
    {
      "epoch": 2.7450980392156863,
      "grad_norm": 0.1455078125,
      "learning_rate": 4.551220387651615e-05,
      "loss": 0.1952,
      "num_input_tokens_seen": 959232,
      "step": 1540
    },
    {
      "epoch": 2.7540106951871657,
      "grad_norm": 0.1279296875,
      "learning_rate": 4.546764206510221e-05,
      "loss": 0.1091,
      "num_input_tokens_seen": 962304,
      "step": 1545
    },
    {
      "epoch": 2.762923351158645,
      "grad_norm": 2.078125,
      "learning_rate": 4.542288214821433e-05,
      "loss": 0.0678,
      "num_input_tokens_seen": 965344,
      "step": 1550
    },
    {
      "epoch": 2.771836007130125,
      "grad_norm": 3.0625,
      "learning_rate": 4.5377924559081946e-05,
      "loss": 0.1246,
      "num_input_tokens_seen": 968032,
      "step": 1555
    },
    {
      "epoch": 2.7807486631016043,
      "grad_norm": 13.125,
      "learning_rate": 4.533276973284771e-05,
      "loss": 0.1625,
      "num_input_tokens_seen": 970624,
      "step": 1560
    },
    {
      "epoch": 2.7896613190730837,
      "grad_norm": 0.03515625,
      "learning_rate": 4.528741810656336e-05,
      "loss": 0.0861,
      "num_input_tokens_seen": 973760,
      "step": 1565
    },
    {
      "epoch": 2.7985739750445635,
      "grad_norm": 16.25,
      "learning_rate": 4.5241870119185426e-05,
      "loss": 0.0216,
      "num_input_tokens_seen": 976480,
      "step": 1570
    },
    {
      "epoch": 2.807486631016043,
      "grad_norm": 0.09912109375,
      "learning_rate": 4.519612621157103e-05,
      "loss": 0.0769,
      "num_input_tokens_seen": 979328,
      "step": 1575
    },
    {
      "epoch": 2.8163992869875223,
      "grad_norm": 2.125,
      "learning_rate": 4.515018682647359e-05,
      "loss": 0.0051,
      "num_input_tokens_seen": 982624,
      "step": 1580
    },
    {
      "epoch": 2.8253119429590017,
      "grad_norm": 0.45703125,
      "learning_rate": 4.510405240853854e-05,
      "loss": 0.1352,
      "num_input_tokens_seen": 985664,
      "step": 1585
    },
    {
      "epoch": 2.834224598930481,
      "grad_norm": 0.0322265625,
      "learning_rate": 4.505772340429905e-05,
      "loss": 0.0066,
      "num_input_tokens_seen": 989024,
      "step": 1590
    },
    {
      "epoch": 2.843137254901961,
      "grad_norm": 21.625,
      "learning_rate": 4.501120026217164e-05,
      "loss": 0.1211,
      "num_input_tokens_seen": 992160,
      "step": 1595
    },
    {
      "epoch": 2.8520499108734403,
      "grad_norm": 0.024169921875,
      "learning_rate": 4.496448343245192e-05,
      "loss": 0.0308,
      "num_input_tokens_seen": 995328,
      "step": 1600
    },
    {
      "epoch": 2.8609625668449197,
      "grad_norm": 5.40625,
      "learning_rate": 4.4917573367310184e-05,
      "loss": 0.011,
      "num_input_tokens_seen": 999136,
      "step": 1605
    },
    {
      "epoch": 2.8698752228163995,
      "grad_norm": 0.08056640625,
      "learning_rate": 4.4870470520787035e-05,
      "loss": 0.1852,
      "num_input_tokens_seen": 1001920,
      "step": 1610
    },
    {
      "epoch": 2.878787878787879,
      "grad_norm": 6.59375,
      "learning_rate": 4.482317534878901e-05,
      "loss": 0.0288,
      "num_input_tokens_seen": 1005632,
      "step": 1615
    },
    {
      "epoch": 2.8877005347593583,
      "grad_norm": 17.375,
      "learning_rate": 4.477568830908415e-05,
      "loss": 0.1688,
      "num_input_tokens_seen": 1009408,
      "step": 1620
    },
    {
      "epoch": 2.8966131907308377,
      "grad_norm": 15.375,
      "learning_rate": 4.4728009861297586e-05,
      "loss": 0.0989,
      "num_input_tokens_seen": 1012448,
      "step": 1625
    },
    {
      "epoch": 2.905525846702317,
      "grad_norm": 22.125,
      "learning_rate": 4.468014046690707e-05,
      "loss": 0.0401,
      "num_input_tokens_seen": 1015616,
      "step": 1630
    },
    {
      "epoch": 2.914438502673797,
      "grad_norm": 5.375,
      "learning_rate": 4.463208058923851e-05,
      "loss": 0.0336,
      "num_input_tokens_seen": 1018944,
      "step": 1635
    },
    {
      "epoch": 2.9233511586452763,
      "grad_norm": 8.9375,
      "learning_rate": 4.458383069346152e-05,
      "loss": 0.0084,
      "num_input_tokens_seen": 1021696,
      "step": 1640
    },
    {
      "epoch": 2.9322638146167557,
      "grad_norm": 0.28515625,
      "learning_rate": 4.453539124658486e-05,
      "loss": 0.1792,
      "num_input_tokens_seen": 1024832,
      "step": 1645
    },
    {
      "epoch": 2.9411764705882355,
      "grad_norm": 0.54296875,
      "learning_rate": 4.4486762717451975e-05,
      "loss": 0.055,
      "num_input_tokens_seen": 1027712,
      "step": 1650
    },
    {
      "epoch": 2.950089126559715,
      "grad_norm": 24.5,
      "learning_rate": 4.443794557673641e-05,
      "loss": 0.2018,
      "num_input_tokens_seen": 1031040,
      "step": 1655
    },
    {
      "epoch": 2.9590017825311943,
      "grad_norm": 5.21875,
      "learning_rate": 4.43889402969373e-05,
      "loss": 0.0196,
      "num_input_tokens_seen": 1033440,
      "step": 1660
    },
    {
      "epoch": 2.9679144385026737,
      "grad_norm": 12.3125,
      "learning_rate": 4.4339747352374726e-05,
      "loss": 0.0174,
      "num_input_tokens_seen": 1036864,
      "step": 1665
    },
    {
      "epoch": 2.976827094474153,
      "grad_norm": 0.03173828125,
      "learning_rate": 4.4290367219185206e-05,
      "loss": 0.0023,
      "num_input_tokens_seen": 1039808,
      "step": 1670
    },
    {
      "epoch": 2.985739750445633,
      "grad_norm": 7.53125,
      "learning_rate": 4.424080037531705e-05,
      "loss": 0.0292,
      "num_input_tokens_seen": 1043200,
      "step": 1675
    },
    {
      "epoch": 2.9946524064171123,
      "grad_norm": 16.875,
      "learning_rate": 4.4191047300525704e-05,
      "loss": 0.0496,
      "num_input_tokens_seen": 1045504,
      "step": 1680
    },
    {
      "epoch": 3.0035650623885917,
      "grad_norm": 5.6875,
      "learning_rate": 4.414110847636916e-05,
      "loss": 0.0335,
      "num_input_tokens_seen": 1047768,
      "step": 1685
    },
    {
      "epoch": 3.0053475935828877,
      "eval_loss": 0.14105600118637085,
      "eval_runtime": 6.283,
      "eval_samples_per_second": 39.631,
      "eval_steps_per_second": 10.027,
      "num_input_tokens_seen": 1048184,
      "step": 1686
    },
    {
      "epoch": 3.0124777183600715,
      "grad_norm": 0.177734375,
      "learning_rate": 4.409098438620326e-05,
      "loss": 0.0955,
      "num_input_tokens_seen": 1050456,
      "step": 1690
    },
    {
      "epoch": 3.021390374331551,
      "grad_norm": 0.01953125,
      "learning_rate": 4.404067551517703e-05,
      "loss": 0.034,
      "num_input_tokens_seen": 1053592,
      "step": 1695
    },
    {
      "epoch": 3.0303030303030303,
      "grad_norm": 0.310546875,
      "learning_rate": 4.399018235022799e-05,
      "loss": 0.0236,
      "num_input_tokens_seen": 1056664,
      "step": 1700
    },
    {
      "epoch": 3.0392156862745097,
      "grad_norm": 0.0228271484375,
      "learning_rate": 4.393950538007743e-05,
      "loss": 0.1618,
      "num_input_tokens_seen": 1059384,
      "step": 1705
    },
    {
      "epoch": 3.0481283422459895,
      "grad_norm": 0.07373046875,
      "learning_rate": 4.3888645095225675e-05,
      "loss": 0.0008,
      "num_input_tokens_seen": 1062168,
      "step": 1710
    },
    {
      "epoch": 3.057040998217469,
      "grad_norm": 0.032958984375,
      "learning_rate": 4.383760198794734e-05,
      "loss": 0.1569,
      "num_input_tokens_seen": 1064952,
      "step": 1715
    },
    {
      "epoch": 3.0659536541889483,
      "grad_norm": 0.03759765625,
      "learning_rate": 4.37863765522866e-05,
      "loss": 0.0255,
      "num_input_tokens_seen": 1067416,
      "step": 1720
    },
    {
      "epoch": 3.0748663101604277,
      "grad_norm": 8.9375,
      "learning_rate": 4.3734969284052345e-05,
      "loss": 0.0085,
      "num_input_tokens_seen": 1070552,
      "step": 1725
    },
    {
      "epoch": 3.0837789661319075,
      "grad_norm": 8.875,
      "learning_rate": 4.368338068081343e-05,
      "loss": 0.0115,
      "num_input_tokens_seen": 1074136,
      "step": 1730
    },
    {
      "epoch": 3.092691622103387,
      "grad_norm": 0.049560546875,
      "learning_rate": 4.3631611241893874e-05,
      "loss": 0.0007,
      "num_input_tokens_seen": 1077848,
      "step": 1735
    },
    {
      "epoch": 3.1016042780748663,
      "grad_norm": 0.0322265625,
      "learning_rate": 4.3579661468367924e-05,
      "loss": 0.0336,
      "num_input_tokens_seen": 1080664,
      "step": 1740
    },
    {
      "epoch": 3.1105169340463457,
      "grad_norm": 25.625,
      "learning_rate": 4.352753186305536e-05,
      "loss": 0.1682,
      "num_input_tokens_seen": 1083992,
      "step": 1745
    },
    {
      "epoch": 3.1194295900178255,
      "grad_norm": 13.5625,
      "learning_rate": 4.347522293051648e-05,
      "loss": 0.2081,
      "num_input_tokens_seen": 1087800,
      "step": 1750
    },
    {
      "epoch": 3.128342245989305,
      "grad_norm": 19.875,
      "learning_rate": 4.3422735177047324e-05,
      "loss": 0.0736,
      "num_input_tokens_seen": 1090776,
      "step": 1755
    },
    {
      "epoch": 3.1372549019607843,
      "grad_norm": 0.036376953125,
      "learning_rate": 4.337006911067473e-05,
      "loss": 0.0485,
      "num_input_tokens_seen": 1093624,
      "step": 1760
    },
    {
      "epoch": 3.1461675579322637,
      "grad_norm": 0.035400390625,
      "learning_rate": 4.331722524115139e-05,
      "loss": 0.0722,
      "num_input_tokens_seen": 1096472,
      "step": 1765
    },
    {
      "epoch": 3.1550802139037435,
      "grad_norm": 1.46875,
      "learning_rate": 4.3264204079950975e-05,
      "loss": 0.1286,
      "num_input_tokens_seen": 1099736,
      "step": 1770
    },
    {
      "epoch": 3.163992869875223,
      "grad_norm": 4.90625,
      "learning_rate": 4.321100614026315e-05,
      "loss": 0.0243,
      "num_input_tokens_seen": 1103384,
      "step": 1775
    },
    {
      "epoch": 3.1729055258467023,
      "grad_norm": 0.02880859375,
      "learning_rate": 4.31576319369886e-05,
      "loss": 0.0609,
      "num_input_tokens_seen": 1106520,
      "step": 1780
    },
    {
      "epoch": 3.1818181818181817,
      "grad_norm": 5.375,
      "learning_rate": 4.310408198673406e-05,
      "loss": 0.0166,
      "num_input_tokens_seen": 1109208,
      "step": 1785
    },
    {
      "epoch": 3.1907308377896615,
      "grad_norm": 0.1318359375,
      "learning_rate": 4.305035680780732e-05,
      "loss": 0.0331,
      "num_input_tokens_seen": 1112536,
      "step": 1790
    },
    {
      "epoch": 3.199643493761141,
      "grad_norm": 0.79296875,
      "learning_rate": 4.299645692021221e-05,
      "loss": 0.1358,
      "num_input_tokens_seen": 1115992,
      "step": 1795
    },
    {
      "epoch": 3.2085561497326203,
      "grad_norm": 0.64453125,
      "learning_rate": 4.294238284564354e-05,
      "loss": 0.0381,
      "num_input_tokens_seen": 1119192,
      "step": 1800
    },
    {
      "epoch": 3.2174688057040997,
      "grad_norm": 25.25,
      "learning_rate": 4.2888135107482067e-05,
      "loss": 0.0293,
      "num_input_tokens_seen": 1122552,
      "step": 1805
    },
    {
      "epoch": 3.2263814616755795,
      "grad_norm": 4.09375,
      "learning_rate": 4.283371423078945e-05,
      "loss": 0.0364,
      "num_input_tokens_seen": 1126072,
      "step": 1810
    },
    {
      "epoch": 3.235294117647059,
      "grad_norm": 16.375,
      "learning_rate": 4.277912074230312e-05,
      "loss": 0.1569,
      "num_input_tokens_seen": 1128792,
      "step": 1815
    },
    {
      "epoch": 3.2442067736185383,
      "grad_norm": 0.02197265625,
      "learning_rate": 4.272435517043125e-05,
      "loss": 0.0284,
      "num_input_tokens_seen": 1132152,
      "step": 1820
    },
    {
      "epoch": 3.2531194295900177,
      "grad_norm": 0.11328125,
      "learning_rate": 4.2669418045247576e-05,
      "loss": 0.0444,
      "num_input_tokens_seen": 1135064,
      "step": 1825
    },
    {
      "epoch": 3.2620320855614975,
      "grad_norm": 14.125,
      "learning_rate": 4.2614309898486297e-05,
      "loss": 0.0877,
      "num_input_tokens_seen": 1137976,
      "step": 1830
    },
    {
      "epoch": 3.270944741532977,
      "grad_norm": 0.06494140625,
      "learning_rate": 4.25590312635369e-05,
      "loss": 0.0404,
      "num_input_tokens_seen": 1141080,
      "step": 1835
    },
    {
      "epoch": 3.2798573975044563,
      "grad_norm": 49.25,
      "learning_rate": 4.250358267543907e-05,
      "loss": 0.0572,
      "num_input_tokens_seen": 1144376,
      "step": 1840
    },
    {
      "epoch": 3.2887700534759357,
      "grad_norm": 7.9375,
      "learning_rate": 4.244796467087741e-05,
      "loss": 0.136,
      "num_input_tokens_seen": 1147224,
      "step": 1845
    },
    {
      "epoch": 3.2976827094474155,
      "grad_norm": 0.02099609375,
      "learning_rate": 4.2392177788176335e-05,
      "loss": 0.0632,
      "num_input_tokens_seen": 1150360,
      "step": 1850
    },
    {
      "epoch": 3.306595365418895,
      "grad_norm": 9.3125,
      "learning_rate": 4.2336222567294804e-05,
      "loss": 0.0068,
      "num_input_tokens_seen": 1153688,
      "step": 1855
    },
    {
      "epoch": 3.3155080213903743,
      "grad_norm": 4.71875,
      "learning_rate": 4.228009954982112e-05,
      "loss": 0.169,
      "num_input_tokens_seen": 1157016,
      "step": 1860
    },
    {
      "epoch": 3.3244206773618536,
      "grad_norm": 0.23828125,
      "learning_rate": 4.22238092789677e-05,
      "loss": 0.0277,
      "num_input_tokens_seen": 1159768,
      "step": 1865
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.1943359375,
      "learning_rate": 4.2167352299565746e-05,
      "loss": 0.0685,
      "num_input_tokens_seen": 1162520,
      "step": 1870
    },
    {
      "epoch": 3.342245989304813,
      "grad_norm": 1.703125,
      "learning_rate": 4.21107291580601e-05,
      "loss": 0.0219,
      "num_input_tokens_seen": 1165336,
      "step": 1875
    },
    {
      "epoch": 3.3511586452762923,
      "grad_norm": 0.051025390625,
      "learning_rate": 4.205394040250382e-05,
      "loss": 0.0469,
      "num_input_tokens_seen": 1168632,
      "step": 1880
    },
    {
      "epoch": 3.3600713012477716,
      "grad_norm": 0.94140625,
      "learning_rate": 4.199698658255298e-05,
      "loss": 0.0652,
      "num_input_tokens_seen": 1171352,
      "step": 1885
    },
    {
      "epoch": 3.3689839572192515,
      "grad_norm": 0.01116943359375,
      "learning_rate": 4.193986824946125e-05,
      "loss": 0.0775,
      "num_input_tokens_seen": 1174360,
      "step": 1890
    },
    {
      "epoch": 3.377896613190731,
      "grad_norm": 0.055419921875,
      "learning_rate": 4.188258595607468e-05,
      "loss": 0.0253,
      "num_input_tokens_seen": 1177368,
      "step": 1895
    },
    {
      "epoch": 3.3868092691622103,
      "grad_norm": 4.625,
      "learning_rate": 4.182514025682625e-05,
      "loss": 0.0291,
      "num_input_tokens_seen": 1180824,
      "step": 1900
    },
    {
      "epoch": 3.3957219251336896,
      "grad_norm": 23.125,
      "learning_rate": 4.176753170773052e-05,
      "loss": 0.1351,
      "num_input_tokens_seen": 1183544,
      "step": 1905
    },
    {
      "epoch": 3.4046345811051695,
      "grad_norm": 0.279296875,
      "learning_rate": 4.170976086637832e-05,
      "loss": 0.0833,
      "num_input_tokens_seen": 1185848,
      "step": 1910
    },
    {
      "epoch": 3.413547237076649,
      "grad_norm": 27.125,
      "learning_rate": 4.1651828291931264e-05,
      "loss": 0.1992,
      "num_input_tokens_seen": 1189176,
      "step": 1915
    },
    {
      "epoch": 3.4224598930481283,
      "grad_norm": 0.1279296875,
      "learning_rate": 4.159373454511636e-05,
      "loss": 0.0105,
      "num_input_tokens_seen": 1192984,
      "step": 1920
    },
    {
      "epoch": 3.431372549019608,
      "grad_norm": 0.07373046875,
      "learning_rate": 4.1535480188220636e-05,
      "loss": 0.1187,
      "num_input_tokens_seen": 1196888,
      "step": 1925
    },
    {
      "epoch": 3.4402852049910875,
      "grad_norm": 0.0098876953125,
      "learning_rate": 4.1477065785085634e-05,
      "loss": 0.1077,
      "num_input_tokens_seen": 1200792,
      "step": 1930
    },
    {
      "epoch": 3.449197860962567,
      "grad_norm": 22.5,
      "learning_rate": 4.141849190110199e-05,
      "loss": 0.0855,
      "num_input_tokens_seen": 1203832,
      "step": 1935
    },
    {
      "epoch": 3.4581105169340463,
      "grad_norm": 0.035888671875,
      "learning_rate": 4.1359759103203935e-05,
      "loss": 0.1094,
      "num_input_tokens_seen": 1207160,
      "step": 1940
    },
    {
      "epoch": 3.4670231729055256,
      "grad_norm": 0.038818359375,
      "learning_rate": 4.130086795986383e-05,
      "loss": 0.083,
      "num_input_tokens_seen": 1210616,
      "step": 1945
    },
    {
      "epoch": 3.4759358288770055,
      "grad_norm": 0.451171875,
      "learning_rate": 4.124181904108664e-05,
      "loss": 0.0812,
      "num_input_tokens_seen": 1213528,
      "step": 1950
    },
    {
      "epoch": 3.484848484848485,
      "grad_norm": 1.0390625,
      "learning_rate": 4.1182612918404466e-05,
      "loss": 0.1351,
      "num_input_tokens_seen": 1216568,
      "step": 1955
    },
    {
      "epoch": 3.4937611408199643,
      "grad_norm": 0.65625,
      "learning_rate": 4.1123250164870955e-05,
      "loss": 0.0371,
      "num_input_tokens_seen": 1219896,
      "step": 1960
    },
    {
      "epoch": 3.502673796791444,
      "grad_norm": 0.68359375,
      "learning_rate": 4.1063731355055763e-05,
      "loss": 0.098,
      "num_input_tokens_seen": 1222904,
      "step": 1965
    },
    {
      "epoch": 3.5062388591800357,
      "eval_loss": 0.11949615180492401,
      "eval_runtime": 6.3174,
      "eval_samples_per_second": 39.415,
      "eval_steps_per_second": 9.972,
      "num_input_tokens_seen": 1223864,
      "step": 1967
    },
    {
      "epoch": 3.5115864527629235,
      "grad_norm": 2.28125,
      "learning_rate": 4.100405706503904e-05,
      "loss": 0.032,
      "num_input_tokens_seen": 1225496,
      "step": 1970
    },
    {
      "epoch": 3.520499108734403,
      "grad_norm": 1.1171875,
      "learning_rate": 4.094422787240581e-05,
      "loss": 0.0615,
      "num_input_tokens_seen": 1228280,
      "step": 1975
    },
    {
      "epoch": 3.5294117647058822,
      "grad_norm": 2.546875,
      "learning_rate": 4.088424435624038e-05,
      "loss": 0.0089,
      "num_input_tokens_seen": 1231288,
      "step": 1980
    },
    {
      "epoch": 3.5383244206773616,
      "grad_norm": 0.11083984375,
      "learning_rate": 4.082410709712077e-05,
      "loss": 0.097,
      "num_input_tokens_seen": 1234456,
      "step": 1985
    },
    {
      "epoch": 3.5472370766488415,
      "grad_norm": 54.5,
      "learning_rate": 4.0763816677113064e-05,
      "loss": 0.0988,
      "num_input_tokens_seen": 1237912,
      "step": 1990
    },
    {
      "epoch": 3.556149732620321,
      "grad_norm": 0.267578125,
      "learning_rate": 4.070337367976578e-05,
      "loss": 0.0549,
      "num_input_tokens_seen": 1240984,
      "step": 1995
    },
    {
      "epoch": 3.5650623885918002,
      "grad_norm": 0.416015625,
      "learning_rate": 4.064277869010421e-05,
      "loss": 0.0523,
      "num_input_tokens_seen": 1244280,
      "step": 2000
    },
    {
      "epoch": 3.57397504456328,
      "grad_norm": 0.015869140625,
      "learning_rate": 4.058203229462482e-05,
      "loss": 0.0064,
      "num_input_tokens_seen": 1246904,
      "step": 2005
    },
    {
      "epoch": 3.5828877005347595,
      "grad_norm": 25.0,
      "learning_rate": 4.052113508128948e-05,
      "loss": 0.2163,
      "num_input_tokens_seen": 1249880,
      "step": 2010
    },
    {
      "epoch": 3.591800356506239,
      "grad_norm": 20.0,
      "learning_rate": 4.0460087639519836e-05,
      "loss": 0.0497,
      "num_input_tokens_seen": 1252408,
      "step": 2015
    },
    {
      "epoch": 3.6007130124777182,
      "grad_norm": 7.5625,
      "learning_rate": 4.039889056019159e-05,
      "loss": 0.0485,
      "num_input_tokens_seen": 1255800,
      "step": 2020
    },
    {
      "epoch": 3.6096256684491976,
      "grad_norm": 0.1689453125,
      "learning_rate": 4.03375444356288e-05,
      "loss": 0.0137,
      "num_input_tokens_seen": 1259160,
      "step": 2025
    },
    {
      "epoch": 3.6185383244206775,
      "grad_norm": 31.5,
      "learning_rate": 4.0276049859598084e-05,
      "loss": 0.153,
      "num_input_tokens_seen": 1262488,
      "step": 2030
    },
    {
      "epoch": 3.627450980392157,
      "grad_norm": 4.46875,
      "learning_rate": 4.021440742730295e-05,
      "loss": 0.0235,
      "num_input_tokens_seen": 1265368,
      "step": 2035
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 14.125,
      "learning_rate": 4.015261773537799e-05,
      "loss": 0.1157,
      "num_input_tokens_seen": 1269112,
      "step": 2040
    },
    {
      "epoch": 3.645276292335116,
      "grad_norm": 0.0869140625,
      "learning_rate": 4.009068138188311e-05,
      "loss": 0.0795,
      "num_input_tokens_seen": 1272408,
      "step": 2045
    },
    {
      "epoch": 3.6541889483065955,
      "grad_norm": 22.625,
      "learning_rate": 4.002859896629776e-05,
      "loss": 0.069,
      "num_input_tokens_seen": 1275640,
      "step": 2050
    },
    {
      "epoch": 3.663101604278075,
      "grad_norm": 0.2080078125,
      "learning_rate": 3.99663710895151e-05,
      "loss": 0.041,
      "num_input_tokens_seen": 1278616,
      "step": 2055
    },
    {
      "epoch": 3.6720142602495542,
      "grad_norm": 12.9375,
      "learning_rate": 3.990399835383623e-05,
      "loss": 0.2036,
      "num_input_tokens_seen": 1281624,
      "step": 2060
    },
    {
      "epoch": 3.6809269162210336,
      "grad_norm": 0.032470703125,
      "learning_rate": 3.984148136296431e-05,
      "loss": 0.1056,
      "num_input_tokens_seen": 1284216,
      "step": 2065
    },
    {
      "epoch": 3.6898395721925135,
      "grad_norm": 0.072265625,
      "learning_rate": 3.977882072199874e-05,
      "loss": 0.0509,
      "num_input_tokens_seen": 1286808,
      "step": 2070
    },
    {
      "epoch": 3.698752228163993,
      "grad_norm": 0.267578125,
      "learning_rate": 3.971601703742932e-05,
      "loss": 0.0604,
      "num_input_tokens_seen": 1289944,
      "step": 2075
    },
    {
      "epoch": 3.7076648841354722,
      "grad_norm": 0.7890625,
      "learning_rate": 3.965307091713037e-05,
      "loss": 0.076,
      "num_input_tokens_seen": 1292856,
      "step": 2080
    },
    {
      "epoch": 3.716577540106952,
      "grad_norm": 0.01251220703125,
      "learning_rate": 3.95899829703548e-05,
      "loss": 0.0219,
      "num_input_tokens_seen": 1296792,
      "step": 2085
    },
    {
      "epoch": 3.7254901960784315,
      "grad_norm": 0.0703125,
      "learning_rate": 3.9526753807728295e-05,
      "loss": 0.0188,
      "num_input_tokens_seen": 1299800,
      "step": 2090
    },
    {
      "epoch": 3.734402852049911,
      "grad_norm": 0.107421875,
      "learning_rate": 3.946338404124334e-05,
      "loss": 0.018,
      "num_input_tokens_seen": 1302648,
      "step": 2095
    },
    {
      "epoch": 3.7433155080213902,
      "grad_norm": 0.44140625,
      "learning_rate": 3.939987428425331e-05,
      "loss": 0.0049,
      "num_input_tokens_seen": 1305016,
      "step": 2100
    },
    {
      "epoch": 3.7522281639928696,
      "grad_norm": 0.48046875,
      "learning_rate": 3.933622515146658e-05,
      "loss": 0.108,
      "num_input_tokens_seen": 1308024,
      "step": 2105
    },
    {
      "epoch": 3.7611408199643495,
      "grad_norm": 0.09765625,
      "learning_rate": 3.9272437258940494e-05,
      "loss": 0.0332,
      "num_input_tokens_seen": 1310552,
      "step": 2110
    },
    {
      "epoch": 3.770053475935829,
      "grad_norm": 2.0,
      "learning_rate": 3.9208511224075484e-05,
      "loss": 0.0833,
      "num_input_tokens_seen": 1313656,
      "step": 2115
    },
    {
      "epoch": 3.7789661319073082,
      "grad_norm": 15.0,
      "learning_rate": 3.914444766560902e-05,
      "loss": 0.0751,
      "num_input_tokens_seen": 1316728,
      "step": 2120
    },
    {
      "epoch": 3.787878787878788,
      "grad_norm": 0.390625,
      "learning_rate": 3.908024720360968e-05,
      "loss": 0.2379,
      "num_input_tokens_seen": 1320344,
      "step": 2125
    },
    {
      "epoch": 3.7967914438502675,
      "grad_norm": 9.1875,
      "learning_rate": 3.9015910459471126e-05,
      "loss": 0.1123,
      "num_input_tokens_seen": 1323416,
      "step": 2130
    },
    {
      "epoch": 3.805704099821747,
      "grad_norm": 3.078125,
      "learning_rate": 3.8951438055906084e-05,
      "loss": 0.0029,
      "num_input_tokens_seen": 1326360,
      "step": 2135
    },
    {
      "epoch": 3.8146167557932262,
      "grad_norm": 28.125,
      "learning_rate": 3.888683061694032e-05,
      "loss": 0.0255,
      "num_input_tokens_seen": 1329944,
      "step": 2140
    },
    {
      "epoch": 3.8235294117647056,
      "grad_norm": 5.28125,
      "learning_rate": 3.882208876790661e-05,
      "loss": 0.0605,
      "num_input_tokens_seen": 1333080,
      "step": 2145
    },
    {
      "epoch": 3.8324420677361855,
      "grad_norm": 13.25,
      "learning_rate": 3.8757213135438655e-05,
      "loss": 0.1111,
      "num_input_tokens_seen": 1336504,
      "step": 2150
    },
    {
      "epoch": 3.841354723707665,
      "grad_norm": 0.020751953125,
      "learning_rate": 3.869220434746509e-05,
      "loss": 0.0818,
      "num_input_tokens_seen": 1339704,
      "step": 2155
    },
    {
      "epoch": 3.8502673796791442,
      "grad_norm": 4.84375,
      "learning_rate": 3.862706303320329e-05,
      "loss": 0.0514,
      "num_input_tokens_seen": 1343032,
      "step": 2160
    },
    {
      "epoch": 3.859180035650624,
      "grad_norm": 0.005950927734375,
      "learning_rate": 3.856178982315342e-05,
      "loss": 0.1118,
      "num_input_tokens_seen": 1346104,
      "step": 2165
    },
    {
      "epoch": 3.8680926916221035,
      "grad_norm": 0.419921875,
      "learning_rate": 3.849638534909219e-05,
      "loss": 0.2059,
      "num_input_tokens_seen": 1348984,
      "step": 2170
    },
    {
      "epoch": 3.877005347593583,
      "grad_norm": 1.484375,
      "learning_rate": 3.843085024406686e-05,
      "loss": 0.0437,
      "num_input_tokens_seen": 1351480,
      "step": 2175
    },
    {
      "epoch": 3.8859180035650622,
      "grad_norm": 0.1337890625,
      "learning_rate": 3.836518514238903e-05,
      "loss": 0.1014,
      "num_input_tokens_seen": 1355448,
      "step": 2180
    },
    {
      "epoch": 3.8948306595365416,
      "grad_norm": 9.125,
      "learning_rate": 3.8299390679628555e-05,
      "loss": 0.0669,
      "num_input_tokens_seen": 1358392,
      "step": 2185
    },
    {
      "epoch": 3.9037433155080214,
      "grad_norm": 17.375,
      "learning_rate": 3.8233467492607354e-05,
      "loss": 0.0723,
      "num_input_tokens_seen": 1361368,
      "step": 2190
    },
    {
      "epoch": 3.912655971479501,
      "grad_norm": 0.025390625,
      "learning_rate": 3.816741621939327e-05,
      "loss": 0.0095,
      "num_input_tokens_seen": 1364536,
      "step": 2195
    },
    {
      "epoch": 3.9215686274509802,
      "grad_norm": 0.2080078125,
      "learning_rate": 3.81012374992939e-05,
      "loss": 0.0823,
      "num_input_tokens_seen": 1367800,
      "step": 2200
    },
    {
      "epoch": 3.93048128342246,
      "grad_norm": 1.4921875,
      "learning_rate": 3.803493197285036e-05,
      "loss": 0.0229,
      "num_input_tokens_seen": 1371224,
      "step": 2205
    },
    {
      "epoch": 3.9393939393939394,
      "grad_norm": 1.1171875,
      "learning_rate": 3.7968500281831146e-05,
      "loss": 0.0061,
      "num_input_tokens_seen": 1373944,
      "step": 2210
    },
    {
      "epoch": 3.948306595365419,
      "grad_norm": 0.0103759765625,
      "learning_rate": 3.79019430692259e-05,
      "loss": 0.0799,
      "num_input_tokens_seen": 1377240,
      "step": 2215
    },
    {
      "epoch": 3.9572192513368982,
      "grad_norm": 0.0257568359375,
      "learning_rate": 3.783526097923915e-05,
      "loss": 0.0031,
      "num_input_tokens_seen": 1380248,
      "step": 2220
    },
    {
      "epoch": 3.966131907308378,
      "grad_norm": 32.0,
      "learning_rate": 3.7768454657284154e-05,
      "loss": 0.1341,
      "num_input_tokens_seen": 1382712,
      "step": 2225
    },
    {
      "epoch": 3.9750445632798574,
      "grad_norm": 15.3125,
      "learning_rate": 3.770152474997657e-05,
      "loss": 0.0958,
      "num_input_tokens_seen": 1385976,
      "step": 2230
    },
    {
      "epoch": 3.983957219251337,
      "grad_norm": 0.38671875,
      "learning_rate": 3.763447190512824e-05,
      "loss": 0.0098,
      "num_input_tokens_seen": 1389624,
      "step": 2235
    },
    {
      "epoch": 3.9928698752228167,
      "grad_norm": 15.25,
      "learning_rate": 3.7567296771740925e-05,
      "loss": 0.102,
      "num_input_tokens_seen": 1392728,
      "step": 2240
    },
    {
      "epoch": 4.001782531194296,
      "grad_norm": 0.0294189453125,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.0022,
      "num_input_tokens_seen": 1395704,
      "step": 2245
    },
    {
      "epoch": 4.007130124777183,
      "eval_loss": 0.14008468389511108,
      "eval_runtime": 6.3066,
      "eval_samples_per_second": 39.482,
      "eval_steps_per_second": 9.99,
      "num_input_tokens_seen": 1397624,
      "step": 2248
    },
    {
      "epoch": 4.010695187165775,
      "grad_norm": 0.0693359375,
      "learning_rate": 3.743258224126819e-05,
      "loss": 0.0091,
      "num_input_tokens_seen": 1398584,
      "step": 2250
    },
    {
      "epoch": 4.019607843137255,
      "grad_norm": 28.375,
      "learning_rate": 3.736504414807922e-05,
      "loss": 0.0594,
      "num_input_tokens_seen": 1401784,
      "step": 2255
    },
    {
      "epoch": 4.028520499108734,
      "grad_norm": 1.03125,
      "learning_rate": 3.729738637413156e-05,
      "loss": 0.0225,
      "num_input_tokens_seen": 1404312,
      "step": 2260
    },
    {
      "epoch": 4.037433155080214,
      "grad_norm": 0.85546875,
      "learning_rate": 3.722960957428203e-05,
      "loss": 0.1722,
      "num_input_tokens_seen": 1407352,
      "step": 2265
    },
    {
      "epoch": 4.046345811051693,
      "grad_norm": 0.03857421875,
      "learning_rate": 3.716171440453952e-05,
      "loss": 0.0917,
      "num_input_tokens_seen": 1410648,
      "step": 2270
    },
    {
      "epoch": 4.055258467023173,
      "grad_norm": 5.375,
      "learning_rate": 3.709370152205863e-05,
      "loss": 0.1049,
      "num_input_tokens_seen": 1413816,
      "step": 2275
    },
    {
      "epoch": 4.064171122994653,
      "grad_norm": 11.0625,
      "learning_rate": 3.7025571585133254e-05,
      "loss": 0.0348,
      "num_input_tokens_seen": 1416024,
      "step": 2280
    },
    {
      "epoch": 4.073083778966132,
      "grad_norm": 9.625,
      "learning_rate": 3.69573252531903e-05,
      "loss": 0.0132,
      "num_input_tokens_seen": 1419128,
      "step": 2285
    },
    {
      "epoch": 4.081996434937611,
      "grad_norm": 0.06396484375,
      "learning_rate": 3.6888963186783224e-05,
      "loss": 0.0493,
      "num_input_tokens_seen": 1421720,
      "step": 2290
    },
    {
      "epoch": 4.090909090909091,
      "grad_norm": 0.6640625,
      "learning_rate": 3.682048604758567e-05,
      "loss": 0.1269,
      "num_input_tokens_seen": 1424632,
      "step": 2295
    },
    {
      "epoch": 4.09982174688057,
      "grad_norm": 0.0247802734375,
      "learning_rate": 3.67518944983851e-05,
      "loss": 0.1213,
      "num_input_tokens_seen": 1427480,
      "step": 2300
    },
    {
      "epoch": 4.10873440285205,
      "grad_norm": 12.0,
      "learning_rate": 3.668318920307632e-05,
      "loss": 0.225,
      "num_input_tokens_seen": 1430296,
      "step": 2305
    },
    {
      "epoch": 4.117647058823529,
      "grad_norm": 49.75,
      "learning_rate": 3.6614370826655074e-05,
      "loss": 0.0336,
      "num_input_tokens_seen": 1432920,
      "step": 2310
    },
    {
      "epoch": 4.126559714795009,
      "grad_norm": 13.375,
      "learning_rate": 3.654544003521164e-05,
      "loss": 0.0948,
      "num_input_tokens_seen": 1435544,
      "step": 2315
    },
    {
      "epoch": 4.135472370766489,
      "grad_norm": 0.07470703125,
      "learning_rate": 3.647639749592433e-05,
      "loss": 0.0656,
      "num_input_tokens_seen": 1438040,
      "step": 2320
    },
    {
      "epoch": 4.144385026737968,
      "grad_norm": 0.49609375,
      "learning_rate": 3.640724387705308e-05,
      "loss": 0.0154,
      "num_input_tokens_seen": 1441528,
      "step": 2325
    },
    {
      "epoch": 4.153297682709447,
      "grad_norm": 0.375,
      "learning_rate": 3.633797984793294e-05,
      "loss": 0.0054,
      "num_input_tokens_seen": 1444920,
      "step": 2330
    },
    {
      "epoch": 4.162210338680927,
      "grad_norm": 23.125,
      "learning_rate": 3.626860607896764e-05,
      "loss": 0.0077,
      "num_input_tokens_seen": 1447896,
      "step": 2335
    },
    {
      "epoch": 4.171122994652406,
      "grad_norm": 0.10205078125,
      "learning_rate": 3.6199123241623046e-05,
      "loss": 0.0303,
      "num_input_tokens_seen": 1451256,
      "step": 2340
    },
    {
      "epoch": 4.180035650623886,
      "grad_norm": 0.053955078125,
      "learning_rate": 3.6129532008420715e-05,
      "loss": 0.1724,
      "num_input_tokens_seen": 1454136,
      "step": 2345
    },
    {
      "epoch": 4.188948306595365,
      "grad_norm": 5.0,
      "learning_rate": 3.605983305293137e-05,
      "loss": 0.1272,
      "num_input_tokens_seen": 1456504,
      "step": 2350
    },
    {
      "epoch": 4.197860962566845,
      "grad_norm": 22.25,
      "learning_rate": 3.599002704976835e-05,
      "loss": 0.0893,
      "num_input_tokens_seen": 1459768,
      "step": 2355
    },
    {
      "epoch": 4.206773618538325,
      "grad_norm": 0.224609375,
      "learning_rate": 3.592011467458113e-05,
      "loss": 0.0046,
      "num_input_tokens_seen": 1462392,
      "step": 2360
    },
    {
      "epoch": 4.215686274509804,
      "grad_norm": 42.0,
      "learning_rate": 3.585009660404873e-05,
      "loss": 0.0609,
      "num_input_tokens_seen": 1466040,
      "step": 2365
    },
    {
      "epoch": 4.224598930481283,
      "grad_norm": 10.75,
      "learning_rate": 3.577997351587322e-05,
      "loss": 0.1036,
      "num_input_tokens_seen": 1469208,
      "step": 2370
    },
    {
      "epoch": 4.233511586452763,
      "grad_norm": 0.01251220703125,
      "learning_rate": 3.5709746088773085e-05,
      "loss": 0.0649,
      "num_input_tokens_seen": 1472536,
      "step": 2375
    },
    {
      "epoch": 4.242424242424242,
      "grad_norm": 4.0625,
      "learning_rate": 3.563941500247676e-05,
      "loss": 0.0242,
      "num_input_tokens_seen": 1475608,
      "step": 2380
    },
    {
      "epoch": 4.251336898395722,
      "grad_norm": 0.09619140625,
      "learning_rate": 3.5568980937715945e-05,
      "loss": 0.0043,
      "num_input_tokens_seen": 1479256,
      "step": 2385
    },
    {
      "epoch": 4.260249554367201,
      "grad_norm": 13.25,
      "learning_rate": 3.54984445762191e-05,
      "loss": 0.0546,
      "num_input_tokens_seen": 1483064,
      "step": 2390
    },
    {
      "epoch": 4.269162210338681,
      "grad_norm": 12.9375,
      "learning_rate": 3.5427806600704785e-05,
      "loss": 0.0818,
      "num_input_tokens_seen": 1485880,
      "step": 2395
    },
    {
      "epoch": 4.278074866310161,
      "grad_norm": 0.5546875,
      "learning_rate": 3.535706769487509e-05,
      "loss": 0.017,
      "num_input_tokens_seen": 1489208,
      "step": 2400
    },
    {
      "epoch": 4.28698752228164,
      "grad_norm": 0.04443359375,
      "learning_rate": 3.5286228543409004e-05,
      "loss": 0.0034,
      "num_input_tokens_seen": 1492216,
      "step": 2405
    },
    {
      "epoch": 4.295900178253119,
      "grad_norm": 0.458984375,
      "learning_rate": 3.5215289831955786e-05,
      "loss": 0.0008,
      "num_input_tokens_seen": 1495960,
      "step": 2410
    },
    {
      "epoch": 4.304812834224599,
      "grad_norm": 1.8125,
      "learning_rate": 3.514425224712835e-05,
      "loss": 0.0146,
      "num_input_tokens_seen": 1498584,
      "step": 2415
    },
    {
      "epoch": 4.313725490196078,
      "grad_norm": 19.125,
      "learning_rate": 3.507311647649657e-05,
      "loss": 0.168,
      "num_input_tokens_seen": 1501880,
      "step": 2420
    },
    {
      "epoch": 4.322638146167558,
      "grad_norm": 0.0712890625,
      "learning_rate": 3.5001883208580665e-05,
      "loss": 0.063,
      "num_input_tokens_seen": 1505112,
      "step": 2425
    },
    {
      "epoch": 4.331550802139038,
      "grad_norm": 0.388671875,
      "learning_rate": 3.493055313284456e-05,
      "loss": 0.1401,
      "num_input_tokens_seen": 1507768,
      "step": 2430
    },
    {
      "epoch": 4.340463458110517,
      "grad_norm": 0.890625,
      "learning_rate": 3.485912693968913e-05,
      "loss": 0.0711,
      "num_input_tokens_seen": 1511224,
      "step": 2435
    },
    {
      "epoch": 4.349376114081997,
      "grad_norm": 0.07763671875,
      "learning_rate": 3.478760532044561e-05,
      "loss": 0.0693,
      "num_input_tokens_seen": 1514456,
      "step": 2440
    },
    {
      "epoch": 4.358288770053476,
      "grad_norm": 0.66796875,
      "learning_rate": 3.471598896736881e-05,
      "loss": 0.0348,
      "num_input_tokens_seen": 1517400,
      "step": 2445
    },
    {
      "epoch": 4.367201426024955,
      "grad_norm": 2.765625,
      "learning_rate": 3.464427857363052e-05,
      "loss": 0.1231,
      "num_input_tokens_seen": 1520664,
      "step": 2450
    },
    {
      "epoch": 4.376114081996435,
      "grad_norm": 0.1826171875,
      "learning_rate": 3.457247483331272e-05,
      "loss": 0.0743,
      "num_input_tokens_seen": 1523960,
      "step": 2455
    },
    {
      "epoch": 4.385026737967914,
      "grad_norm": 19.75,
      "learning_rate": 3.4500578441400876e-05,
      "loss": 0.0609,
      "num_input_tokens_seen": 1526616,
      "step": 2460
    },
    {
      "epoch": 4.393939393939394,
      "grad_norm": 0.051513671875,
      "learning_rate": 3.4428590093777244e-05,
      "loss": 0.1118,
      "num_input_tokens_seen": 1530808,
      "step": 2465
    },
    {
      "epoch": 4.402852049910873,
      "grad_norm": 0.1015625,
      "learning_rate": 3.43565104872141e-05,
      "loss": 0.0235,
      "num_input_tokens_seen": 1533336,
      "step": 2470
    },
    {
      "epoch": 4.411764705882353,
      "grad_norm": 0.060302734375,
      "learning_rate": 3.428434031936704e-05,
      "loss": 0.0906,
      "num_input_tokens_seen": 1535864,
      "step": 2475
    },
    {
      "epoch": 4.420677361853833,
      "grad_norm": 0.0279541015625,
      "learning_rate": 3.421208028876815e-05,
      "loss": 0.0636,
      "num_input_tokens_seen": 1539192,
      "step": 2480
    },
    {
      "epoch": 4.429590017825312,
      "grad_norm": 0.23828125,
      "learning_rate": 3.413973109481935e-05,
      "loss": 0.0018,
      "num_input_tokens_seen": 1542712,
      "step": 2485
    },
    {
      "epoch": 4.438502673796791,
      "grad_norm": 0.12109375,
      "learning_rate": 3.406729343778552e-05,
      "loss": 0.038,
      "num_input_tokens_seen": 1545272,
      "step": 2490
    },
    {
      "epoch": 4.447415329768271,
      "grad_norm": 0.2041015625,
      "learning_rate": 3.3994768018787815e-05,
      "loss": 0.0472,
      "num_input_tokens_seen": 1549464,
      "step": 2495
    },
    {
      "epoch": 4.45632798573975,
      "grad_norm": 0.0191650390625,
      "learning_rate": 3.392215553979679e-05,
      "loss": 0.0049,
      "num_input_tokens_seen": 1552280,
      "step": 2500
    },
    {
      "epoch": 4.46524064171123,
      "grad_norm": 5.53125,
      "learning_rate": 3.38494567036257e-05,
      "loss": 0.1194,
      "num_input_tokens_seen": 1555448,
      "step": 2505
    },
    {
      "epoch": 4.47415329768271,
      "grad_norm": 0.1064453125,
      "learning_rate": 3.3776672213923587e-05,
      "loss": 0.0021,
      "num_input_tokens_seen": 1559480,
      "step": 2510
    },
    {
      "epoch": 4.483065953654189,
      "grad_norm": 0.1435546875,
      "learning_rate": 3.370380277516858e-05,
      "loss": 0.078,
      "num_input_tokens_seen": 1562872,
      "step": 2515
    },
    {
      "epoch": 4.491978609625669,
      "grad_norm": 0.02783203125,
      "learning_rate": 3.3630849092661e-05,
      "loss": 0.0329,
      "num_input_tokens_seen": 1565752,
      "step": 2520
    },
    {
      "epoch": 4.500891265597148,
      "grad_norm": 0.08154296875,
      "learning_rate": 3.355781187251657e-05,
      "loss": 0.0106,
      "num_input_tokens_seen": 1568600,
      "step": 2525
    },
    {
      "epoch": 4.508021390374331,
      "eval_loss": 0.13448920845985413,
      "eval_runtime": 6.3033,
      "eval_samples_per_second": 39.503,
      "eval_steps_per_second": 9.995,
      "num_input_tokens_seen": 1570936,
      "step": 2529
    },
    {
      "epoch": 4.509803921568627,
      "grad_norm": 0.1796875,
      "learning_rate": 3.3484691821659584e-05,
      "loss": 0.0823,
      "num_input_tokens_seen": 1571512,
      "step": 2530
    },
    {
      "epoch": 4.518716577540107,
      "grad_norm": 1.3359375,
      "learning_rate": 3.3411489647816016e-05,
      "loss": 0.076,
      "num_input_tokens_seen": 1574232,
      "step": 2535
    },
    {
      "epoch": 4.527629233511586,
      "grad_norm": 0.08935546875,
      "learning_rate": 3.3338206059506736e-05,
      "loss": 0.0182,
      "num_input_tokens_seen": 1577816,
      "step": 2540
    },
    {
      "epoch": 4.536541889483066,
      "grad_norm": 0.134765625,
      "learning_rate": 3.326484176604061e-05,
      "loss": 0.0118,
      "num_input_tokens_seen": 1581368,
      "step": 2545
    },
    {
      "epoch": 4.545454545454545,
      "grad_norm": 0.02685546875,
      "learning_rate": 3.3191397477507655e-05,
      "loss": 0.0055,
      "num_input_tokens_seen": 1583800,
      "step": 2550
    },
    {
      "epoch": 4.554367201426025,
      "grad_norm": 0.44921875,
      "learning_rate": 3.3117873904772123e-05,
      "loss": 0.0965,
      "num_input_tokens_seen": 1587384,
      "step": 2555
    },
    {
      "epoch": 4.563279857397505,
      "grad_norm": 0.0966796875,
      "learning_rate": 3.30442717594657e-05,
      "loss": 0.0006,
      "num_input_tokens_seen": 1590328,
      "step": 2560
    },
    {
      "epoch": 4.572192513368984,
      "grad_norm": 21.25,
      "learning_rate": 3.297059175398056e-05,
      "loss": 0.0157,
      "num_input_tokens_seen": 1594136,
      "step": 2565
    },
    {
      "epoch": 4.581105169340463,
      "grad_norm": 14.3125,
      "learning_rate": 3.289683460146244e-05,
      "loss": 0.0866,
      "num_input_tokens_seen": 1597656,
      "step": 2570
    },
    {
      "epoch": 4.590017825311943,
      "grad_norm": 0.0908203125,
      "learning_rate": 3.282300101580386e-05,
      "loss": 0.0065,
      "num_input_tokens_seen": 1600536,
      "step": 2575
    },
    {
      "epoch": 4.598930481283422,
      "grad_norm": 0.224609375,
      "learning_rate": 3.274909171163706e-05,
      "loss": 0.0603,
      "num_input_tokens_seen": 1603832,
      "step": 2580
    },
    {
      "epoch": 4.607843137254902,
      "grad_norm": 0.080078125,
      "learning_rate": 3.2675107404327194e-05,
      "loss": 0.2128,
      "num_input_tokens_seen": 1607480,
      "step": 2585
    },
    {
      "epoch": 4.616755793226382,
      "grad_norm": 0.050537109375,
      "learning_rate": 3.2601048809965355e-05,
      "loss": 0.0019,
      "num_input_tokens_seen": 1610296,
      "step": 2590
    },
    {
      "epoch": 4.625668449197861,
      "grad_norm": 0.059326171875,
      "learning_rate": 3.2526916645361666e-05,
      "loss": 0.023,
      "num_input_tokens_seen": 1613336,
      "step": 2595
    },
    {
      "epoch": 4.634581105169341,
      "grad_norm": 1.96875,
      "learning_rate": 3.2452711628038324e-05,
      "loss": 0.1726,
      "num_input_tokens_seen": 1616152,
      "step": 2600
    },
    {
      "epoch": 4.64349376114082,
      "grad_norm": 0.244140625,
      "learning_rate": 3.2378434476222666e-05,
      "loss": 0.0814,
      "num_input_tokens_seen": 1620024,
      "step": 2605
    },
    {
      "epoch": 4.652406417112299,
      "grad_norm": 9.9375,
      "learning_rate": 3.2304085908840244e-05,
      "loss": 0.0128,
      "num_input_tokens_seen": 1623544,
      "step": 2610
    },
    {
      "epoch": 4.661319073083779,
      "grad_norm": 16.875,
      "learning_rate": 3.222966664550777e-05,
      "loss": 0.1778,
      "num_input_tokens_seen": 1626296,
      "step": 2615
    },
    {
      "epoch": 4.670231729055258,
      "grad_norm": 48.75,
      "learning_rate": 3.2155177406526304e-05,
      "loss": 0.0401,
      "num_input_tokens_seen": 1629336,
      "step": 2620
    },
    {
      "epoch": 4.6791443850267385,
      "grad_norm": 0.016845703125,
      "learning_rate": 3.208061891287414e-05,
      "loss": 0.0363,
      "num_input_tokens_seen": 1632888,
      "step": 2625
    },
    {
      "epoch": 4.688057040998218,
      "grad_norm": 39.5,
      "learning_rate": 3.200599188619989e-05,
      "loss": 0.158,
      "num_input_tokens_seen": 1635768,
      "step": 2630
    },
    {
      "epoch": 4.696969696969697,
      "grad_norm": 0.2021484375,
      "learning_rate": 3.1931297048815534e-05,
      "loss": 0.021,
      "num_input_tokens_seen": 1639256,
      "step": 2635
    },
    {
      "epoch": 4.705882352941177,
      "grad_norm": 0.0927734375,
      "learning_rate": 3.185653512368933e-05,
      "loss": 0.0132,
      "num_input_tokens_seen": 1643128,
      "step": 2640
    },
    {
      "epoch": 4.714795008912656,
      "grad_norm": 12.0,
      "learning_rate": 3.178170683443893e-05,
      "loss": 0.0264,
      "num_input_tokens_seen": 1646424,
      "step": 2645
    },
    {
      "epoch": 4.723707664884135,
      "grad_norm": 4.1875,
      "learning_rate": 3.1706812905324276e-05,
      "loss": 0.1036,
      "num_input_tokens_seen": 1649240,
      "step": 2650
    },
    {
      "epoch": 4.732620320855615,
      "grad_norm": 0.484375,
      "learning_rate": 3.1631854061240684e-05,
      "loss": 0.1607,
      "num_input_tokens_seen": 1652184,
      "step": 2655
    },
    {
      "epoch": 4.741532976827094,
      "grad_norm": 0.86328125,
      "learning_rate": 3.155683102771173e-05,
      "loss": 0.0155,
      "num_input_tokens_seen": 1655480,
      "step": 2660
    },
    {
      "epoch": 4.750445632798574,
      "grad_norm": 1.1640625,
      "learning_rate": 3.1481744530882305e-05,
      "loss": 0.0032,
      "num_input_tokens_seen": 1659352,
      "step": 2665
    },
    {
      "epoch": 4.759358288770054,
      "grad_norm": 0.11767578125,
      "learning_rate": 3.1406595297511566e-05,
      "loss": 0.0295,
      "num_input_tokens_seen": 1661976,
      "step": 2670
    },
    {
      "epoch": 4.768270944741533,
      "grad_norm": 0.173828125,
      "learning_rate": 3.133138405496587e-05,
      "loss": 0.0018,
      "num_input_tokens_seen": 1664504,
      "step": 2675
    },
    {
      "epoch": 4.777183600713013,
      "grad_norm": 0.328125,
      "learning_rate": 3.125611153121178e-05,
      "loss": 0.1255,
      "num_input_tokens_seen": 1667288,
      "step": 2680
    },
    {
      "epoch": 4.786096256684492,
      "grad_norm": 26.75,
      "learning_rate": 3.118077845480897e-05,
      "loss": 0.0822,
      "num_input_tokens_seen": 1670360,
      "step": 2685
    },
    {
      "epoch": 4.795008912655971,
      "grad_norm": 0.0966796875,
      "learning_rate": 3.110538555490324e-05,
      "loss": 0.106,
      "num_input_tokens_seen": 1673624,
      "step": 2690
    },
    {
      "epoch": 4.803921568627451,
      "grad_norm": 0.02001953125,
      "learning_rate": 3.1029933561219375e-05,
      "loss": 0.0022,
      "num_input_tokens_seen": 1676440,
      "step": 2695
    },
    {
      "epoch": 4.81283422459893,
      "grad_norm": 22.5,
      "learning_rate": 3.095442320405418e-05,
      "loss": 0.071,
      "num_input_tokens_seen": 1679448,
      "step": 2700
    },
    {
      "epoch": 4.8217468805704105,
      "grad_norm": 1.125,
      "learning_rate": 3.0878855214269293e-05,
      "loss": 0.0424,
      "num_input_tokens_seen": 1682520,
      "step": 2705
    },
    {
      "epoch": 4.83065953654189,
      "grad_norm": 3.625,
      "learning_rate": 3.0803230323284225e-05,
      "loss": 0.0025,
      "num_input_tokens_seen": 1685656,
      "step": 2710
    },
    {
      "epoch": 4.839572192513369,
      "grad_norm": 1.421875,
      "learning_rate": 3.0727549263069224e-05,
      "loss": 0.0027,
      "num_input_tokens_seen": 1688856,
      "step": 2715
    },
    {
      "epoch": 4.848484848484849,
      "grad_norm": 0.39453125,
      "learning_rate": 3.065181276613817e-05,
      "loss": 0.0328,
      "num_input_tokens_seen": 1691768,
      "step": 2720
    },
    {
      "epoch": 4.857397504456328,
      "grad_norm": 0.1767578125,
      "learning_rate": 3.057602156554155e-05,
      "loss": 0.0004,
      "num_input_tokens_seen": 1694488,
      "step": 2725
    },
    {
      "epoch": 4.866310160427807,
      "grad_norm": 18.5,
      "learning_rate": 3.0500176394859293e-05,
      "loss": 0.0695,
      "num_input_tokens_seen": 1697752,
      "step": 2730
    },
    {
      "epoch": 4.875222816399287,
      "grad_norm": 0.041015625,
      "learning_rate": 3.042427798819373e-05,
      "loss": 0.1446,
      "num_input_tokens_seen": 1700408,
      "step": 2735
    },
    {
      "epoch": 4.884135472370766,
      "grad_norm": 0.25,
      "learning_rate": 3.0348327080162435e-05,
      "loss": 0.1087,
      "num_input_tokens_seen": 1703512,
      "step": 2740
    },
    {
      "epoch": 4.893048128342246,
      "grad_norm": 47.0,
      "learning_rate": 3.0272324405891172e-05,
      "loss": 0.0457,
      "num_input_tokens_seen": 1707032,
      "step": 2745
    },
    {
      "epoch": 4.901960784313726,
      "grad_norm": 0.01373291015625,
      "learning_rate": 3.0196270701006706e-05,
      "loss": 0.0006,
      "num_input_tokens_seen": 1710328,
      "step": 2750
    },
    {
      "epoch": 4.910873440285205,
      "grad_norm": 0.045654296875,
      "learning_rate": 3.012016670162977e-05,
      "loss": 0.0688,
      "num_input_tokens_seen": 1712632,
      "step": 2755
    },
    {
      "epoch": 4.919786096256685,
      "grad_norm": 0.026611328125,
      "learning_rate": 3.0044013144367866e-05,
      "loss": 0.1258,
      "num_input_tokens_seen": 1716344,
      "step": 2760
    },
    {
      "epoch": 4.928698752228164,
      "grad_norm": 0.59375,
      "learning_rate": 2.996781076630816e-05,
      "loss": 0.0331,
      "num_input_tokens_seen": 1718712,
      "step": 2765
    },
    {
      "epoch": 4.937611408199643,
      "grad_norm": 4.15625,
      "learning_rate": 2.9891560305010392e-05,
      "loss": 0.0966,
      "num_input_tokens_seen": 1722328,
      "step": 2770
    },
    {
      "epoch": 4.946524064171123,
      "grad_norm": 0.65625,
      "learning_rate": 2.9815262498499657e-05,
      "loss": 0.0005,
      "num_input_tokens_seen": 1725464,
      "step": 2775
    },
    {
      "epoch": 4.955436720142602,
      "grad_norm": 0.007537841796875,
      "learning_rate": 2.9738918085259314e-05,
      "loss": 0.0312,
      "num_input_tokens_seen": 1728472,
      "step": 2780
    },
    {
      "epoch": 4.9643493761140824,
      "grad_norm": 2.625,
      "learning_rate": 2.9662527804223827e-05,
      "loss": 0.1607,
      "num_input_tokens_seen": 1731160,
      "step": 2785
    },
    {
      "epoch": 4.973262032085562,
      "grad_norm": 0.030517578125,
      "learning_rate": 2.9586092394771637e-05,
      "loss": 0.1619,
      "num_input_tokens_seen": 1734264,
      "step": 2790
    },
    {
      "epoch": 4.982174688057041,
      "grad_norm": 33.75,
      "learning_rate": 2.950961259671793e-05,
      "loss": 0.1195,
      "num_input_tokens_seen": 1737144,
      "step": 2795
    },
    {
      "epoch": 4.991087344028521,
      "grad_norm": 1.5234375,
      "learning_rate": 2.943308915030757e-05,
      "loss": 0.0633,
      "num_input_tokens_seen": 1740664,
      "step": 2800
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.51171875,
      "learning_rate": 2.935652279620788e-05,
      "loss": 0.0716,
      "num_input_tokens_seen": 1743216,
      "step": 2805
    },
    {
      "epoch": 5.008912655971479,
      "grad_norm": 0.034912109375,
      "learning_rate": 2.9279914275501473e-05,
      "loss": 0.077,
      "num_input_tokens_seen": 1746384,
      "step": 2810
    },
    {
      "epoch": 5.008912655971479,
      "eval_loss": 0.1459943950176239,
      "eval_runtime": 6.2925,
      "eval_samples_per_second": 39.571,
      "eval_steps_per_second": 10.012,
      "num_input_tokens_seen": 1746384,
      "step": 2810
    },
    {
      "epoch": 5.017825311942959,
      "grad_norm": 0.1953125,
      "learning_rate": 2.9203264329679115e-05,
      "loss": 0.014,
      "num_input_tokens_seen": 1749680,
      "step": 2815
    },
    {
      "epoch": 5.026737967914438,
      "grad_norm": 0.08935546875,
      "learning_rate": 2.9126573700632504e-05,
      "loss": 0.0014,
      "num_input_tokens_seen": 1753104,
      "step": 2820
    },
    {
      "epoch": 5.035650623885918,
      "grad_norm": 0.1865234375,
      "learning_rate": 2.9049843130647112e-05,
      "loss": 0.0459,
      "num_input_tokens_seen": 1756112,
      "step": 2825
    },
    {
      "epoch": 5.044563279857398,
      "grad_norm": 2.21875,
      "learning_rate": 2.8973073362394998e-05,
      "loss": 0.0925,
      "num_input_tokens_seen": 1759344,
      "step": 2830
    },
    {
      "epoch": 5.053475935828877,
      "grad_norm": 2.09375,
      "learning_rate": 2.8896265138927638e-05,
      "loss": 0.0094,
      "num_input_tokens_seen": 1762288,
      "step": 2835
    },
    {
      "epoch": 5.062388591800357,
      "grad_norm": 0.0191650390625,
      "learning_rate": 2.881941920366868e-05,
      "loss": 0.0154,
      "num_input_tokens_seen": 1765072,
      "step": 2840
    },
    {
      "epoch": 5.071301247771836,
      "grad_norm": 1.3203125,
      "learning_rate": 2.8742536300406804e-05,
      "loss": 0.1486,
      "num_input_tokens_seen": 1767952,
      "step": 2845
    },
    {
      "epoch": 5.080213903743315,
      "grad_norm": 1.75,
      "learning_rate": 2.8665617173288516e-05,
      "loss": 0.0356,
      "num_input_tokens_seen": 1770896,
      "step": 2850
    },
    {
      "epoch": 5.089126559714795,
      "grad_norm": 11.4375,
      "learning_rate": 2.8588662566810893e-05,
      "loss": 0.1233,
      "num_input_tokens_seen": 1773840,
      "step": 2855
    },
    {
      "epoch": 5.098039215686274,
      "grad_norm": 6.21875,
      "learning_rate": 2.851167322581445e-05,
      "loss": 0.0232,
      "num_input_tokens_seen": 1776720,
      "step": 2860
    },
    {
      "epoch": 5.106951871657754,
      "grad_norm": 0.022705078125,
      "learning_rate": 2.8434649895475877e-05,
      "loss": 0.044,
      "num_input_tokens_seen": 1779088,
      "step": 2865
    },
    {
      "epoch": 5.115864527629234,
      "grad_norm": 0.04638671875,
      "learning_rate": 2.8357593321300856e-05,
      "loss": 0.0049,
      "num_input_tokens_seen": 1781776,
      "step": 2870
    },
    {
      "epoch": 5.124777183600713,
      "grad_norm": 0.232421875,
      "learning_rate": 2.828050424911683e-05,
      "loss": 0.009,
      "num_input_tokens_seen": 1784720,
      "step": 2875
    },
    {
      "epoch": 5.133689839572193,
      "grad_norm": 2.140625,
      "learning_rate": 2.8203383425065787e-05,
      "loss": 0.0564,
      "num_input_tokens_seen": 1787856,
      "step": 2880
    },
    {
      "epoch": 5.142602495543672,
      "grad_norm": 0.220703125,
      "learning_rate": 2.812623159559704e-05,
      "loss": 0.001,
      "num_input_tokens_seen": 1791088,
      "step": 2885
    },
    {
      "epoch": 5.151515151515151,
      "grad_norm": 0.828125,
      "learning_rate": 2.8049049507460003e-05,
      "loss": 0.0714,
      "num_input_tokens_seen": 1795056,
      "step": 2890
    },
    {
      "epoch": 5.160427807486631,
      "grad_norm": 0.453125,
      "learning_rate": 2.7971837907696973e-05,
      "loss": 0.1153,
      "num_input_tokens_seen": 1798928,
      "step": 2895
    },
    {
      "epoch": 5.16934046345811,
      "grad_norm": 49.75,
      "learning_rate": 2.7894597543635863e-05,
      "loss": 0.0276,
      "num_input_tokens_seen": 1802384,
      "step": 2900
    },
    {
      "epoch": 5.17825311942959,
      "grad_norm": 1.265625,
      "learning_rate": 2.781732916288303e-05,
      "loss": 0.0015,
      "num_input_tokens_seen": 1805616,
      "step": 2905
    },
    {
      "epoch": 5.18716577540107,
      "grad_norm": 0.1865234375,
      "learning_rate": 2.774003351331597e-05,
      "loss": 0.0121,
      "num_input_tokens_seen": 1809008,
      "step": 2910
    },
    {
      "epoch": 5.196078431372549,
      "grad_norm": 29.125,
      "learning_rate": 2.7662711343076135e-05,
      "loss": 0.0308,
      "num_input_tokens_seen": 1812784,
      "step": 2915
    },
    {
      "epoch": 5.204991087344029,
      "grad_norm": 1.0390625,
      "learning_rate": 2.7585363400561658e-05,
      "loss": 0.0108,
      "num_input_tokens_seen": 1815248,
      "step": 2920
    },
    {
      "epoch": 5.213903743315508,
      "grad_norm": 13.8125,
      "learning_rate": 2.7507990434420126e-05,
      "loss": 0.0895,
      "num_input_tokens_seen": 1818032,
      "step": 2925
    },
    {
      "epoch": 5.222816399286987,
      "grad_norm": 0.0238037109375,
      "learning_rate": 2.7430593193541325e-05,
      "loss": 0.0009,
      "num_input_tokens_seen": 1821232,
      "step": 2930
    },
    {
      "epoch": 5.231729055258467,
      "grad_norm": 0.08935546875,
      "learning_rate": 2.7353172427049995e-05,
      "loss": 0.0073,
      "num_input_tokens_seen": 1824784,
      "step": 2935
    },
    {
      "epoch": 5.240641711229946,
      "grad_norm": 1.671875,
      "learning_rate": 2.7275728884298596e-05,
      "loss": 0.0176,
      "num_input_tokens_seen": 1827088,
      "step": 2940
    },
    {
      "epoch": 5.249554367201426,
      "grad_norm": 2.515625,
      "learning_rate": 2.719826331486e-05,
      "loss": 0.0017,
      "num_input_tokens_seen": 1829328,
      "step": 2945
    },
    {
      "epoch": 5.258467023172906,
      "grad_norm": 1.8359375,
      "learning_rate": 2.7120776468520314e-05,
      "loss": 0.0699,
      "num_input_tokens_seen": 1833136,
      "step": 2950
    },
    {
      "epoch": 5.267379679144385,
      "grad_norm": 0.04345703125,
      "learning_rate": 2.7043269095271573e-05,
      "loss": 0.0647,
      "num_input_tokens_seen": 1835632,
      "step": 2955
    },
    {
      "epoch": 5.276292335115865,
      "grad_norm": 0.051025390625,
      "learning_rate": 2.6965741945304467e-05,
      "loss": 0.0064,
      "num_input_tokens_seen": 1838992,
      "step": 2960
    },
    {
      "epoch": 5.285204991087344,
      "grad_norm": 2.15625,
      "learning_rate": 2.6888195769001146e-05,
      "loss": 0.0289,
      "num_input_tokens_seen": 1841840,
      "step": 2965
    },
    {
      "epoch": 5.294117647058823,
      "grad_norm": 0.06591796875,
      "learning_rate": 2.681063131692787e-05,
      "loss": 0.0796,
      "num_input_tokens_seen": 1844560,
      "step": 2970
    },
    {
      "epoch": 5.303030303030303,
      "grad_norm": 26.625,
      "learning_rate": 2.673304933982783e-05,
      "loss": 0.0756,
      "num_input_tokens_seen": 1848624,
      "step": 2975
    },
    {
      "epoch": 5.311942959001782,
      "grad_norm": 66.5,
      "learning_rate": 2.6655450588613806e-05,
      "loss": 0.2812,
      "num_input_tokens_seen": 1851952,
      "step": 2980
    },
    {
      "epoch": 5.320855614973262,
      "grad_norm": 0.380859375,
      "learning_rate": 2.657783581436097e-05,
      "loss": 0.1026,
      "num_input_tokens_seen": 1855696,
      "step": 2985
    },
    {
      "epoch": 5.329768270944742,
      "grad_norm": 20.125,
      "learning_rate": 2.6500205768299535e-05,
      "loss": 0.107,
      "num_input_tokens_seen": 1859408,
      "step": 2990
    },
    {
      "epoch": 5.338680926916221,
      "grad_norm": 0.78125,
      "learning_rate": 2.642256120180758e-05,
      "loss": 0.1455,
      "num_input_tokens_seen": 1861936,
      "step": 2995
    },
    {
      "epoch": 5.347593582887701,
      "grad_norm": 2.296875,
      "learning_rate": 2.6344902866403687e-05,
      "loss": 0.0684,
      "num_input_tokens_seen": 1864624,
      "step": 3000
    },
    {
      "epoch": 5.35650623885918,
      "grad_norm": 13.5625,
      "learning_rate": 2.6267231513739726e-05,
      "loss": 0.1061,
      "num_input_tokens_seen": 1867600,
      "step": 3005
    },
    {
      "epoch": 5.365418894830659,
      "grad_norm": 0.1015625,
      "learning_rate": 2.6189547895593562e-05,
      "loss": 0.0031,
      "num_input_tokens_seen": 1870672,
      "step": 3010
    },
    {
      "epoch": 5.374331550802139,
      "grad_norm": 0.0439453125,
      "learning_rate": 2.611185276386176e-05,
      "loss": 0.0391,
      "num_input_tokens_seen": 1874160,
      "step": 3015
    },
    {
      "epoch": 5.383244206773618,
      "grad_norm": 0.045654296875,
      "learning_rate": 2.6034146870552346e-05,
      "loss": 0.0856,
      "num_input_tokens_seen": 1877616,
      "step": 3020
    },
    {
      "epoch": 5.392156862745098,
      "grad_norm": 0.01116943359375,
      "learning_rate": 2.595643096777748e-05,
      "loss": 0.1954,
      "num_input_tokens_seen": 1880432,
      "step": 3025
    },
    {
      "epoch": 5.401069518716578,
      "grad_norm": 29.125,
      "learning_rate": 2.5878705807746245e-05,
      "loss": 0.0227,
      "num_input_tokens_seen": 1884528,
      "step": 3030
    },
    {
      "epoch": 5.409982174688057,
      "grad_norm": 24.125,
      "learning_rate": 2.580097214275727e-05,
      "loss": 0.1689,
      "num_input_tokens_seen": 1887152,
      "step": 3035
    },
    {
      "epoch": 5.418894830659537,
      "grad_norm": 3.390625,
      "learning_rate": 2.5723230725191554e-05,
      "loss": 0.0036,
      "num_input_tokens_seen": 1890032,
      "step": 3040
    },
    {
      "epoch": 5.427807486631016,
      "grad_norm": 0.0830078125,
      "learning_rate": 2.5645482307505108e-05,
      "loss": 0.0668,
      "num_input_tokens_seen": 1892304,
      "step": 3045
    },
    {
      "epoch": 5.436720142602495,
      "grad_norm": 0.41015625,
      "learning_rate": 2.55677276422217e-05,
      "loss": 0.1114,
      "num_input_tokens_seen": 1895728,
      "step": 3050
    },
    {
      "epoch": 5.445632798573975,
      "grad_norm": 15.8125,
      "learning_rate": 2.548996748192556e-05,
      "loss": 0.0155,
      "num_input_tokens_seen": 1898384,
      "step": 3055
    },
    {
      "epoch": 5.454545454545454,
      "grad_norm": 24.875,
      "learning_rate": 2.541220257925412e-05,
      "loss": 0.0471,
      "num_input_tokens_seen": 1901104,
      "step": 3060
    },
    {
      "epoch": 5.463458110516934,
      "grad_norm": 0.04443359375,
      "learning_rate": 2.5334433686890702e-05,
      "loss": 0.0239,
      "num_input_tokens_seen": 1904976,
      "step": 3065
    },
    {
      "epoch": 5.472370766488414,
      "grad_norm": 0.291015625,
      "learning_rate": 2.5256661557557247e-05,
      "loss": 0.0133,
      "num_input_tokens_seen": 1908688,
      "step": 3070
    },
    {
      "epoch": 5.481283422459893,
      "grad_norm": 0.05224609375,
      "learning_rate": 2.517888694400704e-05,
      "loss": 0.0521,
      "num_input_tokens_seen": 1911792,
      "step": 3075
    },
    {
      "epoch": 5.490196078431373,
      "grad_norm": 6.65625,
      "learning_rate": 2.5101110599017374e-05,
      "loss": 0.0028,
      "num_input_tokens_seen": 1915248,
      "step": 3080
    },
    {
      "epoch": 5.499108734402852,
      "grad_norm": 0.019775390625,
      "learning_rate": 2.502333327538235e-05,
      "loss": 0.0221,
      "num_input_tokens_seen": 1918544,
      "step": 3085
    },
    {
      "epoch": 5.508021390374331,
      "grad_norm": 18.0,
      "learning_rate": 2.4945555725905502e-05,
      "loss": 0.0712,
      "num_input_tokens_seen": 1922032,
      "step": 3090
    },
    {
      "epoch": 5.509803921568627,
      "eval_loss": 0.15239505469799042,
      "eval_runtime": 6.3085,
      "eval_samples_per_second": 39.471,
      "eval_steps_per_second": 9.987,
      "num_input_tokens_seen": 1922384,
      "step": 3091
    },
    {
      "epoch": 5.516934046345811,
      "grad_norm": 73.5,
      "learning_rate": 2.4867778703392554e-05,
      "loss": 0.1049,
      "num_input_tokens_seen": 1924400,
      "step": 3095
    },
    {
      "epoch": 5.52584670231729,
      "grad_norm": 0.0238037109375,
      "learning_rate": 2.479000296064417e-05,
      "loss": 0.0076,
      "num_input_tokens_seen": 1927376,
      "step": 3100
    },
    {
      "epoch": 5.53475935828877,
      "grad_norm": 0.035400390625,
      "learning_rate": 2.4712229250448567e-05,
      "loss": 0.04,
      "num_input_tokens_seen": 1930352,
      "step": 3105
    },
    {
      "epoch": 5.54367201426025,
      "grad_norm": 0.01470947265625,
      "learning_rate": 2.4634458325574323e-05,
      "loss": 0.0013,
      "num_input_tokens_seen": 1933680,
      "step": 3110
    },
    {
      "epoch": 5.552584670231729,
      "grad_norm": 0.11767578125,
      "learning_rate": 2.4556690938763062e-05,
      "loss": 0.0016,
      "num_input_tokens_seen": 1937488,
      "step": 3115
    },
    {
      "epoch": 5.561497326203209,
      "grad_norm": 0.267578125,
      "learning_rate": 2.4478927842722154e-05,
      "loss": 0.0034,
      "num_input_tokens_seen": 1940368,
      "step": 3120
    },
    {
      "epoch": 5.570409982174688,
      "grad_norm": 0.01055908203125,
      "learning_rate": 2.4401169790117427e-05,
      "loss": 0.022,
      "num_input_tokens_seen": 1943728,
      "step": 3125
    },
    {
      "epoch": 5.579322638146167,
      "grad_norm": 36.0,
      "learning_rate": 2.4323417533565916e-05,
      "loss": 0.1843,
      "num_input_tokens_seen": 1946832,
      "step": 3130
    },
    {
      "epoch": 5.588235294117647,
      "grad_norm": 0.0252685546875,
      "learning_rate": 2.424567182562854e-05,
      "loss": 0.0948,
      "num_input_tokens_seen": 1949904,
      "step": 3135
    },
    {
      "epoch": 5.597147950089127,
      "grad_norm": 20.25,
      "learning_rate": 2.4167933418802837e-05,
      "loss": 0.0667,
      "num_input_tokens_seen": 1952432,
      "step": 3140
    },
    {
      "epoch": 5.606060606060606,
      "grad_norm": 24.875,
      "learning_rate": 2.4090203065515695e-05,
      "loss": 0.0711,
      "num_input_tokens_seen": 1955216,
      "step": 3145
    },
    {
      "epoch": 5.614973262032086,
      "grad_norm": 0.053955078125,
      "learning_rate": 2.4012481518116022e-05,
      "loss": 0.0889,
      "num_input_tokens_seen": 1958096,
      "step": 3150
    },
    {
      "epoch": 5.623885918003565,
      "grad_norm": 6.78125,
      "learning_rate": 2.3934769528867513e-05,
      "loss": 0.0592,
      "num_input_tokens_seen": 1961456,
      "step": 3155
    },
    {
      "epoch": 5.632798573975045,
      "grad_norm": 0.03515625,
      "learning_rate": 2.385706784994135e-05,
      "loss": 0.1017,
      "num_input_tokens_seen": 1964272,
      "step": 3160
    },
    {
      "epoch": 5.641711229946524,
      "grad_norm": 0.044189453125,
      "learning_rate": 2.3779377233408923e-05,
      "loss": 0.0376,
      "num_input_tokens_seen": 1967120,
      "step": 3165
    },
    {
      "epoch": 5.650623885918003,
      "grad_norm": 0.058349609375,
      "learning_rate": 2.3701698431234528e-05,
      "loss": 0.1132,
      "num_input_tokens_seen": 1969872,
      "step": 3170
    },
    {
      "epoch": 5.659536541889483,
      "grad_norm": 16.5,
      "learning_rate": 2.362403219526815e-05,
      "loss": 0.1088,
      "num_input_tokens_seen": 1972944,
      "step": 3175
    },
    {
      "epoch": 5.668449197860962,
      "grad_norm": 30.25,
      "learning_rate": 2.3546379277238107e-05,
      "loss": 0.3807,
      "num_input_tokens_seen": 1975888,
      "step": 3180
    },
    {
      "epoch": 5.677361853832442,
      "grad_norm": 0.052978515625,
      "learning_rate": 2.3468740428743833e-05,
      "loss": 0.0442,
      "num_input_tokens_seen": 1979088,
      "step": 3185
    },
    {
      "epoch": 5.686274509803922,
      "grad_norm": 18.375,
      "learning_rate": 2.339111640124859e-05,
      "loss": 0.0184,
      "num_input_tokens_seen": 1981520,
      "step": 3190
    },
    {
      "epoch": 5.695187165775401,
      "grad_norm": 40.25,
      "learning_rate": 2.3313507946072172e-05,
      "loss": 0.0387,
      "num_input_tokens_seen": 1984880,
      "step": 3195
    },
    {
      "epoch": 5.704099821746881,
      "grad_norm": 0.546875,
      "learning_rate": 2.323591581438365e-05,
      "loss": 0.0526,
      "num_input_tokens_seen": 1987440,
      "step": 3200
    },
    {
      "epoch": 5.71301247771836,
      "grad_norm": 5.84375,
      "learning_rate": 2.3158340757194116e-05,
      "loss": 0.0058,
      "num_input_tokens_seen": 1990640,
      "step": 3205
    },
    {
      "epoch": 5.721925133689839,
      "grad_norm": 19.0,
      "learning_rate": 2.3080783525349388e-05,
      "loss": 0.0267,
      "num_input_tokens_seen": 1993808,
      "step": 3210
    },
    {
      "epoch": 5.730837789661319,
      "grad_norm": 21.0,
      "learning_rate": 2.3003244869522743e-05,
      "loss": 0.0237,
      "num_input_tokens_seen": 1996688,
      "step": 3215
    },
    {
      "epoch": 5.739750445632799,
      "grad_norm": 0.12060546875,
      "learning_rate": 2.2925725540207688e-05,
      "loss": 0.0987,
      "num_input_tokens_seen": 1999696,
      "step": 3220
    },
    {
      "epoch": 5.748663101604278,
      "grad_norm": 0.01055908203125,
      "learning_rate": 2.2848226287710645e-05,
      "loss": 0.0499,
      "num_input_tokens_seen": 2002032,
      "step": 3225
    },
    {
      "epoch": 5.757575757575758,
      "grad_norm": 18.0,
      "learning_rate": 2.277074786214372e-05,
      "loss": 0.1424,
      "num_input_tokens_seen": 2005584,
      "step": 3230
    },
    {
      "epoch": 5.766488413547237,
      "grad_norm": 0.1611328125,
      "learning_rate": 2.2693291013417453e-05,
      "loss": 0.097,
      "num_input_tokens_seen": 2008176,
      "step": 3235
    },
    {
      "epoch": 5.775401069518717,
      "grad_norm": 0.8203125,
      "learning_rate": 2.2615856491233513e-05,
      "loss": 0.0027,
      "num_input_tokens_seen": 2011376,
      "step": 3240
    },
    {
      "epoch": 5.784313725490196,
      "grad_norm": 0.171875,
      "learning_rate": 2.2538445045077488e-05,
      "loss": 0.0006,
      "num_input_tokens_seen": 2014224,
      "step": 3245
    },
    {
      "epoch": 5.793226381461675,
      "grad_norm": 7.78125,
      "learning_rate": 2.246105742421162e-05,
      "loss": 0.0911,
      "num_input_tokens_seen": 2016912,
      "step": 3250
    },
    {
      "epoch": 5.802139037433155,
      "grad_norm": 0.134765625,
      "learning_rate": 2.2383694377667543e-05,
      "loss": 0.1258,
      "num_input_tokens_seen": 2020048,
      "step": 3255
    },
    {
      "epoch": 5.811051693404634,
      "grad_norm": 0.0233154296875,
      "learning_rate": 2.2306356654239012e-05,
      "loss": 0.0003,
      "num_input_tokens_seen": 2023216,
      "step": 3260
    },
    {
      "epoch": 5.819964349376114,
      "grad_norm": 0.022216796875,
      "learning_rate": 2.222904500247473e-05,
      "loss": 0.0009,
      "num_input_tokens_seen": 2026928,
      "step": 3265
    },
    {
      "epoch": 5.828877005347594,
      "grad_norm": 6.03125,
      "learning_rate": 2.2151760170671004e-05,
      "loss": 0.004,
      "num_input_tokens_seen": 2029584,
      "step": 3270
    },
    {
      "epoch": 5.837789661319073,
      "grad_norm": 0.10009765625,
      "learning_rate": 2.207450290686458e-05,
      "loss": 0.0417,
      "num_input_tokens_seen": 2032720,
      "step": 3275
    },
    {
      "epoch": 5.846702317290553,
      "grad_norm": 18.0,
      "learning_rate": 2.1997273958825375e-05,
      "loss": 0.088,
      "num_input_tokens_seen": 2036176,
      "step": 3280
    },
    {
      "epoch": 5.855614973262032,
      "grad_norm": 6.84375,
      "learning_rate": 2.1920074074049225e-05,
      "loss": 0.1004,
      "num_input_tokens_seen": 2039632,
      "step": 3285
    },
    {
      "epoch": 5.864527629233511,
      "grad_norm": 17.0,
      "learning_rate": 2.1842903999750665e-05,
      "loss": 0.0648,
      "num_input_tokens_seen": 2043184,
      "step": 3290
    },
    {
      "epoch": 5.873440285204991,
      "grad_norm": 0.01495361328125,
      "learning_rate": 2.1765764482855715e-05,
      "loss": 0.0579,
      "num_input_tokens_seen": 2046416,
      "step": 3295
    },
    {
      "epoch": 5.882352941176471,
      "grad_norm": 0.048095703125,
      "learning_rate": 2.1688656269994612e-05,
      "loss": 0.001,
      "num_input_tokens_seen": 2049008,
      "step": 3300
    },
    {
      "epoch": 5.89126559714795,
      "grad_norm": 0.421875,
      "learning_rate": 2.1611580107494597e-05,
      "loss": 0.0239,
      "num_input_tokens_seen": 2052656,
      "step": 3305
    },
    {
      "epoch": 5.90017825311943,
      "grad_norm": 31.875,
      "learning_rate": 2.153453674137272e-05,
      "loss": 0.0937,
      "num_input_tokens_seen": 2055888,
      "step": 3310
    },
    {
      "epoch": 5.909090909090909,
      "grad_norm": 0.0247802734375,
      "learning_rate": 2.1457526917328588e-05,
      "loss": 0.1453,
      "num_input_tokens_seen": 2059056,
      "step": 3315
    },
    {
      "epoch": 5.918003565062389,
      "grad_norm": 0.0191650390625,
      "learning_rate": 2.1380551380737128e-05,
      "loss": 0.0477,
      "num_input_tokens_seen": 2062096,
      "step": 3320
    },
    {
      "epoch": 5.926916221033868,
      "grad_norm": 0.018798828125,
      "learning_rate": 2.130361087664145e-05,
      "loss": 0.0597,
      "num_input_tokens_seen": 2065168,
      "step": 3325
    },
    {
      "epoch": 5.935828877005347,
      "grad_norm": 0.73828125,
      "learning_rate": 2.122670614974555e-05,
      "loss": 0.0187,
      "num_input_tokens_seen": 2067856,
      "step": 3330
    },
    {
      "epoch": 5.944741532976827,
      "grad_norm": 15.25,
      "learning_rate": 2.1149837944407136e-05,
      "loss": 0.0907,
      "num_input_tokens_seen": 2071056,
      "step": 3335
    },
    {
      "epoch": 5.953654188948306,
      "grad_norm": 0.023681640625,
      "learning_rate": 2.107300700463045e-05,
      "loss": 0.0003,
      "num_input_tokens_seen": 2074192,
      "step": 3340
    },
    {
      "epoch": 5.962566844919786,
      "grad_norm": 0.01068115234375,
      "learning_rate": 2.0996214074059034e-05,
      "loss": 0.0006,
      "num_input_tokens_seen": 2077040,
      "step": 3345
    },
    {
      "epoch": 5.971479500891266,
      "grad_norm": 10.1875,
      "learning_rate": 2.0919459895968517e-05,
      "loss": 0.2074,
      "num_input_tokens_seen": 2079312,
      "step": 3350
    },
    {
      "epoch": 5.980392156862745,
      "grad_norm": 0.126953125,
      "learning_rate": 2.084274521325948e-05,
      "loss": 0.0026,
      "num_input_tokens_seen": 2082864,
      "step": 3355
    },
    {
      "epoch": 5.989304812834225,
      "grad_norm": 0.306640625,
      "learning_rate": 2.0766070768450206e-05,
      "loss": 0.0012,
      "num_input_tokens_seen": 2085872,
      "step": 3360
    },
    {
      "epoch": 5.998217468805704,
      "grad_norm": 0.392578125,
      "learning_rate": 2.0689437303669508e-05,
      "loss": 0.0005,
      "num_input_tokens_seen": 2088272,
      "step": 3365
    },
    {
      "epoch": 6.007130124777183,
      "grad_norm": 0.012451171875,
      "learning_rate": 2.0612845560649603e-05,
      "loss": 0.0046,
      "num_input_tokens_seen": 2091232,
      "step": 3370
    },
    {
      "epoch": 6.010695187165775,
      "eval_loss": 0.15147051215171814,
      "eval_runtime": 6.2854,
      "eval_samples_per_second": 39.616,
      "eval_steps_per_second": 10.023,
      "num_input_tokens_seen": 2092320,
      "step": 3372
    },
    {
      "epoch": 6.016042780748663,
      "grad_norm": 0.3203125,
      "learning_rate": 2.0536296280718825e-05,
      "loss": 0.021,
      "num_input_tokens_seen": 2093952,
      "step": 3375
    },
    {
      "epoch": 6.024955436720143,
      "grad_norm": 35.25,
      "learning_rate": 2.0459790204794545e-05,
      "loss": 0.0566,
      "num_input_tokens_seen": 2097728,
      "step": 3380
    },
    {
      "epoch": 6.033868092691622,
      "grad_norm": 0.09326171875,
      "learning_rate": 2.0383328073375955e-05,
      "loss": 0.0017,
      "num_input_tokens_seen": 2100736,
      "step": 3385
    },
    {
      "epoch": 6.042780748663102,
      "grad_norm": 0.0174560546875,
      "learning_rate": 2.0306910626536926e-05,
      "loss": 0.0363,
      "num_input_tokens_seen": 2104032,
      "step": 3390
    },
    {
      "epoch": 6.051693404634581,
      "grad_norm": 0.072265625,
      "learning_rate": 2.0230538603918787e-05,
      "loss": 0.0004,
      "num_input_tokens_seen": 2107264,
      "step": 3395
    },
    {
      "epoch": 6.0606060606060606,
      "grad_norm": 0.130859375,
      "learning_rate": 2.015421274472325e-05,
      "loss": 0.0012,
      "num_input_tokens_seen": 2110336,
      "step": 3400
    },
    {
      "epoch": 6.06951871657754,
      "grad_norm": 0.048828125,
      "learning_rate": 2.0077933787705204e-05,
      "loss": 0.0009,
      "num_input_tokens_seen": 2113248,
      "step": 3405
    },
    {
      "epoch": 6.078431372549019,
      "grad_norm": 0.09423828125,
      "learning_rate": 2.000170247116554e-05,
      "loss": 0.0291,
      "num_input_tokens_seen": 2116032,
      "step": 3410
    },
    {
      "epoch": 6.087344028520499,
      "grad_norm": 0.01220703125,
      "learning_rate": 1.9925519532944104e-05,
      "loss": 0.0975,
      "num_input_tokens_seen": 2118848,
      "step": 3415
    },
    {
      "epoch": 6.096256684491979,
      "grad_norm": 2.109375,
      "learning_rate": 1.9849385710412424e-05,
      "loss": 0.0013,
      "num_input_tokens_seen": 2122208,
      "step": 3420
    },
    {
      "epoch": 6.105169340463458,
      "grad_norm": 1.359375,
      "learning_rate": 1.977330174046667e-05,
      "loss": 0.0451,
      "num_input_tokens_seen": 2125248,
      "step": 3425
    },
    {
      "epoch": 6.114081996434938,
      "grad_norm": 3.796875,
      "learning_rate": 1.9697268359520506e-05,
      "loss": 0.0574,
      "num_input_tokens_seen": 2129248,
      "step": 3430
    },
    {
      "epoch": 6.122994652406417,
      "grad_norm": 0.06396484375,
      "learning_rate": 1.9621286303497915e-05,
      "loss": 0.0008,
      "num_input_tokens_seen": 2131904,
      "step": 3435
    },
    {
      "epoch": 6.1319073083778965,
      "grad_norm": 0.55859375,
      "learning_rate": 1.954535630782612e-05,
      "loss": 0.0006,
      "num_input_tokens_seen": 2135552,
      "step": 3440
    },
    {
      "epoch": 6.140819964349376,
      "grad_norm": 0.177734375,
      "learning_rate": 1.9469479107428463e-05,
      "loss": 0.0861,
      "num_input_tokens_seen": 2138688,
      "step": 3445
    },
    {
      "epoch": 6.149732620320855,
      "grad_norm": 29.0,
      "learning_rate": 1.9393655436717283e-05,
      "loss": 0.1345,
      "num_input_tokens_seen": 2141248,
      "step": 3450
    },
    {
      "epoch": 6.158645276292335,
      "grad_norm": 45.5,
      "learning_rate": 1.9317886029586778e-05,
      "loss": 0.0748,
      "num_input_tokens_seen": 2144768,
      "step": 3455
    },
    {
      "epoch": 6.167557932263815,
      "grad_norm": 0.0194091796875,
      "learning_rate": 1.9242171619405986e-05,
      "loss": 0.0015,
      "num_input_tokens_seen": 2147552,
      "step": 3460
    },
    {
      "epoch": 6.176470588235294,
      "grad_norm": 17.625,
      "learning_rate": 1.916651293901157e-05,
      "loss": 0.0288,
      "num_input_tokens_seen": 2151040,
      "step": 3465
    },
    {
      "epoch": 6.185383244206774,
      "grad_norm": 21.375,
      "learning_rate": 1.909091072070083e-05,
      "loss": 0.1793,
      "num_input_tokens_seen": 2155040,
      "step": 3470
    },
    {
      "epoch": 6.194295900178253,
      "grad_norm": 0.0810546875,
      "learning_rate": 1.9015365696224564e-05,
      "loss": 0.0771,
      "num_input_tokens_seen": 2157824,
      "step": 3475
    },
    {
      "epoch": 6.2032085561497325,
      "grad_norm": 0.0654296875,
      "learning_rate": 1.893987859677997e-05,
      "loss": 0.0006,
      "num_input_tokens_seen": 2160672,
      "step": 3480
    },
    {
      "epoch": 6.212121212121212,
      "grad_norm": 0.494140625,
      "learning_rate": 1.886445015300362e-05,
      "loss": 0.1115,
      "num_input_tokens_seen": 2163552,
      "step": 3485
    },
    {
      "epoch": 6.221033868092691,
      "grad_norm": 24.875,
      "learning_rate": 1.8789081094964347e-05,
      "loss": 0.0291,
      "num_input_tokens_seen": 2167456,
      "step": 3490
    },
    {
      "epoch": 6.229946524064171,
      "grad_norm": 0.022705078125,
      "learning_rate": 1.8713772152156205e-05,
      "loss": 0.0421,
      "num_input_tokens_seen": 2170560,
      "step": 3495
    },
    {
      "epoch": 6.238859180035651,
      "grad_norm": 37.5,
      "learning_rate": 1.863852405349135e-05,
      "loss": 0.0437,
      "num_input_tokens_seen": 2173152,
      "step": 3500
    },
    {
      "epoch": 6.24777183600713,
      "grad_norm": 22.75,
      "learning_rate": 1.856333752729311e-05,
      "loss": 0.076,
      "num_input_tokens_seen": 2175808,
      "step": 3505
    },
    {
      "epoch": 6.25668449197861,
      "grad_norm": 54.75,
      "learning_rate": 1.848821330128878e-05,
      "loss": 0.1015,
      "num_input_tokens_seen": 2178304,
      "step": 3510
    },
    {
      "epoch": 6.265597147950089,
      "grad_norm": 13.625,
      "learning_rate": 1.8413152102602687e-05,
      "loss": 0.0089,
      "num_input_tokens_seen": 2181312,
      "step": 3515
    },
    {
      "epoch": 6.2745098039215685,
      "grad_norm": 30.125,
      "learning_rate": 1.8338154657749128e-05,
      "loss": 0.008,
      "num_input_tokens_seen": 2184128,
      "step": 3520
    },
    {
      "epoch": 6.283422459893048,
      "grad_norm": 0.8671875,
      "learning_rate": 1.826322169262531e-05,
      "loss": 0.0022,
      "num_input_tokens_seen": 2187584,
      "step": 3525
    },
    {
      "epoch": 6.292335115864527,
      "grad_norm": 0.025634765625,
      "learning_rate": 1.818835393250434e-05,
      "loss": 0.0009,
      "num_input_tokens_seen": 2191168,
      "step": 3530
    },
    {
      "epoch": 6.301247771836007,
      "grad_norm": 0.06787109375,
      "learning_rate": 1.8113552102028236e-05,
      "loss": 0.0091,
      "num_input_tokens_seen": 2194880,
      "step": 3535
    },
    {
      "epoch": 6.310160427807487,
      "grad_norm": 0.2265625,
      "learning_rate": 1.803881692520087e-05,
      "loss": 0.0559,
      "num_input_tokens_seen": 2197184,
      "step": 3540
    },
    {
      "epoch": 6.319073083778966,
      "grad_norm": 0.212890625,
      "learning_rate": 1.796414912538095e-05,
      "loss": 0.0004,
      "num_input_tokens_seen": 2200160,
      "step": 3545
    },
    {
      "epoch": 6.327985739750446,
      "grad_norm": 0.01129150390625,
      "learning_rate": 1.7889549425275093e-05,
      "loss": 0.1829,
      "num_input_tokens_seen": 2203776,
      "step": 3550
    },
    {
      "epoch": 6.336898395721925,
      "grad_norm": 21.625,
      "learning_rate": 1.7815018546930754e-05,
      "loss": 0.176,
      "num_input_tokens_seen": 2207104,
      "step": 3555
    },
    {
      "epoch": 6.3458110516934045,
      "grad_norm": 0.052734375,
      "learning_rate": 1.7740557211729258e-05,
      "loss": 0.0602,
      "num_input_tokens_seen": 2210400,
      "step": 3560
    },
    {
      "epoch": 6.354723707664884,
      "grad_norm": 27.25,
      "learning_rate": 1.7666166140378852e-05,
      "loss": 0.1188,
      "num_input_tokens_seen": 2213728,
      "step": 3565
    },
    {
      "epoch": 6.363636363636363,
      "grad_norm": 37.0,
      "learning_rate": 1.7591846052907673e-05,
      "loss": 0.0346,
      "num_input_tokens_seen": 2216416,
      "step": 3570
    },
    {
      "epoch": 6.372549019607844,
      "grad_norm": 0.2099609375,
      "learning_rate": 1.7517597668656823e-05,
      "loss": 0.0174,
      "num_input_tokens_seen": 2219328,
      "step": 3575
    },
    {
      "epoch": 6.381461675579323,
      "grad_norm": 4.53125,
      "learning_rate": 1.7443421706273395e-05,
      "loss": 0.0271,
      "num_input_tokens_seen": 2222496,
      "step": 3580
    },
    {
      "epoch": 6.390374331550802,
      "grad_norm": 0.08154296875,
      "learning_rate": 1.7369318883703506e-05,
      "loss": 0.1009,
      "num_input_tokens_seen": 2225504,
      "step": 3585
    },
    {
      "epoch": 6.399286987522282,
      "grad_norm": 15.75,
      "learning_rate": 1.7295289918185348e-05,
      "loss": 0.1496,
      "num_input_tokens_seen": 2229312,
      "step": 3590
    },
    {
      "epoch": 6.408199643493761,
      "grad_norm": 0.099609375,
      "learning_rate": 1.722133552624227e-05,
      "loss": 0.0354,
      "num_input_tokens_seen": 2232544,
      "step": 3595
    },
    {
      "epoch": 6.4171122994652405,
      "grad_norm": 0.0927734375,
      "learning_rate": 1.714745642367583e-05,
      "loss": 0.0023,
      "num_input_tokens_seen": 2235808,
      "step": 3600
    },
    {
      "epoch": 6.42602495543672,
      "grad_norm": 0.0322265625,
      "learning_rate": 1.707365332555883e-05,
      "loss": 0.1309,
      "num_input_tokens_seen": 2239040,
      "step": 3605
    },
    {
      "epoch": 6.434937611408199,
      "grad_norm": 0.068359375,
      "learning_rate": 1.699992694622847e-05,
      "loss": 0.1223,
      "num_input_tokens_seen": 2241728,
      "step": 3610
    },
    {
      "epoch": 6.443850267379679,
      "grad_norm": 39.0,
      "learning_rate": 1.6926277999279372e-05,
      "loss": 0.1168,
      "num_input_tokens_seen": 2244928,
      "step": 3615
    },
    {
      "epoch": 6.452762923351159,
      "grad_norm": 0.05908203125,
      "learning_rate": 1.6852707197556677e-05,
      "loss": 0.0013,
      "num_input_tokens_seen": 2247936,
      "step": 3620
    },
    {
      "epoch": 6.461675579322638,
      "grad_norm": 16.375,
      "learning_rate": 1.67792152531492e-05,
      "loss": 0.1828,
      "num_input_tokens_seen": 2250560,
      "step": 3625
    },
    {
      "epoch": 6.470588235294118,
      "grad_norm": 0.9140625,
      "learning_rate": 1.6705802877382464e-05,
      "loss": 0.021,
      "num_input_tokens_seen": 2253248,
      "step": 3630
    },
    {
      "epoch": 6.479500891265597,
      "grad_norm": 1.5,
      "learning_rate": 1.6632470780811866e-05,
      "loss": 0.0776,
      "num_input_tokens_seen": 2256320,
      "step": 3635
    },
    {
      "epoch": 6.4884135472370765,
      "grad_norm": 8.125,
      "learning_rate": 1.6559219673215784e-05,
      "loss": 0.0297,
      "num_input_tokens_seen": 2259168,
      "step": 3640
    },
    {
      "epoch": 6.497326203208556,
      "grad_norm": 27.375,
      "learning_rate": 1.6486050263588702e-05,
      "loss": 0.0353,
      "num_input_tokens_seen": 2262240,
      "step": 3645
    },
    {
      "epoch": 6.506238859180035,
      "grad_norm": 23.375,
      "learning_rate": 1.641296326013436e-05,
      "loss": 0.0674,
      "num_input_tokens_seen": 2265600,
      "step": 3650
    },
    {
      "epoch": 6.5115864527629235,
      "eval_loss": 0.15169650316238403,
      "eval_runtime": 6.2908,
      "eval_samples_per_second": 39.582,
      "eval_steps_per_second": 10.015,
      "num_input_tokens_seen": 2267520,
      "step": 3653
    },
    {
      "epoch": 6.515151515151516,
      "grad_norm": 8.1875,
      "learning_rate": 1.633995937025889e-05,
      "loss": 0.0634,
      "num_input_tokens_seen": 2268768,
      "step": 3655
    },
    {
      "epoch": 6.524064171122995,
      "grad_norm": 2.015625,
      "learning_rate": 1.6267039300563965e-05,
      "loss": 0.0202,
      "num_input_tokens_seen": 2272256,
      "step": 3660
    },
    {
      "epoch": 6.532976827094474,
      "grad_norm": 1.4375,
      "learning_rate": 1.619420375683996e-05,
      "loss": 0.0034,
      "num_input_tokens_seen": 2275968,
      "step": 3665
    },
    {
      "epoch": 6.541889483065954,
      "grad_norm": 2.671875,
      "learning_rate": 1.6121453444059153e-05,
      "loss": 0.008,
      "num_input_tokens_seen": 2278784,
      "step": 3670
    },
    {
      "epoch": 6.550802139037433,
      "grad_norm": 38.25,
      "learning_rate": 1.6048789066368858e-05,
      "loss": 0.0209,
      "num_input_tokens_seen": 2281472,
      "step": 3675
    },
    {
      "epoch": 6.5597147950089125,
      "grad_norm": 0.224609375,
      "learning_rate": 1.5976211327084606e-05,
      "loss": 0.0158,
      "num_input_tokens_seen": 2284608,
      "step": 3680
    },
    {
      "epoch": 6.568627450980392,
      "grad_norm": 0.0546875,
      "learning_rate": 1.59037209286834e-05,
      "loss": 0.0469,
      "num_input_tokens_seen": 2287296,
      "step": 3685
    },
    {
      "epoch": 6.577540106951871,
      "grad_norm": 7.90625,
      "learning_rate": 1.583131857279685e-05,
      "loss": 0.0839,
      "num_input_tokens_seen": 2290176,
      "step": 3690
    },
    {
      "epoch": 6.586452762923351,
      "grad_norm": 18.375,
      "learning_rate": 1.57590049602044e-05,
      "loss": 0.0791,
      "num_input_tokens_seen": 2292960,
      "step": 3695
    },
    {
      "epoch": 6.595365418894831,
      "grad_norm": 0.30078125,
      "learning_rate": 1.5686780790826574e-05,
      "loss": 0.0513,
      "num_input_tokens_seen": 2296192,
      "step": 3700
    },
    {
      "epoch": 6.60427807486631,
      "grad_norm": 0.017333984375,
      "learning_rate": 1.561464676371816e-05,
      "loss": 0.0003,
      "num_input_tokens_seen": 2300224,
      "step": 3705
    },
    {
      "epoch": 6.61319073083779,
      "grad_norm": 27.125,
      "learning_rate": 1.5542603577061464e-05,
      "loss": 0.0658,
      "num_input_tokens_seen": 2303040,
      "step": 3710
    },
    {
      "epoch": 6.622103386809269,
      "grad_norm": 20.625,
      "learning_rate": 1.5470651928159564e-05,
      "loss": 0.01,
      "num_input_tokens_seen": 2305600,
      "step": 3715
    },
    {
      "epoch": 6.6310160427807485,
      "grad_norm": 24.5,
      "learning_rate": 1.539879251342954e-05,
      "loss": 0.0104,
      "num_input_tokens_seen": 2308736,
      "step": 3720
    },
    {
      "epoch": 6.639928698752228,
      "grad_norm": 0.01904296875,
      "learning_rate": 1.5327026028395724e-05,
      "loss": 0.0303,
      "num_input_tokens_seen": 2311840,
      "step": 3725
    },
    {
      "epoch": 6.648841354723707,
      "grad_norm": 0.76953125,
      "learning_rate": 1.5255353167683017e-05,
      "loss": 0.0151,
      "num_input_tokens_seen": 2315808,
      "step": 3730
    },
    {
      "epoch": 6.657754010695188,
      "grad_norm": 16.25,
      "learning_rate": 1.5183774625010119e-05,
      "loss": 0.0755,
      "num_input_tokens_seen": 2319072,
      "step": 3735
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 44.5,
      "learning_rate": 1.5112291093182818e-05,
      "loss": 0.1207,
      "num_input_tokens_seen": 2323104,
      "step": 3740
    },
    {
      "epoch": 6.675579322638146,
      "grad_norm": 0.1904296875,
      "learning_rate": 1.5040903264087328e-05,
      "loss": 0.0103,
      "num_input_tokens_seen": 2325984,
      "step": 3745
    },
    {
      "epoch": 6.684491978609626,
      "grad_norm": 1.765625,
      "learning_rate": 1.4969611828683517e-05,
      "loss": 0.0045,
      "num_input_tokens_seen": 2329152,
      "step": 3750
    },
    {
      "epoch": 6.693404634581105,
      "grad_norm": 20.125,
      "learning_rate": 1.4898417476998289e-05,
      "loss": 0.0771,
      "num_input_tokens_seen": 2332768,
      "step": 3755
    },
    {
      "epoch": 6.7023172905525845,
      "grad_norm": 0.1435546875,
      "learning_rate": 1.4827320898118884e-05,
      "loss": 0.0004,
      "num_input_tokens_seen": 2335680,
      "step": 3760
    },
    {
      "epoch": 6.711229946524064,
      "grad_norm": 0.44140625,
      "learning_rate": 1.4756322780186193e-05,
      "loss": 0.1187,
      "num_input_tokens_seen": 2338656,
      "step": 3765
    },
    {
      "epoch": 6.720142602495543,
      "grad_norm": 0.0167236328125,
      "learning_rate": 1.4685423810388094e-05,
      "loss": 0.0343,
      "num_input_tokens_seen": 2342016,
      "step": 3770
    },
    {
      "epoch": 6.729055258467023,
      "grad_norm": 53.75,
      "learning_rate": 1.4614624674952842e-05,
      "loss": 0.0842,
      "num_input_tokens_seen": 2345120,
      "step": 3775
    },
    {
      "epoch": 6.737967914438503,
      "grad_norm": 0.0400390625,
      "learning_rate": 1.4543926059142379e-05,
      "loss": 0.081,
      "num_input_tokens_seen": 2348512,
      "step": 3780
    },
    {
      "epoch": 6.746880570409982,
      "grad_norm": 0.013916015625,
      "learning_rate": 1.4473328647245726e-05,
      "loss": 0.1136,
      "num_input_tokens_seen": 2350688,
      "step": 3785
    },
    {
      "epoch": 6.755793226381462,
      "grad_norm": 32.25,
      "learning_rate": 1.4402833122572368e-05,
      "loss": 0.0185,
      "num_input_tokens_seen": 2353504,
      "step": 3790
    },
    {
      "epoch": 6.764705882352941,
      "grad_norm": 71.0,
      "learning_rate": 1.4332440167445613e-05,
      "loss": 0.0589,
      "num_input_tokens_seen": 2356672,
      "step": 3795
    },
    {
      "epoch": 6.7736185383244205,
      "grad_norm": 0.0257568359375,
      "learning_rate": 1.4262150463195981e-05,
      "loss": 0.0831,
      "num_input_tokens_seen": 2360288,
      "step": 3800
    },
    {
      "epoch": 6.7825311942959,
      "grad_norm": 0.01312255859375,
      "learning_rate": 1.4191964690154702e-05,
      "loss": 0.0163,
      "num_input_tokens_seen": 2362944,
      "step": 3805
    },
    {
      "epoch": 6.791443850267379,
      "grad_norm": 0.0184326171875,
      "learning_rate": 1.412188352764699e-05,
      "loss": 0.0017,
      "num_input_tokens_seen": 2366080,
      "step": 3810
    },
    {
      "epoch": 6.80035650623886,
      "grad_norm": 2.0625,
      "learning_rate": 1.4051907653985552e-05,
      "loss": 0.1283,
      "num_input_tokens_seen": 2369632,
      "step": 3815
    },
    {
      "epoch": 6.809269162210339,
      "grad_norm": 57.75,
      "learning_rate": 1.3982037746464043e-05,
      "loss": 0.1444,
      "num_input_tokens_seen": 2373504,
      "step": 3820
    },
    {
      "epoch": 6.818181818181818,
      "grad_norm": 5.8125,
      "learning_rate": 1.3912274481350433e-05,
      "loss": 0.0177,
      "num_input_tokens_seen": 2376480,
      "step": 3825
    },
    {
      "epoch": 6.827094474153298,
      "grad_norm": 47.25,
      "learning_rate": 1.3842618533880531e-05,
      "loss": 0.0341,
      "num_input_tokens_seen": 2379488,
      "step": 3830
    },
    {
      "epoch": 6.836007130124777,
      "grad_norm": 49.0,
      "learning_rate": 1.3773070578251424e-05,
      "loss": 0.1742,
      "num_input_tokens_seen": 2382496,
      "step": 3835
    },
    {
      "epoch": 6.8449197860962565,
      "grad_norm": 33.25,
      "learning_rate": 1.3703631287614935e-05,
      "loss": 0.0996,
      "num_input_tokens_seen": 2386304,
      "step": 3840
    },
    {
      "epoch": 6.853832442067736,
      "grad_norm": 28.625,
      "learning_rate": 1.363430133407112e-05,
      "loss": 0.0696,
      "num_input_tokens_seen": 2389504,
      "step": 3845
    },
    {
      "epoch": 6.862745098039216,
      "grad_norm": 1.6875,
      "learning_rate": 1.3565081388661782e-05,
      "loss": 0.0125,
      "num_input_tokens_seen": 2392320,
      "step": 3850
    },
    {
      "epoch": 6.871657754010696,
      "grad_norm": 0.119140625,
      "learning_rate": 1.3495972121363968e-05,
      "loss": 0.1099,
      "num_input_tokens_seen": 2395648,
      "step": 3855
    },
    {
      "epoch": 6.880570409982175,
      "grad_norm": 42.5,
      "learning_rate": 1.3426974201083439e-05,
      "loss": 0.0357,
      "num_input_tokens_seen": 2398080,
      "step": 3860
    },
    {
      "epoch": 6.889483065953654,
      "grad_norm": 0.25390625,
      "learning_rate": 1.3358088295648274e-05,
      "loss": 0.0005,
      "num_input_tokens_seen": 2400448,
      "step": 3865
    },
    {
      "epoch": 6.898395721925134,
      "grad_norm": 0.0400390625,
      "learning_rate": 1.328931507180233e-05,
      "loss": 0.0008,
      "num_input_tokens_seen": 2403424,
      "step": 3870
    },
    {
      "epoch": 6.907308377896613,
      "grad_norm": 1.296875,
      "learning_rate": 1.3220655195198847e-05,
      "loss": 0.0087,
      "num_input_tokens_seen": 2405984,
      "step": 3875
    },
    {
      "epoch": 6.9162210338680925,
      "grad_norm": 36.75,
      "learning_rate": 1.3152109330393985e-05,
      "loss": 0.0832,
      "num_input_tokens_seen": 2409472,
      "step": 3880
    },
    {
      "epoch": 6.925133689839572,
      "grad_norm": 0.04931640625,
      "learning_rate": 1.3083678140840366e-05,
      "loss": 0.0044,
      "num_input_tokens_seen": 2412384,
      "step": 3885
    },
    {
      "epoch": 6.934046345811051,
      "grad_norm": 0.75,
      "learning_rate": 1.3015362288880678e-05,
      "loss": 0.0957,
      "num_input_tokens_seen": 2415328,
      "step": 3890
    },
    {
      "epoch": 6.942959001782532,
      "grad_norm": 0.011962890625,
      "learning_rate": 1.2947162435741278e-05,
      "loss": 0.0202,
      "num_input_tokens_seen": 2418848,
      "step": 3895
    },
    {
      "epoch": 6.951871657754011,
      "grad_norm": 0.47265625,
      "learning_rate": 1.2879079241525783e-05,
      "loss": 0.2008,
      "num_input_tokens_seen": 2421824,
      "step": 3900
    },
    {
      "epoch": 6.96078431372549,
      "grad_norm": 12.25,
      "learning_rate": 1.2811113365208627e-05,
      "loss": 0.2242,
      "num_input_tokens_seen": 2424224,
      "step": 3905
    },
    {
      "epoch": 6.96969696969697,
      "grad_norm": 0.51953125,
      "learning_rate": 1.2743265464628786e-05,
      "loss": 0.0045,
      "num_input_tokens_seen": 2427616,
      "step": 3910
    },
    {
      "epoch": 6.978609625668449,
      "grad_norm": 0.12060546875,
      "learning_rate": 1.2675536196483306e-05,
      "loss": 0.0024,
      "num_input_tokens_seen": 2430368,
      "step": 3915
    },
    {
      "epoch": 6.9875222816399285,
      "grad_norm": 0.0308837890625,
      "learning_rate": 1.260792621632102e-05,
      "loss": 0.0026,
      "num_input_tokens_seen": 2433376,
      "step": 3920
    },
    {
      "epoch": 6.996434937611408,
      "grad_norm": 0.013671875,
      "learning_rate": 1.2540436178536186e-05,
      "loss": 0.003,
      "num_input_tokens_seen": 2436608,
      "step": 3925
    },
    {
      "epoch": 7.005347593582887,
      "grad_norm": 0.1728515625,
      "learning_rate": 1.2473066736362124e-05,
      "loss": 0.0127,
      "num_input_tokens_seen": 2439064,
      "step": 3930
    },
    {
      "epoch": 7.0124777183600715,
      "eval_loss": 0.15626700222492218,
      "eval_runtime": 6.2821,
      "eval_samples_per_second": 39.637,
      "eval_steps_per_second": 10.029,
      "num_input_tokens_seen": 2441688,
      "step": 3934
    },
    {
      "epoch": 7.0142602495543676,
      "grad_norm": 0.021728515625,
      "learning_rate": 1.2405818541864905e-05,
      "loss": 0.0017,
      "num_input_tokens_seen": 2442328,
      "step": 3935
    },
    {
      "epoch": 7.023172905525847,
      "grad_norm": 0.06640625,
      "learning_rate": 1.2338692245937077e-05,
      "loss": 0.1579,
      "num_input_tokens_seen": 2445272,
      "step": 3940
    },
    {
      "epoch": 7.032085561497326,
      "grad_norm": 0.031005859375,
      "learning_rate": 1.2271688498291335e-05,
      "loss": 0.0009,
      "num_input_tokens_seen": 2448216,
      "step": 3945
    },
    {
      "epoch": 7.040998217468806,
      "grad_norm": 33.75,
      "learning_rate": 1.2204807947454203e-05,
      "loss": 0.0329,
      "num_input_tokens_seen": 2451704,
      "step": 3950
    },
    {
      "epoch": 7.049910873440285,
      "grad_norm": 0.01348876953125,
      "learning_rate": 1.2138051240759826e-05,
      "loss": 0.0814,
      "num_input_tokens_seen": 2454392,
      "step": 3955
    },
    {
      "epoch": 7.0588235294117645,
      "grad_norm": 0.86328125,
      "learning_rate": 1.2071419024343633e-05,
      "loss": 0.0202,
      "num_input_tokens_seen": 2457112,
      "step": 3960
    },
    {
      "epoch": 7.067736185383244,
      "grad_norm": 4.46875,
      "learning_rate": 1.2004911943136143e-05,
      "loss": 0.0494,
      "num_input_tokens_seen": 2460312,
      "step": 3965
    },
    {
      "epoch": 7.076648841354723,
      "grad_norm": 0.0125732421875,
      "learning_rate": 1.1938530640856696e-05,
      "loss": 0.0192,
      "num_input_tokens_seen": 2463224,
      "step": 3970
    },
    {
      "epoch": 7.0855614973262036,
      "grad_norm": 0.9609375,
      "learning_rate": 1.1872275760007198e-05,
      "loss": 0.0011,
      "num_input_tokens_seen": 2466008,
      "step": 3975
    },
    {
      "epoch": 7.094474153297683,
      "grad_norm": 0.640625,
      "learning_rate": 1.1806147941865938e-05,
      "loss": 0.0015,
      "num_input_tokens_seen": 2469176,
      "step": 3980
    },
    {
      "epoch": 7.103386809269162,
      "grad_norm": 0.01055908203125,
      "learning_rate": 1.1740147826481385e-05,
      "loss": 0.1977,
      "num_input_tokens_seen": 2472408,
      "step": 3985
    },
    {
      "epoch": 7.112299465240642,
      "grad_norm": 0.177734375,
      "learning_rate": 1.1674276052665973e-05,
      "loss": 0.0318,
      "num_input_tokens_seen": 2475608,
      "step": 3990
    },
    {
      "epoch": 7.121212121212121,
      "grad_norm": 0.01953125,
      "learning_rate": 1.1608533257989901e-05,
      "loss": 0.0146,
      "num_input_tokens_seen": 2478680,
      "step": 3995
    },
    {
      "epoch": 7.1301247771836005,
      "grad_norm": 42.25,
      "learning_rate": 1.1542920078775018e-05,
      "loss": 0.1046,
      "num_input_tokens_seen": 2481592,
      "step": 4000
    },
    {
      "epoch": 7.13903743315508,
      "grad_norm": 32.0,
      "learning_rate": 1.14774371500886e-05,
      "loss": 0.0188,
      "num_input_tokens_seen": 2485176,
      "step": 4005
    },
    {
      "epoch": 7.14795008912656,
      "grad_norm": 0.09765625,
      "learning_rate": 1.141208510573725e-05,
      "loss": 0.0005,
      "num_input_tokens_seen": 2488152,
      "step": 4010
    },
    {
      "epoch": 7.1568627450980395,
      "grad_norm": 10.3125,
      "learning_rate": 1.1346864578260758e-05,
      "loss": 0.0989,
      "num_input_tokens_seen": 2491320,
      "step": 4015
    },
    {
      "epoch": 7.165775401069519,
      "grad_norm": 0.0771484375,
      "learning_rate": 1.1281776198925939e-05,
      "loss": 0.0352,
      "num_input_tokens_seen": 2493944,
      "step": 4020
    },
    {
      "epoch": 7.174688057040998,
      "grad_norm": 0.07373046875,
      "learning_rate": 1.121682059772056e-05,
      "loss": 0.001,
      "num_input_tokens_seen": 2496664,
      "step": 4025
    },
    {
      "epoch": 7.183600713012478,
      "grad_norm": 0.15234375,
      "learning_rate": 1.1151998403347244e-05,
      "loss": 0.0003,
      "num_input_tokens_seen": 2500216,
      "step": 4030
    },
    {
      "epoch": 7.192513368983957,
      "grad_norm": 0.5546875,
      "learning_rate": 1.1087310243217386e-05,
      "loss": 0.0176,
      "num_input_tokens_seen": 2503544,
      "step": 4035
    },
    {
      "epoch": 7.2014260249554365,
      "grad_norm": 0.1025390625,
      "learning_rate": 1.1022756743445028e-05,
      "loss": 0.0753,
      "num_input_tokens_seen": 2507160,
      "step": 4040
    },
    {
      "epoch": 7.210338680926916,
      "grad_norm": 25.625,
      "learning_rate": 1.0958338528840893e-05,
      "loss": 0.1054,
      "num_input_tokens_seen": 2510232,
      "step": 4045
    },
    {
      "epoch": 7.219251336898395,
      "grad_norm": 16.5,
      "learning_rate": 1.0894056222906226e-05,
      "loss": 0.0075,
      "num_input_tokens_seen": 2513144,
      "step": 4050
    },
    {
      "epoch": 7.2281639928698755,
      "grad_norm": 58.5,
      "learning_rate": 1.0829910447826868e-05,
      "loss": 0.0433,
      "num_input_tokens_seen": 2516504,
      "step": 4055
    },
    {
      "epoch": 7.237076648841355,
      "grad_norm": 0.345703125,
      "learning_rate": 1.0765901824467167e-05,
      "loss": 0.0225,
      "num_input_tokens_seen": 2518648,
      "step": 4060
    },
    {
      "epoch": 7.245989304812834,
      "grad_norm": 0.0167236328125,
      "learning_rate": 1.0702030972363963e-05,
      "loss": 0.0218,
      "num_input_tokens_seen": 2521880,
      "step": 4065
    },
    {
      "epoch": 7.254901960784314,
      "grad_norm": 0.0230712890625,
      "learning_rate": 1.063829850972065e-05,
      "loss": 0.0101,
      "num_input_tokens_seen": 2525336,
      "step": 4070
    },
    {
      "epoch": 7.263814616755793,
      "grad_norm": 35.25,
      "learning_rate": 1.0574705053401127e-05,
      "loss": 0.1253,
      "num_input_tokens_seen": 2528184,
      "step": 4075
    },
    {
      "epoch": 7.2727272727272725,
      "grad_norm": 11.125,
      "learning_rate": 1.0511251218923868e-05,
      "loss": 0.0922,
      "num_input_tokens_seen": 2530904,
      "step": 4080
    },
    {
      "epoch": 7.281639928698752,
      "grad_norm": 20.0,
      "learning_rate": 1.0447937620455964e-05,
      "loss": 0.0206,
      "num_input_tokens_seen": 2533656,
      "step": 4085
    },
    {
      "epoch": 7.290552584670232,
      "grad_norm": 0.486328125,
      "learning_rate": 1.0384764870807149e-05,
      "loss": 0.082,
      "num_input_tokens_seen": 2535928,
      "step": 4090
    },
    {
      "epoch": 7.2994652406417115,
      "grad_norm": 0.19140625,
      "learning_rate": 1.0321733581423884e-05,
      "loss": 0.0186,
      "num_input_tokens_seen": 2539352,
      "step": 4095
    },
    {
      "epoch": 7.308377896613191,
      "grad_norm": 23.0,
      "learning_rate": 1.025884436238346e-05,
      "loss": 0.1287,
      "num_input_tokens_seen": 2542456,
      "step": 4100
    },
    {
      "epoch": 7.31729055258467,
      "grad_norm": 0.01953125,
      "learning_rate": 1.0196097822388075e-05,
      "loss": 0.0221,
      "num_input_tokens_seen": 2545816,
      "step": 4105
    },
    {
      "epoch": 7.32620320855615,
      "grad_norm": 0.0390625,
      "learning_rate": 1.013349456875892e-05,
      "loss": 0.0755,
      "num_input_tokens_seen": 2548824,
      "step": 4110
    },
    {
      "epoch": 7.335115864527629,
      "grad_norm": 0.52734375,
      "learning_rate": 1.0071035207430352e-05,
      "loss": 0.0006,
      "num_input_tokens_seen": 2552152,
      "step": 4115
    },
    {
      "epoch": 7.3440285204991085,
      "grad_norm": 0.01708984375,
      "learning_rate": 1.0008720342943966e-05,
      "loss": 0.0005,
      "num_input_tokens_seen": 2555768,
      "step": 4120
    },
    {
      "epoch": 7.352941176470588,
      "grad_norm": 0.259765625,
      "learning_rate": 9.94655057844281e-06,
      "loss": 0.0011,
      "num_input_tokens_seen": 2558328,
      "step": 4125
    },
    {
      "epoch": 7.361853832442068,
      "grad_norm": 2.90625,
      "learning_rate": 9.884526515665508e-06,
      "loss": 0.0646,
      "num_input_tokens_seen": 2561368,
      "step": 4130
    },
    {
      "epoch": 7.3707664884135475,
      "grad_norm": 0.06298828125,
      "learning_rate": 9.822648754940431e-06,
      "loss": 0.0886,
      "num_input_tokens_seen": 2564056,
      "step": 4135
    },
    {
      "epoch": 7.379679144385027,
      "grad_norm": 0.6015625,
      "learning_rate": 9.760917895179894e-06,
      "loss": 0.0008,
      "num_input_tokens_seen": 2566744,
      "step": 4140
    },
    {
      "epoch": 7.388591800356506,
      "grad_norm": 0.228515625,
      "learning_rate": 9.699334533874386e-06,
      "loss": 0.0011,
      "num_input_tokens_seen": 2569656,
      "step": 4145
    },
    {
      "epoch": 7.397504456327986,
      "grad_norm": 26.625,
      "learning_rate": 9.637899267086758e-06,
      "loss": 0.0823,
      "num_input_tokens_seen": 2573112,
      "step": 4150
    },
    {
      "epoch": 7.406417112299465,
      "grad_norm": 0.96484375,
      "learning_rate": 9.576612689446444e-06,
      "loss": 0.123,
      "num_input_tokens_seen": 2576952,
      "step": 4155
    },
    {
      "epoch": 7.4153297682709445,
      "grad_norm": 0.01446533203125,
      "learning_rate": 9.515475394143742e-06,
      "loss": 0.0006,
      "num_input_tokens_seen": 2579896,
      "step": 4160
    },
    {
      "epoch": 7.424242424242424,
      "grad_norm": 31.625,
      "learning_rate": 9.45448797292403e-06,
      "loss": 0.0705,
      "num_input_tokens_seen": 2583544,
      "step": 4165
    },
    {
      "epoch": 7.433155080213904,
      "grad_norm": 41.25,
      "learning_rate": 9.393651016082083e-06,
      "loss": 0.1237,
      "num_input_tokens_seen": 2586200,
      "step": 4170
    },
    {
      "epoch": 7.4420677361853835,
      "grad_norm": 49.0,
      "learning_rate": 9.332965112456337e-06,
      "loss": 0.0532,
      "num_input_tokens_seen": 2589496,
      "step": 4175
    },
    {
      "epoch": 7.450980392156863,
      "grad_norm": 0.1904296875,
      "learning_rate": 9.272430849423174e-06,
      "loss": 0.0375,
      "num_input_tokens_seen": 2591928,
      "step": 4180
    },
    {
      "epoch": 7.459893048128342,
      "grad_norm": 43.0,
      "learning_rate": 9.21204881289125e-06,
      "loss": 0.0419,
      "num_input_tokens_seen": 2595064,
      "step": 4185
    },
    {
      "epoch": 7.468805704099822,
      "grad_norm": 51.25,
      "learning_rate": 9.151819587295845e-06,
      "loss": 0.0159,
      "num_input_tokens_seen": 2597944,
      "step": 4190
    },
    {
      "epoch": 7.477718360071301,
      "grad_norm": 0.314453125,
      "learning_rate": 9.09174375559319e-06,
      "loss": 0.0023,
      "num_input_tokens_seen": 2601656,
      "step": 4195
    },
    {
      "epoch": 7.4866310160427805,
      "grad_norm": 20.875,
      "learning_rate": 9.031821899254796e-06,
      "loss": 0.1474,
      "num_input_tokens_seen": 2604472,
      "step": 4200
    },
    {
      "epoch": 7.49554367201426,
      "grad_norm": 50.5,
      "learning_rate": 8.972054598261892e-06,
      "loss": 0.1761,
      "num_input_tokens_seen": 2607992,
      "step": 4205
    },
    {
      "epoch": 7.50445632798574,
      "grad_norm": 0.0517578125,
      "learning_rate": 8.912442431099724e-06,
      "loss": 0.1577,
      "num_input_tokens_seen": 2611800,
      "step": 4210
    },
    {
      "epoch": 7.5133689839572195,
      "grad_norm": 0.0252685546875,
      "learning_rate": 8.852985974752045e-06,
      "loss": 0.0054,
      "num_input_tokens_seen": 2614936,
      "step": 4215
    },
    {
      "epoch": 7.5133689839572195,
      "eval_loss": 0.1573449969291687,
      "eval_runtime": 6.2868,
      "eval_samples_per_second": 39.607,
      "eval_steps_per_second": 10.021,
      "num_input_tokens_seen": 2614936,
      "step": 4215
    },
    {
      "epoch": 7.522281639928699,
      "grad_norm": 0.25,
      "learning_rate": 8.793685804695482e-06,
      "loss": 0.0687,
      "num_input_tokens_seen": 2618744,
      "step": 4220
    },
    {
      "epoch": 7.531194295900178,
      "grad_norm": 0.0269775390625,
      "learning_rate": 8.734542494893955e-06,
      "loss": 0.1056,
      "num_input_tokens_seen": 2621496,
      "step": 4225
    },
    {
      "epoch": 7.540106951871658,
      "grad_norm": 0.083984375,
      "learning_rate": 8.675556617793143e-06,
      "loss": 0.0174,
      "num_input_tokens_seen": 2624568,
      "step": 4230
    },
    {
      "epoch": 7.549019607843137,
      "grad_norm": 0.337890625,
      "learning_rate": 8.616728744314956e-06,
      "loss": 0.0416,
      "num_input_tokens_seen": 2627832,
      "step": 4235
    },
    {
      "epoch": 7.5579322638146165,
      "grad_norm": 0.057373046875,
      "learning_rate": 8.558059443851998e-06,
      "loss": 0.0847,
      "num_input_tokens_seen": 2631160,
      "step": 4240
    },
    {
      "epoch": 7.566844919786096,
      "grad_norm": 19.25,
      "learning_rate": 8.499549284262017e-06,
      "loss": 0.0763,
      "num_input_tokens_seen": 2634488,
      "step": 4245
    },
    {
      "epoch": 7.575757575757576,
      "grad_norm": 5.3125,
      "learning_rate": 8.441198831862485e-06,
      "loss": 0.05,
      "num_input_tokens_seen": 2637240,
      "step": 4250
    },
    {
      "epoch": 7.5846702317290555,
      "grad_norm": 0.024169921875,
      "learning_rate": 8.383008651425035e-06,
      "loss": 0.005,
      "num_input_tokens_seen": 2639992,
      "step": 4255
    },
    {
      "epoch": 7.593582887700535,
      "grad_norm": 24.25,
      "learning_rate": 8.32497930617006e-06,
      "loss": 0.0504,
      "num_input_tokens_seen": 2643832,
      "step": 4260
    },
    {
      "epoch": 7.602495543672014,
      "grad_norm": 8.0,
      "learning_rate": 8.267111357761243e-06,
      "loss": 0.0035,
      "num_input_tokens_seen": 2646712,
      "step": 4265
    },
    {
      "epoch": 7.611408199643494,
      "grad_norm": 17.125,
      "learning_rate": 8.209405366300088e-06,
      "loss": 0.0828,
      "num_input_tokens_seen": 2650072,
      "step": 4270
    },
    {
      "epoch": 7.620320855614973,
      "grad_norm": 2.90625,
      "learning_rate": 8.151861890320528e-06,
      "loss": 0.0014,
      "num_input_tokens_seen": 2653656,
      "step": 4275
    },
    {
      "epoch": 7.6292335115864525,
      "grad_norm": 1.3984375,
      "learning_rate": 8.094481486783534e-06,
      "loss": 0.0652,
      "num_input_tokens_seen": 2657464,
      "step": 4280
    },
    {
      "epoch": 7.638146167557933,
      "grad_norm": 0.06787109375,
      "learning_rate": 8.0372647110717e-06,
      "loss": 0.1452,
      "num_input_tokens_seen": 2660568,
      "step": 4285
    },
    {
      "epoch": 7.647058823529412,
      "grad_norm": 8.75,
      "learning_rate": 7.98021211698385e-06,
      "loss": 0.0047,
      "num_input_tokens_seen": 2663448,
      "step": 4290
    },
    {
      "epoch": 7.6559714795008915,
      "grad_norm": 17.125,
      "learning_rate": 7.923324256729738e-06,
      "loss": 0.1367,
      "num_input_tokens_seen": 2666136,
      "step": 4295
    },
    {
      "epoch": 7.664884135472371,
      "grad_norm": 0.24609375,
      "learning_rate": 7.866601680924633e-06,
      "loss": 0.0119,
      "num_input_tokens_seen": 2669048,
      "step": 4300
    },
    {
      "epoch": 7.67379679144385,
      "grad_norm": 0.0169677734375,
      "learning_rate": 7.810044938584038e-06,
      "loss": 0.0011,
      "num_input_tokens_seen": 2671800,
      "step": 4305
    },
    {
      "epoch": 7.68270944741533,
      "grad_norm": 0.1376953125,
      "learning_rate": 7.75365457711837e-06,
      "loss": 0.0006,
      "num_input_tokens_seen": 2675448,
      "step": 4310
    },
    {
      "epoch": 7.691622103386809,
      "grad_norm": 0.48828125,
      "learning_rate": 7.697431142327632e-06,
      "loss": 0.0008,
      "num_input_tokens_seen": 2678392,
      "step": 4315
    },
    {
      "epoch": 7.7005347593582885,
      "grad_norm": 1.8984375,
      "learning_rate": 7.641375178396151e-06,
      "loss": 0.0742,
      "num_input_tokens_seen": 2681112,
      "step": 4320
    },
    {
      "epoch": 7.709447415329768,
      "grad_norm": 0.0111083984375,
      "learning_rate": 7.585487227887328e-06,
      "loss": 0.0172,
      "num_input_tokens_seen": 2684856,
      "step": 4325
    },
    {
      "epoch": 7.718360071301248,
      "grad_norm": 0.265625,
      "learning_rate": 7.529767831738366e-06,
      "loss": 0.0057,
      "num_input_tokens_seen": 2687576,
      "step": 4330
    },
    {
      "epoch": 7.7272727272727275,
      "grad_norm": 0.043212890625,
      "learning_rate": 7.474217529255018e-06,
      "loss": 0.057,
      "num_input_tokens_seen": 2690328,
      "step": 4335
    },
    {
      "epoch": 7.736185383244207,
      "grad_norm": 0.034912109375,
      "learning_rate": 7.4188368581064124e-06,
      "loss": 0.0017,
      "num_input_tokens_seen": 2694168,
      "step": 4340
    },
    {
      "epoch": 7.745098039215686,
      "grad_norm": 0.0159912109375,
      "learning_rate": 7.3636263543197945e-06,
      "loss": 0.0594,
      "num_input_tokens_seen": 2697208,
      "step": 4345
    },
    {
      "epoch": 7.754010695187166,
      "grad_norm": 6.625,
      "learning_rate": 7.30858655227539e-06,
      "loss": 0.0541,
      "num_input_tokens_seen": 2700376,
      "step": 4350
    },
    {
      "epoch": 7.762923351158645,
      "grad_norm": 0.08056640625,
      "learning_rate": 7.253717984701208e-06,
      "loss": 0.0507,
      "num_input_tokens_seen": 2703256,
      "step": 4355
    },
    {
      "epoch": 7.7718360071301245,
      "grad_norm": 1.9765625,
      "learning_rate": 7.199021182667873e-06,
      "loss": 0.1346,
      "num_input_tokens_seen": 2705752,
      "step": 4360
    },
    {
      "epoch": 7.780748663101605,
      "grad_norm": 21.75,
      "learning_rate": 7.1444966755834954e-06,
      "loss": 0.1539,
      "num_input_tokens_seen": 2708888,
      "step": 4365
    },
    {
      "epoch": 7.789661319073084,
      "grad_norm": 0.244140625,
      "learning_rate": 7.0901449911885685e-06,
      "loss": 0.0009,
      "num_input_tokens_seen": 2711576,
      "step": 4370
    },
    {
      "epoch": 7.7985739750445635,
      "grad_norm": 21.625,
      "learning_rate": 7.035966655550838e-06,
      "loss": 0.0309,
      "num_input_tokens_seen": 2715000,
      "step": 4375
    },
    {
      "epoch": 7.807486631016043,
      "grad_norm": 0.36328125,
      "learning_rate": 6.98196219306019e-06,
      "loss": 0.0008,
      "num_input_tokens_seen": 2717880,
      "step": 4380
    },
    {
      "epoch": 7.816399286987522,
      "grad_norm": 60.0,
      "learning_rate": 6.928132126423636e-06,
      "loss": 0.0408,
      "num_input_tokens_seen": 2721240,
      "step": 4385
    },
    {
      "epoch": 7.825311942959002,
      "grad_norm": 0.064453125,
      "learning_rate": 6.8744769766601854e-06,
      "loss": 0.1936,
      "num_input_tokens_seen": 2724696,
      "step": 4390
    },
    {
      "epoch": 7.834224598930481,
      "grad_norm": 28.125,
      "learning_rate": 6.820997263095849e-06,
      "loss": 0.0644,
      "num_input_tokens_seen": 2727960,
      "step": 4395
    },
    {
      "epoch": 7.8431372549019605,
      "grad_norm": 4.21875,
      "learning_rate": 6.767693503358608e-06,
      "loss": 0.0025,
      "num_input_tokens_seen": 2731000,
      "step": 4400
    },
    {
      "epoch": 7.85204991087344,
      "grad_norm": 0.08837890625,
      "learning_rate": 6.7145662133733715e-06,
      "loss": 0.0377,
      "num_input_tokens_seen": 2734264,
      "step": 4405
    },
    {
      "epoch": 7.86096256684492,
      "grad_norm": 0.0279541015625,
      "learning_rate": 6.6616159073570135e-06,
      "loss": 0.0446,
      "num_input_tokens_seen": 2736664,
      "step": 4410
    },
    {
      "epoch": 7.8698752228163995,
      "grad_norm": 0.20703125,
      "learning_rate": 6.6088430978133914e-06,
      "loss": 0.0262,
      "num_input_tokens_seen": 2739672,
      "step": 4415
    },
    {
      "epoch": 7.878787878787879,
      "grad_norm": 0.0458984375,
      "learning_rate": 6.556248295528389e-06,
      "loss": 0.0968,
      "num_input_tokens_seen": 2742552,
      "step": 4420
    },
    {
      "epoch": 7.887700534759358,
      "grad_norm": 0.047119140625,
      "learning_rate": 6.5038320095649395e-06,
      "loss": 0.0121,
      "num_input_tokens_seen": 2745880,
      "step": 4425
    },
    {
      "epoch": 7.896613190730838,
      "grad_norm": 0.068359375,
      "learning_rate": 6.451594747258155e-06,
      "loss": 0.0374,
      "num_input_tokens_seen": 2749912,
      "step": 4430
    },
    {
      "epoch": 7.905525846702317,
      "grad_norm": 1.671875,
      "learning_rate": 6.399537014210355e-06,
      "loss": 0.0212,
      "num_input_tokens_seen": 2753368,
      "step": 4435
    },
    {
      "epoch": 7.9144385026737964,
      "grad_norm": 1.3125,
      "learning_rate": 6.3476593142862275e-06,
      "loss": 0.0653,
      "num_input_tokens_seen": 2756568,
      "step": 4440
    },
    {
      "epoch": 7.923351158645277,
      "grad_norm": 18.125,
      "learning_rate": 6.29596214960792e-06,
      "loss": 0.0618,
      "num_input_tokens_seen": 2759704,
      "step": 4445
    },
    {
      "epoch": 7.932263814616756,
      "grad_norm": 0.162109375,
      "learning_rate": 6.244446020550182e-06,
      "loss": 0.0238,
      "num_input_tokens_seen": 2762584,
      "step": 4450
    },
    {
      "epoch": 7.9411764705882355,
      "grad_norm": 7.90625,
      "learning_rate": 6.193111425735515e-06,
      "loss": 0.0035,
      "num_input_tokens_seen": 2765752,
      "step": 4455
    },
    {
      "epoch": 7.950089126559715,
      "grad_norm": 2.453125,
      "learning_rate": 6.141958862029384e-06,
      "loss": 0.0055,
      "num_input_tokens_seen": 2768696,
      "step": 4460
    },
    {
      "epoch": 7.959001782531194,
      "grad_norm": 4.25,
      "learning_rate": 6.090988824535374e-06,
      "loss": 0.062,
      "num_input_tokens_seen": 2772120,
      "step": 4465
    },
    {
      "epoch": 7.967914438502674,
      "grad_norm": 0.080078125,
      "learning_rate": 6.040201806590387e-06,
      "loss": 0.2793,
      "num_input_tokens_seen": 2775384,
      "step": 4470
    },
    {
      "epoch": 7.976827094474153,
      "grad_norm": 0.515625,
      "learning_rate": 5.989598299759919e-06,
      "loss": 0.0087,
      "num_input_tokens_seen": 2778520,
      "step": 4475
    },
    {
      "epoch": 7.9857397504456324,
      "grad_norm": 3.21875,
      "learning_rate": 5.939178793833233e-06,
      "loss": 0.0137,
      "num_input_tokens_seen": 2780888,
      "step": 4480
    },
    {
      "epoch": 7.994652406417112,
      "grad_norm": 0.04833984375,
      "learning_rate": 5.888943776818684e-06,
      "loss": 0.0554,
      "num_input_tokens_seen": 2784312,
      "step": 4485
    },
    {
      "epoch": 8.003565062388592,
      "grad_norm": 0.017578125,
      "learning_rate": 5.83889373493896e-06,
      "loss": 0.0004,
      "num_input_tokens_seen": 2787056,
      "step": 4490
    },
    {
      "epoch": 8.01247771836007,
      "grad_norm": 0.08447265625,
      "learning_rate": 5.789029152626374e-06,
      "loss": 0.0794,
      "num_input_tokens_seen": 2790288,
      "step": 4495
    },
    {
      "epoch": 8.014260249554367,
      "eval_loss": 0.1566127985715866,
      "eval_runtime": 6.2879,
      "eval_samples_per_second": 39.6,
      "eval_steps_per_second": 10.019,
      "num_input_tokens_seen": 2790832,
      "step": 4496
    },
    {
      "epoch": 8.02139037433155,
      "grad_norm": 0.0693359375,
      "learning_rate": 5.73935051251818e-06,
      "loss": 0.0913,
      "num_input_tokens_seen": 2793136,
      "step": 4500
    },
    {
      "epoch": 8.030303030303031,
      "grad_norm": 2.171875,
      "learning_rate": 5.689858295451914e-06,
      "loss": 0.0212,
      "num_input_tokens_seen": 2796464,
      "step": 4505
    },
    {
      "epoch": 8.03921568627451,
      "grad_norm": 0.025634765625,
      "learning_rate": 5.640552980460742e-06,
      "loss": 0.0003,
      "num_input_tokens_seen": 2799344,
      "step": 4510
    },
    {
      "epoch": 8.04812834224599,
      "grad_norm": 3.34375,
      "learning_rate": 5.591435044768783e-06,
      "loss": 0.1263,
      "num_input_tokens_seen": 2801648,
      "step": 4515
    },
    {
      "epoch": 8.057040998217468,
      "grad_norm": 29.75,
      "learning_rate": 5.542504963786552e-06,
      "loss": 0.0286,
      "num_input_tokens_seen": 2804976,
      "step": 4520
    },
    {
      "epoch": 8.065953654188949,
      "grad_norm": 0.02587890625,
      "learning_rate": 5.493763211106293e-06,
      "loss": 0.0051,
      "num_input_tokens_seen": 2807472,
      "step": 4525
    },
    {
      "epoch": 8.074866310160427,
      "grad_norm": 0.1875,
      "learning_rate": 5.4452102584974545e-06,
      "loss": 0.0008,
      "num_input_tokens_seen": 2810768,
      "step": 4530
    },
    {
      "epoch": 8.083778966131907,
      "grad_norm": 0.1513671875,
      "learning_rate": 5.396846575902095e-06,
      "loss": 0.0798,
      "num_input_tokens_seen": 2814480,
      "step": 4535
    },
    {
      "epoch": 8.092691622103386,
      "grad_norm": 0.04541015625,
      "learning_rate": 5.348672631430318e-06,
      "loss": 0.0871,
      "num_input_tokens_seen": 2817968,
      "step": 4540
    },
    {
      "epoch": 8.101604278074866,
      "grad_norm": 0.04443359375,
      "learning_rate": 5.300688891355765e-06,
      "loss": 0.0298,
      "num_input_tokens_seen": 2820784,
      "step": 4545
    },
    {
      "epoch": 8.110516934046347,
      "grad_norm": 2.203125,
      "learning_rate": 5.252895820111112e-06,
      "loss": 0.0013,
      "num_input_tokens_seen": 2823824,
      "step": 4550
    },
    {
      "epoch": 8.119429590017825,
      "grad_norm": 24.75,
      "learning_rate": 5.205293880283552e-06,
      "loss": 0.0577,
      "num_input_tokens_seen": 2826832,
      "step": 4555
    },
    {
      "epoch": 8.128342245989305,
      "grad_norm": 17.375,
      "learning_rate": 5.157883532610305e-06,
      "loss": 0.1202,
      "num_input_tokens_seen": 2830256,
      "step": 4560
    },
    {
      "epoch": 8.137254901960784,
      "grad_norm": 0.1416015625,
      "learning_rate": 5.110665235974219e-06,
      "loss": 0.0166,
      "num_input_tokens_seen": 2832848,
      "step": 4565
    },
    {
      "epoch": 8.146167557932264,
      "grad_norm": 5.5625,
      "learning_rate": 5.06363944739924e-06,
      "loss": 0.0079,
      "num_input_tokens_seen": 2835664,
      "step": 4570
    },
    {
      "epoch": 8.155080213903743,
      "grad_norm": 0.0294189453125,
      "learning_rate": 5.0168066220460715e-06,
      "loss": 0.0216,
      "num_input_tokens_seen": 2838864,
      "step": 4575
    },
    {
      "epoch": 8.163992869875223,
      "grad_norm": 0.0155029296875,
      "learning_rate": 4.97016721320773e-06,
      "loss": 0.0084,
      "num_input_tokens_seen": 2841840,
      "step": 4580
    },
    {
      "epoch": 8.172905525846703,
      "grad_norm": 0.212890625,
      "learning_rate": 4.9237216723051485e-06,
      "loss": 0.0398,
      "num_input_tokens_seen": 2844976,
      "step": 4585
    },
    {
      "epoch": 8.181818181818182,
      "grad_norm": 33.0,
      "learning_rate": 4.877470448882815e-06,
      "loss": 0.1104,
      "num_input_tokens_seen": 2847856,
      "step": 4590
    },
    {
      "epoch": 8.190730837789662,
      "grad_norm": 4.21875,
      "learning_rate": 4.831413990604447e-06,
      "loss": 0.005,
      "num_input_tokens_seen": 2850192,
      "step": 4595
    },
    {
      "epoch": 8.19964349376114,
      "grad_norm": 0.63671875,
      "learning_rate": 4.7855527432486336e-06,
      "loss": 0.0028,
      "num_input_tokens_seen": 2853008,
      "step": 4600
    },
    {
      "epoch": 8.20855614973262,
      "grad_norm": 0.287109375,
      "learning_rate": 4.739887150704508e-06,
      "loss": 0.0306,
      "num_input_tokens_seen": 2856464,
      "step": 4605
    },
    {
      "epoch": 8.2174688057041,
      "grad_norm": 4.90625,
      "learning_rate": 4.694417654967492e-06,
      "loss": 0.1142,
      "num_input_tokens_seen": 2858864,
      "step": 4610
    },
    {
      "epoch": 8.22638146167558,
      "grad_norm": 1.0703125,
      "learning_rate": 4.649144696134972e-06,
      "loss": 0.1039,
      "num_input_tokens_seen": 2861488,
      "step": 4615
    },
    {
      "epoch": 8.235294117647058,
      "grad_norm": 0.2255859375,
      "learning_rate": 4.6040687124020794e-06,
      "loss": 0.0168,
      "num_input_tokens_seen": 2865136,
      "step": 4620
    },
    {
      "epoch": 8.244206773618538,
      "grad_norm": 30.5,
      "learning_rate": 4.5591901400574285e-06,
      "loss": 0.0975,
      "num_input_tokens_seen": 2867984,
      "step": 4625
    },
    {
      "epoch": 8.253119429590019,
      "grad_norm": 0.07763671875,
      "learning_rate": 4.514509413478888e-06,
      "loss": 0.045,
      "num_input_tokens_seen": 2871088,
      "step": 4630
    },
    {
      "epoch": 8.262032085561497,
      "grad_norm": 0.224609375,
      "learning_rate": 4.470026965129384e-06,
      "loss": 0.0047,
      "num_input_tokens_seen": 2874352,
      "step": 4635
    },
    {
      "epoch": 8.270944741532977,
      "grad_norm": 0.1669921875,
      "learning_rate": 4.425743225552731e-06,
      "loss": 0.1698,
      "num_input_tokens_seen": 2877840,
      "step": 4640
    },
    {
      "epoch": 8.279857397504456,
      "grad_norm": 0.01251220703125,
      "learning_rate": 4.381658623369445e-06,
      "loss": 0.04,
      "num_input_tokens_seen": 2881456,
      "step": 4645
    },
    {
      "epoch": 8.288770053475936,
      "grad_norm": 0.01470947265625,
      "learning_rate": 4.337773585272581e-06,
      "loss": 0.0493,
      "num_input_tokens_seen": 2884400,
      "step": 4650
    },
    {
      "epoch": 8.297682709447415,
      "grad_norm": 0.0225830078125,
      "learning_rate": 4.294088536023652e-06,
      "loss": 0.0691,
      "num_input_tokens_seen": 2887536,
      "step": 4655
    },
    {
      "epoch": 8.306595365418895,
      "grad_norm": 0.029541015625,
      "learning_rate": 4.250603898448455e-06,
      "loss": 0.0404,
      "num_input_tokens_seen": 2890352,
      "step": 4660
    },
    {
      "epoch": 8.315508021390375,
      "grad_norm": 11.625,
      "learning_rate": 4.2073200934330315e-06,
      "loss": 0.0727,
      "num_input_tokens_seen": 2893520,
      "step": 4665
    },
    {
      "epoch": 8.324420677361854,
      "grad_norm": 38.25,
      "learning_rate": 4.164237539919577e-06,
      "loss": 0.1344,
      "num_input_tokens_seen": 2896048,
      "step": 4670
    },
    {
      "epoch": 8.333333333333334,
      "grad_norm": 0.265625,
      "learning_rate": 4.121356654902364e-06,
      "loss": 0.0411,
      "num_input_tokens_seen": 2899472,
      "step": 4675
    },
    {
      "epoch": 8.342245989304812,
      "grad_norm": 0.333984375,
      "learning_rate": 4.078677853423724e-06,
      "loss": 0.0012,
      "num_input_tokens_seen": 2902832,
      "step": 4680
    },
    {
      "epoch": 8.351158645276293,
      "grad_norm": 0.34765625,
      "learning_rate": 4.036201548570049e-06,
      "loss": 0.08,
      "num_input_tokens_seen": 2906576,
      "step": 4685
    },
    {
      "epoch": 8.360071301247771,
      "grad_norm": 0.04345703125,
      "learning_rate": 3.993928151467766e-06,
      "loss": 0.0402,
      "num_input_tokens_seen": 2909840,
      "step": 4690
    },
    {
      "epoch": 8.368983957219251,
      "grad_norm": 10.8125,
      "learning_rate": 3.951858071279352e-06,
      "loss": 0.0201,
      "num_input_tokens_seen": 2912752,
      "step": 4695
    },
    {
      "epoch": 8.37789661319073,
      "grad_norm": 0.259765625,
      "learning_rate": 3.909991715199412e-06,
      "loss": 0.0012,
      "num_input_tokens_seen": 2915024,
      "step": 4700
    },
    {
      "epoch": 8.38680926916221,
      "grad_norm": 0.033203125,
      "learning_rate": 3.8683294884506945e-06,
      "loss": 0.0012,
      "num_input_tokens_seen": 2918480,
      "step": 4705
    },
    {
      "epoch": 8.39572192513369,
      "grad_norm": 23.5,
      "learning_rate": 3.826871794280193e-06,
      "loss": 0.0716,
      "num_input_tokens_seen": 2921712,
      "step": 4710
    },
    {
      "epoch": 8.404634581105169,
      "grad_norm": 28.125,
      "learning_rate": 3.7856190339552513e-06,
      "loss": 0.0972,
      "num_input_tokens_seen": 2925040,
      "step": 4715
    },
    {
      "epoch": 8.41354723707665,
      "grad_norm": 0.97265625,
      "learning_rate": 3.7445716067596503e-06,
      "loss": 0.0277,
      "num_input_tokens_seen": 2928112,
      "step": 4720
    },
    {
      "epoch": 8.422459893048128,
      "grad_norm": 0.041259765625,
      "learning_rate": 3.7037299099897586e-06,
      "loss": 0.0825,
      "num_input_tokens_seen": 2932368,
      "step": 4725
    },
    {
      "epoch": 8.431372549019608,
      "grad_norm": 0.0277099609375,
      "learning_rate": 3.663094338950704e-06,
      "loss": 0.055,
      "num_input_tokens_seen": 2935088,
      "step": 4730
    },
    {
      "epoch": 8.440285204991087,
      "grad_norm": 40.0,
      "learning_rate": 3.6226652869525285e-06,
      "loss": 0.0191,
      "num_input_tokens_seen": 2937840,
      "step": 4735
    },
    {
      "epoch": 8.449197860962567,
      "grad_norm": 3.609375,
      "learning_rate": 3.5824431453063662e-06,
      "loss": 0.0527,
      "num_input_tokens_seen": 2941008,
      "step": 4740
    },
    {
      "epoch": 8.458110516934047,
      "grad_norm": 0.0252685546875,
      "learning_rate": 3.5424283033207024e-06,
      "loss": 0.1851,
      "num_input_tokens_seen": 2944464,
      "step": 4745
    },
    {
      "epoch": 8.467023172905526,
      "grad_norm": 30.875,
      "learning_rate": 3.5026211482975497e-06,
      "loss": 0.0429,
      "num_input_tokens_seen": 2947376,
      "step": 4750
    },
    {
      "epoch": 8.475935828877006,
      "grad_norm": 0.45703125,
      "learning_rate": 3.463022065528748e-06,
      "loss": 0.0462,
      "num_input_tokens_seen": 2950480,
      "step": 4755
    },
    {
      "epoch": 8.484848484848484,
      "grad_norm": 0.40625,
      "learning_rate": 3.4236314382922103e-06,
      "loss": 0.0008,
      "num_input_tokens_seen": 2953392,
      "step": 4760
    },
    {
      "epoch": 8.493761140819965,
      "grad_norm": 0.02978515625,
      "learning_rate": 3.3844496478482064e-06,
      "loss": 0.0007,
      "num_input_tokens_seen": 2956272,
      "step": 4765
    },
    {
      "epoch": 8.502673796791443,
      "grad_norm": 0.01043701171875,
      "learning_rate": 3.345477073435685e-06,
      "loss": 0.1504,
      "num_input_tokens_seen": 2959056,
      "step": 4770
    },
    {
      "epoch": 8.511586452762923,
      "grad_norm": 14.75,
      "learning_rate": 3.3067140922686174e-06,
      "loss": 0.0063,
      "num_input_tokens_seen": 2962480,
      "step": 4775
    },
    {
      "epoch": 8.515151515151516,
      "eval_loss": 0.1583101898431778,
      "eval_runtime": 6.2815,
      "eval_samples_per_second": 39.64,
      "eval_steps_per_second": 10.03,
      "num_input_tokens_seen": 2963888,
      "step": 4777
    },
    {
      "epoch": 8.520499108734402,
      "grad_norm": 0.01312255859375,
      "learning_rate": 3.268161079532317e-06,
      "loss": 0.0027,
      "num_input_tokens_seen": 2965360,
      "step": 4780
    },
    {
      "epoch": 8.529411764705882,
      "grad_norm": 0.26953125,
      "learning_rate": 3.22981840837982e-06,
      "loss": 0.0006,
      "num_input_tokens_seen": 2968464,
      "step": 4785
    },
    {
      "epoch": 8.538324420677363,
      "grad_norm": 0.2021484375,
      "learning_rate": 3.1916864499282856e-06,
      "loss": 0.0389,
      "num_input_tokens_seen": 2972144,
      "step": 4790
    },
    {
      "epoch": 8.547237076648841,
      "grad_norm": 0.62109375,
      "learning_rate": 3.1537655732553768e-06,
      "loss": 0.0008,
      "num_input_tokens_seen": 2974384,
      "step": 4795
    },
    {
      "epoch": 8.556149732620321,
      "grad_norm": 38.25,
      "learning_rate": 3.1160561453957183e-06,
      "loss": 0.063,
      "num_input_tokens_seen": 2977104,
      "step": 4800
    },
    {
      "epoch": 8.5650623885918,
      "grad_norm": 0.169921875,
      "learning_rate": 3.078558531337336e-06,
      "loss": 0.0639,
      "num_input_tokens_seen": 2980464,
      "step": 4805
    },
    {
      "epoch": 8.57397504456328,
      "grad_norm": 0.033935546875,
      "learning_rate": 3.0412730940181015e-06,
      "loss": 0.0005,
      "num_input_tokens_seen": 2983248,
      "step": 4810
    },
    {
      "epoch": 8.582887700534759,
      "grad_norm": 0.0556640625,
      "learning_rate": 3.0042001943222376e-06,
      "loss": 0.0015,
      "num_input_tokens_seen": 2986256,
      "step": 4815
    },
    {
      "epoch": 8.591800356506239,
      "grad_norm": 0.029541015625,
      "learning_rate": 2.967340191076834e-06,
      "loss": 0.0002,
      "num_input_tokens_seen": 2990256,
      "step": 4820
    },
    {
      "epoch": 8.60071301247772,
      "grad_norm": 3.4375,
      "learning_rate": 2.930693441048371e-06,
      "loss": 0.0588,
      "num_input_tokens_seen": 2992592,
      "step": 4825
    },
    {
      "epoch": 8.609625668449198,
      "grad_norm": 0.05517578125,
      "learning_rate": 2.8942602989392386e-06,
      "loss": 0.1028,
      "num_input_tokens_seen": 2995888,
      "step": 4830
    },
    {
      "epoch": 8.618538324420678,
      "grad_norm": 0.051025390625,
      "learning_rate": 2.858041117384341e-06,
      "loss": 0.0167,
      "num_input_tokens_seen": 2999280,
      "step": 4835
    },
    {
      "epoch": 8.627450980392156,
      "grad_norm": 0.0157470703125,
      "learning_rate": 2.8220362469476624e-06,
      "loss": 0.1074,
      "num_input_tokens_seen": 3002864,
      "step": 4840
    },
    {
      "epoch": 8.636363636363637,
      "grad_norm": 37.5,
      "learning_rate": 2.7862460361188614e-06,
      "loss": 0.0163,
      "num_input_tokens_seen": 3004944,
      "step": 4845
    },
    {
      "epoch": 8.645276292335115,
      "grad_norm": 0.88671875,
      "learning_rate": 2.750670831309957e-06,
      "loss": 0.0012,
      "num_input_tokens_seen": 3008464,
      "step": 4850
    },
    {
      "epoch": 8.654188948306595,
      "grad_norm": 9.875,
      "learning_rate": 2.7153109768518925e-06,
      "loss": 0.0027,
      "num_input_tokens_seen": 3012240,
      "step": 4855
    },
    {
      "epoch": 8.663101604278076,
      "grad_norm": 0.0279541015625,
      "learning_rate": 2.680166814991256e-06,
      "loss": 0.0144,
      "num_input_tokens_seen": 3015056,
      "step": 4860
    },
    {
      "epoch": 8.672014260249554,
      "grad_norm": 0.08447265625,
      "learning_rate": 2.645238685886961e-06,
      "loss": 0.0146,
      "num_input_tokens_seen": 3018160,
      "step": 4865
    },
    {
      "epoch": 8.680926916221035,
      "grad_norm": 26.5,
      "learning_rate": 2.6105269276069573e-06,
      "loss": 0.1039,
      "num_input_tokens_seen": 3021392,
      "step": 4870
    },
    {
      "epoch": 8.689839572192513,
      "grad_norm": 0.95703125,
      "learning_rate": 2.5760318761249263e-06,
      "loss": 0.0219,
      "num_input_tokens_seen": 3024240,
      "step": 4875
    },
    {
      "epoch": 8.698752228163993,
      "grad_norm": 0.072265625,
      "learning_rate": 2.541753865317076e-06,
      "loss": 0.0881,
      "num_input_tokens_seen": 3026800,
      "step": 4880
    },
    {
      "epoch": 8.707664884135472,
      "grad_norm": 0.48828125,
      "learning_rate": 2.507693226958871e-06,
      "loss": 0.0007,
      "num_input_tokens_seen": 3029968,
      "step": 4885
    },
    {
      "epoch": 8.716577540106952,
      "grad_norm": 3.859375,
      "learning_rate": 2.473850290721838e-06,
      "loss": 0.1465,
      "num_input_tokens_seen": 3032656,
      "step": 4890
    },
    {
      "epoch": 8.72549019607843,
      "grad_norm": 0.453125,
      "learning_rate": 2.4402253841703914e-06,
      "loss": 0.0205,
      "num_input_tokens_seen": 3035376,
      "step": 4895
    },
    {
      "epoch": 8.73440285204991,
      "grad_norm": 0.01275634765625,
      "learning_rate": 2.4068188327586257e-06,
      "loss": 0.1735,
      "num_input_tokens_seen": 3038512,
      "step": 4900
    },
    {
      "epoch": 8.743315508021391,
      "grad_norm": 0.94921875,
      "learning_rate": 2.373630959827186e-06,
      "loss": 0.0055,
      "num_input_tokens_seen": 3041744,
      "step": 4905
    },
    {
      "epoch": 8.75222816399287,
      "grad_norm": 0.056884765625,
      "learning_rate": 2.3406620866001485e-06,
      "loss": 0.0884,
      "num_input_tokens_seen": 3045232,
      "step": 4910
    },
    {
      "epoch": 8.76114081996435,
      "grad_norm": 0.54296875,
      "learning_rate": 2.3079125321818996e-06,
      "loss": 0.0012,
      "num_input_tokens_seen": 3047728,
      "step": 4915
    },
    {
      "epoch": 8.770053475935828,
      "grad_norm": 0.040771484375,
      "learning_rate": 2.275382613554031e-06,
      "loss": 0.1771,
      "num_input_tokens_seen": 3050864,
      "step": 4920
    },
    {
      "epoch": 8.778966131907309,
      "grad_norm": 0.2373046875,
      "learning_rate": 2.2430726455723113e-06,
      "loss": 0.0071,
      "num_input_tokens_seen": 3053680,
      "step": 4925
    },
    {
      "epoch": 8.787878787878787,
      "grad_norm": 0.10888671875,
      "learning_rate": 2.210982940963596e-06,
      "loss": 0.0768,
      "num_input_tokens_seen": 3057136,
      "step": 4930
    },
    {
      "epoch": 8.796791443850267,
      "grad_norm": 37.75,
      "learning_rate": 2.1791138103228275e-06,
      "loss": 0.0193,
      "num_input_tokens_seen": 3060144,
      "step": 4935
    },
    {
      "epoch": 8.805704099821746,
      "grad_norm": 0.283203125,
      "learning_rate": 2.1474655621100347e-06,
      "loss": 0.1267,
      "num_input_tokens_seen": 3063024,
      "step": 4940
    },
    {
      "epoch": 8.814616755793226,
      "grad_norm": 0.12890625,
      "learning_rate": 2.116038502647319e-06,
      "loss": 0.001,
      "num_input_tokens_seen": 3066320,
      "step": 4945
    },
    {
      "epoch": 8.823529411764707,
      "grad_norm": 0.0174560546875,
      "learning_rate": 2.084832936115902e-06,
      "loss": 0.0392,
      "num_input_tokens_seen": 3069296,
      "step": 4950
    },
    {
      "epoch": 8.832442067736185,
      "grad_norm": 0.02392578125,
      "learning_rate": 2.0538491645531982e-06,
      "loss": 0.0302,
      "num_input_tokens_seen": 3071888,
      "step": 4955
    },
    {
      "epoch": 8.841354723707665,
      "grad_norm": 27.75,
      "learning_rate": 2.0230874878498648e-06,
      "loss": 0.0071,
      "num_input_tokens_seen": 3075984,
      "step": 4960
    },
    {
      "epoch": 8.850267379679144,
      "grad_norm": 0.19140625,
      "learning_rate": 1.9925482037469188e-06,
      "loss": 0.024,
      "num_input_tokens_seen": 3079152,
      "step": 4965
    },
    {
      "epoch": 8.859180035650624,
      "grad_norm": 11.4375,
      "learning_rate": 1.9622316078328566e-06,
      "loss": 0.0991,
      "num_input_tokens_seen": 3082544,
      "step": 4970
    },
    {
      "epoch": 8.868092691622103,
      "grad_norm": 0.2109375,
      "learning_rate": 1.9321379935407697e-06,
      "loss": 0.0211,
      "num_input_tokens_seen": 3085680,
      "step": 4975
    },
    {
      "epoch": 8.877005347593583,
      "grad_norm": 11.0,
      "learning_rate": 1.9022676521455117e-06,
      "loss": 0.1267,
      "num_input_tokens_seen": 3089392,
      "step": 4980
    },
    {
      "epoch": 8.885918003565063,
      "grad_norm": 0.06298828125,
      "learning_rate": 1.8726208727609219e-06,
      "loss": 0.0369,
      "num_input_tokens_seen": 3092656,
      "step": 4985
    },
    {
      "epoch": 8.894830659536542,
      "grad_norm": 0.072265625,
      "learning_rate": 1.8431979423369604e-06,
      "loss": 0.0336,
      "num_input_tokens_seen": 3095600,
      "step": 4990
    },
    {
      "epoch": 8.903743315508022,
      "grad_norm": 0.31640625,
      "learning_rate": 1.8139991456569694e-06,
      "loss": 0.0862,
      "num_input_tokens_seen": 3098320,
      "step": 4995
    },
    {
      "epoch": 8.9126559714795,
      "grad_norm": 38.25,
      "learning_rate": 1.7850247653349223e-06,
      "loss": 0.0456,
      "num_input_tokens_seen": 3101520,
      "step": 5000
    },
    {
      "epoch": 8.92156862745098,
      "grad_norm": 20.875,
      "learning_rate": 1.7562750818126556e-06,
      "loss": 0.0171,
      "num_input_tokens_seen": 3104816,
      "step": 5005
    },
    {
      "epoch": 8.93048128342246,
      "grad_norm": 0.0238037109375,
      "learning_rate": 1.727750373357187e-06,
      "loss": 0.0012,
      "num_input_tokens_seen": 3108176,
      "step": 5010
    },
    {
      "epoch": 8.93939393939394,
      "grad_norm": 57.0,
      "learning_rate": 1.699450916058018e-06,
      "loss": 0.1332,
      "num_input_tokens_seen": 3111248,
      "step": 5015
    },
    {
      "epoch": 8.94830659536542,
      "grad_norm": 0.0693359375,
      "learning_rate": 1.6713769838244325e-06,
      "loss": 0.1009,
      "num_input_tokens_seen": 3114224,
      "step": 5020
    },
    {
      "epoch": 8.957219251336898,
      "grad_norm": 0.027099609375,
      "learning_rate": 1.6435288483828748e-06,
      "loss": 0.0024,
      "num_input_tokens_seen": 3117232,
      "step": 5025
    },
    {
      "epoch": 8.966131907308379,
      "grad_norm": 0.017822265625,
      "learning_rate": 1.615906779274326e-06,
      "loss": 0.0626,
      "num_input_tokens_seen": 3120240,
      "step": 5030
    },
    {
      "epoch": 8.975044563279857,
      "grad_norm": 16.25,
      "learning_rate": 1.588511043851662e-06,
      "loss": 0.2263,
      "num_input_tokens_seen": 3123792,
      "step": 5035
    },
    {
      "epoch": 8.983957219251337,
      "grad_norm": 0.01556396484375,
      "learning_rate": 1.5613419072770864e-06,
      "loss": 0.0026,
      "num_input_tokens_seen": 3127184,
      "step": 5040
    },
    {
      "epoch": 8.992869875222816,
      "grad_norm": 0.04052734375,
      "learning_rate": 1.534399632519573e-06,
      "loss": 0.0026,
      "num_input_tokens_seen": 3130480,
      "step": 5045
    },
    {
      "epoch": 9.001782531194296,
      "grad_norm": 0.291015625,
      "learning_rate": 1.5076844803522922e-06,
      "loss": 0.0006,
      "num_input_tokens_seen": 3132712,
      "step": 5050
    },
    {
      "epoch": 9.010695187165775,
      "grad_norm": 0.0390625,
      "learning_rate": 1.4811967093501189e-06,
      "loss": 0.0007,
      "num_input_tokens_seen": 3135400,
      "step": 5055
    },
    {
      "epoch": 9.016042780748663,
      "eval_loss": 0.15806140005588531,
      "eval_runtime": 6.2781,
      "eval_samples_per_second": 39.662,
      "eval_steps_per_second": 10.035,
      "num_input_tokens_seen": 3137352,
      "step": 5058
    },
    {
      "epoch": 9.019607843137255,
      "grad_norm": 0.1650390625,
      "learning_rate": 1.4549365758871142e-06,
      "loss": 0.0742,
      "num_input_tokens_seen": 3138248,
      "step": 5060
    },
    {
      "epoch": 9.028520499108735,
      "grad_norm": 0.8046875,
      "learning_rate": 1.4289043341340375e-06,
      "loss": 0.0338,
      "num_input_tokens_seen": 3141480,
      "step": 5065
    },
    {
      "epoch": 9.037433155080214,
      "grad_norm": 13.1875,
      "learning_rate": 1.4031002360558849e-06,
      "loss": 0.0035,
      "num_input_tokens_seen": 3144904,
      "step": 5070
    },
    {
      "epoch": 9.046345811051694,
      "grad_norm": 0.19921875,
      "learning_rate": 1.377524531409491e-06,
      "loss": 0.0448,
      "num_input_tokens_seen": 3148968,
      "step": 5075
    },
    {
      "epoch": 9.055258467023172,
      "grad_norm": 0.064453125,
      "learning_rate": 1.3521774677410476e-06,
      "loss": 0.0376,
      "num_input_tokens_seen": 3151912,
      "step": 5080
    },
    {
      "epoch": 9.064171122994653,
      "grad_norm": 0.0211181640625,
      "learning_rate": 1.3270592903837503e-06,
      "loss": 0.0003,
      "num_input_tokens_seen": 3155080,
      "step": 5085
    },
    {
      "epoch": 9.073083778966131,
      "grad_norm": 0.01611328125,
      "learning_rate": 1.3021702424554221e-06,
      "loss": 0.0624,
      "num_input_tokens_seen": 3157768,
      "step": 5090
    },
    {
      "epoch": 9.081996434937611,
      "grad_norm": 2.171875,
      "learning_rate": 1.2775105648561352e-06,
      "loss": 0.0089,
      "num_input_tokens_seen": 3161224,
      "step": 5095
    },
    {
      "epoch": 9.090909090909092,
      "grad_norm": 16.0,
      "learning_rate": 1.2530804962659098e-06,
      "loss": 0.0148,
      "num_input_tokens_seen": 3163944,
      "step": 5100
    },
    {
      "epoch": 9.09982174688057,
      "grad_norm": 0.921875,
      "learning_rate": 1.2288802731423883e-06,
      "loss": 0.0874,
      "num_input_tokens_seen": 3166728,
      "step": 5105
    },
    {
      "epoch": 9.10873440285205,
      "grad_norm": 6.0,
      "learning_rate": 1.2049101297185422e-06,
      "loss": 0.0666,
      "num_input_tokens_seen": 3170120,
      "step": 5110
    },
    {
      "epoch": 9.117647058823529,
      "grad_norm": 0.010498046875,
      "learning_rate": 1.1811702980004058e-06,
      "loss": 0.0025,
      "num_input_tokens_seen": 3173000,
      "step": 5115
    },
    {
      "epoch": 9.12655971479501,
      "grad_norm": 0.1171875,
      "learning_rate": 1.1576610077648513e-06,
      "loss": 0.0083,
      "num_input_tokens_seen": 3176520,
      "step": 5120
    },
    {
      "epoch": 9.135472370766488,
      "grad_norm": 0.064453125,
      "learning_rate": 1.134382486557342e-06,
      "loss": 0.0005,
      "num_input_tokens_seen": 3179496,
      "step": 5125
    },
    {
      "epoch": 9.144385026737968,
      "grad_norm": 33.5,
      "learning_rate": 1.1113349596897331e-06,
      "loss": 0.0083,
      "num_input_tokens_seen": 3182248,
      "step": 5130
    },
    {
      "epoch": 9.153297682709447,
      "grad_norm": 40.0,
      "learning_rate": 1.0885186502381017e-06,
      "loss": 0.0882,
      "num_input_tokens_seen": 3184840,
      "step": 5135
    },
    {
      "epoch": 9.162210338680927,
      "grad_norm": 0.9609375,
      "learning_rate": 1.0659337790405704e-06,
      "loss": 0.0108,
      "num_input_tokens_seen": 3187720,
      "step": 5140
    },
    {
      "epoch": 9.171122994652407,
      "grad_norm": 0.01080322265625,
      "learning_rate": 1.0435805646951958e-06,
      "loss": 0.0015,
      "num_input_tokens_seen": 3190536,
      "step": 5145
    },
    {
      "epoch": 9.180035650623886,
      "grad_norm": 0.11767578125,
      "learning_rate": 1.0214592235578274e-06,
      "loss": 0.0013,
      "num_input_tokens_seen": 3193608,
      "step": 5150
    },
    {
      "epoch": 9.188948306595366,
      "grad_norm": 2.984375,
      "learning_rate": 9.995699697400247e-07,
      "loss": 0.0014,
      "num_input_tokens_seen": 3196936,
      "step": 5155
    },
    {
      "epoch": 9.197860962566844,
      "grad_norm": 0.01416015625,
      "learning_rate": 9.77913015106982e-07,
      "loss": 0.0024,
      "num_input_tokens_seen": 3200040,
      "step": 5160
    },
    {
      "epoch": 9.206773618538325,
      "grad_norm": 1.09375,
      "learning_rate": 9.564885692754793e-07,
      "loss": 0.0527,
      "num_input_tokens_seen": 3203240,
      "step": 5165
    },
    {
      "epoch": 9.215686274509803,
      "grad_norm": 0.01129150390625,
      "learning_rate": 9.352968396118628e-07,
      "loss": 0.0041,
      "num_input_tokens_seen": 3206376,
      "step": 5170
    },
    {
      "epoch": 9.224598930481283,
      "grad_norm": 35.25,
      "learning_rate": 9.143380312300137e-07,
      "loss": 0.0784,
      "num_input_tokens_seen": 3209480,
      "step": 5175
    },
    {
      "epoch": 9.233511586452764,
      "grad_norm": 0.0166015625,
      "learning_rate": 8.936123469893892e-07,
      "loss": 0.0373,
      "num_input_tokens_seen": 3213448,
      "step": 5180
    },
    {
      "epoch": 9.242424242424242,
      "grad_norm": 0.018798828125,
      "learning_rate": 8.731199874930374e-07,
      "loss": 0.126,
      "num_input_tokens_seen": 3216776,
      "step": 5185
    },
    {
      "epoch": 9.251336898395722,
      "grad_norm": 0.671875,
      "learning_rate": 8.528611510856766e-07,
      "loss": 0.0006,
      "num_input_tokens_seen": 3219752,
      "step": 5190
    },
    {
      "epoch": 9.260249554367201,
      "grad_norm": 0.01708984375,
      "learning_rate": 8.328360338517583e-07,
      "loss": 0.1164,
      "num_input_tokens_seen": 3223048,
      "step": 5195
    },
    {
      "epoch": 9.269162210338681,
      "grad_norm": 0.019775390625,
      "learning_rate": 8.130448296135768e-07,
      "loss": 0.0017,
      "num_input_tokens_seen": 3226984,
      "step": 5200
    },
    {
      "epoch": 9.27807486631016,
      "grad_norm": 0.05810546875,
      "learning_rate": 7.934877299293875e-07,
      "loss": 0.035,
      "num_input_tokens_seen": 3230088,
      "step": 5205
    },
    {
      "epoch": 9.28698752228164,
      "grad_norm": 0.69140625,
      "learning_rate": 7.741649240915666e-07,
      "loss": 0.0006,
      "num_input_tokens_seen": 3232840,
      "step": 5210
    },
    {
      "epoch": 9.29590017825312,
      "grad_norm": 0.055419921875,
      "learning_rate": 7.550765991247654e-07,
      "loss": 0.0003,
      "num_input_tokens_seen": 3235944,
      "step": 5215
    },
    {
      "epoch": 9.304812834224599,
      "grad_norm": 0.470703125,
      "learning_rate": 7.362229397840981e-07,
      "loss": 0.0231,
      "num_input_tokens_seen": 3238728,
      "step": 5220
    },
    {
      "epoch": 9.313725490196079,
      "grad_norm": 13.9375,
      "learning_rate": 7.17604128553373e-07,
      "loss": 0.2057,
      "num_input_tokens_seen": 3241256,
      "step": 5225
    },
    {
      "epoch": 9.322638146167558,
      "grad_norm": 0.1240234375,
      "learning_rate": 6.992203456432977e-07,
      "loss": 0.0871,
      "num_input_tokens_seen": 3244680,
      "step": 5230
    },
    {
      "epoch": 9.331550802139038,
      "grad_norm": 0.0135498046875,
      "learning_rate": 6.810717689897633e-07,
      "loss": 0.0041,
      "num_input_tokens_seen": 3247560,
      "step": 5235
    },
    {
      "epoch": 9.340463458110516,
      "grad_norm": 0.3359375,
      "learning_rate": 6.631585742521068e-07,
      "loss": 0.0077,
      "num_input_tokens_seen": 3251176,
      "step": 5240
    },
    {
      "epoch": 9.349376114081997,
      "grad_norm": 0.0181884765625,
      "learning_rate": 6.454809348114044e-07,
      "loss": 0.0529,
      "num_input_tokens_seen": 3254152,
      "step": 5245
    },
    {
      "epoch": 9.358288770053475,
      "grad_norm": 0.0179443359375,
      "learning_rate": 6.280390217688114e-07,
      "loss": 0.0004,
      "num_input_tokens_seen": 3256744,
      "step": 5250
    },
    {
      "epoch": 9.367201426024955,
      "grad_norm": 0.90234375,
      "learning_rate": 6.108330039438892e-07,
      "loss": 0.032,
      "num_input_tokens_seen": 3259400,
      "step": 5255
    },
    {
      "epoch": 9.376114081996436,
      "grad_norm": 0.5078125,
      "learning_rate": 5.938630478729917e-07,
      "loss": 0.0469,
      "num_input_tokens_seen": 3262728,
      "step": 5260
    },
    {
      "epoch": 9.385026737967914,
      "grad_norm": 0.11083984375,
      "learning_rate": 5.771293178076286e-07,
      "loss": 0.0692,
      "num_input_tokens_seen": 3266376,
      "step": 5265
    },
    {
      "epoch": 9.393939393939394,
      "grad_norm": 2.03125,
      "learning_rate": 5.606319757128914e-07,
      "loss": 0.0088,
      "num_input_tokens_seen": 3268808,
      "step": 5270
    },
    {
      "epoch": 9.402852049910873,
      "grad_norm": 22.5,
      "learning_rate": 5.443711812658792e-07,
      "loss": 0.0079,
      "num_input_tokens_seen": 3272008,
      "step": 5275
    },
    {
      "epoch": 9.411764705882353,
      "grad_norm": 0.18359375,
      "learning_rate": 5.283470918541616e-07,
      "loss": 0.0696,
      "num_input_tokens_seen": 3274920,
      "step": 5280
    },
    {
      "epoch": 9.420677361853832,
      "grad_norm": 0.025390625,
      "learning_rate": 5.125598625742523e-07,
      "loss": 0.0042,
      "num_input_tokens_seen": 3278376,
      "step": 5285
    },
    {
      "epoch": 9.429590017825312,
      "grad_norm": 0.0361328125,
      "learning_rate": 4.970096462300927e-07,
      "loss": 0.1221,
      "num_input_tokens_seen": 3281704,
      "step": 5290
    },
    {
      "epoch": 9.43850267379679,
      "grad_norm": 29.625,
      "learning_rate": 4.816965933315987e-07,
      "loss": 0.0225,
      "num_input_tokens_seen": 3285256,
      "step": 5295
    },
    {
      "epoch": 9.44741532976827,
      "grad_norm": 0.318359375,
      "learning_rate": 4.6662085209318305e-07,
      "loss": 0.0314,
      "num_input_tokens_seen": 3288616,
      "step": 5300
    },
    {
      "epoch": 9.456327985739751,
      "grad_norm": 15.125,
      "learning_rate": 4.517825684323324e-07,
      "loss": 0.0592,
      "num_input_tokens_seen": 3291752,
      "step": 5305
    },
    {
      "epoch": 9.46524064171123,
      "grad_norm": 19.0,
      "learning_rate": 4.3718188596819086e-07,
      "loss": 0.1233,
      "num_input_tokens_seen": 3294344,
      "step": 5310
    },
    {
      "epoch": 9.47415329768271,
      "grad_norm": 0.018798828125,
      "learning_rate": 4.228189460201676e-07,
      "loss": 0.0009,
      "num_input_tokens_seen": 3297512,
      "step": 5315
    },
    {
      "epoch": 9.483065953654188,
      "grad_norm": 26.0,
      "learning_rate": 4.086938876065732e-07,
      "loss": 0.0773,
      "num_input_tokens_seen": 3300296,
      "step": 5320
    },
    {
      "epoch": 9.491978609625669,
      "grad_norm": 0.10400390625,
      "learning_rate": 3.948068474432715e-07,
      "loss": 0.006,
      "num_input_tokens_seen": 3304360,
      "step": 5325
    },
    {
      "epoch": 9.500891265597147,
      "grad_norm": 0.07177734375,
      "learning_rate": 3.8115795994236313e-07,
      "loss": 0.0812,
      "num_input_tokens_seen": 3307304,
      "step": 5330
    },
    {
      "epoch": 9.509803921568627,
      "grad_norm": 0.047119140625,
      "learning_rate": 3.6774735721087085e-07,
      "loss": 0.0782,
      "num_input_tokens_seen": 3310536,
      "step": 5335
    },
    {
      "epoch": 9.516934046345812,
      "eval_loss": 0.15913818776607513,
      "eval_runtime": 6.277,
      "eval_samples_per_second": 39.669,
      "eval_steps_per_second": 10.037,
      "num_input_tokens_seen": 3312648,
      "step": 5339
    },
    {
      "epoch": 9.518716577540108,
      "grad_norm": 15.0625,
      "learning_rate": 3.5457516904947587e-07,
      "loss": 0.0046,
      "num_input_tokens_seen": 3313672,
      "step": 5340
    },
    {
      "epoch": 9.527629233511586,
      "grad_norm": 2.90625,
      "learning_rate": 3.416415229512443e-07,
      "loss": 0.1425,
      "num_input_tokens_seen": 3317224,
      "step": 5345
    },
    {
      "epoch": 9.536541889483066,
      "grad_norm": 0.02783203125,
      "learning_rate": 3.2894654410041417e-07,
      "loss": 0.0392,
      "num_input_tokens_seen": 3319848,
      "step": 5350
    },
    {
      "epoch": 9.545454545454545,
      "grad_norm": 0.0274658203125,
      "learning_rate": 3.1649035537117123e-07,
      "loss": 0.0162,
      "num_input_tokens_seen": 3322664,
      "step": 5355
    },
    {
      "epoch": 9.554367201426025,
      "grad_norm": 0.04345703125,
      "learning_rate": 3.042730773264557e-07,
      "loss": 0.0243,
      "num_input_tokens_seen": 3325928,
      "step": 5360
    },
    {
      "epoch": 9.563279857397504,
      "grad_norm": 1.7734375,
      "learning_rate": 2.9229482821680197e-07,
      "loss": 0.0205,
      "num_input_tokens_seen": 3328680,
      "step": 5365
    },
    {
      "epoch": 9.572192513368984,
      "grad_norm": 31.0,
      "learning_rate": 2.8055572397919784e-07,
      "loss": 0.0256,
      "num_input_tokens_seen": 3331976,
      "step": 5370
    },
    {
      "epoch": 9.581105169340464,
      "grad_norm": 0.021484375,
      "learning_rate": 2.690558782359576e-07,
      "loss": 0.0021,
      "num_input_tokens_seen": 3334888,
      "step": 5375
    },
    {
      "epoch": 9.590017825311943,
      "grad_norm": 38.25,
      "learning_rate": 2.5779540229361745e-07,
      "loss": 0.0778,
      "num_input_tokens_seen": 3337960,
      "step": 5380
    },
    {
      "epoch": 9.598930481283423,
      "grad_norm": 0.271484375,
      "learning_rate": 2.467744051418641e-07,
      "loss": 0.0009,
      "num_input_tokens_seen": 3340936,
      "step": 5385
    },
    {
      "epoch": 9.607843137254902,
      "grad_norm": 52.75,
      "learning_rate": 2.3599299345248292e-07,
      "loss": 0.021,
      "num_input_tokens_seen": 3343784,
      "step": 5390
    },
    {
      "epoch": 9.616755793226382,
      "grad_norm": 34.75,
      "learning_rate": 2.2545127157831413e-07,
      "loss": 0.0243,
      "num_input_tokens_seen": 3347016,
      "step": 5395
    },
    {
      "epoch": 9.62566844919786,
      "grad_norm": 0.1318359375,
      "learning_rate": 2.1514934155226208e-07,
      "loss": 0.0374,
      "num_input_tokens_seen": 3349800,
      "step": 5400
    },
    {
      "epoch": 9.63458110516934,
      "grad_norm": 0.0302734375,
      "learning_rate": 2.0508730308627933e-07,
      "loss": 0.0016,
      "num_input_tokens_seen": 3353640,
      "step": 5405
    },
    {
      "epoch": 9.643493761140821,
      "grad_norm": 0.02294921875,
      "learning_rate": 1.9526525357043136e-07,
      "loss": 0.0963,
      "num_input_tokens_seen": 3356904,
      "step": 5410
    },
    {
      "epoch": 9.6524064171123,
      "grad_norm": 0.26171875,
      "learning_rate": 1.8568328807193337e-07,
      "loss": 0.1029,
      "num_input_tokens_seen": 3360232,
      "step": 5415
    },
    {
      "epoch": 9.66131907308378,
      "grad_norm": 11.75,
      "learning_rate": 1.7634149933423993e-07,
      "loss": 0.1306,
      "num_input_tokens_seen": 3362824,
      "step": 5420
    },
    {
      "epoch": 9.670231729055258,
      "grad_norm": 0.08154296875,
      "learning_rate": 1.6723997777614574e-07,
      "loss": 0.1203,
      "num_input_tokens_seen": 3366152,
      "step": 5425
    },
    {
      "epoch": 9.679144385026738,
      "grad_norm": 0.0194091796875,
      "learning_rate": 1.5837881149090294e-07,
      "loss": 0.0005,
      "num_input_tokens_seen": 3369192,
      "step": 5430
    },
    {
      "epoch": 9.688057040998217,
      "grad_norm": 0.09716796875,
      "learning_rate": 1.497580862453829e-07,
      "loss": 0.2422,
      "num_input_tokens_seen": 3372776,
      "step": 5435
    },
    {
      "epoch": 9.696969696969697,
      "grad_norm": 19.5,
      "learning_rate": 1.4137788547923246e-07,
      "loss": 0.0752,
      "num_input_tokens_seen": 3376232,
      "step": 5440
    },
    {
      "epoch": 9.705882352941176,
      "grad_norm": 1.8359375,
      "learning_rate": 1.3323829030407465e-07,
      "loss": 0.067,
      "num_input_tokens_seen": 3379912,
      "step": 5445
    },
    {
      "epoch": 9.714795008912656,
      "grad_norm": 29.75,
      "learning_rate": 1.2533937950272023e-07,
      "loss": 0.1758,
      "num_input_tokens_seen": 3382824,
      "step": 5450
    },
    {
      "epoch": 9.723707664884136,
      "grad_norm": 0.29296875,
      "learning_rate": 1.176812295283991e-07,
      "loss": 0.0762,
      "num_input_tokens_seen": 3385640,
      "step": 5455
    },
    {
      "epoch": 9.732620320855615,
      "grad_norm": 0.13671875,
      "learning_rate": 1.1026391450404128e-07,
      "loss": 0.0657,
      "num_input_tokens_seen": 3389672,
      "step": 5460
    },
    {
      "epoch": 9.741532976827095,
      "grad_norm": 0.359375,
      "learning_rate": 1.0308750622153307e-07,
      "loss": 0.1683,
      "num_input_tokens_seen": 3393096,
      "step": 5465
    },
    {
      "epoch": 9.750445632798574,
      "grad_norm": 0.05859375,
      "learning_rate": 9.615207414103434e-08,
      "loss": 0.0005,
      "num_input_tokens_seen": 3396136,
      "step": 5470
    },
    {
      "epoch": 9.759358288770054,
      "grad_norm": 20.125,
      "learning_rate": 8.945768539031785e-08,
      "loss": 0.147,
      "num_input_tokens_seen": 3399304,
      "step": 5475
    },
    {
      "epoch": 9.768270944741532,
      "grad_norm": 0.1484375,
      "learning_rate": 8.30044047640921e-08,
      "loss": 0.0234,
      "num_input_tokens_seen": 3402216,
      "step": 5480
    },
    {
      "epoch": 9.777183600713013,
      "grad_norm": 0.11572265625,
      "learning_rate": 7.679229472340176e-08,
      "loss": 0.0406,
      "num_input_tokens_seen": 3405096,
      "step": 5485
    },
    {
      "epoch": 9.786096256684491,
      "grad_norm": 26.25,
      "learning_rate": 7.082141539500597e-08,
      "loss": 0.1852,
      "num_input_tokens_seen": 3407912,
      "step": 5490
    },
    {
      "epoch": 9.795008912655971,
      "grad_norm": 0.011962890625,
      "learning_rate": 6.509182457080376e-08,
      "loss": 0.0003,
      "num_input_tokens_seen": 3410856,
      "step": 5495
    },
    {
      "epoch": 9.803921568627452,
      "grad_norm": 0.052978515625,
      "learning_rate": 5.9603577707267875e-08,
      "loss": 0.0123,
      "num_input_tokens_seen": 3413928,
      "step": 5500
    },
    {
      "epoch": 9.81283422459893,
      "grad_norm": 0.0220947265625,
      "learning_rate": 5.435672792491742e-08,
      "loss": 0.0107,
      "num_input_tokens_seen": 3417416,
      "step": 5505
    },
    {
      "epoch": 9.82174688057041,
      "grad_norm": 0.388671875,
      "learning_rate": 4.935132600780157e-08,
      "loss": 0.0133,
      "num_input_tokens_seen": 3420136,
      "step": 5510
    },
    {
      "epoch": 9.830659536541889,
      "grad_norm": 18.375,
      "learning_rate": 4.4587420402997235e-08,
      "loss": 0.0713,
      "num_input_tokens_seen": 3423272,
      "step": 5515
    },
    {
      "epoch": 9.83957219251337,
      "grad_norm": 0.2216796875,
      "learning_rate": 4.006505722015386e-08,
      "loss": 0.0024,
      "num_input_tokens_seen": 3426472,
      "step": 5520
    },
    {
      "epoch": 9.848484848484848,
      "grad_norm": 0.1748046875,
      "learning_rate": 3.578428023103819e-08,
      "loss": 0.024,
      "num_input_tokens_seen": 3429992,
      "step": 5525
    },
    {
      "epoch": 9.857397504456328,
      "grad_norm": 0.1357421875,
      "learning_rate": 3.1745130869123566e-08,
      "loss": 0.0727,
      "num_input_tokens_seen": 3432456,
      "step": 5530
    },
    {
      "epoch": 9.866310160427808,
      "grad_norm": 1.1484375,
      "learning_rate": 2.794764822916518e-08,
      "loss": 0.1221,
      "num_input_tokens_seen": 3434888,
      "step": 5535
    },
    {
      "epoch": 9.875222816399287,
      "grad_norm": 10.25,
      "learning_rate": 2.4391869066844874e-08,
      "loss": 0.1701,
      "num_input_tokens_seen": 3437832,
      "step": 5540
    },
    {
      "epoch": 9.884135472370767,
      "grad_norm": 0.01239013671875,
      "learning_rate": 2.1077827798404726e-08,
      "loss": 0.0002,
      "num_input_tokens_seen": 3440872,
      "step": 5545
    },
    {
      "epoch": 9.893048128342246,
      "grad_norm": 0.019287109375,
      "learning_rate": 1.8005556500313993e-08,
      "loss": 0.0609,
      "num_input_tokens_seen": 3443784,
      "step": 5550
    },
    {
      "epoch": 9.901960784313726,
      "grad_norm": 0.0634765625,
      "learning_rate": 1.51750849089638e-08,
      "loss": 0.095,
      "num_input_tokens_seen": 3447592,
      "step": 5555
    },
    {
      "epoch": 9.910873440285204,
      "grad_norm": 7.875,
      "learning_rate": 1.2586440420372936e-08,
      "loss": 0.0665,
      "num_input_tokens_seen": 3451048,
      "step": 5560
    },
    {
      "epoch": 9.919786096256685,
      "grad_norm": 0.1435546875,
      "learning_rate": 1.023964808992417e-08,
      "loss": 0.1243,
      "num_input_tokens_seen": 3453928,
      "step": 5565
    },
    {
      "epoch": 9.928698752228165,
      "grad_norm": 0.09423828125,
      "learning_rate": 8.134730632125554e-09,
      "loss": 0.0082,
      "num_input_tokens_seen": 3456968,
      "step": 5570
    },
    {
      "epoch": 9.937611408199643,
      "grad_norm": 0.0279541015625,
      "learning_rate": 6.271708420385603e-09,
      "loss": 0.1228,
      "num_input_tokens_seen": 3460616,
      "step": 5575
    },
    {
      "epoch": 9.946524064171124,
      "grad_norm": 43.25,
      "learning_rate": 4.650599486827334e-09,
      "loss": 0.0615,
      "num_input_tokens_seen": 3463592,
      "step": 5580
    },
    {
      "epoch": 9.955436720142602,
      "grad_norm": 1.9453125,
      "learning_rate": 3.2714195220912013e-09,
      "loss": 0.0408,
      "num_input_tokens_seen": 3466888,
      "step": 5585
    },
    {
      "epoch": 9.964349376114082,
      "grad_norm": 0.1787109375,
      "learning_rate": 2.134181875204644e-09,
      "loss": 0.0049,
      "num_input_tokens_seen": 3470408,
      "step": 5590
    },
    {
      "epoch": 9.973262032085561,
      "grad_norm": 0.037841796875,
      "learning_rate": 1.2388975534460834e-09,
      "loss": 0.1008,
      "num_input_tokens_seen": 3473608,
      "step": 5595
    },
    {
      "epoch": 9.982174688057041,
      "grad_norm": 0.01239013671875,
      "learning_rate": 5.855752222366783e-10,
      "loss": 0.0005,
      "num_input_tokens_seen": 3476616,
      "step": 5600
    },
    {
      "epoch": 9.99108734402852,
      "grad_norm": 0.009765625,
      "learning_rate": 1.7422120505705686e-10,
      "loss": 0.0557,
      "num_input_tokens_seen": 3479624,
      "step": 5605
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.0159912109375,
      "learning_rate": 4.839483383478616e-12,
      "loss": 0.0016,
      "num_input_tokens_seen": 3481336,
      "step": 5610
    },
    {
      "epoch": 10.0,
      "num_input_tokens_seen": 3481336,
      "step": 5610,
      "total_flos": 1.5676298662753075e+17,
      "train_loss": 0.07543940818015696,
      "train_runtime": 1380.2197,
      "train_samples_per_second": 16.237,
      "train_steps_per_second": 4.065
    }
  ],
  "logging_steps": 5,
  "max_steps": 5610,
  "num_input_tokens_seen": 3481336,
  "num_train_epochs": 10,
  "save_steps": 281,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.5676298662753075e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}