diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,61563 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 8790,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00034129692832764505,
+      "grad_norm": 8.071894645690918,
+      "learning_rate": 0.0009998862343572241,
+      "loss": 12.1418,
+      "step": 1
+    },
+    {
+      "epoch": 0.0006825938566552901,
+      "grad_norm": 5.984577178955078,
+      "learning_rate": 0.0009997724687144482,
+      "loss": 10.9212,
+      "step": 2
+    },
+    {
+      "epoch": 0.0010238907849829352,
+      "grad_norm": 5.766450881958008,
+      "learning_rate": 0.0009996587030716723,
+      "loss": 10.3976,
+      "step": 3
+    },
+    {
+      "epoch": 0.0013651877133105802,
+      "grad_norm": 4.486149787902832,
+      "learning_rate": 0.0009995449374288964,
+      "loss": 9.6927,
+      "step": 4
+    },
+    {
+      "epoch": 0.0017064846416382253,
+      "grad_norm": 4.054387092590332,
+      "learning_rate": 0.0009994311717861205,
+      "loss": 9.4759,
+      "step": 5
+    },
+    {
+      "epoch": 0.0020477815699658703,
+      "grad_norm": 4.101088523864746,
+      "learning_rate": 0.0009993174061433449,
+      "loss": 9.2064,
+      "step": 6
+    },
+    {
+      "epoch": 0.002389078498293515,
+      "grad_norm": 4.606311798095703,
+      "learning_rate": 0.000999203640500569,
+      "loss": 8.5577,
+      "step": 7
+    },
+    {
+      "epoch": 0.0027303754266211604,
+      "grad_norm": 4.251020431518555,
+      "learning_rate": 0.000999089874857793,
+      "loss": 8.6099,
+      "step": 8
+    },
+    {
+      "epoch": 0.0030716723549488053,
+      "grad_norm": 4.296706199645996,
+      "learning_rate": 0.0009989761092150172,
+      "loss": 8.6605,
+      "step": 9
+    },
+    {
+      "epoch": 0.0034129692832764505,
+      "grad_norm": 4.185203552246094,
+      "learning_rate": 0.000998862343572241,
+      "loss": 8.4278,
+      "step": 10
+    },
+    {
+      "epoch": 0.0037542662116040954,
+      "grad_norm": 19.910564422607422,
+      "learning_rate": 0.0009987485779294652,
+      "loss": 6.7563,
+      "step": 11
+    },
+    {
+      "epoch": 0.004095563139931741,
+      "grad_norm": 4.908084392547607,
+      "learning_rate": 0.0009986348122866895,
+      "loss": 8.5025,
+      "step": 12
+    },
+    {
+      "epoch": 0.004436860068259386,
+      "grad_norm": 4.213999271392822,
+      "learning_rate": 0.0009985210466439136,
+      "loss": 8.3884,
+      "step": 13
+    },
+    {
+      "epoch": 0.00477815699658703,
+      "grad_norm": 5.283620834350586,
+      "learning_rate": 0.0009984072810011377,
+      "loss": 7.6757,
+      "step": 14
+    },
+    {
+      "epoch": 0.005119453924914676,
+      "grad_norm": 4.11642599105835,
+      "learning_rate": 0.0009982935153583618,
+      "loss": 8.2126,
+      "step": 15
+    },
+    {
+      "epoch": 0.005460750853242321,
+      "grad_norm": 3.9162049293518066,
+      "learning_rate": 0.000998179749715586,
+      "loss": 8.1421,
+      "step": 16
+    },
+    {
+      "epoch": 0.005802047781569966,
+      "grad_norm": 3.7291955947875977,
+      "learning_rate": 0.00099806598407281,
+      "loss": 8.4332,
+      "step": 17
+    },
+    {
+      "epoch": 0.0061433447098976105,
+      "grad_norm": 4.422487258911133,
+      "learning_rate": 0.0009979522184300341,
+      "loss": 7.373,
+      "step": 18
+    },
+    {
+      "epoch": 0.006484641638225256,
+      "grad_norm": 3.9871609210968018,
+      "learning_rate": 0.0009978384527872582,
+      "loss": 7.8555,
+      "step": 19
+    },
+    {
+      "epoch": 0.006825938566552901,
+      "grad_norm": 4.012209415435791,
+      "learning_rate": 0.0009977246871444823,
+      "loss": 7.9322,
+      "step": 20
+    },
+    {
+      "epoch": 0.007167235494880546,
+      "grad_norm": 3.9194388389587402,
+      "learning_rate": 0.0009976109215017064,
+      "loss": 7.5508,
+      "step": 21
+    },
+    {
+      "epoch": 0.007508532423208191,
+      "grad_norm": 3.990365743637085,
+      "learning_rate": 0.0009974971558589305,
+      "loss": 7.8027,
+      "step": 22
+    },
+    {
+      "epoch": 0.007849829351535836,
+      "grad_norm": 4.236116409301758,
+      "learning_rate": 0.0009973833902161549,
+      "loss": 7.9234,
+      "step": 23
+    },
+    {
+      "epoch": 0.008191126279863481,
+      "grad_norm": 9.924948692321777,
+      "learning_rate": 0.000997269624573379,
+      "loss": 7.0817,
+      "step": 24
+    },
+    {
+      "epoch": 0.008532423208191127,
+      "grad_norm": 7.816974639892578,
+      "learning_rate": 0.000997155858930603,
+      "loss": 6.7681,
+      "step": 25
+    },
+    {
+      "epoch": 0.008873720136518772,
+      "grad_norm": 4.295772075653076,
+      "learning_rate": 0.0009970420932878272,
+      "loss": 7.8656,
+      "step": 26
+    },
+    {
+      "epoch": 0.009215017064846417,
+      "grad_norm": 8.219182968139648,
+      "learning_rate": 0.0009969283276450513,
+      "loss": 5.6737,
+      "step": 27
+    },
+    {
+      "epoch": 0.00955631399317406,
+      "grad_norm": 6.722366809844971,
+      "learning_rate": 0.0009968145620022754,
+      "loss": 6.44,
+      "step": 28
+    },
+    {
+      "epoch": 0.009897610921501706,
+      "grad_norm": 4.0717597007751465,
+      "learning_rate": 0.0009967007963594995,
+      "loss": 7.4495,
+      "step": 29
+    },
+    {
+      "epoch": 0.010238907849829351,
+      "grad_norm": 3.9176180362701416,
+      "learning_rate": 0.0009965870307167236,
+      "loss": 7.2357,
+      "step": 30
+    },
+    {
+      "epoch": 0.010580204778156996,
+      "grad_norm": 4.101375102996826,
+      "learning_rate": 0.0009964732650739477,
+      "loss": 7.068,
+      "step": 31
+    },
+    {
+      "epoch": 0.010921501706484642,
+      "grad_norm": 3.5856902599334717,
+      "learning_rate": 0.0009963594994311718,
+      "loss": 7.6736,
+      "step": 32
+    },
+    {
+      "epoch": 0.011262798634812287,
+      "grad_norm": 4.2619194984436035,
+      "learning_rate": 0.000996245733788396,
+      "loss": 7.0183,
+      "step": 33
+    },
+    {
+      "epoch": 0.011604095563139932,
+      "grad_norm": 3.7279083728790283,
+      "learning_rate": 0.00099613196814562,
+      "loss": 7.1387,
+      "step": 34
+    },
+    {
+      "epoch": 0.011945392491467578,
+      "grad_norm": 3.478987216949463,
+      "learning_rate": 0.0009960182025028441,
+      "loss": 7.4298,
+      "step": 35
+    },
+    {
+      "epoch": 0.012286689419795221,
+      "grad_norm": 3.570570707321167,
+      "learning_rate": 0.0009959044368600682,
+      "loss": 7.0718,
+      "step": 36
+    },
+    {
+      "epoch": 0.012627986348122866,
+      "grad_norm": 3.5319583415985107,
+      "learning_rate": 0.0009957906712172923,
+      "loss": 7.0964,
+      "step": 37
+    },
+    {
+      "epoch": 0.012969283276450512,
+      "grad_norm": 3.5056099891662598,
+      "learning_rate": 0.0009956769055745164,
+      "loss": 7.1669,
+      "step": 38
+    },
+    {
+      "epoch": 0.013310580204778157,
+      "grad_norm": 3.48520827293396,
+      "learning_rate": 0.0009955631399317405,
+      "loss": 7.5944,
+      "step": 39
+    },
+    {
+      "epoch": 0.013651877133105802,
+      "grad_norm": 3.7257206439971924,
+      "learning_rate": 0.0009954493742889647,
+      "loss": 7.0551,
+      "step": 40
+    },
+    {
+      "epoch": 0.013993174061433447,
+      "grad_norm": 4.222928524017334,
+      "learning_rate": 0.000995335608646189,
+      "loss": 6.9363,
+      "step": 41
+    },
+    {
+      "epoch": 0.014334470989761093,
+      "grad_norm": 4.023824214935303,
+      "learning_rate": 0.000995221843003413,
+      "loss": 7.2418,
+      "step": 42
+    },
+    {
+      "epoch": 0.014675767918088738,
+      "grad_norm": 3.970456600189209,
+      "learning_rate": 0.0009951080773606372,
+      "loss": 6.9394,
+      "step": 43
+    },
+    {
+      "epoch": 0.015017064846416382,
+      "grad_norm": 3.697723627090454,
+      "learning_rate": 0.0009949943117178613,
+      "loss": 7.1638,
+      "step": 44
+    },
+    {
+      "epoch": 0.015358361774744027,
+      "grad_norm": 9.966476440429688,
+      "learning_rate": 0.0009948805460750854,
+      "loss": 6.5215,
+      "step": 45
+    },
+    {
+      "epoch": 0.015699658703071672,
+      "grad_norm": 4.007918357849121,
+      "learning_rate": 0.0009947667804323095,
+      "loss": 7.0861,
+      "step": 46
+    },
+    {
+      "epoch": 0.016040955631399317,
+      "grad_norm": 4.113846778869629,
+      "learning_rate": 0.0009946530147895336,
+      "loss": 7.4096,
+      "step": 47
+    },
+    {
+      "epoch": 0.016382252559726963,
+      "grad_norm": 3.8716022968292236,
+      "learning_rate": 0.0009945392491467577,
+      "loss": 7.1557,
+      "step": 48
+    },
+    {
+      "epoch": 0.016723549488054608,
+      "grad_norm": 3.5497727394104004,
+      "learning_rate": 0.0009944254835039818,
+      "loss": 7.3205,
+      "step": 49
+    },
+    {
+      "epoch": 0.017064846416382253,
+      "grad_norm": 3.501861810684204,
+      "learning_rate": 0.000994311717861206,
+      "loss": 7.1354,
+      "step": 50
+    },
+    {
+      "epoch": 0.0174061433447099,
+      "grad_norm": 3.8758769035339355,
+      "learning_rate": 0.00099419795221843,
+      "loss": 6.743,
+      "step": 51
+    },
+    {
+      "epoch": 0.017747440273037544,
+      "grad_norm": 4.309325218200684,
+      "learning_rate": 0.0009940841865756541,
+      "loss": 6.7807,
+      "step": 52
+    },
+    {
+      "epoch": 0.01808873720136519,
+      "grad_norm": 3.619455337524414,
+      "learning_rate": 0.0009939704209328782,
+      "loss": 7.1105,
+      "step": 53
+    },
+    {
+      "epoch": 0.018430034129692834,
+      "grad_norm": 3.5727694034576416,
+      "learning_rate": 0.0009938566552901023,
+      "loss": 7.1482,
+      "step": 54
+    },
+    {
+      "epoch": 0.01877133105802048,
+      "grad_norm": 11.776286125183105,
+      "learning_rate": 0.0009937428896473264,
+      "loss": 5.6497,
+      "step": 55
+    },
+    {
+      "epoch": 0.01911262798634812,
+      "grad_norm": 4.078109264373779,
+      "learning_rate": 0.0009936291240045505,
+      "loss": 7.2197,
+      "step": 56
+    },
+    {
+      "epoch": 0.019453924914675767,
+      "grad_norm": 4.522604942321777,
+      "learning_rate": 0.0009935153583617747,
+      "loss": 6.3764,
+      "step": 57
+    },
+    {
+      "epoch": 0.019795221843003412,
+      "grad_norm": 3.437112808227539,
+      "learning_rate": 0.000993401592718999,
+      "loss": 6.8425,
+      "step": 58
+    },
+    {
+      "epoch": 0.020136518771331057,
+      "grad_norm": 3.7361385822296143,
+      "learning_rate": 0.000993287827076223,
+      "loss": 7.3883,
+      "step": 59
+    },
+    {
+      "epoch": 0.020477815699658702,
+      "grad_norm": 3.4807488918304443,
+      "learning_rate": 0.0009931740614334472,
+      "loss": 7.353,
+      "step": 60
+    },
+    {
+      "epoch": 0.020819112627986348,
+      "grad_norm": 3.5185515880584717,
+      "learning_rate": 0.0009930602957906713,
+      "loss": 7.1191,
+      "step": 61
+    },
+    {
+      "epoch": 0.021160409556313993,
+      "grad_norm": 3.42722749710083,
+      "learning_rate": 0.0009929465301478954,
+      "loss": 7.0395,
+      "step": 62
+    },
+    {
+      "epoch": 0.021501706484641638,
+      "grad_norm": 3.5530056953430176,
+      "learning_rate": 0.0009928327645051195,
+      "loss": 7.005,
+      "step": 63
+    },
+    {
+      "epoch": 0.021843003412969283,
+      "grad_norm": 3.577638626098633,
+      "learning_rate": 0.0009927189988623436,
+      "loss": 7.1476,
+      "step": 64
+    },
+    {
+      "epoch": 0.02218430034129693,
+      "grad_norm": 4.04377555847168,
+      "learning_rate": 0.0009926052332195677,
+      "loss": 7.1881,
+      "step": 65
+    },
+    {
+      "epoch": 0.022525597269624574,
+      "grad_norm": 4.380084037780762,
+      "learning_rate": 0.0009924914675767918,
+      "loss": 6.6989,
+      "step": 66
+    },
+    {
+      "epoch": 0.02286689419795222,
+      "grad_norm": 4.379724502563477,
+      "learning_rate": 0.000992377701934016,
+      "loss": 7.276,
+      "step": 67
+    },
+    {
+      "epoch": 0.023208191126279865,
+      "grad_norm": 3.8870277404785156,
+      "learning_rate": 0.00099226393629124,
+      "loss": 7.2336,
+      "step": 68
+    },
+    {
+      "epoch": 0.02354948805460751,
+      "grad_norm": 3.6682233810424805,
+      "learning_rate": 0.0009921501706484641,
+      "loss": 6.7899,
+      "step": 69
+    },
+    {
+      "epoch": 0.023890784982935155,
+      "grad_norm": 3.732855796813965,
+      "learning_rate": 0.0009920364050056882,
+      "loss": 6.5689,
+      "step": 70
+    },
+    {
+      "epoch": 0.024232081911262797,
+      "grad_norm": 3.7168567180633545,
+      "learning_rate": 0.0009919226393629123,
+      "loss": 6.9181,
+      "step": 71
+    },
+    {
+      "epoch": 0.024573378839590442,
+      "grad_norm": 3.8344621658325195,
+      "learning_rate": 0.0009918088737201364,
+      "loss": 7.4924,
+      "step": 72
+    },
+    {
+      "epoch": 0.024914675767918087,
+      "grad_norm": 5.3828606605529785,
+      "learning_rate": 0.0009916951080773605,
+      "loss": 6.9739,
+      "step": 73
+    },
+    {
+      "epoch": 0.025255972696245733,
+      "grad_norm": 3.8628880977630615,
+      "learning_rate": 0.0009915813424345847,
+      "loss": 7.7034,
+      "step": 74
+    },
+    {
+      "epoch": 0.025597269624573378,
+      "grad_norm": 4.138500213623047,
+      "learning_rate": 0.000991467576791809,
+      "loss": 7.1116,
+      "step": 75
+    },
+    {
+      "epoch": 0.025938566552901023,
+      "grad_norm": 3.994074821472168,
+      "learning_rate": 0.000991353811149033,
+      "loss": 6.6436,
+      "step": 76
+    },
+    {
+      "epoch": 0.02627986348122867,
+      "grad_norm": 3.761909246444702,
+      "learning_rate": 0.0009912400455062572,
+      "loss": 6.9494,
+      "step": 77
+    },
+    {
+      "epoch": 0.026621160409556314,
+      "grad_norm": 3.97804594039917,
+      "learning_rate": 0.0009911262798634813,
+      "loss": 6.7642,
+      "step": 78
+    },
+    {
+      "epoch": 0.02696245733788396,
+      "grad_norm": 12.43903923034668,
+      "learning_rate": 0.0009910125142207054,
+      "loss": 7.3236,
+      "step": 79
+    },
+    {
+      "epoch": 0.027303754266211604,
+      "grad_norm": 4.261969566345215,
+      "learning_rate": 0.0009908987485779295,
+      "loss": 6.7744,
+      "step": 80
+    },
+    {
+      "epoch": 0.02764505119453925,
+      "grad_norm": 7.615736484527588,
+      "learning_rate": 0.0009907849829351536,
+      "loss": 6.103,
+      "step": 81
+    },
+    {
+      "epoch": 0.027986348122866895,
+      "grad_norm": 3.833221197128296,
+      "learning_rate": 0.0009906712172923777,
+      "loss": 6.8201,
+      "step": 82
+    },
+    {
+      "epoch": 0.02832764505119454,
+      "grad_norm": 3.7682363986968994,
+      "learning_rate": 0.0009905574516496018,
+      "loss": 7.2732,
+      "step": 83
+    },
+    {
+      "epoch": 0.028668941979522185,
+      "grad_norm": 3.788194179534912,
+      "learning_rate": 0.000990443686006826,
+      "loss": 6.8951,
+      "step": 84
+    },
+    {
+      "epoch": 0.02901023890784983,
+      "grad_norm": 3.743475914001465,
+      "learning_rate": 0.00099032992036405,
+      "loss": 7.4113,
+      "step": 85
+    },
+    {
+      "epoch": 0.029351535836177476,
+      "grad_norm": 6.027090072631836,
+      "learning_rate": 0.0009902161547212743,
+      "loss": 6.7511,
+      "step": 86
+    },
+    {
+      "epoch": 0.029692832764505118,
+      "grad_norm": 11.415129661560059,
+      "learning_rate": 0.0009901023890784985,
+      "loss": 6.9999,
+      "step": 87
+    },
+    {
+      "epoch": 0.030034129692832763,
+      "grad_norm": 3.9674999713897705,
+      "learning_rate": 0.0009899886234357223,
+      "loss": 7.0998,
+      "step": 88
+    },
+    {
+      "epoch": 0.03037542662116041,
+      "grad_norm": 2.7847208976745605,
+      "learning_rate": 0.0009898748577929464,
+      "loss": 3.5909,
+      "step": 89
+    },
+    {
+      "epoch": 0.030716723549488054,
+      "grad_norm": 4.207309722900391,
+      "learning_rate": 0.0009897610921501705,
+      "loss": 6.5897,
+      "step": 90
+    },
+    {
+      "epoch": 0.0310580204778157,
+      "grad_norm": 3.6647539138793945,
+      "learning_rate": 0.0009896473265073947,
+      "loss": 7.064,
+      "step": 91
+    },
+    {
+      "epoch": 0.031399317406143344,
+      "grad_norm": 3.664642572402954,
+      "learning_rate": 0.000989533560864619,
+      "loss": 6.9888,
+      "step": 92
+    },
+    {
+      "epoch": 0.03174061433447099,
+      "grad_norm": 3.5732812881469727,
+      "learning_rate": 0.000989419795221843,
+      "loss": 6.7395,
+      "step": 93
+    },
+    {
+      "epoch": 0.032081911262798635,
+      "grad_norm": 3.5871763229370117,
+      "learning_rate": 0.0009893060295790672,
+      "loss": 7.214,
+      "step": 94
+    },
+    {
+      "epoch": 0.032423208191126277,
+      "grad_norm": 7.773305892944336,
+      "learning_rate": 0.0009891922639362913,
+      "loss": 6.2924,
+      "step": 95
+    },
+    {
+      "epoch": 0.032764505119453925,
+      "grad_norm": 4.1082563400268555,
+      "learning_rate": 0.0009890784982935154,
+      "loss": 7.0112,
+      "step": 96
+    },
+    {
+      "epoch": 0.03310580204778157,
+      "grad_norm": 3.709672212600708,
+      "learning_rate": 0.0009889647326507395,
+      "loss": 6.6266,
+      "step": 97
+    },
+    {
+      "epoch": 0.033447098976109216,
+      "grad_norm": 5.293731212615967,
+      "learning_rate": 0.0009888509670079636,
+      "loss": 5.1656,
+      "step": 98
+    },
+    {
+      "epoch": 0.03378839590443686,
+      "grad_norm": 3.9541916847229004,
+      "learning_rate": 0.0009887372013651877,
+      "loss": 7.2169,
+      "step": 99
+    },
+    {
+      "epoch": 0.034129692832764506,
+      "grad_norm": 4.045575141906738,
+      "learning_rate": 0.0009886234357224118,
+      "loss": 6.9998,
+      "step": 100
+    },
+    {
+      "epoch": 0.03447098976109215,
+      "grad_norm": 3.9652297496795654,
+      "learning_rate": 0.000988509670079636,
+      "loss": 7.1183,
+      "step": 101
+    },
+    {
+      "epoch": 0.0348122866894198,
+      "grad_norm": 3.934983730316162,
+      "learning_rate": 0.00098839590443686,
+      "loss": 7.133,
+      "step": 102
+    },
+    {
+      "epoch": 0.03515358361774744,
+      "grad_norm": 3.6193904876708984,
+      "learning_rate": 0.0009882821387940843,
+      "loss": 7.1528,
+      "step": 103
+    },
+    {
+      "epoch": 0.03549488054607509,
+      "grad_norm": 3.532212018966675,
+      "learning_rate": 0.0009881683731513085,
+      "loss": 7.3217,
+      "step": 104
+    },
+    {
+      "epoch": 0.03583617747440273,
+      "grad_norm": 5.649550437927246,
+      "learning_rate": 0.0009880546075085326,
+      "loss": 6.6923,
+      "step": 105
+    },
+    {
+      "epoch": 0.03617747440273038,
+      "grad_norm": 3.867431402206421,
+      "learning_rate": 0.0009879408418657567,
+      "loss": 6.7362,
+      "step": 106
+    },
+    {
+      "epoch": 0.03651877133105802,
+      "grad_norm": 3.762444019317627,
+      "learning_rate": 0.0009878270762229806,
+      "loss": 7.6837,
+      "step": 107
+    },
+    {
+      "epoch": 0.03686006825938567,
+      "grad_norm": 3.932126522064209,
+      "learning_rate": 0.0009877133105802047,
+      "loss": 7.0018,
+      "step": 108
+    },
+    {
+      "epoch": 0.03720136518771331,
+      "grad_norm": 5.752960205078125,
+      "learning_rate": 0.000987599544937429,
+      "loss": 6.8083,
+      "step": 109
+    },
+    {
+      "epoch": 0.03754266211604096,
+      "grad_norm": 3.6438000202178955,
+      "learning_rate": 0.000987485779294653,
+      "loss": 6.7968,
+      "step": 110
+    },
+    {
+      "epoch": 0.0378839590443686,
+      "grad_norm": 3.9477944374084473,
+      "learning_rate": 0.0009873720136518772,
+      "loss": 6.9717,
+      "step": 111
+    },
+    {
+      "epoch": 0.03822525597269624,
+      "grad_norm": 3.776455879211426,
+      "learning_rate": 0.0009872582480091013,
+      "loss": 6.5417,
+      "step": 112
+    },
+    {
+      "epoch": 0.03856655290102389,
+      "grad_norm": 4.05007791519165,
+      "learning_rate": 0.0009871444823663254,
+      "loss": 7.0061,
+      "step": 113
+    },
+    {
+      "epoch": 0.03890784982935153,
+      "grad_norm": 3.8773951530456543,
+      "learning_rate": 0.0009870307167235495,
+      "loss": 7.1258,
+      "step": 114
+    },
+    {
+      "epoch": 0.03924914675767918,
+      "grad_norm": 3.6618783473968506,
+      "learning_rate": 0.0009869169510807736,
+      "loss": 6.6578,
+      "step": 115
+    },
+    {
+      "epoch": 0.039590443686006824,
+      "grad_norm": 3.705451726913452,
+      "learning_rate": 0.0009868031854379977,
+      "loss": 7.15,
+      "step": 116
+    },
+    {
+      "epoch": 0.03993174061433447,
+      "grad_norm": 3.6103193759918213,
+      "learning_rate": 0.0009866894197952218,
+      "loss": 7.2358,
+      "step": 117
+    },
+    {
+      "epoch": 0.040273037542662114,
+      "grad_norm": 3.8052639961242676,
+      "learning_rate": 0.000986575654152446,
+      "loss": 6.7413,
+      "step": 118
+    },
+    {
+      "epoch": 0.04061433447098976,
+      "grad_norm": 3.7359206676483154,
+      "learning_rate": 0.00098646188850967,
+      "loss": 6.8834,
+      "step": 119
+    },
+    {
+      "epoch": 0.040955631399317405,
+      "grad_norm": 3.7333805561065674,
+      "learning_rate": 0.0009863481228668941,
+      "loss": 6.7141,
+      "step": 120
+    },
+    {
+      "epoch": 0.041296928327645054,
+      "grad_norm": 3.592845916748047,
+      "learning_rate": 0.0009862343572241185,
+      "loss": 6.9167,
+      "step": 121
+    },
+    {
+      "epoch": 0.041638225255972695,
+      "grad_norm": 3.8049752712249756,
+      "learning_rate": 0.0009861205915813426,
+      "loss": 7.1732,
+      "step": 122
+    },
+    {
+      "epoch": 0.041979522184300344,
+      "grad_norm": 3.7017672061920166,
+      "learning_rate": 0.0009860068259385667,
+      "loss": 7.1409,
+      "step": 123
+    },
+    {
+      "epoch": 0.042320819112627986,
+      "grad_norm": 3.581944704055786,
+      "learning_rate": 0.0009858930602957908,
+      "loss": 7.401,
+      "step": 124
+    },
+    {
+      "epoch": 0.042662116040955635,
+      "grad_norm": 3.60552716255188,
+      "learning_rate": 0.0009857792946530149,
+      "loss": 7.1492,
+      "step": 125
+    },
+    {
+      "epoch": 0.043003412969283276,
+      "grad_norm": 3.5588576793670654,
+      "learning_rate": 0.000985665529010239,
+      "loss": 7.2996,
+      "step": 126
+    },
+    {
+      "epoch": 0.04334470989761092,
+      "grad_norm": 3.566678524017334,
+      "learning_rate": 0.000985551763367463,
+      "loss": 6.7319,
+      "step": 127
+    },
+    {
+      "epoch": 0.04368600682593857,
+      "grad_norm": 5.839478492736816,
+      "learning_rate": 0.0009854379977246872,
+      "loss": 6.2529,
+      "step": 128
+    },
+    {
+      "epoch": 0.04402730375426621,
+      "grad_norm": 5.420695781707764,
+      "learning_rate": 0.0009853242320819113,
+      "loss": 6.3413,
+      "step": 129
+    },
+    {
+      "epoch": 0.04436860068259386,
+      "grad_norm": 4.142327785491943,
+      "learning_rate": 0.0009852104664391354,
+      "loss": 7.3027,
+      "step": 130
+    },
+    {
+      "epoch": 0.0447098976109215,
+      "grad_norm": 3.7694830894470215,
+      "learning_rate": 0.0009850967007963595,
+      "loss": 6.5612,
+      "step": 131
+    },
+    {
+      "epoch": 0.04505119453924915,
+      "grad_norm": 3.7021262645721436,
+      "learning_rate": 0.0009849829351535836,
+      "loss": 7.1486,
+      "step": 132
+    },
+    {
+      "epoch": 0.04539249146757679,
+      "grad_norm": 3.4137067794799805,
+      "learning_rate": 0.0009848691695108077,
+      "loss": 7.3512,
+      "step": 133
+    },
+    {
+      "epoch": 0.04573378839590444,
+      "grad_norm": 3.9189088344573975,
+      "learning_rate": 0.0009847554038680318,
+      "loss": 6.877,
+      "step": 134
+    },
+    {
+      "epoch": 0.04607508532423208,
+      "grad_norm": 3.598252534866333,
+      "learning_rate": 0.000984641638225256,
+      "loss": 7.1513,
+      "step": 135
+    },
+    {
+      "epoch": 0.04641638225255973,
+      "grad_norm": 3.869384527206421,
+      "learning_rate": 0.00098452787258248,
+      "loss": 6.7868,
+      "step": 136
+    },
+    {
+      "epoch": 0.04675767918088737,
+      "grad_norm": 4.416536808013916,
+      "learning_rate": 0.0009844141069397041,
+      "loss": 6.3465,
+      "step": 137
+    },
+    {
+      "epoch": 0.04709897610921502,
+      "grad_norm": 3.654963493347168,
+      "learning_rate": 0.0009843003412969285,
+      "loss": 7.4472,
+      "step": 138
+    },
+    {
+      "epoch": 0.04744027303754266,
+      "grad_norm": 3.6647439002990723,
+      "learning_rate": 0.0009841865756541526,
+      "loss": 7.0507,
+      "step": 139
+    },
+    {
+      "epoch": 0.04778156996587031,
+      "grad_norm": 3.4522414207458496,
+      "learning_rate": 0.0009840728100113767,
+      "loss": 7.1764,
+      "step": 140
+    },
+    {
+      "epoch": 0.04812286689419795,
+      "grad_norm": 3.3954365253448486,
+      "learning_rate": 0.0009839590443686008,
+      "loss": 6.9876,
+      "step": 141
+    },
+    {
+      "epoch": 0.048464163822525594,
+      "grad_norm": 3.482323169708252,
+      "learning_rate": 0.0009838452787258249,
+      "loss": 6.5746,
+      "step": 142
+    },
+    {
+      "epoch": 0.04880546075085324,
+      "grad_norm": 3.477513551712036,
+      "learning_rate": 0.000983731513083049,
+      "loss": 7.2068,
+      "step": 143
+    },
+    {
+      "epoch": 0.049146757679180884,
+      "grad_norm": 3.6850838661193848,
+      "learning_rate": 0.000983617747440273,
+      "loss": 6.8805,
+      "step": 144
+    },
+    {
+      "epoch": 0.04948805460750853,
+      "grad_norm": 6.7132673263549805,
+      "learning_rate": 0.0009835039817974972,
+      "loss": 6.5654,
+      "step": 145
+    },
+    {
+      "epoch": 0.049829351535836175,
+      "grad_norm": 3.8716394901275635,
+      "learning_rate": 0.0009833902161547213,
+      "loss": 6.9398,
+      "step": 146
+    },
+    {
+      "epoch": 0.050170648464163824,
+      "grad_norm": 3.562126874923706,
+      "learning_rate": 0.0009832764505119454,
+      "loss": 6.804,
+      "step": 147
+    },
+    {
+      "epoch": 0.050511945392491465,
+      "grad_norm": 3.6806352138519287,
+      "learning_rate": 0.0009831626848691695,
+      "loss": 7.0423,
+      "step": 148
+    },
+    {
+      "epoch": 0.050853242320819114,
+      "grad_norm": 5.191007614135742,
+      "learning_rate": 0.0009830489192263936,
+      "loss": 5.9842,
+      "step": 149
+    },
+    {
+      "epoch": 0.051194539249146756,
+      "grad_norm": 3.838003396987915,
+      "learning_rate": 0.0009829351535836177,
+      "loss": 7.0365,
+      "step": 150
+    },
+    {
+      "epoch": 0.051535836177474405,
+      "grad_norm": 4.209732532501221,
+      "learning_rate": 0.0009828213879408418,
+      "loss": 6.7704,
+      "step": 151
+    },
+    {
+      "epoch": 0.05187713310580205,
+      "grad_norm": 3.973942995071411,
+      "learning_rate": 0.000982707622298066,
+      "loss": 6.5295,
+      "step": 152
+    },
+    {
+      "epoch": 0.052218430034129695,
+      "grad_norm": 3.3907649517059326,
+      "learning_rate": 0.00098259385665529,
+      "loss": 7.0298,
+      "step": 153
+    },
+    {
+      "epoch": 0.05255972696245734,
+      "grad_norm": 3.6388776302337646,
+      "learning_rate": 0.0009824800910125141,
+      "loss": 6.7746,
+      "step": 154
+    },
+    {
+      "epoch": 0.052901023890784986,
+      "grad_norm": 3.418466091156006,
+      "learning_rate": 0.0009823663253697385,
+      "loss": 6.9274,
+      "step": 155
+    },
+    {
+      "epoch": 0.05324232081911263,
+      "grad_norm": 4.016181945800781,
+      "learning_rate": 0.0009822525597269626,
+      "loss": 6.647,
+      "step": 156
+    },
+    {
+      "epoch": 0.053583617747440276,
+      "grad_norm": 3.4281997680664062,
+      "learning_rate": 0.0009821387940841867,
+      "loss": 6.594,
+      "step": 157
+    },
+    {
+      "epoch": 0.05392491467576792,
+      "grad_norm": 3.6327078342437744,
+      "learning_rate": 0.0009820250284414108,
+      "loss": 6.8371,
+      "step": 158
+    },
+    {
+      "epoch": 0.05426621160409556,
+      "grad_norm": 3.937331438064575,
+      "learning_rate": 0.0009819112627986349,
+      "loss": 6.6783,
+      "step": 159
+    },
+    {
+      "epoch": 0.05460750853242321,
+      "grad_norm": 5.26201057434082,
+      "learning_rate": 0.000981797497155859,
+      "loss": 5.6406,
+      "step": 160
+    },
+    {
+      "epoch": 0.05494880546075085,
+      "grad_norm": 3.728435516357422,
+      "learning_rate": 0.000981683731513083,
+      "loss": 7.1284,
+      "step": 161
+    },
+    {
+      "epoch": 0.0552901023890785,
+      "grad_norm": 3.79921555519104,
+      "learning_rate": 0.0009815699658703072,
+      "loss": 6.7235,
+      "step": 162
+    },
+    {
+      "epoch": 0.05563139931740614,
+      "grad_norm": 3.5397472381591797,
+      "learning_rate": 0.0009814562002275313,
+      "loss": 6.6777,
+      "step": 163
+    },
+    {
+      "epoch": 0.05597269624573379,
+      "grad_norm": 3.5530202388763428,
+      "learning_rate": 0.0009813424345847554,
+      "loss": 6.4635,
+      "step": 164
+    },
+    {
+      "epoch": 0.05631399317406143,
+      "grad_norm": 4.582598686218262,
+      "learning_rate": 0.0009812286689419795,
+      "loss": 6.3199,
+      "step": 165
+    },
+    {
+      "epoch": 0.05665529010238908,
+      "grad_norm": 4.460880279541016,
+      "learning_rate": 0.0009811149032992036,
+      "loss": 5.5054,
+      "step": 166
+    },
+    {
+      "epoch": 0.05699658703071672,
+      "grad_norm": 3.812800168991089,
+      "learning_rate": 0.0009810011376564277,
+      "loss": 6.8298,
+      "step": 167
+    },
+    {
+      "epoch": 0.05733788395904437,
+      "grad_norm": 3.747919797897339,
+      "learning_rate": 0.0009808873720136518,
+      "loss": 6.7751,
+      "step": 168
+    },
+    {
+      "epoch": 0.05767918088737201,
+      "grad_norm": 6.039458274841309,
+      "learning_rate": 0.000980773606370876,
+      "loss": 5.7825,
+      "step": 169
+    },
+    {
+      "epoch": 0.05802047781569966,
+      "grad_norm": 3.8591084480285645,
+      "learning_rate": 0.0009806598407281,
+      "loss": 6.8238,
+      "step": 170
+    },
+    {
+      "epoch": 0.0583617747440273,
+      "grad_norm": 3.8271124362945557,
+      "learning_rate": 0.0009805460750853241,
+      "loss": 6.3032,
+      "step": 171
+    },
+    {
+      "epoch": 0.05870307167235495,
+      "grad_norm": 3.730949640274048,
+      "learning_rate": 0.0009804323094425485,
+      "loss": 6.9856,
+      "step": 172
+    },
+    {
+      "epoch": 0.059044368600682594,
+      "grad_norm": 5.848387718200684,
+      "learning_rate": 0.0009803185437997726,
+      "loss": 5.8503,
+      "step": 173
+    },
+    {
+      "epoch": 0.059385665529010236,
+      "grad_norm": 5.451254844665527,
+      "learning_rate": 0.0009802047781569967,
+      "loss": 6.6104,
+      "step": 174
+    },
+    {
+      "epoch": 0.059726962457337884,
+      "grad_norm": 4.356447219848633,
+      "learning_rate": 0.0009800910125142208,
+      "loss": 7.2519,
+      "step": 175
+    },
+    {
+      "epoch": 0.060068259385665526,
+      "grad_norm": 3.5853893756866455,
+      "learning_rate": 0.0009799772468714449,
+      "loss": 6.8436,
+      "step": 176
+    },
+    {
+      "epoch": 0.060409556313993175,
+      "grad_norm": 3.9135236740112305,
+      "learning_rate": 0.000979863481228669,
+      "loss": 6.7005,
+      "step": 177
+    },
+    {
+      "epoch": 0.06075085324232082,
+      "grad_norm": 3.59726881980896,
+      "learning_rate": 0.000979749715585893,
+      "loss": 6.747,
+      "step": 178
+    },
+    {
+      "epoch": 0.061092150170648465,
+      "grad_norm": 4.7263593673706055,
+      "learning_rate": 0.0009796359499431172,
+      "loss": 5.5305,
+      "step": 179
+    },
+    {
+      "epoch": 0.06143344709897611,
+      "grad_norm": 3.808228015899658,
+      "learning_rate": 0.0009795221843003413,
+      "loss": 7.0182,
+      "step": 180
+    },
+    {
+      "epoch": 0.061774744027303756,
+      "grad_norm": 3.7145042419433594,
+      "learning_rate": 0.0009794084186575654,
+      "loss": 6.8614,
+      "step": 181
+    },
+    {
+      "epoch": 0.0621160409556314,
+      "grad_norm": 3.560530424118042,
+      "learning_rate": 0.0009792946530147895,
+      "loss": 6.8768,
+      "step": 182
+    },
+    {
+      "epoch": 0.062457337883959047,
+      "grad_norm": 3.480214834213257,
+      "learning_rate": 0.0009791808873720136,
+      "loss": 6.9845,
+      "step": 183
+    },
+    {
+      "epoch": 0.06279863481228669,
+      "grad_norm": 3.4854085445404053,
+      "learning_rate": 0.000979067121729238,
+      "loss": 6.9272,
+      "step": 184
+    },
+    {
+      "epoch": 0.06313993174061433,
+      "grad_norm": 3.636730432510376,
+      "learning_rate": 0.0009789533560864618,
+      "loss": 6.84,
+      "step": 185
+    },
+    {
+      "epoch": 0.06348122866894199,
+      "grad_norm": 3.543924331665039,
+      "learning_rate": 0.000978839590443686,
+      "loss": 6.9787,
+      "step": 186
+    },
+    {
+      "epoch": 0.06382252559726963,
+      "grad_norm": 3.629248857498169,
+      "learning_rate": 0.00097872582480091,
+      "loss": 6.7263,
+      "step": 187
+    },
+    {
+      "epoch": 0.06416382252559727,
+      "grad_norm": 3.5820744037628174,
+      "learning_rate": 0.0009786120591581341,
+      "loss": 6.6454,
+      "step": 188
+    },
+    {
+      "epoch": 0.06450511945392491,
+      "grad_norm": 6.293400764465332,
+      "learning_rate": 0.0009784982935153585,
+      "loss": 4.73,
+      "step": 189
+    },
+    {
+      "epoch": 0.06484641638225255,
+      "grad_norm": 6.925587177276611,
+      "learning_rate": 0.0009783845278725826,
+      "loss": 4.6797,
+      "step": 190
+    },
+    {
+      "epoch": 0.06518771331058021,
+      "grad_norm": 4.270597457885742,
+      "learning_rate": 0.0009782707622298067,
+      "loss": 6.9059,
+      "step": 191
+    },
+    {
+      "epoch": 0.06552901023890785,
+      "grad_norm": 3.773348808288574,
+      "learning_rate": 0.0009781569965870308,
+      "loss": 7.0965,
+      "step": 192
+    },
+    {
+      "epoch": 0.06587030716723549,
+      "grad_norm": 3.767024517059326,
+      "learning_rate": 0.0009780432309442549,
+      "loss": 6.7144,
+      "step": 193
+    },
+    {
+      "epoch": 0.06621160409556313,
+      "grad_norm": 3.9505908489227295,
+      "learning_rate": 0.000977929465301479,
+      "loss": 6.8606,
+      "step": 194
+    },
+    {
+      "epoch": 0.06655290102389079,
+      "grad_norm": 3.617802381515503,
+      "learning_rate": 0.000977815699658703,
+      "loss": 6.7616,
+      "step": 195
+    },
+    {
+      "epoch": 0.06689419795221843,
+      "grad_norm": 3.433324098587036,
+      "learning_rate": 0.0009777019340159272,
+      "loss": 6.7047,
+      "step": 196
+    },
+    {
+      "epoch": 0.06723549488054607,
+      "grad_norm": 3.5975823402404785,
+      "learning_rate": 0.0009775881683731513,
+      "loss": 6.5858,
+      "step": 197
+    },
+    {
+      "epoch": 0.06757679180887372,
+      "grad_norm": 3.548548936843872,
+      "learning_rate": 0.0009774744027303754,
+      "loss": 7.1213,
+      "step": 198
+    },
+    {
+      "epoch": 0.06791808873720137,
+      "grad_norm": 3.8470911979675293,
+      "learning_rate": 0.0009773606370875995,
+      "loss": 6.3278,
+      "step": 199
+    },
+    {
+      "epoch": 0.06825938566552901,
+      "grad_norm": 3.527210235595703,
+      "learning_rate": 0.0009772468714448236,
+      "loss": 6.8947,
+      "step": 200
+    },
+    {
+      "epoch": 0.06860068259385665,
+      "grad_norm": 3.4403419494628906,
+      "learning_rate": 0.000977133105802048,
+      "loss": 6.738,
+      "step": 201
+    },
+    {
+      "epoch": 0.0689419795221843,
+      "grad_norm": 3.733076810836792,
+      "learning_rate": 0.000977019340159272,
+      "loss": 6.9729,
+      "step": 202
+    },
+    {
+      "epoch": 0.06928327645051195,
+      "grad_norm": 3.4491777420043945,
+      "learning_rate": 0.0009769055745164961,
+      "loss": 7.0452,
+      "step": 203
+    },
+    {
+      "epoch": 0.0696245733788396,
+      "grad_norm": 12.765624046325684,
+      "learning_rate": 0.0009767918088737202,
+      "loss": 6.6967,
+      "step": 204
+    },
+    {
+      "epoch": 0.06996587030716724,
+      "grad_norm": 4.026791095733643,
+      "learning_rate": 0.0009766780432309441,
+      "loss": 6.9201,
+      "step": 205
+    },
+    {
+      "epoch": 0.07030716723549488,
+      "grad_norm": 6.290328502655029,
+      "learning_rate": 0.0009765642775881682,
+      "loss": 6.3519,
+      "step": 206
+    },
+    {
+      "epoch": 0.07064846416382252,
+      "grad_norm": 3.867109775543213,
+      "learning_rate": 0.0009764505119453925,
+      "loss": 7.309,
+      "step": 207
+    },
+    {
+      "epoch": 0.07098976109215017,
+      "grad_norm": 3.9246108531951904,
+      "learning_rate": 0.0009763367463026166,
+      "loss": 6.7805,
+      "step": 208
+    },
+    {
+      "epoch": 0.07133105802047782,
+      "grad_norm": 3.378211498260498,
+      "learning_rate": 0.0009762229806598408,
+      "loss": 7.007,
+      "step": 209
+    },
+    {
+      "epoch": 0.07167235494880546,
+      "grad_norm": 3.3665435314178467,
+      "learning_rate": 0.0009761092150170649,
+      "loss": 6.7892,
+      "step": 210
+    },
+    {
+      "epoch": 0.0720136518771331,
+      "grad_norm": 3.457585096359253,
+      "learning_rate": 0.000975995449374289,
+      "loss": 6.8816,
+      "step": 211
+    },
+    {
+      "epoch": 0.07235494880546076,
+      "grad_norm": 3.859353542327881,
+      "learning_rate": 0.0009758816837315131,
+      "loss": 5.9202,
+      "step": 212
+    },
+    {
+      "epoch": 0.0726962457337884,
+      "grad_norm": 5.0818190574646,
+      "learning_rate": 0.0009757679180887372,
+      "loss": 6.0906,
+      "step": 213
+    },
+    {
+      "epoch": 0.07303754266211604,
+      "grad_norm": 3.940415620803833,
+      "learning_rate": 0.0009756541524459613,
+      "loss": 7.1606,
+      "step": 214
+    },
+    {
+      "epoch": 0.07337883959044368,
+      "grad_norm": 9.891190528869629,
+      "learning_rate": 0.0009755403868031855,
+      "loss": 4.2678,
+      "step": 215
+    },
+    {
+      "epoch": 0.07372013651877134,
+      "grad_norm": 4.019010543823242,
+      "learning_rate": 0.0009754266211604096,
+      "loss": 6.3178,
+      "step": 216
+    },
+    {
+      "epoch": 0.07406143344709898,
+      "grad_norm": 3.7409043312072754,
+      "learning_rate": 0.0009753128555176337,
+      "loss": 6.9846,
+      "step": 217
+    },
+    {
+      "epoch": 0.07440273037542662,
+      "grad_norm": 3.7472164630889893,
+      "learning_rate": 0.0009751990898748578,
+      "loss": 6.3952,
+      "step": 218
+    },
+    {
+      "epoch": 0.07474402730375426,
+      "grad_norm": 3.981301784515381,
+      "learning_rate": 0.0009750853242320819,
+      "loss": 6.888,
+      "step": 219
+    },
+    {
+      "epoch": 0.07508532423208192,
+      "grad_norm": 7.142719268798828,
+      "learning_rate": 0.000974971558589306,
+      "loss": 6.4133,
+      "step": 220
+    },
+    {
+      "epoch": 0.07542662116040956,
+      "grad_norm": 3.712938070297241,
+      "learning_rate": 0.0009748577929465302,
+      "loss": 7.2119,
+      "step": 221
+    },
+    {
+      "epoch": 0.0757679180887372,
+      "grad_norm": 3.7851550579071045,
+      "learning_rate": 0.0009747440273037544,
+      "loss": 6.5805,
+      "step": 222
+    },
+    {
+      "epoch": 0.07610921501706484,
+      "grad_norm": 3.5392749309539795,
+      "learning_rate": 0.0009746302616609785,
+      "loss": 6.6098,
+      "step": 223
+    },
+    {
+      "epoch": 0.07645051194539249,
+      "grad_norm": 4.032394886016846,
+      "learning_rate": 0.0009745164960182025,
+      "loss": 6.7128,
+      "step": 224
+    },
+    {
+      "epoch": 0.07679180887372014,
+      "grad_norm": 3.6618125438690186,
+      "learning_rate": 0.0009744027303754266,
+      "loss": 6.6606,
+      "step": 225
+    },
+    {
+      "epoch": 0.07713310580204778,
+      "grad_norm": 3.458958148956299,
+      "learning_rate": 0.0009742889647326507,
+      "loss": 6.8682,
+      "step": 226
+    },
+    {
+      "epoch": 0.07747440273037542,
+      "grad_norm": 4.641584396362305,
+      "learning_rate": 0.0009741751990898749,
+      "loss": 6.1076,
+      "step": 227
+    },
+    {
+      "epoch": 0.07781569965870307,
+      "grad_norm": 3.7606310844421387,
+      "learning_rate": 0.000974061433447099,
+      "loss": 6.9889,
+      "step": 228
+    },
+    {
+      "epoch": 0.07815699658703072,
+      "grad_norm": 3.448821783065796,
+      "learning_rate": 0.0009739476678043231,
+      "loss": 6.9313,
+      "step": 229
+    },
+    {
+      "epoch": 0.07849829351535836,
+      "grad_norm": 3.49123477935791,
+      "learning_rate": 0.0009738339021615472,
+      "loss": 6.9526,
+      "step": 230
+    },
+    {
+      "epoch": 0.078839590443686,
+      "grad_norm": 3.5688812732696533,
+      "learning_rate": 0.0009737201365187713,
+      "loss": 6.5122,
+      "step": 231
+    },
+    {
+      "epoch": 0.07918088737201365,
+      "grad_norm": 3.416281223297119,
+      "learning_rate": 0.0009736063708759955,
+      "loss": 6.7054,
+      "step": 232
+    },
+    {
+      "epoch": 0.0795221843003413,
+      "grad_norm": 3.3496880531311035,
+      "learning_rate": 0.0009734926052332196,
+      "loss": 6.6948,
+      "step": 233
+    },
+    {
+      "epoch": 0.07986348122866894,
+      "grad_norm": 3.9867589473724365,
+      "learning_rate": 0.0009733788395904437,
+      "loss": 6.8821,
+      "step": 234
+    },
+    {
+      "epoch": 0.08020477815699659,
+      "grad_norm": 3.6225593090057373,
+      "learning_rate": 0.0009732650739476678,
+      "loss": 6.706,
+      "step": 235
+    },
+    {
+      "epoch": 0.08054607508532423,
+      "grad_norm": 3.7255892753601074,
+      "learning_rate": 0.0009731513083048919,
+      "loss": 7.0775,
+      "step": 236
+    },
+    {
+      "epoch": 0.08088737201365187,
+      "grad_norm": 8.664634704589844,
+      "learning_rate": 0.000973037542662116,
+      "loss": 5.4271,
+      "step": 237
+    },
+    {
+      "epoch": 0.08122866894197953,
+      "grad_norm": 4.9100117683410645,
+      "learning_rate": 0.0009729237770193402,
+      "loss": 6.4401,
+      "step": 238
+    },
+    {
+      "epoch": 0.08156996587030717,
+      "grad_norm": 3.7028720378875732,
+      "learning_rate": 0.0009728100113765644,
+      "loss": 6.3939,
+      "step": 239
+    },
+    {
+      "epoch": 0.08191126279863481,
+      "grad_norm": 3.713730573654175,
+      "learning_rate": 0.0009726962457337885,
+      "loss": 6.8395,
+      "step": 240
+    },
+    {
+      "epoch": 0.08225255972696245,
+      "grad_norm": 3.7303857803344727,
+      "learning_rate": 0.0009725824800910126,
+      "loss": 6.5482,
+      "step": 241
+    },
+    {
+      "epoch": 0.08259385665529011,
+      "grad_norm": 3.4762961864471436,
+      "learning_rate": 0.0009724687144482367,
+      "loss": 7.0664,
+      "step": 242
+    },
+    {
+      "epoch": 0.08293515358361775,
+      "grad_norm": 3.4974660873413086,
+      "learning_rate": 0.0009723549488054608,
+      "loss": 6.7011,
+      "step": 243
+    },
+    {
+      "epoch": 0.08327645051194539,
+      "grad_norm": 3.515742540359497,
+      "learning_rate": 0.0009722411831626849,
+      "loss": 7.2582,
+      "step": 244
+    },
+    {
+      "epoch": 0.08361774744027303,
+      "grad_norm": 3.654632806777954,
+      "learning_rate": 0.000972127417519909,
+      "loss": 6.6722,
+      "step": 245
+    },
+    {
+      "epoch": 0.08395904436860069,
+      "grad_norm": 3.580826759338379,
+      "learning_rate": 0.0009720136518771331,
+      "loss": 6.6658,
+      "step": 246
+    },
+    {
+      "epoch": 0.08430034129692833,
+      "grad_norm": 3.44158673286438,
+      "learning_rate": 0.0009718998862343572,
+      "loss": 6.6069,
+      "step": 247
+    },
+    {
+      "epoch": 0.08464163822525597,
+      "grad_norm": 3.856389045715332,
+      "learning_rate": 0.0009717861205915813,
+      "loss": 6.7927,
+      "step": 248
+    },
+    {
+      "epoch": 0.08498293515358361,
+      "grad_norm": 3.6192235946655273,
+      "learning_rate": 0.0009716723549488055,
+      "loss": 6.9655,
+      "step": 249
+    },
+    {
+      "epoch": 0.08532423208191127,
+      "grad_norm": 3.7248499393463135,
+      "learning_rate": 0.0009715585893060296,
+      "loss": 7.0733,
+      "step": 250
+    },
+    {
+      "epoch": 0.08566552901023891,
+      "grad_norm": 3.485734701156616,
+      "learning_rate": 0.0009714448236632537,
+      "loss": 6.7725,
+      "step": 251
+    },
+    {
+      "epoch": 0.08600682593856655,
+      "grad_norm": 3.442128896713257,
+      "learning_rate": 0.0009713310580204778,
+      "loss": 6.8307,
+      "step": 252
+    },
+    {
+      "epoch": 0.0863481228668942,
+      "grad_norm": 4.0273308753967285,
+      "learning_rate": 0.0009712172923777019,
+      "loss": 6.2919,
+      "step": 253
+    },
+    {
+      "epoch": 0.08668941979522184,
+      "grad_norm": 3.7515664100646973,
+      "learning_rate": 0.000971103526734926,
+      "loss": 6.7646,
+      "step": 254
+    },
+    {
+      "epoch": 0.08703071672354949,
+      "grad_norm": 3.6255035400390625,
+      "learning_rate": 0.0009709897610921503,
+      "loss": 7.1922,
+      "step": 255
+    },
+    {
+      "epoch": 0.08737201365187713,
+      "grad_norm": 3.9632773399353027,
+      "learning_rate": 0.0009708759954493744,
+      "loss": 6.6199,
+      "step": 256
+    },
+    {
+      "epoch": 0.08771331058020478,
+      "grad_norm": 3.5362298488616943,
+      "learning_rate": 0.0009707622298065985,
+      "loss": 6.9201,
+      "step": 257
+    },
+    {
+      "epoch": 0.08805460750853242,
+      "grad_norm": 3.4530086517333984,
+      "learning_rate": 0.0009706484641638226,
+      "loss": 6.8544,
+      "step": 258
+    },
+    {
+      "epoch": 0.08839590443686007,
+      "grad_norm": 3.586812973022461,
+      "learning_rate": 0.0009705346985210467,
+      "loss": 6.6807,
+      "step": 259
+    },
+    {
+      "epoch": 0.08873720136518772,
+      "grad_norm": 3.553516387939453,
+      "learning_rate": 0.0009704209328782708,
+      "loss": 7.0369,
+      "step": 260
+    },
+    {
+      "epoch": 0.08907849829351536,
+      "grad_norm": 3.6874141693115234,
+      "learning_rate": 0.000970307167235495,
+      "loss": 6.7392,
+      "step": 261
+    },
+    {
+      "epoch": 0.089419795221843,
+      "grad_norm": 3.5124759674072266,
+      "learning_rate": 0.0009701934015927191,
+      "loss": 6.6912,
+      "step": 262
+    },
+    {
+      "epoch": 0.08976109215017065,
+      "grad_norm": 3.7072601318359375,
+      "learning_rate": 0.0009700796359499431,
+      "loss": 6.5064,
+      "step": 263
+    },
+    {
+      "epoch": 0.0901023890784983,
+      "grad_norm": 4.079402446746826,
+      "learning_rate": 0.0009699658703071672,
+      "loss": 6.487,
+      "step": 264
+    },
+    {
+      "epoch": 0.09044368600682594,
+      "grad_norm": 3.6425037384033203,
+      "learning_rate": 0.0009698521046643913,
+      "loss": 6.7205,
+      "step": 265
+    },
+    {
+      "epoch": 0.09078498293515358,
+      "grad_norm": 3.5927422046661377,
+      "learning_rate": 0.0009697383390216154,
+      "loss": 6.6607,
+      "step": 266
+    },
+    {
+      "epoch": 0.09112627986348124,
+      "grad_norm": 3.606675624847412,
+      "learning_rate": 0.0009696245733788396,
+      "loss": 7.0384,
+      "step": 267
+    },
+    {
+      "epoch": 0.09146757679180888,
+      "grad_norm": 4.241934776306152,
+      "learning_rate": 0.0009695108077360637,
+      "loss": 6.8748,
+      "step": 268
+    },
+    {
+      "epoch": 0.09180887372013652,
+      "grad_norm": 3.801959276199341,
+      "learning_rate": 0.0009693970420932878,
+      "loss": 7.1313,
+      "step": 269
+    },
+    {
+      "epoch": 0.09215017064846416,
+      "grad_norm": 3.597132921218872,
+      "learning_rate": 0.0009692832764505119,
+      "loss": 6.954,
+      "step": 270
+    },
+    {
+      "epoch": 0.0924914675767918,
+      "grad_norm": 3.9115519523620605,
+      "learning_rate": 0.000969169510807736,
+      "loss": 6.5418,
+      "step": 271
+    },
+    {
+      "epoch": 0.09283276450511946,
+      "grad_norm": 4.029146671295166,
+      "learning_rate": 0.0009690557451649603,
+      "loss": 6.626,
+      "step": 272
+    },
+    {
+      "epoch": 0.0931740614334471,
+      "grad_norm": 3.7334775924682617,
+      "learning_rate": 0.0009689419795221844,
+      "loss": 6.3663,
+      "step": 273
+    },
+    {
+      "epoch": 0.09351535836177474,
+      "grad_norm": 3.6482112407684326,
+      "learning_rate": 0.0009688282138794085,
+      "loss": 6.9092,
+      "step": 274
+    },
+    {
+      "epoch": 0.09385665529010238,
+      "grad_norm": 8.627721786499023,
+      "learning_rate": 0.0009687144482366326,
+      "loss": 5.9086,
+      "step": 275
+    },
+    {
+      "epoch": 0.09419795221843004,
+      "grad_norm": 3.9426357746124268,
+      "learning_rate": 0.0009686006825938567,
+      "loss": 7.169,
+      "step": 276
+    },
+    {
+      "epoch": 0.09453924914675768,
+      "grad_norm": 4.135240077972412,
+      "learning_rate": 0.0009684869169510808,
+      "loss": 6.7811,
+      "step": 277
+    },
+    {
+      "epoch": 0.09488054607508532,
+      "grad_norm": 4.837364673614502,
+      "learning_rate": 0.000968373151308305,
+      "loss": 6.8342,
+      "step": 278
+    },
+    {
+      "epoch": 0.09522184300341296,
+      "grad_norm": 3.7553555965423584,
+      "learning_rate": 0.0009682593856655291,
+      "loss": 6.4767,
+      "step": 279
+    },
+    {
+      "epoch": 0.09556313993174062,
+      "grad_norm": 3.947141647338867,
+      "learning_rate": 0.0009681456200227532,
+      "loss": 6.2992,
+      "step": 280
+    },
+    {
+      "epoch": 0.09590443686006826,
+      "grad_norm": 4.0404486656188965,
+      "learning_rate": 0.0009680318543799773,
+      "loss": 6.9428,
+      "step": 281
+    },
+    {
+      "epoch": 0.0962457337883959,
+      "grad_norm": 3.603900194168091,
+      "learning_rate": 0.0009679180887372013,
+      "loss": 7.2719,
+      "step": 282
+    },
+    {
+      "epoch": 0.09658703071672355,
+      "grad_norm": 5.454912185668945,
+      "learning_rate": 0.0009678043230944254,
+      "loss": 6.1625,
+      "step": 283
+    },
+    {
+      "epoch": 0.09692832764505119,
+      "grad_norm": 4.6887736320495605,
+      "learning_rate": 0.0009676905574516496,
+      "loss": 4.2823,
+      "step": 284
+    },
+    {
+      "epoch": 0.09726962457337884,
+      "grad_norm": 4.27936315536499,
+      "learning_rate": 0.0009675767918088737,
+      "loss": 6.8959,
+      "step": 285
+    },
+    {
+      "epoch": 0.09761092150170649,
+      "grad_norm": 3.739509105682373,
+      "learning_rate": 0.0009674630261660978,
+      "loss": 7.2816,
+      "step": 286
+    },
+    {
+      "epoch": 0.09795221843003413,
+      "grad_norm": 3.4555413722991943,
+      "learning_rate": 0.0009673492605233219,
+      "loss": 7.0285,
+      "step": 287
+    },
+    {
+      "epoch": 0.09829351535836177,
+      "grad_norm": 3.6354432106018066,
+      "learning_rate": 0.000967235494880546,
+      "loss": 6.5368,
+      "step": 288
+    },
+    {
+      "epoch": 0.09863481228668942,
+      "grad_norm": 3.558715343475342,
+      "learning_rate": 0.0009671217292377701,
+      "loss": 7.1075,
+      "step": 289
+    },
+    {
+      "epoch": 0.09897610921501707,
+      "grad_norm": 9.740402221679688,
+      "learning_rate": 0.0009670079635949944,
+      "loss": 6.4126,
+      "step": 290
+    },
+    {
+      "epoch": 0.09931740614334471,
+      "grad_norm": 3.963900089263916,
+      "learning_rate": 0.0009668941979522185,
+      "loss": 6.8788,
+      "step": 291
+    },
+    {
+      "epoch": 0.09965870307167235,
+      "grad_norm": 3.8590564727783203,
+      "learning_rate": 0.0009667804323094426,
+      "loss": 7.1995,
+      "step": 292
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5563805103302,
+      "learning_rate": 0.0009666666666666667,
+      "loss": 7.3654,
+      "step": 293
+    },
+    {
+      "epoch": 0.10034129692832765,
+      "grad_norm": 3.6024012565612793,
+      "learning_rate": 0.0009665529010238908,
+      "loss": 6.7006,
+      "step": 294
+    },
+    {
+      "epoch": 0.10068259385665529,
+      "grad_norm": 3.463623523712158,
+      "learning_rate": 0.000966439135381115,
+      "loss": 6.8185,
+      "step": 295
+    },
+    {
+      "epoch": 0.10102389078498293,
+      "grad_norm": 5.106781005859375,
+      "learning_rate": 0.0009663253697383391,
+      "loss": 5.5867,
+      "step": 296
+    },
+    {
+      "epoch": 0.10136518771331059,
+      "grad_norm": 3.8914577960968018,
+      "learning_rate": 0.0009662116040955632,
+      "loss": 7.6307,
+      "step": 297
+    },
+    {
+      "epoch": 0.10170648464163823,
+      "grad_norm": 4.023569583892822,
+      "learning_rate": 0.0009660978384527873,
+      "loss": 6.4419,
+      "step": 298
+    },
+    {
+      "epoch": 0.10204778156996587,
+      "grad_norm": 4.755953788757324,
+      "learning_rate": 0.0009659840728100114,
+      "loss": 6.7173,
+      "step": 299
+    },
+    {
+      "epoch": 0.10238907849829351,
+      "grad_norm": 3.9473185539245605,
+      "learning_rate": 0.0009658703071672355,
+      "loss": 4.9165,
+      "step": 300
+    },
+    {
+      "epoch": 0.10273037542662115,
+      "grad_norm": 4.2228684425354,
+      "learning_rate": 0.0009657565415244597,
+      "loss": 6.7526,
+      "step": 301
+    },
+    {
+      "epoch": 0.10307167235494881,
+      "grad_norm": 3.5894923210144043,
+      "learning_rate": 0.0009656427758816837,
+      "loss": 7.1576,
+      "step": 302
+    },
+    {
+      "epoch": 0.10341296928327645,
+      "grad_norm": 4.090893745422363,
+      "learning_rate": 0.0009655290102389078,
+      "loss": 6.4097,
+      "step": 303
+    },
+    {
+      "epoch": 0.1037542662116041,
+      "grad_norm": 3.7067759037017822,
+      "learning_rate": 0.0009654152445961319,
+      "loss": 7.3378,
+      "step": 304
+    },
+    {
+      "epoch": 0.10409556313993173,
+      "grad_norm": 3.431490182876587,
+      "learning_rate": 0.000965301478953356,
+      "loss": 7.1981,
+      "step": 305
+    },
+    {
+      "epoch": 0.10443686006825939,
+      "grad_norm": 3.6670467853546143,
+      "learning_rate": 0.0009651877133105801,
+      "loss": 6.5654,
+      "step": 306
+    },
+    {
+      "epoch": 0.10477815699658703,
+      "grad_norm": 3.554365634918213,
+      "learning_rate": 0.0009650739476678044,
+      "loss": 6.481,
+      "step": 307
+    },
+    {
+      "epoch": 0.10511945392491467,
+      "grad_norm": 3.5014779567718506,
+      "learning_rate": 0.0009649601820250285,
+      "loss": 6.8385,
+      "step": 308
+    },
+    {
+      "epoch": 0.10546075085324232,
+      "grad_norm": 10.873976707458496,
+      "learning_rate": 0.0009648464163822526,
+      "loss": 6.5274,
+      "step": 309
+    },
+    {
+      "epoch": 0.10580204778156997,
+      "grad_norm": 3.9372551441192627,
+      "learning_rate": 0.0009647326507394767,
+      "loss": 7.2976,
+      "step": 310
+    },
+    {
+      "epoch": 0.10614334470989761,
+      "grad_norm": 3.7556705474853516,
+      "learning_rate": 0.0009646188850967008,
+      "loss": 7.2823,
+      "step": 311
+    },
+    {
+      "epoch": 0.10648464163822526,
+      "grad_norm": 5.166213512420654,
+      "learning_rate": 0.000964505119453925,
+      "loss": 6.1935,
+      "step": 312
+    },
+    {
+      "epoch": 0.1068259385665529,
+      "grad_norm": 4.041624546051025,
+      "learning_rate": 0.0009643913538111491,
+      "loss": 7.1861,
+      "step": 313
+    },
+    {
+      "epoch": 0.10716723549488055,
+      "grad_norm": 4.318609714508057,
+      "learning_rate": 0.0009642775881683732,
+      "loss": 6.4683,
+      "step": 314
+    },
+    {
+      "epoch": 0.1075085324232082,
+      "grad_norm": 3.769338607788086,
+      "learning_rate": 0.0009641638225255973,
+      "loss": 6.7583,
+      "step": 315
+    },
+    {
+      "epoch": 0.10784982935153584,
+      "grad_norm": 3.6845760345458984,
+      "learning_rate": 0.0009640500568828214,
+      "loss": 6.8184,
+      "step": 316
+    },
+    {
+      "epoch": 0.10819112627986348,
+      "grad_norm": 3.5176687240600586,
+      "learning_rate": 0.0009639362912400455,
+      "loss": 6.7214,
+      "step": 317
+    },
+    {
+      "epoch": 0.10853242320819112,
+      "grad_norm": 3.728588581085205,
+      "learning_rate": 0.0009638225255972697,
+      "loss": 7.0418,
+      "step": 318
+    },
+    {
+      "epoch": 0.10887372013651878,
+      "grad_norm": 3.9855856895446777,
+      "learning_rate": 0.0009637087599544938,
+      "loss": 6.6959,
+      "step": 319
+    },
+    {
+      "epoch": 0.10921501706484642,
+      "grad_norm": 4.362178802490234,
+      "learning_rate": 0.0009635949943117179,
+      "loss": 6.0196,
+      "step": 320
+    },
+    {
+      "epoch": 0.10955631399317406,
+      "grad_norm": 5.292174816131592,
+      "learning_rate": 0.0009634812286689419,
+      "loss": 5.8822,
+      "step": 321
+    },
+    {
+      "epoch": 0.1098976109215017,
+      "grad_norm": 4.121539115905762,
+      "learning_rate": 0.000963367463026166,
+      "loss": 7.0859,
+      "step": 322
+    },
+    {
+      "epoch": 0.11023890784982936,
+      "grad_norm": 4.680109024047852,
+      "learning_rate": 0.0009632536973833901,
+      "loss": 6.4441,
+      "step": 323
+    },
+    {
+      "epoch": 0.110580204778157,
+      "grad_norm": 3.6892054080963135,
+      "learning_rate": 0.0009631399317406144,
+      "loss": 7.0459,
+      "step": 324
+    },
+    {
+      "epoch": 0.11092150170648464,
+      "grad_norm": 3.639540672302246,
+      "learning_rate": 0.0009630261660978385,
+      "loss": 7.1954,
+      "step": 325
+    },
+    {
+      "epoch": 0.11126279863481228,
+      "grad_norm": 3.4080095291137695,
+      "learning_rate": 0.0009629124004550626,
+      "loss": 7.0554,
+      "step": 326
+    },
+    {
+      "epoch": 0.11160409556313994,
+      "grad_norm": 3.4967737197875977,
+      "learning_rate": 0.0009627986348122867,
+      "loss": 7.17,
+      "step": 327
+    },
+    {
+      "epoch": 0.11194539249146758,
+      "grad_norm": 3.8794071674346924,
+      "learning_rate": 0.0009626848691695108,
+      "loss": 6.759,
+      "step": 328
+    },
+    {
+      "epoch": 0.11228668941979522,
+      "grad_norm": 3.408348321914673,
+      "learning_rate": 0.0009625711035267349,
+      "loss": 6.7917,
+      "step": 329
+    },
+    {
+      "epoch": 0.11262798634812286,
+      "grad_norm": 3.614760637283325,
+      "learning_rate": 0.0009624573378839591,
+      "loss": 6.6399,
+      "step": 330
+    },
+    {
+      "epoch": 0.1129692832764505,
+      "grad_norm": 3.6130752563476562,
+      "learning_rate": 0.0009623435722411832,
+      "loss": 6.8108,
+      "step": 331
+    },
+    {
+      "epoch": 0.11331058020477816,
+      "grad_norm": 3.7819409370422363,
+      "learning_rate": 0.0009622298065984073,
+      "loss": 7.3848,
+      "step": 332
+    },
+    {
+      "epoch": 0.1136518771331058,
+      "grad_norm": 3.626450300216675,
+      "learning_rate": 0.0009621160409556314,
+      "loss": 7.0753,
+      "step": 333
+    },
+    {
+      "epoch": 0.11399317406143344,
+      "grad_norm": 3.5669236183166504,
+      "learning_rate": 0.0009620022753128555,
+      "loss": 6.5975,
+      "step": 334
+    },
+    {
+      "epoch": 0.11433447098976109,
+      "grad_norm": 3.4628782272338867,
+      "learning_rate": 0.0009618885096700797,
+      "loss": 6.7684,
+      "step": 335
+    },
+    {
+      "epoch": 0.11467576791808874,
+      "grad_norm": 7.344601154327393,
+      "learning_rate": 0.0009617747440273038,
+      "loss": 6.0625,
+      "step": 336
+    },
+    {
+      "epoch": 0.11501706484641638,
+      "grad_norm": 3.9138948917388916,
+      "learning_rate": 0.0009616609783845279,
+      "loss": 6.9233,
+      "step": 337
+    },
+    {
+      "epoch": 0.11535836177474403,
+      "grad_norm": 3.7263455390930176,
+      "learning_rate": 0.000961547212741752,
+      "loss": 7.1069,
+      "step": 338
+    },
+    {
+      "epoch": 0.11569965870307167,
+      "grad_norm": 3.638993740081787,
+      "learning_rate": 0.0009614334470989762,
+      "loss": 6.5446,
+      "step": 339
+    },
+    {
+      "epoch": 0.11604095563139932,
+      "grad_norm": 3.507223129272461,
+      "learning_rate": 0.0009613196814562003,
+      "loss": 6.8566,
+      "step": 340
+    },
+    {
+      "epoch": 0.11638225255972696,
+      "grad_norm": 3.5542092323303223,
+      "learning_rate": 0.0009612059158134244,
+      "loss": 7.013,
+      "step": 341
+    },
+    {
+      "epoch": 0.1167235494880546,
+      "grad_norm": 3.579059600830078,
+      "learning_rate": 0.0009610921501706485,
+      "loss": 7.4816,
+      "step": 342
+    },
+    {
+      "epoch": 0.11706484641638225,
+      "grad_norm": 3.462669849395752,
+      "learning_rate": 0.0009609783845278726,
+      "loss": 6.7983,
+      "step": 343
+    },
+    {
+      "epoch": 0.1174061433447099,
+      "grad_norm": 3.597429037094116,
+      "learning_rate": 0.0009608646188850967,
+      "loss": 6.9345,
+      "step": 344
+    },
+    {
+      "epoch": 0.11774744027303755,
+      "grad_norm": 15.800237655639648,
+      "learning_rate": 0.0009607508532423208,
+      "loss": 6.5234,
+      "step": 345
+    },
+    {
+      "epoch": 0.11808873720136519,
+      "grad_norm": 4.462514400482178,
+      "learning_rate": 0.0009606370875995449,
+      "loss": 7.2368,
+      "step": 346
+    },
+    {
+      "epoch": 0.11843003412969283,
+      "grad_norm": 4.498167514801025,
+      "learning_rate": 0.0009605233219567691,
+      "loss": 6.6885,
+      "step": 347
+    },
+    {
+      "epoch": 0.11877133105802047,
+      "grad_norm": 4.717837810516357,
+      "learning_rate": 0.0009604095563139932,
+      "loss": 5.435,
+      "step": 348
+    },
+    {
+      "epoch": 0.11911262798634813,
+      "grad_norm": 3.9620487689971924,
+      "learning_rate": 0.0009602957906712173,
+      "loss": 7.6661,
+      "step": 349
+    },
+    {
+      "epoch": 0.11945392491467577,
+      "grad_norm": 3.7153236865997314,
+      "learning_rate": 0.0009601820250284414,
+      "loss": 6.6377,
+      "step": 350
+    },
+    {
+      "epoch": 0.11979522184300341,
+      "grad_norm": 4.2678632736206055,
+      "learning_rate": 0.0009600682593856655,
+      "loss": 6.5929,
+      "step": 351
+    },
+    {
+      "epoch": 0.12013651877133105,
+      "grad_norm": 3.488931894302368,
+      "learning_rate": 0.0009599544937428897,
+      "loss": 6.7701,
+      "step": 352
+    },
+    {
+      "epoch": 0.12047781569965871,
+      "grad_norm": 3.7950785160064697,
+      "learning_rate": 0.0009598407281001138,
+      "loss": 6.3283,
+      "step": 353
+    },
+    {
+      "epoch": 0.12081911262798635,
+      "grad_norm": 3.611812114715576,
+      "learning_rate": 0.0009597269624573379,
+      "loss": 7.0418,
+      "step": 354
+    },
+    {
+      "epoch": 0.12116040955631399,
+      "grad_norm": 3.4913434982299805,
+      "learning_rate": 0.000959613196814562,
+      "loss": 7.1764,
+      "step": 355
+    },
+    {
+      "epoch": 0.12150170648464163,
+      "grad_norm": 3.7057766914367676,
+      "learning_rate": 0.0009594994311717862,
+      "loss": 6.6347,
+      "step": 356
+    },
+    {
+      "epoch": 0.12184300341296929,
+      "grad_norm": 3.8991594314575195,
+      "learning_rate": 0.0009593856655290103,
+      "loss": 6.7696,
+      "step": 357
+    },
+    {
+      "epoch": 0.12218430034129693,
+      "grad_norm": 3.863154172897339,
+      "learning_rate": 0.0009592718998862345,
+      "loss": 7.317,
+      "step": 358
+    },
+    {
+      "epoch": 0.12252559726962457,
+      "grad_norm": 3.521376848220825,
+      "learning_rate": 0.0009591581342434586,
+      "loss": 7.5248,
+      "step": 359
+    },
+    {
+      "epoch": 0.12286689419795221,
+      "grad_norm": 3.6891329288482666,
+      "learning_rate": 0.0009590443686006826,
+      "loss": 6.9287,
+      "step": 360
+    },
+    {
+      "epoch": 0.12320819112627987,
+      "grad_norm": 3.7071328163146973,
+      "learning_rate": 0.0009589306029579067,
+      "loss": 7.086,
+      "step": 361
+    },
+    {
+      "epoch": 0.12354948805460751,
+      "grad_norm": 5.102900981903076,
+      "learning_rate": 0.0009588168373151308,
+      "loss": 6.361,
+      "step": 362
+    },
+    {
+      "epoch": 0.12389078498293515,
+      "grad_norm": 4.640048503875732,
+      "learning_rate": 0.0009587030716723549,
+      "loss": 5.9508,
+      "step": 363
+    },
+    {
+      "epoch": 0.1242320819112628,
+      "grad_norm": 7.667442321777344,
+      "learning_rate": 0.0009585893060295791,
+      "loss": 7.4653,
+      "step": 364
+    },
+    {
+      "epoch": 0.12457337883959044,
+      "grad_norm": 4.129282474517822,
+      "learning_rate": 0.0009584755403868032,
+      "loss": 7.1397,
+      "step": 365
+    },
+    {
+      "epoch": 0.12491467576791809,
+      "grad_norm": 3.7277729511260986,
+      "learning_rate": 0.0009583617747440273,
+      "loss": 6.2993,
+      "step": 366
+    },
+    {
+      "epoch": 0.12525597269624572,
+      "grad_norm": 3.472531795501709,
+      "learning_rate": 0.0009582480091012514,
+      "loss": 6.7378,
+      "step": 367
+    },
+    {
+      "epoch": 0.12559726962457338,
+      "grad_norm": 4.953517913818359,
+      "learning_rate": 0.0009581342434584755,
+      "loss": 6.3897,
+      "step": 368
+    },
+    {
+      "epoch": 0.12593856655290103,
+      "grad_norm": 3.745403289794922,
+      "learning_rate": 0.0009580204778156996,
+      "loss": 6.8016,
+      "step": 369
+    },
+    {
+      "epoch": 0.12627986348122866,
+      "grad_norm": 3.9581098556518555,
+      "learning_rate": 0.0009579067121729238,
+      "loss": 6.9358,
+      "step": 370
+    },
+    {
+      "epoch": 0.12662116040955632,
+      "grad_norm": 3.588125467300415,
+      "learning_rate": 0.0009577929465301479,
+      "loss": 6.5954,
+      "step": 371
+    },
+    {
+      "epoch": 0.12696245733788397,
+      "grad_norm": 3.592097282409668,
+      "learning_rate": 0.000957679180887372,
+      "loss": 7.2236,
+      "step": 372
+    },
+    {
+      "epoch": 0.1273037542662116,
+      "grad_norm": 3.409364700317383,
+      "learning_rate": 0.0009575654152445962,
+      "loss": 7.0846,
+      "step": 373
+    },
+    {
+      "epoch": 0.12764505119453926,
+      "grad_norm": 4.143648147583008,
+      "learning_rate": 0.0009574516496018203,
+      "loss": 6.6434,
+      "step": 374
+    },
+    {
+      "epoch": 0.12798634812286688,
+      "grad_norm": 4.170565128326416,
+      "learning_rate": 0.0009573378839590445,
+      "loss": 6.5485,
+      "step": 375
+    },
+    {
+      "epoch": 0.12832764505119454,
+      "grad_norm": 3.7150259017944336,
+      "learning_rate": 0.0009572241183162686,
+      "loss": 6.7944,
+      "step": 376
+    },
+    {
+      "epoch": 0.1286689419795222,
+      "grad_norm": 9.187422752380371,
+      "learning_rate": 0.0009571103526734927,
+      "loss": 5.1412,
+      "step": 377
+    },
+    {
+      "epoch": 0.12901023890784982,
+      "grad_norm": 3.895047187805176,
+      "learning_rate": 0.0009569965870307168,
+      "loss": 6.4227,
+      "step": 378
+    },
+    {
+      "epoch": 0.12935153583617748,
+      "grad_norm": 3.7599642276763916,
+      "learning_rate": 0.0009568828213879409,
+      "loss": 6.9761,
+      "step": 379
+    },
+    {
+      "epoch": 0.1296928327645051,
+      "grad_norm": 3.6376614570617676,
+      "learning_rate": 0.0009567690557451649,
+      "loss": 7.0387,
+      "step": 380
+    },
+    {
+      "epoch": 0.13003412969283276,
+      "grad_norm": 3.673825263977051,
+      "learning_rate": 0.0009566552901023891,
+      "loss": 6.6699,
+      "step": 381
+    },
+    {
+      "epoch": 0.13037542662116042,
+      "grad_norm": 3.4586920738220215,
+      "learning_rate": 0.0009565415244596132,
+      "loss": 6.8109,
+      "step": 382
+    },
+    {
+      "epoch": 0.13071672354948805,
+      "grad_norm": 3.3592073917388916,
+      "learning_rate": 0.0009564277588168373,
+      "loss": 7.0395,
+      "step": 383
+    },
+    {
+      "epoch": 0.1310580204778157,
+      "grad_norm": 3.409188985824585,
+      "learning_rate": 0.0009563139931740614,
+      "loss": 6.7086,
+      "step": 384
+    },
+    {
+      "epoch": 0.13139931740614336,
+      "grad_norm": 3.571364164352417,
+      "learning_rate": 0.0009562002275312855,
+      "loss": 7.0613,
+      "step": 385
+    },
+    {
+      "epoch": 0.13174061433447098,
+      "grad_norm": 3.6313486099243164,
+      "learning_rate": 0.0009560864618885096,
+      "loss": 6.8737,
+      "step": 386
+    },
+    {
+      "epoch": 0.13208191126279864,
+      "grad_norm": 5.386178970336914,
+      "learning_rate": 0.0009559726962457338,
+      "loss": 6.8494,
+      "step": 387
+    },
+    {
+      "epoch": 0.13242320819112627,
+      "grad_norm": 3.736448287963867,
+      "learning_rate": 0.0009558589306029579,
+      "loss": 6.805,
+      "step": 388
+    },
+    {
+      "epoch": 0.13276450511945392,
+      "grad_norm": 4.280710697174072,
+      "learning_rate": 0.000955745164960182,
+      "loss": 6.5263,
+      "step": 389
+    },
+    {
+      "epoch": 0.13310580204778158,
+      "grad_norm": 3.7368850708007812,
+      "learning_rate": 0.0009556313993174062,
+      "loss": 6.8247,
+      "step": 390
+    },
+    {
+      "epoch": 0.1334470989761092,
+      "grad_norm": 3.579983711242676,
+      "learning_rate": 0.0009555176336746303,
+      "loss": 7.1728,
+      "step": 391
+    },
+    {
+      "epoch": 0.13378839590443686,
+      "grad_norm": 3.969280242919922,
+      "learning_rate": 0.0009554038680318545,
+      "loss": 6.2745,
+      "step": 392
+    },
+    {
+      "epoch": 0.13412969283276452,
+      "grad_norm": 3.6488418579101562,
+      "learning_rate": 0.0009552901023890786,
+      "loss": 7.0,
+      "step": 393
+    },
+    {
+      "epoch": 0.13447098976109215,
+      "grad_norm": 3.3817570209503174,
+      "learning_rate": 0.0009551763367463027,
+      "loss": 6.2211,
+      "step": 394
+    },
+    {
+      "epoch": 0.1348122866894198,
+      "grad_norm": 3.5612990856170654,
+      "learning_rate": 0.0009550625711035268,
+      "loss": 6.9262,
+      "step": 395
+    },
+    {
+      "epoch": 0.13515358361774743,
+      "grad_norm": 3.475904703140259,
+      "learning_rate": 0.0009549488054607509,
+      "loss": 7.1462,
+      "step": 396
+    },
+    {
+      "epoch": 0.13549488054607509,
+      "grad_norm": 3.9977428913116455,
+      "learning_rate": 0.000954835039817975,
+      "loss": 6.8524,
+      "step": 397
+    },
+    {
+      "epoch": 0.13583617747440274,
+      "grad_norm": 3.5686051845550537,
+      "learning_rate": 0.0009547212741751992,
+      "loss": 7.1196,
+      "step": 398
+    },
+    {
+      "epoch": 0.13617747440273037,
+      "grad_norm": 3.616010904312134,
+      "learning_rate": 0.0009546075085324232,
+      "loss": 6.8889,
+      "step": 399
+    },
+    {
+      "epoch": 0.13651877133105803,
+      "grad_norm": 3.5153353214263916,
+      "learning_rate": 0.0009544937428896473,
+      "loss": 7.0783,
+      "step": 400
+    },
+    {
+      "epoch": 0.13686006825938565,
+      "grad_norm": 3.46384596824646,
+      "learning_rate": 0.0009543799772468714,
+      "loss": 7.2027,
+      "step": 401
+    },
+    {
+      "epoch": 0.1372013651877133,
+      "grad_norm": 3.6632463932037354,
+      "learning_rate": 0.0009542662116040955,
+      "loss": 6.8579,
+      "step": 402
+    },
+    {
+      "epoch": 0.13754266211604096,
+      "grad_norm": 3.654392719268799,
+      "learning_rate": 0.0009541524459613196,
+      "loss": 6.9971,
+      "step": 403
+    },
+    {
+      "epoch": 0.1378839590443686,
+      "grad_norm": 3.7651548385620117,
+      "learning_rate": 0.0009540386803185438,
+      "loss": 6.9558,
+      "step": 404
+    },
+    {
+      "epoch": 0.13822525597269625,
+      "grad_norm": 3.590897560119629,
+      "learning_rate": 0.0009539249146757679,
+      "loss": 6.9441,
+      "step": 405
+    },
+    {
+      "epoch": 0.1385665529010239,
+      "grad_norm": 3.5330758094787598,
+      "learning_rate": 0.000953811149032992,
+      "loss": 7.3093,
+      "step": 406
+    },
+    {
+      "epoch": 0.13890784982935153,
+      "grad_norm": 3.9493212699890137,
+      "learning_rate": 0.0009536973833902162,
+      "loss": 6.1789,
+      "step": 407
+    },
+    {
+      "epoch": 0.1392491467576792,
+      "grad_norm": 3.6030216217041016,
+      "learning_rate": 0.0009535836177474403,
+      "loss": 6.7268,
+      "step": 408
+    },
+    {
+      "epoch": 0.13959044368600682,
+      "grad_norm": 3.7132041454315186,
+      "learning_rate": 0.0009534698521046644,
+      "loss": 6.8771,
+      "step": 409
+    },
+    {
+      "epoch": 0.13993174061433447,
+      "grad_norm": 3.7219643592834473,
+      "learning_rate": 0.0009533560864618886,
+      "loss": 6.7741,
+      "step": 410
+    },
+    {
+      "epoch": 0.14027303754266213,
+      "grad_norm": 3.658827066421509,
+      "learning_rate": 0.0009532423208191127,
+      "loss": 7.0821,
+      "step": 411
+    },
+    {
+      "epoch": 0.14061433447098975,
+      "grad_norm": 3.503436326980591,
+      "learning_rate": 0.0009531285551763368,
+      "loss": 7.0266,
+      "step": 412
+    },
+    {
+      "epoch": 0.1409556313993174,
+      "grad_norm": 3.4202098846435547,
+      "learning_rate": 0.0009530147895335609,
+      "loss": 6.7716,
+      "step": 413
+    },
+    {
+      "epoch": 0.14129692832764504,
+      "grad_norm": 3.497220993041992,
+      "learning_rate": 0.000952901023890785,
+      "loss": 7.1303,
+      "step": 414
+    },
+    {
+      "epoch": 0.1416382252559727,
+      "grad_norm": 3.7073585987091064,
+      "learning_rate": 0.0009527872582480092,
+      "loss": 6.9418,
+      "step": 415
+    },
+    {
+      "epoch": 0.14197952218430035,
+      "grad_norm": 3.451277256011963,
+      "learning_rate": 0.0009526734926052333,
+      "loss": 6.9495,
+      "step": 416
+    },
+    {
+      "epoch": 0.14232081911262798,
+      "grad_norm": 3.564136505126953,
+      "learning_rate": 0.0009525597269624574,
+      "loss": 6.8177,
+      "step": 417
+    },
+    {
+      "epoch": 0.14266211604095563,
+      "grad_norm": 4.293824672698975,
+      "learning_rate": 0.0009524459613196815,
+      "loss": 7.1537,
+      "step": 418
+    },
+    {
+      "epoch": 0.1430034129692833,
+      "grad_norm": 4.002580165863037,
+      "learning_rate": 0.0009523321956769055,
+      "loss": 7.0574,
+      "step": 419
+    },
+    {
+      "epoch": 0.14334470989761092,
+      "grad_norm": 3.631091833114624,
+      "learning_rate": 0.0009522184300341296,
+      "loss": 6.552,
+      "step": 420
+    },
+    {
+      "epoch": 0.14368600682593857,
+      "grad_norm": 3.5741636753082275,
+      "learning_rate": 0.0009521046643913538,
+      "loss": 6.7028,
+      "step": 421
+    },
+    {
+      "epoch": 0.1440273037542662,
+      "grad_norm": 3.8410189151763916,
+      "learning_rate": 0.0009519908987485779,
+      "loss": 7.0462,
+      "step": 422
+    },
+    {
+      "epoch": 0.14436860068259386,
+      "grad_norm": 3.5977299213409424,
+      "learning_rate": 0.000951877133105802,
+      "loss": 7.1202,
+      "step": 423
+    },
+    {
+      "epoch": 0.1447098976109215,
+      "grad_norm": 3.541287899017334,
+      "learning_rate": 0.0009517633674630262,
+      "loss": 6.6661,
+      "step": 424
+    },
+    {
+      "epoch": 0.14505119453924914,
+      "grad_norm": 3.6210668087005615,
+      "learning_rate": 0.0009516496018202503,
+      "loss": 7.1838,
+      "step": 425
+    },
+    {
+      "epoch": 0.1453924914675768,
+      "grad_norm": 4.462888717651367,
+      "learning_rate": 0.0009515358361774744,
+      "loss": 5.8725,
+      "step": 426
+    },
+    {
+      "epoch": 0.14573378839590442,
+      "grad_norm": 3.566049814224243,
+      "learning_rate": 0.0009514220705346986,
+      "loss": 6.7938,
+      "step": 427
+    },
+    {
+      "epoch": 0.14607508532423208,
+      "grad_norm": 3.4675862789154053,
+      "learning_rate": 0.0009513083048919227,
+      "loss": 6.5886,
+      "step": 428
+    },
+    {
+      "epoch": 0.14641638225255973,
+      "grad_norm": 3.4165515899658203,
+      "learning_rate": 0.0009511945392491468,
+      "loss": 6.9825,
+      "step": 429
+    },
+    {
+      "epoch": 0.14675767918088736,
+      "grad_norm": 3.5711846351623535,
+      "learning_rate": 0.0009510807736063709,
+      "loss": 6.9881,
+      "step": 430
+    },
+    {
+      "epoch": 0.14709897610921502,
+      "grad_norm": 3.523361921310425,
+      "learning_rate": 0.000950967007963595,
+      "loss": 6.6169,
+      "step": 431
+    },
+    {
+      "epoch": 0.14744027303754267,
+      "grad_norm": 3.6487302780151367,
+      "learning_rate": 0.0009508532423208191,
+      "loss": 7.0033,
+      "step": 432
+    },
+    {
+      "epoch": 0.1477815699658703,
+      "grad_norm": 4.30921745300293,
+      "learning_rate": 0.0009507394766780433,
+      "loss": 6.4524,
+      "step": 433
+    },
+    {
+      "epoch": 0.14812286689419796,
+      "grad_norm": 3.7454891204833984,
+      "learning_rate": 0.0009506257110352674,
+      "loss": 6.6826,
+      "step": 434
+    },
+    {
+      "epoch": 0.14846416382252559,
+      "grad_norm": 3.676464796066284,
+      "learning_rate": 0.0009505119453924915,
+      "loss": 6.8758,
+      "step": 435
+    },
+    {
+      "epoch": 0.14880546075085324,
+      "grad_norm": 3.738009452819824,
+      "learning_rate": 0.0009503981797497156,
+      "loss": 6.9021,
+      "step": 436
+    },
+    {
+      "epoch": 0.1491467576791809,
+      "grad_norm": 3.388455629348755,
+      "learning_rate": 0.0009502844141069397,
+      "loss": 6.7852,
+      "step": 437
+    },
+    {
+      "epoch": 0.14948805460750852,
+      "grad_norm": 5.163676738739014,
+      "learning_rate": 0.0009501706484641638,
+      "loss": 6.4292,
+      "step": 438
+    },
+    {
+      "epoch": 0.14982935153583618,
+      "grad_norm": 3.6834332942962646,
+      "learning_rate": 0.0009500568828213879,
+      "loss": 6.8296,
+      "step": 439
+    },
+    {
+      "epoch": 0.15017064846416384,
+      "grad_norm": 3.6888158321380615,
+      "learning_rate": 0.000949943117178612,
+      "loss": 6.7824,
+      "step": 440
+    },
+    {
+      "epoch": 0.15051194539249146,
+      "grad_norm": 3.5720479488372803,
+      "learning_rate": 0.0009498293515358362,
+      "loss": 6.973,
+      "step": 441
+    },
+    {
+      "epoch": 0.15085324232081912,
+      "grad_norm": 3.5570552349090576,
+      "learning_rate": 0.0009497155858930603,
+      "loss": 6.264,
+      "step": 442
+    },
+    {
+      "epoch": 0.15119453924914675,
+      "grad_norm": 3.610084056854248,
+      "learning_rate": 0.0009496018202502844,
+      "loss": 6.2526,
+      "step": 443
+    },
+    {
+      "epoch": 0.1515358361774744,
+      "grad_norm": 3.445772647857666,
+      "learning_rate": 0.0009494880546075086,
+      "loss": 6.888,
+      "step": 444
+    },
+    {
+      "epoch": 0.15187713310580206,
+      "grad_norm": 3.443215847015381,
+      "learning_rate": 0.0009493742889647327,
+      "loss": 7.2213,
+      "step": 445
+    },
+    {
+      "epoch": 0.1522184300341297,
+      "grad_norm": 3.4744131565093994,
+      "learning_rate": 0.0009492605233219568,
+      "loss": 6.1855,
+      "step": 446
+    },
+    {
+      "epoch": 0.15255972696245734,
+      "grad_norm": 3.616422176361084,
+      "learning_rate": 0.0009491467576791809,
+      "loss": 6.7537,
+      "step": 447
+    },
+    {
+      "epoch": 0.15290102389078497,
+      "grad_norm": 3.427217483520508,
+      "learning_rate": 0.000949032992036405,
+      "loss": 6.5635,
+      "step": 448
+    },
+    {
+      "epoch": 0.15324232081911263,
+      "grad_norm": 3.655622720718384,
+      "learning_rate": 0.0009489192263936291,
+      "loss": 6.4325,
+      "step": 449
+    },
+    {
+      "epoch": 0.15358361774744028,
+      "grad_norm": 3.9160969257354736,
+      "learning_rate": 0.0009488054607508533,
+      "loss": 6.4151,
+      "step": 450
+    },
+    {
+      "epoch": 0.1539249146757679,
+      "grad_norm": 4.333205223083496,
+      "learning_rate": 0.0009486916951080774,
+      "loss": 6.0077,
+      "step": 451
+    },
+    {
+      "epoch": 0.15426621160409557,
+      "grad_norm": 3.7891650199890137,
+      "learning_rate": 0.0009485779294653015,
+      "loss": 6.3463,
+      "step": 452
+    },
+    {
+      "epoch": 0.15460750853242322,
+      "grad_norm": 3.907245635986328,
+      "learning_rate": 0.0009484641638225256,
+      "loss": 5.503,
+      "step": 453
+    },
+    {
+      "epoch": 0.15494880546075085,
+      "grad_norm": 3.9546966552734375,
+      "learning_rate": 0.0009483503981797497,
+      "loss": 6.6111,
+      "step": 454
+    },
+    {
+      "epoch": 0.1552901023890785,
+      "grad_norm": 3.712440252304077,
+      "learning_rate": 0.000948236632536974,
+      "loss": 6.6064,
+      "step": 455
+    },
+    {
+      "epoch": 0.15563139931740613,
+      "grad_norm": 3.5256028175354004,
+      "learning_rate": 0.0009481228668941981,
+      "loss": 6.9209,
+      "step": 456
+    },
+    {
+      "epoch": 0.1559726962457338,
+      "grad_norm": 4.240937232971191,
+      "learning_rate": 0.0009480091012514222,
+      "loss": 6.3623,
+      "step": 457
+    },
+    {
+      "epoch": 0.15631399317406144,
+      "grad_norm": 3.398322820663452,
+      "learning_rate": 0.0009478953356086462,
+      "loss": 6.574,
+      "step": 458
+    },
+    {
+      "epoch": 0.15665529010238907,
+      "grad_norm": 3.7059226036071777,
+      "learning_rate": 0.0009477815699658703,
+      "loss": 6.9253,
+      "step": 459
+    },
+    {
+      "epoch": 0.15699658703071673,
+      "grad_norm": 4.11593770980835,
+      "learning_rate": 0.0009476678043230944,
+      "loss": 6.4204,
+      "step": 460
+    },
+    {
+      "epoch": 0.15733788395904436,
+      "grad_norm": 3.5810751914978027,
+      "learning_rate": 0.0009475540386803186,
+      "loss": 6.8083,
+      "step": 461
+    },
+    {
+      "epoch": 0.157679180887372,
+      "grad_norm": 3.538257360458374,
+      "learning_rate": 0.0009474402730375427,
+      "loss": 7.1334,
+      "step": 462
+    },
+    {
+      "epoch": 0.15802047781569967,
+      "grad_norm": 3.456045627593994,
+      "learning_rate": 0.0009473265073947668,
+      "loss": 7.0928,
+      "step": 463
+    },
+    {
+      "epoch": 0.1583617747440273,
+      "grad_norm": 3.6278865337371826,
+      "learning_rate": 0.0009472127417519909,
+      "loss": 6.8821,
+      "step": 464
+    },
+    {
+      "epoch": 0.15870307167235495,
+      "grad_norm": 3.6520791053771973,
+      "learning_rate": 0.000947098976109215,
+      "loss": 6.5972,
+      "step": 465
+    },
+    {
+      "epoch": 0.1590443686006826,
+      "grad_norm": 3.5707054138183594,
+      "learning_rate": 0.0009469852104664391,
+      "loss": 7.1507,
+      "step": 466
+    },
+    {
+      "epoch": 0.15938566552901023,
+      "grad_norm": 3.7022809982299805,
+      "learning_rate": 0.0009468714448236633,
+      "loss": 6.5752,
+      "step": 467
+    },
+    {
+      "epoch": 0.1597269624573379,
+      "grad_norm": 3.426180839538574,
+      "learning_rate": 0.0009467576791808874,
+      "loss": 6.6443,
+      "step": 468
+    },
+    {
+      "epoch": 0.16006825938566552,
+      "grad_norm": 3.5428926944732666,
+      "learning_rate": 0.0009466439135381115,
+      "loss": 6.5234,
+      "step": 469
+    },
+    {
+      "epoch": 0.16040955631399317,
+      "grad_norm": 3.7080366611480713,
+      "learning_rate": 0.0009465301478953356,
+      "loss": 7.1171,
+      "step": 470
+    },
+    {
+      "epoch": 0.16075085324232083,
+      "grad_norm": 3.589245080947876,
+      "learning_rate": 0.0009464163822525597,
+      "loss": 6.8297,
+      "step": 471
+    },
+    {
+      "epoch": 0.16109215017064846,
+      "grad_norm": 3.5839226245880127,
+      "learning_rate": 0.0009463026166097838,
+      "loss": 6.6766,
+      "step": 472
+    },
+    {
+      "epoch": 0.1614334470989761,
+      "grad_norm": 3.460501194000244,
+      "learning_rate": 0.0009461888509670081,
+      "loss": 6.9474,
+      "step": 473
+    },
+    {
+      "epoch": 0.16177474402730374,
+      "grad_norm": 3.6447677612304688,
+      "learning_rate": 0.0009460750853242322,
+      "loss": 7.2087,
+      "step": 474
+    },
+    {
+      "epoch": 0.1621160409556314,
+      "grad_norm": 6.186277866363525,
+      "learning_rate": 0.0009459613196814563,
+      "loss": 5.5688,
+      "step": 475
+    },
+    {
+      "epoch": 0.16245733788395905,
+      "grad_norm": 3.7447400093078613,
+      "learning_rate": 0.0009458475540386804,
+      "loss": 6.5,
+      "step": 476
+    },
+    {
+      "epoch": 0.16279863481228668,
+      "grad_norm": 3.6991195678710938,
+      "learning_rate": 0.0009457337883959044,
+      "loss": 6.7563,
+      "step": 477
+    },
+    {
+      "epoch": 0.16313993174061434,
+      "grad_norm": 3.4451916217803955,
+      "learning_rate": 0.0009456200227531286,
+      "loss": 7.0189,
+      "step": 478
+    },
+    {
+      "epoch": 0.163481228668942,
+      "grad_norm": 17.707456588745117,
+      "learning_rate": 0.0009455062571103527,
+      "loss": 5.872,
+      "step": 479
+    },
+    {
+      "epoch": 0.16382252559726962,
+      "grad_norm": 4.85144567489624,
+      "learning_rate": 0.0009453924914675768,
+      "loss": 6.0977,
+      "step": 480
+    },
+    {
+      "epoch": 0.16416382252559727,
+      "grad_norm": 3.997605085372925,
+      "learning_rate": 0.0009452787258248009,
+      "loss": 7.2862,
+      "step": 481
+    },
+    {
+      "epoch": 0.1645051194539249,
+      "grad_norm": 4.731887340545654,
+      "learning_rate": 0.000945164960182025,
+      "loss": 7.1271,
+      "step": 482
+    },
+    {
+      "epoch": 0.16484641638225256,
+      "grad_norm": 3.7682464122772217,
+      "learning_rate": 0.0009450511945392491,
+      "loss": 7.1285,
+      "step": 483
+    },
+    {
+      "epoch": 0.16518771331058021,
+      "grad_norm": 3.5759470462799072,
+      "learning_rate": 0.0009449374288964733,
+      "loss": 7.0162,
+      "step": 484
+    },
+    {
+      "epoch": 0.16552901023890784,
+      "grad_norm": 3.409883499145508,
+      "learning_rate": 0.0009448236632536974,
+      "loss": 7.237,
+      "step": 485
+    },
+    {
+      "epoch": 0.1658703071672355,
+      "grad_norm": 3.6246159076690674,
+      "learning_rate": 0.0009447098976109215,
+      "loss": 6.7235,
+      "step": 486
+    },
+    {
+      "epoch": 0.16621160409556315,
+      "grad_norm": 3.5993361473083496,
+      "learning_rate": 0.0009445961319681456,
+      "loss": 6.7617,
+      "step": 487
+    },
+    {
+      "epoch": 0.16655290102389078,
+      "grad_norm": 3.4608912467956543,
+      "learning_rate": 0.0009444823663253697,
+      "loss": 6.7189,
+      "step": 488
+    },
+    {
+      "epoch": 0.16689419795221844,
+      "grad_norm": 4.949087619781494,
+      "learning_rate": 0.0009443686006825938,
+      "loss": 6.1499,
+      "step": 489
+    },
+    {
+      "epoch": 0.16723549488054607,
+      "grad_norm": 3.6994855403900146,
+      "learning_rate": 0.0009442548350398181,
+      "loss": 7.1528,
+      "step": 490
+    },
+    {
+      "epoch": 0.16757679180887372,
+      "grad_norm": 3.868940591812134,
+      "learning_rate": 0.0009441410693970422,
+      "loss": 7.0506,
+      "step": 491
+    },
+    {
+      "epoch": 0.16791808873720138,
+      "grad_norm": 3.722750663757324,
+      "learning_rate": 0.0009440273037542663,
+      "loss": 7.0174,
+      "step": 492
+    },
+    {
+      "epoch": 0.168259385665529,
+      "grad_norm": 7.372501373291016,
+      "learning_rate": 0.0009439135381114904,
+      "loss": 5.1276,
+      "step": 493
+    },
+    {
+      "epoch": 0.16860068259385666,
+      "grad_norm": 5.211971282958984,
+      "learning_rate": 0.0009437997724687145,
+      "loss": 6.1224,
+      "step": 494
+    },
+    {
+      "epoch": 0.1689419795221843,
+      "grad_norm": 3.662050485610962,
+      "learning_rate": 0.0009436860068259387,
+      "loss": 6.9951,
+      "step": 495
+    },
+    {
+      "epoch": 0.16928327645051194,
+      "grad_norm": 3.8950295448303223,
+      "learning_rate": 0.0009435722411831627,
+      "loss": 6.9776,
+      "step": 496
+    },
+    {
+      "epoch": 0.1696245733788396,
+      "grad_norm": 3.697416067123413,
+      "learning_rate": 0.0009434584755403868,
+      "loss": 6.5273,
+      "step": 497
+    },
+    {
+      "epoch": 0.16996587030716723,
+      "grad_norm": 3.4824891090393066,
+      "learning_rate": 0.0009433447098976109,
+      "loss": 6.9454,
+      "step": 498
+    },
+    {
+      "epoch": 0.17030716723549488,
+      "grad_norm": 3.859316825866699,
+      "learning_rate": 0.000943230944254835,
+      "loss": 5.7186,
+      "step": 499
+    },
+    {
+      "epoch": 0.17064846416382254,
+      "grad_norm": 8.799308776855469,
+      "learning_rate": 0.0009431171786120591,
+      "loss": 6.3941,
+      "step": 500
+    },
+    {
+      "epoch": 0.17098976109215017,
+      "grad_norm": 3.745943069458008,
+      "learning_rate": 0.0009430034129692833,
+      "loss": 7.0287,
+      "step": 501
+    },
+    {
+      "epoch": 0.17133105802047782,
+      "grad_norm": 3.7031750679016113,
+      "learning_rate": 0.0009428896473265074,
+      "loss": 6.6921,
+      "step": 502
+    },
+    {
+      "epoch": 0.17167235494880545,
+      "grad_norm": 3.5170814990997314,
+      "learning_rate": 0.0009427758816837315,
+      "loss": 7.0773,
+      "step": 503
+    },
+    {
+      "epoch": 0.1720136518771331,
+      "grad_norm": 3.4774396419525146,
+      "learning_rate": 0.0009426621160409556,
+      "loss": 6.9226,
+      "step": 504
+    },
+    {
+      "epoch": 0.17235494880546076,
+      "grad_norm": 3.5223052501678467,
+      "learning_rate": 0.0009425483503981797,
+      "loss": 6.7229,
+      "step": 505
+    },
+    {
+      "epoch": 0.1726962457337884,
+      "grad_norm": 4.429044723510742,
+      "learning_rate": 0.0009424345847554038,
+      "loss": 6.1489,
+      "step": 506
+    },
+    {
+      "epoch": 0.17303754266211605,
+      "grad_norm": 3.946762800216675,
+      "learning_rate": 0.0009423208191126281,
+      "loss": 6.3177,
+      "step": 507
+    },
+    {
+      "epoch": 0.17337883959044367,
+      "grad_norm": 4.3716278076171875,
+      "learning_rate": 0.0009422070534698522,
+      "loss": 7.4142,
+      "step": 508
+    },
+    {
+      "epoch": 0.17372013651877133,
+      "grad_norm": 4.144838809967041,
+      "learning_rate": 0.0009420932878270763,
+      "loss": 6.5615,
+      "step": 509
+    },
+    {
+      "epoch": 0.17406143344709898,
+      "grad_norm": 3.442716121673584,
+      "learning_rate": 0.0009419795221843004,
+      "loss": 7.1919,
+      "step": 510
+    },
+    {
+      "epoch": 0.1744027303754266,
+      "grad_norm": 3.616173505783081,
+      "learning_rate": 0.0009418657565415245,
+      "loss": 7.0364,
+      "step": 511
+    },
+    {
+      "epoch": 0.17474402730375427,
+      "grad_norm": 3.2651257514953613,
+      "learning_rate": 0.0009417519908987486,
+      "loss": 7.1869,
+      "step": 512
+    },
+    {
+      "epoch": 0.17508532423208192,
+      "grad_norm": 3.429654836654663,
+      "learning_rate": 0.0009416382252559728,
+      "loss": 7.3759,
+      "step": 513
+    },
+    {
+      "epoch": 0.17542662116040955,
+      "grad_norm": 3.8366029262542725,
+      "learning_rate": 0.0009415244596131969,
+      "loss": 6.5422,
+      "step": 514
+    },
+    {
+      "epoch": 0.1757679180887372,
+      "grad_norm": 6.700514793395996,
+      "learning_rate": 0.000941410693970421,
+      "loss": 5.9699,
+      "step": 515
+    },
+    {
+      "epoch": 0.17610921501706484,
+      "grad_norm": 3.9318792819976807,
+      "learning_rate": 0.000941296928327645,
+      "loss": 7.1496,
+      "step": 516
+    },
+    {
+      "epoch": 0.1764505119453925,
+      "grad_norm": 3.7701475620269775,
+      "learning_rate": 0.0009411831626848691,
+      "loss": 7.1363,
+      "step": 517
+    },
+    {
+      "epoch": 0.17679180887372015,
+      "grad_norm": 3.5054924488067627,
+      "learning_rate": 0.0009410693970420933,
+      "loss": 6.9795,
+      "step": 518
+    },
+    {
+      "epoch": 0.17713310580204777,
+      "grad_norm": 3.7713727951049805,
+      "learning_rate": 0.0009409556313993174,
+      "loss": 6.6172,
+      "step": 519
+    },
+    {
+      "epoch": 0.17747440273037543,
+      "grad_norm": 8.710354804992676,
+      "learning_rate": 0.0009408418657565415,
+      "loss": 6.0337,
+      "step": 520
+    },
+    {
+      "epoch": 0.17781569965870306,
+      "grad_norm": 4.916598320007324,
+      "learning_rate": 0.0009407281001137656,
+      "loss": 6.0356,
+      "step": 521
+    },
+    {
+      "epoch": 0.1781569965870307,
+      "grad_norm": 3.8350720405578613,
+      "learning_rate": 0.0009406143344709897,
+      "loss": 7.201,
+      "step": 522
+    },
+    {
+      "epoch": 0.17849829351535837,
+      "grad_norm": 3.7248024940490723,
+      "learning_rate": 0.0009405005688282138,
+      "loss": 6.5138,
+      "step": 523
+    },
+    {
+      "epoch": 0.178839590443686,
+      "grad_norm": 3.730102062225342,
+      "learning_rate": 0.0009403868031854381,
+      "loss": 6.3692,
+      "step": 524
+    },
+    {
+      "epoch": 0.17918088737201365,
+      "grad_norm": 3.4296181201934814,
+      "learning_rate": 0.0009402730375426622,
+      "loss": 6.7243,
+      "step": 525
+    },
+    {
+      "epoch": 0.1795221843003413,
+      "grad_norm": 3.4603865146636963,
+      "learning_rate": 0.0009401592718998863,
+      "loss": 7.2087,
+      "step": 526
+    },
+    {
+      "epoch": 0.17986348122866894,
+      "grad_norm": 3.4255411624908447,
+      "learning_rate": 0.0009400455062571104,
+      "loss": 6.6606,
+      "step": 527
+    },
+    {
+      "epoch": 0.1802047781569966,
+      "grad_norm": 4.901156902313232,
+      "learning_rate": 0.0009399317406143345,
+      "loss": 6.4545,
+      "step": 528
+    },
+    {
+      "epoch": 0.18054607508532422,
+      "grad_norm": 3.5273005962371826,
+      "learning_rate": 0.0009398179749715586,
+      "loss": 6.8521,
+      "step": 529
+    },
+    {
+      "epoch": 0.18088737201365188,
+      "grad_norm": 3.794410467147827,
+      "learning_rate": 0.0009397042093287828,
+      "loss": 6.5531,
+      "step": 530
+    },
+    {
+      "epoch": 0.18122866894197953,
+      "grad_norm": 3.8027398586273193,
+      "learning_rate": 0.0009395904436860069,
+      "loss": 6.9897,
+      "step": 531
+    },
+    {
+      "epoch": 0.18156996587030716,
+      "grad_norm": 3.6400845050811768,
+      "learning_rate": 0.000939476678043231,
+      "loss": 6.4599,
+      "step": 532
+    },
+    {
+      "epoch": 0.18191126279863482,
+      "grad_norm": 3.6795458793640137,
+      "learning_rate": 0.0009393629124004551,
+      "loss": 7.0436,
+      "step": 533
+    },
+    {
+      "epoch": 0.18225255972696247,
+      "grad_norm": 3.549872398376465,
+      "learning_rate": 0.0009392491467576792,
+      "loss": 6.8872,
+      "step": 534
+    },
+    {
+      "epoch": 0.1825938566552901,
+      "grad_norm": 3.6060047149658203,
+      "learning_rate": 0.0009391353811149032,
+      "loss": 6.8889,
+      "step": 535
+    },
+    {
+      "epoch": 0.18293515358361775,
+      "grad_norm": 3.5164926052093506,
+      "learning_rate": 0.0009390216154721274,
+      "loss": 7.2238,
+      "step": 536
+    },
+    {
+      "epoch": 0.18327645051194538,
+      "grad_norm": 4.204543590545654,
+      "learning_rate": 0.0009389078498293515,
+      "loss": 6.8331,
+      "step": 537
+    },
+    {
+      "epoch": 0.18361774744027304,
+      "grad_norm": 3.713944673538208,
+      "learning_rate": 0.0009387940841865756,
+      "loss": 6.3832,
+      "step": 538
+    },
+    {
+      "epoch": 0.1839590443686007,
+      "grad_norm": 3.504955530166626,
+      "learning_rate": 0.0009386803185437997,
+      "loss": 6.9558,
+      "step": 539
+    },
+    {
+      "epoch": 0.18430034129692832,
+      "grad_norm": 3.7779481410980225,
+      "learning_rate": 0.0009385665529010238,
+      "loss": 6.7189,
+      "step": 540
+    },
+    {
+      "epoch": 0.18464163822525598,
+      "grad_norm": 3.7830755710601807,
+      "learning_rate": 0.0009384527872582481,
+      "loss": 7.0508,
+      "step": 541
+    },
+    {
+      "epoch": 0.1849829351535836,
+      "grad_norm": 3.6746010780334473,
+      "learning_rate": 0.0009383390216154722,
+      "loss": 6.6615,
+      "step": 542
+    },
+    {
+      "epoch": 0.18532423208191126,
+      "grad_norm": 3.6559464931488037,
+      "learning_rate": 0.0009382252559726963,
+      "loss": 7.2021,
+      "step": 543
+    },
+    {
+      "epoch": 0.18566552901023892,
+      "grad_norm": 3.4099745750427246,
+      "learning_rate": 0.0009381114903299204,
+      "loss": 6.8867,
+      "step": 544
+    },
+    {
+      "epoch": 0.18600682593856654,
+      "grad_norm": 3.7949607372283936,
+      "learning_rate": 0.0009379977246871445,
+      "loss": 6.3174,
+      "step": 545
+    },
+    {
+      "epoch": 0.1863481228668942,
+      "grad_norm": 3.6113579273223877,
+      "learning_rate": 0.0009378839590443686,
+      "loss": 7.2956,
+      "step": 546
+    },
+    {
+      "epoch": 0.18668941979522186,
+      "grad_norm": 4.320201396942139,
+      "learning_rate": 0.0009377701934015928,
+      "loss": 7.0727,
+      "step": 547
+    },
+    {
+      "epoch": 0.18703071672354948,
+      "grad_norm": 3.824106454849243,
+      "learning_rate": 0.0009376564277588169,
+      "loss": 6.5791,
+      "step": 548
+    },
+    {
+      "epoch": 0.18737201365187714,
+      "grad_norm": 3.5648560523986816,
+      "learning_rate": 0.000937542662116041,
+      "loss": 7.0372,
+      "step": 549
+    },
+    {
+      "epoch": 0.18771331058020477,
+      "grad_norm": 6.90482234954834,
+      "learning_rate": 0.0009374288964732651,
+      "loss": 6.3724,
+      "step": 550
+    },
+    {
+      "epoch": 0.18805460750853242,
+      "grad_norm": 3.8881046772003174,
+      "learning_rate": 0.0009373151308304892,
+      "loss": 6.7275,
+      "step": 551
+    },
+    {
+      "epoch": 0.18839590443686008,
+      "grad_norm": 3.4772567749023438,
+      "learning_rate": 0.0009372013651877133,
+      "loss": 7.1439,
+      "step": 552
+    },
+    {
+      "epoch": 0.1887372013651877,
+      "grad_norm": 3.674238681793213,
+      "learning_rate": 0.0009370875995449375,
+      "loss": 7.3201,
+      "step": 553
+    },
+    {
+      "epoch": 0.18907849829351536,
+      "grad_norm": 3.3719608783721924,
+      "learning_rate": 0.0009369738339021616,
+      "loss": 6.646,
+      "step": 554
+    },
+    {
+      "epoch": 0.189419795221843,
+      "grad_norm": 3.6703298091888428,
+      "learning_rate": 0.0009368600682593856,
+      "loss": 6.7161,
+      "step": 555
+    },
+    {
+      "epoch": 0.18976109215017065,
+      "grad_norm": 3.6245388984680176,
+      "learning_rate": 0.0009367463026166097,
+      "loss": 6.3849,
+      "step": 556
+    },
+    {
+      "epoch": 0.1901023890784983,
+      "grad_norm": 3.5778818130493164,
+      "learning_rate": 0.0009366325369738338,
+      "loss": 7.4046,
+      "step": 557
+    },
+    {
+      "epoch": 0.19044368600682593,
+      "grad_norm": 4.9418253898620605,
+      "learning_rate": 0.000936518771331058,
+      "loss": 6.5024,
+      "step": 558
+    },
+    {
+      "epoch": 0.19078498293515359,
+      "grad_norm": 3.6862545013427734,
+      "learning_rate": 0.0009364050056882822,
+      "loss": 6.7619,
+      "step": 559
+    },
+    {
+      "epoch": 0.19112627986348124,
+      "grad_norm": 4.22722864151001,
+      "learning_rate": 0.0009362912400455063,
+      "loss": 6.5308,
+      "step": 560
+    },
+    {
+      "epoch": 0.19146757679180887,
+      "grad_norm": 3.733459949493408,
+      "learning_rate": 0.0009361774744027304,
+      "loss": 6.5565,
+      "step": 561
+    },
+    {
+      "epoch": 0.19180887372013652,
+      "grad_norm": 9.069499015808105,
+      "learning_rate": 0.0009360637087599545,
+      "loss": 4.9708,
+      "step": 562
+    },
+    {
+      "epoch": 0.19215017064846415,
+      "grad_norm": 3.968690872192383,
+      "learning_rate": 0.0009359499431171786,
+      "loss": 7.3263,
+      "step": 563
+    },
+    {
+      "epoch": 0.1924914675767918,
+      "grad_norm": 3.5820865631103516,
+      "learning_rate": 0.0009358361774744028,
+      "loss": 6.7364,
+      "step": 564
+    },
+    {
+      "epoch": 0.19283276450511946,
+      "grad_norm": 3.6914074420928955,
+      "learning_rate": 0.0009357224118316269,
+      "loss": 6.8361,
+      "step": 565
+    },
+    {
+      "epoch": 0.1931740614334471,
+      "grad_norm": 3.580321788787842,
+      "learning_rate": 0.000935608646188851,
+      "loss": 6.4783,
+      "step": 566
+    },
+    {
+      "epoch": 0.19351535836177475,
+      "grad_norm": 3.5576207637786865,
+      "learning_rate": 0.0009354948805460751,
+      "loss": 6.5367,
+      "step": 567
+    },
+    {
+      "epoch": 0.19385665529010238,
+      "grad_norm": 3.515730619430542,
+      "learning_rate": 0.0009353811149032992,
+      "loss": 6.481,
+      "step": 568
+    },
+    {
+      "epoch": 0.19419795221843003,
+      "grad_norm": 3.8840041160583496,
+      "learning_rate": 0.0009352673492605233,
+      "loss": 6.4732,
+      "step": 569
+    },
+    {
+      "epoch": 0.1945392491467577,
+      "grad_norm": 3.75571346282959,
+      "learning_rate": 0.0009351535836177475,
+      "loss": 7.1131,
+      "step": 570
+    },
+    {
+      "epoch": 0.19488054607508531,
+      "grad_norm": 4.804229259490967,
+      "learning_rate": 0.0009350398179749716,
+      "loss": 6.1646,
+      "step": 571
+    },
+    {
+      "epoch": 0.19522184300341297,
+      "grad_norm": 3.6777968406677246,
+      "learning_rate": 0.0009349260523321957,
+      "loss": 6.8409,
+      "step": 572
+    },
+    {
+      "epoch": 0.19556313993174063,
+      "grad_norm": 3.6401546001434326,
+      "learning_rate": 0.0009348122866894199,
+      "loss": 7.1515,
+      "step": 573
+    },
+    {
+      "epoch": 0.19590443686006825,
+      "grad_norm": 3.532172679901123,
+      "learning_rate": 0.0009346985210466438,
+      "loss": 6.3835,
+      "step": 574
+    },
+    {
+      "epoch": 0.1962457337883959,
+      "grad_norm": 3.4800662994384766,
+      "learning_rate": 0.000934584755403868,
+      "loss": 7.2508,
+      "step": 575
+    },
+    {
+      "epoch": 0.19658703071672354,
+      "grad_norm": 3.7157084941864014,
+      "learning_rate": 0.0009344709897610922,
+      "loss": 6.508,
+      "step": 576
+    },
+    {
+      "epoch": 0.1969283276450512,
+      "grad_norm": 3.7525947093963623,
+      "learning_rate": 0.0009343572241183163,
+      "loss": 6.3877,
+      "step": 577
+    },
+    {
+      "epoch": 0.19726962457337885,
+      "grad_norm": 3.6418375968933105,
+      "learning_rate": 0.0009342434584755404,
+      "loss": 6.7508,
+      "step": 578
+    },
+    {
+      "epoch": 0.19761092150170648,
+      "grad_norm": 4.113407135009766,
+      "learning_rate": 0.0009341296928327645,
+      "loss": 6.676,
+      "step": 579
+    },
+    {
+      "epoch": 0.19795221843003413,
+      "grad_norm": 3.6704797744750977,
+      "learning_rate": 0.0009340159271899886,
+      "loss": 6.9985,
+      "step": 580
+    },
+    {
+      "epoch": 0.1982935153583618,
+      "grad_norm": 3.590165138244629,
+      "learning_rate": 0.0009339021615472128,
+      "loss": 6.6784,
+      "step": 581
+    },
+    {
+      "epoch": 0.19863481228668942,
+      "grad_norm": 3.575233221054077,
+      "learning_rate": 0.0009337883959044369,
+      "loss": 6.4992,
+      "step": 582
+    },
+    {
+      "epoch": 0.19897610921501707,
+      "grad_norm": 3.404895782470703,
+      "learning_rate": 0.000933674630261661,
+      "loss": 6.9338,
+      "step": 583
+    },
+    {
+      "epoch": 0.1993174061433447,
+      "grad_norm": 3.5058460235595703,
+      "learning_rate": 0.0009335608646188851,
+      "loss": 7.2116,
+      "step": 584
+    },
+    {
+      "epoch": 0.19965870307167236,
+      "grad_norm": 3.462622880935669,
+      "learning_rate": 0.0009334470989761092,
+      "loss": 7.0462,
+      "step": 585
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.042704105377197,
+      "learning_rate": 0.0009333333333333333,
+      "loss": 5.7069,
+      "step": 586
+    },
+    {
+      "epoch": 0.20034129692832764,
+      "grad_norm": 3.7907888889312744,
+      "learning_rate": 0.0009332195676905575,
+      "loss": 6.7271,
+      "step": 587
+    },
+    {
+      "epoch": 0.2006825938566553,
+      "grad_norm": 3.531925916671753,
+      "learning_rate": 0.0009331058020477816,
+      "loss": 7.0678,
+      "step": 588
+    },
+    {
+      "epoch": 0.20102389078498292,
+      "grad_norm": 3.567275285720825,
+      "learning_rate": 0.0009329920364050057,
+      "loss": 6.9811,
+      "step": 589
+    },
+    {
+      "epoch": 0.20136518771331058,
+      "grad_norm": 3.8175978660583496,
+      "learning_rate": 0.0009328782707622299,
+      "loss": 6.6779,
+      "step": 590
+    },
+    {
+      "epoch": 0.20170648464163823,
+      "grad_norm": 3.672842502593994,
+      "learning_rate": 0.000932764505119454,
+      "loss": 6.6118,
+      "step": 591
+    },
+    {
+      "epoch": 0.20204778156996586,
+      "grad_norm": 3.625286102294922,
+      "learning_rate": 0.0009326507394766781,
+      "loss": 7.1495,
+      "step": 592
+    },
+    {
+      "epoch": 0.20238907849829352,
+      "grad_norm": 3.4718549251556396,
+      "learning_rate": 0.0009325369738339023,
+      "loss": 7.5187,
+      "step": 593
+    },
+    {
+      "epoch": 0.20273037542662117,
+      "grad_norm": 3.5792765617370605,
+      "learning_rate": 0.0009324232081911263,
+      "loss": 6.7316,
+      "step": 594
+    },
+    {
+      "epoch": 0.2030716723549488,
+      "grad_norm": 19.332000732421875,
+      "learning_rate": 0.0009323094425483504,
+      "loss": 6.0635,
+      "step": 595
+    },
+    {
+      "epoch": 0.20341296928327646,
+      "grad_norm": 3.741669178009033,
+      "learning_rate": 0.0009321956769055745,
+      "loss": 7.1053,
+      "step": 596
+    },
+    {
+      "epoch": 0.20375426621160408,
+      "grad_norm": 4.053689002990723,
+      "learning_rate": 0.0009320819112627986,
+      "loss": 6.9079,
+      "step": 597
+    },
+    {
+      "epoch": 0.20409556313993174,
+      "grad_norm": 3.5358211994171143,
+      "learning_rate": 0.0009319681456200227,
+      "loss": 6.9465,
+      "step": 598
+    },
+    {
+      "epoch": 0.2044368600682594,
+      "grad_norm": 3.4627411365509033,
+      "learning_rate": 0.0009318543799772469,
+      "loss": 7.1842,
+      "step": 599
+    },
+    {
+      "epoch": 0.20477815699658702,
+      "grad_norm": 3.443898916244507,
+      "learning_rate": 0.000931740614334471,
+      "loss": 6.5099,
+      "step": 600
+    },
+    {
+      "epoch": 0.20511945392491468,
+      "grad_norm": 3.4302115440368652,
+      "learning_rate": 0.0009316268486916951,
+      "loss": 6.896,
+      "step": 601
+    },
+    {
+      "epoch": 0.2054607508532423,
+      "grad_norm": 3.346616268157959,
+      "learning_rate": 0.0009315130830489192,
+      "loss": 6.8985,
+      "step": 602
+    },
+    {
+      "epoch": 0.20580204778156996,
+      "grad_norm": 3.753207206726074,
+      "learning_rate": 0.0009313993174061433,
+      "loss": 6.4971,
+      "step": 603
+    },
+    {
+      "epoch": 0.20614334470989762,
+      "grad_norm": 3.675739049911499,
+      "learning_rate": 0.0009312855517633675,
+      "loss": 6.6612,
+      "step": 604
+    },
+    {
+      "epoch": 0.20648464163822525,
+      "grad_norm": 3.6538069248199463,
+      "learning_rate": 0.0009311717861205916,
+      "loss": 7.0008,
+      "step": 605
+    },
+    {
+      "epoch": 0.2068259385665529,
+      "grad_norm": 3.7137982845306396,
+      "learning_rate": 0.0009310580204778157,
+      "loss": 7.0696,
+      "step": 606
+    },
+    {
+      "epoch": 0.20716723549488056,
+      "grad_norm": 3.7337138652801514,
+      "learning_rate": 0.0009309442548350399,
+      "loss": 6.686,
+      "step": 607
+    },
+    {
+      "epoch": 0.2075085324232082,
+      "grad_norm": 7.5925374031066895,
+      "learning_rate": 0.000930830489192264,
+      "loss": 6.9714,
+      "step": 608
+    },
+    {
+      "epoch": 0.20784982935153584,
+      "grad_norm": 3.7731120586395264,
+      "learning_rate": 0.0009307167235494881,
+      "loss": 6.924,
+      "step": 609
+    },
+    {
+      "epoch": 0.20819112627986347,
+      "grad_norm": 6.881464004516602,
+      "learning_rate": 0.0009306029579067123,
+      "loss": 6.9183,
+      "step": 610
+    },
+    {
+      "epoch": 0.20853242320819113,
+      "grad_norm": 3.973215103149414,
+      "learning_rate": 0.0009304891922639364,
+      "loss": 6.7891,
+      "step": 611
+    },
+    {
+      "epoch": 0.20887372013651878,
+      "grad_norm": 3.4739882946014404,
+      "learning_rate": 0.0009303754266211605,
+      "loss": 7.1653,
+      "step": 612
+    },
+    {
+      "epoch": 0.2092150170648464,
+      "grad_norm": 3.3610455989837646,
+      "learning_rate": 0.0009302616609783845,
+      "loss": 6.6231,
+      "step": 613
+    },
+    {
+      "epoch": 0.20955631399317406,
+      "grad_norm": 3.4226975440979004,
+      "learning_rate": 0.0009301478953356086,
+      "loss": 7.0169,
+      "step": 614
+    },
+    {
+      "epoch": 0.2098976109215017,
+      "grad_norm": 4.903579235076904,
+      "learning_rate": 0.0009300341296928327,
+      "loss": 6.194,
+      "step": 615
+    },
+    {
+      "epoch": 0.21023890784982935,
+      "grad_norm": 3.5574898719787598,
+      "learning_rate": 0.0009299203640500569,
+      "loss": 7.0527,
+      "step": 616
+    },
+    {
+      "epoch": 0.210580204778157,
+      "grad_norm": 3.566174030303955,
+      "learning_rate": 0.000929806598407281,
+      "loss": 6.8922,
+      "step": 617
+    },
+    {
+      "epoch": 0.21092150170648463,
+      "grad_norm": 4.031206130981445,
+      "learning_rate": 0.0009296928327645051,
+      "loss": 5.8383,
+      "step": 618
+    },
+    {
+      "epoch": 0.2112627986348123,
+      "grad_norm": 3.552882671356201,
+      "learning_rate": 0.0009295790671217292,
+      "loss": 6.6364,
+      "step": 619
+    },
+    {
+      "epoch": 0.21160409556313994,
+      "grad_norm": 6.967950820922852,
+      "learning_rate": 0.0009294653014789533,
+      "loss": 6.5468,
+      "step": 620
+    },
+    {
+      "epoch": 0.21194539249146757,
+      "grad_norm": 3.7092936038970947,
+      "learning_rate": 0.0009293515358361775,
+      "loss": 7.2022,
+      "step": 621
+    },
+    {
+      "epoch": 0.21228668941979523,
+      "grad_norm": 3.346576690673828,
+      "learning_rate": 0.0009292377701934016,
+      "loss": 6.4981,
+      "step": 622
+    },
+    {
+      "epoch": 0.21262798634812285,
+      "grad_norm": 3.6055033206939697,
+      "learning_rate": 0.0009291240045506257,
+      "loss": 6.9378,
+      "step": 623
+    },
+    {
+      "epoch": 0.2129692832764505,
+      "grad_norm": 3.5705137252807617,
+      "learning_rate": 0.0009290102389078499,
+      "loss": 7.0167,
+      "step": 624
+    },
+    {
+      "epoch": 0.21331058020477817,
+      "grad_norm": 4.07545280456543,
+      "learning_rate": 0.000928896473265074,
+      "loss": 6.4161,
+      "step": 625
+    },
+    {
+      "epoch": 0.2136518771331058,
+      "grad_norm": 3.458583116531372,
+      "learning_rate": 0.0009287827076222981,
+      "loss": 6.7419,
+      "step": 626
+    },
+    {
+      "epoch": 0.21399317406143345,
+      "grad_norm": 3.4590044021606445,
+      "learning_rate": 0.0009286689419795223,
+      "loss": 6.5505,
+      "step": 627
+    },
+    {
+      "epoch": 0.2143344709897611,
+      "grad_norm": 3.5659596920013428,
+      "learning_rate": 0.0009285551763367464,
+      "loss": 6.6099,
+      "step": 628
+    },
+    {
+      "epoch": 0.21467576791808873,
+      "grad_norm": 3.5950722694396973,
+      "learning_rate": 0.0009284414106939705,
+      "loss": 6.6846,
+      "step": 629
+    },
+    {
+      "epoch": 0.2150170648464164,
+      "grad_norm": 3.622309923171997,
+      "learning_rate": 0.0009283276450511946,
+      "loss": 6.7184,
+      "step": 630
+    },
+    {
+      "epoch": 0.21535836177474402,
+      "grad_norm": 3.5090525150299072,
+      "learning_rate": 0.0009282138794084187,
+      "loss": 7.0747,
+      "step": 631
+    },
+    {
+      "epoch": 0.21569965870307167,
+      "grad_norm": 3.4692866802215576,
+      "learning_rate": 0.0009281001137656428,
+      "loss": 7.4671,
+      "step": 632
+    },
+    {
+      "epoch": 0.21604095563139933,
+      "grad_norm": 3.454890489578247,
+      "learning_rate": 0.0009279863481228669,
+      "loss": 7.0538,
+      "step": 633
+    },
+    {
+      "epoch": 0.21638225255972696,
+      "grad_norm": 5.090976715087891,
+      "learning_rate": 0.000927872582480091,
+      "loss": 6.5268,
+      "step": 634
+    },
+    {
+      "epoch": 0.2167235494880546,
+      "grad_norm": 3.852503776550293,
+      "learning_rate": 0.0009277588168373151,
+      "loss": 6.9396,
+      "step": 635
+    },
+    {
+      "epoch": 0.21706484641638224,
+      "grad_norm": 3.7737314701080322,
+      "learning_rate": 0.0009276450511945392,
+      "loss": 6.7992,
+      "step": 636
+    },
+    {
+      "epoch": 0.2174061433447099,
+      "grad_norm": 3.6202504634857178,
+      "learning_rate": 0.0009275312855517633,
+      "loss": 7.4372,
+      "step": 637
+    },
+    {
+      "epoch": 0.21774744027303755,
+      "grad_norm": 3.5939218997955322,
+      "learning_rate": 0.0009274175199089874,
+      "loss": 6.6953,
+      "step": 638
+    },
+    {
+      "epoch": 0.21808873720136518,
+      "grad_norm": 3.323547601699829,
+      "learning_rate": 0.0009273037542662116,
+      "loss": 6.7816,
+      "step": 639
+    },
+    {
+      "epoch": 0.21843003412969283,
+      "grad_norm": 3.6134841442108154,
+      "learning_rate": 0.0009271899886234357,
+      "loss": 6.6858,
+      "step": 640
+    },
+    {
+      "epoch": 0.2187713310580205,
+      "grad_norm": 3.660320520401001,
+      "learning_rate": 0.0009270762229806599,
+      "loss": 7.0058,
+      "step": 641
+    },
+    {
+      "epoch": 0.21911262798634812,
+      "grad_norm": 3.380748748779297,
+      "learning_rate": 0.000926962457337884,
+      "loss": 6.735,
+      "step": 642
+    },
+    {
+      "epoch": 0.21945392491467577,
+      "grad_norm": 6.264308452606201,
+      "learning_rate": 0.0009268486916951081,
+      "loss": 5.8211,
+      "step": 643
+    },
+    {
+      "epoch": 0.2197952218430034,
+      "grad_norm": 3.7562389373779297,
+      "learning_rate": 0.0009267349260523323,
+      "loss": 6.7083,
+      "step": 644
+    },
+    {
+      "epoch": 0.22013651877133106,
+      "grad_norm": 3.7538766860961914,
+      "learning_rate": 0.0009266211604095564,
+      "loss": 6.8536,
+      "step": 645
+    },
+    {
+      "epoch": 0.2204778156996587,
+      "grad_norm": 3.648890972137451,
+      "learning_rate": 0.0009265073947667805,
+      "loss": 6.6039,
+      "step": 646
+    },
+    {
+      "epoch": 0.22081911262798634,
+      "grad_norm": 3.616142988204956,
+      "learning_rate": 0.0009263936291240046,
+      "loss": 6.0385,
+      "step": 647
+    },
+    {
+      "epoch": 0.221160409556314,
+      "grad_norm": 3.6815671920776367,
+      "learning_rate": 0.0009262798634812287,
+      "loss": 7.124,
+      "step": 648
+    },
+    {
+      "epoch": 0.22150170648464163,
+      "grad_norm": 3.596876382827759,
+      "learning_rate": 0.0009261660978384528,
+      "loss": 6.5317,
+      "step": 649
+    },
+    {
+      "epoch": 0.22184300341296928,
+      "grad_norm": 3.4325168132781982,
+      "learning_rate": 0.000926052332195677,
+      "loss": 6.7055,
+      "step": 650
+    },
+    {
+      "epoch": 0.22218430034129694,
+      "grad_norm": 3.43967604637146,
+      "learning_rate": 0.0009259385665529011,
+      "loss": 6.9872,
+      "step": 651
+    },
+    {
+      "epoch": 0.22252559726962456,
+      "grad_norm": 3.4720215797424316,
+      "learning_rate": 0.0009258248009101251,
+      "loss": 6.807,
+      "step": 652
+    },
+    {
+      "epoch": 0.22286689419795222,
+      "grad_norm": 4.370832920074463,
+      "learning_rate": 0.0009257110352673492,
+      "loss": 5.8649,
+      "step": 653
+    },
+    {
+      "epoch": 0.22320819112627988,
+      "grad_norm": 3.566333293914795,
+      "learning_rate": 0.0009255972696245733,
+      "loss": 6.6208,
+      "step": 654
+    },
+    {
+      "epoch": 0.2235494880546075,
+      "grad_norm": 3.5970399379730225,
+      "learning_rate": 0.0009254835039817974,
+      "loss": 7.0027,
+      "step": 655
+    },
+    {
+      "epoch": 0.22389078498293516,
+      "grad_norm": 3.698763370513916,
+      "learning_rate": 0.0009253697383390216,
+      "loss": 7.0716,
+      "step": 656
+    },
+    {
+      "epoch": 0.2242320819112628,
+      "grad_norm": 3.4119129180908203,
+      "learning_rate": 0.0009252559726962458,
+      "loss": 6.8046,
+      "step": 657
+    },
+    {
+      "epoch": 0.22457337883959044,
+      "grad_norm": 3.6655516624450684,
+      "learning_rate": 0.0009251422070534699,
+      "loss": 6.9696,
+      "step": 658
+    },
+    {
+      "epoch": 0.2249146757679181,
+      "grad_norm": 5.750580787658691,
+      "learning_rate": 0.000925028441410694,
+      "loss": 5.6451,
+      "step": 659
+    },
+    {
+      "epoch": 0.22525597269624573,
+      "grad_norm": 3.9216561317443848,
+      "learning_rate": 0.0009249146757679181,
+      "loss": 6.4134,
+      "step": 660
+    },
+    {
+      "epoch": 0.22559726962457338,
+      "grad_norm": 4.239558696746826,
+      "learning_rate": 0.0009248009101251423,
+      "loss": 5.6066,
+      "step": 661
+    },
+    {
+      "epoch": 0.225938566552901,
+      "grad_norm": 3.7342917919158936,
+      "learning_rate": 0.0009246871444823664,
+      "loss": 7.0818,
+      "step": 662
+    },
+    {
+      "epoch": 0.22627986348122867,
+      "grad_norm": 5.248586177825928,
+      "learning_rate": 0.0009245733788395905,
+      "loss": 5.4571,
+      "step": 663
+    },
+    {
+      "epoch": 0.22662116040955632,
+      "grad_norm": 3.6075022220611572,
+      "learning_rate": 0.0009244596131968146,
+      "loss": 6.9643,
+      "step": 664
+    },
+    {
+      "epoch": 0.22696245733788395,
+      "grad_norm": 3.528815269470215,
+      "learning_rate": 0.0009243458475540387,
+      "loss": 7.1822,
+      "step": 665
+    },
+    {
+      "epoch": 0.2273037542662116,
+      "grad_norm": 3.4905171394348145,
+      "learning_rate": 0.0009242320819112628,
+      "loss": 7.0671,
+      "step": 666
+    },
+    {
+      "epoch": 0.22764505119453926,
+      "grad_norm": 3.64202880859375,
+      "learning_rate": 0.000924118316268487,
+      "loss": 6.6831,
+      "step": 667
+    },
+    {
+      "epoch": 0.2279863481228669,
+      "grad_norm": 6.533376216888428,
+      "learning_rate": 0.0009240045506257111,
+      "loss": 6.1337,
+      "step": 668
+    },
+    {
+      "epoch": 0.22832764505119454,
+      "grad_norm": 3.9284870624542236,
+      "learning_rate": 0.0009238907849829352,
+      "loss": 6.5213,
+      "step": 669
+    },
+    {
+      "epoch": 0.22866894197952217,
+      "grad_norm": 3.7156975269317627,
+      "learning_rate": 0.0009237770193401593,
+      "loss": 6.5675,
+      "step": 670
+    },
+    {
+      "epoch": 0.22901023890784983,
+      "grad_norm": 4.064757347106934,
+      "learning_rate": 0.0009236632536973833,
+      "loss": 7.1138,
+      "step": 671
+    },
+    {
+      "epoch": 0.22935153583617748,
+      "grad_norm": 5.838778018951416,
+      "learning_rate": 0.0009235494880546074,
+      "loss": 6.6873,
+      "step": 672
+    },
+    {
+      "epoch": 0.2296928327645051,
+      "grad_norm": 3.700157642364502,
+      "learning_rate": 0.0009234357224118316,
+      "loss": 6.5795,
+      "step": 673
+    },
+    {
+      "epoch": 0.23003412969283277,
+      "grad_norm": 3.6335175037384033,
+      "learning_rate": 0.0009233219567690558,
+      "loss": 6.5419,
+      "step": 674
+    },
+    {
+      "epoch": 0.23037542662116042,
+      "grad_norm": 3.073715925216675,
+      "learning_rate": 0.0009232081911262799,
+      "loss": 6.6049,
+      "step": 675
+    },
+    {
+      "epoch": 0.23071672354948805,
+      "grad_norm": 3.373309850692749,
+      "learning_rate": 0.000923094425483504,
+      "loss": 6.7097,
+      "step": 676
+    },
+    {
+      "epoch": 0.2310580204778157,
+      "grad_norm": 3.3596081733703613,
+      "learning_rate": 0.0009229806598407281,
+      "loss": 6.8117,
+      "step": 677
+    },
+    {
+      "epoch": 0.23139931740614333,
+      "grad_norm": 3.4036943912506104,
+      "learning_rate": 0.0009228668941979522,
+      "loss": 7.1459,
+      "step": 678
+    },
+    {
+      "epoch": 0.231740614334471,
+      "grad_norm": 5.635969638824463,
+      "learning_rate": 0.0009227531285551764,
+      "loss": 5.9685,
+      "step": 679
+    },
+    {
+      "epoch": 0.23208191126279865,
+      "grad_norm": 4.183934688568115,
+      "learning_rate": 0.0009226393629124005,
+      "loss": 6.5457,
+      "step": 680
+    },
+    {
+      "epoch": 0.23242320819112627,
+      "grad_norm": 3.8760788440704346,
+      "learning_rate": 0.0009225255972696246,
+      "loss": 7.0518,
+      "step": 681
+    },
+    {
+      "epoch": 0.23276450511945393,
+      "grad_norm": 3.5754668712615967,
+      "learning_rate": 0.0009224118316268487,
+      "loss": 6.5852,
+      "step": 682
+    },
+    {
+      "epoch": 0.23310580204778156,
+      "grad_norm": 3.5699808597564697,
+      "learning_rate": 0.0009222980659840728,
+      "loss": 7.1735,
+      "step": 683
+    },
+    {
+      "epoch": 0.2334470989761092,
+      "grad_norm": 3.3643417358398438,
+      "learning_rate": 0.000922184300341297,
+      "loss": 6.8809,
+      "step": 684
+    },
+    {
+      "epoch": 0.23378839590443687,
+      "grad_norm": 4.393078804016113,
+      "learning_rate": 0.0009220705346985211,
+      "loss": 6.2643,
+      "step": 685
+    },
+    {
+      "epoch": 0.2341296928327645,
+      "grad_norm": 3.7724368572235107,
+      "learning_rate": 0.0009219567690557452,
+      "loss": 6.6099,
+      "step": 686
+    },
+    {
+      "epoch": 0.23447098976109215,
+      "grad_norm": 3.7385144233703613,
+      "learning_rate": 0.0009218430034129693,
+      "loss": 6.8247,
+      "step": 687
+    },
+    {
+      "epoch": 0.2348122866894198,
+      "grad_norm": 3.5062074661254883,
+      "learning_rate": 0.0009217292377701934,
+      "loss": 7.1594,
+      "step": 688
+    },
+    {
+      "epoch": 0.23515358361774744,
+      "grad_norm": 4.198519229888916,
+      "learning_rate": 0.0009216154721274175,
+      "loss": 6.0545,
+      "step": 689
+    },
+    {
+      "epoch": 0.2354948805460751,
+      "grad_norm": 3.814720392227173,
+      "learning_rate": 0.0009215017064846418,
+      "loss": 7.1292,
+      "step": 690
+    },
+    {
+      "epoch": 0.23583617747440272,
+      "grad_norm": 3.642256498336792,
+      "learning_rate": 0.0009213879408418658,
+      "loss": 7.0849,
+      "step": 691
+    },
+    {
+      "epoch": 0.23617747440273038,
+      "grad_norm": 4.872190475463867,
+      "learning_rate": 0.0009212741751990899,
+      "loss": 6.8024,
+      "step": 692
+    },
+    {
+      "epoch": 0.23651877133105803,
+      "grad_norm": 3.5935611724853516,
+      "learning_rate": 0.000921160409556314,
+      "loss": 6.6592,
+      "step": 693
+    },
+    {
+      "epoch": 0.23686006825938566,
+      "grad_norm": 4.270242214202881,
+      "learning_rate": 0.0009210466439135381,
+      "loss": 6.636,
+      "step": 694
+    },
+    {
+      "epoch": 0.23720136518771331,
+      "grad_norm": 3.73964524269104,
+      "learning_rate": 0.0009209328782707622,
+      "loss": 7.3186,
+      "step": 695
+    },
+    {
+      "epoch": 0.23754266211604094,
+      "grad_norm": 3.55539870262146,
+      "learning_rate": 0.0009208191126279864,
+      "loss": 7.008,
+      "step": 696
+    },
+    {
+      "epoch": 0.2378839590443686,
+      "grad_norm": 3.5897581577301025,
+      "learning_rate": 0.0009207053469852105,
+      "loss": 6.8969,
+      "step": 697
+    },
+    {
+      "epoch": 0.23822525597269625,
+      "grad_norm": 3.5224783420562744,
+      "learning_rate": 0.0009205915813424346,
+      "loss": 6.6032,
+      "step": 698
+    },
+    {
+      "epoch": 0.23856655290102388,
+      "grad_norm": 3.5760772228240967,
+      "learning_rate": 0.0009204778156996587,
+      "loss": 6.8774,
+      "step": 699
+    },
+    {
+      "epoch": 0.23890784982935154,
+      "grad_norm": 3.4625370502471924,
+      "learning_rate": 0.0009203640500568828,
+      "loss": 6.934,
+      "step": 700
+    },
+    {
+      "epoch": 0.2392491467576792,
+      "grad_norm": 3.7003684043884277,
+      "learning_rate": 0.0009202502844141069,
+      "loss": 6.7796,
+      "step": 701
+    },
+    {
+      "epoch": 0.23959044368600682,
+      "grad_norm": 3.4892773628234863,
+      "learning_rate": 0.0009201365187713311,
+      "loss": 6.6771,
+      "step": 702
+    },
+    {
+      "epoch": 0.23993174061433448,
+      "grad_norm": 3.626009941101074,
+      "learning_rate": 0.0009200227531285552,
+      "loss": 6.4784,
+      "step": 703
+    },
+    {
+      "epoch": 0.2402730375426621,
+      "grad_norm": 4.080516815185547,
+      "learning_rate": 0.0009199089874857793,
+      "loss": 6.844,
+      "step": 704
+    },
+    {
+      "epoch": 0.24061433447098976,
+      "grad_norm": 3.7570433616638184,
+      "learning_rate": 0.0009197952218430034,
+      "loss": 6.3753,
+      "step": 705
+    },
+    {
+      "epoch": 0.24095563139931742,
+      "grad_norm": 3.6478946208953857,
+      "learning_rate": 0.0009196814562002275,
+      "loss": 6.4808,
+      "step": 706
+    },
+    {
+      "epoch": 0.24129692832764504,
+      "grad_norm": 3.6579360961914062,
+      "learning_rate": 0.0009195676905574518,
+      "loss": 6.345,
+      "step": 707
+    },
+    {
+      "epoch": 0.2416382252559727,
+      "grad_norm": 3.3532590866088867,
+      "learning_rate": 0.0009194539249146759,
+      "loss": 6.8791,
+      "step": 708
+    },
+    {
+      "epoch": 0.24197952218430033,
+      "grad_norm": 3.424499273300171,
+      "learning_rate": 0.0009193401592719,
+      "loss": 6.5488,
+      "step": 709
+    },
+    {
+      "epoch": 0.24232081911262798,
+      "grad_norm": 3.450228452682495,
+      "learning_rate": 0.000919226393629124,
+      "loss": 6.8405,
+      "step": 710
+    },
+    {
+      "epoch": 0.24266211604095564,
+      "grad_norm": 3.4876630306243896,
+      "learning_rate": 0.0009191126279863481,
+      "loss": 6.9152,
+      "step": 711
+    },
+    {
+      "epoch": 0.24300341296928327,
+      "grad_norm": 3.635850429534912,
+      "learning_rate": 0.0009189988623435722,
+      "loss": 6.6264,
+      "step": 712
+    },
+    {
+      "epoch": 0.24334470989761092,
+      "grad_norm": 3.780963897705078,
+      "learning_rate": 0.0009188850967007964,
+      "loss": 6.1287,
+      "step": 713
+    },
+    {
+      "epoch": 0.24368600682593858,
+      "grad_norm": 3.4603798389434814,
+      "learning_rate": 0.0009187713310580205,
+      "loss": 6.6569,
+      "step": 714
+    },
+    {
+      "epoch": 0.2440273037542662,
+      "grad_norm": 3.917168378829956,
+      "learning_rate": 0.0009186575654152446,
+      "loss": 6.2551,
+      "step": 715
+    },
+    {
+      "epoch": 0.24436860068259386,
+      "grad_norm": 3.537386417388916,
+      "learning_rate": 0.0009185437997724687,
+      "loss": 7.0423,
+      "step": 716
+    },
+    {
+      "epoch": 0.2447098976109215,
+      "grad_norm": 3.577162027359009,
+      "learning_rate": 0.0009184300341296928,
+      "loss": 6.9338,
+      "step": 717
+    },
+    {
+      "epoch": 0.24505119453924915,
+      "grad_norm": 3.8202831745147705,
+      "learning_rate": 0.0009183162684869169,
+      "loss": 6.0808,
+      "step": 718
+    },
+    {
+      "epoch": 0.2453924914675768,
+      "grad_norm": 3.6418538093566895,
+      "learning_rate": 0.0009182025028441411,
+      "loss": 7.1966,
+      "step": 719
+    },
+    {
+      "epoch": 0.24573378839590443,
+      "grad_norm": 3.42926025390625,
+      "learning_rate": 0.0009180887372013652,
+      "loss": 7.0308,
+      "step": 720
+    },
+    {
+      "epoch": 0.24607508532423208,
+      "grad_norm": 3.68467378616333,
+      "learning_rate": 0.0009179749715585893,
+      "loss": 6.9812,
+      "step": 721
+    },
+    {
+      "epoch": 0.24641638225255974,
+      "grad_norm": 3.54465651512146,
+      "learning_rate": 0.0009178612059158134,
+      "loss": 6.8175,
+      "step": 722
+    },
+    {
+      "epoch": 0.24675767918088737,
+      "grad_norm": 3.897510051727295,
+      "learning_rate": 0.0009177474402730375,
+      "loss": 6.6669,
+      "step": 723
+    },
+    {
+      "epoch": 0.24709897610921502,
+      "grad_norm": 3.5380194187164307,
+      "learning_rate": 0.0009176336746302618,
+      "loss": 6.7135,
+      "step": 724
+    },
+    {
+      "epoch": 0.24744027303754265,
+      "grad_norm": 4.017455577850342,
+      "learning_rate": 0.0009175199089874859,
+      "loss": 6.2163,
+      "step": 725
+    },
+    {
+      "epoch": 0.2477815699658703,
+      "grad_norm": 3.646085739135742,
+      "learning_rate": 0.00091740614334471,
+      "loss": 6.6526,
+      "step": 726
+    },
+    {
+      "epoch": 0.24812286689419796,
+      "grad_norm": 3.8709123134613037,
+      "learning_rate": 0.0009172923777019341,
+      "loss": 6.7876,
+      "step": 727
+    },
+    {
+      "epoch": 0.2484641638225256,
+      "grad_norm": 3.960822582244873,
+      "learning_rate": 0.0009171786120591582,
+      "loss": 6.7348,
+      "step": 728
+    },
+    {
+      "epoch": 0.24880546075085325,
+      "grad_norm": 3.8198039531707764,
+      "learning_rate": 0.0009170648464163823,
+      "loss": 6.4908,
+      "step": 729
+    },
+    {
+      "epoch": 0.24914675767918087,
+      "grad_norm": 4.048033237457275,
+      "learning_rate": 0.0009169510807736064,
+      "loss": 7.083,
+      "step": 730
+    },
+    {
+      "epoch": 0.24948805460750853,
+      "grad_norm": 3.5977671146392822,
+      "learning_rate": 0.0009168373151308305,
+      "loss": 7.1354,
+      "step": 731
+    },
+    {
+      "epoch": 0.24982935153583619,
+      "grad_norm": 3.9307358264923096,
+      "learning_rate": 0.0009167235494880546,
+      "loss": 6.8698,
+      "step": 732
+    },
+    {
+      "epoch": 0.2501706484641638,
+      "grad_norm": 3.6373484134674072,
+      "learning_rate": 0.0009166097838452787,
+      "loss": 6.9548,
+      "step": 733
+    },
+    {
+      "epoch": 0.25051194539249144,
+      "grad_norm": 3.8144493103027344,
+      "learning_rate": 0.0009164960182025028,
+      "loss": 6.875,
+      "step": 734
+    },
+    {
+      "epoch": 0.2508532423208191,
+      "grad_norm": 3.9613707065582275,
+      "learning_rate": 0.0009163822525597269,
+      "loss": 6.3705,
+      "step": 735
+    },
+    {
+      "epoch": 0.25119453924914675,
+      "grad_norm": 4.828775882720947,
+      "learning_rate": 0.0009162684869169511,
+      "loss": 6.0426,
+      "step": 736
+    },
+    {
+      "epoch": 0.2515358361774744,
+      "grad_norm": 3.8135440349578857,
+      "learning_rate": 0.0009161547212741752,
+      "loss": 6.7698,
+      "step": 737
+    },
+    {
+      "epoch": 0.25187713310580206,
+      "grad_norm": 3.477957248687744,
+      "learning_rate": 0.0009160409556313993,
+      "loss": 6.7878,
+      "step": 738
+    },
+    {
+      "epoch": 0.2522184300341297,
+      "grad_norm": 3.6808841228485107,
+      "learning_rate": 0.0009159271899886234,
+      "loss": 6.9381,
+      "step": 739
+    },
+    {
+      "epoch": 0.2525597269624573,
+      "grad_norm": 3.500927448272705,
+      "learning_rate": 0.0009158134243458475,
+      "loss": 6.7481,
+      "step": 740
+    },
+    {
+      "epoch": 0.252901023890785,
+      "grad_norm": 3.291139602661133,
+      "learning_rate": 0.0009156996587030717,
+      "loss": 6.6917,
+      "step": 741
+    },
+    {
+      "epoch": 0.25324232081911263,
+      "grad_norm": 3.652759552001953,
+      "learning_rate": 0.0009155858930602959,
+      "loss": 7.108,
+      "step": 742
+    },
+    {
+      "epoch": 0.25358361774744026,
+      "grad_norm": 3.6527302265167236,
+      "learning_rate": 0.00091547212741752,
+      "loss": 6.6711,
+      "step": 743
+    },
+    {
+      "epoch": 0.25392491467576794,
+      "grad_norm": 3.4871606826782227,
+      "learning_rate": 0.0009153583617747441,
+      "loss": 7.0744,
+      "step": 744
+    },
+    {
+      "epoch": 0.25426621160409557,
+      "grad_norm": 5.465394973754883,
+      "learning_rate": 0.0009152445961319682,
+      "loss": 6.4962,
+      "step": 745
+    },
+    {
+      "epoch": 0.2546075085324232,
+      "grad_norm": 3.8844666481018066,
+      "learning_rate": 0.0009151308304891923,
+      "loss": 6.8589,
+      "step": 746
+    },
+    {
+      "epoch": 0.2549488054607508,
+      "grad_norm": 3.7383275032043457,
+      "learning_rate": 0.0009150170648464165,
+      "loss": 7.1996,
+      "step": 747
+    },
+    {
+      "epoch": 0.2552901023890785,
+      "grad_norm": 3.918121814727783,
+      "learning_rate": 0.0009149032992036406,
+      "loss": 6.5667,
+      "step": 748
+    },
+    {
+      "epoch": 0.25563139931740614,
+      "grad_norm": 3.829987049102783,
+      "learning_rate": 0.0009147895335608646,
+      "loss": 6.5303,
+      "step": 749
+    },
+    {
+      "epoch": 0.25597269624573377,
+      "grad_norm": 3.3914873600006104,
+      "learning_rate": 0.0009146757679180887,
+      "loss": 7.1671,
+      "step": 750
+    },
+    {
+      "epoch": 0.25631399317406145,
+      "grad_norm": 3.518707275390625,
+      "learning_rate": 0.0009145620022753128,
+      "loss": 6.7905,
+      "step": 751
+    },
+    {
+      "epoch": 0.2566552901023891,
+      "grad_norm": 3.641456127166748,
+      "learning_rate": 0.0009144482366325369,
+      "loss": 6.7043,
+      "step": 752
+    },
+    {
+      "epoch": 0.2569965870307167,
+      "grad_norm": 3.5336403846740723,
+      "learning_rate": 0.0009143344709897611,
+      "loss": 6.4704,
+      "step": 753
+    },
+    {
+      "epoch": 0.2573378839590444,
+      "grad_norm": 3.45283579826355,
+      "learning_rate": 0.0009142207053469852,
+      "loss": 6.7629,
+      "step": 754
+    },
+    {
+      "epoch": 0.257679180887372,
+      "grad_norm": 3.4973959922790527,
+      "learning_rate": 0.0009141069397042093,
+      "loss": 6.82,
+      "step": 755
+    },
+    {
+      "epoch": 0.25802047781569964,
+      "grad_norm": 3.7503435611724854,
+      "learning_rate": 0.0009139931740614334,
+      "loss": 6.2377,
+      "step": 756
+    },
+    {
+      "epoch": 0.25836177474402733,
+      "grad_norm": 3.7249867916107178,
+      "learning_rate": 0.0009138794084186575,
+      "loss": 7.3,
+      "step": 757
+    },
+    {
+      "epoch": 0.25870307167235496,
+      "grad_norm": 3.5649001598358154,
+      "learning_rate": 0.0009137656427758817,
+      "loss": 6.6075,
+      "step": 758
+    },
+    {
+      "epoch": 0.2590443686006826,
+      "grad_norm": 4.057753562927246,
+      "learning_rate": 0.0009136518771331059,
+      "loss": 6.6559,
+      "step": 759
+    },
+    {
+      "epoch": 0.2593856655290102,
+      "grad_norm": 3.949220657348633,
+      "learning_rate": 0.00091353811149033,
+      "loss": 6.8383,
+      "step": 760
+    },
+    {
+      "epoch": 0.2597269624573379,
+      "grad_norm": 3.460750102996826,
+      "learning_rate": 0.0009134243458475541,
+      "loss": 6.892,
+      "step": 761
+    },
+    {
+      "epoch": 0.2600682593856655,
+      "grad_norm": 7.605457305908203,
+      "learning_rate": 0.0009133105802047782,
+      "loss": 6.766,
+      "step": 762
+    },
+    {
+      "epoch": 0.26040955631399315,
+      "grad_norm": 3.6430115699768066,
+      "learning_rate": 0.0009131968145620023,
+      "loss": 6.9554,
+      "step": 763
+    },
+    {
+      "epoch": 0.26075085324232083,
+      "grad_norm": 3.728748083114624,
+      "learning_rate": 0.0009130830489192265,
+      "loss": 7.2722,
+      "step": 764
+    },
+    {
+      "epoch": 0.26109215017064846,
+      "grad_norm": 3.6627445220947266,
+      "learning_rate": 0.0009129692832764506,
+      "loss": 6.8527,
+      "step": 765
+    },
+    {
+      "epoch": 0.2614334470989761,
+      "grad_norm": 3.310154676437378,
+      "learning_rate": 0.0009128555176336747,
+      "loss": 7.1432,
+      "step": 766
+    },
+    {
+      "epoch": 0.2617747440273038,
+      "grad_norm": 3.2910642623901367,
+      "learning_rate": 0.0009127417519908988,
+      "loss": 7.0282,
+      "step": 767
+    },
+    {
+      "epoch": 0.2621160409556314,
+      "grad_norm": 3.301368236541748,
+      "learning_rate": 0.0009126279863481229,
+      "loss": 6.8306,
+      "step": 768
+    },
+    {
+      "epoch": 0.26245733788395903,
+      "grad_norm": 6.695905685424805,
+      "learning_rate": 0.0009125142207053469,
+      "loss": 5.8919,
+      "step": 769
+    },
+    {
+      "epoch": 0.2627986348122867,
+      "grad_norm": 3.5930447578430176,
+      "learning_rate": 0.0009124004550625711,
+      "loss": 6.3776,
+      "step": 770
+    },
+    {
+      "epoch": 0.26313993174061434,
+      "grad_norm": 3.598477363586426,
+      "learning_rate": 0.0009122866894197952,
+      "loss": 6.9116,
+      "step": 771
+    },
+    {
+      "epoch": 0.26348122866894197,
+      "grad_norm": 3.46647572517395,
+      "learning_rate": 0.0009121729237770193,
+      "loss": 6.6275,
+      "step": 772
+    },
+    {
+      "epoch": 0.2638225255972696,
+      "grad_norm": 3.6713485717773438,
+      "learning_rate": 0.0009120591581342434,
+      "loss": 6.3535,
+      "step": 773
+    },
+    {
+      "epoch": 0.2641638225255973,
+      "grad_norm": 4.576082229614258,
+      "learning_rate": 0.0009119453924914675,
+      "loss": 6.585,
+      "step": 774
+    },
+    {
+      "epoch": 0.2645051194539249,
+      "grad_norm": 3.976733922958374,
+      "learning_rate": 0.0009118316268486917,
+      "loss": 6.479,
+      "step": 775
+    },
+    {
+      "epoch": 0.26484641638225254,
+      "grad_norm": 3.6250696182250977,
+      "learning_rate": 0.0009117178612059159,
+      "loss": 6.4861,
+      "step": 776
+    },
+    {
+      "epoch": 0.2651877133105802,
+      "grad_norm": 3.517406702041626,
+      "learning_rate": 0.00091160409556314,
+      "loss": 7.2256,
+      "step": 777
+    },
+    {
+      "epoch": 0.26552901023890785,
+      "grad_norm": 3.710671901702881,
+      "learning_rate": 0.0009114903299203641,
+      "loss": 6.5771,
+      "step": 778
+    },
+    {
+      "epoch": 0.2658703071672355,
+      "grad_norm": 3.4462730884552,
+      "learning_rate": 0.0009113765642775882,
+      "loss": 6.6247,
+      "step": 779
+    },
+    {
+      "epoch": 0.26621160409556316,
+      "grad_norm": 3.5399367809295654,
+      "learning_rate": 0.0009112627986348123,
+      "loss": 6.4941,
+      "step": 780
+    },
+    {
+      "epoch": 0.2665529010238908,
+      "grad_norm": 3.5168278217315674,
+      "learning_rate": 0.0009111490329920364,
+      "loss": 6.8313,
+      "step": 781
+    },
+    {
+      "epoch": 0.2668941979522184,
+      "grad_norm": 3.445139169692993,
+      "learning_rate": 0.0009110352673492606,
+      "loss": 6.7253,
+      "step": 782
+    },
+    {
+      "epoch": 0.2672354948805461,
+      "grad_norm": 3.9971537590026855,
+      "learning_rate": 0.0009109215017064847,
+      "loss": 6.8024,
+      "step": 783
+    },
+    {
+      "epoch": 0.2675767918088737,
+      "grad_norm": 3.8834123611450195,
+      "learning_rate": 0.0009108077360637088,
+      "loss": 6.729,
+      "step": 784
+    },
+    {
+      "epoch": 0.26791808873720135,
+      "grad_norm": 4.05747652053833,
+      "learning_rate": 0.0009106939704209329,
+      "loss": 6.3846,
+      "step": 785
+    },
+    {
+      "epoch": 0.26825938566552904,
+      "grad_norm": 3.5583267211914062,
+      "learning_rate": 0.000910580204778157,
+      "loss": 7.0615,
+      "step": 786
+    },
+    {
+      "epoch": 0.26860068259385667,
+      "grad_norm": 3.9900383949279785,
+      "learning_rate": 0.0009104664391353812,
+      "loss": 6.3194,
+      "step": 787
+    },
+    {
+      "epoch": 0.2689419795221843,
+      "grad_norm": 3.564735174179077,
+      "learning_rate": 0.0009103526734926052,
+      "loss": 6.736,
+      "step": 788
+    },
+    {
+      "epoch": 0.2692832764505119,
+      "grad_norm": 3.4203550815582275,
+      "learning_rate": 0.0009102389078498293,
+      "loss": 6.8528,
+      "step": 789
+    },
+    {
+      "epoch": 0.2696245733788396,
+      "grad_norm": 3.507297992706299,
+      "learning_rate": 0.0009101251422070534,
+      "loss": 6.7446,
+      "step": 790
+    },
+    {
+      "epoch": 0.26996587030716723,
+      "grad_norm": 3.489607095718384,
+      "learning_rate": 0.0009100113765642775,
+      "loss": 7.1027,
+      "step": 791
+    },
+    {
+      "epoch": 0.27030716723549486,
+      "grad_norm": 3.3406436443328857,
+      "learning_rate": 0.0009098976109215017,
+      "loss": 6.9971,
+      "step": 792
+    },
+    {
+      "epoch": 0.27064846416382254,
+      "grad_norm": 3.441478967666626,
+      "learning_rate": 0.0009097838452787259,
+      "loss": 6.8602,
+      "step": 793
+    },
+    {
+      "epoch": 0.27098976109215017,
+      "grad_norm": 3.7188546657562256,
+      "learning_rate": 0.00090967007963595,
+      "loss": 6.4872,
+      "step": 794
+    },
+    {
+      "epoch": 0.2713310580204778,
+      "grad_norm": 3.50281023979187,
+      "learning_rate": 0.0009095563139931741,
+      "loss": 6.6612,
+      "step": 795
+    },
+    {
+      "epoch": 0.2716723549488055,
+      "grad_norm": 3.8761279582977295,
+      "learning_rate": 0.0009094425483503982,
+      "loss": 6.5065,
+      "step": 796
+    },
+    {
+      "epoch": 0.2720136518771331,
+      "grad_norm": 5.031599521636963,
+      "learning_rate": 0.0009093287827076223,
+      "loss": 5.0877,
+      "step": 797
+    },
+    {
+      "epoch": 0.27235494880546074,
+      "grad_norm": 3.8323252201080322,
+      "learning_rate": 0.0009092150170648464,
+      "loss": 6.7722,
+      "step": 798
+    },
+    {
+      "epoch": 0.2726962457337884,
+      "grad_norm": 5.560990333557129,
+      "learning_rate": 0.0009091012514220706,
+      "loss": 5.4916,
+      "step": 799
+    },
+    {
+      "epoch": 0.27303754266211605,
+      "grad_norm": 3.516458511352539,
+      "learning_rate": 0.0009089874857792947,
+      "loss": 6.7099,
+      "step": 800
+    },
+    {
+      "epoch": 0.2733788395904437,
+      "grad_norm": 3.5895726680755615,
+      "learning_rate": 0.0009088737201365188,
+      "loss": 6.3197,
+      "step": 801
+    },
+    {
+      "epoch": 0.2737201365187713,
+      "grad_norm": 3.489635467529297,
+      "learning_rate": 0.0009087599544937429,
+      "loss": 7.0153,
+      "step": 802
+    },
+    {
+      "epoch": 0.274061433447099,
+      "grad_norm": 3.5357353687286377,
+      "learning_rate": 0.000908646188850967,
+      "loss": 6.392,
+      "step": 803
+    },
+    {
+      "epoch": 0.2744027303754266,
+      "grad_norm": 3.435361623764038,
+      "learning_rate": 0.0009085324232081912,
+      "loss": 6.7289,
+      "step": 804
+    },
+    {
+      "epoch": 0.27474402730375425,
+      "grad_norm": 3.4798481464385986,
+      "learning_rate": 0.0009084186575654153,
+      "loss": 7.1502,
+      "step": 805
+    },
+    {
+      "epoch": 0.27508532423208193,
+      "grad_norm": 4.012742042541504,
+      "learning_rate": 0.0009083048919226394,
+      "loss": 5.9575,
+      "step": 806
+    },
+    {
+      "epoch": 0.27542662116040956,
+      "grad_norm": 3.503770112991333,
+      "learning_rate": 0.0009081911262798636,
+      "loss": 6.7986,
+      "step": 807
+    },
+    {
+      "epoch": 0.2757679180887372,
+      "grad_norm": 3.8158676624298096,
+      "learning_rate": 0.0009080773606370875,
+      "loss": 6.6379,
+      "step": 808
+    },
+    {
+      "epoch": 0.27610921501706487,
+      "grad_norm": 3.9384515285491943,
+      "learning_rate": 0.0009079635949943117,
+      "loss": 6.6877,
+      "step": 809
+    },
+    {
+      "epoch": 0.2764505119453925,
+      "grad_norm": 4.289785861968994,
+      "learning_rate": 0.0009078498293515359,
+      "loss": 6.4728,
+      "step": 810
+    },
+    {
+      "epoch": 0.2767918088737201,
+      "grad_norm": 3.6413400173187256,
+      "learning_rate": 0.00090773606370876,
+      "loss": 6.9374,
+      "step": 811
+    },
+    {
+      "epoch": 0.2771331058020478,
+      "grad_norm": 3.771023988723755,
+      "learning_rate": 0.0009076222980659841,
+      "loss": 7.0035,
+      "step": 812
+    },
+    {
+      "epoch": 0.27747440273037544,
+      "grad_norm": 3.7076187133789062,
+      "learning_rate": 0.0009075085324232082,
+      "loss": 6.5376,
+      "step": 813
+    },
+    {
+      "epoch": 0.27781569965870306,
+      "grad_norm": 4.858947277069092,
+      "learning_rate": 0.0009073947667804323,
+      "loss": 6.2922,
+      "step": 814
+    },
+    {
+      "epoch": 0.2781569965870307,
+      "grad_norm": 3.7647488117218018,
+      "learning_rate": 0.0009072810011376564,
+      "loss": 7.0143,
+      "step": 815
+    },
+    {
+      "epoch": 0.2784982935153584,
+      "grad_norm": 4.246391773223877,
+      "learning_rate": 0.0009071672354948806,
+      "loss": 6.8741,
+      "step": 816
+    },
+    {
+      "epoch": 0.278839590443686,
+      "grad_norm": 3.448207139968872,
+      "learning_rate": 0.0009070534698521047,
+      "loss": 7.1626,
+      "step": 817
+    },
+    {
+      "epoch": 0.27918088737201363,
+      "grad_norm": 3.4277663230895996,
+      "learning_rate": 0.0009069397042093288,
+      "loss": 7.1784,
+      "step": 818
+    },
+    {
+      "epoch": 0.2795221843003413,
+      "grad_norm": 3.3655548095703125,
+      "learning_rate": 0.0009068259385665529,
+      "loss": 6.8989,
+      "step": 819
+    },
+    {
+      "epoch": 0.27986348122866894,
+      "grad_norm": 15.125733375549316,
+      "learning_rate": 0.000906712172923777,
+      "loss": 6.2618,
+      "step": 820
+    },
+    {
+      "epoch": 0.28020477815699657,
+      "grad_norm": 3.803508996963501,
+      "learning_rate": 0.0009065984072810011,
+      "loss": 7.2742,
+      "step": 821
+    },
+    {
+      "epoch": 0.28054607508532425,
+      "grad_norm": 4.941516399383545,
+      "learning_rate": 0.0009064846416382253,
+      "loss": 6.3995,
+      "step": 822
+    },
+    {
+      "epoch": 0.2808873720136519,
+      "grad_norm": 4.072309494018555,
+      "learning_rate": 0.0009063708759954494,
+      "loss": 6.865,
+      "step": 823
+    },
+    {
+      "epoch": 0.2812286689419795,
+      "grad_norm": 3.5273547172546387,
+      "learning_rate": 0.0009062571103526736,
+      "loss": 6.8719,
+      "step": 824
+    },
+    {
+      "epoch": 0.2815699658703072,
+      "grad_norm": 3.4187939167022705,
+      "learning_rate": 0.0009061433447098977,
+      "loss": 6.6182,
+      "step": 825
+    },
+    {
+      "epoch": 0.2819112627986348,
+      "grad_norm": 5.67510461807251,
+      "learning_rate": 0.0009060295790671218,
+      "loss": 5.9724,
+      "step": 826
+    },
+    {
+      "epoch": 0.28225255972696245,
+      "grad_norm": 3.5598597526550293,
+      "learning_rate": 0.0009059158134243459,
+      "loss": 7.0266,
+      "step": 827
+    },
+    {
+      "epoch": 0.2825938566552901,
+      "grad_norm": 4.407433986663818,
+      "learning_rate": 0.00090580204778157,
+      "loss": 6.3405,
+      "step": 828
+    },
+    {
+      "epoch": 0.28293515358361776,
+      "grad_norm": 3.8481311798095703,
+      "learning_rate": 0.0009056882821387941,
+      "loss": 6.6051,
+      "step": 829
+    },
+    {
+      "epoch": 0.2832764505119454,
+      "grad_norm": 3.4790523052215576,
+      "learning_rate": 0.0009055745164960182,
+      "loss": 6.5085,
+      "step": 830
+    },
+    {
+      "epoch": 0.283617747440273,
+      "grad_norm": 3.397590398788452,
+      "learning_rate": 0.0009054607508532423,
+      "loss": 7.1648,
+      "step": 831
+    },
+    {
+      "epoch": 0.2839590443686007,
+      "grad_norm": 3.374161958694458,
+      "learning_rate": 0.0009053469852104664,
+      "loss": 6.8825,
+      "step": 832
+    },
+    {
+      "epoch": 0.2843003412969283,
+      "grad_norm": 3.5871407985687256,
+      "learning_rate": 0.0009052332195676906,
+      "loss": 6.7742,
+      "step": 833
+    },
+    {
+      "epoch": 0.28464163822525596,
+      "grad_norm": 5.533285140991211,
+      "learning_rate": 0.0009051194539249147,
+      "loss": 6.1539,
+      "step": 834
+    },
+    {
+      "epoch": 0.28498293515358364,
+      "grad_norm": 4.258869647979736,
+      "learning_rate": 0.0009050056882821388,
+      "loss": 6.2765,
+      "step": 835
+    },
+    {
+      "epoch": 0.28532423208191127,
+      "grad_norm": 3.785416603088379,
+      "learning_rate": 0.0009048919226393629,
+      "loss": 7.2967,
+      "step": 836
+    },
+    {
+      "epoch": 0.2856655290102389,
+      "grad_norm": 3.6271724700927734,
+      "learning_rate": 0.000904778156996587,
+      "loss": 7.1172,
+      "step": 837
+    },
+    {
+      "epoch": 0.2860068259385666,
+      "grad_norm": 3.934699296951294,
+      "learning_rate": 0.0009046643913538111,
+      "loss": 6.4742,
+      "step": 838
+    },
+    {
+      "epoch": 0.2863481228668942,
+      "grad_norm": 3.78275990486145,
+      "learning_rate": 0.0009045506257110353,
+      "loss": 6.3398,
+      "step": 839
+    },
+    {
+      "epoch": 0.28668941979522183,
+      "grad_norm": 4.096293926239014,
+      "learning_rate": 0.0009044368600682594,
+      "loss": 6.7516,
+      "step": 840
+    },
+    {
+      "epoch": 0.28703071672354946,
+      "grad_norm": 9.30138111114502,
+      "learning_rate": 0.0009043230944254836,
+      "loss": 7.8543,
+      "step": 841
+    },
+    {
+      "epoch": 0.28737201365187715,
+      "grad_norm": 3.7339835166931152,
+      "learning_rate": 0.0009042093287827077,
+      "loss": 7.1888,
+      "step": 842
+    },
+    {
+      "epoch": 0.2877133105802048,
+      "grad_norm": 3.5420281887054443,
+      "learning_rate": 0.0009040955631399318,
+      "loss": 6.9893,
+      "step": 843
+    },
+    {
+      "epoch": 0.2880546075085324,
+      "grad_norm": 3.5827724933624268,
+      "learning_rate": 0.0009039817974971559,
+      "loss": 6.8226,
+      "step": 844
+    },
+    {
+      "epoch": 0.2883959044368601,
+      "grad_norm": 22.785751342773438,
+      "learning_rate": 0.0009038680318543801,
+      "loss": 6.6574,
+      "step": 845
+    },
+    {
+      "epoch": 0.2887372013651877,
+      "grad_norm": 3.686525583267212,
+      "learning_rate": 0.0009037542662116041,
+      "loss": 6.6054,
+      "step": 846
+    },
+    {
+      "epoch": 0.28907849829351534,
+      "grad_norm": 3.8692002296447754,
+      "learning_rate": 0.0009036405005688282,
+      "loss": 6.785,
+      "step": 847
+    },
+    {
+      "epoch": 0.289419795221843,
+      "grad_norm": 5.953486919403076,
+      "learning_rate": 0.0009035267349260523,
+      "loss": 4.0383,
+      "step": 848
+    },
+    {
+      "epoch": 0.28976109215017065,
+      "grad_norm": 3.760619878768921,
+      "learning_rate": 0.0009034129692832764,
+      "loss": 6.9499,
+      "step": 849
+    },
+    {
+      "epoch": 0.2901023890784983,
+      "grad_norm": 4.419182300567627,
+      "learning_rate": 0.0009032992036405006,
+      "loss": 6.7118,
+      "step": 850
+    },
+    {
+      "epoch": 0.29044368600682596,
+      "grad_norm": 3.5752365589141846,
+      "learning_rate": 0.0009031854379977247,
+      "loss": 6.9143,
+      "step": 851
+    },
+    {
+      "epoch": 0.2907849829351536,
+      "grad_norm": 3.5920283794403076,
+      "learning_rate": 0.0009030716723549488,
+      "loss": 7.0298,
+      "step": 852
+    },
+    {
+      "epoch": 0.2911262798634812,
+      "grad_norm": 4.226795196533203,
+      "learning_rate": 0.0009029579067121729,
+      "loss": 5.9417,
+      "step": 853
+    },
+    {
+      "epoch": 0.29146757679180885,
+      "grad_norm": 3.6170947551727295,
+      "learning_rate": 0.000902844141069397,
+      "loss": 6.926,
+      "step": 854
+    },
+    {
+      "epoch": 0.29180887372013653,
+      "grad_norm": 3.58585786819458,
+      "learning_rate": 0.0009027303754266211,
+      "loss": 6.7011,
+      "step": 855
+    },
+    {
+      "epoch": 0.29215017064846416,
+      "grad_norm": 3.7158362865448,
+      "learning_rate": 0.0009026166097838453,
+      "loss": 6.3346,
+      "step": 856
+    },
+    {
+      "epoch": 0.2924914675767918,
+      "grad_norm": 3.4805827140808105,
+      "learning_rate": 0.0009025028441410694,
+      "loss": 7.1524,
+      "step": 857
+    },
+    {
+      "epoch": 0.29283276450511947,
+      "grad_norm": 4.309206485748291,
+      "learning_rate": 0.0009023890784982936,
+      "loss": 6.79,
+      "step": 858
+    },
+    {
+      "epoch": 0.2931740614334471,
+      "grad_norm": 4.295877456665039,
+      "learning_rate": 0.0009022753128555177,
+      "loss": 6.5865,
+      "step": 859
+    },
+    {
+      "epoch": 0.2935153583617747,
+      "grad_norm": 3.572010040283203,
+      "learning_rate": 0.0009021615472127418,
+      "loss": 6.9353,
+      "step": 860
+    },
+    {
+      "epoch": 0.2938566552901024,
+      "grad_norm": 3.4424901008605957,
+      "learning_rate": 0.0009020477815699659,
+      "loss": 6.7194,
+      "step": 861
+    },
+    {
+      "epoch": 0.29419795221843004,
+      "grad_norm": 3.855348825454712,
+      "learning_rate": 0.0009019340159271901,
+      "loss": 6.9222,
+      "step": 862
+    },
+    {
+      "epoch": 0.29453924914675766,
+      "grad_norm": 6.261025428771973,
+      "learning_rate": 0.0009018202502844142,
+      "loss": 5.8834,
+      "step": 863
+    },
+    {
+      "epoch": 0.29488054607508535,
+      "grad_norm": 4.185757160186768,
+      "learning_rate": 0.0009017064846416383,
+      "loss": 6.8247,
+      "step": 864
+    },
+    {
+      "epoch": 0.295221843003413,
+      "grad_norm": 3.7310314178466797,
+      "learning_rate": 0.0009015927189988624,
+      "loss": 7.1868,
+      "step": 865
+    },
+    {
+      "epoch": 0.2955631399317406,
+      "grad_norm": 3.530855655670166,
+      "learning_rate": 0.0009014789533560864,
+      "loss": 7.2622,
+      "step": 866
+    },
+    {
+      "epoch": 0.29590443686006823,
+      "grad_norm": 7.029987335205078,
+      "learning_rate": 0.0009013651877133105,
+      "loss": 6.3624,
+      "step": 867
+    },
+    {
+      "epoch": 0.2962457337883959,
+      "grad_norm": 3.599027633666992,
+      "learning_rate": 0.0009012514220705347,
+      "loss": 7.0364,
+      "step": 868
+    },
+    {
+      "epoch": 0.29658703071672354,
+      "grad_norm": 3.781937599182129,
+      "learning_rate": 0.0009011376564277588,
+      "loss": 6.8376,
+      "step": 869
+    },
+    {
+      "epoch": 0.29692832764505117,
+      "grad_norm": 3.625401258468628,
+      "learning_rate": 0.0009010238907849829,
+      "loss": 7.1427,
+      "step": 870
+    },
+    {
+      "epoch": 0.29726962457337885,
+      "grad_norm": 3.468726396560669,
+      "learning_rate": 0.000900910125142207,
+      "loss": 7.1632,
+      "step": 871
+    },
+    {
+      "epoch": 0.2976109215017065,
+      "grad_norm": 5.193005084991455,
+      "learning_rate": 0.0009007963594994311,
+      "loss": 6.2791,
+      "step": 872
+    },
+    {
+      "epoch": 0.2979522184300341,
+      "grad_norm": 3.5303850173950195,
+      "learning_rate": 0.0009006825938566553,
+      "loss": 6.9128,
+      "step": 873
+    },
+    {
+      "epoch": 0.2982935153583618,
+      "grad_norm": 3.5984513759613037,
+      "learning_rate": 0.0009005688282138795,
+      "loss": 6.8501,
+      "step": 874
+    },
+    {
+      "epoch": 0.2986348122866894,
+      "grad_norm": 13.128771781921387,
+      "learning_rate": 0.0009004550625711036,
+      "loss": 4.9908,
+      "step": 875
+    },
+    {
+      "epoch": 0.29897610921501705,
+      "grad_norm": 4.513108730316162,
+      "learning_rate": 0.0009003412969283277,
+      "loss": 6.2214,
+      "step": 876
+    },
+    {
+      "epoch": 0.29931740614334473,
+      "grad_norm": 3.9287192821502686,
+      "learning_rate": 0.0009002275312855518,
+      "loss": 6.5736,
+      "step": 877
+    },
+    {
+      "epoch": 0.29965870307167236,
+      "grad_norm": 3.6785993576049805,
+      "learning_rate": 0.0009001137656427759,
+      "loss": 6.5955,
+      "step": 878
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5705087184906006,
+      "learning_rate": 0.0009000000000000001,
+      "loss": 6.6474,
+      "step": 879
+    },
+    {
+      "epoch": 0.3003412969283277,
+      "grad_norm": 3.4256176948547363,
+      "learning_rate": 0.0008998862343572242,
+      "loss": 7.1297,
+      "step": 880
+    },
+    {
+      "epoch": 0.3006825938566553,
+      "grad_norm": 3.45796275138855,
+      "learning_rate": 0.0008997724687144483,
+      "loss": 7.0026,
+      "step": 881
+    },
+    {
+      "epoch": 0.30102389078498293,
+      "grad_norm": 3.569291353225708,
+      "learning_rate": 0.0008996587030716724,
+      "loss": 6.8185,
+      "step": 882
+    },
+    {
+      "epoch": 0.30136518771331056,
+      "grad_norm": 4.46466588973999,
+      "learning_rate": 0.0008995449374288965,
+      "loss": 6.4208,
+      "step": 883
+    },
+    {
+      "epoch": 0.30170648464163824,
+      "grad_norm": 3.295574903488159,
+      "learning_rate": 0.0008994311717861206,
+      "loss": 6.6947,
+      "step": 884
+    },
+    {
+      "epoch": 0.30204778156996587,
+      "grad_norm": 3.447225570678711,
+      "learning_rate": 0.0008993174061433447,
+      "loss": 7.3734,
+      "step": 885
+    },
+    {
+      "epoch": 0.3023890784982935,
+      "grad_norm": 3.348236322402954,
+      "learning_rate": 0.0008992036405005688,
+      "loss": 6.6871,
+      "step": 886
+    },
+    {
+      "epoch": 0.3027303754266212,
+      "grad_norm": 3.5283188819885254,
+      "learning_rate": 0.0008990898748577929,
+      "loss": 6.9201,
+      "step": 887
+    },
+    {
+      "epoch": 0.3030716723549488,
+      "grad_norm": 3.4299886226654053,
+      "learning_rate": 0.000898976109215017,
+      "loss": 6.8192,
+      "step": 888
+    },
+    {
+      "epoch": 0.30341296928327643,
+      "grad_norm": 3.3818650245666504,
+      "learning_rate": 0.0008988623435722411,
+      "loss": 6.4265,
+      "step": 889
+    },
+    {
+      "epoch": 0.3037542662116041,
+      "grad_norm": 3.4895620346069336,
+      "learning_rate": 0.0008987485779294653,
+      "loss": 6.9777,
+      "step": 890
+    },
+    {
+      "epoch": 0.30409556313993175,
+      "grad_norm": 3.668978691101074,
+      "learning_rate": 0.0008986348122866895,
+      "loss": 6.9788,
+      "step": 891
+    },
+    {
+      "epoch": 0.3044368600682594,
+      "grad_norm": 3.538581132888794,
+      "learning_rate": 0.0008985210466439136,
+      "loss": 6.7965,
+      "step": 892
+    },
+    {
+      "epoch": 0.30477815699658706,
+      "grad_norm": 3.531942844390869,
+      "learning_rate": 0.0008984072810011377,
+      "loss": 7.2188,
+      "step": 893
+    },
+    {
+      "epoch": 0.3051194539249147,
+      "grad_norm": 3.566673755645752,
+      "learning_rate": 0.0008982935153583618,
+      "loss": 6.6428,
+      "step": 894
+    },
+    {
+      "epoch": 0.3054607508532423,
+      "grad_norm": 3.5514180660247803,
+      "learning_rate": 0.0008981797497155859,
+      "loss": 7.0462,
+      "step": 895
+    },
+    {
+      "epoch": 0.30580204778156994,
+      "grad_norm": 9.032792091369629,
+      "learning_rate": 0.0008980659840728101,
+      "loss": 8.609,
+      "step": 896
+    },
+    {
+      "epoch": 0.3061433447098976,
+      "grad_norm": 3.692873001098633,
+      "learning_rate": 0.0008979522184300342,
+      "loss": 6.7491,
+      "step": 897
+    },
+    {
+      "epoch": 0.30648464163822525,
+      "grad_norm": 3.670551061630249,
+      "learning_rate": 0.0008978384527872583,
+      "loss": 6.3838,
+      "step": 898
+    },
+    {
+      "epoch": 0.3068259385665529,
+      "grad_norm": 5.768166542053223,
+      "learning_rate": 0.0008977246871444824,
+      "loss": 6.4646,
+      "step": 899
+    },
+    {
+      "epoch": 0.30716723549488056,
+      "grad_norm": 3.830822706222534,
+      "learning_rate": 0.0008976109215017065,
+      "loss": 6.4905,
+      "step": 900
+    },
+    {
+      "epoch": 0.3075085324232082,
+      "grad_norm": 3.606807231903076,
+      "learning_rate": 0.0008974971558589306,
+      "loss": 6.8074,
+      "step": 901
+    },
+    {
+      "epoch": 0.3078498293515358,
+      "grad_norm": 3.6105079650878906,
+      "learning_rate": 0.0008973833902161548,
+      "loss": 6.8121,
+      "step": 902
+    },
+    {
+      "epoch": 0.3081911262798635,
+      "grad_norm": 3.3105735778808594,
+      "learning_rate": 0.0008972696245733789,
+      "loss": 6.5316,
+      "step": 903
+    },
+    {
+      "epoch": 0.30853242320819113,
+      "grad_norm": 4.859966278076172,
+      "learning_rate": 0.000897155858930603,
+      "loss": 6.0759,
+      "step": 904
+    },
+    {
+      "epoch": 0.30887372013651876,
+      "grad_norm": 3.6393473148345947,
+      "learning_rate": 0.000897042093287827,
+      "loss": 6.8395,
+      "step": 905
+    },
+    {
+      "epoch": 0.30921501706484644,
+      "grad_norm": 3.572521686553955,
+      "learning_rate": 0.0008969283276450511,
+      "loss": 6.8854,
+      "step": 906
+    },
+    {
+      "epoch": 0.30955631399317407,
+      "grad_norm": 3.761270046234131,
+      "learning_rate": 0.0008968145620022752,
+      "loss": 6.2211,
+      "step": 907
+    },
+    {
+      "epoch": 0.3098976109215017,
+      "grad_norm": 4.0054192543029785,
+      "learning_rate": 0.0008967007963594995,
+      "loss": 6.4794,
+      "step": 908
+    },
+    {
+      "epoch": 0.3102389078498293,
+      "grad_norm": 3.401998519897461,
+      "learning_rate": 0.0008965870307167236,
+      "loss": 7.0889,
+      "step": 909
+    },
+    {
+      "epoch": 0.310580204778157,
+      "grad_norm": 7.148408889770508,
+      "learning_rate": 0.0008964732650739477,
+      "loss": 6.3543,
+      "step": 910
+    },
+    {
+      "epoch": 0.31092150170648464,
+      "grad_norm": 3.945793628692627,
+      "learning_rate": 0.0008963594994311718,
+      "loss": 7.0554,
+      "step": 911
+    },
+    {
+      "epoch": 0.31126279863481227,
+      "grad_norm": 4.088592052459717,
+      "learning_rate": 0.0008962457337883959,
+      "loss": 6.5995,
+      "step": 912
+    },
+    {
+      "epoch": 0.31160409556313995,
+      "grad_norm": 4.394853591918945,
+      "learning_rate": 0.0008961319681456201,
+      "loss": 7.0796,
+      "step": 913
+    },
+    {
+      "epoch": 0.3119453924914676,
+      "grad_norm": 3.678943395614624,
+      "learning_rate": 0.0008960182025028442,
+      "loss": 7.3075,
+      "step": 914
+    },
+    {
+      "epoch": 0.3122866894197952,
+      "grad_norm": 3.424969434738159,
+      "learning_rate": 0.0008959044368600683,
+      "loss": 6.9823,
+      "step": 915
+    },
+    {
+      "epoch": 0.3126279863481229,
+      "grad_norm": 3.4337804317474365,
+      "learning_rate": 0.0008957906712172924,
+      "loss": 6.8313,
+      "step": 916
+    },
+    {
+      "epoch": 0.3129692832764505,
+      "grad_norm": 3.492877960205078,
+      "learning_rate": 0.0008956769055745165,
+      "loss": 6.7776,
+      "step": 917
+    },
+    {
+      "epoch": 0.31331058020477814,
+      "grad_norm": 3.4652328491210938,
+      "learning_rate": 0.0008955631399317406,
+      "loss": 6.5177,
+      "step": 918
+    },
+    {
+      "epoch": 0.3136518771331058,
+      "grad_norm": 3.4624969959259033,
+      "learning_rate": 0.0008954493742889648,
+      "loss": 6.8393,
+      "step": 919
+    },
+    {
+      "epoch": 0.31399317406143346,
+      "grad_norm": 3.3327879905700684,
+      "learning_rate": 0.0008953356086461889,
+      "loss": 6.9991,
+      "step": 920
+    },
+    {
+      "epoch": 0.3143344709897611,
+      "grad_norm": 3.756620168685913,
+      "learning_rate": 0.000895221843003413,
+      "loss": 6.8865,
+      "step": 921
+    },
+    {
+      "epoch": 0.3146757679180887,
+      "grad_norm": 8.067540168762207,
+      "learning_rate": 0.0008951080773606371,
+      "loss": 6.0514,
+      "step": 922
+    },
+    {
+      "epoch": 0.3150170648464164,
+      "grad_norm": 3.782060384750366,
+      "learning_rate": 0.0008949943117178612,
+      "loss": 7.1974,
+      "step": 923
+    },
+    {
+      "epoch": 0.315358361774744,
+      "grad_norm": 3.9047977924346924,
+      "learning_rate": 0.0008948805460750852,
+      "loss": 6.9444,
+      "step": 924
+    },
+    {
+      "epoch": 0.31569965870307165,
+      "grad_norm": 3.4596681594848633,
+      "learning_rate": 0.0008947667804323095,
+      "loss": 6.9334,
+      "step": 925
+    },
+    {
+      "epoch": 0.31604095563139933,
+      "grad_norm": 3.472452163696289,
+      "learning_rate": 0.0008946530147895336,
+      "loss": 6.5,
+      "step": 926
+    },
+    {
+      "epoch": 0.31638225255972696,
+      "grad_norm": 3.3752787113189697,
+      "learning_rate": 0.0008945392491467577,
+      "loss": 6.3694,
+      "step": 927
+    },
+    {
+      "epoch": 0.3167235494880546,
+      "grad_norm": 15.716917037963867,
+      "learning_rate": 0.0008944254835039818,
+      "loss": 9.5123,
+      "step": 928
+    },
+    {
+      "epoch": 0.3170648464163823,
+      "grad_norm": 4.5218353271484375,
+      "learning_rate": 0.0008943117178612059,
+      "loss": 6.7683,
+      "step": 929
+    },
+    {
+      "epoch": 0.3174061433447099,
+      "grad_norm": 3.797961711883545,
+      "learning_rate": 0.0008941979522184301,
+      "loss": 6.4766,
+      "step": 930
+    },
+    {
+      "epoch": 0.31774744027303753,
+      "grad_norm": 5.070075035095215,
+      "learning_rate": 0.0008940841865756542,
+      "loss": 6.3685,
+      "step": 931
+    },
+    {
+      "epoch": 0.3180887372013652,
+      "grad_norm": 3.9309263229370117,
+      "learning_rate": 0.0008939704209328783,
+      "loss": 5.8371,
+      "step": 932
+    },
+    {
+      "epoch": 0.31843003412969284,
+      "grad_norm": 4.169372081756592,
+      "learning_rate": 0.0008938566552901024,
+      "loss": 7.4385,
+      "step": 933
+    },
+    {
+      "epoch": 0.31877133105802047,
+      "grad_norm": 3.563783884048462,
+      "learning_rate": 0.0008937428896473265,
+      "loss": 6.448,
+      "step": 934
+    },
+    {
+      "epoch": 0.3191126279863481,
+      "grad_norm": 3.728722333908081,
+      "learning_rate": 0.0008936291240045506,
+      "loss": 6.3533,
+      "step": 935
+    },
+    {
+      "epoch": 0.3194539249146758,
+      "grad_norm": 3.3905997276306152,
+      "learning_rate": 0.0008935153583617748,
+      "loss": 6.6507,
+      "step": 936
+    },
+    {
+      "epoch": 0.3197952218430034,
+      "grad_norm": 3.2747488021850586,
+      "learning_rate": 0.0008934015927189989,
+      "loss": 7.0518,
+      "step": 937
+    },
+    {
+      "epoch": 0.32013651877133104,
+      "grad_norm": 3.3199024200439453,
+      "learning_rate": 0.000893287827076223,
+      "loss": 6.4047,
+      "step": 938
+    },
+    {
+      "epoch": 0.3204778156996587,
+      "grad_norm": 6.756965637207031,
+      "learning_rate": 0.0008931740614334471,
+      "loss": 6.8004,
+      "step": 939
+    },
+    {
+      "epoch": 0.32081911262798635,
+      "grad_norm": 3.5635828971862793,
+      "learning_rate": 0.0008930602957906712,
+      "loss": 6.7656,
+      "step": 940
+    },
+    {
+      "epoch": 0.321160409556314,
+      "grad_norm": 3.9271492958068848,
+      "learning_rate": 0.0008929465301478953,
+      "loss": 6.823,
+      "step": 941
+    },
+    {
+      "epoch": 0.32150170648464166,
+      "grad_norm": 3.3149237632751465,
+      "learning_rate": 0.0008928327645051196,
+      "loss": 4.3894,
+      "step": 942
+    },
+    {
+      "epoch": 0.3218430034129693,
+      "grad_norm": 4.048464298248291,
+      "learning_rate": 0.0008927189988623437,
+      "loss": 7.0178,
+      "step": 943
+    },
+    {
+      "epoch": 0.3221843003412969,
+      "grad_norm": 4.173979759216309,
+      "learning_rate": 0.0008926052332195677,
+      "loss": 6.5905,
+      "step": 944
+    },
+    {
+      "epoch": 0.3225255972696246,
+      "grad_norm": 3.4944701194763184,
+      "learning_rate": 0.0008924914675767918,
+      "loss": 6.7574,
+      "step": 945
+    },
+    {
+      "epoch": 0.3228668941979522,
+      "grad_norm": 3.6318819522857666,
+      "learning_rate": 0.0008923777019340159,
+      "loss": 6.7549,
+      "step": 946
+    },
+    {
+      "epoch": 0.32320819112627985,
+      "grad_norm": 3.452150821685791,
+      "learning_rate": 0.00089226393629124,
+      "loss": 6.6196,
+      "step": 947
+    },
+    {
+      "epoch": 0.3235494880546075,
+      "grad_norm": 5.811069488525391,
+      "learning_rate": 0.0008921501706484642,
+      "loss": 5.2026,
+      "step": 948
+    },
+    {
+      "epoch": 0.32389078498293516,
+      "grad_norm": 3.539780378341675,
+      "learning_rate": 0.0008920364050056883,
+      "loss": 6.6506,
+      "step": 949
+    },
+    {
+      "epoch": 0.3242320819112628,
+      "grad_norm": 4.116481304168701,
+      "learning_rate": 0.0008919226393629124,
+      "loss": 6.993,
+      "step": 950
+    },
+    {
+      "epoch": 0.3245733788395904,
+      "grad_norm": 3.6477298736572266,
+      "learning_rate": 0.0008918088737201365,
+      "loss": 6.3687,
+      "step": 951
+    },
+    {
+      "epoch": 0.3249146757679181,
+      "grad_norm": 3.5045762062072754,
+      "learning_rate": 0.0008916951080773606,
+      "loss": 6.5074,
+      "step": 952
+    },
+    {
+      "epoch": 0.32525597269624573,
+      "grad_norm": 4.331373691558838,
+      "learning_rate": 0.0008915813424345848,
+      "loss": 6.8514,
+      "step": 953
+    },
+    {
+      "epoch": 0.32559726962457336,
+      "grad_norm": 3.383375644683838,
+      "learning_rate": 0.0008914675767918089,
+      "loss": 6.7091,
+      "step": 954
+    },
+    {
+      "epoch": 0.32593856655290104,
+      "grad_norm": 3.8869664669036865,
+      "learning_rate": 0.000891353811149033,
+      "loss": 6.7753,
+      "step": 955
+    },
+    {
+      "epoch": 0.32627986348122867,
+      "grad_norm": 5.349417209625244,
+      "learning_rate": 0.0008912400455062571,
+      "loss": 6.3789,
+      "step": 956
+    },
+    {
+      "epoch": 0.3266211604095563,
+      "grad_norm": 3.5549848079681396,
+      "learning_rate": 0.0008911262798634812,
+      "loss": 6.8974,
+      "step": 957
+    },
+    {
+      "epoch": 0.326962457337884,
+      "grad_norm": 3.521125555038452,
+      "learning_rate": 0.0008910125142207054,
+      "loss": 6.947,
+      "step": 958
+    },
+    {
+      "epoch": 0.3273037542662116,
+      "grad_norm": 5.296824932098389,
+      "learning_rate": 0.0008908987485779296,
+      "loss": 5.9691,
+      "step": 959
+    },
+    {
+      "epoch": 0.32764505119453924,
+      "grad_norm": 3.733386516571045,
+      "learning_rate": 0.0008907849829351537,
+      "loss": 6.672,
+      "step": 960
+    },
+    {
+      "epoch": 0.32798634812286687,
+      "grad_norm": 4.317715644836426,
+      "learning_rate": 0.0008906712172923778,
+      "loss": 6.1089,
+      "step": 961
+    },
+    {
+      "epoch": 0.32832764505119455,
+      "grad_norm": 3.5799124240875244,
+      "learning_rate": 0.0008905574516496019,
+      "loss": 6.7116,
+      "step": 962
+    },
+    {
+      "epoch": 0.3286689419795222,
+      "grad_norm": 3.54365873336792,
+      "learning_rate": 0.0008904436860068259,
+      "loss": 6.6934,
+      "step": 963
+    },
+    {
+      "epoch": 0.3290102389078498,
+      "grad_norm": 3.48354172706604,
+      "learning_rate": 0.00089032992036405,
+      "loss": 6.7771,
+      "step": 964
+    },
+    {
+      "epoch": 0.3293515358361775,
+      "grad_norm": 3.8075222969055176,
+      "learning_rate": 0.0008902161547212742,
+      "loss": 6.726,
+      "step": 965
+    },
+    {
+      "epoch": 0.3296928327645051,
+      "grad_norm": 3.549077033996582,
+      "learning_rate": 0.0008901023890784983,
+      "loss": 6.8491,
+      "step": 966
+    },
+    {
+      "epoch": 0.33003412969283275,
+      "grad_norm": 3.487042188644409,
+      "learning_rate": 0.0008899886234357224,
+      "loss": 6.9951,
+      "step": 967
+    },
+    {
+      "epoch": 0.33037542662116043,
+      "grad_norm": 3.4379420280456543,
+      "learning_rate": 0.0008898748577929465,
+      "loss": 6.7213,
+      "step": 968
+    },
+    {
+      "epoch": 0.33071672354948806,
+      "grad_norm": 3.3688578605651855,
+      "learning_rate": 0.0008897610921501706,
+      "loss": 6.7468,
+      "step": 969
+    },
+    {
+      "epoch": 0.3310580204778157,
+      "grad_norm": 4.88122034072876,
+      "learning_rate": 0.0008896473265073947,
+      "loss": 6.8637,
+      "step": 970
+    },
+    {
+      "epoch": 0.33139931740614337,
+      "grad_norm": 3.590115547180176,
+      "learning_rate": 0.0008895335608646189,
+      "loss": 6.4559,
+      "step": 971
+    },
+    {
+      "epoch": 0.331740614334471,
+      "grad_norm": 3.8333377838134766,
+      "learning_rate": 0.000889419795221843,
+      "loss": 6.0111,
+      "step": 972
+    },
+    {
+      "epoch": 0.3320819112627986,
+      "grad_norm": 3.6459643840789795,
+      "learning_rate": 0.0008893060295790671,
+      "loss": 5.3953,
+      "step": 973
+    },
+    {
+      "epoch": 0.3324232081911263,
+      "grad_norm": 3.799060583114624,
+      "learning_rate": 0.0008891922639362912,
+      "loss": 6.8401,
+      "step": 974
+    },
+    {
+      "epoch": 0.33276450511945393,
+      "grad_norm": 3.553802728652954,
+      "learning_rate": 0.0008890784982935154,
+      "loss": 6.763,
+      "step": 975
+    },
+    {
+      "epoch": 0.33310580204778156,
+      "grad_norm": 5.171194553375244,
+      "learning_rate": 0.0008889647326507396,
+      "loss": 6.4421,
+      "step": 976
+    },
+    {
+      "epoch": 0.3334470989761092,
+      "grad_norm": 4.729731559753418,
+      "learning_rate": 0.0008888509670079637,
+      "loss": 4.2545,
+      "step": 977
+    },
+    {
+      "epoch": 0.3337883959044369,
+      "grad_norm": 4.018588066101074,
+      "learning_rate": 0.0008887372013651878,
+      "loss": 6.8516,
+      "step": 978
+    },
+    {
+      "epoch": 0.3341296928327645,
+      "grad_norm": 3.8841042518615723,
+      "learning_rate": 0.0008886234357224119,
+      "loss": 7.2299,
+      "step": 979
+    },
+    {
+      "epoch": 0.33447098976109213,
+      "grad_norm": 3.4637234210968018,
+      "learning_rate": 0.000888509670079636,
+      "loss": 6.5481,
+      "step": 980
+    },
+    {
+      "epoch": 0.3348122866894198,
+      "grad_norm": 3.2626147270202637,
+      "learning_rate": 0.0008883959044368601,
+      "loss": 7.2605,
+      "step": 981
+    },
+    {
+      "epoch": 0.33515358361774744,
+      "grad_norm": 3.390256643295288,
+      "learning_rate": 0.0008882821387940843,
+      "loss": 6.8236,
+      "step": 982
+    },
+    {
+      "epoch": 0.33549488054607507,
+      "grad_norm": 3.666612386703491,
+      "learning_rate": 0.0008881683731513083,
+      "loss": 6.838,
+      "step": 983
+    },
+    {
+      "epoch": 0.33583617747440275,
+      "grad_norm": 3.5521507263183594,
+      "learning_rate": 0.0008880546075085324,
+      "loss": 6.6424,
+      "step": 984
+    },
+    {
+      "epoch": 0.3361774744027304,
+      "grad_norm": 4.2973551750183105,
+      "learning_rate": 0.0008879408418657565,
+      "loss": 6.7528,
+      "step": 985
+    },
+    {
+      "epoch": 0.336518771331058,
+      "grad_norm": 3.663874864578247,
+      "learning_rate": 0.0008878270762229806,
+      "loss": 6.7709,
+      "step": 986
+    },
+    {
+      "epoch": 0.3368600682593857,
+      "grad_norm": 3.8758318424224854,
+      "learning_rate": 0.0008877133105802047,
+      "loss": 6.5531,
+      "step": 987
+    },
+    {
+      "epoch": 0.3372013651877133,
+      "grad_norm": 3.551164388656616,
+      "learning_rate": 0.0008875995449374289,
+      "loss": 6.9101,
+      "step": 988
+    },
+    {
+      "epoch": 0.33754266211604095,
+      "grad_norm": 3.607560873031616,
+      "learning_rate": 0.000887485779294653,
+      "loss": 6.5752,
+      "step": 989
+    },
+    {
+      "epoch": 0.3378839590443686,
+      "grad_norm": 3.6682469844818115,
+      "learning_rate": 0.0008873720136518771,
+      "loss": 6.3309,
+      "step": 990
+    },
+    {
+      "epoch": 0.33822525597269626,
+      "grad_norm": 3.57940411567688,
+      "learning_rate": 0.0008872582480091012,
+      "loss": 7.2817,
+      "step": 991
+    },
+    {
+      "epoch": 0.3385665529010239,
+      "grad_norm": 3.5571844577789307,
+      "learning_rate": 0.0008871444823663254,
+      "loss": 6.6641,
+      "step": 992
+    },
+    {
+      "epoch": 0.3389078498293515,
+      "grad_norm": 3.4492499828338623,
+      "learning_rate": 0.0008870307167235496,
+      "loss": 6.9242,
+      "step": 993
+    },
+    {
+      "epoch": 0.3392491467576792,
+      "grad_norm": 4.031134128570557,
+      "learning_rate": 0.0008869169510807737,
+      "loss": 6.8263,
+      "step": 994
+    },
+    {
+      "epoch": 0.3395904436860068,
+      "grad_norm": 3.3851895332336426,
+      "learning_rate": 0.0008868031854379978,
+      "loss": 6.847,
+      "step": 995
+    },
+    {
+      "epoch": 0.33993174061433445,
+      "grad_norm": 4.721614837646484,
+      "learning_rate": 0.0008866894197952219,
+      "loss": 6.7878,
+      "step": 996
+    },
+    {
+      "epoch": 0.34027303754266214,
+      "grad_norm": 3.3920321464538574,
+      "learning_rate": 0.000886575654152446,
+      "loss": 6.6165,
+      "step": 997
+    },
+    {
+      "epoch": 0.34061433447098977,
+      "grad_norm": 3.5880556106567383,
+      "learning_rate": 0.0008864618885096701,
+      "loss": 6.3648,
+      "step": 998
+    },
+    {
+      "epoch": 0.3409556313993174,
+      "grad_norm": 3.7084357738494873,
+      "learning_rate": 0.0008863481228668943,
+      "loss": 6.5777,
+      "step": 999
+    },
+    {
+      "epoch": 0.3412969283276451,
+      "grad_norm": 3.5012943744659424,
+      "learning_rate": 0.0008862343572241184,
+      "loss": 6.7863,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3416382252559727,
+      "grad_norm": 3.5657830238342285,
+      "learning_rate": 0.0008861205915813425,
+      "loss": 7.3395,
+      "step": 1001
+    },
+    {
+      "epoch": 0.34197952218430033,
+      "grad_norm": 4.714262008666992,
+      "learning_rate": 0.0008860068259385665,
+      "loss": 6.4638,
+      "step": 1002
+    },
+    {
+      "epoch": 0.34232081911262796,
+      "grad_norm": 3.793743848800659,
+      "learning_rate": 0.0008858930602957906,
+      "loss": 6.6217,
+      "step": 1003
+    },
+    {
+      "epoch": 0.34266211604095564,
+      "grad_norm": 6.239543914794922,
+      "learning_rate": 0.0008857792946530147,
+      "loss": 6.3399,
+      "step": 1004
+    },
+    {
+      "epoch": 0.3430034129692833,
+      "grad_norm": 3.838210344314575,
+      "learning_rate": 0.0008856655290102389,
+      "loss": 7.1438,
+      "step": 1005
+    },
+    {
+      "epoch": 0.3433447098976109,
+      "grad_norm": 4.131673812866211,
+      "learning_rate": 0.000885551763367463,
+      "loss": 6.3229,
+      "step": 1006
+    },
+    {
+      "epoch": 0.3436860068259386,
+      "grad_norm": 3.4897634983062744,
+      "learning_rate": 0.0008854379977246871,
+      "loss": 7.3418,
+      "step": 1007
+    },
+    {
+      "epoch": 0.3440273037542662,
+      "grad_norm": 3.3156917095184326,
+      "learning_rate": 0.0008853242320819112,
+      "loss": 6.6131,
+      "step": 1008
+    },
+    {
+      "epoch": 0.34436860068259384,
+      "grad_norm": 5.101655006408691,
+      "learning_rate": 0.0008852104664391354,
+      "loss": 6.4129,
+      "step": 1009
+    },
+    {
+      "epoch": 0.3447098976109215,
+      "grad_norm": 3.4084413051605225,
+      "learning_rate": 0.0008850967007963595,
+      "loss": 6.8648,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34505119453924915,
+      "grad_norm": 3.558993101119995,
+      "learning_rate": 0.0008849829351535837,
+      "loss": 6.5622,
+      "step": 1011
+    },
+    {
+      "epoch": 0.3453924914675768,
+      "grad_norm": 3.5753562450408936,
+      "learning_rate": 0.0008848691695108078,
+      "loss": 6.7141,
+      "step": 1012
+    },
+    {
+      "epoch": 0.34573378839590446,
+      "grad_norm": 3.5802392959594727,
+      "learning_rate": 0.0008847554038680319,
+      "loss": 6.9107,
+      "step": 1013
+    },
+    {
+      "epoch": 0.3460750853242321,
+      "grad_norm": 3.4724674224853516,
+      "learning_rate": 0.000884641638225256,
+      "loss": 6.6967,
+      "step": 1014
+    },
+    {
+      "epoch": 0.3464163822525597,
+      "grad_norm": 3.293440103530884,
+      "learning_rate": 0.0008845278725824801,
+      "loss": 6.4138,
+      "step": 1015
+    },
+    {
+      "epoch": 0.34675767918088735,
+      "grad_norm": 3.4762954711914062,
+      "learning_rate": 0.0008844141069397043,
+      "loss": 6.8232,
+      "step": 1016
+    },
+    {
+      "epoch": 0.34709897610921503,
+      "grad_norm": 3.6370632648468018,
+      "learning_rate": 0.0008843003412969284,
+      "loss": 6.9722,
+      "step": 1017
+    },
+    {
+      "epoch": 0.34744027303754266,
+      "grad_norm": 3.511749029159546,
+      "learning_rate": 0.0008841865756541525,
+      "loss": 6.8501,
+      "step": 1018
+    },
+    {
+      "epoch": 0.3477815699658703,
+      "grad_norm": 5.163912296295166,
+      "learning_rate": 0.0008840728100113766,
+      "loss": 6.1093,
+      "step": 1019
+    },
+    {
+      "epoch": 0.34812286689419797,
+      "grad_norm": 8.192817687988281,
+      "learning_rate": 0.0008839590443686007,
+      "loss": 6.5018,
+      "step": 1020
+    },
+    {
+      "epoch": 0.3484641638225256,
+      "grad_norm": 3.8490147590637207,
+      "learning_rate": 0.0008838452787258247,
+      "loss": 6.8653,
+      "step": 1021
+    },
+    {
+      "epoch": 0.3488054607508532,
+      "grad_norm": 3.7160000801086426,
+      "learning_rate": 0.0008837315130830489,
+      "loss": 6.9512,
+      "step": 1022
+    },
+    {
+      "epoch": 0.3491467576791809,
+      "grad_norm": 3.8101956844329834,
+      "learning_rate": 0.000883617747440273,
+      "loss": 6.2376,
+      "step": 1023
+    },
+    {
+      "epoch": 0.34948805460750854,
+      "grad_norm": 3.3378794193267822,
+      "learning_rate": 0.0008835039817974971,
+      "loss": 6.8794,
+      "step": 1024
+    },
+    {
+      "epoch": 0.34982935153583616,
+      "grad_norm": 3.5934832096099854,
+      "learning_rate": 0.0008833902161547212,
+      "loss": 6.1176,
+      "step": 1025
+    },
+    {
+      "epoch": 0.35017064846416385,
+      "grad_norm": 3.461996078491211,
+      "learning_rate": 0.0008832764505119454,
+      "loss": 6.5096,
+      "step": 1026
+    },
+    {
+      "epoch": 0.3505119453924915,
+      "grad_norm": 3.272123098373413,
+      "learning_rate": 0.0008831626848691695,
+      "loss": 6.7895,
+      "step": 1027
+    },
+    {
+      "epoch": 0.3508532423208191,
+      "grad_norm": 3.3164021968841553,
+      "learning_rate": 0.0008830489192263937,
+      "loss": 6.7558,
+      "step": 1028
+    },
+    {
+      "epoch": 0.35119453924914673,
+      "grad_norm": 3.2223832607269287,
+      "learning_rate": 0.0008829351535836178,
+      "loss": 6.5788,
+      "step": 1029
+    },
+    {
+      "epoch": 0.3515358361774744,
+      "grad_norm": 5.8226237297058105,
+      "learning_rate": 0.0008828213879408419,
+      "loss": 3.4241,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35187713310580204,
+      "grad_norm": 3.6470441818237305,
+      "learning_rate": 0.000882707622298066,
+      "loss": 6.7667,
+      "step": 1031
+    },
+    {
+      "epoch": 0.35221843003412967,
+      "grad_norm": 3.8023180961608887,
+      "learning_rate": 0.0008825938566552901,
+      "loss": 6.8807,
+      "step": 1032
+    },
+    {
+      "epoch": 0.35255972696245735,
+      "grad_norm": 3.8079946041107178,
+      "learning_rate": 0.0008824800910125143,
+      "loss": 6.7188,
+      "step": 1033
+    },
+    {
+      "epoch": 0.352901023890785,
+      "grad_norm": 3.5138330459594727,
+      "learning_rate": 0.0008823663253697384,
+      "loss": 7.1953,
+      "step": 1034
+    },
+    {
+      "epoch": 0.3532423208191126,
+      "grad_norm": 3.8133950233459473,
+      "learning_rate": 0.0008822525597269625,
+      "loss": 7.1188,
+      "step": 1035
+    },
+    {
+      "epoch": 0.3535836177474403,
+      "grad_norm": 3.640202522277832,
+      "learning_rate": 0.0008821387940841866,
+      "loss": 6.739,
+      "step": 1036
+    },
+    {
+      "epoch": 0.3539249146757679,
+      "grad_norm": 3.518843173980713,
+      "learning_rate": 0.0008820250284414107,
+      "loss": 6.8397,
+      "step": 1037
+    },
+    {
+      "epoch": 0.35426621160409555,
+      "grad_norm": 3.5482702255249023,
+      "learning_rate": 0.0008819112627986348,
+      "loss": 6.618,
+      "step": 1038
+    },
+    {
+      "epoch": 0.35460750853242323,
+      "grad_norm": 3.419344902038574,
+      "learning_rate": 0.000881797497155859,
+      "loss": 6.9977,
+      "step": 1039
+    },
+    {
+      "epoch": 0.35494880546075086,
+      "grad_norm": 3.4673240184783936,
+      "learning_rate": 0.0008816837315130831,
+      "loss": 6.4868,
+      "step": 1040
+    },
+    {
+      "epoch": 0.3552901023890785,
+      "grad_norm": 8.60751724243164,
+      "learning_rate": 0.0008815699658703071,
+      "loss": 4.6019,
+      "step": 1041
+    },
+    {
+      "epoch": 0.3556313993174061,
+      "grad_norm": 4.347542762756348,
+      "learning_rate": 0.0008814562002275312,
+      "loss": 6.7114,
+      "step": 1042
+    },
+    {
+      "epoch": 0.3559726962457338,
+      "grad_norm": 5.797741413116455,
+      "learning_rate": 0.0008813424345847554,
+      "loss": 4.9746,
+      "step": 1043
+    },
+    {
+      "epoch": 0.3563139931740614,
+      "grad_norm": 4.145489692687988,
+      "learning_rate": 0.0008812286689419795,
+      "loss": 6.8995,
+      "step": 1044
+    },
+    {
+      "epoch": 0.35665529010238906,
+      "grad_norm": 4.579039096832275,
+      "learning_rate": 0.0008811149032992037,
+      "loss": 7.1214,
+      "step": 1045
+    },
+    {
+      "epoch": 0.35699658703071674,
+      "grad_norm": 3.6013543605804443,
+      "learning_rate": 0.0008810011376564278,
+      "loss": 6.9495,
+      "step": 1046
+    },
+    {
+      "epoch": 0.35733788395904437,
+      "grad_norm": 3.570963144302368,
+      "learning_rate": 0.0008808873720136519,
+      "loss": 7.0384,
+      "step": 1047
+    },
+    {
+      "epoch": 0.357679180887372,
+      "grad_norm": 3.448312759399414,
+      "learning_rate": 0.000880773606370876,
+      "loss": 6.8964,
+      "step": 1048
+    },
+    {
+      "epoch": 0.3580204778156997,
+      "grad_norm": 3.4679174423217773,
+      "learning_rate": 0.0008806598407281001,
+      "loss": 6.5713,
+      "step": 1049
+    },
+    {
+      "epoch": 0.3583617747440273,
+      "grad_norm": 3.8704769611358643,
+      "learning_rate": 0.0008805460750853242,
+      "loss": 6.424,
+      "step": 1050
+    },
+    {
+      "epoch": 0.35870307167235493,
+      "grad_norm": 3.3437047004699707,
+      "learning_rate": 0.0008804323094425484,
+      "loss": 7.1021,
+      "step": 1051
+    },
+    {
+      "epoch": 0.3590443686006826,
+      "grad_norm": 7.765682220458984,
+      "learning_rate": 0.0008803185437997725,
+      "loss": 5.6959,
+      "step": 1052
+    },
+    {
+      "epoch": 0.35938566552901025,
+      "grad_norm": 4.041772842407227,
+      "learning_rate": 0.0008802047781569966,
+      "loss": 6.5353,
+      "step": 1053
+    },
+    {
+      "epoch": 0.3597269624573379,
+      "grad_norm": 4.038043022155762,
+      "learning_rate": 0.0008800910125142207,
+      "loss": 6.8547,
+      "step": 1054
+    },
+    {
+      "epoch": 0.36006825938566556,
+      "grad_norm": 3.7071590423583984,
+      "learning_rate": 0.0008799772468714448,
+      "loss": 6.4563,
+      "step": 1055
+    },
+    {
+      "epoch": 0.3604095563139932,
+      "grad_norm": 4.166858196258545,
+      "learning_rate": 0.000879863481228669,
+      "loss": 6.431,
+      "step": 1056
+    },
+    {
+      "epoch": 0.3607508532423208,
+      "grad_norm": 3.6435773372650146,
+      "learning_rate": 0.0008797497155858931,
+      "loss": 6.6037,
+      "step": 1057
+    },
+    {
+      "epoch": 0.36109215017064844,
+      "grad_norm": 14.491841316223145,
+      "learning_rate": 0.0008796359499431173,
+      "loss": 7.0323,
+      "step": 1058
+    },
+    {
+      "epoch": 0.3614334470989761,
+      "grad_norm": 3.81062388420105,
+      "learning_rate": 0.0008795221843003414,
+      "loss": 7.0748,
+      "step": 1059
+    },
+    {
+      "epoch": 0.36177474402730375,
+      "grad_norm": 3.9405431747436523,
+      "learning_rate": 0.0008794084186575654,
+      "loss": 6.9883,
+      "step": 1060
+    },
+    {
+      "epoch": 0.3621160409556314,
+      "grad_norm": 3.7249462604522705,
+      "learning_rate": 0.0008792946530147895,
+      "loss": 6.7057,
+      "step": 1061
+    },
+    {
+      "epoch": 0.36245733788395906,
+      "grad_norm": 4.359280109405518,
+      "learning_rate": 0.0008791808873720137,
+      "loss": 6.2134,
+      "step": 1062
+    },
+    {
+      "epoch": 0.3627986348122867,
+      "grad_norm": 3.8895323276519775,
+      "learning_rate": 0.0008790671217292378,
+      "loss": 6.3199,
+      "step": 1063
+    },
+    {
+      "epoch": 0.3631399317406143,
+      "grad_norm": 3.3901989459991455,
+      "learning_rate": 0.0008789533560864619,
+      "loss": 6.822,
+      "step": 1064
+    },
+    {
+      "epoch": 0.363481228668942,
+      "grad_norm": 3.3694682121276855,
+      "learning_rate": 0.000878839590443686,
+      "loss": 6.8209,
+      "step": 1065
+    },
+    {
+      "epoch": 0.36382252559726963,
+      "grad_norm": 3.7551913261413574,
+      "learning_rate": 0.0008787258248009101,
+      "loss": 6.7343,
+      "step": 1066
+    },
+    {
+      "epoch": 0.36416382252559726,
+      "grad_norm": 3.3077569007873535,
+      "learning_rate": 0.0008786120591581342,
+      "loss": 6.944,
+      "step": 1067
+    },
+    {
+      "epoch": 0.36450511945392494,
+      "grad_norm": 3.292142629623413,
+      "learning_rate": 0.0008784982935153584,
+      "loss": 7.1756,
+      "step": 1068
+    },
+    {
+      "epoch": 0.36484641638225257,
+      "grad_norm": 4.133096218109131,
+      "learning_rate": 0.0008783845278725825,
+      "loss": 6.7308,
+      "step": 1069
+    },
+    {
+      "epoch": 0.3651877133105802,
+      "grad_norm": 3.361743927001953,
+      "learning_rate": 0.0008782707622298066,
+      "loss": 7.252,
+      "step": 1070
+    },
+    {
+      "epoch": 0.3655290102389078,
+      "grad_norm": 3.582728624343872,
+      "learning_rate": 0.0008781569965870307,
+      "loss": 6.8102,
+      "step": 1071
+    },
+    {
+      "epoch": 0.3658703071672355,
+      "grad_norm": 4.287242412567139,
+      "learning_rate": 0.0008780432309442548,
+      "loss": 6.4504,
+      "step": 1072
+    },
+    {
+      "epoch": 0.36621160409556314,
+      "grad_norm": 3.59372878074646,
+      "learning_rate": 0.000877929465301479,
+      "loss": 7.1913,
+      "step": 1073
+    },
+    {
+      "epoch": 0.36655290102389076,
+      "grad_norm": 3.4123799800872803,
+      "learning_rate": 0.0008778156996587031,
+      "loss": 6.5238,
+      "step": 1074
+    },
+    {
+      "epoch": 0.36689419795221845,
+      "grad_norm": 3.3152260780334473,
+      "learning_rate": 0.0008777019340159273,
+      "loss": 6.9748,
+      "step": 1075
+    },
+    {
+      "epoch": 0.3672354948805461,
+      "grad_norm": 3.3980987071990967,
+      "learning_rate": 0.0008775881683731514,
+      "loss": 6.8588,
+      "step": 1076
+    },
+    {
+      "epoch": 0.3675767918088737,
+      "grad_norm": 3.447199583053589,
+      "learning_rate": 0.0008774744027303755,
+      "loss": 6.3821,
+      "step": 1077
+    },
+    {
+      "epoch": 0.3679180887372014,
+      "grad_norm": 3.4618401527404785,
+      "learning_rate": 0.0008773606370875996,
+      "loss": 6.95,
+      "step": 1078
+    },
+    {
+      "epoch": 0.368259385665529,
+      "grad_norm": 3.597135543823242,
+      "learning_rate": 0.0008772468714448238,
+      "loss": 6.4812,
+      "step": 1079
+    },
+    {
+      "epoch": 0.36860068259385664,
+      "grad_norm": 3.396935224533081,
+      "learning_rate": 0.0008771331058020478,
+      "loss": 6.7553,
+      "step": 1080
+    },
+    {
+      "epoch": 0.3689419795221843,
+      "grad_norm": 3.4151859283447266,
+      "learning_rate": 0.0008770193401592719,
+      "loss": 6.4082,
+      "step": 1081
+    },
+    {
+      "epoch": 0.36928327645051195,
+      "grad_norm": 3.4455440044403076,
+      "learning_rate": 0.000876905574516496,
+      "loss": 6.7333,
+      "step": 1082
+    },
+    {
+      "epoch": 0.3696245733788396,
+      "grad_norm": 3.539586067199707,
+      "learning_rate": 0.0008767918088737201,
+      "loss": 6.5843,
+      "step": 1083
+    },
+    {
+      "epoch": 0.3699658703071672,
+      "grad_norm": 6.984851360321045,
+      "learning_rate": 0.0008766780432309442,
+      "loss": 5.8875,
+      "step": 1084
+    },
+    {
+      "epoch": 0.3703071672354949,
+      "grad_norm": 3.949312448501587,
+      "learning_rate": 0.0008765642775881684,
+      "loss": 6.2822,
+      "step": 1085
+    },
+    {
+      "epoch": 0.3706484641638225,
+      "grad_norm": 3.787013530731201,
+      "learning_rate": 0.0008764505119453925,
+      "loss": 6.7024,
+      "step": 1086
+    },
+    {
+      "epoch": 0.37098976109215015,
+      "grad_norm": 3.797456741333008,
+      "learning_rate": 0.0008763367463026166,
+      "loss": 6.6641,
+      "step": 1087
+    },
+    {
+      "epoch": 0.37133105802047783,
+      "grad_norm": 3.9768006801605225,
+      "learning_rate": 0.0008762229806598407,
+      "loss": 6.8357,
+      "step": 1088
+    },
+    {
+      "epoch": 0.37167235494880546,
+      "grad_norm": 3.4279286861419678,
+      "learning_rate": 0.0008761092150170648,
+      "loss": 6.5971,
+      "step": 1089
+    },
+    {
+      "epoch": 0.3720136518771331,
+      "grad_norm": 3.417083501815796,
+      "learning_rate": 0.0008759954493742889,
+      "loss": 6.818,
+      "step": 1090
+    },
+    {
+      "epoch": 0.3723549488054608,
+      "grad_norm": 3.553903818130493,
+      "learning_rate": 0.0008758816837315132,
+      "loss": 6.8396,
+      "step": 1091
+    },
+    {
+      "epoch": 0.3726962457337884,
+      "grad_norm": 3.620215892791748,
+      "learning_rate": 0.0008757679180887373,
+      "loss": 6.2553,
+      "step": 1092
+    },
+    {
+      "epoch": 0.37303754266211603,
+      "grad_norm": 5.6876935958862305,
+      "learning_rate": 0.0008756541524459614,
+      "loss": 6.035,
+      "step": 1093
+    },
+    {
+      "epoch": 0.3733788395904437,
+      "grad_norm": 3.6916663646698,
+      "learning_rate": 0.0008755403868031855,
+      "loss": 6.9809,
+      "step": 1094
+    },
+    {
+      "epoch": 0.37372013651877134,
+      "grad_norm": 3.829458236694336,
+      "learning_rate": 0.0008754266211604096,
+      "loss": 6.3949,
+      "step": 1095
+    },
+    {
+      "epoch": 0.37406143344709897,
+      "grad_norm": 3.719882011413574,
+      "learning_rate": 0.0008753128555176338,
+      "loss": 7.0602,
+      "step": 1096
+    },
+    {
+      "epoch": 0.3744027303754266,
+      "grad_norm": 4.063767433166504,
+      "learning_rate": 0.0008751990898748579,
+      "loss": 6.713,
+      "step": 1097
+    },
+    {
+      "epoch": 0.3747440273037543,
+      "grad_norm": 3.6009387969970703,
+      "learning_rate": 0.000875085324232082,
+      "loss": 6.6959,
+      "step": 1098
+    },
+    {
+      "epoch": 0.3750853242320819,
+      "grad_norm": 3.6646664142608643,
+      "learning_rate": 0.000874971558589306,
+      "loss": 6.8591,
+      "step": 1099
+    },
+    {
+      "epoch": 0.37542662116040953,
+      "grad_norm": 3.4815189838409424,
+      "learning_rate": 0.0008748577929465301,
+      "loss": 6.6657,
+      "step": 1100
+    },
+    {
+      "epoch": 0.3757679180887372,
+      "grad_norm": 3.6855881214141846,
+      "learning_rate": 0.0008747440273037542,
+      "loss": 6.7854,
+      "step": 1101
+    },
+    {
+      "epoch": 0.37610921501706485,
+      "grad_norm": 3.5180046558380127,
+      "learning_rate": 0.0008746302616609784,
+      "loss": 6.8922,
+      "step": 1102
+    },
+    {
+      "epoch": 0.3764505119453925,
+      "grad_norm": 3.3709070682525635,
+      "learning_rate": 0.0008745164960182025,
+      "loss": 7.0109,
+      "step": 1103
+    },
+    {
+      "epoch": 0.37679180887372016,
+      "grad_norm": 3.4868783950805664,
+      "learning_rate": 0.0008744027303754266,
+      "loss": 7.1033,
+      "step": 1104
+    },
+    {
+      "epoch": 0.3771331058020478,
+      "grad_norm": 3.5890045166015625,
+      "learning_rate": 0.0008742889647326507,
+      "loss": 6.5034,
+      "step": 1105
+    },
+    {
+      "epoch": 0.3774744027303754,
+      "grad_norm": 5.1739726066589355,
+      "learning_rate": 0.0008741751990898748,
+      "loss": 6.2063,
+      "step": 1106
+    },
+    {
+      "epoch": 0.3778156996587031,
+      "grad_norm": 3.525132179260254,
+      "learning_rate": 0.0008740614334470989,
+      "loss": 7.3972,
+      "step": 1107
+    },
+    {
+      "epoch": 0.3781569965870307,
+      "grad_norm": 3.622143268585205,
+      "learning_rate": 0.0008739476678043232,
+      "loss": 7.0089,
+      "step": 1108
+    },
+    {
+      "epoch": 0.37849829351535835,
+      "grad_norm": 3.494596481323242,
+      "learning_rate": 0.0008738339021615473,
+      "loss": 6.5724,
+      "step": 1109
+    },
+    {
+      "epoch": 0.378839590443686,
+      "grad_norm": 3.3533830642700195,
+      "learning_rate": 0.0008737201365187714,
+      "loss": 7.1964,
+      "step": 1110
+    },
+    {
+      "epoch": 0.37918088737201366,
+      "grad_norm": 3.518562078475952,
+      "learning_rate": 0.0008736063708759955,
+      "loss": 7.2307,
+      "step": 1111
+    },
+    {
+      "epoch": 0.3795221843003413,
+      "grad_norm": 3.4765775203704834,
+      "learning_rate": 0.0008734926052332196,
+      "loss": 6.6891,
+      "step": 1112
+    },
+    {
+      "epoch": 0.3798634812286689,
+      "grad_norm": 3.379734754562378,
+      "learning_rate": 0.0008733788395904437,
+      "loss": 6.883,
+      "step": 1113
+    },
+    {
+      "epoch": 0.3802047781569966,
+      "grad_norm": 3.9166646003723145,
+      "learning_rate": 0.0008732650739476679,
+      "loss": 6.7689,
+      "step": 1114
+    },
+    {
+      "epoch": 0.38054607508532423,
+      "grad_norm": 3.344590663909912,
+      "learning_rate": 0.000873151308304892,
+      "loss": 6.579,
+      "step": 1115
+    },
+    {
+      "epoch": 0.38088737201365186,
+      "grad_norm": 3.7421436309814453,
+      "learning_rate": 0.0008730375426621161,
+      "loss": 7.2903,
+      "step": 1116
+    },
+    {
+      "epoch": 0.38122866894197954,
+      "grad_norm": 3.9341204166412354,
+      "learning_rate": 0.0008729237770193402,
+      "loss": 6.4053,
+      "step": 1117
+    },
+    {
+      "epoch": 0.38156996587030717,
+      "grad_norm": 3.6717798709869385,
+      "learning_rate": 0.0008728100113765643,
+      "loss": 6.56,
+      "step": 1118
+    },
+    {
+      "epoch": 0.3819112627986348,
+      "grad_norm": 3.7355589866638184,
+      "learning_rate": 0.0008726962457337884,
+      "loss": 6.1227,
+      "step": 1119
+    },
+    {
+      "epoch": 0.3822525597269625,
+      "grad_norm": 3.6662402153015137,
+      "learning_rate": 0.0008725824800910125,
+      "loss": 6.8174,
+      "step": 1120
+    },
+    {
+      "epoch": 0.3825938566552901,
+      "grad_norm": 3.7742650508880615,
+      "learning_rate": 0.0008724687144482366,
+      "loss": 6.9281,
+      "step": 1121
+    },
+    {
+      "epoch": 0.38293515358361774,
+      "grad_norm": 3.4731431007385254,
+      "learning_rate": 0.0008723549488054607,
+      "loss": 6.5173,
+      "step": 1122
+    },
+    {
+      "epoch": 0.38327645051194537,
+      "grad_norm": 3.55572247505188,
+      "learning_rate": 0.0008722411831626848,
+      "loss": 6.8068,
+      "step": 1123
+    },
+    {
+      "epoch": 0.38361774744027305,
+      "grad_norm": 3.307126522064209,
+      "learning_rate": 0.0008721274175199089,
+      "loss": 6.2692,
+      "step": 1124
+    },
+    {
+      "epoch": 0.3839590443686007,
+      "grad_norm": 3.444089889526367,
+      "learning_rate": 0.0008720136518771332,
+      "loss": 6.4222,
+      "step": 1125
+    },
+    {
+      "epoch": 0.3843003412969283,
+      "grad_norm": 3.731123685836792,
+      "learning_rate": 0.0008718998862343573,
+      "loss": 6.7032,
+      "step": 1126
+    },
+    {
+      "epoch": 0.384641638225256,
+      "grad_norm": 3.5673561096191406,
+      "learning_rate": 0.0008717861205915814,
+      "loss": 6.7617,
+      "step": 1127
+    },
+    {
+      "epoch": 0.3849829351535836,
+      "grad_norm": 4.335822105407715,
+      "learning_rate": 0.0008716723549488055,
+      "loss": 6.0413,
+      "step": 1128
+    },
+    {
+      "epoch": 0.38532423208191124,
+      "grad_norm": 3.483842372894287,
+      "learning_rate": 0.0008715585893060296,
+      "loss": 6.902,
+      "step": 1129
+    },
+    {
+      "epoch": 0.3856655290102389,
+      "grad_norm": 3.3141043186187744,
+      "learning_rate": 0.0008714448236632537,
+      "loss": 6.5926,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38600682593856656,
+      "grad_norm": 3.5295164585113525,
+      "learning_rate": 0.0008713310580204779,
+      "loss": 7.411,
+      "step": 1131
+    },
+    {
+      "epoch": 0.3863481228668942,
+      "grad_norm": 3.6749589443206787,
+      "learning_rate": 0.000871217292377702,
+      "loss": 6.4381,
+      "step": 1132
+    },
+    {
+      "epoch": 0.38668941979522187,
+      "grad_norm": 3.7729690074920654,
+      "learning_rate": 0.0008711035267349261,
+      "loss": 6.2913,
+      "step": 1133
+    },
+    {
+      "epoch": 0.3870307167235495,
+      "grad_norm": 3.4314284324645996,
+      "learning_rate": 0.0008709897610921502,
+      "loss": 7.1253,
+      "step": 1134
+    },
+    {
+      "epoch": 0.3873720136518771,
+      "grad_norm": 3.633718252182007,
+      "learning_rate": 0.0008708759954493743,
+      "loss": 6.378,
+      "step": 1135
+    },
+    {
+      "epoch": 0.38771331058020475,
+      "grad_norm": 3.4309566020965576,
+      "learning_rate": 0.0008707622298065985,
+      "loss": 6.7567,
+      "step": 1136
+    },
+    {
+      "epoch": 0.38805460750853243,
+      "grad_norm": 5.337145805358887,
+      "learning_rate": 0.0008706484641638226,
+      "loss": 6.1113,
+      "step": 1137
+    },
+    {
+      "epoch": 0.38839590443686006,
+      "grad_norm": 3.6700313091278076,
+      "learning_rate": 0.0008705346985210466,
+      "loss": 7.0188,
+      "step": 1138
+    },
+    {
+      "epoch": 0.3887372013651877,
+      "grad_norm": 3.756471633911133,
+      "learning_rate": 0.0008704209328782707,
+      "loss": 6.3016,
+      "step": 1139
+    },
+    {
+      "epoch": 0.3890784982935154,
+      "grad_norm": 3.3960928916931152,
+      "learning_rate": 0.0008703071672354948,
+      "loss": 6.6751,
+      "step": 1140
+    },
+    {
+      "epoch": 0.389419795221843,
+      "grad_norm": 3.354475736618042,
+      "learning_rate": 0.0008701934015927189,
+      "loss": 6.9203,
+      "step": 1141
+    },
+    {
+      "epoch": 0.38976109215017063,
+      "grad_norm": 4.095553398132324,
+      "learning_rate": 0.0008700796359499432,
+      "loss": 6.2154,
+      "step": 1142
+    },
+    {
+      "epoch": 0.3901023890784983,
+      "grad_norm": 3.592214584350586,
+      "learning_rate": 0.0008699658703071673,
+      "loss": 6.5693,
+      "step": 1143
+    },
+    {
+      "epoch": 0.39044368600682594,
+      "grad_norm": 3.5223283767700195,
+      "learning_rate": 0.0008698521046643914,
+      "loss": 7.3168,
+      "step": 1144
+    },
+    {
+      "epoch": 0.39078498293515357,
+      "grad_norm": 3.3714346885681152,
+      "learning_rate": 0.0008697383390216155,
+      "loss": 6.5233,
+      "step": 1145
+    },
+    {
+      "epoch": 0.39112627986348125,
+      "grad_norm": 4.475141525268555,
+      "learning_rate": 0.0008696245733788396,
+      "loss": 6.3844,
+      "step": 1146
+    },
+    {
+      "epoch": 0.3914675767918089,
+      "grad_norm": 3.5419039726257324,
+      "learning_rate": 0.0008695108077360637,
+      "loss": 6.718,
+      "step": 1147
+    },
+    {
+      "epoch": 0.3918088737201365,
+      "grad_norm": 3.6940994262695312,
+      "learning_rate": 0.0008693970420932879,
+      "loss": 6.9183,
+      "step": 1148
+    },
+    {
+      "epoch": 0.3921501706484642,
+      "grad_norm": 3.406237840652466,
+      "learning_rate": 0.000869283276450512,
+      "loss": 6.817,
+      "step": 1149
+    },
+    {
+      "epoch": 0.3924914675767918,
+      "grad_norm": 3.5237555503845215,
+      "learning_rate": 0.0008691695108077361,
+      "loss": 6.565,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39283276450511945,
+      "grad_norm": 3.400852918624878,
+      "learning_rate": 0.0008690557451649602,
+      "loss": 6.5612,
+      "step": 1151
+    },
+    {
+      "epoch": 0.3931740614334471,
+      "grad_norm": 3.451266050338745,
+      "learning_rate": 0.0008689419795221843,
+      "loss": 6.5092,
+      "step": 1152
+    },
+    {
+      "epoch": 0.39351535836177476,
+      "grad_norm": 4.814857006072998,
+      "learning_rate": 0.0008688282138794084,
+      "loss": 6.1531,
+      "step": 1153
+    },
+    {
+      "epoch": 0.3938566552901024,
+      "grad_norm": 3.619635820388794,
+      "learning_rate": 0.0008687144482366326,
+      "loss": 6.4214,
+      "step": 1154
+    },
+    {
+      "epoch": 0.39419795221843,
+      "grad_norm": 2.678661346435547,
+      "learning_rate": 0.0008686006825938567,
+      "loss": 3.9171,
+      "step": 1155
+    },
+    {
+      "epoch": 0.3945392491467577,
+      "grad_norm": 4.169159889221191,
+      "learning_rate": 0.0008684869169510808,
+      "loss": 6.2438,
+      "step": 1156
+    },
+    {
+      "epoch": 0.3948805460750853,
+      "grad_norm": 3.618849515914917,
+      "learning_rate": 0.0008683731513083049,
+      "loss": 6.7297,
+      "step": 1157
+    },
+    {
+      "epoch": 0.39522184300341295,
+      "grad_norm": 5.10459566116333,
+      "learning_rate": 0.0008682593856655289,
+      "loss": 6.2313,
+      "step": 1158
+    },
+    {
+      "epoch": 0.39556313993174064,
+      "grad_norm": 3.637228012084961,
+      "learning_rate": 0.0008681456200227532,
+      "loss": 6.6383,
+      "step": 1159
+    },
+    {
+      "epoch": 0.39590443686006827,
+      "grad_norm": 3.539729356765747,
+      "learning_rate": 0.0008680318543799773,
+      "loss": 6.6497,
+      "step": 1160
+    },
+    {
+      "epoch": 0.3962457337883959,
+      "grad_norm": 3.3988003730773926,
+      "learning_rate": 0.0008679180887372014,
+      "loss": 6.2502,
+      "step": 1161
+    },
+    {
+      "epoch": 0.3965870307167236,
+      "grad_norm": 3.7295820713043213,
+      "learning_rate": 0.0008678043230944255,
+      "loss": 6.354,
+      "step": 1162
+    },
+    {
+      "epoch": 0.3969283276450512,
+      "grad_norm": 3.539332151412964,
+      "learning_rate": 0.0008676905574516496,
+      "loss": 7.1846,
+      "step": 1163
+    },
+    {
+      "epoch": 0.39726962457337883,
+      "grad_norm": 3.5238754749298096,
+      "learning_rate": 0.0008675767918088737,
+      "loss": 6.2985,
+      "step": 1164
+    },
+    {
+      "epoch": 0.39761092150170646,
+      "grad_norm": 7.77903938293457,
+      "learning_rate": 0.0008674630261660979,
+      "loss": 4.739,
+      "step": 1165
+    },
+    {
+      "epoch": 0.39795221843003414,
+      "grad_norm": 3.561558485031128,
+      "learning_rate": 0.000867349260523322,
+      "loss": 6.7075,
+      "step": 1166
+    },
+    {
+      "epoch": 0.39829351535836177,
+      "grad_norm": 3.6842966079711914,
+      "learning_rate": 0.0008672354948805461,
+      "loss": 6.934,
+      "step": 1167
+    },
+    {
+      "epoch": 0.3986348122866894,
+      "grad_norm": 3.6015913486480713,
+      "learning_rate": 0.0008671217292377702,
+      "loss": 7.0431,
+      "step": 1168
+    },
+    {
+      "epoch": 0.3989761092150171,
+      "grad_norm": 3.3261101245880127,
+      "learning_rate": 0.0008670079635949943,
+      "loss": 6.5901,
+      "step": 1169
+    },
+    {
+      "epoch": 0.3993174061433447,
+      "grad_norm": 3.6874048709869385,
+      "learning_rate": 0.0008668941979522184,
+      "loss": 6.8816,
+      "step": 1170
+    },
+    {
+      "epoch": 0.39965870307167234,
+      "grad_norm": 3.367588996887207,
+      "learning_rate": 0.0008667804323094426,
+      "loss": 6.8432,
+      "step": 1171
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2957537174224854,
+      "learning_rate": 0.0008666666666666667,
+      "loss": 6.7904,
+      "step": 1172
+    },
+    {
+      "epoch": 0.40034129692832765,
+      "grad_norm": 3.3574087619781494,
+      "learning_rate": 0.0008665529010238908,
+      "loss": 6.7804,
+      "step": 1173
+    },
+    {
+      "epoch": 0.4006825938566553,
+      "grad_norm": 3.550178050994873,
+      "learning_rate": 0.0008664391353811149,
+      "loss": 6.5741,
+      "step": 1174
+    },
+    {
+      "epoch": 0.40102389078498296,
+      "grad_norm": 3.4762377738952637,
+      "learning_rate": 0.000866325369738339,
+      "loss": 7.1227,
+      "step": 1175
+    },
+    {
+      "epoch": 0.4013651877133106,
+      "grad_norm": 3.523291826248169,
+      "learning_rate": 0.0008662116040955633,
+      "loss": 6.8638,
+      "step": 1176
+    },
+    {
+      "epoch": 0.4017064846416382,
+      "grad_norm": 3.267287492752075,
+      "learning_rate": 0.0008660978384527873,
+      "loss": 6.6402,
+      "step": 1177
+    },
+    {
+      "epoch": 0.40204778156996585,
+      "grad_norm": 3.369250774383545,
+      "learning_rate": 0.0008659840728100114,
+      "loss": 6.683,
+      "step": 1178
+    },
+    {
+      "epoch": 0.40238907849829353,
+      "grad_norm": 3.4328765869140625,
+      "learning_rate": 0.0008658703071672355,
+      "loss": 6.6393,
+      "step": 1179
+    },
+    {
+      "epoch": 0.40273037542662116,
+      "grad_norm": 3.6571414470672607,
+      "learning_rate": 0.0008657565415244596,
+      "loss": 6.5947,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4030716723549488,
+      "grad_norm": 3.311383008956909,
+      "learning_rate": 0.0008656427758816837,
+      "loss": 6.7028,
+      "step": 1181
+    },
+    {
+      "epoch": 0.40341296928327647,
+      "grad_norm": 3.5350563526153564,
+      "learning_rate": 0.0008655290102389079,
+      "loss": 6.2399,
+      "step": 1182
+    },
+    {
+      "epoch": 0.4037542662116041,
+      "grad_norm": 3.58445143699646,
+      "learning_rate": 0.000865415244596132,
+      "loss": 6.4364,
+      "step": 1183
+    },
+    {
+      "epoch": 0.4040955631399317,
+      "grad_norm": 3.439987897872925,
+      "learning_rate": 0.0008653014789533561,
+      "loss": 6.5951,
+      "step": 1184
+    },
+    {
+      "epoch": 0.4044368600682594,
+      "grad_norm": 8.74139404296875,
+      "learning_rate": 0.0008651877133105802,
+      "loss": 6.3451,
+      "step": 1185
+    },
+    {
+      "epoch": 0.40477815699658704,
+      "grad_norm": 3.636200428009033,
+      "learning_rate": 0.0008650739476678043,
+      "loss": 7.1951,
+      "step": 1186
+    },
+    {
+      "epoch": 0.40511945392491466,
+      "grad_norm": 3.539865493774414,
+      "learning_rate": 0.0008649601820250284,
+      "loss": 7.0252,
+      "step": 1187
+    },
+    {
+      "epoch": 0.40546075085324235,
+      "grad_norm": 3.5502421855926514,
+      "learning_rate": 0.0008648464163822526,
+      "loss": 6.9277,
+      "step": 1188
+    },
+    {
+      "epoch": 0.40580204778157,
+      "grad_norm": 6.065408706665039,
+      "learning_rate": 0.0008647326507394767,
+      "loss": 6.2756,
+      "step": 1189
+    },
+    {
+      "epoch": 0.4061433447098976,
+      "grad_norm": 3.5880606174468994,
+      "learning_rate": 0.0008646188850967008,
+      "loss": 6.609,
+      "step": 1190
+    },
+    {
+      "epoch": 0.40648464163822523,
+      "grad_norm": 3.5269694328308105,
+      "learning_rate": 0.000864505119453925,
+      "loss": 6.6192,
+      "step": 1191
+    },
+    {
+      "epoch": 0.4068259385665529,
+      "grad_norm": 3.216411828994751,
+      "learning_rate": 0.000864391353811149,
+      "loss": 6.6329,
+      "step": 1192
+    },
+    {
+      "epoch": 0.40716723549488054,
+      "grad_norm": 3.540984630584717,
+      "learning_rate": 0.0008642775881683732,
+      "loss": 6.8023,
+      "step": 1193
+    },
+    {
+      "epoch": 0.40750853242320817,
+      "grad_norm": 3.430344581604004,
+      "learning_rate": 0.0008641638225255974,
+      "loss": 6.4494,
+      "step": 1194
+    },
+    {
+      "epoch": 0.40784982935153585,
+      "grad_norm": 3.64399790763855,
+      "learning_rate": 0.0008640500568828215,
+      "loss": 6.0932,
+      "step": 1195
+    },
+    {
+      "epoch": 0.4081911262798635,
+      "grad_norm": 3.293858528137207,
+      "learning_rate": 0.0008639362912400455,
+      "loss": 6.4266,
+      "step": 1196
+    },
+    {
+      "epoch": 0.4085324232081911,
+      "grad_norm": 3.498997449874878,
+      "learning_rate": 0.0008638225255972696,
+      "loss": 6.4859,
+      "step": 1197
+    },
+    {
+      "epoch": 0.4088737201365188,
+      "grad_norm": 3.547935724258423,
+      "learning_rate": 0.0008637087599544937,
+      "loss": 7.0903,
+      "step": 1198
+    },
+    {
+      "epoch": 0.4092150170648464,
+      "grad_norm": 3.3775041103363037,
+      "learning_rate": 0.0008635949943117179,
+      "loss": 6.6445,
+      "step": 1199
+    },
+    {
+      "epoch": 0.40955631399317405,
+      "grad_norm": 3.678861379623413,
+      "learning_rate": 0.000863481228668942,
+      "loss": 6.3223,
+      "step": 1200
+    },
+    {
+      "epoch": 0.40989761092150173,
+      "grad_norm": 5.478918075561523,
+      "learning_rate": 0.0008633674630261661,
+      "loss": 6.726,
+      "step": 1201
+    },
+    {
+      "epoch": 0.41023890784982936,
+      "grad_norm": 5.6262407302856445,
+      "learning_rate": 0.0008632536973833902,
+      "loss": 6.1447,
+      "step": 1202
+    },
+    {
+      "epoch": 0.410580204778157,
+      "grad_norm": 3.8595380783081055,
+      "learning_rate": 0.0008631399317406143,
+      "loss": 6.7759,
+      "step": 1203
+    },
+    {
+      "epoch": 0.4109215017064846,
+      "grad_norm": 3.590745449066162,
+      "learning_rate": 0.0008630261660978384,
+      "loss": 6.6065,
+      "step": 1204
+    },
+    {
+      "epoch": 0.4112627986348123,
+      "grad_norm": 3.531480550765991,
+      "learning_rate": 0.0008629124004550626,
+      "loss": 6.3746,
+      "step": 1205
+    },
+    {
+      "epoch": 0.4116040955631399,
+      "grad_norm": 3.606606960296631,
+      "learning_rate": 0.0008627986348122867,
+      "loss": 6.3677,
+      "step": 1206
+    },
+    {
+      "epoch": 0.41194539249146755,
+      "grad_norm": 3.5986878871917725,
+      "learning_rate": 0.0008626848691695108,
+      "loss": 6.2718,
+      "step": 1207
+    },
+    {
+      "epoch": 0.41228668941979524,
+      "grad_norm": 3.399188995361328,
+      "learning_rate": 0.000862571103526735,
+      "loss": 6.8167,
+      "step": 1208
+    },
+    {
+      "epoch": 0.41262798634812287,
+      "grad_norm": 3.2899904251098633,
+      "learning_rate": 0.000862457337883959,
+      "loss": 7.0081,
+      "step": 1209
+    },
+    {
+      "epoch": 0.4129692832764505,
+      "grad_norm": 3.522502899169922,
+      "learning_rate": 0.0008623435722411832,
+      "loss": 6.6561,
+      "step": 1210
+    },
+    {
+      "epoch": 0.4133105802047782,
+      "grad_norm": 3.5413362979888916,
+      "learning_rate": 0.0008622298065984074,
+      "loss": 6.8759,
+      "step": 1211
+    },
+    {
+      "epoch": 0.4136518771331058,
+      "grad_norm": 3.4126598834991455,
+      "learning_rate": 0.0008621160409556315,
+      "loss": 7.0551,
+      "step": 1212
+    },
+    {
+      "epoch": 0.41399317406143343,
+      "grad_norm": 3.658048391342163,
+      "learning_rate": 0.0008620022753128556,
+      "loss": 6.7476,
+      "step": 1213
+    },
+    {
+      "epoch": 0.4143344709897611,
+      "grad_norm": 3.4803097248077393,
+      "learning_rate": 0.0008618885096700797,
+      "loss": 6.717,
+      "step": 1214
+    },
+    {
+      "epoch": 0.41467576791808874,
+      "grad_norm": 3.52303147315979,
+      "learning_rate": 0.0008617747440273038,
+      "loss": 6.6623,
+      "step": 1215
+    },
+    {
+      "epoch": 0.4150170648464164,
+      "grad_norm": 3.362966537475586,
+      "learning_rate": 0.0008616609783845278,
+      "loss": 6.8483,
+      "step": 1216
+    },
+    {
+      "epoch": 0.415358361774744,
+      "grad_norm": 4.339871883392334,
+      "learning_rate": 0.000861547212741752,
+      "loss": 6.3404,
+      "step": 1217
+    },
+    {
+      "epoch": 0.4156996587030717,
+      "grad_norm": 3.4630043506622314,
+      "learning_rate": 0.0008614334470989761,
+      "loss": 7.0581,
+      "step": 1218
+    },
+    {
+      "epoch": 0.4160409556313993,
+      "grad_norm": 4.27215576171875,
+      "learning_rate": 0.0008613196814562002,
+      "loss": 6.5148,
+      "step": 1219
+    },
+    {
+      "epoch": 0.41638225255972694,
+      "grad_norm": 3.6531379222869873,
+      "learning_rate": 0.0008612059158134243,
+      "loss": 6.1321,
+      "step": 1220
+    },
+    {
+      "epoch": 0.4167235494880546,
+      "grad_norm": 3.4010026454925537,
+      "learning_rate": 0.0008610921501706484,
+      "loss": 6.827,
+      "step": 1221
+    },
+    {
+      "epoch": 0.41706484641638225,
+      "grad_norm": 3.467449903488159,
+      "learning_rate": 0.0008609783845278726,
+      "loss": 7.0066,
+      "step": 1222
+    },
+    {
+      "epoch": 0.4174061433447099,
+      "grad_norm": 3.3873372077941895,
+      "learning_rate": 0.0008608646188850967,
+      "loss": 6.1582,
+      "step": 1223
+    },
+    {
+      "epoch": 0.41774744027303756,
+      "grad_norm": 3.483137607574463,
+      "learning_rate": 0.0008607508532423208,
+      "loss": 6.3962,
+      "step": 1224
+    },
+    {
+      "epoch": 0.4180887372013652,
+      "grad_norm": 3.4427895545959473,
+      "learning_rate": 0.000860637087599545,
+      "loss": 6.5905,
+      "step": 1225
+    },
+    {
+      "epoch": 0.4184300341296928,
+      "grad_norm": 3.545475721359253,
+      "learning_rate": 0.000860523321956769,
+      "loss": 7.2328,
+      "step": 1226
+    },
+    {
+      "epoch": 0.4187713310580205,
+      "grad_norm": 3.3596596717834473,
+      "learning_rate": 0.0008604095563139932,
+      "loss": 6.6329,
+      "step": 1227
+    },
+    {
+      "epoch": 0.41911262798634813,
+      "grad_norm": 3.347614049911499,
+      "learning_rate": 0.0008602957906712174,
+      "loss": 6.6236,
+      "step": 1228
+    },
+    {
+      "epoch": 0.41945392491467576,
+      "grad_norm": 3.284127950668335,
+      "learning_rate": 0.0008601820250284415,
+      "loss": 6.7784,
+      "step": 1229
+    },
+    {
+      "epoch": 0.4197952218430034,
+      "grad_norm": 3.3968138694763184,
+      "learning_rate": 0.0008600682593856656,
+      "loss": 6.8942,
+      "step": 1230
+    },
+    {
+      "epoch": 0.42013651877133107,
+      "grad_norm": 3.375905752182007,
+      "learning_rate": 0.0008599544937428897,
+      "loss": 6.6425,
+      "step": 1231
+    },
+    {
+      "epoch": 0.4204778156996587,
+      "grad_norm": 3.409838914871216,
+      "learning_rate": 0.0008598407281001138,
+      "loss": 6.6804,
+      "step": 1232
+    },
+    {
+      "epoch": 0.4208191126279863,
+      "grad_norm": 4.324724197387695,
+      "learning_rate": 0.0008597269624573379,
+      "loss": 6.1592,
+      "step": 1233
+    },
+    {
+      "epoch": 0.421160409556314,
+      "grad_norm": 5.707208156585693,
+      "learning_rate": 0.0008596131968145621,
+      "loss": 5.9943,
+      "step": 1234
+    },
+    {
+      "epoch": 0.42150170648464164,
+      "grad_norm": 3.6807117462158203,
+      "learning_rate": 0.0008594994311717861,
+      "loss": 6.7764,
+      "step": 1235
+    },
+    {
+      "epoch": 0.42184300341296926,
+      "grad_norm": 3.467085361480713,
+      "learning_rate": 0.0008593856655290102,
+      "loss": 6.5814,
+      "step": 1236
+    },
+    {
+      "epoch": 0.42218430034129695,
+      "grad_norm": 3.4565463066101074,
+      "learning_rate": 0.0008592718998862343,
+      "loss": 6.9017,
+      "step": 1237
+    },
+    {
+      "epoch": 0.4225255972696246,
+      "grad_norm": 4.344884395599365,
+      "learning_rate": 0.0008591581342434584,
+      "loss": 6.5987,
+      "step": 1238
+    },
+    {
+      "epoch": 0.4228668941979522,
+      "grad_norm": 4.411989212036133,
+      "learning_rate": 0.0008590443686006826,
+      "loss": 6.741,
+      "step": 1239
+    },
+    {
+      "epoch": 0.4232081911262799,
+      "grad_norm": 3.775130271911621,
+      "learning_rate": 0.0008589306029579067,
+      "loss": 6.3065,
+      "step": 1240
+    },
+    {
+      "epoch": 0.4235494880546075,
+      "grad_norm": 3.5393505096435547,
+      "learning_rate": 0.0008588168373151308,
+      "loss": 6.9163,
+      "step": 1241
+    },
+    {
+      "epoch": 0.42389078498293514,
+      "grad_norm": 3.7094571590423584,
+      "learning_rate": 0.000858703071672355,
+      "loss": 5.9449,
+      "step": 1242
+    },
+    {
+      "epoch": 0.4242320819112628,
+      "grad_norm": 3.3689160346984863,
+      "learning_rate": 0.000858589306029579,
+      "loss": 6.5366,
+      "step": 1243
+    },
+    {
+      "epoch": 0.42457337883959045,
+      "grad_norm": 4.066009521484375,
+      "learning_rate": 0.0008584755403868032,
+      "loss": 6.2055,
+      "step": 1244
+    },
+    {
+      "epoch": 0.4249146757679181,
+      "grad_norm": 3.611774444580078,
+      "learning_rate": 0.0008583617747440274,
+      "loss": 6.6451,
+      "step": 1245
+    },
+    {
+      "epoch": 0.4252559726962457,
+      "grad_norm": 3.587705373764038,
+      "learning_rate": 0.0008582480091012515,
+      "loss": 6.6259,
+      "step": 1246
+    },
+    {
+      "epoch": 0.4255972696245734,
+      "grad_norm": 3.442603349685669,
+      "learning_rate": 0.0008581342434584756,
+      "loss": 6.8989,
+      "step": 1247
+    },
+    {
+      "epoch": 0.425938566552901,
+      "grad_norm": 4.580451488494873,
+      "learning_rate": 0.0008580204778156997,
+      "loss": 6.2377,
+      "step": 1248
+    },
+    {
+      "epoch": 0.42627986348122865,
+      "grad_norm": 3.4543490409851074,
+      "learning_rate": 0.0008579067121729238,
+      "loss": 6.2214,
+      "step": 1249
+    },
+    {
+      "epoch": 0.42662116040955633,
+      "grad_norm": 5.877060890197754,
+      "learning_rate": 0.0008577929465301479,
+      "loss": 5.4147,
+      "step": 1250
+    },
+    {
+      "epoch": 0.42696245733788396,
+      "grad_norm": 4.635700225830078,
+      "learning_rate": 0.0008576791808873721,
+      "loss": 6.1787,
+      "step": 1251
+    },
+    {
+      "epoch": 0.4273037542662116,
+      "grad_norm": 3.6666946411132812,
+      "learning_rate": 0.0008575654152445962,
+      "loss": 7.006,
+      "step": 1252
+    },
+    {
+      "epoch": 0.42764505119453927,
+      "grad_norm": 3.5804553031921387,
+      "learning_rate": 0.0008574516496018203,
+      "loss": 6.8813,
+      "step": 1253
+    },
+    {
+      "epoch": 0.4279863481228669,
+      "grad_norm": 3.4612207412719727,
+      "learning_rate": 0.0008573378839590444,
+      "loss": 6.6291,
+      "step": 1254
+    },
+    {
+      "epoch": 0.4283276450511945,
+      "grad_norm": 3.6497762203216553,
+      "learning_rate": 0.0008572241183162684,
+      "loss": 6.9736,
+      "step": 1255
+    },
+    {
+      "epoch": 0.4286689419795222,
+      "grad_norm": 3.296250104904175,
+      "learning_rate": 0.0008571103526734925,
+      "loss": 6.5129,
+      "step": 1256
+    },
+    {
+      "epoch": 0.42901023890784984,
+      "grad_norm": 3.4279532432556152,
+      "learning_rate": 0.0008569965870307167,
+      "loss": 6.4369,
+      "step": 1257
+    },
+    {
+      "epoch": 0.42935153583617747,
+      "grad_norm": 3.448242664337158,
+      "learning_rate": 0.0008568828213879408,
+      "loss": 6.6171,
+      "step": 1258
+    },
+    {
+      "epoch": 0.4296928327645051,
+      "grad_norm": 3.353353977203369,
+      "learning_rate": 0.000856769055745165,
+      "loss": 6.5279,
+      "step": 1259
+    },
+    {
+      "epoch": 0.4300341296928328,
+      "grad_norm": 3.488964319229126,
+      "learning_rate": 0.0008566552901023891,
+      "loss": 6.8257,
+      "step": 1260
+    },
+    {
+      "epoch": 0.4303754266211604,
+      "grad_norm": 3.5172197818756104,
+      "learning_rate": 0.0008565415244596132,
+      "loss": 6.1531,
+      "step": 1261
+    },
+    {
+      "epoch": 0.43071672354948803,
+      "grad_norm": 3.3878061771392822,
+      "learning_rate": 0.0008564277588168374,
+      "loss": 6.8535,
+      "step": 1262
+    },
+    {
+      "epoch": 0.4310580204778157,
+      "grad_norm": 3.500743865966797,
+      "learning_rate": 0.0008563139931740615,
+      "loss": 6.9892,
+      "step": 1263
+    },
+    {
+      "epoch": 0.43139931740614335,
+      "grad_norm": 7.8660736083984375,
+      "learning_rate": 0.0008562002275312856,
+      "loss": 5.9798,
+      "step": 1264
+    },
+    {
+      "epoch": 0.431740614334471,
+      "grad_norm": 3.737638473510742,
+      "learning_rate": 0.0008560864618885097,
+      "loss": 6.8881,
+      "step": 1265
+    },
+    {
+      "epoch": 0.43208191126279866,
+      "grad_norm": 3.8500828742980957,
+      "learning_rate": 0.0008559726962457338,
+      "loss": 6.8304,
+      "step": 1266
+    },
+    {
+      "epoch": 0.4324232081911263,
+      "grad_norm": 3.5614750385284424,
+      "learning_rate": 0.0008558589306029579,
+      "loss": 6.4568,
+      "step": 1267
+    },
+    {
+      "epoch": 0.4327645051194539,
+      "grad_norm": 3.2859628200531006,
+      "learning_rate": 0.0008557451649601821,
+      "loss": 6.977,
+      "step": 1268
+    },
+    {
+      "epoch": 0.4331058020477816,
+      "grad_norm": 3.4138572216033936,
+      "learning_rate": 0.0008556313993174062,
+      "loss": 5.7698,
+      "step": 1269
+    },
+    {
+      "epoch": 0.4334470989761092,
+      "grad_norm": 3.3377716541290283,
+      "learning_rate": 0.0008555176336746303,
+      "loss": 6.5715,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43378839590443685,
+      "grad_norm": 3.5361649990081787,
+      "learning_rate": 0.0008554038680318544,
+      "loss": 6.546,
+      "step": 1271
+    },
+    {
+      "epoch": 0.4341296928327645,
+      "grad_norm": 3.4464402198791504,
+      "learning_rate": 0.0008552901023890785,
+      "loss": 6.2318,
+      "step": 1272
+    },
+    {
+      "epoch": 0.43447098976109216,
+      "grad_norm": 3.289402484893799,
+      "learning_rate": 0.0008551763367463026,
+      "loss": 6.7913,
+      "step": 1273
+    },
+    {
+      "epoch": 0.4348122866894198,
+      "grad_norm": 3.665785312652588,
+      "learning_rate": 0.0008550625711035267,
+      "loss": 6.6228,
+      "step": 1274
+    },
+    {
+      "epoch": 0.4351535836177474,
+      "grad_norm": 5.519403457641602,
+      "learning_rate": 0.0008549488054607508,
+      "loss": 6.0677,
+      "step": 1275
+    },
+    {
+      "epoch": 0.4354948805460751,
+      "grad_norm": 3.639775276184082,
+      "learning_rate": 0.000854835039817975,
+      "loss": 6.8561,
+      "step": 1276
+    },
+    {
+      "epoch": 0.43583617747440273,
+      "grad_norm": 3.529599905014038,
+      "learning_rate": 0.0008547212741751991,
+      "loss": 6.6314,
+      "step": 1277
+    },
+    {
+      "epoch": 0.43617747440273036,
+      "grad_norm": 3.8909149169921875,
+      "learning_rate": 0.0008546075085324232,
+      "loss": 6.7449,
+      "step": 1278
+    },
+    {
+      "epoch": 0.43651877133105804,
+      "grad_norm": 4.182558059692383,
+      "learning_rate": 0.0008544937428896473,
+      "loss": 6.1402,
+      "step": 1279
+    },
+    {
+      "epoch": 0.43686006825938567,
+      "grad_norm": 3.6773123741149902,
+      "learning_rate": 0.0008543799772468715,
+      "loss": 6.5567,
+      "step": 1280
+    },
+    {
+      "epoch": 0.4372013651877133,
+      "grad_norm": 3.5389983654022217,
+      "learning_rate": 0.0008542662116040956,
+      "loss": 6.847,
+      "step": 1281
+    },
+    {
+      "epoch": 0.437542662116041,
+      "grad_norm": 3.5662002563476562,
+      "learning_rate": 0.0008541524459613197,
+      "loss": 6.242,
+      "step": 1282
+    },
+    {
+      "epoch": 0.4378839590443686,
+      "grad_norm": 3.623596668243408,
+      "learning_rate": 0.0008540386803185438,
+      "loss": 6.4278,
+      "step": 1283
+    },
+    {
+      "epoch": 0.43822525597269624,
+      "grad_norm": 3.6582117080688477,
+      "learning_rate": 0.0008539249146757679,
+      "loss": 6.266,
+      "step": 1284
+    },
+    {
+      "epoch": 0.43856655290102387,
+      "grad_norm": 5.415266990661621,
+      "learning_rate": 0.0008538111490329921,
+      "loss": 4.8563,
+      "step": 1285
+    },
+    {
+      "epoch": 0.43890784982935155,
+      "grad_norm": 3.88291335105896,
+      "learning_rate": 0.0008536973833902162,
+      "loss": 7.0351,
+      "step": 1286
+    },
+    {
+      "epoch": 0.4392491467576792,
+      "grad_norm": 3.7798752784729004,
+      "learning_rate": 0.0008535836177474403,
+      "loss": 6.9712,
+      "step": 1287
+    },
+    {
+      "epoch": 0.4395904436860068,
+      "grad_norm": 4.600677490234375,
+      "learning_rate": 0.0008534698521046644,
+      "loss": 6.4982,
+      "step": 1288
+    },
+    {
+      "epoch": 0.4399317406143345,
+      "grad_norm": 3.6498169898986816,
+      "learning_rate": 0.0008533560864618885,
+      "loss": 6.7925,
+      "step": 1289
+    },
+    {
+      "epoch": 0.4402730375426621,
+      "grad_norm": 3.712674617767334,
+      "learning_rate": 0.0008532423208191126,
+      "loss": 6.5274,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44061433447098974,
+      "grad_norm": 3.2611215114593506,
+      "learning_rate": 0.0008531285551763368,
+      "loss": 6.954,
+      "step": 1291
+    },
+    {
+      "epoch": 0.4409556313993174,
+      "grad_norm": 3.1427042484283447,
+      "learning_rate": 0.000853014789533561,
+      "loss": 6.6816,
+      "step": 1292
+    },
+    {
+      "epoch": 0.44129692832764505,
+      "grad_norm": 3.4534733295440674,
+      "learning_rate": 0.0008529010238907851,
+      "loss": 6.4714,
+      "step": 1293
+    },
+    {
+      "epoch": 0.4416382252559727,
+      "grad_norm": 3.25132155418396,
+      "learning_rate": 0.0008527872582480091,
+      "loss": 6.3441,
+      "step": 1294
+    },
+    {
+      "epoch": 0.44197952218430037,
+      "grad_norm": 3.3785088062286377,
+      "learning_rate": 0.0008526734926052332,
+      "loss": 6.8156,
+      "step": 1295
+    },
+    {
+      "epoch": 0.442320819112628,
+      "grad_norm": 3.79790997505188,
+      "learning_rate": 0.0008525597269624573,
+      "loss": 6.3941,
+      "step": 1296
+    },
+    {
+      "epoch": 0.4426621160409556,
+      "grad_norm": 10.423567771911621,
+      "learning_rate": 0.0008524459613196815,
+      "loss": 5.7454,
+      "step": 1297
+    },
+    {
+      "epoch": 0.44300341296928325,
+      "grad_norm": 6.871647357940674,
+      "learning_rate": 0.0008523321956769056,
+      "loss": 6.844,
+      "step": 1298
+    },
+    {
+      "epoch": 0.44334470989761093,
+      "grad_norm": 4.372817039489746,
+      "learning_rate": 0.0008522184300341297,
+      "loss": 6.4617,
+      "step": 1299
+    },
+    {
+      "epoch": 0.44368600682593856,
+      "grad_norm": 3.725336790084839,
+      "learning_rate": 0.0008521046643913538,
+      "loss": 6.9746,
+      "step": 1300
+    },
+    {
+      "epoch": 0.4440273037542662,
+      "grad_norm": 3.9999923706054688,
+      "learning_rate": 0.0008519908987485779,
+      "loss": 6.1832,
+      "step": 1301
+    },
+    {
+      "epoch": 0.4443686006825939,
+      "grad_norm": 8.845232963562012,
+      "learning_rate": 0.0008518771331058021,
+      "loss": 4.6078,
+      "step": 1302
+    },
+    {
+      "epoch": 0.4447098976109215,
+      "grad_norm": 3.548273801803589,
+      "learning_rate": 0.0008517633674630262,
+      "loss": 7.1117,
+      "step": 1303
+    },
+    {
+      "epoch": 0.44505119453924913,
+      "grad_norm": 3.508362054824829,
+      "learning_rate": 0.0008516496018202503,
+      "loss": 7.0278,
+      "step": 1304
+    },
+    {
+      "epoch": 0.4453924914675768,
+      "grad_norm": 5.351772785186768,
+      "learning_rate": 0.0008515358361774744,
+      "loss": 6.3568,
+      "step": 1305
+    },
+    {
+      "epoch": 0.44573378839590444,
+      "grad_norm": 3.509801149368286,
+      "learning_rate": 0.0008514220705346985,
+      "loss": 6.6052,
+      "step": 1306
+    },
+    {
+      "epoch": 0.44607508532423207,
+      "grad_norm": 3.358114719390869,
+      "learning_rate": 0.0008513083048919226,
+      "loss": 6.3912,
+      "step": 1307
+    },
+    {
+      "epoch": 0.44641638225255975,
+      "grad_norm": 5.949875354766846,
+      "learning_rate": 0.0008511945392491469,
+      "loss": 6.5835,
+      "step": 1308
+    },
+    {
+      "epoch": 0.4467576791808874,
+      "grad_norm": 3.4979476928710938,
+      "learning_rate": 0.000851080773606371,
+      "loss": 6.1771,
+      "step": 1309
+    },
+    {
+      "epoch": 0.447098976109215,
+      "grad_norm": 3.5438613891601562,
+      "learning_rate": 0.0008509670079635951,
+      "loss": 6.2541,
+      "step": 1310
+    },
+    {
+      "epoch": 0.44744027303754264,
+      "grad_norm": 3.3947629928588867,
+      "learning_rate": 0.0008508532423208192,
+      "loss": 6.2373,
+      "step": 1311
+    },
+    {
+      "epoch": 0.4477815699658703,
+      "grad_norm": 3.5880777835845947,
+      "learning_rate": 0.0008507394766780433,
+      "loss": 6.3093,
+      "step": 1312
+    },
+    {
+      "epoch": 0.44812286689419795,
+      "grad_norm": 3.6189653873443604,
+      "learning_rate": 0.0008506257110352673,
+      "loss": 6.1852,
+      "step": 1313
+    },
+    {
+      "epoch": 0.4484641638225256,
+      "grad_norm": 3.5925283432006836,
+      "learning_rate": 0.0008505119453924915,
+      "loss": 6.7835,
+      "step": 1314
+    },
+    {
+      "epoch": 0.44880546075085326,
+      "grad_norm": 3.7551400661468506,
+      "learning_rate": 0.0008503981797497156,
+      "loss": 6.3399,
+      "step": 1315
+    },
+    {
+      "epoch": 0.4491467576791809,
+      "grad_norm": 3.489022731781006,
+      "learning_rate": 0.0008502844141069397,
+      "loss": 6.8905,
+      "step": 1316
+    },
+    {
+      "epoch": 0.4494880546075085,
+      "grad_norm": 3.6786651611328125,
+      "learning_rate": 0.0008501706484641638,
+      "loss": 6.74,
+      "step": 1317
+    },
+    {
+      "epoch": 0.4498293515358362,
+      "grad_norm": 3.3341808319091797,
+      "learning_rate": 0.0008500568828213879,
+      "loss": 6.4478,
+      "step": 1318
+    },
+    {
+      "epoch": 0.4501706484641638,
+      "grad_norm": 8.599390983581543,
+      "learning_rate": 0.000849943117178612,
+      "loss": 6.5547,
+      "step": 1319
+    },
+    {
+      "epoch": 0.45051194539249145,
+      "grad_norm": 4.72025728225708,
+      "learning_rate": 0.0008498293515358362,
+      "loss": 6.7761,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45085324232081914,
+      "grad_norm": 3.8403704166412354,
+      "learning_rate": 0.0008497155858930603,
+      "loss": 6.6878,
+      "step": 1321
+    },
+    {
+      "epoch": 0.45119453924914676,
+      "grad_norm": 3.509477376937866,
+      "learning_rate": 0.0008496018202502844,
+      "loss": 7.0436,
+      "step": 1322
+    },
+    {
+      "epoch": 0.4515358361774744,
+      "grad_norm": 3.821585178375244,
+      "learning_rate": 0.0008494880546075085,
+      "loss": 6.573,
+      "step": 1323
+    },
+    {
+      "epoch": 0.451877133105802,
+      "grad_norm": 3.739370584487915,
+      "learning_rate": 0.0008493742889647326,
+      "loss": 6.3931,
+      "step": 1324
+    },
+    {
+      "epoch": 0.4522184300341297,
+      "grad_norm": 3.2959489822387695,
+      "learning_rate": 0.0008492605233219569,
+      "loss": 6.8188,
+      "step": 1325
+    },
+    {
+      "epoch": 0.45255972696245733,
+      "grad_norm": 3.6209750175476074,
+      "learning_rate": 0.000849146757679181,
+      "loss": 6.6649,
+      "step": 1326
+    },
+    {
+      "epoch": 0.45290102389078496,
+      "grad_norm": 3.411741018295288,
+      "learning_rate": 0.0008490329920364051,
+      "loss": 6.2213,
+      "step": 1327
+    },
+    {
+      "epoch": 0.45324232081911264,
+      "grad_norm": 3.494786262512207,
+      "learning_rate": 0.0008489192263936292,
+      "loss": 6.9988,
+      "step": 1328
+    },
+    {
+      "epoch": 0.45358361774744027,
+      "grad_norm": 3.3790512084960938,
+      "learning_rate": 0.0008488054607508533,
+      "loss": 6.7258,
+      "step": 1329
+    },
+    {
+      "epoch": 0.4539249146757679,
+      "grad_norm": 3.412260055541992,
+      "learning_rate": 0.0008486916951080774,
+      "loss": 6.8724,
+      "step": 1330
+    },
+    {
+      "epoch": 0.4542662116040956,
+      "grad_norm": 3.655362844467163,
+      "learning_rate": 0.0008485779294653016,
+      "loss": 6.644,
+      "step": 1331
+    },
+    {
+      "epoch": 0.4546075085324232,
+      "grad_norm": 3.687267780303955,
+      "learning_rate": 0.0008484641638225257,
+      "loss": 6.2437,
+      "step": 1332
+    },
+    {
+      "epoch": 0.45494880546075084,
+      "grad_norm": 3.81823468208313,
+      "learning_rate": 0.0008483503981797497,
+      "loss": 6.98,
+      "step": 1333
+    },
+    {
+      "epoch": 0.4552901023890785,
+      "grad_norm": 4.4424967765808105,
+      "learning_rate": 0.0008482366325369738,
+      "loss": 6.6895,
+      "step": 1334
+    },
+    {
+      "epoch": 0.45563139931740615,
+      "grad_norm": 3.6372294425964355,
+      "learning_rate": 0.0008481228668941979,
+      "loss": 6.4528,
+      "step": 1335
+    },
+    {
+      "epoch": 0.4559726962457338,
+      "grad_norm": 4.353541851043701,
+      "learning_rate": 0.000848009101251422,
+      "loss": 6.8672,
+      "step": 1336
+    },
+    {
+      "epoch": 0.45631399317406146,
+      "grad_norm": 3.73286509513855,
+      "learning_rate": 0.0008478953356086462,
+      "loss": 6.2034,
+      "step": 1337
+    },
+    {
+      "epoch": 0.4566552901023891,
+      "grad_norm": 3.300739049911499,
+      "learning_rate": 0.0008477815699658703,
+      "loss": 6.6602,
+      "step": 1338
+    },
+    {
+      "epoch": 0.4569965870307167,
+      "grad_norm": 3.4340107440948486,
+      "learning_rate": 0.0008476678043230944,
+      "loss": 6.3554,
+      "step": 1339
+    },
+    {
+      "epoch": 0.45733788395904434,
+      "grad_norm": 3.3781628608703613,
+      "learning_rate": 0.0008475540386803185,
+      "loss": 6.4238,
+      "step": 1340
+    },
+    {
+      "epoch": 0.45767918088737203,
+      "grad_norm": 3.5280208587646484,
+      "learning_rate": 0.0008474402730375426,
+      "loss": 6.8517,
+      "step": 1341
+    },
+    {
+      "epoch": 0.45802047781569966,
+      "grad_norm": 3.4973056316375732,
+      "learning_rate": 0.0008473265073947669,
+      "loss": 6.3462,
+      "step": 1342
+    },
+    {
+      "epoch": 0.4583617747440273,
+      "grad_norm": 3.7946386337280273,
+      "learning_rate": 0.000847212741751991,
+      "loss": 6.7902,
+      "step": 1343
+    },
+    {
+      "epoch": 0.45870307167235497,
+      "grad_norm": 3.5301756858825684,
+      "learning_rate": 0.0008470989761092151,
+      "loss": 6.9557,
+      "step": 1344
+    },
+    {
+      "epoch": 0.4590443686006826,
+      "grad_norm": 3.576582431793213,
+      "learning_rate": 0.0008469852104664392,
+      "loss": 7.1152,
+      "step": 1345
+    },
+    {
+      "epoch": 0.4593856655290102,
+      "grad_norm": 3.645031452178955,
+      "learning_rate": 0.0008468714448236633,
+      "loss": 6.5057,
+      "step": 1346
+    },
+    {
+      "epoch": 0.4597269624573379,
+      "grad_norm": 3.4032137393951416,
+      "learning_rate": 0.0008467576791808874,
+      "loss": 6.8364,
+      "step": 1347
+    },
+    {
+      "epoch": 0.46006825938566553,
+      "grad_norm": 3.5544800758361816,
+      "learning_rate": 0.0008466439135381116,
+      "loss": 6.8789,
+      "step": 1348
+    },
+    {
+      "epoch": 0.46040955631399316,
+      "grad_norm": 3.475367784500122,
+      "learning_rate": 0.0008465301478953357,
+      "loss": 6.9583,
+      "step": 1349
+    },
+    {
+      "epoch": 0.46075085324232085,
+      "grad_norm": 3.5934853553771973,
+      "learning_rate": 0.0008464163822525598,
+      "loss": 7.1558,
+      "step": 1350
+    },
+    {
+      "epoch": 0.4610921501706485,
+      "grad_norm": 3.3591737747192383,
+      "learning_rate": 0.0008463026166097839,
+      "loss": 6.709,
+      "step": 1351
+    },
+    {
+      "epoch": 0.4614334470989761,
+      "grad_norm": 3.4733426570892334,
+      "learning_rate": 0.0008461888509670079,
+      "loss": 6.7051,
+      "step": 1352
+    },
+    {
+      "epoch": 0.46177474402730373,
+      "grad_norm": 3.335206985473633,
+      "learning_rate": 0.000846075085324232,
+      "loss": 6.4163,
+      "step": 1353
+    },
+    {
+      "epoch": 0.4621160409556314,
+      "grad_norm": 3.442255973815918,
+      "learning_rate": 0.0008459613196814562,
+      "loss": 6.4429,
+      "step": 1354
+    },
+    {
+      "epoch": 0.46245733788395904,
+      "grad_norm": 3.5777900218963623,
+      "learning_rate": 0.0008458475540386803,
+      "loss": 6.9031,
+      "step": 1355
+    },
+    {
+      "epoch": 0.46279863481228667,
+      "grad_norm": 3.4293618202209473,
+      "learning_rate": 0.0008457337883959044,
+      "loss": 6.5918,
+      "step": 1356
+    },
+    {
+      "epoch": 0.46313993174061435,
+      "grad_norm": 3.3813259601593018,
+      "learning_rate": 0.0008456200227531285,
+      "loss": 6.4456,
+      "step": 1357
+    },
+    {
+      "epoch": 0.463481228668942,
+      "grad_norm": 4.471978664398193,
+      "learning_rate": 0.0008455062571103526,
+      "loss": 5.7945,
+      "step": 1358
+    },
+    {
+      "epoch": 0.4638225255972696,
+      "grad_norm": 4.107702255249023,
+      "learning_rate": 0.0008453924914675767,
+      "loss": 6.4136,
+      "step": 1359
+    },
+    {
+      "epoch": 0.4641638225255973,
+      "grad_norm": 3.807112216949463,
+      "learning_rate": 0.000845278725824801,
+      "loss": 6.8553,
+      "step": 1360
+    },
+    {
+      "epoch": 0.4645051194539249,
+      "grad_norm": 3.7887256145477295,
+      "learning_rate": 0.0008451649601820251,
+      "loss": 6.4577,
+      "step": 1361
+    },
+    {
+      "epoch": 0.46484641638225255,
+      "grad_norm": 3.356083393096924,
+      "learning_rate": 0.0008450511945392492,
+      "loss": 6.4253,
+      "step": 1362
+    },
+    {
+      "epoch": 0.46518771331058023,
+      "grad_norm": 3.5079097747802734,
+      "learning_rate": 0.0008449374288964733,
+      "loss": 6.9637,
+      "step": 1363
+    },
+    {
+      "epoch": 0.46552901023890786,
+      "grad_norm": 3.414142370223999,
+      "learning_rate": 0.0008448236632536974,
+      "loss": 6.5443,
+      "step": 1364
+    },
+    {
+      "epoch": 0.4658703071672355,
+      "grad_norm": 7.6356587409973145,
+      "learning_rate": 0.0008447098976109216,
+      "loss": 6.1728,
+      "step": 1365
+    },
+    {
+      "epoch": 0.4662116040955631,
+      "grad_norm": 3.9965884685516357,
+      "learning_rate": 0.0008445961319681457,
+      "loss": 6.5908,
+      "step": 1366
+    },
+    {
+      "epoch": 0.4665529010238908,
+      "grad_norm": 3.6052424907684326,
+      "learning_rate": 0.0008444823663253698,
+      "loss": 6.8923,
+      "step": 1367
+    },
+    {
+      "epoch": 0.4668941979522184,
+      "grad_norm": 3.6835129261016846,
+      "learning_rate": 0.0008443686006825939,
+      "loss": 6.4614,
+      "step": 1368
+    },
+    {
+      "epoch": 0.46723549488054605,
+      "grad_norm": 3.5373635292053223,
+      "learning_rate": 0.000844254835039818,
+      "loss": 6.6617,
+      "step": 1369
+    },
+    {
+      "epoch": 0.46757679180887374,
+      "grad_norm": 4.537879467010498,
+      "learning_rate": 0.0008441410693970421,
+      "loss": 5.4222,
+      "step": 1370
+    },
+    {
+      "epoch": 0.46791808873720137,
+      "grad_norm": 3.356199026107788,
+      "learning_rate": 0.0008440273037542663,
+      "loss": 6.792,
+      "step": 1371
+    },
+    {
+      "epoch": 0.468259385665529,
+      "grad_norm": 3.830232858657837,
+      "learning_rate": 0.0008439135381114903,
+      "loss": 6.6182,
+      "step": 1372
+    },
+    {
+      "epoch": 0.4686006825938567,
+      "grad_norm": 3.4203712940216064,
+      "learning_rate": 0.0008437997724687144,
+      "loss": 6.5533,
+      "step": 1373
+    },
+    {
+      "epoch": 0.4689419795221843,
+      "grad_norm": 3.5441665649414062,
+      "learning_rate": 0.0008436860068259385,
+      "loss": 6.5479,
+      "step": 1374
+    },
+    {
+      "epoch": 0.46928327645051193,
+      "grad_norm": 3.4909310340881348,
+      "learning_rate": 0.0008435722411831626,
+      "loss": 6.1958,
+      "step": 1375
+    },
+    {
+      "epoch": 0.4696245733788396,
+      "grad_norm": 3.6683506965637207,
+      "learning_rate": 0.0008434584755403867,
+      "loss": 6.6273,
+      "step": 1376
+    },
+    {
+      "epoch": 0.46996587030716724,
+      "grad_norm": 4.192461013793945,
+      "learning_rate": 0.000843344709897611,
+      "loss": 6.0141,
+      "step": 1377
+    },
+    {
+      "epoch": 0.47030716723549487,
+      "grad_norm": 3.5298233032226562,
+      "learning_rate": 0.0008432309442548351,
+      "loss": 6.9406,
+      "step": 1378
+    },
+    {
+      "epoch": 0.4706484641638225,
+      "grad_norm": 3.549668550491333,
+      "learning_rate": 0.0008431171786120592,
+      "loss": 6.8979,
+      "step": 1379
+    },
+    {
+      "epoch": 0.4709897610921502,
+      "grad_norm": 3.3617653846740723,
+      "learning_rate": 0.0008430034129692833,
+      "loss": 6.4427,
+      "step": 1380
+    },
+    {
+      "epoch": 0.4713310580204778,
+      "grad_norm": 3.317580461502075,
+      "learning_rate": 0.0008428896473265074,
+      "loss": 7.0131,
+      "step": 1381
+    },
+    {
+      "epoch": 0.47167235494880544,
+      "grad_norm": 3.481401205062866,
+      "learning_rate": 0.0008427758816837315,
+      "loss": 6.6665,
+      "step": 1382
+    },
+    {
+      "epoch": 0.4720136518771331,
+      "grad_norm": 3.4528825283050537,
+      "learning_rate": 0.0008426621160409557,
+      "loss": 6.6673,
+      "step": 1383
+    },
+    {
+      "epoch": 0.47235494880546075,
+      "grad_norm": 4.272742748260498,
+      "learning_rate": 0.0008425483503981798,
+      "loss": 5.9544,
+      "step": 1384
+    },
+    {
+      "epoch": 0.4726962457337884,
+      "grad_norm": 3.5958762168884277,
+      "learning_rate": 0.0008424345847554039,
+      "loss": 6.8734,
+      "step": 1385
+    },
+    {
+      "epoch": 0.47303754266211606,
+      "grad_norm": 3.564652681350708,
+      "learning_rate": 0.000842320819112628,
+      "loss": 6.846,
+      "step": 1386
+    },
+    {
+      "epoch": 0.4733788395904437,
+      "grad_norm": 3.3910999298095703,
+      "learning_rate": 0.0008422070534698521,
+      "loss": 6.3894,
+      "step": 1387
+    },
+    {
+      "epoch": 0.4737201365187713,
+      "grad_norm": 4.498713493347168,
+      "learning_rate": 0.0008420932878270763,
+      "loss": 6.7253,
+      "step": 1388
+    },
+    {
+      "epoch": 0.474061433447099,
+      "grad_norm": 3.539781332015991,
+      "learning_rate": 0.0008419795221843004,
+      "loss": 7.1748,
+      "step": 1389
+    },
+    {
+      "epoch": 0.47440273037542663,
+      "grad_norm": 3.6499907970428467,
+      "learning_rate": 0.0008418657565415245,
+      "loss": 7.0704,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47474402730375426,
+      "grad_norm": 3.4689838886260986,
+      "learning_rate": 0.0008417519908987485,
+      "loss": 6.6226,
+      "step": 1391
+    },
+    {
+      "epoch": 0.4750853242320819,
+      "grad_norm": 3.3279478549957275,
+      "learning_rate": 0.0008416382252559726,
+      "loss": 6.6018,
+      "step": 1392
+    },
+    {
+      "epoch": 0.47542662116040957,
+      "grad_norm": 3.3303651809692383,
+      "learning_rate": 0.0008415244596131967,
+      "loss": 6.6632,
+      "step": 1393
+    },
+    {
+      "epoch": 0.4757679180887372,
+      "grad_norm": 3.5589680671691895,
+      "learning_rate": 0.000841410693970421,
+      "loss": 6.7845,
+      "step": 1394
+    },
+    {
+      "epoch": 0.4761092150170648,
+      "grad_norm": 3.484956741333008,
+      "learning_rate": 0.0008412969283276451,
+      "loss": 6.8862,
+      "step": 1395
+    },
+    {
+      "epoch": 0.4764505119453925,
+      "grad_norm": 4.03617525100708,
+      "learning_rate": 0.0008411831626848692,
+      "loss": 6.6096,
+      "step": 1396
+    },
+    {
+      "epoch": 0.47679180887372014,
+      "grad_norm": 3.343886375427246,
+      "learning_rate": 0.0008410693970420933,
+      "loss": 6.322,
+      "step": 1397
+    },
+    {
+      "epoch": 0.47713310580204776,
+      "grad_norm": 3.6128275394439697,
+      "learning_rate": 0.0008409556313993174,
+      "loss": 6.9015,
+      "step": 1398
+    },
+    {
+      "epoch": 0.47747440273037545,
+      "grad_norm": 3.4812707901000977,
+      "learning_rate": 0.0008408418657565415,
+      "loss": 6.8565,
+      "step": 1399
+    },
+    {
+      "epoch": 0.4778156996587031,
+      "grad_norm": 3.590444326400757,
+      "learning_rate": 0.0008407281001137657,
+      "loss": 6.534,
+      "step": 1400
+    },
+    {
+      "epoch": 0.4781569965870307,
+      "grad_norm": 3.251206159591675,
+      "learning_rate": 0.0008406143344709898,
+      "loss": 6.7066,
+      "step": 1401
+    },
+    {
+      "epoch": 0.4784982935153584,
+      "grad_norm": 3.3419792652130127,
+      "learning_rate": 0.0008405005688282139,
+      "loss": 6.9371,
+      "step": 1402
+    },
+    {
+      "epoch": 0.478839590443686,
+      "grad_norm": 3.3254618644714355,
+      "learning_rate": 0.000840386803185438,
+      "loss": 6.8618,
+      "step": 1403
+    },
+    {
+      "epoch": 0.47918088737201364,
+      "grad_norm": 5.28602409362793,
+      "learning_rate": 0.0008402730375426621,
+      "loss": 5.7656,
+      "step": 1404
+    },
+    {
+      "epoch": 0.47952218430034127,
+      "grad_norm": 4.599923610687256,
+      "learning_rate": 0.0008401592718998863,
+      "loss": 5.8085,
+      "step": 1405
+    },
+    {
+      "epoch": 0.47986348122866895,
+      "grad_norm": 3.606318712234497,
+      "learning_rate": 0.0008400455062571104,
+      "loss": 6.6817,
+      "step": 1406
+    },
+    {
+      "epoch": 0.4802047781569966,
+      "grad_norm": 3.839956283569336,
+      "learning_rate": 0.0008399317406143345,
+      "loss": 7.2384,
+      "step": 1407
+    },
+    {
+      "epoch": 0.4805460750853242,
+      "grad_norm": 3.5408711433410645,
+      "learning_rate": 0.0008398179749715586,
+      "loss": 6.7327,
+      "step": 1408
+    },
+    {
+      "epoch": 0.4808873720136519,
+      "grad_norm": 3.496644973754883,
+      "learning_rate": 0.0008397042093287828,
+      "loss": 6.9381,
+      "step": 1409
+    },
+    {
+      "epoch": 0.4812286689419795,
+      "grad_norm": 3.4285078048706055,
+      "learning_rate": 0.0008395904436860067,
+      "loss": 6.5183,
+      "step": 1410
+    },
+    {
+      "epoch": 0.48156996587030715,
+      "grad_norm": 3.322774648666382,
+      "learning_rate": 0.000839476678043231,
+      "loss": 6.3513,
+      "step": 1411
+    },
+    {
+      "epoch": 0.48191126279863483,
+      "grad_norm": 3.334423542022705,
+      "learning_rate": 0.0008393629124004551,
+      "loss": 6.9284,
+      "step": 1412
+    },
+    {
+      "epoch": 0.48225255972696246,
+      "grad_norm": 3.4194326400756836,
+      "learning_rate": 0.0008392491467576792,
+      "loss": 7.0258,
+      "step": 1413
+    },
+    {
+      "epoch": 0.4825938566552901,
+      "grad_norm": 5.742277145385742,
+      "learning_rate": 0.0008391353811149033,
+      "loss": 5.9465,
+      "step": 1414
+    },
+    {
+      "epoch": 0.48293515358361777,
+      "grad_norm": 3.6702628135681152,
+      "learning_rate": 0.0008390216154721274,
+      "loss": 6.7509,
+      "step": 1415
+    },
+    {
+      "epoch": 0.4832764505119454,
+      "grad_norm": 3.5624711513519287,
+      "learning_rate": 0.0008389078498293515,
+      "loss": 7.3361,
+      "step": 1416
+    },
+    {
+      "epoch": 0.483617747440273,
+      "grad_norm": 3.3893463611602783,
+      "learning_rate": 0.0008387940841865757,
+      "loss": 6.5935,
+      "step": 1417
+    },
+    {
+      "epoch": 0.48395904436860065,
+      "grad_norm": 4.454052448272705,
+      "learning_rate": 0.0008386803185437998,
+      "loss": 6.352,
+      "step": 1418
+    },
+    {
+      "epoch": 0.48430034129692834,
+      "grad_norm": 3.6366562843322754,
+      "learning_rate": 0.0008385665529010239,
+      "loss": 6.7991,
+      "step": 1419
+    },
+    {
+      "epoch": 0.48464163822525597,
+      "grad_norm": 3.4529809951782227,
+      "learning_rate": 0.000838452787258248,
+      "loss": 6.6057,
+      "step": 1420
+    },
+    {
+      "epoch": 0.4849829351535836,
+      "grad_norm": 3.475712776184082,
+      "learning_rate": 0.0008383390216154721,
+      "loss": 5.9796,
+      "step": 1421
+    },
+    {
+      "epoch": 0.4853242320819113,
+      "grad_norm": 3.537789821624756,
+      "learning_rate": 0.0008382252559726962,
+      "loss": 7.112,
+      "step": 1422
+    },
+    {
+      "epoch": 0.4856655290102389,
+      "grad_norm": 3.504626750946045,
+      "learning_rate": 0.0008381114903299204,
+      "loss": 6.8114,
+      "step": 1423
+    },
+    {
+      "epoch": 0.48600682593856653,
+      "grad_norm": 3.580801486968994,
+      "learning_rate": 0.0008379977246871445,
+      "loss": 6.4932,
+      "step": 1424
+    },
+    {
+      "epoch": 0.4863481228668942,
+      "grad_norm": 3.4925484657287598,
+      "learning_rate": 0.0008378839590443686,
+      "loss": 6.7872,
+      "step": 1425
+    },
+    {
+      "epoch": 0.48668941979522184,
+      "grad_norm": 3.814373254776001,
+      "learning_rate": 0.0008377701934015928,
+      "loss": 6.9086,
+      "step": 1426
+    },
+    {
+      "epoch": 0.4870307167235495,
+      "grad_norm": 3.2810380458831787,
+      "learning_rate": 0.0008376564277588169,
+      "loss": 6.6771,
+      "step": 1427
+    },
+    {
+      "epoch": 0.48737201365187716,
+      "grad_norm": 3.3674893379211426,
+      "learning_rate": 0.0008375426621160411,
+      "loss": 6.7597,
+      "step": 1428
+    },
+    {
+      "epoch": 0.4877133105802048,
+      "grad_norm": 3.3947198390960693,
+      "learning_rate": 0.0008374288964732652,
+      "loss": 6.5867,
+      "step": 1429
+    },
+    {
+      "epoch": 0.4880546075085324,
+      "grad_norm": 3.6930675506591797,
+      "learning_rate": 0.0008373151308304892,
+      "loss": 6.7218,
+      "step": 1430
+    },
+    {
+      "epoch": 0.4883959044368601,
+      "grad_norm": 3.436976909637451,
+      "learning_rate": 0.0008372013651877133,
+      "loss": 6.906,
+      "step": 1431
+    },
+    {
+      "epoch": 0.4887372013651877,
+      "grad_norm": 3.462581157684326,
+      "learning_rate": 0.0008370875995449374,
+      "loss": 6.7478,
+      "step": 1432
+    },
+    {
+      "epoch": 0.48907849829351535,
+      "grad_norm": 3.483604907989502,
+      "learning_rate": 0.0008369738339021615,
+      "loss": 6.3906,
+      "step": 1433
+    },
+    {
+      "epoch": 0.489419795221843,
+      "grad_norm": 3.3858816623687744,
+      "learning_rate": 0.0008368600682593857,
+      "loss": 6.3999,
+      "step": 1434
+    },
+    {
+      "epoch": 0.48976109215017066,
+      "grad_norm": 3.541221857070923,
+      "learning_rate": 0.0008367463026166098,
+      "loss": 7.3143,
+      "step": 1435
+    },
+    {
+      "epoch": 0.4901023890784983,
+      "grad_norm": 3.380033254623413,
+      "learning_rate": 0.0008366325369738339,
+      "loss": 6.7311,
+      "step": 1436
+    },
+    {
+      "epoch": 0.4904436860068259,
+      "grad_norm": 3.423128604888916,
+      "learning_rate": 0.000836518771331058,
+      "loss": 6.5591,
+      "step": 1437
+    },
+    {
+      "epoch": 0.4907849829351536,
+      "grad_norm": 3.3924520015716553,
+      "learning_rate": 0.0008364050056882821,
+      "loss": 6.8674,
+      "step": 1438
+    },
+    {
+      "epoch": 0.49112627986348123,
+      "grad_norm": 3.7471725940704346,
+      "learning_rate": 0.0008362912400455062,
+      "loss": 6.6909,
+      "step": 1439
+    },
+    {
+      "epoch": 0.49146757679180886,
+      "grad_norm": 3.535061836242676,
+      "learning_rate": 0.0008361774744027304,
+      "loss": 6.5537,
+      "step": 1440
+    },
+    {
+      "epoch": 0.49180887372013654,
+      "grad_norm": 3.317924976348877,
+      "learning_rate": 0.0008360637087599545,
+      "loss": 6.6799,
+      "step": 1441
+    },
+    {
+      "epoch": 0.49215017064846417,
+      "grad_norm": 5.021495342254639,
+      "learning_rate": 0.0008359499431171786,
+      "loss": 5.363,
+      "step": 1442
+    },
+    {
+      "epoch": 0.4924914675767918,
+      "grad_norm": 3.5335588455200195,
+      "learning_rate": 0.0008358361774744028,
+      "loss": 6.8348,
+      "step": 1443
+    },
+    {
+      "epoch": 0.4928327645051195,
+      "grad_norm": 3.5306732654571533,
+      "learning_rate": 0.0008357224118316269,
+      "loss": 6.6957,
+      "step": 1444
+    },
+    {
+      "epoch": 0.4931740614334471,
+      "grad_norm": 3.4954781532287598,
+      "learning_rate": 0.0008356086461888511,
+      "loss": 6.8805,
+      "step": 1445
+    },
+    {
+      "epoch": 0.49351535836177474,
+      "grad_norm": 3.4456214904785156,
+      "learning_rate": 0.0008354948805460752,
+      "loss": 6.6546,
+      "step": 1446
+    },
+    {
+      "epoch": 0.49385665529010236,
+      "grad_norm": 3.323965311050415,
+      "learning_rate": 0.0008353811149032993,
+      "loss": 6.8501,
+      "step": 1447
+    },
+    {
+      "epoch": 0.49419795221843005,
+      "grad_norm": 3.692003011703491,
+      "learning_rate": 0.0008352673492605234,
+      "loss": 6.282,
+      "step": 1448
+    },
+    {
+      "epoch": 0.4945392491467577,
+      "grad_norm": 3.3409910202026367,
+      "learning_rate": 0.0008351535836177474,
+      "loss": 6.6912,
+      "step": 1449
+    },
+    {
+      "epoch": 0.4948805460750853,
+      "grad_norm": 3.5133414268493652,
+      "learning_rate": 0.0008350398179749715,
+      "loss": 6.0178,
+      "step": 1450
+    },
+    {
+      "epoch": 0.495221843003413,
+      "grad_norm": 4.341811656951904,
+      "learning_rate": 0.0008349260523321957,
+      "loss": 6.4182,
+      "step": 1451
+    },
+    {
+      "epoch": 0.4955631399317406,
+      "grad_norm": 4.636251926422119,
+      "learning_rate": 0.0008348122866894198,
+      "loss": 6.824,
+      "step": 1452
+    },
+    {
+      "epoch": 0.49590443686006824,
+      "grad_norm": 5.0915632247924805,
+      "learning_rate": 0.0008346985210466439,
+      "loss": 5.3702,
+      "step": 1453
+    },
+    {
+      "epoch": 0.4962457337883959,
+      "grad_norm": 3.465383529663086,
+      "learning_rate": 0.000834584755403868,
+      "loss": 6.642,
+      "step": 1454
+    },
+    {
+      "epoch": 0.49658703071672355,
+      "grad_norm": 3.462819814682007,
+      "learning_rate": 0.0008344709897610921,
+      "loss": 6.5499,
+      "step": 1455
+    },
+    {
+      "epoch": 0.4969283276450512,
+      "grad_norm": 3.4404056072235107,
+      "learning_rate": 0.0008343572241183162,
+      "loss": 6.8499,
+      "step": 1456
+    },
+    {
+      "epoch": 0.49726962457337887,
+      "grad_norm": 3.342381238937378,
+      "learning_rate": 0.0008342434584755404,
+      "loss": 6.5637,
+      "step": 1457
+    },
+    {
+      "epoch": 0.4976109215017065,
+      "grad_norm": 3.599907875061035,
+      "learning_rate": 0.0008341296928327645,
+      "loss": 6.5076,
+      "step": 1458
+    },
+    {
+      "epoch": 0.4979522184300341,
+      "grad_norm": 3.521761178970337,
+      "learning_rate": 0.0008340159271899886,
+      "loss": 6.557,
+      "step": 1459
+    },
+    {
+      "epoch": 0.49829351535836175,
+      "grad_norm": 4.715250015258789,
+      "learning_rate": 0.0008339021615472128,
+      "loss": 5.1495,
+      "step": 1460
+    },
+    {
+      "epoch": 0.49863481228668943,
+      "grad_norm": 3.4854562282562256,
+      "learning_rate": 0.0008337883959044369,
+      "loss": 6.7593,
+      "step": 1461
+    },
+    {
+      "epoch": 0.49897610921501706,
+      "grad_norm": 3.903296947479248,
+      "learning_rate": 0.000833674630261661,
+      "loss": 6.1174,
+      "step": 1462
+    },
+    {
+      "epoch": 0.4993174061433447,
+      "grad_norm": 3.4270193576812744,
+      "learning_rate": 0.0008335608646188852,
+      "loss": 6.5968,
+      "step": 1463
+    },
+    {
+      "epoch": 0.49965870307167237,
+      "grad_norm": 8.496833801269531,
+      "learning_rate": 0.0008334470989761093,
+      "loss": 6.2964,
+      "step": 1464
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6175074577331543,
+      "learning_rate": 0.0008333333333333334,
+      "loss": 6.989,
+      "step": 1465
+    },
+    {
+      "epoch": 0.5003412969283276,
+      "grad_norm": 3.5683434009552,
+      "learning_rate": 0.0008332195676905575,
+      "loss": 7.0006,
+      "step": 1466
+    },
+    {
+      "epoch": 0.5006825938566553,
+      "grad_norm": 3.887437582015991,
+      "learning_rate": 0.0008331058020477816,
+      "loss": 6.1925,
+      "step": 1467
+    },
+    {
+      "epoch": 0.5010238907849829,
+      "grad_norm": 3.302295207977295,
+      "learning_rate": 0.0008329920364050058,
+      "loss": 6.4899,
+      "step": 1468
+    },
+    {
+      "epoch": 0.5013651877133106,
+      "grad_norm": 3.6436431407928467,
+      "learning_rate": 0.0008328782707622298,
+      "loss": 6.103,
+      "step": 1469
+    },
+    {
+      "epoch": 0.5017064846416383,
+      "grad_norm": 3.531428337097168,
+      "learning_rate": 0.0008327645051194539,
+      "loss": 6.1611,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5020477815699659,
+      "grad_norm": 3.350801944732666,
+      "learning_rate": 0.000832650739476678,
+      "loss": 6.4045,
+      "step": 1471
+    },
+    {
+      "epoch": 0.5023890784982935,
+      "grad_norm": 4.109683513641357,
+      "learning_rate": 0.0008325369738339021,
+      "loss": 3.9738,
+      "step": 1472
+    },
+    {
+      "epoch": 0.5027303754266211,
+      "grad_norm": 3.4948461055755615,
+      "learning_rate": 0.0008324232081911262,
+      "loss": 6.8257,
+      "step": 1473
+    },
+    {
+      "epoch": 0.5030716723549488,
+      "grad_norm": 3.710834264755249,
+      "learning_rate": 0.0008323094425483504,
+      "loss": 6.6345,
+      "step": 1474
+    },
+    {
+      "epoch": 0.5034129692832765,
+      "grad_norm": 3.6197144985198975,
+      "learning_rate": 0.0008321956769055745,
+      "loss": 6.386,
+      "step": 1475
+    },
+    {
+      "epoch": 0.5037542662116041,
+      "grad_norm": 3.6927449703216553,
+      "learning_rate": 0.0008320819112627986,
+      "loss": 6.3214,
+      "step": 1476
+    },
+    {
+      "epoch": 0.5040955631399318,
+      "grad_norm": 3.6893727779388428,
+      "learning_rate": 0.0008319681456200228,
+      "loss": 6.8485,
+      "step": 1477
+    },
+    {
+      "epoch": 0.5044368600682594,
+      "grad_norm": 4.398456573486328,
+      "learning_rate": 0.0008318543799772469,
+      "loss": 6.3134,
+      "step": 1478
+    },
+    {
+      "epoch": 0.504778156996587,
+      "grad_norm": 3.5371034145355225,
+      "learning_rate": 0.000831740614334471,
+      "loss": 6.685,
+      "step": 1479
+    },
+    {
+      "epoch": 0.5051194539249146,
+      "grad_norm": 3.924985885620117,
+      "learning_rate": 0.0008316268486916952,
+      "loss": 6.4405,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5054607508532423,
+      "grad_norm": 3.409550666809082,
+      "learning_rate": 0.0008315130830489193,
+      "loss": 6.5593,
+      "step": 1481
+    },
+    {
+      "epoch": 0.50580204778157,
+      "grad_norm": 3.491574287414551,
+      "learning_rate": 0.0008313993174061434,
+      "loss": 6.2359,
+      "step": 1482
+    },
+    {
+      "epoch": 0.5061433447098976,
+      "grad_norm": 3.836836338043213,
+      "learning_rate": 0.0008312855517633675,
+      "loss": 6.3855,
+      "step": 1483
+    },
+    {
+      "epoch": 0.5064846416382253,
+      "grad_norm": 5.168152809143066,
+      "learning_rate": 0.0008311717861205916,
+      "loss": 4.9784,
+      "step": 1484
+    },
+    {
+      "epoch": 0.5068259385665529,
+      "grad_norm": 9.451974868774414,
+      "learning_rate": 0.0008310580204778158,
+      "loss": 6.8587,
+      "step": 1485
+    },
+    {
+      "epoch": 0.5071672354948805,
+      "grad_norm": 3.954226493835449,
+      "learning_rate": 0.0008309442548350399,
+      "loss": 7.1898,
+      "step": 1486
+    },
+    {
+      "epoch": 0.5075085324232081,
+      "grad_norm": 5.50667142868042,
+      "learning_rate": 0.000830830489192264,
+      "loss": 6.3041,
+      "step": 1487
+    },
+    {
+      "epoch": 0.5078498293515359,
+      "grad_norm": 3.840693235397339,
+      "learning_rate": 0.000830716723549488,
+      "loss": 6.9871,
+      "step": 1488
+    },
+    {
+      "epoch": 0.5081911262798635,
+      "grad_norm": 8.774356842041016,
+      "learning_rate": 0.0008306029579067121,
+      "loss": 8.2422,
+      "step": 1489
+    },
+    {
+      "epoch": 0.5085324232081911,
+      "grad_norm": 3.4625966548919678,
+      "learning_rate": 0.0008304891922639362,
+      "loss": 6.9081,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5088737201365188,
+      "grad_norm": 3.4914445877075195,
+      "learning_rate": 0.0008303754266211604,
+      "loss": 6.7927,
+      "step": 1491
+    },
+    {
+      "epoch": 0.5092150170648464,
+      "grad_norm": 3.4084482192993164,
+      "learning_rate": 0.0008302616609783845,
+      "loss": 7.4133,
+      "step": 1492
+    },
+    {
+      "epoch": 0.509556313993174,
+      "grad_norm": 7.066652774810791,
+      "learning_rate": 0.0008301478953356087,
+      "loss": 6.348,
+      "step": 1493
+    },
+    {
+      "epoch": 0.5098976109215017,
+      "grad_norm": 4.37141227722168,
+      "learning_rate": 0.0008300341296928328,
+      "loss": 6.462,
+      "step": 1494
+    },
+    {
+      "epoch": 0.5102389078498294,
+      "grad_norm": 5.614762783050537,
+      "learning_rate": 0.0008299203640500569,
+      "loss": 6.7,
+      "step": 1495
+    },
+    {
+      "epoch": 0.510580204778157,
+      "grad_norm": 4.1260271072387695,
+      "learning_rate": 0.000829806598407281,
+      "loss": 6.0831,
+      "step": 1496
+    },
+    {
+      "epoch": 0.5109215017064846,
+      "grad_norm": 3.777806282043457,
+      "learning_rate": 0.0008296928327645052,
+      "loss": 7.0977,
+      "step": 1497
+    },
+    {
+      "epoch": 0.5112627986348123,
+      "grad_norm": 3.448523759841919,
+      "learning_rate": 0.0008295790671217293,
+      "loss": 6.7309,
+      "step": 1498
+    },
+    {
+      "epoch": 0.5116040955631399,
+      "grad_norm": 3.3855607509613037,
+      "learning_rate": 0.0008294653014789534,
+      "loss": 6.2709,
+      "step": 1499
+    },
+    {
+      "epoch": 0.5119453924914675,
+      "grad_norm": 3.435232162475586,
+      "learning_rate": 0.0008293515358361775,
+      "loss": 6.3555,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5122866894197953,
+      "grad_norm": 3.2398738861083984,
+      "learning_rate": 0.0008292377701934016,
+      "loss": 3.8611,
+      "step": 1501
+    },
+    {
+      "epoch": 0.5126279863481229,
+      "grad_norm": 4.865748882293701,
+      "learning_rate": 0.0008291240045506257,
+      "loss": 6.4543,
+      "step": 1502
+    },
+    {
+      "epoch": 0.5129692832764505,
+      "grad_norm": 4.460921287536621,
+      "learning_rate": 0.0008290102389078499,
+      "loss": 6.4598,
+      "step": 1503
+    },
+    {
+      "epoch": 0.5133105802047782,
+      "grad_norm": 4.03560209274292,
+      "learning_rate": 0.000828896473265074,
+      "loss": 6.5586,
+      "step": 1504
+    },
+    {
+      "epoch": 0.5136518771331058,
+      "grad_norm": 3.64315128326416,
+      "learning_rate": 0.0008287827076222981,
+      "loss": 7.1097,
+      "step": 1505
+    },
+    {
+      "epoch": 0.5139931740614334,
+      "grad_norm": 4.334589004516602,
+      "learning_rate": 0.0008286689419795222,
+      "loss": 6.402,
+      "step": 1506
+    },
+    {
+      "epoch": 0.514334470989761,
+      "grad_norm": 27.29024314880371,
+      "learning_rate": 0.0008285551763367463,
+      "loss": 7.7064,
+      "step": 1507
+    },
+    {
+      "epoch": 0.5146757679180888,
+      "grad_norm": 4.106762886047363,
+      "learning_rate": 0.0008284414106939704,
+      "loss": 6.5886,
+      "step": 1508
+    },
+    {
+      "epoch": 0.5150170648464164,
+      "grad_norm": 4.14035701751709,
+      "learning_rate": 0.0008283276450511945,
+      "loss": 7.2057,
+      "step": 1509
+    },
+    {
+      "epoch": 0.515358361774744,
+      "grad_norm": 3.5756776332855225,
+      "learning_rate": 0.0008282138794084187,
+      "loss": 6.7014,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5156996587030717,
+      "grad_norm": 3.2851498126983643,
+      "learning_rate": 0.0008281001137656428,
+      "loss": 6.6437,
+      "step": 1511
+    },
+    {
+      "epoch": 0.5160409556313993,
+      "grad_norm": 3.490445375442505,
+      "learning_rate": 0.0008279863481228669,
+      "loss": 6.464,
+      "step": 1512
+    },
+    {
+      "epoch": 0.5163822525597269,
+      "grad_norm": 3.3551950454711914,
+      "learning_rate": 0.000827872582480091,
+      "loss": 6.5594,
+      "step": 1513
+    },
+    {
+      "epoch": 0.5167235494880547,
+      "grad_norm": 3.4405341148376465,
+      "learning_rate": 0.0008277588168373152,
+      "loss": 6.8159,
+      "step": 1514
+    },
+    {
+      "epoch": 0.5170648464163823,
+      "grad_norm": 3.30784010887146,
+      "learning_rate": 0.0008276450511945393,
+      "loss": 6.8372,
+      "step": 1515
+    },
+    {
+      "epoch": 0.5174061433447099,
+      "grad_norm": 3.6393938064575195,
+      "learning_rate": 0.0008275312855517634,
+      "loss": 6.9488,
+      "step": 1516
+    },
+    {
+      "epoch": 0.5177474402730375,
+      "grad_norm": 3.558758497238159,
+      "learning_rate": 0.0008274175199089875,
+      "loss": 6.452,
+      "step": 1517
+    },
+    {
+      "epoch": 0.5180887372013652,
+      "grad_norm": 3.8486268520355225,
+      "learning_rate": 0.0008273037542662116,
+      "loss": 6.601,
+      "step": 1518
+    },
+    {
+      "epoch": 0.5184300341296928,
+      "grad_norm": 3.421262741088867,
+      "learning_rate": 0.0008271899886234357,
+      "loss": 6.6491,
+      "step": 1519
+    },
+    {
+      "epoch": 0.5187713310580204,
+      "grad_norm": 3.4998815059661865,
+      "learning_rate": 0.0008270762229806599,
+      "loss": 6.5504,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5191126279863482,
+      "grad_norm": 3.533386707305908,
+      "learning_rate": 0.000826962457337884,
+      "loss": 6.797,
+      "step": 1521
+    },
+    {
+      "epoch": 0.5194539249146758,
+      "grad_norm": 3.6013705730438232,
+      "learning_rate": 0.0008268486916951081,
+      "loss": 7.1215,
+      "step": 1522
+    },
+    {
+      "epoch": 0.5197952218430034,
+      "grad_norm": 3.9131691455841064,
+      "learning_rate": 0.0008267349260523322,
+      "loss": 6.4156,
+      "step": 1523
+    },
+    {
+      "epoch": 0.520136518771331,
+      "grad_norm": 3.614016532897949,
+      "learning_rate": 0.0008266211604095563,
+      "loss": 6.9913,
+      "step": 1524
+    },
+    {
+      "epoch": 0.5204778156996587,
+      "grad_norm": 4.081565856933594,
+      "learning_rate": 0.0008265073947667804,
+      "loss": 6.7432,
+      "step": 1525
+    },
+    {
+      "epoch": 0.5208191126279863,
+      "grad_norm": 3.426879405975342,
+      "learning_rate": 0.0008263936291240047,
+      "loss": 7.093,
+      "step": 1526
+    },
+    {
+      "epoch": 0.521160409556314,
+      "grad_norm": 3.377384662628174,
+      "learning_rate": 0.0008262798634812287,
+      "loss": 6.8175,
+      "step": 1527
+    },
+    {
+      "epoch": 0.5215017064846417,
+      "grad_norm": 3.400791883468628,
+      "learning_rate": 0.0008261660978384528,
+      "loss": 6.8663,
+      "step": 1528
+    },
+    {
+      "epoch": 0.5218430034129693,
+      "grad_norm": 3.3176791667938232,
+      "learning_rate": 0.0008260523321956769,
+      "loss": 7.0355,
+      "step": 1529
+    },
+    {
+      "epoch": 0.5221843003412969,
+      "grad_norm": 3.3764326572418213,
+      "learning_rate": 0.000825938566552901,
+      "loss": 6.8315,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5225255972696246,
+      "grad_norm": 3.2921946048736572,
+      "learning_rate": 0.0008258248009101252,
+      "loss": 6.869,
+      "step": 1531
+    },
+    {
+      "epoch": 0.5228668941979522,
+      "grad_norm": 3.3354740142822266,
+      "learning_rate": 0.0008257110352673493,
+      "loss": 6.8916,
+      "step": 1532
+    },
+    {
+      "epoch": 0.5232081911262798,
+      "grad_norm": 7.8725199699401855,
+      "learning_rate": 0.0008255972696245734,
+      "loss": 5.065,
+      "step": 1533
+    },
+    {
+      "epoch": 0.5235494880546075,
+      "grad_norm": 3.697795867919922,
+      "learning_rate": 0.0008254835039817975,
+      "loss": 6.9579,
+      "step": 1534
+    },
+    {
+      "epoch": 0.5238907849829352,
+      "grad_norm": 8.069738388061523,
+      "learning_rate": 0.0008253697383390216,
+      "loss": 6.3024,
+      "step": 1535
+    },
+    {
+      "epoch": 0.5242320819112628,
+      "grad_norm": 4.431558132171631,
+      "learning_rate": 0.0008252559726962457,
+      "loss": 6.1488,
+      "step": 1536
+    },
+    {
+      "epoch": 0.5245733788395904,
+      "grad_norm": 4.185642719268799,
+      "learning_rate": 0.0008251422070534699,
+      "loss": 7.1044,
+      "step": 1537
+    },
+    {
+      "epoch": 0.5249146757679181,
+      "grad_norm": 3.705733060836792,
+      "learning_rate": 0.000825028441410694,
+      "loss": 6.8109,
+      "step": 1538
+    },
+    {
+      "epoch": 0.5252559726962457,
+      "grad_norm": 3.4032537937164307,
+      "learning_rate": 0.0008249146757679181,
+      "loss": 6.6804,
+      "step": 1539
+    },
+    {
+      "epoch": 0.5255972696245734,
+      "grad_norm": 3.183061122894287,
+      "learning_rate": 0.0008248009101251422,
+      "loss": 6.3562,
+      "step": 1540
+    },
+    {
+      "epoch": 0.525938566552901,
+      "grad_norm": 3.406890630722046,
+      "learning_rate": 0.0008246871444823663,
+      "loss": 6.4419,
+      "step": 1541
+    },
+    {
+      "epoch": 0.5262798634812287,
+      "grad_norm": 3.5061683654785156,
+      "learning_rate": 0.0008245733788395904,
+      "loss": 6.7721,
+      "step": 1542
+    },
+    {
+      "epoch": 0.5266211604095563,
+      "grad_norm": 3.441354751586914,
+      "learning_rate": 0.0008244596131968147,
+      "loss": 6.1766,
+      "step": 1543
+    },
+    {
+      "epoch": 0.5269624573378839,
+      "grad_norm": 4.2335429191589355,
+      "learning_rate": 0.0008243458475540388,
+      "loss": 6.2502,
+      "step": 1544
+    },
+    {
+      "epoch": 0.5273037542662116,
+      "grad_norm": 3.5456383228302,
+      "learning_rate": 0.0008242320819112629,
+      "loss": 6.5149,
+      "step": 1545
+    },
+    {
+      "epoch": 0.5276450511945392,
+      "grad_norm": 3.58383846282959,
+      "learning_rate": 0.000824118316268487,
+      "loss": 6.4269,
+      "step": 1546
+    },
+    {
+      "epoch": 0.5279863481228669,
+      "grad_norm": 4.371920585632324,
+      "learning_rate": 0.000824004550625711,
+      "loss": 6.4354,
+      "step": 1547
+    },
+    {
+      "epoch": 0.5283276450511946,
+      "grad_norm": 3.588916063308716,
+      "learning_rate": 0.0008238907849829351,
+      "loss": 6.6855,
+      "step": 1548
+    },
+    {
+      "epoch": 0.5286689419795222,
+      "grad_norm": 4.130184650421143,
+      "learning_rate": 0.0008237770193401593,
+      "loss": 6.1499,
+      "step": 1549
+    },
+    {
+      "epoch": 0.5290102389078498,
+      "grad_norm": 4.080915927886963,
+      "learning_rate": 0.0008236632536973834,
+      "loss": 5.4587,
+      "step": 1550
+    },
+    {
+      "epoch": 0.5293515358361774,
+      "grad_norm": 3.7122843265533447,
+      "learning_rate": 0.0008235494880546075,
+      "loss": 6.3315,
+      "step": 1551
+    },
+    {
+      "epoch": 0.5296928327645051,
+      "grad_norm": 3.8710949420928955,
+      "learning_rate": 0.0008234357224118316,
+      "loss": 6.4841,
+      "step": 1552
+    },
+    {
+      "epoch": 0.5300341296928328,
+      "grad_norm": 3.440357208251953,
+      "learning_rate": 0.0008233219567690557,
+      "loss": 6.9585,
+      "step": 1553
+    },
+    {
+      "epoch": 0.5303754266211604,
+      "grad_norm": 3.3393759727478027,
+      "learning_rate": 0.0008232081911262799,
+      "loss": 6.8411,
+      "step": 1554
+    },
+    {
+      "epoch": 0.5307167235494881,
+      "grad_norm": 3.5455715656280518,
+      "learning_rate": 0.000823094425483504,
+      "loss": 6.6823,
+      "step": 1555
+    },
+    {
+      "epoch": 0.5310580204778157,
+      "grad_norm": 3.815316677093506,
+      "learning_rate": 0.0008229806598407281,
+      "loss": 6.1607,
+      "step": 1556
+    },
+    {
+      "epoch": 0.5313993174061433,
+      "grad_norm": 3.953657388687134,
+      "learning_rate": 0.0008228668941979522,
+      "loss": 6.6442,
+      "step": 1557
+    },
+    {
+      "epoch": 0.531740614334471,
+      "grad_norm": 3.4265329837799072,
+      "learning_rate": 0.0008227531285551763,
+      "loss": 6.6399,
+      "step": 1558
+    },
+    {
+      "epoch": 0.5320819112627987,
+      "grad_norm": 3.6078526973724365,
+      "learning_rate": 0.0008226393629124004,
+      "loss": 6.654,
+      "step": 1559
+    },
+    {
+      "epoch": 0.5324232081911263,
+      "grad_norm": 3.5532915592193604,
+      "learning_rate": 0.0008225255972696247,
+      "loss": 6.0045,
+      "step": 1560
+    },
+    {
+      "epoch": 0.532764505119454,
+      "grad_norm": 3.4579410552978516,
+      "learning_rate": 0.0008224118316268488,
+      "loss": 6.6127,
+      "step": 1561
+    },
+    {
+      "epoch": 0.5331058020477816,
+      "grad_norm": 3.489680051803589,
+      "learning_rate": 0.0008222980659840729,
+      "loss": 7.2289,
+      "step": 1562
+    },
+    {
+      "epoch": 0.5334470989761092,
+      "grad_norm": 3.981506824493408,
+      "learning_rate": 0.000822184300341297,
+      "loss": 6.8315,
+      "step": 1563
+    },
+    {
+      "epoch": 0.5337883959044368,
+      "grad_norm": 3.5434958934783936,
+      "learning_rate": 0.0008220705346985211,
+      "loss": 6.4467,
+      "step": 1564
+    },
+    {
+      "epoch": 0.5341296928327645,
+      "grad_norm": 20.380250930786133,
+      "learning_rate": 0.0008219567690557452,
+      "loss": 5.3638,
+      "step": 1565
+    },
+    {
+      "epoch": 0.5344709897610922,
+      "grad_norm": 3.482201337814331,
+      "learning_rate": 0.0008218430034129693,
+      "loss": 6.8673,
+      "step": 1566
+    },
+    {
+      "epoch": 0.5348122866894198,
+      "grad_norm": 3.6969149112701416,
+      "learning_rate": 0.0008217292377701934,
+      "loss": 6.7083,
+      "step": 1567
+    },
+    {
+      "epoch": 0.5351535836177475,
+      "grad_norm": 3.4104528427124023,
+      "learning_rate": 0.0008216154721274175,
+      "loss": 6.7722,
+      "step": 1568
+    },
+    {
+      "epoch": 0.5354948805460751,
+      "grad_norm": 3.611492872238159,
+      "learning_rate": 0.0008215017064846416,
+      "loss": 6.3124,
+      "step": 1569
+    },
+    {
+      "epoch": 0.5358361774744027,
+      "grad_norm": 16.186843872070312,
+      "learning_rate": 0.0008213879408418657,
+      "loss": 8.7558,
+      "step": 1570
+    },
+    {
+      "epoch": 0.5361774744027303,
+      "grad_norm": 3.5169475078582764,
+      "learning_rate": 0.0008212741751990899,
+      "loss": 7.0439,
+      "step": 1571
+    },
+    {
+      "epoch": 0.5365187713310581,
+      "grad_norm": 3.61405611038208,
+      "learning_rate": 0.000821160409556314,
+      "loss": 6.2176,
+      "step": 1572
+    },
+    {
+      "epoch": 0.5368600682593857,
+      "grad_norm": 3.3224093914031982,
+      "learning_rate": 0.0008210466439135381,
+      "loss": 6.693,
+      "step": 1573
+    },
+    {
+      "epoch": 0.5372013651877133,
+      "grad_norm": 3.339376211166382,
+      "learning_rate": 0.0008209328782707622,
+      "loss": 6.3451,
+      "step": 1574
+    },
+    {
+      "epoch": 0.537542662116041,
+      "grad_norm": 3.3219499588012695,
+      "learning_rate": 0.0008208191126279863,
+      "loss": 6.9464,
+      "step": 1575
+    },
+    {
+      "epoch": 0.5378839590443686,
+      "grad_norm": 3.3583548069000244,
+      "learning_rate": 0.0008207053469852104,
+      "loss": 6.9545,
+      "step": 1576
+    },
+    {
+      "epoch": 0.5382252559726962,
+      "grad_norm": 3.8789596557617188,
+      "learning_rate": 0.0008205915813424347,
+      "loss": 6.5906,
+      "step": 1577
+    },
+    {
+      "epoch": 0.5385665529010238,
+      "grad_norm": 3.4868991374969482,
+      "learning_rate": 0.0008204778156996588,
+      "loss": 6.8179,
+      "step": 1578
+    },
+    {
+      "epoch": 0.5389078498293516,
+      "grad_norm": 3.5278048515319824,
+      "learning_rate": 0.0008203640500568829,
+      "loss": 6.8537,
+      "step": 1579
+    },
+    {
+      "epoch": 0.5392491467576792,
+      "grad_norm": 3.1563076972961426,
+      "learning_rate": 0.000820250284414107,
+      "loss": 6.7337,
+      "step": 1580
+    },
+    {
+      "epoch": 0.5395904436860068,
+      "grad_norm": 3.347745418548584,
+      "learning_rate": 0.0008201365187713311,
+      "loss": 6.8519,
+      "step": 1581
+    },
+    {
+      "epoch": 0.5399317406143345,
+      "grad_norm": 3.3623850345611572,
+      "learning_rate": 0.0008200227531285552,
+      "loss": 6.7128,
+      "step": 1582
+    },
+    {
+      "epoch": 0.5402730375426621,
+      "grad_norm": 3.4226489067077637,
+      "learning_rate": 0.0008199089874857794,
+      "loss": 6.9467,
+      "step": 1583
+    },
+    {
+      "epoch": 0.5406143344709897,
+      "grad_norm": 3.467519521713257,
+      "learning_rate": 0.0008197952218430035,
+      "loss": 7.0993,
+      "step": 1584
+    },
+    {
+      "epoch": 0.5409556313993175,
+      "grad_norm": 4.463557720184326,
+      "learning_rate": 0.0008196814562002275,
+      "loss": 6.207,
+      "step": 1585
+    },
+    {
+      "epoch": 0.5412969283276451,
+      "grad_norm": 7.1371846199035645,
+      "learning_rate": 0.0008195676905574516,
+      "loss": 6.115,
+      "step": 1586
+    },
+    {
+      "epoch": 0.5416382252559727,
+      "grad_norm": 4.31298828125,
+      "learning_rate": 0.0008194539249146757,
+      "loss": 6.4027,
+      "step": 1587
+    },
+    {
+      "epoch": 0.5419795221843003,
+      "grad_norm": 3.982283115386963,
+      "learning_rate": 0.0008193401592718998,
+      "loss": 6.9022,
+      "step": 1588
+    },
+    {
+      "epoch": 0.542320819112628,
+      "grad_norm": 3.6711089611053467,
+      "learning_rate": 0.000819226393629124,
+      "loss": 6.7532,
+      "step": 1589
+    },
+    {
+      "epoch": 0.5426621160409556,
+      "grad_norm": 3.598482370376587,
+      "learning_rate": 0.0008191126279863481,
+      "loss": 6.0211,
+      "step": 1590
+    },
+    {
+      "epoch": 0.5430034129692832,
+      "grad_norm": 3.518918514251709,
+      "learning_rate": 0.0008189988623435722,
+      "loss": 6.7096,
+      "step": 1591
+    },
+    {
+      "epoch": 0.543344709897611,
+      "grad_norm": 3.3870673179626465,
+      "learning_rate": 0.0008188850967007963,
+      "loss": 6.8977,
+      "step": 1592
+    },
+    {
+      "epoch": 0.5436860068259386,
+      "grad_norm": 3.3564047813415527,
+      "learning_rate": 0.0008187713310580204,
+      "loss": 6.5459,
+      "step": 1593
+    },
+    {
+      "epoch": 0.5440273037542662,
+      "grad_norm": 3.4745521545410156,
+      "learning_rate": 0.0008186575654152447,
+      "loss": 6.3275,
+      "step": 1594
+    },
+    {
+      "epoch": 0.5443686006825939,
+      "grad_norm": 3.4397785663604736,
+      "learning_rate": 0.0008185437997724688,
+      "loss": 6.644,
+      "step": 1595
+    },
+    {
+      "epoch": 0.5447098976109215,
+      "grad_norm": 3.6341989040374756,
+      "learning_rate": 0.0008184300341296929,
+      "loss": 6.3724,
+      "step": 1596
+    },
+    {
+      "epoch": 0.5450511945392491,
+      "grad_norm": 3.4546432495117188,
+      "learning_rate": 0.000818316268486917,
+      "loss": 6.0687,
+      "step": 1597
+    },
+    {
+      "epoch": 0.5453924914675768,
+      "grad_norm": 3.2902228832244873,
+      "learning_rate": 0.0008182025028441411,
+      "loss": 6.7731,
+      "step": 1598
+    },
+    {
+      "epoch": 0.5457337883959045,
+      "grad_norm": 21.545167922973633,
+      "learning_rate": 0.0008180887372013652,
+      "loss": 9.0693,
+      "step": 1599
+    },
+    {
+      "epoch": 0.5460750853242321,
+      "grad_norm": 4.197141170501709,
+      "learning_rate": 0.0008179749715585894,
+      "loss": 6.2883,
+      "step": 1600
+    },
+    {
+      "epoch": 0.5464163822525597,
+      "grad_norm": 3.763883113861084,
+      "learning_rate": 0.0008178612059158135,
+      "loss": 6.7896,
+      "step": 1601
+    },
+    {
+      "epoch": 0.5467576791808874,
+      "grad_norm": 3.678478717803955,
+      "learning_rate": 0.0008177474402730376,
+      "loss": 6.6968,
+      "step": 1602
+    },
+    {
+      "epoch": 0.547098976109215,
+      "grad_norm": 4.786699295043945,
+      "learning_rate": 0.0008176336746302617,
+      "loss": 6.342,
+      "step": 1603
+    },
+    {
+      "epoch": 0.5474402730375426,
+      "grad_norm": 2.2955498695373535,
+      "learning_rate": 0.0008175199089874858,
+      "loss": 3.6378,
+      "step": 1604
+    },
+    {
+      "epoch": 0.5477815699658704,
+      "grad_norm": 3.439753293991089,
+      "learning_rate": 0.0008174061433447098,
+      "loss": 6.7798,
+      "step": 1605
+    },
+    {
+      "epoch": 0.548122866894198,
+      "grad_norm": 4.1457839012146,
+      "learning_rate": 0.000817292377701934,
+      "loss": 6.3028,
+      "step": 1606
+    },
+    {
+      "epoch": 0.5484641638225256,
+      "grad_norm": 3.4417805671691895,
+      "learning_rate": 0.0008171786120591581,
+      "loss": 7.0572,
+      "step": 1607
+    },
+    {
+      "epoch": 0.5488054607508532,
+      "grad_norm": 3.348179578781128,
+      "learning_rate": 0.0008170648464163822,
+      "loss": 6.7471,
+      "step": 1608
+    },
+    {
+      "epoch": 0.5491467576791809,
+      "grad_norm": 3.4743990898132324,
+      "learning_rate": 0.0008169510807736063,
+      "loss": 7.0257,
+      "step": 1609
+    },
+    {
+      "epoch": 0.5494880546075085,
+      "grad_norm": 3.894490957260132,
+      "learning_rate": 0.0008168373151308304,
+      "loss": 6.1222,
+      "step": 1610
+    },
+    {
+      "epoch": 0.5498293515358362,
+      "grad_norm": 3.6673929691314697,
+      "learning_rate": 0.0008167235494880547,
+      "loss": 6.5822,
+      "step": 1611
+    },
+    {
+      "epoch": 0.5501706484641639,
+      "grad_norm": 3.5163426399230957,
+      "learning_rate": 0.0008166097838452788,
+      "loss": 6.7198,
+      "step": 1612
+    },
+    {
+      "epoch": 0.5505119453924915,
+      "grad_norm": 4.025699138641357,
+      "learning_rate": 0.0008164960182025029,
+      "loss": 6.3125,
+      "step": 1613
+    },
+    {
+      "epoch": 0.5508532423208191,
+      "grad_norm": 4.222406387329102,
+      "learning_rate": 0.000816382252559727,
+      "loss": 7.1646,
+      "step": 1614
+    },
+    {
+      "epoch": 0.5511945392491467,
+      "grad_norm": 6.100660800933838,
+      "learning_rate": 0.0008162684869169511,
+      "loss": 6.3055,
+      "step": 1615
+    },
+    {
+      "epoch": 0.5515358361774744,
+      "grad_norm": 3.6640098094940186,
+      "learning_rate": 0.0008161547212741752,
+      "loss": 6.3726,
+      "step": 1616
+    },
+    {
+      "epoch": 0.551877133105802,
+      "grad_norm": 3.549154281616211,
+      "learning_rate": 0.0008160409556313994,
+      "loss": 6.8686,
+      "step": 1617
+    },
+    {
+      "epoch": 0.5522184300341297,
+      "grad_norm": 3.3445968627929688,
+      "learning_rate": 0.0008159271899886235,
+      "loss": 6.7752,
+      "step": 1618
+    },
+    {
+      "epoch": 0.5525597269624574,
+      "grad_norm": 3.3909804821014404,
+      "learning_rate": 0.0008158134243458476,
+      "loss": 6.7392,
+      "step": 1619
+    },
+    {
+      "epoch": 0.552901023890785,
+      "grad_norm": 3.3068623542785645,
+      "learning_rate": 0.0008156996587030717,
+      "loss": 7.0023,
+      "step": 1620
+    },
+    {
+      "epoch": 0.5532423208191126,
+      "grad_norm": 5.286759376525879,
+      "learning_rate": 0.0008155858930602958,
+      "loss": 5.7817,
+      "step": 1621
+    },
+    {
+      "epoch": 0.5535836177474402,
+      "grad_norm": 4.21416711807251,
+      "learning_rate": 0.0008154721274175199,
+      "loss": 6.6314,
+      "step": 1622
+    },
+    {
+      "epoch": 0.5539249146757679,
+      "grad_norm": 3.586740016937256,
+      "learning_rate": 0.0008153583617747441,
+      "loss": 6.7685,
+      "step": 1623
+    },
+    {
+      "epoch": 0.5542662116040956,
+      "grad_norm": 3.7087645530700684,
+      "learning_rate": 0.0008152445961319681,
+      "loss": 5.8739,
+      "step": 1624
+    },
+    {
+      "epoch": 0.5546075085324232,
+      "grad_norm": 3.9392848014831543,
+      "learning_rate": 0.0008151308304891922,
+      "loss": 6.6611,
+      "step": 1625
+    },
+    {
+      "epoch": 0.5549488054607509,
+      "grad_norm": 3.521493673324585,
+      "learning_rate": 0.0008150170648464163,
+      "loss": 7.1034,
+      "step": 1626
+    },
+    {
+      "epoch": 0.5552901023890785,
+      "grad_norm": 4.224700927734375,
+      "learning_rate": 0.0008149032992036404,
+      "loss": 5.8669,
+      "step": 1627
+    },
+    {
+      "epoch": 0.5556313993174061,
+      "grad_norm": 3.757660388946533,
+      "learning_rate": 0.0008147895335608646,
+      "loss": 6.5412,
+      "step": 1628
+    },
+    {
+      "epoch": 0.5559726962457338,
+      "grad_norm": 3.58388614654541,
+      "learning_rate": 0.0008146757679180888,
+      "loss": 6.7925,
+      "step": 1629
+    },
+    {
+      "epoch": 0.5563139931740614,
+      "grad_norm": 3.373109817504883,
+      "learning_rate": 0.0008145620022753129,
+      "loss": 6.5338,
+      "step": 1630
+    },
+    {
+      "epoch": 0.5566552901023891,
+      "grad_norm": 3.3776378631591797,
+      "learning_rate": 0.000814448236632537,
+      "loss": 6.6037,
+      "step": 1631
+    },
+    {
+      "epoch": 0.5569965870307167,
+      "grad_norm": 3.9116930961608887,
+      "learning_rate": 0.0008143344709897611,
+      "loss": 5.6358,
+      "step": 1632
+    },
+    {
+      "epoch": 0.5573378839590444,
+      "grad_norm": 3.494377851486206,
+      "learning_rate": 0.0008142207053469852,
+      "loss": 6.9675,
+      "step": 1633
+    },
+    {
+      "epoch": 0.557679180887372,
+      "grad_norm": 3.6268115043640137,
+      "learning_rate": 0.0008141069397042094,
+      "loss": 6.7061,
+      "step": 1634
+    },
+    {
+      "epoch": 0.5580204778156996,
+      "grad_norm": 4.0463409423828125,
+      "learning_rate": 0.0008139931740614335,
+      "loss": 5.8411,
+      "step": 1635
+    },
+    {
+      "epoch": 0.5583617747440273,
+      "grad_norm": 3.519866466522217,
+      "learning_rate": 0.0008138794084186576,
+      "loss": 6.6951,
+      "step": 1636
+    },
+    {
+      "epoch": 0.558703071672355,
+      "grad_norm": 3.481750965118408,
+      "learning_rate": 0.0008137656427758817,
+      "loss": 7.011,
+      "step": 1637
+    },
+    {
+      "epoch": 0.5590443686006826,
+      "grad_norm": 3.268798828125,
+      "learning_rate": 0.0008136518771331058,
+      "loss": 6.7461,
+      "step": 1638
+    },
+    {
+      "epoch": 0.5593856655290103,
+      "grad_norm": 3.8919827938079834,
+      "learning_rate": 0.0008135381114903299,
+      "loss": 6.4497,
+      "step": 1639
+    },
+    {
+      "epoch": 0.5597269624573379,
+      "grad_norm": 3.408388614654541,
+      "learning_rate": 0.0008134243458475541,
+      "loss": 7.0488,
+      "step": 1640
+    },
+    {
+      "epoch": 0.5600682593856655,
+      "grad_norm": 3.337852954864502,
+      "learning_rate": 0.0008133105802047782,
+      "loss": 6.6489,
+      "step": 1641
+    },
+    {
+      "epoch": 0.5604095563139931,
+      "grad_norm": 3.176896095275879,
+      "learning_rate": 0.0008131968145620023,
+      "loss": 7.0354,
+      "step": 1642
+    },
+    {
+      "epoch": 0.5607508532423208,
+      "grad_norm": 3.4171290397644043,
+      "learning_rate": 0.0008130830489192265,
+      "loss": 6.6106,
+      "step": 1643
+    },
+    {
+      "epoch": 0.5610921501706485,
+      "grad_norm": 3.348039150238037,
+      "learning_rate": 0.0008129692832764504,
+      "loss": 6.4115,
+      "step": 1644
+    },
+    {
+      "epoch": 0.5614334470989761,
+      "grad_norm": 3.897597074508667,
+      "learning_rate": 0.0008128555176336746,
+      "loss": 6.5893,
+      "step": 1645
+    },
+    {
+      "epoch": 0.5617747440273038,
+      "grad_norm": 3.424443006515503,
+      "learning_rate": 0.0008127417519908988,
+      "loss": 6.4034,
+      "step": 1646
+    },
+    {
+      "epoch": 0.5621160409556314,
+      "grad_norm": 3.577786445617676,
+      "learning_rate": 0.0008126279863481229,
+      "loss": 6.916,
+      "step": 1647
+    },
+    {
+      "epoch": 0.562457337883959,
+      "grad_norm": 3.442438840866089,
+      "learning_rate": 0.000812514220705347,
+      "loss": 7.1346,
+      "step": 1648
+    },
+    {
+      "epoch": 0.5627986348122866,
+      "grad_norm": 3.3798537254333496,
+      "learning_rate": 0.0008124004550625711,
+      "loss": 6.4942,
+      "step": 1649
+    },
+    {
+      "epoch": 0.5631399317406144,
+      "grad_norm": 3.885824680328369,
+      "learning_rate": 0.0008122866894197952,
+      "loss": 6.8912,
+      "step": 1650
+    },
+    {
+      "epoch": 0.563481228668942,
+      "grad_norm": 3.910099744796753,
+      "learning_rate": 0.0008121729237770193,
+      "loss": 6.3676,
+      "step": 1651
+    },
+    {
+      "epoch": 0.5638225255972696,
+      "grad_norm": 3.5344743728637695,
+      "learning_rate": 0.0008120591581342435,
+      "loss": 6.8041,
+      "step": 1652
+    },
+    {
+      "epoch": 0.5641638225255973,
+      "grad_norm": 7.247477054595947,
+      "learning_rate": 0.0008119453924914676,
+      "loss": 5.6654,
+      "step": 1653
+    },
+    {
+      "epoch": 0.5645051194539249,
+      "grad_norm": 3.664292097091675,
+      "learning_rate": 0.0008118316268486917,
+      "loss": 7.0846,
+      "step": 1654
+    },
+    {
+      "epoch": 0.5648464163822525,
+      "grad_norm": 5.331287384033203,
+      "learning_rate": 0.0008117178612059158,
+      "loss": 5.1035,
+      "step": 1655
+    },
+    {
+      "epoch": 0.5651877133105802,
+      "grad_norm": 4.2963175773620605,
+      "learning_rate": 0.0008116040955631399,
+      "loss": 6.543,
+      "step": 1656
+    },
+    {
+      "epoch": 0.5655290102389079,
+      "grad_norm": 3.7694435119628906,
+      "learning_rate": 0.0008114903299203641,
+      "loss": 6.5782,
+      "step": 1657
+    },
+    {
+      "epoch": 0.5658703071672355,
+      "grad_norm": 3.3305928707122803,
+      "learning_rate": 0.0008113765642775882,
+      "loss": 6.992,
+      "step": 1658
+    },
+    {
+      "epoch": 0.5662116040955631,
+      "grad_norm": 3.3793208599090576,
+      "learning_rate": 0.0008112627986348123,
+      "loss": 6.3376,
+      "step": 1659
+    },
+    {
+      "epoch": 0.5665529010238908,
+      "grad_norm": 3.6285388469696045,
+      "learning_rate": 0.0008111490329920365,
+      "loss": 6.6463,
+      "step": 1660
+    },
+    {
+      "epoch": 0.5668941979522184,
+      "grad_norm": 3.3783798217773438,
+      "learning_rate": 0.0008110352673492606,
+      "loss": 6.4926,
+      "step": 1661
+    },
+    {
+      "epoch": 0.567235494880546,
+      "grad_norm": 3.489332675933838,
+      "learning_rate": 0.0008109215017064847,
+      "loss": 6.9067,
+      "step": 1662
+    },
+    {
+      "epoch": 0.5675767918088738,
+      "grad_norm": 3.4197375774383545,
+      "learning_rate": 0.0008108077360637088,
+      "loss": 6.8896,
+      "step": 1663
+    },
+    {
+      "epoch": 0.5679180887372014,
+      "grad_norm": 5.391114234924316,
+      "learning_rate": 0.0008106939704209329,
+      "loss": 6.6774,
+      "step": 1664
+    },
+    {
+      "epoch": 0.568259385665529,
+      "grad_norm": 3.594270706176758,
+      "learning_rate": 0.000810580204778157,
+      "loss": 7.0118,
+      "step": 1665
+    },
+    {
+      "epoch": 0.5686006825938567,
+      "grad_norm": 3.6209936141967773,
+      "learning_rate": 0.0008104664391353811,
+      "loss": 5.8403,
+      "step": 1666
+    },
+    {
+      "epoch": 0.5689419795221843,
+      "grad_norm": 3.4565608501434326,
+      "learning_rate": 0.0008103526734926052,
+      "loss": 7.2969,
+      "step": 1667
+    },
+    {
+      "epoch": 0.5692832764505119,
+      "grad_norm": 3.351238489151001,
+      "learning_rate": 0.0008102389078498293,
+      "loss": 7.1469,
+      "step": 1668
+    },
+    {
+      "epoch": 0.5696245733788395,
+      "grad_norm": 3.340491771697998,
+      "learning_rate": 0.0008101251422070535,
+      "loss": 6.5831,
+      "step": 1669
+    },
+    {
+      "epoch": 0.5699658703071673,
+      "grad_norm": 3.3697030544281006,
+      "learning_rate": 0.0008100113765642776,
+      "loss": 6.5447,
+      "step": 1670
+    },
+    {
+      "epoch": 0.5703071672354949,
+      "grad_norm": 3.3213913440704346,
+      "learning_rate": 0.0008098976109215017,
+      "loss": 6.7253,
+      "step": 1671
+    },
+    {
+      "epoch": 0.5706484641638225,
+      "grad_norm": 3.31199312210083,
+      "learning_rate": 0.0008097838452787258,
+      "loss": 6.7037,
+      "step": 1672
+    },
+    {
+      "epoch": 0.5709897610921502,
+      "grad_norm": 3.239823341369629,
+      "learning_rate": 0.0008096700796359499,
+      "loss": 6.2837,
+      "step": 1673
+    },
+    {
+      "epoch": 0.5713310580204778,
+      "grad_norm": 3.233081102371216,
+      "learning_rate": 0.0008095563139931741,
+      "loss": 6.4038,
+      "step": 1674
+    },
+    {
+      "epoch": 0.5716723549488054,
+      "grad_norm": 3.2402610778808594,
+      "learning_rate": 0.0008094425483503982,
+      "loss": 6.8486,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5720136518771332,
+      "grad_norm": 3.5634925365448,
+      "learning_rate": 0.0008093287827076223,
+      "loss": 5.8242,
+      "step": 1676
+    },
+    {
+      "epoch": 0.5723549488054608,
+      "grad_norm": 3.4806721210479736,
+      "learning_rate": 0.0008092150170648465,
+      "loss": 7.0927,
+      "step": 1677
+    },
+    {
+      "epoch": 0.5726962457337884,
+      "grad_norm": 3.4350132942199707,
+      "learning_rate": 0.0008091012514220706,
+      "loss": 6.6534,
+      "step": 1678
+    },
+    {
+      "epoch": 0.573037542662116,
+      "grad_norm": 3.3814663887023926,
+      "learning_rate": 0.0008089874857792947,
+      "loss": 6.8434,
+      "step": 1679
+    },
+    {
+      "epoch": 0.5733788395904437,
+      "grad_norm": 3.420792579650879,
+      "learning_rate": 0.0008088737201365189,
+      "loss": 6.8301,
+      "step": 1680
+    },
+    {
+      "epoch": 0.5737201365187713,
+      "grad_norm": 3.6617331504821777,
+      "learning_rate": 0.000808759954493743,
+      "loss": 7.1931,
+      "step": 1681
+    },
+    {
+      "epoch": 0.5740614334470989,
+      "grad_norm": 3.417377233505249,
+      "learning_rate": 0.0008086461888509671,
+      "loss": 6.8856,
+      "step": 1682
+    },
+    {
+      "epoch": 0.5744027303754267,
+      "grad_norm": 3.4047882556915283,
+      "learning_rate": 0.0008085324232081911,
+      "loss": 6.7193,
+      "step": 1683
+    },
+    {
+      "epoch": 0.5747440273037543,
+      "grad_norm": 5.160583019256592,
+      "learning_rate": 0.0008084186575654152,
+      "loss": 6.3314,
+      "step": 1684
+    },
+    {
+      "epoch": 0.5750853242320819,
+      "grad_norm": 3.442270517349243,
+      "learning_rate": 0.0008083048919226393,
+      "loss": 6.5523,
+      "step": 1685
+    },
+    {
+      "epoch": 0.5754266211604095,
+      "grad_norm": 3.4969444274902344,
+      "learning_rate": 0.0008081911262798635,
+      "loss": 6.1442,
+      "step": 1686
+    },
+    {
+      "epoch": 0.5757679180887372,
+      "grad_norm": 3.283874988555908,
+      "learning_rate": 0.0008080773606370876,
+      "loss": 6.5447,
+      "step": 1687
+    },
+    {
+      "epoch": 0.5761092150170648,
+      "grad_norm": 3.373081922531128,
+      "learning_rate": 0.0008079635949943117,
+      "loss": 6.9056,
+      "step": 1688
+    },
+    {
+      "epoch": 0.5764505119453925,
+      "grad_norm": 3.2919609546661377,
+      "learning_rate": 0.0008078498293515358,
+      "loss": 6.6918,
+      "step": 1689
+    },
+    {
+      "epoch": 0.5767918088737202,
+      "grad_norm": 3.302234411239624,
+      "learning_rate": 0.0008077360637087599,
+      "loss": 6.5728,
+      "step": 1690
+    },
+    {
+      "epoch": 0.5771331058020478,
+      "grad_norm": 3.6976983547210693,
+      "learning_rate": 0.000807622298065984,
+      "loss": 6.1227,
+      "step": 1691
+    },
+    {
+      "epoch": 0.5774744027303754,
+      "grad_norm": 3.5474138259887695,
+      "learning_rate": 0.0008075085324232082,
+      "loss": 6.2463,
+      "step": 1692
+    },
+    {
+      "epoch": 0.577815699658703,
+      "grad_norm": 3.680959701538086,
+      "learning_rate": 0.0008073947667804323,
+      "loss": 6.6472,
+      "step": 1693
+    },
+    {
+      "epoch": 0.5781569965870307,
+      "grad_norm": 3.4441518783569336,
+      "learning_rate": 0.0008072810011376565,
+      "loss": 6.5921,
+      "step": 1694
+    },
+    {
+      "epoch": 0.5784982935153583,
+      "grad_norm": 4.723526477813721,
+      "learning_rate": 0.0008071672354948806,
+      "loss": 6.2168,
+      "step": 1695
+    },
+    {
+      "epoch": 0.578839590443686,
+      "grad_norm": 3.377566337585449,
+      "learning_rate": 0.0008070534698521047,
+      "loss": 6.7912,
+      "step": 1696
+    },
+    {
+      "epoch": 0.5791808873720137,
+      "grad_norm": 3.3538033962249756,
+      "learning_rate": 0.0008069397042093289,
+      "loss": 6.7811,
+      "step": 1697
+    },
+    {
+      "epoch": 0.5795221843003413,
+      "grad_norm": 3.2516143321990967,
+      "learning_rate": 0.000806825938566553,
+      "loss": 6.6028,
+      "step": 1698
+    },
+    {
+      "epoch": 0.5798634812286689,
+      "grad_norm": 3.5221121311187744,
+      "learning_rate": 0.0008067121729237771,
+      "loss": 6.8532,
+      "step": 1699
+    },
+    {
+      "epoch": 0.5802047781569966,
+      "grad_norm": 3.5948259830474854,
+      "learning_rate": 0.0008065984072810012,
+      "loss": 6.5451,
+      "step": 1700
+    },
+    {
+      "epoch": 0.5805460750853242,
+      "grad_norm": 3.5543367862701416,
+      "learning_rate": 0.0008064846416382253,
+      "loss": 6.5479,
+      "step": 1701
+    },
+    {
+      "epoch": 0.5808873720136519,
+      "grad_norm": 3.5869266986846924,
+      "learning_rate": 0.0008063708759954493,
+      "loss": 6.4638,
+      "step": 1702
+    },
+    {
+      "epoch": 0.5812286689419796,
+      "grad_norm": 3.233880043029785,
+      "learning_rate": 0.0008062571103526735,
+      "loss": 6.6942,
+      "step": 1703
+    },
+    {
+      "epoch": 0.5815699658703072,
+      "grad_norm": 3.558081865310669,
+      "learning_rate": 0.0008061433447098976,
+      "loss": 6.5371,
+      "step": 1704
+    },
+    {
+      "epoch": 0.5819112627986348,
+      "grad_norm": 3.4222469329833984,
+      "learning_rate": 0.0008060295790671217,
+      "loss": 6.667,
+      "step": 1705
+    },
+    {
+      "epoch": 0.5822525597269624,
+      "grad_norm": 3.447845697402954,
+      "learning_rate": 0.0008059158134243458,
+      "loss": 6.8876,
+      "step": 1706
+    },
+    {
+      "epoch": 0.5825938566552901,
+      "grad_norm": 5.436509132385254,
+      "learning_rate": 0.0008058020477815699,
+      "loss": 6.4207,
+      "step": 1707
+    },
+    {
+      "epoch": 0.5829351535836177,
+      "grad_norm": 3.4778072834014893,
+      "learning_rate": 0.000805688282138794,
+      "loss": 6.5555,
+      "step": 1708
+    },
+    {
+      "epoch": 0.5832764505119454,
+      "grad_norm": 3.7661755084991455,
+      "learning_rate": 0.0008055745164960182,
+      "loss": 6.6308,
+      "step": 1709
+    },
+    {
+      "epoch": 0.5836177474402731,
+      "grad_norm": 3.373845100402832,
+      "learning_rate": 0.0008054607508532424,
+      "loss": 7.044,
+      "step": 1710
+    },
+    {
+      "epoch": 0.5839590443686007,
+      "grad_norm": 3.3945181369781494,
+      "learning_rate": 0.0008053469852104665,
+      "loss": 6.8728,
+      "step": 1711
+    },
+    {
+      "epoch": 0.5843003412969283,
+      "grad_norm": 3.244920015335083,
+      "learning_rate": 0.0008052332195676906,
+      "loss": 6.4727,
+      "step": 1712
+    },
+    {
+      "epoch": 0.5846416382252559,
+      "grad_norm": 3.3084375858306885,
+      "learning_rate": 0.0008051194539249147,
+      "loss": 6.6563,
+      "step": 1713
+    },
+    {
+      "epoch": 0.5849829351535836,
+      "grad_norm": 3.3623743057250977,
+      "learning_rate": 0.0008050056882821389,
+      "loss": 6.8908,
+      "step": 1714
+    },
+    {
+      "epoch": 0.5853242320819113,
+      "grad_norm": 3.3226821422576904,
+      "learning_rate": 0.000804891922639363,
+      "loss": 6.7584,
+      "step": 1715
+    },
+    {
+      "epoch": 0.5856655290102389,
+      "grad_norm": 3.3474953174591064,
+      "learning_rate": 0.0008047781569965871,
+      "loss": 7.0349,
+      "step": 1716
+    },
+    {
+      "epoch": 0.5860068259385666,
+      "grad_norm": 3.4366891384124756,
+      "learning_rate": 0.0008046643913538112,
+      "loss": 6.7988,
+      "step": 1717
+    },
+    {
+      "epoch": 0.5863481228668942,
+      "grad_norm": 3.3742077350616455,
+      "learning_rate": 0.0008045506257110353,
+      "loss": 6.3303,
+      "step": 1718
+    },
+    {
+      "epoch": 0.5866894197952218,
+      "grad_norm": 3.367166042327881,
+      "learning_rate": 0.0008044368600682594,
+      "loss": 6.9229,
+      "step": 1719
+    },
+    {
+      "epoch": 0.5870307167235495,
+      "grad_norm": 3.3801698684692383,
+      "learning_rate": 0.0008043230944254836,
+      "loss": 6.4958,
+      "step": 1720
+    },
+    {
+      "epoch": 0.5873720136518771,
+      "grad_norm": 3.513958692550659,
+      "learning_rate": 0.0008042093287827077,
+      "loss": 6.4753,
+      "step": 1721
+    },
+    {
+      "epoch": 0.5877133105802048,
+      "grad_norm": 3.514188051223755,
+      "learning_rate": 0.0008040955631399317,
+      "loss": 6.6555,
+      "step": 1722
+    },
+    {
+      "epoch": 0.5880546075085324,
+      "grad_norm": 3.736156702041626,
+      "learning_rate": 0.0008039817974971558,
+      "loss": 6.2383,
+      "step": 1723
+    },
+    {
+      "epoch": 0.5883959044368601,
+      "grad_norm": 3.569303274154663,
+      "learning_rate": 0.0008038680318543799,
+      "loss": 6.6994,
+      "step": 1724
+    },
+    {
+      "epoch": 0.5887372013651877,
+      "grad_norm": 3.693387031555176,
+      "learning_rate": 0.000803754266211604,
+      "loss": 7.2297,
+      "step": 1725
+    },
+    {
+      "epoch": 0.5890784982935153,
+      "grad_norm": 3.4730892181396484,
+      "learning_rate": 0.0008036405005688282,
+      "loss": 6.9965,
+      "step": 1726
+    },
+    {
+      "epoch": 0.589419795221843,
+      "grad_norm": 3.8345372676849365,
+      "learning_rate": 0.0008035267349260524,
+      "loss": 6.1234,
+      "step": 1727
+    },
+    {
+      "epoch": 0.5897610921501707,
+      "grad_norm": 3.4538676738739014,
+      "learning_rate": 0.0008034129692832765,
+      "loss": 6.5552,
+      "step": 1728
+    },
+    {
+      "epoch": 0.5901023890784983,
+      "grad_norm": 3.578291654586792,
+      "learning_rate": 0.0008032992036405006,
+      "loss": 7.2901,
+      "step": 1729
+    },
+    {
+      "epoch": 0.590443686006826,
+      "grad_norm": 3.2786262035369873,
+      "learning_rate": 0.0008031854379977247,
+      "loss": 6.9642,
+      "step": 1730
+    },
+    {
+      "epoch": 0.5907849829351536,
+      "grad_norm": 3.3056552410125732,
+      "learning_rate": 0.0008030716723549488,
+      "loss": 6.7187,
+      "step": 1731
+    },
+    {
+      "epoch": 0.5911262798634812,
+      "grad_norm": 3.6989173889160156,
+      "learning_rate": 0.000802957906712173,
+      "loss": 6.7993,
+      "step": 1732
+    },
+    {
+      "epoch": 0.5914675767918088,
+      "grad_norm": 3.3782973289489746,
+      "learning_rate": 0.0008028441410693971,
+      "loss": 6.8576,
+      "step": 1733
+    },
+    {
+      "epoch": 0.5918088737201365,
+      "grad_norm": 3.2711713314056396,
+      "learning_rate": 0.0008027303754266212,
+      "loss": 6.8393,
+      "step": 1734
+    },
+    {
+      "epoch": 0.5921501706484642,
+      "grad_norm": 3.265575885772705,
+      "learning_rate": 0.0008026166097838453,
+      "loss": 6.5095,
+      "step": 1735
+    },
+    {
+      "epoch": 0.5924914675767918,
+      "grad_norm": 3.287745237350464,
+      "learning_rate": 0.0008025028441410694,
+      "loss": 6.8591,
+      "step": 1736
+    },
+    {
+      "epoch": 0.5928327645051195,
+      "grad_norm": 4.958430290222168,
+      "learning_rate": 0.0008023890784982936,
+      "loss": 6.7966,
+      "step": 1737
+    },
+    {
+      "epoch": 0.5931740614334471,
+      "grad_norm": 4.398024082183838,
+      "learning_rate": 0.0008022753128555177,
+      "loss": 5.6212,
+      "step": 1738
+    },
+    {
+      "epoch": 0.5935153583617747,
+      "grad_norm": 3.486335277557373,
+      "learning_rate": 0.0008021615472127418,
+      "loss": 6.9339,
+      "step": 1739
+    },
+    {
+      "epoch": 0.5938566552901023,
+      "grad_norm": 3.2780847549438477,
+      "learning_rate": 0.0008020477815699659,
+      "loss": 6.4291,
+      "step": 1740
+    },
+    {
+      "epoch": 0.5941979522184301,
+      "grad_norm": 8.394953727722168,
+      "learning_rate": 0.0008019340159271899,
+      "loss": 6.1876,
+      "step": 1741
+    },
+    {
+      "epoch": 0.5945392491467577,
+      "grad_norm": 3.4329168796539307,
+      "learning_rate": 0.000801820250284414,
+      "loss": 4.1804,
+      "step": 1742
+    },
+    {
+      "epoch": 0.5948805460750853,
+      "grad_norm": 4.278625011444092,
+      "learning_rate": 0.0008017064846416382,
+      "loss": 6.3612,
+      "step": 1743
+    },
+    {
+      "epoch": 0.595221843003413,
+      "grad_norm": 3.5409955978393555,
+      "learning_rate": 0.0008015927189988624,
+      "loss": 6.4157,
+      "step": 1744
+    },
+    {
+      "epoch": 0.5955631399317406,
+      "grad_norm": 5.1169891357421875,
+      "learning_rate": 0.0008014789533560865,
+      "loss": 5.6104,
+      "step": 1745
+    },
+    {
+      "epoch": 0.5959044368600682,
+      "grad_norm": 3.445643901824951,
+      "learning_rate": 0.0008013651877133106,
+      "loss": 6.0491,
+      "step": 1746
+    },
+    {
+      "epoch": 0.596245733788396,
+      "grad_norm": 3.8756022453308105,
+      "learning_rate": 0.0008012514220705347,
+      "loss": 5.9952,
+      "step": 1747
+    },
+    {
+      "epoch": 0.5965870307167236,
+      "grad_norm": 3.3911144733428955,
+      "learning_rate": 0.0008011376564277588,
+      "loss": 7.2222,
+      "step": 1748
+    },
+    {
+      "epoch": 0.5969283276450512,
+      "grad_norm": 3.516190528869629,
+      "learning_rate": 0.000801023890784983,
+      "loss": 6.4552,
+      "step": 1749
+    },
+    {
+      "epoch": 0.5972696245733788,
+      "grad_norm": 3.3756418228149414,
+      "learning_rate": 0.0008009101251422071,
+      "loss": 6.643,
+      "step": 1750
+    },
+    {
+      "epoch": 0.5976109215017065,
+      "grad_norm": 6.02211856842041,
+      "learning_rate": 0.0008007963594994312,
+      "loss": 6.1575,
+      "step": 1751
+    },
+    {
+      "epoch": 0.5979522184300341,
+      "grad_norm": 3.416879177093506,
+      "learning_rate": 0.0008006825938566553,
+      "loss": 6.7652,
+      "step": 1752
+    },
+    {
+      "epoch": 0.5982935153583617,
+      "grad_norm": 4.541624069213867,
+      "learning_rate": 0.0008005688282138794,
+      "loss": 6.4852,
+      "step": 1753
+    },
+    {
+      "epoch": 0.5986348122866895,
+      "grad_norm": 12.214677810668945,
+      "learning_rate": 0.0008004550625711036,
+      "loss": 6.6869,
+      "step": 1754
+    },
+    {
+      "epoch": 0.5989761092150171,
+      "grad_norm": 3.7498533725738525,
+      "learning_rate": 0.0008003412969283277,
+      "loss": 6.6797,
+      "step": 1755
+    },
+    {
+      "epoch": 0.5993174061433447,
+      "grad_norm": 7.0607523918151855,
+      "learning_rate": 0.0008002275312855518,
+      "loss": 6.0239,
+      "step": 1756
+    },
+    {
+      "epoch": 0.5996587030716723,
+      "grad_norm": 5.875295639038086,
+      "learning_rate": 0.0008001137656427759,
+      "loss": 6.777,
+      "step": 1757
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.4712796211242676,
+      "learning_rate": 0.0008,
+      "loss": 6.5113,
+      "step": 1758
+    },
+    {
+      "epoch": 0.6003412969283276,
+      "grad_norm": 3.289867877960205,
+      "learning_rate": 0.0007998862343572241,
+      "loss": 6.271,
+      "step": 1759
+    },
+    {
+      "epoch": 0.6006825938566553,
+      "grad_norm": 3.281529664993286,
+      "learning_rate": 0.0007997724687144482,
+      "loss": 6.068,
+      "step": 1760
+    },
+    {
+      "epoch": 0.601023890784983,
+      "grad_norm": 3.141273260116577,
+      "learning_rate": 0.0007996587030716724,
+      "loss": 6.6534,
+      "step": 1761
+    },
+    {
+      "epoch": 0.6013651877133106,
+      "grad_norm": 3.538984537124634,
+      "learning_rate": 0.0007995449374288965,
+      "loss": 6.7176,
+      "step": 1762
+    },
+    {
+      "epoch": 0.6017064846416382,
+      "grad_norm": 3.208937644958496,
+      "learning_rate": 0.0007994311717861206,
+      "loss": 6.2085,
+      "step": 1763
+    },
+    {
+      "epoch": 0.6020477815699659,
+      "grad_norm": 3.4167351722717285,
+      "learning_rate": 0.0007993174061433447,
+      "loss": 6.4306,
+      "step": 1764
+    },
+    {
+      "epoch": 0.6023890784982935,
+      "grad_norm": 3.342594861984253,
+      "learning_rate": 0.0007992036405005688,
+      "loss": 6.9312,
+      "step": 1765
+    },
+    {
+      "epoch": 0.6027303754266211,
+      "grad_norm": 3.4137015342712402,
+      "learning_rate": 0.000799089874857793,
+      "loss": 6.6204,
+      "step": 1766
+    },
+    {
+      "epoch": 0.6030716723549489,
+      "grad_norm": 3.446514368057251,
+      "learning_rate": 0.0007989761092150171,
+      "loss": 6.866,
+      "step": 1767
+    },
+    {
+      "epoch": 0.6034129692832765,
+      "grad_norm": 3.3566482067108154,
+      "learning_rate": 0.0007988623435722412,
+      "loss": 6.1879,
+      "step": 1768
+    },
+    {
+      "epoch": 0.6037542662116041,
+      "grad_norm": 3.4932174682617188,
+      "learning_rate": 0.0007987485779294653,
+      "loss": 7.2527,
+      "step": 1769
+    },
+    {
+      "epoch": 0.6040955631399317,
+      "grad_norm": 3.385852098464966,
+      "learning_rate": 0.0007986348122866894,
+      "loss": 7.0579,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6044368600682594,
+      "grad_norm": 3.311905860900879,
+      "learning_rate": 0.0007985210466439135,
+      "loss": 7.0279,
+      "step": 1771
+    },
+    {
+      "epoch": 0.604778156996587,
+      "grad_norm": 3.295506238937378,
+      "learning_rate": 0.0007984072810011377,
+      "loss": 6.3771,
+      "step": 1772
+    },
+    {
+      "epoch": 0.6051194539249147,
+      "grad_norm": 3.4223151206970215,
+      "learning_rate": 0.0007982935153583618,
+      "loss": 6.1649,
+      "step": 1773
+    },
+    {
+      "epoch": 0.6054607508532424,
+      "grad_norm": 3.4446609020233154,
+      "learning_rate": 0.0007981797497155859,
+      "loss": 5.7184,
+      "step": 1774
+    },
+    {
+      "epoch": 0.60580204778157,
+      "grad_norm": 3.3032233715057373,
+      "learning_rate": 0.00079806598407281,
+      "loss": 6.4993,
+      "step": 1775
+    },
+    {
+      "epoch": 0.6061433447098976,
+      "grad_norm": 4.309022426605225,
+      "learning_rate": 0.0007979522184300341,
+      "loss": 5.8551,
+      "step": 1776
+    },
+    {
+      "epoch": 0.6064846416382252,
+      "grad_norm": 5.289666175842285,
+      "learning_rate": 0.0007978384527872584,
+      "loss": 6.1384,
+      "step": 1777
+    },
+    {
+      "epoch": 0.6068259385665529,
+      "grad_norm": 3.589069366455078,
+      "learning_rate": 0.0007977246871444825,
+      "loss": 6.6301,
+      "step": 1778
+    },
+    {
+      "epoch": 0.6071672354948805,
+      "grad_norm": 4.165103435516357,
+      "learning_rate": 0.0007976109215017066,
+      "loss": 6.2209,
+      "step": 1779
+    },
+    {
+      "epoch": 0.6075085324232082,
+      "grad_norm": 3.654651641845703,
+      "learning_rate": 0.0007974971558589306,
+      "loss": 6.8522,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6078498293515359,
+      "grad_norm": 3.5410244464874268,
+      "learning_rate": 0.0007973833902161547,
+      "loss": 6.7159,
+      "step": 1781
+    },
+    {
+      "epoch": 0.6081911262798635,
+      "grad_norm": 3.301004648208618,
+      "learning_rate": 0.0007972696245733788,
+      "loss": 6.7727,
+      "step": 1782
+    },
+    {
+      "epoch": 0.6085324232081911,
+      "grad_norm": 8.049127578735352,
+      "learning_rate": 0.000797155858930603,
+      "loss": 5.6872,
+      "step": 1783
+    },
+    {
+      "epoch": 0.6088737201365187,
+      "grad_norm": 3.7068381309509277,
+      "learning_rate": 0.0007970420932878271,
+      "loss": 6.6709,
+      "step": 1784
+    },
+    {
+      "epoch": 0.6092150170648464,
+      "grad_norm": 3.3772456645965576,
+      "learning_rate": 0.0007969283276450512,
+      "loss": 6.8441,
+      "step": 1785
+    },
+    {
+      "epoch": 0.6095563139931741,
+      "grad_norm": 3.3938705921173096,
+      "learning_rate": 0.0007968145620022753,
+      "loss": 6.7171,
+      "step": 1786
+    },
+    {
+      "epoch": 0.6098976109215017,
+      "grad_norm": 3.674920082092285,
+      "learning_rate": 0.0007967007963594994,
+      "loss": 6.424,
+      "step": 1787
+    },
+    {
+      "epoch": 0.6102389078498294,
+      "grad_norm": 5.99582576751709,
+      "learning_rate": 0.0007965870307167235,
+      "loss": 6.1769,
+      "step": 1788
+    },
+    {
+      "epoch": 0.610580204778157,
+      "grad_norm": 3.5924954414367676,
+      "learning_rate": 0.0007964732650739477,
+      "loss": 6.5125,
+      "step": 1789
+    },
+    {
+      "epoch": 0.6109215017064846,
+      "grad_norm": 3.4204015731811523,
+      "learning_rate": 0.0007963594994311718,
+      "loss": 6.6597,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6112627986348123,
+      "grad_norm": 3.3368844985961914,
+      "learning_rate": 0.0007962457337883959,
+      "loss": 6.2667,
+      "step": 1791
+    },
+    {
+      "epoch": 0.6116040955631399,
+      "grad_norm": 3.3641443252563477,
+      "learning_rate": 0.00079613196814562,
+      "loss": 6.5083,
+      "step": 1792
+    },
+    {
+      "epoch": 0.6119453924914676,
+      "grad_norm": 8.113044738769531,
+      "learning_rate": 0.0007960182025028441,
+      "loss": 6.1718,
+      "step": 1793
+    },
+    {
+      "epoch": 0.6122866894197952,
+      "grad_norm": 3.4572935104370117,
+      "learning_rate": 0.0007959044368600682,
+      "loss": 6.3586,
+      "step": 1794
+    },
+    {
+      "epoch": 0.6126279863481229,
+      "grad_norm": 3.6194283962249756,
+      "learning_rate": 0.0007957906712172925,
+      "loss": 6.8134,
+      "step": 1795
+    },
+    {
+      "epoch": 0.6129692832764505,
+      "grad_norm": 3.5017333030700684,
+      "learning_rate": 0.0007956769055745166,
+      "loss": 7.0632,
+      "step": 1796
+    },
+    {
+      "epoch": 0.6133105802047781,
+      "grad_norm": 3.5431251525878906,
+      "learning_rate": 0.0007955631399317407,
+      "loss": 7.0487,
+      "step": 1797
+    },
+    {
+      "epoch": 0.6136518771331058,
+      "grad_norm": 3.3673455715179443,
+      "learning_rate": 0.0007954493742889648,
+      "loss": 6.7595,
+      "step": 1798
+    },
+    {
+      "epoch": 0.6139931740614335,
+      "grad_norm": 5.616215705871582,
+      "learning_rate": 0.0007953356086461888,
+      "loss": 5.7906,
+      "step": 1799
+    },
+    {
+      "epoch": 0.6143344709897611,
+      "grad_norm": 3.828660488128662,
+      "learning_rate": 0.000795221843003413,
+      "loss": 6.9296,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6146757679180888,
+      "grad_norm": 3.748854637145996,
+      "learning_rate": 0.0007951080773606371,
+      "loss": 6.3692,
+      "step": 1801
+    },
+    {
+      "epoch": 0.6150170648464164,
+      "grad_norm": 3.4253838062286377,
+      "learning_rate": 0.0007949943117178612,
+      "loss": 6.5113,
+      "step": 1802
+    },
+    {
+      "epoch": 0.615358361774744,
+      "grad_norm": 5.2447357177734375,
+      "learning_rate": 0.0007948805460750853,
+      "loss": 6.4785,
+      "step": 1803
+    },
+    {
+      "epoch": 0.6156996587030716,
+      "grad_norm": 3.3891453742980957,
+      "learning_rate": 0.0007947667804323094,
+      "loss": 7.1645,
+      "step": 1804
+    },
+    {
+      "epoch": 0.6160409556313993,
+      "grad_norm": 3.5664119720458984,
+      "learning_rate": 0.0007946530147895335,
+      "loss": 6.733,
+      "step": 1805
+    },
+    {
+      "epoch": 0.616382252559727,
+      "grad_norm": 4.1726298332214355,
+      "learning_rate": 0.0007945392491467577,
+      "loss": 5.8798,
+      "step": 1806
+    },
+    {
+      "epoch": 0.6167235494880546,
+      "grad_norm": 4.10631799697876,
+      "learning_rate": 0.0007944254835039818,
+      "loss": 6.4036,
+      "step": 1807
+    },
+    {
+      "epoch": 0.6170648464163823,
+      "grad_norm": 4.00150728225708,
+      "learning_rate": 0.0007943117178612059,
+      "loss": 6.7822,
+      "step": 1808
+    },
+    {
+      "epoch": 0.6174061433447099,
+      "grad_norm": 3.687084674835205,
+      "learning_rate": 0.00079419795221843,
+      "loss": 6.653,
+      "step": 1809
+    },
+    {
+      "epoch": 0.6177474402730375,
+      "grad_norm": 3.6721670627593994,
+      "learning_rate": 0.0007940841865756541,
+      "loss": 6.7528,
+      "step": 1810
+    },
+    {
+      "epoch": 0.6180887372013651,
+      "grad_norm": 3.374575138092041,
+      "learning_rate": 0.0007939704209328783,
+      "loss": 6.6044,
+      "step": 1811
+    },
+    {
+      "epoch": 0.6184300341296929,
+      "grad_norm": 3.891878843307495,
+      "learning_rate": 0.0007938566552901025,
+      "loss": 6.495,
+      "step": 1812
+    },
+    {
+      "epoch": 0.6187713310580205,
+      "grad_norm": 3.5438005924224854,
+      "learning_rate": 0.0007937428896473266,
+      "loss": 6.2664,
+      "step": 1813
+    },
+    {
+      "epoch": 0.6191126279863481,
+      "grad_norm": 3.501797676086426,
+      "learning_rate": 0.0007936291240045507,
+      "loss": 6.7247,
+      "step": 1814
+    },
+    {
+      "epoch": 0.6194539249146758,
+      "grad_norm": 3.600700855255127,
+      "learning_rate": 0.0007935153583617748,
+      "loss": 6.8725,
+      "step": 1815
+    },
+    {
+      "epoch": 0.6197952218430034,
+      "grad_norm": 3.210486650466919,
+      "learning_rate": 0.0007934015927189989,
+      "loss": 6.9718,
+      "step": 1816
+    },
+    {
+      "epoch": 0.620136518771331,
+      "grad_norm": 3.1832642555236816,
+      "learning_rate": 0.0007932878270762231,
+      "loss": 6.8543,
+      "step": 1817
+    },
+    {
+      "epoch": 0.6204778156996587,
+      "grad_norm": 4.302059173583984,
+      "learning_rate": 0.0007931740614334472,
+      "loss": 5.7968,
+      "step": 1818
+    },
+    {
+      "epoch": 0.6208191126279864,
+      "grad_norm": 3.688284158706665,
+      "learning_rate": 0.0007930602957906712,
+      "loss": 6.4013,
+      "step": 1819
+    },
+    {
+      "epoch": 0.621160409556314,
+      "grad_norm": 3.592975616455078,
+      "learning_rate": 0.0007929465301478953,
+      "loss": 7.0964,
+      "step": 1820
+    },
+    {
+      "epoch": 0.6215017064846416,
+      "grad_norm": 3.5165815353393555,
+      "learning_rate": 0.0007928327645051194,
+      "loss": 7.1356,
+      "step": 1821
+    },
+    {
+      "epoch": 0.6218430034129693,
+      "grad_norm": 3.4153037071228027,
+      "learning_rate": 0.0007927189988623435,
+      "loss": 6.2598,
+      "step": 1822
+    },
+    {
+      "epoch": 0.6221843003412969,
+      "grad_norm": 3.3510231971740723,
+      "learning_rate": 0.0007926052332195677,
+      "loss": 6.149,
+      "step": 1823
+    },
+    {
+      "epoch": 0.6225255972696245,
+      "grad_norm": 3.2553350925445557,
+      "learning_rate": 0.0007924914675767918,
+      "loss": 6.9447,
+      "step": 1824
+    },
+    {
+      "epoch": 0.6228668941979523,
+      "grad_norm": 3.4982099533081055,
+      "learning_rate": 0.0007923777019340159,
+      "loss": 6.8259,
+      "step": 1825
+    },
+    {
+      "epoch": 0.6232081911262799,
+      "grad_norm": 3.446209192276001,
+      "learning_rate": 0.00079226393629124,
+      "loss": 6.4785,
+      "step": 1826
+    },
+    {
+      "epoch": 0.6235494880546075,
+      "grad_norm": 3.815865993499756,
+      "learning_rate": 0.0007921501706484641,
+      "loss": 6.1361,
+      "step": 1827
+    },
+    {
+      "epoch": 0.6238907849829352,
+      "grad_norm": 3.4647111892700195,
+      "learning_rate": 0.0007920364050056883,
+      "loss": 6.4204,
+      "step": 1828
+    },
+    {
+      "epoch": 0.6242320819112628,
+      "grad_norm": 5.167119026184082,
+      "learning_rate": 0.0007919226393629125,
+      "loss": 6.0301,
+      "step": 1829
+    },
+    {
+      "epoch": 0.6245733788395904,
+      "grad_norm": 3.3964574337005615,
+      "learning_rate": 0.0007918088737201366,
+      "loss": 6.1847,
+      "step": 1830
+    },
+    {
+      "epoch": 0.624914675767918,
+      "grad_norm": 3.4840471744537354,
+      "learning_rate": 0.0007916951080773607,
+      "loss": 6.8604,
+      "step": 1831
+    },
+    {
+      "epoch": 0.6252559726962458,
+      "grad_norm": 3.4988765716552734,
+      "learning_rate": 0.0007915813424345848,
+      "loss": 6.5158,
+      "step": 1832
+    },
+    {
+      "epoch": 0.6255972696245734,
+      "grad_norm": 3.5883636474609375,
+      "learning_rate": 0.0007914675767918089,
+      "loss": 6.6659,
+      "step": 1833
+    },
+    {
+      "epoch": 0.625938566552901,
+      "grad_norm": 3.3029944896698,
+      "learning_rate": 0.000791353811149033,
+      "loss": 6.7029,
+      "step": 1834
+    },
+    {
+      "epoch": 0.6262798634812287,
+      "grad_norm": 3.6475093364715576,
+      "learning_rate": 0.0007912400455062572,
+      "loss": 6.6529,
+      "step": 1835
+    },
+    {
+      "epoch": 0.6266211604095563,
+      "grad_norm": 3.320028305053711,
+      "learning_rate": 0.0007911262798634813,
+      "loss": 6.3984,
+      "step": 1836
+    },
+    {
+      "epoch": 0.6269624573378839,
+      "grad_norm": 3.4347753524780273,
+      "learning_rate": 0.0007910125142207054,
+      "loss": 6.7391,
+      "step": 1837
+    },
+    {
+      "epoch": 0.6273037542662117,
+      "grad_norm": 3.7779273986816406,
+      "learning_rate": 0.0007908987485779294,
+      "loss": 6.4521,
+      "step": 1838
+    },
+    {
+      "epoch": 0.6276450511945393,
+      "grad_norm": 3.466188907623291,
+      "learning_rate": 0.0007907849829351535,
+      "loss": 6.3165,
+      "step": 1839
+    },
+    {
+      "epoch": 0.6279863481228669,
+      "grad_norm": 3.9799110889434814,
+      "learning_rate": 0.0007906712172923777,
+      "loss": 6.4507,
+      "step": 1840
+    },
+    {
+      "epoch": 0.6283276450511945,
+      "grad_norm": 3.497555732727051,
+      "learning_rate": 0.0007905574516496018,
+      "loss": 6.819,
+      "step": 1841
+    },
+    {
+      "epoch": 0.6286689419795222,
+      "grad_norm": 3.5930631160736084,
+      "learning_rate": 0.0007904436860068259,
+      "loss": 6.5014,
+      "step": 1842
+    },
+    {
+      "epoch": 0.6290102389078498,
+      "grad_norm": 3.517838478088379,
+      "learning_rate": 0.00079032992036405,
+      "loss": 6.4191,
+      "step": 1843
+    },
+    {
+      "epoch": 0.6293515358361774,
+      "grad_norm": 3.4117469787597656,
+      "learning_rate": 0.0007902161547212741,
+      "loss": 6.6453,
+      "step": 1844
+    },
+    {
+      "epoch": 0.6296928327645052,
+      "grad_norm": 3.3217053413391113,
+      "learning_rate": 0.0007901023890784983,
+      "loss": 6.5161,
+      "step": 1845
+    },
+    {
+      "epoch": 0.6300341296928328,
+      "grad_norm": 3.3784968852996826,
+      "learning_rate": 0.0007899886234357225,
+      "loss": 6.4929,
+      "step": 1846
+    },
+    {
+      "epoch": 0.6303754266211604,
+      "grad_norm": 3.2478907108306885,
+      "learning_rate": 0.0007898748577929466,
+      "loss": 6.4198,
+      "step": 1847
+    },
+    {
+      "epoch": 0.630716723549488,
+      "grad_norm": 3.4220783710479736,
+      "learning_rate": 0.0007897610921501707,
+      "loss": 6.9195,
+      "step": 1848
+    },
+    {
+      "epoch": 0.6310580204778157,
+      "grad_norm": 3.5167126655578613,
+      "learning_rate": 0.0007896473265073948,
+      "loss": 6.4107,
+      "step": 1849
+    },
+    {
+      "epoch": 0.6313993174061433,
+      "grad_norm": 3.388826608657837,
+      "learning_rate": 0.0007895335608646189,
+      "loss": 6.7458,
+      "step": 1850
+    },
+    {
+      "epoch": 0.631740614334471,
+      "grad_norm": 3.440612316131592,
+      "learning_rate": 0.000789419795221843,
+      "loss": 6.2142,
+      "step": 1851
+    },
+    {
+      "epoch": 0.6320819112627987,
+      "grad_norm": 3.389021873474121,
+      "learning_rate": 0.0007893060295790672,
+      "loss": 6.7827,
+      "step": 1852
+    },
+    {
+      "epoch": 0.6324232081911263,
+      "grad_norm": 3.3972480297088623,
+      "learning_rate": 0.0007891922639362913,
+      "loss": 6.7742,
+      "step": 1853
+    },
+    {
+      "epoch": 0.6327645051194539,
+      "grad_norm": 4.108426094055176,
+      "learning_rate": 0.0007890784982935154,
+      "loss": 6.5065,
+      "step": 1854
+    },
+    {
+      "epoch": 0.6331058020477816,
+      "grad_norm": 3.433762311935425,
+      "learning_rate": 0.0007889647326507395,
+      "loss": 6.8187,
+      "step": 1855
+    },
+    {
+      "epoch": 0.6334470989761092,
+      "grad_norm": 4.6445746421813965,
+      "learning_rate": 0.0007888509670079636,
+      "loss": 5.8341,
+      "step": 1856
+    },
+    {
+      "epoch": 0.6337883959044368,
+      "grad_norm": 4.046677589416504,
+      "learning_rate": 0.0007887372013651878,
+      "loss": 5.9867,
+      "step": 1857
+    },
+    {
+      "epoch": 0.6341296928327645,
+      "grad_norm": 3.4539291858673096,
+      "learning_rate": 0.0007886234357224118,
+      "loss": 6.8548,
+      "step": 1858
+    },
+    {
+      "epoch": 0.6344709897610922,
+      "grad_norm": 8.063450813293457,
+      "learning_rate": 0.0007885096700796359,
+      "loss": 7.0799,
+      "step": 1859
+    },
+    {
+      "epoch": 0.6348122866894198,
+      "grad_norm": 4.95775032043457,
+      "learning_rate": 0.00078839590443686,
+      "loss": 6.2372,
+      "step": 1860
+    },
+    {
+      "epoch": 0.6351535836177474,
+      "grad_norm": 3.3882105350494385,
+      "learning_rate": 0.0007882821387940841,
+      "loss": 6.8213,
+      "step": 1861
+    },
+    {
+      "epoch": 0.6354948805460751,
+      "grad_norm": 3.2759015560150146,
+      "learning_rate": 0.0007881683731513083,
+      "loss": 6.717,
+      "step": 1862
+    },
+    {
+      "epoch": 0.6358361774744027,
+      "grad_norm": 3.300447463989258,
+      "learning_rate": 0.0007880546075085325,
+      "loss": 6.194,
+      "step": 1863
+    },
+    {
+      "epoch": 0.6361774744027304,
+      "grad_norm": 3.613112688064575,
+      "learning_rate": 0.0007879408418657566,
+      "loss": 6.6778,
+      "step": 1864
+    },
+    {
+      "epoch": 0.636518771331058,
+      "grad_norm": 3.3438189029693604,
+      "learning_rate": 0.0007878270762229807,
+      "loss": 6.2512,
+      "step": 1865
+    },
+    {
+      "epoch": 0.6368600682593857,
+      "grad_norm": 3.257805347442627,
+      "learning_rate": 0.0007877133105802048,
+      "loss": 6.7676,
+      "step": 1866
+    },
+    {
+      "epoch": 0.6372013651877133,
+      "grad_norm": 3.257349729537964,
+      "learning_rate": 0.0007875995449374289,
+      "loss": 6.7774,
+      "step": 1867
+    },
+    {
+      "epoch": 0.6375426621160409,
+      "grad_norm": 9.708579063415527,
+      "learning_rate": 0.000787485779294653,
+      "loss": 6.7028,
+      "step": 1868
+    },
+    {
+      "epoch": 0.6378839590443686,
+      "grad_norm": 3.499809503555298,
+      "learning_rate": 0.0007873720136518772,
+      "loss": 6.5104,
+      "step": 1869
+    },
+    {
+      "epoch": 0.6382252559726962,
+      "grad_norm": 3.5874903202056885,
+      "learning_rate": 0.0007872582480091013,
+      "loss": 7.1038,
+      "step": 1870
+    },
+    {
+      "epoch": 0.6385665529010239,
+      "grad_norm": 3.483231544494629,
+      "learning_rate": 0.0007871444823663254,
+      "loss": 6.7159,
+      "step": 1871
+    },
+    {
+      "epoch": 0.6389078498293516,
+      "grad_norm": 3.406229257583618,
+      "learning_rate": 0.0007870307167235495,
+      "loss": 6.5912,
+      "step": 1872
+    },
+    {
+      "epoch": 0.6392491467576792,
+      "grad_norm": 3.2390456199645996,
+      "learning_rate": 0.0007869169510807736,
+      "loss": 6.7661,
+      "step": 1873
+    },
+    {
+      "epoch": 0.6395904436860068,
+      "grad_norm": 5.70449161529541,
+      "learning_rate": 0.0007868031854379977,
+      "loss": 5.6305,
+      "step": 1874
+    },
+    {
+      "epoch": 0.6399317406143344,
+      "grad_norm": 3.646918535232544,
+      "learning_rate": 0.0007866894197952219,
+      "loss": 6.772,
+      "step": 1875
+    },
+    {
+      "epoch": 0.6402730375426621,
+      "grad_norm": 3.5219454765319824,
+      "learning_rate": 0.000786575654152446,
+      "loss": 6.6647,
+      "step": 1876
+    },
+    {
+      "epoch": 0.6406143344709898,
+      "grad_norm": 3.4784984588623047,
+      "learning_rate": 0.00078646188850967,
+      "loss": 6.8152,
+      "step": 1877
+    },
+    {
+      "epoch": 0.6409556313993174,
+      "grad_norm": 3.503469467163086,
+      "learning_rate": 0.0007863481228668941,
+      "loss": 6.5573,
+      "step": 1878
+    },
+    {
+      "epoch": 0.6412969283276451,
+      "grad_norm": 3.3382554054260254,
+      "learning_rate": 0.0007862343572241183,
+      "loss": 7.0561,
+      "step": 1879
+    },
+    {
+      "epoch": 0.6416382252559727,
+      "grad_norm": 3.2195024490356445,
+      "learning_rate": 0.0007861205915813425,
+      "loss": 6.5636,
+      "step": 1880
+    },
+    {
+      "epoch": 0.6419795221843003,
+      "grad_norm": 3.3817391395568848,
+      "learning_rate": 0.0007860068259385666,
+      "loss": 6.0123,
+      "step": 1881
+    },
+    {
+      "epoch": 0.642320819112628,
+      "grad_norm": 5.903642177581787,
+      "learning_rate": 0.0007858930602957907,
+      "loss": 5.0916,
+      "step": 1882
+    },
+    {
+      "epoch": 0.6426621160409556,
+      "grad_norm": 3.8153889179229736,
+      "learning_rate": 0.0007857792946530148,
+      "loss": 6.7254,
+      "step": 1883
+    },
+    {
+      "epoch": 0.6430034129692833,
+      "grad_norm": 3.6918563842773438,
+      "learning_rate": 0.0007856655290102389,
+      "loss": 6.8791,
+      "step": 1884
+    },
+    {
+      "epoch": 0.643344709897611,
+      "grad_norm": 3.476710796356201,
+      "learning_rate": 0.000785551763367463,
+      "loss": 6.2294,
+      "step": 1885
+    },
+    {
+      "epoch": 0.6436860068259386,
+      "grad_norm": 3.537196397781372,
+      "learning_rate": 0.0007854379977246872,
+      "loss": 6.5357,
+      "step": 1886
+    },
+    {
+      "epoch": 0.6440273037542662,
+      "grad_norm": 4.697029113769531,
+      "learning_rate": 0.0007853242320819113,
+      "loss": 6.3114,
+      "step": 1887
+    },
+    {
+      "epoch": 0.6443686006825938,
+      "grad_norm": 3.5125951766967773,
+      "learning_rate": 0.0007852104664391354,
+      "loss": 6.6103,
+      "step": 1888
+    },
+    {
+      "epoch": 0.6447098976109215,
+      "grad_norm": 3.5048978328704834,
+      "learning_rate": 0.0007850967007963595,
+      "loss": 7.0019,
+      "step": 1889
+    },
+    {
+      "epoch": 0.6450511945392492,
+      "grad_norm": 3.2914698123931885,
+      "learning_rate": 0.0007849829351535836,
+      "loss": 6.8293,
+      "step": 1890
+    },
+    {
+      "epoch": 0.6453924914675768,
+      "grad_norm": 5.254654407501221,
+      "learning_rate": 0.0007848691695108077,
+      "loss": 6.2791,
+      "step": 1891
+    },
+    {
+      "epoch": 0.6457337883959045,
+      "grad_norm": 3.4875900745391846,
+      "learning_rate": 0.0007847554038680319,
+      "loss": 7.0699,
+      "step": 1892
+    },
+    {
+      "epoch": 0.6460750853242321,
+      "grad_norm": 7.53113317489624,
+      "learning_rate": 0.000784641638225256,
+      "loss": 5.7548,
+      "step": 1893
+    },
+    {
+      "epoch": 0.6464163822525597,
+      "grad_norm": 3.497556209564209,
+      "learning_rate": 0.0007845278725824802,
+      "loss": 6.2634,
+      "step": 1894
+    },
+    {
+      "epoch": 0.6467576791808873,
+      "grad_norm": 5.653488636016846,
+      "learning_rate": 0.0007844141069397043,
+      "loss": 5.9736,
+      "step": 1895
+    },
+    {
+      "epoch": 0.647098976109215,
+      "grad_norm": 3.3323814868927,
+      "learning_rate": 0.0007843003412969284,
+      "loss": 6.4086,
+      "step": 1896
+    },
+    {
+      "epoch": 0.6474402730375427,
+      "grad_norm": 5.373404026031494,
+      "learning_rate": 0.0007841865756541524,
+      "loss": 6.2502,
+      "step": 1897
+    },
+    {
+      "epoch": 0.6477815699658703,
+      "grad_norm": 4.002843379974365,
+      "learning_rate": 0.0007840728100113766,
+      "loss": 6.3939,
+      "step": 1898
+    },
+    {
+      "epoch": 0.648122866894198,
+      "grad_norm": 3.4836678504943848,
+      "learning_rate": 0.0007839590443686007,
+      "loss": 6.5778,
+      "step": 1899
+    },
+    {
+      "epoch": 0.6484641638225256,
+      "grad_norm": 5.268570899963379,
+      "learning_rate": 0.0007838452787258248,
+      "loss": 6.1417,
+      "step": 1900
+    },
+    {
+      "epoch": 0.6488054607508532,
+      "grad_norm": 3.3893330097198486,
+      "learning_rate": 0.0007837315130830489,
+      "loss": 6.579,
+      "step": 1901
+    },
+    {
+      "epoch": 0.6491467576791808,
+      "grad_norm": 3.871875762939453,
+      "learning_rate": 0.000783617747440273,
+      "loss": 6.7902,
+      "step": 1902
+    },
+    {
+      "epoch": 0.6494880546075086,
+      "grad_norm": 3.5524959564208984,
+      "learning_rate": 0.0007835039817974972,
+      "loss": 6.6698,
+      "step": 1903
+    },
+    {
+      "epoch": 0.6498293515358362,
+      "grad_norm": 3.273620128631592,
+      "learning_rate": 0.0007833902161547213,
+      "loss": 6.2347,
+      "step": 1904
+    },
+    {
+      "epoch": 0.6501706484641638,
+      "grad_norm": 3.283205986022949,
+      "learning_rate": 0.0007832764505119454,
+      "loss": 5.978,
+      "step": 1905
+    },
+    {
+      "epoch": 0.6505119453924915,
+      "grad_norm": 3.333651304244995,
+      "learning_rate": 0.0007831626848691695,
+      "loss": 6.5876,
+      "step": 1906
+    },
+    {
+      "epoch": 0.6508532423208191,
+      "grad_norm": 6.1999030113220215,
+      "learning_rate": 0.0007830489192263936,
+      "loss": 6.6631,
+      "step": 1907
+    },
+    {
+      "epoch": 0.6511945392491467,
+      "grad_norm": 3.410543203353882,
+      "learning_rate": 0.0007829351535836177,
+      "loss": 6.5243,
+      "step": 1908
+    },
+    {
+      "epoch": 0.6515358361774743,
+      "grad_norm": 4.422999858856201,
+      "learning_rate": 0.0007828213879408419,
+      "loss": 6.4633,
+      "step": 1909
+    },
+    {
+      "epoch": 0.6518771331058021,
+      "grad_norm": 3.5655972957611084,
+      "learning_rate": 0.000782707622298066,
+      "loss": 6.524,
+      "step": 1910
+    },
+    {
+      "epoch": 0.6522184300341297,
+      "grad_norm": 3.3885929584503174,
+      "learning_rate": 0.0007825938566552902,
+      "loss": 6.7883,
+      "step": 1911
+    },
+    {
+      "epoch": 0.6525597269624573,
+      "grad_norm": 3.3488664627075195,
+      "learning_rate": 0.0007824800910125143,
+      "loss": 6.917,
+      "step": 1912
+    },
+    {
+      "epoch": 0.652901023890785,
+      "grad_norm": 3.5068624019622803,
+      "learning_rate": 0.0007823663253697384,
+      "loss": 6.6262,
+      "step": 1913
+    },
+    {
+      "epoch": 0.6532423208191126,
+      "grad_norm": 3.233506202697754,
+      "learning_rate": 0.0007822525597269625,
+      "loss": 6.6697,
+      "step": 1914
+    },
+    {
+      "epoch": 0.6535836177474402,
+      "grad_norm": 3.320382833480835,
+      "learning_rate": 0.0007821387940841867,
+      "loss": 6.9184,
+      "step": 1915
+    },
+    {
+      "epoch": 0.653924914675768,
+      "grad_norm": 3.2479734420776367,
+      "learning_rate": 0.0007820250284414107,
+      "loss": 6.8682,
+      "step": 1916
+    },
+    {
+      "epoch": 0.6542662116040956,
+      "grad_norm": 3.2496206760406494,
+      "learning_rate": 0.0007819112627986348,
+      "loss": 6.945,
+      "step": 1917
+    },
+    {
+      "epoch": 0.6546075085324232,
+      "grad_norm": 3.3230035305023193,
+      "learning_rate": 0.0007817974971558589,
+      "loss": 6.6453,
+      "step": 1918
+    },
+    {
+      "epoch": 0.6549488054607508,
+      "grad_norm": 4.4280619621276855,
+      "learning_rate": 0.000781683731513083,
+      "loss": 5.9717,
+      "step": 1919
+    },
+    {
+      "epoch": 0.6552901023890785,
+      "grad_norm": 3.414978265762329,
+      "learning_rate": 0.0007815699658703072,
+      "loss": 7.0985,
+      "step": 1920
+    },
+    {
+      "epoch": 0.6556313993174061,
+      "grad_norm": 3.3314380645751953,
+      "learning_rate": 0.0007814562002275313,
+      "loss": 6.9524,
+      "step": 1921
+    },
+    {
+      "epoch": 0.6559726962457337,
+      "grad_norm": 3.4111015796661377,
+      "learning_rate": 0.0007813424345847554,
+      "loss": 6.455,
+      "step": 1922
+    },
+    {
+      "epoch": 0.6563139931740615,
+      "grad_norm": 3.4409379959106445,
+      "learning_rate": 0.0007812286689419795,
+      "loss": 6.6308,
+      "step": 1923
+    },
+    {
+      "epoch": 0.6566552901023891,
+      "grad_norm": 3.246084451675415,
+      "learning_rate": 0.0007811149032992036,
+      "loss": 6.8811,
+      "step": 1924
+    },
+    {
+      "epoch": 0.6569965870307167,
+      "grad_norm": 3.421894073486328,
+      "learning_rate": 0.0007810011376564277,
+      "loss": 6.9111,
+      "step": 1925
+    },
+    {
+      "epoch": 0.6573378839590444,
+      "grad_norm": 3.2279820442199707,
+      "learning_rate": 0.0007808873720136519,
+      "loss": 6.8361,
+      "step": 1926
+    },
+    {
+      "epoch": 0.657679180887372,
+      "grad_norm": 3.337752103805542,
+      "learning_rate": 0.000780773606370876,
+      "loss": 7.0658,
+      "step": 1927
+    },
+    {
+      "epoch": 0.6580204778156996,
+      "grad_norm": 3.4158434867858887,
+      "learning_rate": 0.0007806598407281002,
+      "loss": 6.6466,
+      "step": 1928
+    },
+    {
+      "epoch": 0.6583617747440274,
+      "grad_norm": 3.2831523418426514,
+      "learning_rate": 0.0007805460750853243,
+      "loss": 6.6213,
+      "step": 1929
+    },
+    {
+      "epoch": 0.658703071672355,
+      "grad_norm": 3.8880624771118164,
+      "learning_rate": 0.0007804323094425484,
+      "loss": 6.2975,
+      "step": 1930
+    },
+    {
+      "epoch": 0.6590443686006826,
+      "grad_norm": 3.35299015045166,
+      "learning_rate": 0.0007803185437997725,
+      "loss": 7.1754,
+      "step": 1931
+    },
+    {
+      "epoch": 0.6593856655290102,
+      "grad_norm": 3.4394142627716064,
+      "learning_rate": 0.0007802047781569967,
+      "loss": 6.6023,
+      "step": 1932
+    },
+    {
+      "epoch": 0.6597269624573379,
+      "grad_norm": 3.386638641357422,
+      "learning_rate": 0.0007800910125142208,
+      "loss": 6.9677,
+      "step": 1933
+    },
+    {
+      "epoch": 0.6600682593856655,
+      "grad_norm": 3.344113826751709,
+      "learning_rate": 0.0007799772468714449,
+      "loss": 6.9545,
+      "step": 1934
+    },
+    {
+      "epoch": 0.6604095563139932,
+      "grad_norm": 3.3587405681610107,
+      "learning_rate": 0.0007798634812286689,
+      "loss": 7.0659,
+      "step": 1935
+    },
+    {
+      "epoch": 0.6607508532423209,
+      "grad_norm": 3.9267964363098145,
+      "learning_rate": 0.000779749715585893,
+      "loss": 4.5715,
+      "step": 1936
+    },
+    {
+      "epoch": 0.6610921501706485,
+      "grad_norm": 3.7344627380371094,
+      "learning_rate": 0.0007796359499431171,
+      "loss": 6.0021,
+      "step": 1937
+    },
+    {
+      "epoch": 0.6614334470989761,
+      "grad_norm": 3.6474874019622803,
+      "learning_rate": 0.0007795221843003413,
+      "loss": 6.2885,
+      "step": 1938
+    },
+    {
+      "epoch": 0.6617747440273037,
+      "grad_norm": 6.553812026977539,
+      "learning_rate": 0.0007794084186575654,
+      "loss": 5.7511,
+      "step": 1939
+    },
+    {
+      "epoch": 0.6621160409556314,
+      "grad_norm": 3.5463573932647705,
+      "learning_rate": 0.0007792946530147895,
+      "loss": 6.4974,
+      "step": 1940
+    },
+    {
+      "epoch": 0.662457337883959,
+      "grad_norm": 3.536761999130249,
+      "learning_rate": 0.0007791808873720136,
+      "loss": 6.4656,
+      "step": 1941
+    },
+    {
+      "epoch": 0.6627986348122867,
+      "grad_norm": 3.480790615081787,
+      "learning_rate": 0.0007790671217292377,
+      "loss": 5.9657,
+      "step": 1942
+    },
+    {
+      "epoch": 0.6631399317406144,
+      "grad_norm": 3.640864849090576,
+      "learning_rate": 0.000778953356086462,
+      "loss": 7.0483,
+      "step": 1943
+    },
+    {
+      "epoch": 0.663481228668942,
+      "grad_norm": 3.7452423572540283,
+      "learning_rate": 0.000778839590443686,
+      "loss": 6.3805,
+      "step": 1944
+    },
+    {
+      "epoch": 0.6638225255972696,
+      "grad_norm": 3.4025330543518066,
+      "learning_rate": 0.0007787258248009102,
+      "loss": 6.5517,
+      "step": 1945
+    },
+    {
+      "epoch": 0.6641638225255972,
+      "grad_norm": 3.4373252391815186,
+      "learning_rate": 0.0007786120591581343,
+      "loss": 6.8531,
+      "step": 1946
+    },
+    {
+      "epoch": 0.6645051194539249,
+      "grad_norm": 3.367783546447754,
+      "learning_rate": 0.0007784982935153584,
+      "loss": 6.8904,
+      "step": 1947
+    },
+    {
+      "epoch": 0.6648464163822526,
+      "grad_norm": 3.365324020385742,
+      "learning_rate": 0.0007783845278725825,
+      "loss": 6.4844,
+      "step": 1948
+    },
+    {
+      "epoch": 0.6651877133105802,
+      "grad_norm": 3.2752928733825684,
+      "learning_rate": 0.0007782707622298067,
+      "loss": 6.9888,
+      "step": 1949
+    },
+    {
+      "epoch": 0.6655290102389079,
+      "grad_norm": 3.411865711212158,
+      "learning_rate": 0.0007781569965870308,
+      "loss": 6.9369,
+      "step": 1950
+    },
+    {
+      "epoch": 0.6658703071672355,
+      "grad_norm": 3.8111231327056885,
+      "learning_rate": 0.0007780432309442549,
+      "loss": 5.9984,
+      "step": 1951
+    },
+    {
+      "epoch": 0.6662116040955631,
+      "grad_norm": 3.5093679428100586,
+      "learning_rate": 0.000777929465301479,
+      "loss": 7.1369,
+      "step": 1952
+    },
+    {
+      "epoch": 0.6665529010238908,
+      "grad_norm": 8.49619197845459,
+      "learning_rate": 0.0007778156996587031,
+      "loss": 6.2056,
+      "step": 1953
+    },
+    {
+      "epoch": 0.6668941979522184,
+      "grad_norm": 3.572247266769409,
+      "learning_rate": 0.0007777019340159272,
+      "loss": 7.0313,
+      "step": 1954
+    },
+    {
+      "epoch": 0.6672354948805461,
+      "grad_norm": 6.142834663391113,
+      "learning_rate": 0.0007775881683731513,
+      "loss": 4.7668,
+      "step": 1955
+    },
+    {
+      "epoch": 0.6675767918088737,
+      "grad_norm": 3.7518608570098877,
+      "learning_rate": 0.0007774744027303754,
+      "loss": 7.0454,
+      "step": 1956
+    },
+    {
+      "epoch": 0.6679180887372014,
+      "grad_norm": 3.596379280090332,
+      "learning_rate": 0.0007773606370875995,
+      "loss": 6.8423,
+      "step": 1957
+    },
+    {
+      "epoch": 0.668259385665529,
+      "grad_norm": 3.454772710800171,
+      "learning_rate": 0.0007772468714448236,
+      "loss": 6.3305,
+      "step": 1958
+    },
+    {
+      "epoch": 0.6686006825938566,
+      "grad_norm": 3.3043911457061768,
+      "learning_rate": 0.0007771331058020477,
+      "loss": 6.575,
+      "step": 1959
+    },
+    {
+      "epoch": 0.6689419795221843,
+      "grad_norm": 3.314772844314575,
+      "learning_rate": 0.0007770193401592718,
+      "loss": 6.6896,
+      "step": 1960
+    },
+    {
+      "epoch": 0.669283276450512,
+      "grad_norm": 3.414109230041504,
+      "learning_rate": 0.000776905574516496,
+      "loss": 6.3657,
+      "step": 1961
+    },
+    {
+      "epoch": 0.6696245733788396,
+      "grad_norm": 8.866144180297852,
+      "learning_rate": 0.0007767918088737202,
+      "loss": 6.4915,
+      "step": 1962
+    },
+    {
+      "epoch": 0.6699658703071673,
+      "grad_norm": 5.978751182556152,
+      "learning_rate": 0.0007766780432309443,
+      "loss": 4.808,
+      "step": 1963
+    },
+    {
+      "epoch": 0.6703071672354949,
+      "grad_norm": 4.224075794219971,
+      "learning_rate": 0.0007765642775881684,
+      "loss": 6.2096,
+      "step": 1964
+    },
+    {
+      "epoch": 0.6706484641638225,
+      "grad_norm": 3.7597286701202393,
+      "learning_rate": 0.0007764505119453925,
+      "loss": 6.6962,
+      "step": 1965
+    },
+    {
+      "epoch": 0.6709897610921501,
+      "grad_norm": 3.6143290996551514,
+      "learning_rate": 0.0007763367463026167,
+      "loss": 6.2249,
+      "step": 1966
+    },
+    {
+      "epoch": 0.6713310580204778,
+      "grad_norm": 3.733955144882202,
+      "learning_rate": 0.0007762229806598408,
+      "loss": 6.1922,
+      "step": 1967
+    },
+    {
+      "epoch": 0.6716723549488055,
+      "grad_norm": 3.268876791000366,
+      "learning_rate": 0.0007761092150170649,
+      "loss": 6.6224,
+      "step": 1968
+    },
+    {
+      "epoch": 0.6720136518771331,
+      "grad_norm": 3.7937519550323486,
+      "learning_rate": 0.000775995449374289,
+      "loss": 6.1537,
+      "step": 1969
+    },
+    {
+      "epoch": 0.6723549488054608,
+      "grad_norm": 3.96634578704834,
+      "learning_rate": 0.0007758816837315131,
+      "loss": 5.2715,
+      "step": 1970
+    },
+    {
+      "epoch": 0.6726962457337884,
+      "grad_norm": 3.32902455329895,
+      "learning_rate": 0.0007757679180887372,
+      "loss": 7.1139,
+      "step": 1971
+    },
+    {
+      "epoch": 0.673037542662116,
+      "grad_norm": 3.3868730068206787,
+      "learning_rate": 0.0007756541524459614,
+      "loss": 7.0923,
+      "step": 1972
+    },
+    {
+      "epoch": 0.6733788395904436,
+      "grad_norm": 3.4353487491607666,
+      "learning_rate": 0.0007755403868031855,
+      "loss": 6.6146,
+      "step": 1973
+    },
+    {
+      "epoch": 0.6737201365187714,
+      "grad_norm": 4.369273662567139,
+      "learning_rate": 0.0007754266211604095,
+      "loss": 6.2917,
+      "step": 1974
+    },
+    {
+      "epoch": 0.674061433447099,
+      "grad_norm": 3.72794508934021,
+      "learning_rate": 0.0007753128555176336,
+      "loss": 6.1588,
+      "step": 1975
+    },
+    {
+      "epoch": 0.6744027303754266,
+      "grad_norm": 3.5048928260803223,
+      "learning_rate": 0.0007751990898748577,
+      "loss": 6.603,
+      "step": 1976
+    },
+    {
+      "epoch": 0.6747440273037543,
+      "grad_norm": 4.048648834228516,
+      "learning_rate": 0.0007750853242320818,
+      "loss": 5.5782,
+      "step": 1977
+    },
+    {
+      "epoch": 0.6750853242320819,
+      "grad_norm": 3.9075467586517334,
+      "learning_rate": 0.000774971558589306,
+      "loss": 5.8681,
+      "step": 1978
+    },
+    {
+      "epoch": 0.6754266211604095,
+      "grad_norm": 3.3288211822509766,
+      "learning_rate": 0.0007748577929465302,
+      "loss": 6.6653,
+      "step": 1979
+    },
+    {
+      "epoch": 0.6757679180887372,
+      "grad_norm": 3.535240650177002,
+      "learning_rate": 0.0007747440273037543,
+      "loss": 6.4881,
+      "step": 1980
+    },
+    {
+      "epoch": 0.6761092150170649,
+      "grad_norm": 3.2548129558563232,
+      "learning_rate": 0.0007746302616609784,
+      "loss": 6.792,
+      "step": 1981
+    },
+    {
+      "epoch": 0.6764505119453925,
+      "grad_norm": 3.5258853435516357,
+      "learning_rate": 0.0007745164960182025,
+      "loss": 6.5358,
+      "step": 1982
+    },
+    {
+      "epoch": 0.6767918088737201,
+      "grad_norm": 5.013880729675293,
+      "learning_rate": 0.0007744027303754267,
+      "loss": 6.1146,
+      "step": 1983
+    },
+    {
+      "epoch": 0.6771331058020478,
+      "grad_norm": 3.550227165222168,
+      "learning_rate": 0.0007742889647326508,
+      "loss": 6.9435,
+      "step": 1984
+    },
+    {
+      "epoch": 0.6774744027303754,
+      "grad_norm": 3.3671066761016846,
+      "learning_rate": 0.0007741751990898749,
+      "loss": 6.4325,
+      "step": 1985
+    },
+    {
+      "epoch": 0.677815699658703,
+      "grad_norm": 4.051577091217041,
+      "learning_rate": 0.000774061433447099,
+      "loss": 6.3684,
+      "step": 1986
+    },
+    {
+      "epoch": 0.6781569965870308,
+      "grad_norm": 3.442668914794922,
+      "learning_rate": 0.0007739476678043231,
+      "loss": 6.7317,
+      "step": 1987
+    },
+    {
+      "epoch": 0.6784982935153584,
+      "grad_norm": 3.2804269790649414,
+      "learning_rate": 0.0007738339021615472,
+      "loss": 6.6793,
+      "step": 1988
+    },
+    {
+      "epoch": 0.678839590443686,
+      "grad_norm": 4.920018196105957,
+      "learning_rate": 0.0007737201365187714,
+      "loss": 5.9878,
+      "step": 1989
+    },
+    {
+      "epoch": 0.6791808873720137,
+      "grad_norm": 3.5679967403411865,
+      "learning_rate": 0.0007736063708759955,
+      "loss": 6.9321,
+      "step": 1990
+    },
+    {
+      "epoch": 0.6795221843003413,
+      "grad_norm": 3.628213405609131,
+      "learning_rate": 0.0007734926052332196,
+      "loss": 6.5297,
+      "step": 1991
+    },
+    {
+      "epoch": 0.6798634812286689,
+      "grad_norm": 3.5422585010528564,
+      "learning_rate": 0.0007733788395904437,
+      "loss": 6.1211,
+      "step": 1992
+    },
+    {
+      "epoch": 0.6802047781569965,
+      "grad_norm": 3.3573875427246094,
+      "learning_rate": 0.0007732650739476678,
+      "loss": 7.0201,
+      "step": 1993
+    },
+    {
+      "epoch": 0.6805460750853243,
+      "grad_norm": 3.2984368801116943,
+      "learning_rate": 0.0007731513083048918,
+      "loss": 6.5626,
+      "step": 1994
+    },
+    {
+      "epoch": 0.6808873720136519,
+      "grad_norm": 3.394038200378418,
+      "learning_rate": 0.000773037542662116,
+      "loss": 6.9126,
+      "step": 1995
+    },
+    {
+      "epoch": 0.6812286689419795,
+      "grad_norm": 3.2731823921203613,
+      "learning_rate": 0.0007729237770193402,
+      "loss": 6.8367,
+      "step": 1996
+    },
+    {
+      "epoch": 0.6815699658703072,
+      "grad_norm": 3.3242337703704834,
+      "learning_rate": 0.0007728100113765643,
+      "loss": 6.6835,
+      "step": 1997
+    },
+    {
+      "epoch": 0.6819112627986348,
+      "grad_norm": 3.444890260696411,
+      "learning_rate": 0.0007726962457337884,
+      "loss": 6.7718,
+      "step": 1998
+    },
+    {
+      "epoch": 0.6822525597269624,
+      "grad_norm": 3.4120771884918213,
+      "learning_rate": 0.0007725824800910125,
+      "loss": 6.6572,
+      "step": 1999
+    },
+    {
+      "epoch": 0.6825938566552902,
+      "grad_norm": 4.183150768280029,
+      "learning_rate": 0.0007724687144482366,
+      "loss": 6.0941,
+      "step": 2000
+    },
+    {
+      "epoch": 0.6829351535836178,
+      "grad_norm": 3.6355671882629395,
+      "learning_rate": 0.0007723549488054608,
+      "loss": 5.8741,
+      "step": 2001
+    },
+    {
+      "epoch": 0.6832764505119454,
+      "grad_norm": 6.971808433532715,
+      "learning_rate": 0.0007722411831626849,
+      "loss": 4.9892,
+      "step": 2002
+    },
+    {
+      "epoch": 0.683617747440273,
+      "grad_norm": 4.916726589202881,
+      "learning_rate": 0.000772127417519909,
+      "loss": 6.2013,
+      "step": 2003
+    },
+    {
+      "epoch": 0.6839590443686007,
+      "grad_norm": 3.724529981613159,
+      "learning_rate": 0.0007720136518771331,
+      "loss": 6.086,
+      "step": 2004
+    },
+    {
+      "epoch": 0.6843003412969283,
+      "grad_norm": 3.7401139736175537,
+      "learning_rate": 0.0007718998862343572,
+      "loss": 6.875,
+      "step": 2005
+    },
+    {
+      "epoch": 0.6846416382252559,
+      "grad_norm": 3.41847562789917,
+      "learning_rate": 0.0007717861205915814,
+      "loss": 6.9408,
+      "step": 2006
+    },
+    {
+      "epoch": 0.6849829351535837,
+      "grad_norm": 3.425471782684326,
+      "learning_rate": 0.0007716723549488055,
+      "loss": 6.7579,
+      "step": 2007
+    },
+    {
+      "epoch": 0.6853242320819113,
+      "grad_norm": 3.197573661804199,
+      "learning_rate": 0.0007715585893060296,
+      "loss": 6.6664,
+      "step": 2008
+    },
+    {
+      "epoch": 0.6856655290102389,
+      "grad_norm": 3.4145679473876953,
+      "learning_rate": 0.0007714448236632537,
+      "loss": 6.515,
+      "step": 2009
+    },
+    {
+      "epoch": 0.6860068259385665,
+      "grad_norm": 3.9534404277801514,
+      "learning_rate": 0.0007713310580204778,
+      "loss": 6.2037,
+      "step": 2010
+    },
+    {
+      "epoch": 0.6863481228668942,
+      "grad_norm": 3.274080514907837,
+      "learning_rate": 0.000771217292377702,
+      "loss": 6.4559,
+      "step": 2011
+    },
+    {
+      "epoch": 0.6866894197952218,
+      "grad_norm": 3.604434013366699,
+      "learning_rate": 0.0007711035267349262,
+      "loss": 7.1227,
+      "step": 2012
+    },
+    {
+      "epoch": 0.6870307167235495,
+      "grad_norm": 3.635999917984009,
+      "learning_rate": 0.0007709897610921502,
+      "loss": 6.5486,
+      "step": 2013
+    },
+    {
+      "epoch": 0.6873720136518772,
+      "grad_norm": 3.22017765045166,
+      "learning_rate": 0.0007708759954493743,
+      "loss": 6.9555,
+      "step": 2014
+    },
+    {
+      "epoch": 0.6877133105802048,
+      "grad_norm": 3.5242180824279785,
+      "learning_rate": 0.0007707622298065984,
+      "loss": 6.6827,
+      "step": 2015
+    },
+    {
+      "epoch": 0.6880546075085324,
+      "grad_norm": 3.7955944538116455,
+      "learning_rate": 0.0007706484641638225,
+      "loss": 5.7521,
+      "step": 2016
+    },
+    {
+      "epoch": 0.68839590443686,
+      "grad_norm": 3.422865629196167,
+      "learning_rate": 0.0007705346985210466,
+      "loss": 6.7181,
+      "step": 2017
+    },
+    {
+      "epoch": 0.6887372013651877,
+      "grad_norm": 3.7344348430633545,
+      "learning_rate": 0.0007704209328782708,
+      "loss": 6.6848,
+      "step": 2018
+    },
+    {
+      "epoch": 0.6890784982935153,
+      "grad_norm": 3.6247267723083496,
+      "learning_rate": 0.0007703071672354949,
+      "loss": 6.3779,
+      "step": 2019
+    },
+    {
+      "epoch": 0.689419795221843,
+      "grad_norm": 3.669395923614502,
+      "learning_rate": 0.000770193401592719,
+      "loss": 6.2553,
+      "step": 2020
+    },
+    {
+      "epoch": 0.6897610921501707,
+      "grad_norm": 3.4517054557800293,
+      "learning_rate": 0.0007700796359499431,
+      "loss": 6.3794,
+      "step": 2021
+    },
+    {
+      "epoch": 0.6901023890784983,
+      "grad_norm": 3.477140188217163,
+      "learning_rate": 0.0007699658703071672,
+      "loss": 7.0087,
+      "step": 2022
+    },
+    {
+      "epoch": 0.6904436860068259,
+      "grad_norm": 3.354229688644409,
+      "learning_rate": 0.0007698521046643914,
+      "loss": 6.3737,
+      "step": 2023
+    },
+    {
+      "epoch": 0.6907849829351536,
+      "grad_norm": 3.4395453929901123,
+      "learning_rate": 0.0007697383390216155,
+      "loss": 6.6592,
+      "step": 2024
+    },
+    {
+      "epoch": 0.6911262798634812,
+      "grad_norm": 3.5157663822174072,
+      "learning_rate": 0.0007696245733788396,
+      "loss": 6.6477,
+      "step": 2025
+    },
+    {
+      "epoch": 0.6914675767918089,
+      "grad_norm": 3.3607208728790283,
+      "learning_rate": 0.0007695108077360637,
+      "loss": 6.3353,
+      "step": 2026
+    },
+    {
+      "epoch": 0.6918088737201366,
+      "grad_norm": 5.396403789520264,
+      "learning_rate": 0.0007693970420932878,
+      "loss": 6.0246,
+      "step": 2027
+    },
+    {
+      "epoch": 0.6921501706484642,
+      "grad_norm": 3.4905757904052734,
+      "learning_rate": 0.000769283276450512,
+      "loss": 6.1745,
+      "step": 2028
+    },
+    {
+      "epoch": 0.6924914675767918,
+      "grad_norm": 3.5405080318450928,
+      "learning_rate": 0.0007691695108077362,
+      "loss": 6.5764,
+      "step": 2029
+    },
+    {
+      "epoch": 0.6928327645051194,
+      "grad_norm": 4.243968963623047,
+      "learning_rate": 0.0007690557451649603,
+      "loss": 6.2635,
+      "step": 2030
+    },
+    {
+      "epoch": 0.6931740614334471,
+      "grad_norm": 6.531406402587891,
+      "learning_rate": 0.0007689419795221844,
+      "loss": 5.7165,
+      "step": 2031
+    },
+    {
+      "epoch": 0.6935153583617747,
+      "grad_norm": 3.5809173583984375,
+      "learning_rate": 0.0007688282138794085,
+      "loss": 6.8084,
+      "step": 2032
+    },
+    {
+      "epoch": 0.6938566552901024,
+      "grad_norm": 3.7270376682281494,
+      "learning_rate": 0.0007687144482366325,
+      "loss": 6.7852,
+      "step": 2033
+    },
+    {
+      "epoch": 0.6941979522184301,
+      "grad_norm": 3.384864330291748,
+      "learning_rate": 0.0007686006825938566,
+      "loss": 6.6244,
+      "step": 2034
+    },
+    {
+      "epoch": 0.6945392491467577,
+      "grad_norm": 4.850109100341797,
+      "learning_rate": 0.0007684869169510808,
+      "loss": 6.7117,
+      "step": 2035
+    },
+    {
+      "epoch": 0.6948805460750853,
+      "grad_norm": 6.860106468200684,
+      "learning_rate": 0.0007683731513083049,
+      "loss": 6.1666,
+      "step": 2036
+    },
+    {
+      "epoch": 0.6952218430034129,
+      "grad_norm": 3.350128412246704,
+      "learning_rate": 0.000768259385665529,
+      "loss": 6.9472,
+      "step": 2037
+    },
+    {
+      "epoch": 0.6955631399317406,
+      "grad_norm": 3.3778512477874756,
+      "learning_rate": 0.0007681456200227531,
+      "loss": 6.8923,
+      "step": 2038
+    },
+    {
+      "epoch": 0.6959044368600683,
+      "grad_norm": 3.257622003555298,
+      "learning_rate": 0.0007680318543799772,
+      "loss": 6.6449,
+      "step": 2039
+    },
+    {
+      "epoch": 0.6962457337883959,
+      "grad_norm": 3.383254051208496,
+      "learning_rate": 0.0007679180887372013,
+      "loss": 5.932,
+      "step": 2040
+    },
+    {
+      "epoch": 0.6965870307167236,
+      "grad_norm": 3.2423672676086426,
+      "learning_rate": 0.0007678043230944255,
+      "loss": 6.2983,
+      "step": 2041
+    },
+    {
+      "epoch": 0.6969283276450512,
+      "grad_norm": 3.3049263954162598,
+      "learning_rate": 0.0007676905574516496,
+      "loss": 6.4907,
+      "step": 2042
+    },
+    {
+      "epoch": 0.6972696245733788,
+      "grad_norm": 3.3618760108947754,
+      "learning_rate": 0.0007675767918088737,
+      "loss": 6.322,
+      "step": 2043
+    },
+    {
+      "epoch": 0.6976109215017064,
+      "grad_norm": 3.2208092212677,
+      "learning_rate": 0.0007674630261660978,
+      "loss": 6.8044,
+      "step": 2044
+    },
+    {
+      "epoch": 0.6979522184300341,
+      "grad_norm": 3.3235998153686523,
+      "learning_rate": 0.000767349260523322,
+      "loss": 5.9484,
+      "step": 2045
+    },
+    {
+      "epoch": 0.6982935153583618,
+      "grad_norm": 3.5273094177246094,
+      "learning_rate": 0.0007672354948805462,
+      "loss": 6.9184,
+      "step": 2046
+    },
+    {
+      "epoch": 0.6986348122866894,
+      "grad_norm": 3.4016122817993164,
+      "learning_rate": 0.0007671217292377703,
+      "loss": 6.5703,
+      "step": 2047
+    },
+    {
+      "epoch": 0.6989761092150171,
+      "grad_norm": 5.7711405754089355,
+      "learning_rate": 0.0007670079635949944,
+      "loss": 5.5968,
+      "step": 2048
+    },
+    {
+      "epoch": 0.6993174061433447,
+      "grad_norm": 7.544925689697266,
+      "learning_rate": 0.0007668941979522185,
+      "loss": 6.2352,
+      "step": 2049
+    },
+    {
+      "epoch": 0.6996587030716723,
+      "grad_norm": 3.5243000984191895,
+      "learning_rate": 0.0007667804323094426,
+      "loss": 6.3483,
+      "step": 2050
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.510413885116577,
+      "learning_rate": 0.0007666666666666667,
+      "loss": 6.7339,
+      "step": 2051
+    },
+    {
+      "epoch": 0.7003412969283277,
+      "grad_norm": 3.272040843963623,
+      "learning_rate": 0.0007665529010238908,
+      "loss": 6.3808,
+      "step": 2052
+    },
+    {
+      "epoch": 0.7006825938566553,
+      "grad_norm": 3.230138063430786,
+      "learning_rate": 0.0007664391353811149,
+      "loss": 6.4907,
+      "step": 2053
+    },
+    {
+      "epoch": 0.701023890784983,
+      "grad_norm": 3.2666988372802734,
+      "learning_rate": 0.000766325369738339,
+      "loss": 6.5341,
+      "step": 2054
+    },
+    {
+      "epoch": 0.7013651877133106,
+      "grad_norm": 3.246366262435913,
+      "learning_rate": 0.0007662116040955631,
+      "loss": 6.4526,
+      "step": 2055
+    },
+    {
+      "epoch": 0.7017064846416382,
+      "grad_norm": 3.516317844390869,
+      "learning_rate": 0.0007660978384527872,
+      "loss": 7.0432,
+      "step": 2056
+    },
+    {
+      "epoch": 0.7020477815699658,
+      "grad_norm": 3.530691146850586,
+      "learning_rate": 0.0007659840728100113,
+      "loss": 6.7314,
+      "step": 2057
+    },
+    {
+      "epoch": 0.7023890784982935,
+      "grad_norm": 3.3396496772766113,
+      "learning_rate": 0.0007658703071672355,
+      "loss": 6.9605,
+      "step": 2058
+    },
+    {
+      "epoch": 0.7027303754266212,
+      "grad_norm": 3.306985378265381,
+      "learning_rate": 0.0007657565415244596,
+      "loss": 6.488,
+      "step": 2059
+    },
+    {
+      "epoch": 0.7030716723549488,
+      "grad_norm": 3.3087706565856934,
+      "learning_rate": 0.0007656427758816837,
+      "loss": 5.9937,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7034129692832765,
+      "grad_norm": 3.159797191619873,
+      "learning_rate": 0.0007655290102389078,
+      "loss": 6.5233,
+      "step": 2061
+    },
+    {
+      "epoch": 0.7037542662116041,
+      "grad_norm": 3.4056601524353027,
+      "learning_rate": 0.000765415244596132,
+      "loss": 6.7224,
+      "step": 2062
+    },
+    {
+      "epoch": 0.7040955631399317,
+      "grad_norm": 3.676870822906494,
+      "learning_rate": 0.0007653014789533561,
+      "loss": 6.3446,
+      "step": 2063
+    },
+    {
+      "epoch": 0.7044368600682593,
+      "grad_norm": 3.3793466091156006,
+      "learning_rate": 0.0007651877133105803,
+      "loss": 6.8083,
+      "step": 2064
+    },
+    {
+      "epoch": 0.7047781569965871,
+      "grad_norm": 3.507800340652466,
+      "learning_rate": 0.0007650739476678044,
+      "loss": 7.0639,
+      "step": 2065
+    },
+    {
+      "epoch": 0.7051194539249147,
+      "grad_norm": 6.222537994384766,
+      "learning_rate": 0.0007649601820250285,
+      "loss": 6.6979,
+      "step": 2066
+    },
+    {
+      "epoch": 0.7054607508532423,
+      "grad_norm": 3.477949380874634,
+      "learning_rate": 0.0007648464163822526,
+      "loss": 6.6638,
+      "step": 2067
+    },
+    {
+      "epoch": 0.70580204778157,
+      "grad_norm": 3.4923906326293945,
+      "learning_rate": 0.0007647326507394767,
+      "loss": 6.8871,
+      "step": 2068
+    },
+    {
+      "epoch": 0.7061433447098976,
+      "grad_norm": 3.818601608276367,
+      "learning_rate": 0.0007646188850967009,
+      "loss": 6.0797,
+      "step": 2069
+    },
+    {
+      "epoch": 0.7064846416382252,
+      "grad_norm": 3.1661059856414795,
+      "learning_rate": 0.000764505119453925,
+      "loss": 6.2969,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7068259385665528,
+      "grad_norm": 3.237128496170044,
+      "learning_rate": 0.0007643913538111491,
+      "loss": 6.4024,
+      "step": 2071
+    },
+    {
+      "epoch": 0.7071672354948806,
+      "grad_norm": 3.7169055938720703,
+      "learning_rate": 0.0007642775881683731,
+      "loss": 5.803,
+      "step": 2072
+    },
+    {
+      "epoch": 0.7075085324232082,
+      "grad_norm": 3.3643290996551514,
+      "learning_rate": 0.0007641638225255972,
+      "loss": 6.5204,
+      "step": 2073
+    },
+    {
+      "epoch": 0.7078498293515358,
+      "grad_norm": 3.5020577907562256,
+      "learning_rate": 0.0007640500568828213,
+      "loss": 6.4471,
+      "step": 2074
+    },
+    {
+      "epoch": 0.7081911262798635,
+      "grad_norm": 3.174459934234619,
+      "learning_rate": 0.0007639362912400455,
+      "loss": 6.4608,
+      "step": 2075
+    },
+    {
+      "epoch": 0.7085324232081911,
+      "grad_norm": 3.408348321914673,
+      "learning_rate": 0.0007638225255972696,
+      "loss": 6.2846,
+      "step": 2076
+    },
+    {
+      "epoch": 0.7088737201365187,
+      "grad_norm": 3.218968152999878,
+      "learning_rate": 0.0007637087599544937,
+      "loss": 6.5383,
+      "step": 2077
+    },
+    {
+      "epoch": 0.7092150170648465,
+      "grad_norm": 3.2198712825775146,
+      "learning_rate": 0.0007635949943117178,
+      "loss": 6.9809,
+      "step": 2078
+    },
+    {
+      "epoch": 0.7095563139931741,
+      "grad_norm": 3.364302635192871,
+      "learning_rate": 0.000763481228668942,
+      "loss": 6.0809,
+      "step": 2079
+    },
+    {
+      "epoch": 0.7098976109215017,
+      "grad_norm": 3.22308611869812,
+      "learning_rate": 0.0007633674630261661,
+      "loss": 6.7824,
+      "step": 2080
+    },
+    {
+      "epoch": 0.7102389078498293,
+      "grad_norm": 3.4846742153167725,
+      "learning_rate": 0.0007632536973833903,
+      "loss": 6.8048,
+      "step": 2081
+    },
+    {
+      "epoch": 0.710580204778157,
+      "grad_norm": 3.32481050491333,
+      "learning_rate": 0.0007631399317406144,
+      "loss": 6.4347,
+      "step": 2082
+    },
+    {
+      "epoch": 0.7109215017064846,
+      "grad_norm": 3.340768814086914,
+      "learning_rate": 0.0007630261660978385,
+      "loss": 6.7707,
+      "step": 2083
+    },
+    {
+      "epoch": 0.7112627986348122,
+      "grad_norm": 5.2757086753845215,
+      "learning_rate": 0.0007629124004550626,
+      "loss": 5.6813,
+      "step": 2084
+    },
+    {
+      "epoch": 0.71160409556314,
+      "grad_norm": 3.4196841716766357,
+      "learning_rate": 0.0007627986348122867,
+      "loss": 6.8647,
+      "step": 2085
+    },
+    {
+      "epoch": 0.7119453924914676,
+      "grad_norm": 3.5735483169555664,
+      "learning_rate": 0.0007626848691695109,
+      "loss": 6.4418,
+      "step": 2086
+    },
+    {
+      "epoch": 0.7122866894197952,
+      "grad_norm": 3.366319179534912,
+      "learning_rate": 0.000762571103526735,
+      "loss": 6.5476,
+      "step": 2087
+    },
+    {
+      "epoch": 0.7126279863481229,
+      "grad_norm": 3.2403366565704346,
+      "learning_rate": 0.0007624573378839591,
+      "loss": 6.5685,
+      "step": 2088
+    },
+    {
+      "epoch": 0.7129692832764505,
+      "grad_norm": 3.317497730255127,
+      "learning_rate": 0.0007623435722411832,
+      "loss": 6.5655,
+      "step": 2089
+    },
+    {
+      "epoch": 0.7133105802047781,
+      "grad_norm": 3.3788607120513916,
+      "learning_rate": 0.0007622298065984073,
+      "loss": 6.5866,
+      "step": 2090
+    },
+    {
+      "epoch": 0.7136518771331058,
+      "grad_norm": 7.484004020690918,
+      "learning_rate": 0.0007621160409556313,
+      "loss": 5.0795,
+      "step": 2091
+    },
+    {
+      "epoch": 0.7139931740614335,
+      "grad_norm": 3.5781850814819336,
+      "learning_rate": 0.0007620022753128555,
+      "loss": 6.0668,
+      "step": 2092
+    },
+    {
+      "epoch": 0.7143344709897611,
+      "grad_norm": 3.5423996448516846,
+      "learning_rate": 0.0007618885096700796,
+      "loss": 6.4591,
+      "step": 2093
+    },
+    {
+      "epoch": 0.7146757679180887,
+      "grad_norm": 3.2912189960479736,
+      "learning_rate": 0.0007617747440273037,
+      "loss": 6.3944,
+      "step": 2094
+    },
+    {
+      "epoch": 0.7150170648464164,
+      "grad_norm": 3.3964452743530273,
+      "learning_rate": 0.0007616609783845278,
+      "loss": 6.0996,
+      "step": 2095
+    },
+    {
+      "epoch": 0.715358361774744,
+      "grad_norm": 3.4383544921875,
+      "learning_rate": 0.000761547212741752,
+      "loss": 6.8264,
+      "step": 2096
+    },
+    {
+      "epoch": 0.7156996587030716,
+      "grad_norm": 3.80399227142334,
+      "learning_rate": 0.0007614334470989761,
+      "loss": 5.6848,
+      "step": 2097
+    },
+    {
+      "epoch": 0.7160409556313994,
+      "grad_norm": 3.4346396923065186,
+      "learning_rate": 0.0007613196814562003,
+      "loss": 6.3825,
+      "step": 2098
+    },
+    {
+      "epoch": 0.716382252559727,
+      "grad_norm": 4.535429000854492,
+      "learning_rate": 0.0007612059158134244,
+      "loss": 5.9877,
+      "step": 2099
+    },
+    {
+      "epoch": 0.7167235494880546,
+      "grad_norm": 4.623420238494873,
+      "learning_rate": 0.0007610921501706485,
+      "loss": 6.6088,
+      "step": 2100
+    },
+    {
+      "epoch": 0.7170648464163822,
+      "grad_norm": 3.520456552505493,
+      "learning_rate": 0.0007609783845278726,
+      "loss": 6.8445,
+      "step": 2101
+    },
+    {
+      "epoch": 0.7174061433447099,
+      "grad_norm": 3.4850854873657227,
+      "learning_rate": 0.0007608646188850967,
+      "loss": 6.5181,
+      "step": 2102
+    },
+    {
+      "epoch": 0.7177474402730375,
+      "grad_norm": 3.35722279548645,
+      "learning_rate": 0.0007607508532423208,
+      "loss": 7.1329,
+      "step": 2103
+    },
+    {
+      "epoch": 0.7180887372013652,
+      "grad_norm": 3.2173945903778076,
+      "learning_rate": 0.000760637087599545,
+      "loss": 6.9774,
+      "step": 2104
+    },
+    {
+      "epoch": 0.7184300341296929,
+      "grad_norm": 3.337705373764038,
+      "learning_rate": 0.0007605233219567691,
+      "loss": 7.1209,
+      "step": 2105
+    },
+    {
+      "epoch": 0.7187713310580205,
+      "grad_norm": 3.361863136291504,
+      "learning_rate": 0.0007604095563139932,
+      "loss": 6.4386,
+      "step": 2106
+    },
+    {
+      "epoch": 0.7191126279863481,
+      "grad_norm": 3.371455669403076,
+      "learning_rate": 0.0007602957906712173,
+      "loss": 7.0502,
+      "step": 2107
+    },
+    {
+      "epoch": 0.7194539249146757,
+      "grad_norm": 3.6178102493286133,
+      "learning_rate": 0.0007601820250284414,
+      "loss": 6.33,
+      "step": 2108
+    },
+    {
+      "epoch": 0.7197952218430034,
+      "grad_norm": 3.337318181991577,
+      "learning_rate": 0.0007600682593856656,
+      "loss": 7.0613,
+      "step": 2109
+    },
+    {
+      "epoch": 0.7201365187713311,
+      "grad_norm": 3.516834020614624,
+      "learning_rate": 0.0007599544937428896,
+      "loss": 6.2131,
+      "step": 2110
+    },
+    {
+      "epoch": 0.7204778156996587,
+      "grad_norm": 3.4340057373046875,
+      "learning_rate": 0.0007598407281001137,
+      "loss": 6.4354,
+      "step": 2111
+    },
+    {
+      "epoch": 0.7208191126279864,
+      "grad_norm": 3.5667688846588135,
+      "learning_rate": 0.0007597269624573379,
+      "loss": 6.841,
+      "step": 2112
+    },
+    {
+      "epoch": 0.721160409556314,
+      "grad_norm": 3.2933144569396973,
+      "learning_rate": 0.000759613196814562,
+      "loss": 6.542,
+      "step": 2113
+    },
+    {
+      "epoch": 0.7215017064846416,
+      "grad_norm": 3.4875848293304443,
+      "learning_rate": 0.0007594994311717861,
+      "loss": 6.9857,
+      "step": 2114
+    },
+    {
+      "epoch": 0.7218430034129693,
+      "grad_norm": 4.947911739349365,
+      "learning_rate": 0.0007593856655290103,
+      "loss": 5.5756,
+      "step": 2115
+    },
+    {
+      "epoch": 0.7221843003412969,
+      "grad_norm": 3.511719226837158,
+      "learning_rate": 0.0007592718998862344,
+      "loss": 6.6189,
+      "step": 2116
+    },
+    {
+      "epoch": 0.7225255972696246,
+      "grad_norm": 3.4902215003967285,
+      "learning_rate": 0.0007591581342434585,
+      "loss": 6.9783,
+      "step": 2117
+    },
+    {
+      "epoch": 0.7228668941979522,
+      "grad_norm": 3.5169901847839355,
+      "learning_rate": 0.0007590443686006826,
+      "loss": 6.8164,
+      "step": 2118
+    },
+    {
+      "epoch": 0.7232081911262799,
+      "grad_norm": 3.3471617698669434,
+      "learning_rate": 0.0007589306029579067,
+      "loss": 6.8412,
+      "step": 2119
+    },
+    {
+      "epoch": 0.7235494880546075,
+      "grad_norm": 3.3342645168304443,
+      "learning_rate": 0.0007588168373151308,
+      "loss": 6.4882,
+      "step": 2120
+    },
+    {
+      "epoch": 0.7238907849829351,
+      "grad_norm": 3.29974627494812,
+      "learning_rate": 0.000758703071672355,
+      "loss": 6.5342,
+      "step": 2121
+    },
+    {
+      "epoch": 0.7242320819112628,
+      "grad_norm": 5.739980697631836,
+      "learning_rate": 0.0007585893060295791,
+      "loss": 6.6463,
+      "step": 2122
+    },
+    {
+      "epoch": 0.7245733788395905,
+      "grad_norm": 4.314258575439453,
+      "learning_rate": 0.0007584755403868032,
+      "loss": 5.9307,
+      "step": 2123
+    },
+    {
+      "epoch": 0.7249146757679181,
+      "grad_norm": 5.394867897033691,
+      "learning_rate": 0.0007583617747440273,
+      "loss": 6.2618,
+      "step": 2124
+    },
+    {
+      "epoch": 0.7252559726962458,
+      "grad_norm": 3.7781617641448975,
+      "learning_rate": 0.0007582480091012514,
+      "loss": 6.4021,
+      "step": 2125
+    },
+    {
+      "epoch": 0.7255972696245734,
+      "grad_norm": 3.883280038833618,
+      "learning_rate": 0.0007581342434584756,
+      "loss": 6.2862,
+      "step": 2126
+    },
+    {
+      "epoch": 0.725938566552901,
+      "grad_norm": 3.35640025138855,
+      "learning_rate": 0.0007580204778156997,
+      "loss": 6.9783,
+      "step": 2127
+    },
+    {
+      "epoch": 0.7262798634812286,
+      "grad_norm": 3.3483872413635254,
+      "learning_rate": 0.0007579067121729239,
+      "loss": 6.6387,
+      "step": 2128
+    },
+    {
+      "epoch": 0.7266211604095563,
+      "grad_norm": 3.295193672180176,
+      "learning_rate": 0.000757792946530148,
+      "loss": 6.7598,
+      "step": 2129
+    },
+    {
+      "epoch": 0.726962457337884,
+      "grad_norm": 5.115669250488281,
+      "learning_rate": 0.000757679180887372,
+      "loss": 5.9118,
+      "step": 2130
+    },
+    {
+      "epoch": 0.7273037542662116,
+      "grad_norm": 3.72269606590271,
+      "learning_rate": 0.0007575654152445961,
+      "loss": 6.5912,
+      "step": 2131
+    },
+    {
+      "epoch": 0.7276450511945393,
+      "grad_norm": 3.6045024394989014,
+      "learning_rate": 0.0007574516496018203,
+      "loss": 6.7721,
+      "step": 2132
+    },
+    {
+      "epoch": 0.7279863481228669,
+      "grad_norm": 3.3702828884124756,
+      "learning_rate": 0.0007573378839590444,
+      "loss": 6.1741,
+      "step": 2133
+    },
+    {
+      "epoch": 0.7283276450511945,
+      "grad_norm": 3.3547866344451904,
+      "learning_rate": 0.0007572241183162685,
+      "loss": 6.5913,
+      "step": 2134
+    },
+    {
+      "epoch": 0.7286689419795221,
+      "grad_norm": 3.366942882537842,
+      "learning_rate": 0.0007571103526734926,
+      "loss": 6.519,
+      "step": 2135
+    },
+    {
+      "epoch": 0.7290102389078499,
+      "grad_norm": 3.3425605297088623,
+      "learning_rate": 0.0007569965870307167,
+      "loss": 7.0345,
+      "step": 2136
+    },
+    {
+      "epoch": 0.7293515358361775,
+      "grad_norm": 6.479069232940674,
+      "learning_rate": 0.0007568828213879408,
+      "loss": 6.3204,
+      "step": 2137
+    },
+    {
+      "epoch": 0.7296928327645051,
+      "grad_norm": 3.5541889667510986,
+      "learning_rate": 0.000756769055745165,
+      "loss": 6.6065,
+      "step": 2138
+    },
+    {
+      "epoch": 0.7300341296928328,
+      "grad_norm": 3.4596619606018066,
+      "learning_rate": 0.0007566552901023891,
+      "loss": 6.6794,
+      "step": 2139
+    },
+    {
+      "epoch": 0.7303754266211604,
+      "grad_norm": 3.4551479816436768,
+      "learning_rate": 0.0007565415244596132,
+      "loss": 6.7369,
+      "step": 2140
+    },
+    {
+      "epoch": 0.730716723549488,
+      "grad_norm": 3.431898832321167,
+      "learning_rate": 0.0007564277588168373,
+      "loss": 6.1183,
+      "step": 2141
+    },
+    {
+      "epoch": 0.7310580204778157,
+      "grad_norm": 3.323843002319336,
+      "learning_rate": 0.0007563139931740614,
+      "loss": 6.919,
+      "step": 2142
+    },
+    {
+      "epoch": 0.7313993174061434,
+      "grad_norm": 3.4173998832702637,
+      "learning_rate": 0.0007562002275312855,
+      "loss": 6.7127,
+      "step": 2143
+    },
+    {
+      "epoch": 0.731740614334471,
+      "grad_norm": 3.3746798038482666,
+      "learning_rate": 0.0007560864618885098,
+      "loss": 6.8423,
+      "step": 2144
+    },
+    {
+      "epoch": 0.7320819112627986,
+      "grad_norm": 3.6130669116973877,
+      "learning_rate": 0.0007559726962457339,
+      "loss": 6.3245,
+      "step": 2145
+    },
+    {
+      "epoch": 0.7324232081911263,
+      "grad_norm": 3.3271877765655518,
+      "learning_rate": 0.000755858930602958,
+      "loss": 6.885,
+      "step": 2146
+    },
+    {
+      "epoch": 0.7327645051194539,
+      "grad_norm": 4.328009605407715,
+      "learning_rate": 0.0007557451649601821,
+      "loss": 6.3472,
+      "step": 2147
+    },
+    {
+      "epoch": 0.7331058020477815,
+      "grad_norm": 3.4987595081329346,
+      "learning_rate": 0.0007556313993174062,
+      "loss": 6.632,
+      "step": 2148
+    },
+    {
+      "epoch": 0.7334470989761093,
+      "grad_norm": 3.967522144317627,
+      "learning_rate": 0.0007555176336746303,
+      "loss": 6.3009,
+      "step": 2149
+    },
+    {
+      "epoch": 0.7337883959044369,
+      "grad_norm": 4.190941333770752,
+      "learning_rate": 0.0007554038680318544,
+      "loss": 5.315,
+      "step": 2150
+    },
+    {
+      "epoch": 0.7341296928327645,
+      "grad_norm": 3.2750062942504883,
+      "learning_rate": 0.0007552901023890785,
+      "loss": 6.4132,
+      "step": 2151
+    },
+    {
+      "epoch": 0.7344709897610922,
+      "grad_norm": 3.1912786960601807,
+      "learning_rate": 0.0007551763367463026,
+      "loss": 6.4303,
+      "step": 2152
+    },
+    {
+      "epoch": 0.7348122866894198,
+      "grad_norm": 4.60476016998291,
+      "learning_rate": 0.0007550625711035267,
+      "loss": 5.7562,
+      "step": 2153
+    },
+    {
+      "epoch": 0.7351535836177474,
+      "grad_norm": 3.1308608055114746,
+      "learning_rate": 0.0007549488054607508,
+      "loss": 6.6797,
+      "step": 2154
+    },
+    {
+      "epoch": 0.735494880546075,
+      "grad_norm": 3.2094202041625977,
+      "learning_rate": 0.000754835039817975,
+      "loss": 6.6746,
+      "step": 2155
+    },
+    {
+      "epoch": 0.7358361774744028,
+      "grad_norm": 3.3604657649993896,
+      "learning_rate": 0.0007547212741751991,
+      "loss": 7.061,
+      "step": 2156
+    },
+    {
+      "epoch": 0.7361774744027304,
+      "grad_norm": 3.310800075531006,
+      "learning_rate": 0.0007546075085324232,
+      "loss": 6.4595,
+      "step": 2157
+    },
+    {
+      "epoch": 0.736518771331058,
+      "grad_norm": 4.224414348602295,
+      "learning_rate": 0.0007544937428896473,
+      "loss": 6.3834,
+      "step": 2158
+    },
+    {
+      "epoch": 0.7368600682593857,
+      "grad_norm": 3.4265730381011963,
+      "learning_rate": 0.0007543799772468714,
+      "loss": 6.8672,
+      "step": 2159
+    },
+    {
+      "epoch": 0.7372013651877133,
+      "grad_norm": 3.1829991340637207,
+      "learning_rate": 0.0007542662116040955,
+      "loss": 6.7422,
+      "step": 2160
+    },
+    {
+      "epoch": 0.7375426621160409,
+      "grad_norm": 3.3258779048919678,
+      "learning_rate": 0.0007541524459613198,
+      "loss": 6.8123,
+      "step": 2161
+    },
+    {
+      "epoch": 0.7378839590443687,
+      "grad_norm": 3.4259634017944336,
+      "learning_rate": 0.0007540386803185439,
+      "loss": 7.0327,
+      "step": 2162
+    },
+    {
+      "epoch": 0.7382252559726963,
+      "grad_norm": 6.536485195159912,
+      "learning_rate": 0.000753924914675768,
+      "loss": 6.6945,
+      "step": 2163
+    },
+    {
+      "epoch": 0.7385665529010239,
+      "grad_norm": 3.6783645153045654,
+      "learning_rate": 0.0007538111490329921,
+      "loss": 7.1871,
+      "step": 2164
+    },
+    {
+      "epoch": 0.7389078498293515,
+      "grad_norm": 3.5718960762023926,
+      "learning_rate": 0.0007536973833902162,
+      "loss": 6.322,
+      "step": 2165
+    },
+    {
+      "epoch": 0.7392491467576792,
+      "grad_norm": 3.513333320617676,
+      "learning_rate": 0.0007535836177474404,
+      "loss": 6.5015,
+      "step": 2166
+    },
+    {
+      "epoch": 0.7395904436860068,
+      "grad_norm": 3.306096315383911,
+      "learning_rate": 0.0007534698521046645,
+      "loss": 6.2813,
+      "step": 2167
+    },
+    {
+      "epoch": 0.7399317406143344,
+      "grad_norm": 4.2289557456970215,
+      "learning_rate": 0.0007533560864618886,
+      "loss": 6.128,
+      "step": 2168
+    },
+    {
+      "epoch": 0.7402730375426622,
+      "grad_norm": 3.4186182022094727,
+      "learning_rate": 0.0007532423208191126,
+      "loss": 6.4153,
+      "step": 2169
+    },
+    {
+      "epoch": 0.7406143344709898,
+      "grad_norm": 3.4397335052490234,
+      "learning_rate": 0.0007531285551763367,
+      "loss": 6.4691,
+      "step": 2170
+    },
+    {
+      "epoch": 0.7409556313993174,
+      "grad_norm": 3.2782340049743652,
+      "learning_rate": 0.0007530147895335608,
+      "loss": 6.8505,
+      "step": 2171
+    },
+    {
+      "epoch": 0.741296928327645,
+      "grad_norm": 3.247593879699707,
+      "learning_rate": 0.000752901023890785,
+      "loss": 6.429,
+      "step": 2172
+    },
+    {
+      "epoch": 0.7416382252559727,
+      "grad_norm": 3.29437518119812,
+      "learning_rate": 0.0007527872582480091,
+      "loss": 6.7258,
+      "step": 2173
+    },
+    {
+      "epoch": 0.7419795221843003,
+      "grad_norm": 3.4087376594543457,
+      "learning_rate": 0.0007526734926052332,
+      "loss": 7.0698,
+      "step": 2174
+    },
+    {
+      "epoch": 0.742320819112628,
+      "grad_norm": 3.289987564086914,
+      "learning_rate": 0.0007525597269624573,
+      "loss": 6.3238,
+      "step": 2175
+    },
+    {
+      "epoch": 0.7426621160409557,
+      "grad_norm": 3.370927095413208,
+      "learning_rate": 0.0007524459613196814,
+      "loss": 6.6566,
+      "step": 2176
+    },
+    {
+      "epoch": 0.7430034129692833,
+      "grad_norm": 3.518704891204834,
+      "learning_rate": 0.0007523321956769055,
+      "loss": 6.1031,
+      "step": 2177
+    },
+    {
+      "epoch": 0.7433447098976109,
+      "grad_norm": 3.2359862327575684,
+      "learning_rate": 0.0007522184300341298,
+      "loss": 6.4414,
+      "step": 2178
+    },
+    {
+      "epoch": 0.7436860068259386,
+      "grad_norm": 3.503610134124756,
+      "learning_rate": 0.0007521046643913539,
+      "loss": 6.473,
+      "step": 2179
+    },
+    {
+      "epoch": 0.7440273037542662,
+      "grad_norm": 3.644157886505127,
+      "learning_rate": 0.000751990898748578,
+      "loss": 6.7016,
+      "step": 2180
+    },
+    {
+      "epoch": 0.7443686006825938,
+      "grad_norm": 3.6388585567474365,
+      "learning_rate": 0.0007518771331058021,
+      "loss": 6.0456,
+      "step": 2181
+    },
+    {
+      "epoch": 0.7447098976109215,
+      "grad_norm": 3.494711399078369,
+      "learning_rate": 0.0007517633674630262,
+      "loss": 6.3996,
+      "step": 2182
+    },
+    {
+      "epoch": 0.7450511945392492,
+      "grad_norm": 3.3580422401428223,
+      "learning_rate": 0.0007516496018202503,
+      "loss": 6.791,
+      "step": 2183
+    },
+    {
+      "epoch": 0.7453924914675768,
+      "grad_norm": 3.2778096199035645,
+      "learning_rate": 0.0007515358361774745,
+      "loss": 6.8585,
+      "step": 2184
+    },
+    {
+      "epoch": 0.7457337883959044,
+      "grad_norm": 4.171144485473633,
+      "learning_rate": 0.0007514220705346986,
+      "loss": 6.5811,
+      "step": 2185
+    },
+    {
+      "epoch": 0.7460750853242321,
+      "grad_norm": 3.2704648971557617,
+      "learning_rate": 0.0007513083048919227,
+      "loss": 6.4436,
+      "step": 2186
+    },
+    {
+      "epoch": 0.7464163822525597,
+      "grad_norm": 3.2986881732940674,
+      "learning_rate": 0.0007511945392491468,
+      "loss": 6.3978,
+      "step": 2187
+    },
+    {
+      "epoch": 0.7467576791808874,
+      "grad_norm": 3.2707314491271973,
+      "learning_rate": 0.0007510807736063708,
+      "loss": 6.8469,
+      "step": 2188
+    },
+    {
+      "epoch": 0.747098976109215,
+      "grad_norm": 3.3278586864471436,
+      "learning_rate": 0.000750967007963595,
+      "loss": 6.7223,
+      "step": 2189
+    },
+    {
+      "epoch": 0.7474402730375427,
+      "grad_norm": 3.3085014820098877,
+      "learning_rate": 0.0007508532423208191,
+      "loss": 6.732,
+      "step": 2190
+    },
+    {
+      "epoch": 0.7477815699658703,
+      "grad_norm": 3.2527832984924316,
+      "learning_rate": 0.0007507394766780432,
+      "loss": 6.7635,
+      "step": 2191
+    },
+    {
+      "epoch": 0.7481228668941979,
+      "grad_norm": 4.035043239593506,
+      "learning_rate": 0.0007506257110352673,
+      "loss": 6.5922,
+      "step": 2192
+    },
+    {
+      "epoch": 0.7484641638225256,
+      "grad_norm": 3.383103370666504,
+      "learning_rate": 0.0007505119453924914,
+      "loss": 6.6097,
+      "step": 2193
+    },
+    {
+      "epoch": 0.7488054607508532,
+      "grad_norm": 3.518967628479004,
+      "learning_rate": 0.0007503981797497155,
+      "loss": 6.7073,
+      "step": 2194
+    },
+    {
+      "epoch": 0.7491467576791809,
+      "grad_norm": 3.142347574234009,
+      "learning_rate": 0.0007502844141069398,
+      "loss": 6.7196,
+      "step": 2195
+    },
+    {
+      "epoch": 0.7494880546075086,
+      "grad_norm": 3.3072586059570312,
+      "learning_rate": 0.0007501706484641639,
+      "loss": 7.2352,
+      "step": 2196
+    },
+    {
+      "epoch": 0.7498293515358362,
+      "grad_norm": 3.449594020843506,
+      "learning_rate": 0.000750056882821388,
+      "loss": 6.746,
+      "step": 2197
+    },
+    {
+      "epoch": 0.7501706484641638,
+      "grad_norm": 3.179112672805786,
+      "learning_rate": 0.0007499431171786121,
+      "loss": 6.3185,
+      "step": 2198
+    },
+    {
+      "epoch": 0.7505119453924914,
+      "grad_norm": 3.2998242378234863,
+      "learning_rate": 0.0007498293515358362,
+      "loss": 6.7703,
+      "step": 2199
+    },
+    {
+      "epoch": 0.7508532423208191,
+      "grad_norm": 5.338634967803955,
+      "learning_rate": 0.0007497155858930603,
+      "loss": 5.8161,
+      "step": 2200
+    },
+    {
+      "epoch": 0.7511945392491468,
+      "grad_norm": 4.291819095611572,
+      "learning_rate": 0.0007496018202502845,
+      "loss": 6.3253,
+      "step": 2201
+    },
+    {
+      "epoch": 0.7515358361774744,
+      "grad_norm": 3.579894542694092,
+      "learning_rate": 0.0007494880546075086,
+      "loss": 6.7618,
+      "step": 2202
+    },
+    {
+      "epoch": 0.7518771331058021,
+      "grad_norm": 3.397998809814453,
+      "learning_rate": 0.0007493742889647327,
+      "loss": 6.4193,
+      "step": 2203
+    },
+    {
+      "epoch": 0.7522184300341297,
+      "grad_norm": 3.181931972503662,
+      "learning_rate": 0.0007492605233219568,
+      "loss": 6.1566,
+      "step": 2204
+    },
+    {
+      "epoch": 0.7525597269624573,
+      "grad_norm": 3.316357374191284,
+      "learning_rate": 0.0007491467576791809,
+      "loss": 6.9517,
+      "step": 2205
+    },
+    {
+      "epoch": 0.752901023890785,
+      "grad_norm": 3.6711394786834717,
+      "learning_rate": 0.000749032992036405,
+      "loss": 6.699,
+      "step": 2206
+    },
+    {
+      "epoch": 0.7532423208191126,
+      "grad_norm": 3.2248027324676514,
+      "learning_rate": 0.0007489192263936292,
+      "loss": 6.7637,
+      "step": 2207
+    },
+    {
+      "epoch": 0.7535836177474403,
+      "grad_norm": 6.611885070800781,
+      "learning_rate": 0.0007488054607508532,
+      "loss": 5.6556,
+      "step": 2208
+    },
+    {
+      "epoch": 0.7539249146757679,
+      "grad_norm": 3.5705573558807373,
+      "learning_rate": 0.0007486916951080773,
+      "loss": 6.6186,
+      "step": 2209
+    },
+    {
+      "epoch": 0.7542662116040956,
+      "grad_norm": 3.3954081535339355,
+      "learning_rate": 0.0007485779294653014,
+      "loss": 6.6294,
+      "step": 2210
+    },
+    {
+      "epoch": 0.7546075085324232,
+      "grad_norm": 4.374399185180664,
+      "learning_rate": 0.0007484641638225255,
+      "loss": 5.2589,
+      "step": 2211
+    },
+    {
+      "epoch": 0.7549488054607508,
+      "grad_norm": 3.399153232574463,
+      "learning_rate": 0.0007483503981797498,
+      "loss": 7.0186,
+      "step": 2212
+    },
+    {
+      "epoch": 0.7552901023890785,
+      "grad_norm": 3.427119016647339,
+      "learning_rate": 0.0007482366325369739,
+      "loss": 6.4384,
+      "step": 2213
+    },
+    {
+      "epoch": 0.7556313993174062,
+      "grad_norm": 3.3054943084716797,
+      "learning_rate": 0.000748122866894198,
+      "loss": 6.8928,
+      "step": 2214
+    },
+    {
+      "epoch": 0.7559726962457338,
+      "grad_norm": 3.1258208751678467,
+      "learning_rate": 0.0007480091012514221,
+      "loss": 7.0228,
+      "step": 2215
+    },
+    {
+      "epoch": 0.7563139931740614,
+      "grad_norm": 3.197582721710205,
+      "learning_rate": 0.0007478953356086462,
+      "loss": 6.4303,
+      "step": 2216
+    },
+    {
+      "epoch": 0.7566552901023891,
+      "grad_norm": 3.482532262802124,
+      "learning_rate": 0.0007477815699658703,
+      "loss": 6.5166,
+      "step": 2217
+    },
+    {
+      "epoch": 0.7569965870307167,
+      "grad_norm": 3.3333423137664795,
+      "learning_rate": 0.0007476678043230945,
+      "loss": 6.8339,
+      "step": 2218
+    },
+    {
+      "epoch": 0.7573378839590443,
+      "grad_norm": 3.8367319107055664,
+      "learning_rate": 0.0007475540386803186,
+      "loss": 4.8938,
+      "step": 2219
+    },
+    {
+      "epoch": 0.757679180887372,
+      "grad_norm": 3.483358383178711,
+      "learning_rate": 0.0007474402730375427,
+      "loss": 6.3541,
+      "step": 2220
+    },
+    {
+      "epoch": 0.7580204778156997,
+      "grad_norm": 3.646254539489746,
+      "learning_rate": 0.0007473265073947668,
+      "loss": 6.2515,
+      "step": 2221
+    },
+    {
+      "epoch": 0.7583617747440273,
+      "grad_norm": 3.3919055461883545,
+      "learning_rate": 0.0007472127417519909,
+      "loss": 6.4064,
+      "step": 2222
+    },
+    {
+      "epoch": 0.758703071672355,
+      "grad_norm": 3.8102240562438965,
+      "learning_rate": 0.000747098976109215,
+      "loss": 6.0644,
+      "step": 2223
+    },
+    {
+      "epoch": 0.7590443686006826,
+      "grad_norm": 4.608096599578857,
+      "learning_rate": 0.0007469852104664392,
+      "loss": 6.7343,
+      "step": 2224
+    },
+    {
+      "epoch": 0.7593856655290102,
+      "grad_norm": 4.500402450561523,
+      "learning_rate": 0.0007468714448236633,
+      "loss": 6.7885,
+      "step": 2225
+    },
+    {
+      "epoch": 0.7597269624573378,
+      "grad_norm": 3.641897439956665,
+      "learning_rate": 0.0007467576791808874,
+      "loss": 7.0954,
+      "step": 2226
+    },
+    {
+      "epoch": 0.7600682593856656,
+      "grad_norm": 3.584850788116455,
+      "learning_rate": 0.0007466439135381114,
+      "loss": 6.7046,
+      "step": 2227
+    },
+    {
+      "epoch": 0.7604095563139932,
+      "grad_norm": 3.864504337310791,
+      "learning_rate": 0.0007465301478953355,
+      "loss": 6.564,
+      "step": 2228
+    },
+    {
+      "epoch": 0.7607508532423208,
+      "grad_norm": 3.3200836181640625,
+      "learning_rate": 0.0007464163822525596,
+      "loss": 6.8344,
+      "step": 2229
+    },
+    {
+      "epoch": 0.7610921501706485,
+      "grad_norm": 3.698155641555786,
+      "learning_rate": 0.0007463026166097839,
+      "loss": 6.0784,
+      "step": 2230
+    },
+    {
+      "epoch": 0.7614334470989761,
+      "grad_norm": 3.2318496704101562,
+      "learning_rate": 0.000746188850967008,
+      "loss": 6.6715,
+      "step": 2231
+    },
+    {
+      "epoch": 0.7617747440273037,
+      "grad_norm": 3.3073530197143555,
+      "learning_rate": 0.0007460750853242321,
+      "loss": 6.779,
+      "step": 2232
+    },
+    {
+      "epoch": 0.7621160409556313,
+      "grad_norm": 3.1656546592712402,
+      "learning_rate": 0.0007459613196814562,
+      "loss": 6.1155,
+      "step": 2233
+    },
+    {
+      "epoch": 0.7624573378839591,
+      "grad_norm": 3.251887798309326,
+      "learning_rate": 0.0007458475540386803,
+      "loss": 6.747,
+      "step": 2234
+    },
+    {
+      "epoch": 0.7627986348122867,
+      "grad_norm": 3.452404022216797,
+      "learning_rate": 0.0007457337883959045,
+      "loss": 6.3328,
+      "step": 2235
+    },
+    {
+      "epoch": 0.7631399317406143,
+      "grad_norm": 10.201861381530762,
+      "learning_rate": 0.0007456200227531286,
+      "loss": 5.8861,
+      "step": 2236
+    },
+    {
+      "epoch": 0.763481228668942,
+      "grad_norm": 6.577554225921631,
+      "learning_rate": 0.0007455062571103527,
+      "loss": 6.1329,
+      "step": 2237
+    },
+    {
+      "epoch": 0.7638225255972696,
+      "grad_norm": 4.222815036773682,
+      "learning_rate": 0.0007453924914675768,
+      "loss": 6.9536,
+      "step": 2238
+    },
+    {
+      "epoch": 0.7641638225255972,
+      "grad_norm": 4.045281887054443,
+      "learning_rate": 0.0007452787258248009,
+      "loss": 6.7163,
+      "step": 2239
+    },
+    {
+      "epoch": 0.764505119453925,
+      "grad_norm": 3.9914534091949463,
+      "learning_rate": 0.000745164960182025,
+      "loss": 6.5436,
+      "step": 2240
+    },
+    {
+      "epoch": 0.7648464163822526,
+      "grad_norm": 3.4769484996795654,
+      "learning_rate": 0.0007450511945392492,
+      "loss": 6.695,
+      "step": 2241
+    },
+    {
+      "epoch": 0.7651877133105802,
+      "grad_norm": 3.3896267414093018,
+      "learning_rate": 0.0007449374288964733,
+      "loss": 6.5388,
+      "step": 2242
+    },
+    {
+      "epoch": 0.7655290102389078,
+      "grad_norm": 3.0643441677093506,
+      "learning_rate": 0.0007448236632536974,
+      "loss": 6.0324,
+      "step": 2243
+    },
+    {
+      "epoch": 0.7658703071672355,
+      "grad_norm": 3.558314800262451,
+      "learning_rate": 0.0007447098976109215,
+      "loss": 6.1635,
+      "step": 2244
+    },
+    {
+      "epoch": 0.7662116040955631,
+      "grad_norm": 3.5593554973602295,
+      "learning_rate": 0.0007445961319681457,
+      "loss": 7.1386,
+      "step": 2245
+    },
+    {
+      "epoch": 0.7665529010238907,
+      "grad_norm": 7.941109657287598,
+      "learning_rate": 0.0007444823663253698,
+      "loss": 5.8192,
+      "step": 2246
+    },
+    {
+      "epoch": 0.7668941979522185,
+      "grad_norm": 3.61783504486084,
+      "learning_rate": 0.0007443686006825939,
+      "loss": 6.484,
+      "step": 2247
+    },
+    {
+      "epoch": 0.7672354948805461,
+      "grad_norm": 3.6058130264282227,
+      "learning_rate": 0.000744254835039818,
+      "loss": 6.908,
+      "step": 2248
+    },
+    {
+      "epoch": 0.7675767918088737,
+      "grad_norm": 3.377002477645874,
+      "learning_rate": 0.0007441410693970421,
+      "loss": 6.7491,
+      "step": 2249
+    },
+    {
+      "epoch": 0.7679180887372014,
+      "grad_norm": 3.3072400093078613,
+      "learning_rate": 0.0007440273037542662,
+      "loss": 6.5535,
+      "step": 2250
+    },
+    {
+      "epoch": 0.768259385665529,
+      "grad_norm": 5.5078020095825195,
+      "learning_rate": 0.0007439135381114903,
+      "loss": 6.8116,
+      "step": 2251
+    },
+    {
+      "epoch": 0.7686006825938566,
+      "grad_norm": 3.3346757888793945,
+      "learning_rate": 0.0007437997724687145,
+      "loss": 6.6911,
+      "step": 2252
+    },
+    {
+      "epoch": 0.7689419795221843,
+      "grad_norm": 3.3978588581085205,
+      "learning_rate": 0.0007436860068259386,
+      "loss": 6.6,
+      "step": 2253
+    },
+    {
+      "epoch": 0.769283276450512,
+      "grad_norm": 3.0739238262176514,
+      "learning_rate": 0.0007435722411831627,
+      "loss": 6.3782,
+      "step": 2254
+    },
+    {
+      "epoch": 0.7696245733788396,
+      "grad_norm": 3.2551000118255615,
+      "learning_rate": 0.0007434584755403868,
+      "loss": 6.6849,
+      "step": 2255
+    },
+    {
+      "epoch": 0.7699658703071672,
+      "grad_norm": 3.300963878631592,
+      "learning_rate": 0.0007433447098976109,
+      "loss": 6.8587,
+      "step": 2256
+    },
+    {
+      "epoch": 0.7703071672354949,
+      "grad_norm": 3.314493179321289,
+      "learning_rate": 0.000743230944254835,
+      "loss": 6.3657,
+      "step": 2257
+    },
+    {
+      "epoch": 0.7706484641638225,
+      "grad_norm": 3.863924264907837,
+      "learning_rate": 0.0007431171786120592,
+      "loss": 6.2232,
+      "step": 2258
+    },
+    {
+      "epoch": 0.7709897610921501,
+      "grad_norm": 3.727370023727417,
+      "learning_rate": 0.0007430034129692833,
+      "loss": 6.1062,
+      "step": 2259
+    },
+    {
+      "epoch": 0.7713310580204779,
+      "grad_norm": 3.3532567024230957,
+      "learning_rate": 0.0007428896473265074,
+      "loss": 6.0705,
+      "step": 2260
+    },
+    {
+      "epoch": 0.7716723549488055,
+      "grad_norm": 3.544203281402588,
+      "learning_rate": 0.0007427758816837315,
+      "loss": 6.8618,
+      "step": 2261
+    },
+    {
+      "epoch": 0.7720136518771331,
+      "grad_norm": 3.5925631523132324,
+      "learning_rate": 0.0007426621160409557,
+      "loss": 6.7984,
+      "step": 2262
+    },
+    {
+      "epoch": 0.7723549488054607,
+      "grad_norm": 11.028677940368652,
+      "learning_rate": 0.0007425483503981798,
+      "loss": 7.319,
+      "step": 2263
+    },
+    {
+      "epoch": 0.7726962457337884,
+      "grad_norm": 4.054958343505859,
+      "learning_rate": 0.000742434584755404,
+      "loss": 3.7261,
+      "step": 2264
+    },
+    {
+      "epoch": 0.773037542662116,
+      "grad_norm": 3.978557825088501,
+      "learning_rate": 0.0007423208191126281,
+      "loss": 6.3767,
+      "step": 2265
+    },
+    {
+      "epoch": 0.7733788395904437,
+      "grad_norm": 3.5264103412628174,
+      "learning_rate": 0.0007422070534698521,
+      "loss": 6.3895,
+      "step": 2266
+    },
+    {
+      "epoch": 0.7737201365187714,
+      "grad_norm": 5.270613193511963,
+      "learning_rate": 0.0007420932878270762,
+      "loss": 7.091,
+      "step": 2267
+    },
+    {
+      "epoch": 0.774061433447099,
+      "grad_norm": 9.161282539367676,
+      "learning_rate": 0.0007419795221843003,
+      "loss": 6.3023,
+      "step": 2268
+    },
+    {
+      "epoch": 0.7744027303754266,
+      "grad_norm": 3.5067977905273438,
+      "learning_rate": 0.0007418657565415244,
+      "loss": 7.0963,
+      "step": 2269
+    },
+    {
+      "epoch": 0.7747440273037542,
+      "grad_norm": 3.9335129261016846,
+      "learning_rate": 0.0007417519908987486,
+      "loss": 6.6442,
+      "step": 2270
+    },
+    {
+      "epoch": 0.7750853242320819,
+      "grad_norm": 3.2506704330444336,
+      "learning_rate": 0.0007416382252559727,
+      "loss": 6.592,
+      "step": 2271
+    },
+    {
+      "epoch": 0.7754266211604095,
+      "grad_norm": 3.383512258529663,
+      "learning_rate": 0.0007415244596131968,
+      "loss": 6.3122,
+      "step": 2272
+    },
+    {
+      "epoch": 0.7757679180887372,
+      "grad_norm": 4.869744300842285,
+      "learning_rate": 0.0007414106939704209,
+      "loss": 6.1108,
+      "step": 2273
+    },
+    {
+      "epoch": 0.7761092150170649,
+      "grad_norm": 9.724971771240234,
+      "learning_rate": 0.000741296928327645,
+      "loss": 6.1679,
+      "step": 2274
+    },
+    {
+      "epoch": 0.7764505119453925,
+      "grad_norm": 3.5860729217529297,
+      "learning_rate": 0.0007411831626848692,
+      "loss": 6.8647,
+      "step": 2275
+    },
+    {
+      "epoch": 0.7767918088737201,
+      "grad_norm": 3.8942580223083496,
+      "learning_rate": 0.0007410693970420933,
+      "loss": 6.8968,
+      "step": 2276
+    },
+    {
+      "epoch": 0.7771331058020478,
+      "grad_norm": 3.5359604358673096,
+      "learning_rate": 0.0007409556313993174,
+      "loss": 6.515,
+      "step": 2277
+    },
+    {
+      "epoch": 0.7774744027303754,
+      "grad_norm": 5.767436981201172,
+      "learning_rate": 0.0007408418657565415,
+      "loss": 6.4321,
+      "step": 2278
+    },
+    {
+      "epoch": 0.7778156996587031,
+      "grad_norm": 3.3587779998779297,
+      "learning_rate": 0.0007407281001137657,
+      "loss": 6.6002,
+      "step": 2279
+    },
+    {
+      "epoch": 0.7781569965870307,
+      "grad_norm": 3.4271020889282227,
+      "learning_rate": 0.0007406143344709898,
+      "loss": 6.5293,
+      "step": 2280
+    },
+    {
+      "epoch": 0.7784982935153584,
+      "grad_norm": 3.1101016998291016,
+      "learning_rate": 0.000740500568828214,
+      "loss": 6.7376,
+      "step": 2281
+    },
+    {
+      "epoch": 0.778839590443686,
+      "grad_norm": 9.123248100280762,
+      "learning_rate": 0.0007403868031854381,
+      "loss": 6.5933,
+      "step": 2282
+    },
+    {
+      "epoch": 0.7791808873720136,
+      "grad_norm": 3.5987279415130615,
+      "learning_rate": 0.0007402730375426622,
+      "loss": 5.8846,
+      "step": 2283
+    },
+    {
+      "epoch": 0.7795221843003413,
+      "grad_norm": 3.3567636013031006,
+      "learning_rate": 0.0007401592718998863,
+      "loss": 6.7901,
+      "step": 2284
+    },
+    {
+      "epoch": 0.7798634812286689,
+      "grad_norm": 3.2160773277282715,
+      "learning_rate": 0.0007400455062571104,
+      "loss": 6.4096,
+      "step": 2285
+    },
+    {
+      "epoch": 0.7802047781569966,
+      "grad_norm": 3.893007278442383,
+      "learning_rate": 0.0007399317406143344,
+      "loss": 6.9421,
+      "step": 2286
+    },
+    {
+      "epoch": 0.7805460750853243,
+      "grad_norm": 4.227537631988525,
+      "learning_rate": 0.0007398179749715586,
+      "loss": 6.5151,
+      "step": 2287
+    },
+    {
+      "epoch": 0.7808873720136519,
+      "grad_norm": 3.350599527359009,
+      "learning_rate": 0.0007397042093287827,
+      "loss": 6.3571,
+      "step": 2288
+    },
+    {
+      "epoch": 0.7812286689419795,
+      "grad_norm": 3.571381092071533,
+      "learning_rate": 0.0007395904436860068,
+      "loss": 6.446,
+      "step": 2289
+    },
+    {
+      "epoch": 0.7815699658703071,
+      "grad_norm": 3.227069139480591,
+      "learning_rate": 0.0007394766780432309,
+      "loss": 6.7134,
+      "step": 2290
+    },
+    {
+      "epoch": 0.7819112627986348,
+      "grad_norm": 5.10248327255249,
+      "learning_rate": 0.000739362912400455,
+      "loss": 6.0105,
+      "step": 2291
+    },
+    {
+      "epoch": 0.7822525597269625,
+      "grad_norm": 12.228768348693848,
+      "learning_rate": 0.0007392491467576792,
+      "loss": 6.8161,
+      "step": 2292
+    },
+    {
+      "epoch": 0.7825938566552901,
+      "grad_norm": 3.5377376079559326,
+      "learning_rate": 0.0007391353811149033,
+      "loss": 6.5361,
+      "step": 2293
+    },
+    {
+      "epoch": 0.7829351535836178,
+      "grad_norm": 3.385505437850952,
+      "learning_rate": 0.0007390216154721274,
+      "loss": 6.5231,
+      "step": 2294
+    },
+    {
+      "epoch": 0.7832764505119454,
+      "grad_norm": 3.796246290206909,
+      "learning_rate": 0.0007389078498293515,
+      "loss": 6.8123,
+      "step": 2295
+    },
+    {
+      "epoch": 0.783617747440273,
+      "grad_norm": 3.4816348552703857,
+      "learning_rate": 0.0007387940841865757,
+      "loss": 6.6924,
+      "step": 2296
+    },
+    {
+      "epoch": 0.7839590443686006,
+      "grad_norm": 3.1583478450775146,
+      "learning_rate": 0.0007386803185437998,
+      "loss": 6.6573,
+      "step": 2297
+    },
+    {
+      "epoch": 0.7843003412969284,
+      "grad_norm": 3.1136934757232666,
+      "learning_rate": 0.000738566552901024,
+      "loss": 6.679,
+      "step": 2298
+    },
+    {
+      "epoch": 0.784641638225256,
+      "grad_norm": 3.1441729068756104,
+      "learning_rate": 0.0007384527872582481,
+      "loss": 6.7502,
+      "step": 2299
+    },
+    {
+      "epoch": 0.7849829351535836,
+      "grad_norm": 3.35772442817688,
+      "learning_rate": 0.0007383390216154722,
+      "loss": 6.6187,
+      "step": 2300
+    },
+    {
+      "epoch": 0.7853242320819113,
+      "grad_norm": 3.171649694442749,
+      "learning_rate": 0.0007382252559726963,
+      "loss": 6.6412,
+      "step": 2301
+    },
+    {
+      "epoch": 0.7856655290102389,
+      "grad_norm": 3.1992034912109375,
+      "learning_rate": 0.0007381114903299204,
+      "loss": 6.5126,
+      "step": 2302
+    },
+    {
+      "epoch": 0.7860068259385665,
+      "grad_norm": 3.29266619682312,
+      "learning_rate": 0.0007379977246871445,
+      "loss": 6.8787,
+      "step": 2303
+    },
+    {
+      "epoch": 0.7863481228668942,
+      "grad_norm": 3.2478480339050293,
+      "learning_rate": 0.0007378839590443687,
+      "loss": 6.6584,
+      "step": 2304
+    },
+    {
+      "epoch": 0.7866894197952219,
+      "grad_norm": 4.051445007324219,
+      "learning_rate": 0.0007377701934015927,
+      "loss": 5.3193,
+      "step": 2305
+    },
+    {
+      "epoch": 0.7870307167235495,
+      "grad_norm": 3.5747928619384766,
+      "learning_rate": 0.0007376564277588168,
+      "loss": 6.6164,
+      "step": 2306
+    },
+    {
+      "epoch": 0.7873720136518771,
+      "grad_norm": 3.6790921688079834,
+      "learning_rate": 0.0007375426621160409,
+      "loss": 6.2069,
+      "step": 2307
+    },
+    {
+      "epoch": 0.7877133105802048,
+      "grad_norm": 3.410109519958496,
+      "learning_rate": 0.000737428896473265,
+      "loss": 6.6221,
+      "step": 2308
+    },
+    {
+      "epoch": 0.7880546075085324,
+      "grad_norm": 3.273460626602173,
+      "learning_rate": 0.0007373151308304891,
+      "loss": 6.8308,
+      "step": 2309
+    },
+    {
+      "epoch": 0.78839590443686,
+      "grad_norm": 3.7715392112731934,
+      "learning_rate": 0.0007372013651877133,
+      "loss": 6.3757,
+      "step": 2310
+    },
+    {
+      "epoch": 0.7887372013651878,
+      "grad_norm": 3.6163434982299805,
+      "learning_rate": 0.0007370875995449374,
+      "loss": 6.4616,
+      "step": 2311
+    },
+    {
+      "epoch": 0.7890784982935154,
+      "grad_norm": 3.2773282527923584,
+      "learning_rate": 0.0007369738339021615,
+      "loss": 6.6696,
+      "step": 2312
+    },
+    {
+      "epoch": 0.789419795221843,
+      "grad_norm": 3.1374948024749756,
+      "learning_rate": 0.0007368600682593857,
+      "loss": 6.7808,
+      "step": 2313
+    },
+    {
+      "epoch": 0.7897610921501707,
+      "grad_norm": 3.342297315597534,
+      "learning_rate": 0.0007367463026166098,
+      "loss": 6.3873,
+      "step": 2314
+    },
+    {
+      "epoch": 0.7901023890784983,
+      "grad_norm": 3.1699624061584473,
+      "learning_rate": 0.000736632536973834,
+      "loss": 6.458,
+      "step": 2315
+    },
+    {
+      "epoch": 0.7904436860068259,
+      "grad_norm": 3.4422099590301514,
+      "learning_rate": 0.0007365187713310581,
+      "loss": 6.8495,
+      "step": 2316
+    },
+    {
+      "epoch": 0.7907849829351535,
+      "grad_norm": 3.30281400680542,
+      "learning_rate": 0.0007364050056882822,
+      "loss": 6.2791,
+      "step": 2317
+    },
+    {
+      "epoch": 0.7911262798634813,
+      "grad_norm": 3.281485080718994,
+      "learning_rate": 0.0007362912400455063,
+      "loss": 6.733,
+      "step": 2318
+    },
+    {
+      "epoch": 0.7914675767918089,
+      "grad_norm": 3.2476160526275635,
+      "learning_rate": 0.0007361774744027304,
+      "loss": 7.0849,
+      "step": 2319
+    },
+    {
+      "epoch": 0.7918088737201365,
+      "grad_norm": 5.255532264709473,
+      "learning_rate": 0.0007360637087599545,
+      "loss": 5.6159,
+      "step": 2320
+    },
+    {
+      "epoch": 0.7921501706484642,
+      "grad_norm": 3.1768462657928467,
+      "learning_rate": 0.0007359499431171787,
+      "loss": 6.4978,
+      "step": 2321
+    },
+    {
+      "epoch": 0.7924914675767918,
+      "grad_norm": 3.386536121368408,
+      "learning_rate": 0.0007358361774744028,
+      "loss": 6.5782,
+      "step": 2322
+    },
+    {
+      "epoch": 0.7928327645051194,
+      "grad_norm": 3.406545877456665,
+      "learning_rate": 0.0007357224118316269,
+      "loss": 6.5096,
+      "step": 2323
+    },
+    {
+      "epoch": 0.7931740614334472,
+      "grad_norm": 3.1887378692626953,
+      "learning_rate": 0.0007356086461888509,
+      "loss": 6.5397,
+      "step": 2324
+    },
+    {
+      "epoch": 0.7935153583617748,
+      "grad_norm": 3.7656588554382324,
+      "learning_rate": 0.000735494880546075,
+      "loss": 6.4162,
+      "step": 2325
+    },
+    {
+      "epoch": 0.7938566552901024,
+      "grad_norm": 4.206430912017822,
+      "learning_rate": 0.0007353811149032991,
+      "loss": 6.1527,
+      "step": 2326
+    },
+    {
+      "epoch": 0.79419795221843,
+      "grad_norm": 3.5633111000061035,
+      "learning_rate": 0.0007352673492605233,
+      "loss": 6.1671,
+      "step": 2327
+    },
+    {
+      "epoch": 0.7945392491467577,
+      "grad_norm": 3.431990623474121,
+      "learning_rate": 0.0007351535836177474,
+      "loss": 6.4424,
+      "step": 2328
+    },
+    {
+      "epoch": 0.7948805460750853,
+      "grad_norm": 3.3288986682891846,
+      "learning_rate": 0.0007350398179749716,
+      "loss": 6.4327,
+      "step": 2329
+    },
+    {
+      "epoch": 0.7952218430034129,
+      "grad_norm": 5.098928928375244,
+      "learning_rate": 0.0007349260523321957,
+      "loss": 6.9967,
+      "step": 2330
+    },
+    {
+      "epoch": 0.7955631399317407,
+      "grad_norm": 10.33311939239502,
+      "learning_rate": 0.0007348122866894198,
+      "loss": 7.4892,
+      "step": 2331
+    },
+    {
+      "epoch": 0.7959044368600683,
+      "grad_norm": 3.6178691387176514,
+      "learning_rate": 0.000734698521046644,
+      "loss": 6.2764,
+      "step": 2332
+    },
+    {
+      "epoch": 0.7962457337883959,
+      "grad_norm": 3.6678335666656494,
+      "learning_rate": 0.0007345847554038681,
+      "loss": 6.4699,
+      "step": 2333
+    },
+    {
+      "epoch": 0.7965870307167235,
+      "grad_norm": 3.2506070137023926,
+      "learning_rate": 0.0007344709897610922,
+      "loss": 6.8747,
+      "step": 2334
+    },
+    {
+      "epoch": 0.7969283276450512,
+      "grad_norm": 12.323904037475586,
+      "learning_rate": 0.0007343572241183163,
+      "loss": 7.0181,
+      "step": 2335
+    },
+    {
+      "epoch": 0.7972696245733788,
+      "grad_norm": 3.266211986541748,
+      "learning_rate": 0.0007342434584755404,
+      "loss": 6.9827,
+      "step": 2336
+    },
+    {
+      "epoch": 0.7976109215017065,
+      "grad_norm": 3.527411699295044,
+      "learning_rate": 0.0007341296928327645,
+      "loss": 7.1384,
+      "step": 2337
+    },
+    {
+      "epoch": 0.7979522184300342,
+      "grad_norm": 3.395819902420044,
+      "learning_rate": 0.0007340159271899887,
+      "loss": 6.9804,
+      "step": 2338
+    },
+    {
+      "epoch": 0.7982935153583618,
+      "grad_norm": 3.189687728881836,
+      "learning_rate": 0.0007339021615472128,
+      "loss": 6.748,
+      "step": 2339
+    },
+    {
+      "epoch": 0.7986348122866894,
+      "grad_norm": 3.2057018280029297,
+      "learning_rate": 0.0007337883959044369,
+      "loss": 7.0284,
+      "step": 2340
+    },
+    {
+      "epoch": 0.798976109215017,
+      "grad_norm": 3.4170916080474854,
+      "learning_rate": 0.000733674630261661,
+      "loss": 6.6718,
+      "step": 2341
+    },
+    {
+      "epoch": 0.7993174061433447,
+      "grad_norm": 5.019681930541992,
+      "learning_rate": 0.0007335608646188851,
+      "loss": 4.0973,
+      "step": 2342
+    },
+    {
+      "epoch": 0.7996587030716723,
+      "grad_norm": 3.8981494903564453,
+      "learning_rate": 0.0007334470989761092,
+      "loss": 6.8531,
+      "step": 2343
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.444851875305176,
+      "learning_rate": 0.0007333333333333333,
+      "loss": 6.7379,
+      "step": 2344
+    },
+    {
+      "epoch": 0.8003412969283277,
+      "grad_norm": 3.404766082763672,
+      "learning_rate": 0.0007332195676905574,
+      "loss": 6.8825,
+      "step": 2345
+    },
+    {
+      "epoch": 0.8006825938566553,
+      "grad_norm": 3.660458564758301,
+      "learning_rate": 0.0007331058020477816,
+      "loss": 6.5001,
+      "step": 2346
+    },
+    {
+      "epoch": 0.8010238907849829,
+      "grad_norm": 3.563992738723755,
+      "learning_rate": 0.0007329920364050057,
+      "loss": 7.2701,
+      "step": 2347
+    },
+    {
+      "epoch": 0.8013651877133106,
+      "grad_norm": 3.2425544261932373,
+      "learning_rate": 0.0007328782707622298,
+      "loss": 6.3414,
+      "step": 2348
+    },
+    {
+      "epoch": 0.8017064846416382,
+      "grad_norm": 3.2002296447753906,
+      "learning_rate": 0.0007327645051194539,
+      "loss": 6.5811,
+      "step": 2349
+    },
+    {
+      "epoch": 0.8020477815699659,
+      "grad_norm": 6.365078926086426,
+      "learning_rate": 0.0007326507394766781,
+      "loss": 5.5698,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8023890784982936,
+      "grad_norm": 3.3614845275878906,
+      "learning_rate": 0.0007325369738339022,
+      "loss": 6.7194,
+      "step": 2351
+    },
+    {
+      "epoch": 0.8027303754266212,
+      "grad_norm": 3.386559247970581,
+      "learning_rate": 0.0007324232081911263,
+      "loss": 6.7674,
+      "step": 2352
+    },
+    {
+      "epoch": 0.8030716723549488,
+      "grad_norm": 3.318051338195801,
+      "learning_rate": 0.0007323094425483504,
+      "loss": 6.8814,
+      "step": 2353
+    },
+    {
+      "epoch": 0.8034129692832764,
+      "grad_norm": 3.2678332328796387,
+      "learning_rate": 0.0007321956769055745,
+      "loss": 6.7354,
+      "step": 2354
+    },
+    {
+      "epoch": 0.8037542662116041,
+      "grad_norm": 3.1607394218444824,
+      "learning_rate": 0.0007320819112627987,
+      "loss": 6.4444,
+      "step": 2355
+    },
+    {
+      "epoch": 0.8040955631399317,
+      "grad_norm": 3.887561798095703,
+      "learning_rate": 0.0007319681456200228,
+      "loss": 6.402,
+      "step": 2356
+    },
+    {
+      "epoch": 0.8044368600682594,
+      "grad_norm": 3.33502459526062,
+      "learning_rate": 0.0007318543799772469,
+      "loss": 6.8184,
+      "step": 2357
+    },
+    {
+      "epoch": 0.8047781569965871,
+      "grad_norm": 3.411867380142212,
+      "learning_rate": 0.000731740614334471,
+      "loss": 6.712,
+      "step": 2358
+    },
+    {
+      "epoch": 0.8051194539249147,
+      "grad_norm": 3.4380221366882324,
+      "learning_rate": 0.0007316268486916951,
+      "loss": 6.7389,
+      "step": 2359
+    },
+    {
+      "epoch": 0.8054607508532423,
+      "grad_norm": 4.988135814666748,
+      "learning_rate": 0.0007315130830489192,
+      "loss": 5.3108,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8058020477815699,
+      "grad_norm": 3.4688353538513184,
+      "learning_rate": 0.0007313993174061435,
+      "loss": 6.6813,
+      "step": 2361
+    },
+    {
+      "epoch": 0.8061433447098976,
+      "grad_norm": 3.7119762897491455,
+      "learning_rate": 0.0007312855517633676,
+      "loss": 6.419,
+      "step": 2362
+    },
+    {
+      "epoch": 0.8064846416382253,
+      "grad_norm": 4.025672912597656,
+      "learning_rate": 0.0007311717861205916,
+      "loss": 6.0607,
+      "step": 2363
+    },
+    {
+      "epoch": 0.8068259385665529,
+      "grad_norm": 3.4323062896728516,
+      "learning_rate": 0.0007310580204778157,
+      "loss": 6.8558,
+      "step": 2364
+    },
+    {
+      "epoch": 0.8071672354948806,
+      "grad_norm": 3.3362526893615723,
+      "learning_rate": 0.0007309442548350398,
+      "loss": 6.2351,
+      "step": 2365
+    },
+    {
+      "epoch": 0.8075085324232082,
+      "grad_norm": 3.44800066947937,
+      "learning_rate": 0.0007308304891922639,
+      "loss": 6.2169,
+      "step": 2366
+    },
+    {
+      "epoch": 0.8078498293515358,
+      "grad_norm": 3.3122382164001465,
+      "learning_rate": 0.0007307167235494881,
+      "loss": 7.1405,
+      "step": 2367
+    },
+    {
+      "epoch": 0.8081911262798634,
+      "grad_norm": 3.3105790615081787,
+      "learning_rate": 0.0007306029579067122,
+      "loss": 6.7058,
+      "step": 2368
+    },
+    {
+      "epoch": 0.8085324232081911,
+      "grad_norm": 3.940232038497925,
+      "learning_rate": 0.0007304891922639363,
+      "loss": 6.7658,
+      "step": 2369
+    },
+    {
+      "epoch": 0.8088737201365188,
+      "grad_norm": 3.7123522758483887,
+      "learning_rate": 0.0007303754266211604,
+      "loss": 6.2908,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8092150170648464,
+      "grad_norm": 3.907029151916504,
+      "learning_rate": 0.0007302616609783845,
+      "loss": 6.2083,
+      "step": 2371
+    },
+    {
+      "epoch": 0.8095563139931741,
+      "grad_norm": 3.455446481704712,
+      "learning_rate": 0.0007301478953356086,
+      "loss": 6.6809,
+      "step": 2372
+    },
+    {
+      "epoch": 0.8098976109215017,
+      "grad_norm": 3.3250789642333984,
+      "learning_rate": 0.0007300341296928328,
+      "loss": 6.6732,
+      "step": 2373
+    },
+    {
+      "epoch": 0.8102389078498293,
+      "grad_norm": 3.413405656814575,
+      "learning_rate": 0.0007299203640500569,
+      "loss": 6.0811,
+      "step": 2374
+    },
+    {
+      "epoch": 0.810580204778157,
+      "grad_norm": 3.436432361602783,
+      "learning_rate": 0.000729806598407281,
+      "loss": 6.8951,
+      "step": 2375
+    },
+    {
+      "epoch": 0.8109215017064847,
+      "grad_norm": 3.539832592010498,
+      "learning_rate": 0.0007296928327645051,
+      "loss": 6.9036,
+      "step": 2376
+    },
+    {
+      "epoch": 0.8112627986348123,
+      "grad_norm": 3.5589230060577393,
+      "learning_rate": 0.0007295790671217292,
+      "loss": 6.6657,
+      "step": 2377
+    },
+    {
+      "epoch": 0.81160409556314,
+      "grad_norm": 5.298614025115967,
+      "learning_rate": 0.0007294653014789535,
+      "loss": 5.9221,
+      "step": 2378
+    },
+    {
+      "epoch": 0.8119453924914676,
+      "grad_norm": 3.4280951023101807,
+      "learning_rate": 0.0007293515358361776,
+      "loss": 6.775,
+      "step": 2379
+    },
+    {
+      "epoch": 0.8122866894197952,
+      "grad_norm": 3.2987239360809326,
+      "learning_rate": 0.0007292377701934017,
+      "loss": 6.5316,
+      "step": 2380
+    },
+    {
+      "epoch": 0.8126279863481228,
+      "grad_norm": 4.67313289642334,
+      "learning_rate": 0.0007291240045506258,
+      "loss": 6.9242,
+      "step": 2381
+    },
+    {
+      "epoch": 0.8129692832764505,
+      "grad_norm": 3.3883895874023438,
+      "learning_rate": 0.0007290102389078499,
+      "loss": 6.9792,
+      "step": 2382
+    },
+    {
+      "epoch": 0.8133105802047782,
+      "grad_norm": 3.6097097396850586,
+      "learning_rate": 0.0007288964732650739,
+      "loss": 6.0067,
+      "step": 2383
+    },
+    {
+      "epoch": 0.8136518771331058,
+      "grad_norm": 3.708559989929199,
+      "learning_rate": 0.0007287827076222981,
+      "loss": 6.6121,
+      "step": 2384
+    },
+    {
+      "epoch": 0.8139931740614335,
+      "grad_norm": 3.4102509021759033,
+      "learning_rate": 0.0007286689419795222,
+      "loss": 6.7611,
+      "step": 2385
+    },
+    {
+      "epoch": 0.8143344709897611,
+      "grad_norm": 3.181281566619873,
+      "learning_rate": 0.0007285551763367463,
+      "loss": 6.5479,
+      "step": 2386
+    },
+    {
+      "epoch": 0.8146757679180887,
+      "grad_norm": 3.2941017150878906,
+      "learning_rate": 0.0007284414106939704,
+      "loss": 6.9345,
+      "step": 2387
+    },
+    {
+      "epoch": 0.8150170648464163,
+      "grad_norm": 3.4857585430145264,
+      "learning_rate": 0.0007283276450511945,
+      "loss": 6.2137,
+      "step": 2388
+    },
+    {
+      "epoch": 0.8153583617747441,
+      "grad_norm": 3.3792757987976074,
+      "learning_rate": 0.0007282138794084186,
+      "loss": 6.9875,
+      "step": 2389
+    },
+    {
+      "epoch": 0.8156996587030717,
+      "grad_norm": 3.3115673065185547,
+      "learning_rate": 0.0007281001137656428,
+      "loss": 6.9524,
+      "step": 2390
+    },
+    {
+      "epoch": 0.8160409556313993,
+      "grad_norm": 3.5123884677886963,
+      "learning_rate": 0.0007279863481228669,
+      "loss": 6.3792,
+      "step": 2391
+    },
+    {
+      "epoch": 0.816382252559727,
+      "grad_norm": 3.4690353870391846,
+      "learning_rate": 0.000727872582480091,
+      "loss": 7.0668,
+      "step": 2392
+    },
+    {
+      "epoch": 0.8167235494880546,
+      "grad_norm": 3.470229387283325,
+      "learning_rate": 0.0007277588168373151,
+      "loss": 6.5134,
+      "step": 2393
+    },
+    {
+      "epoch": 0.8170648464163822,
+      "grad_norm": 3.3032960891723633,
+      "learning_rate": 0.0007276450511945392,
+      "loss": 7.0573,
+      "step": 2394
+    },
+    {
+      "epoch": 0.8174061433447098,
+      "grad_norm": 6.7078094482421875,
+      "learning_rate": 0.0007275312855517635,
+      "loss": 6.4783,
+      "step": 2395
+    },
+    {
+      "epoch": 0.8177474402730376,
+      "grad_norm": 3.469536304473877,
+      "learning_rate": 0.0007274175199089876,
+      "loss": 6.6776,
+      "step": 2396
+    },
+    {
+      "epoch": 0.8180887372013652,
+      "grad_norm": 3.5298237800598145,
+      "learning_rate": 0.0007273037542662117,
+      "loss": 6.3748,
+      "step": 2397
+    },
+    {
+      "epoch": 0.8184300341296928,
+      "grad_norm": 3.280069589614868,
+      "learning_rate": 0.0007271899886234358,
+      "loss": 6.5264,
+      "step": 2398
+    },
+    {
+      "epoch": 0.8187713310580205,
+      "grad_norm": 3.1752591133117676,
+      "learning_rate": 0.0007270762229806599,
+      "loss": 6.656,
+      "step": 2399
+    },
+    {
+      "epoch": 0.8191126279863481,
+      "grad_norm": 3.6475753784179688,
+      "learning_rate": 0.000726962457337884,
+      "loss": 6.703,
+      "step": 2400
+    },
+    {
+      "epoch": 0.8194539249146757,
+      "grad_norm": 3.4017770290374756,
+      "learning_rate": 0.0007268486916951082,
+      "loss": 6.6566,
+      "step": 2401
+    },
+    {
+      "epoch": 0.8197952218430035,
+      "grad_norm": 3.364267349243164,
+      "learning_rate": 0.0007267349260523322,
+      "loss": 6.5412,
+      "step": 2402
+    },
+    {
+      "epoch": 0.8201365187713311,
+      "grad_norm": 3.2577364444732666,
+      "learning_rate": 0.0007266211604095563,
+      "loss": 6.4592,
+      "step": 2403
+    },
+    {
+      "epoch": 0.8204778156996587,
+      "grad_norm": 3.6079461574554443,
+      "learning_rate": 0.0007265073947667804,
+      "loss": 6.4137,
+      "step": 2404
+    },
+    {
+      "epoch": 0.8208191126279863,
+      "grad_norm": 3.259861946105957,
+      "learning_rate": 0.0007263936291240045,
+      "loss": 6.5796,
+      "step": 2405
+    },
+    {
+      "epoch": 0.821160409556314,
+      "grad_norm": 3.173610210418701,
+      "learning_rate": 0.0007262798634812286,
+      "loss": 6.9958,
+      "step": 2406
+    },
+    {
+      "epoch": 0.8215017064846416,
+      "grad_norm": 3.753018856048584,
+      "learning_rate": 0.0007261660978384528,
+      "loss": 6.7856,
+      "step": 2407
+    },
+    {
+      "epoch": 0.8218430034129692,
+      "grad_norm": 4.427280426025391,
+      "learning_rate": 0.0007260523321956769,
+      "loss": 6.5343,
+      "step": 2408
+    },
+    {
+      "epoch": 0.822184300341297,
+      "grad_norm": 3.921279191970825,
+      "learning_rate": 0.000725938566552901,
+      "loss": 6.1591,
+      "step": 2409
+    },
+    {
+      "epoch": 0.8225255972696246,
+      "grad_norm": 3.6298179626464844,
+      "learning_rate": 0.0007258248009101251,
+      "loss": 6.6855,
+      "step": 2410
+    },
+    {
+      "epoch": 0.8228668941979522,
+      "grad_norm": 4.846485137939453,
+      "learning_rate": 0.0007257110352673492,
+      "loss": 5.8797,
+      "step": 2411
+    },
+    {
+      "epoch": 0.8232081911262799,
+      "grad_norm": 3.467717409133911,
+      "learning_rate": 0.0007255972696245733,
+      "loss": 6.3911,
+      "step": 2412
+    },
+    {
+      "epoch": 0.8235494880546075,
+      "grad_norm": 3.697725534439087,
+      "learning_rate": 0.0007254835039817976,
+      "loss": 6.4286,
+      "step": 2413
+    },
+    {
+      "epoch": 0.8238907849829351,
+      "grad_norm": 3.694342851638794,
+      "learning_rate": 0.0007253697383390217,
+      "loss": 6.1186,
+      "step": 2414
+    },
+    {
+      "epoch": 0.8242320819112628,
+      "grad_norm": 3.3936257362365723,
+      "learning_rate": 0.0007252559726962458,
+      "loss": 6.7766,
+      "step": 2415
+    },
+    {
+      "epoch": 0.8245733788395905,
+      "grad_norm": 3.418151617050171,
+      "learning_rate": 0.0007251422070534699,
+      "loss": 6.3386,
+      "step": 2416
+    },
+    {
+      "epoch": 0.8249146757679181,
+      "grad_norm": 3.6463325023651123,
+      "learning_rate": 0.000725028441410694,
+      "loss": 6.1039,
+      "step": 2417
+    },
+    {
+      "epoch": 0.8252559726962457,
+      "grad_norm": 3.3325319290161133,
+      "learning_rate": 0.0007249146757679182,
+      "loss": 6.4488,
+      "step": 2418
+    },
+    {
+      "epoch": 0.8255972696245734,
+      "grad_norm": 3.4702980518341064,
+      "learning_rate": 0.0007248009101251423,
+      "loss": 5.9765,
+      "step": 2419
+    },
+    {
+      "epoch": 0.825938566552901,
+      "grad_norm": 3.6678733825683594,
+      "learning_rate": 0.0007246871444823664,
+      "loss": 6.4048,
+      "step": 2420
+    },
+    {
+      "epoch": 0.8262798634812286,
+      "grad_norm": 3.4993784427642822,
+      "learning_rate": 0.0007245733788395905,
+      "loss": 7.3037,
+      "step": 2421
+    },
+    {
+      "epoch": 0.8266211604095564,
+      "grad_norm": 3.750591516494751,
+      "learning_rate": 0.0007244596131968145,
+      "loss": 6.4685,
+      "step": 2422
+    },
+    {
+      "epoch": 0.826962457337884,
+      "grad_norm": 3.406001567840576,
+      "learning_rate": 0.0007243458475540386,
+      "loss": 6.5744,
+      "step": 2423
+    },
+    {
+      "epoch": 0.8273037542662116,
+      "grad_norm": 3.754840135574341,
+      "learning_rate": 0.0007242320819112628,
+      "loss": 6.0161,
+      "step": 2424
+    },
+    {
+      "epoch": 0.8276450511945392,
+      "grad_norm": 4.734508514404297,
+      "learning_rate": 0.0007241183162684869,
+      "loss": 6.0864,
+      "step": 2425
+    },
+    {
+      "epoch": 0.8279863481228669,
+      "grad_norm": 3.3035054206848145,
+      "learning_rate": 0.000724004550625711,
+      "loss": 6.7669,
+      "step": 2426
+    },
+    {
+      "epoch": 0.8283276450511945,
+      "grad_norm": 3.275688648223877,
+      "learning_rate": 0.0007238907849829351,
+      "loss": 6.5271,
+      "step": 2427
+    },
+    {
+      "epoch": 0.8286689419795222,
+      "grad_norm": 3.2510900497436523,
+      "learning_rate": 0.0007237770193401592,
+      "loss": 6.3089,
+      "step": 2428
+    },
+    {
+      "epoch": 0.8290102389078499,
+      "grad_norm": 4.378868579864502,
+      "learning_rate": 0.0007236632536973833,
+      "loss": 5.3725,
+      "step": 2429
+    },
+    {
+      "epoch": 0.8293515358361775,
+      "grad_norm": 3.4092857837677,
+      "learning_rate": 0.0007235494880546076,
+      "loss": 6.4763,
+      "step": 2430
+    },
+    {
+      "epoch": 0.8296928327645051,
+      "grad_norm": 3.2129056453704834,
+      "learning_rate": 0.0007234357224118317,
+      "loss": 6.4405,
+      "step": 2431
+    },
+    {
+      "epoch": 0.8300341296928327,
+      "grad_norm": 3.3604624271392822,
+      "learning_rate": 0.0007233219567690558,
+      "loss": 6.5199,
+      "step": 2432
+    },
+    {
+      "epoch": 0.8303754266211604,
+      "grad_norm": 3.6818511486053467,
+      "learning_rate": 0.0007232081911262799,
+      "loss": 6.1862,
+      "step": 2433
+    },
+    {
+      "epoch": 0.830716723549488,
+      "grad_norm": 3.174171209335327,
+      "learning_rate": 0.000723094425483504,
+      "loss": 6.5523,
+      "step": 2434
+    },
+    {
+      "epoch": 0.8310580204778157,
+      "grad_norm": 3.3606784343719482,
+      "learning_rate": 0.0007229806598407282,
+      "loss": 6.6429,
+      "step": 2435
+    },
+    {
+      "epoch": 0.8313993174061434,
+      "grad_norm": 3.436796188354492,
+      "learning_rate": 0.0007228668941979523,
+      "loss": 6.749,
+      "step": 2436
+    },
+    {
+      "epoch": 0.831740614334471,
+      "grad_norm": 3.20074462890625,
+      "learning_rate": 0.0007227531285551764,
+      "loss": 6.6947,
+      "step": 2437
+    },
+    {
+      "epoch": 0.8320819112627986,
+      "grad_norm": 3.3540759086608887,
+      "learning_rate": 0.0007226393629124005,
+      "loss": 6.7936,
+      "step": 2438
+    },
+    {
+      "epoch": 0.8324232081911263,
+      "grad_norm": 12.003594398498535,
+      "learning_rate": 0.0007225255972696246,
+      "loss": 6.054,
+      "step": 2439
+    },
+    {
+      "epoch": 0.8327645051194539,
+      "grad_norm": 3.418896198272705,
+      "learning_rate": 0.0007224118316268487,
+      "loss": 6.8313,
+      "step": 2440
+    },
+    {
+      "epoch": 0.8331058020477816,
+      "grad_norm": 3.5334503650665283,
+      "learning_rate": 0.0007222980659840728,
+      "loss": 6.7967,
+      "step": 2441
+    },
+    {
+      "epoch": 0.8334470989761092,
+      "grad_norm": 3.854482412338257,
+      "learning_rate": 0.0007221843003412969,
+      "loss": 5.6593,
+      "step": 2442
+    },
+    {
+      "epoch": 0.8337883959044369,
+      "grad_norm": 3.3690783977508545,
+      "learning_rate": 0.000722070534698521,
+      "loss": 6.9381,
+      "step": 2443
+    },
+    {
+      "epoch": 0.8341296928327645,
+      "grad_norm": 4.998669624328613,
+      "learning_rate": 0.0007219567690557451,
+      "loss": 5.7073,
+      "step": 2444
+    },
+    {
+      "epoch": 0.8344709897610921,
+      "grad_norm": 3.3754959106445312,
+      "learning_rate": 0.0007218430034129692,
+      "loss": 6.9455,
+      "step": 2445
+    },
+    {
+      "epoch": 0.8348122866894198,
+      "grad_norm": 3.3395328521728516,
+      "learning_rate": 0.0007217292377701933,
+      "loss": 6.6405,
+      "step": 2446
+    },
+    {
+      "epoch": 0.8351535836177474,
+      "grad_norm": 3.234178066253662,
+      "learning_rate": 0.0007216154721274176,
+      "loss": 7.0086,
+      "step": 2447
+    },
+    {
+      "epoch": 0.8354948805460751,
+      "grad_norm": 3.2380173206329346,
+      "learning_rate": 0.0007215017064846417,
+      "loss": 6.4784,
+      "step": 2448
+    },
+    {
+      "epoch": 0.8358361774744028,
+      "grad_norm": 3.333740472793579,
+      "learning_rate": 0.0007213879408418658,
+      "loss": 6.5965,
+      "step": 2449
+    },
+    {
+      "epoch": 0.8361774744027304,
+      "grad_norm": 3.263247489929199,
+      "learning_rate": 0.0007212741751990899,
+      "loss": 6.589,
+      "step": 2450
+    },
+    {
+      "epoch": 0.836518771331058,
+      "grad_norm": 3.6883716583251953,
+      "learning_rate": 0.000721160409556314,
+      "loss": 6.2031,
+      "step": 2451
+    },
+    {
+      "epoch": 0.8368600682593856,
+      "grad_norm": 3.3364152908325195,
+      "learning_rate": 0.0007210466439135381,
+      "loss": 6.6068,
+      "step": 2452
+    },
+    {
+      "epoch": 0.8372013651877133,
+      "grad_norm": 3.7093443870544434,
+      "learning_rate": 0.0007209328782707623,
+      "loss": 6.4188,
+      "step": 2453
+    },
+    {
+      "epoch": 0.837542662116041,
+      "grad_norm": 3.4039225578308105,
+      "learning_rate": 0.0007208191126279864,
+      "loss": 6.5616,
+      "step": 2454
+    },
+    {
+      "epoch": 0.8378839590443686,
+      "grad_norm": 3.357640027999878,
+      "learning_rate": 0.0007207053469852105,
+      "loss": 6.5854,
+      "step": 2455
+    },
+    {
+      "epoch": 0.8382252559726963,
+      "grad_norm": 4.375115871429443,
+      "learning_rate": 0.0007205915813424346,
+      "loss": 5.801,
+      "step": 2456
+    },
+    {
+      "epoch": 0.8385665529010239,
+      "grad_norm": 3.4280662536621094,
+      "learning_rate": 0.0007204778156996587,
+      "loss": 6.8118,
+      "step": 2457
+    },
+    {
+      "epoch": 0.8389078498293515,
+      "grad_norm": 3.6396632194519043,
+      "learning_rate": 0.0007203640500568829,
+      "loss": 6.6882,
+      "step": 2458
+    },
+    {
+      "epoch": 0.8392491467576791,
+      "grad_norm": 3.238856792449951,
+      "learning_rate": 0.000720250284414107,
+      "loss": 6.3282,
+      "step": 2459
+    },
+    {
+      "epoch": 0.8395904436860068,
+      "grad_norm": 3.151153087615967,
+      "learning_rate": 0.0007201365187713311,
+      "loss": 6.816,
+      "step": 2460
+    },
+    {
+      "epoch": 0.8399317406143345,
+      "grad_norm": 4.557154178619385,
+      "learning_rate": 0.0007200227531285551,
+      "loss": 6.6596,
+      "step": 2461
+    },
+    {
+      "epoch": 0.8402730375426621,
+      "grad_norm": 3.225273609161377,
+      "learning_rate": 0.0007199089874857792,
+      "loss": 6.5537,
+      "step": 2462
+    },
+    {
+      "epoch": 0.8406143344709898,
+      "grad_norm": 3.4817261695861816,
+      "learning_rate": 0.0007197952218430033,
+      "loss": 6.567,
+      "step": 2463
+    },
+    {
+      "epoch": 0.8409556313993174,
+      "grad_norm": 3.3386945724487305,
+      "learning_rate": 0.0007196814562002276,
+      "loss": 7.1524,
+      "step": 2464
+    },
+    {
+      "epoch": 0.841296928327645,
+      "grad_norm": 3.4559755325317383,
+      "learning_rate": 0.0007195676905574517,
+      "loss": 6.4858,
+      "step": 2465
+    },
+    {
+      "epoch": 0.8416382252559726,
+      "grad_norm": 3.230203866958618,
+      "learning_rate": 0.0007194539249146758,
+      "loss": 6.6456,
+      "step": 2466
+    },
+    {
+      "epoch": 0.8419795221843004,
+      "grad_norm": 3.321943998336792,
+      "learning_rate": 0.0007193401592718999,
+      "loss": 6.6494,
+      "step": 2467
+    },
+    {
+      "epoch": 0.842320819112628,
+      "grad_norm": 3.304291248321533,
+      "learning_rate": 0.000719226393629124,
+      "loss": 6.7956,
+      "step": 2468
+    },
+    {
+      "epoch": 0.8426621160409556,
+      "grad_norm": 3.3571488857269287,
+      "learning_rate": 0.0007191126279863481,
+      "loss": 6.8112,
+      "step": 2469
+    },
+    {
+      "epoch": 0.8430034129692833,
+      "grad_norm": 3.388721466064453,
+      "learning_rate": 0.0007189988623435723,
+      "loss": 6.8564,
+      "step": 2470
+    },
+    {
+      "epoch": 0.8433447098976109,
+      "grad_norm": 3.430446147918701,
+      "learning_rate": 0.0007188850967007964,
+      "loss": 6.7175,
+      "step": 2471
+    },
+    {
+      "epoch": 0.8436860068259385,
+      "grad_norm": 3.3103549480438232,
+      "learning_rate": 0.0007187713310580205,
+      "loss": 6.5384,
+      "step": 2472
+    },
+    {
+      "epoch": 0.8440273037542663,
+      "grad_norm": 7.369909763336182,
+      "learning_rate": 0.0007186575654152446,
+      "loss": 5.398,
+      "step": 2473
+    },
+    {
+      "epoch": 0.8443686006825939,
+      "grad_norm": 3.683732748031616,
+      "learning_rate": 0.0007185437997724687,
+      "loss": 6.4451,
+      "step": 2474
+    },
+    {
+      "epoch": 0.8447098976109215,
+      "grad_norm": 4.542445659637451,
+      "learning_rate": 0.0007184300341296928,
+      "loss": 6.101,
+      "step": 2475
+    },
+    {
+      "epoch": 0.8450511945392492,
+      "grad_norm": 4.248722553253174,
+      "learning_rate": 0.000718316268486917,
+      "loss": 6.0299,
+      "step": 2476
+    },
+    {
+      "epoch": 0.8453924914675768,
+      "grad_norm": 3.940197467803955,
+      "learning_rate": 0.0007182025028441411,
+      "loss": 5.5218,
+      "step": 2477
+    },
+    {
+      "epoch": 0.8457337883959044,
+      "grad_norm": 3.645324945449829,
+      "learning_rate": 0.0007180887372013652,
+      "loss": 6.9645,
+      "step": 2478
+    },
+    {
+      "epoch": 0.846075085324232,
+      "grad_norm": 5.713097095489502,
+      "learning_rate": 0.0007179749715585894,
+      "loss": 5.7437,
+      "step": 2479
+    },
+    {
+      "epoch": 0.8464163822525598,
+      "grad_norm": 3.8885574340820312,
+      "learning_rate": 0.0007178612059158133,
+      "loss": 6.7958,
+      "step": 2480
+    },
+    {
+      "epoch": 0.8467576791808874,
+      "grad_norm": 3.430229902267456,
+      "learning_rate": 0.0007177474402730376,
+      "loss": 6.7616,
+      "step": 2481
+    },
+    {
+      "epoch": 0.847098976109215,
+      "grad_norm": 3.286057472229004,
+      "learning_rate": 0.0007176336746302617,
+      "loss": 6.6936,
+      "step": 2482
+    },
+    {
+      "epoch": 0.8474402730375427,
+      "grad_norm": 6.800687313079834,
+      "learning_rate": 0.0007175199089874858,
+      "loss": 5.4928,
+      "step": 2483
+    },
+    {
+      "epoch": 0.8477815699658703,
+      "grad_norm": 6.051274299621582,
+      "learning_rate": 0.0007174061433447099,
+      "loss": 6.3143,
+      "step": 2484
+    },
+    {
+      "epoch": 0.8481228668941979,
+      "grad_norm": 3.3994340896606445,
+      "learning_rate": 0.000717292377701934,
+      "loss": 6.5889,
+      "step": 2485
+    },
+    {
+      "epoch": 0.8484641638225257,
+      "grad_norm": 3.665647029876709,
+      "learning_rate": 0.0007171786120591581,
+      "loss": 5.8222,
+      "step": 2486
+    },
+    {
+      "epoch": 0.8488054607508533,
+      "grad_norm": 3.3410794734954834,
+      "learning_rate": 0.0007170648464163823,
+      "loss": 6.8781,
+      "step": 2487
+    },
+    {
+      "epoch": 0.8491467576791809,
+      "grad_norm": 3.42598819732666,
+      "learning_rate": 0.0007169510807736064,
+      "loss": 6.8923,
+      "step": 2488
+    },
+    {
+      "epoch": 0.8494880546075085,
+      "grad_norm": 7.492861270904541,
+      "learning_rate": 0.0007168373151308305,
+      "loss": 7.0957,
+      "step": 2489
+    },
+    {
+      "epoch": 0.8498293515358362,
+      "grad_norm": 3.137038469314575,
+      "learning_rate": 0.0007167235494880546,
+      "loss": 6.6471,
+      "step": 2490
+    },
+    {
+      "epoch": 0.8501706484641638,
+      "grad_norm": 3.3882358074188232,
+      "learning_rate": 0.0007166097838452787,
+      "loss": 6.782,
+      "step": 2491
+    },
+    {
+      "epoch": 0.8505119453924914,
+      "grad_norm": 3.2121896743774414,
+      "learning_rate": 0.0007164960182025028,
+      "loss": 7.047,
+      "step": 2492
+    },
+    {
+      "epoch": 0.8508532423208192,
+      "grad_norm": 3.2568767070770264,
+      "learning_rate": 0.000716382252559727,
+      "loss": 6.6304,
+      "step": 2493
+    },
+    {
+      "epoch": 0.8511945392491468,
+      "grad_norm": 3.2692127227783203,
+      "learning_rate": 0.0007162684869169511,
+      "loss": 6.4022,
+      "step": 2494
+    },
+    {
+      "epoch": 0.8515358361774744,
+      "grad_norm": 3.5089681148529053,
+      "learning_rate": 0.0007161547212741752,
+      "loss": 6.5384,
+      "step": 2495
+    },
+    {
+      "epoch": 0.851877133105802,
+      "grad_norm": 3.226477861404419,
+      "learning_rate": 0.0007160409556313994,
+      "loss": 6.6573,
+      "step": 2496
+    },
+    {
+      "epoch": 0.8522184300341297,
+      "grad_norm": 3.2870302200317383,
+      "learning_rate": 0.0007159271899886235,
+      "loss": 6.9091,
+      "step": 2497
+    },
+    {
+      "epoch": 0.8525597269624573,
+      "grad_norm": 3.368769884109497,
+      "learning_rate": 0.0007158134243458477,
+      "loss": 6.54,
+      "step": 2498
+    },
+    {
+      "epoch": 0.852901023890785,
+      "grad_norm": 7.26986837387085,
+      "learning_rate": 0.0007156996587030717,
+      "loss": 5.2813,
+      "step": 2499
+    },
+    {
+      "epoch": 0.8532423208191127,
+      "grad_norm": 3.4235429763793945,
+      "learning_rate": 0.0007155858930602958,
+      "loss": 6.5904,
+      "step": 2500
+    },
+    {
+      "epoch": 0.8535836177474403,
+      "grad_norm": 3.4683103561401367,
+      "learning_rate": 0.0007154721274175199,
+      "loss": 6.8818,
+      "step": 2501
+    },
+    {
+      "epoch": 0.8539249146757679,
+      "grad_norm": 3.259256362915039,
+      "learning_rate": 0.000715358361774744,
+      "loss": 6.5643,
+      "step": 2502
+    },
+    {
+      "epoch": 0.8542662116040955,
+      "grad_norm": 3.19038724899292,
+      "learning_rate": 0.0007152445961319681,
+      "loss": 6.4574,
+      "step": 2503
+    },
+    {
+      "epoch": 0.8546075085324232,
+      "grad_norm": 4.036156177520752,
+      "learning_rate": 0.0007151308304891923,
+      "loss": 6.0228,
+      "step": 2504
+    },
+    {
+      "epoch": 0.8549488054607508,
+      "grad_norm": 3.6171653270721436,
+      "learning_rate": 0.0007150170648464164,
+      "loss": 6.4731,
+      "step": 2505
+    },
+    {
+      "epoch": 0.8552901023890785,
+      "grad_norm": 3.450333595275879,
+      "learning_rate": 0.0007149032992036405,
+      "loss": 6.3563,
+      "step": 2506
+    },
+    {
+      "epoch": 0.8556313993174062,
+      "grad_norm": 3.4450273513793945,
+      "learning_rate": 0.0007147895335608646,
+      "loss": 6.9957,
+      "step": 2507
+    },
+    {
+      "epoch": 0.8559726962457338,
+      "grad_norm": 3.228724241256714,
+      "learning_rate": 0.0007146757679180887,
+      "loss": 6.4982,
+      "step": 2508
+    },
+    {
+      "epoch": 0.8563139931740614,
+      "grad_norm": 3.283799409866333,
+      "learning_rate": 0.0007145620022753128,
+      "loss": 6.301,
+      "step": 2509
+    },
+    {
+      "epoch": 0.856655290102389,
+      "grad_norm": 3.258451223373413,
+      "learning_rate": 0.000714448236632537,
+      "loss": 6.2914,
+      "step": 2510
+    },
+    {
+      "epoch": 0.8569965870307167,
+      "grad_norm": 5.3775153160095215,
+      "learning_rate": 0.0007143344709897611,
+      "loss": 6.2267,
+      "step": 2511
+    },
+    {
+      "epoch": 0.8573378839590444,
+      "grad_norm": 3.5916664600372314,
+      "learning_rate": 0.0007142207053469852,
+      "loss": 6.7227,
+      "step": 2512
+    },
+    {
+      "epoch": 0.857679180887372,
+      "grad_norm": 5.10945987701416,
+      "learning_rate": 0.0007141069397042094,
+      "loss": 6.2029,
+      "step": 2513
+    },
+    {
+      "epoch": 0.8580204778156997,
+      "grad_norm": 3.6190226078033447,
+      "learning_rate": 0.0007139931740614335,
+      "loss": 6.0994,
+      "step": 2514
+    },
+    {
+      "epoch": 0.8583617747440273,
+      "grad_norm": 3.3382725715637207,
+      "learning_rate": 0.0007138794084186576,
+      "loss": 6.891,
+      "step": 2515
+    },
+    {
+      "epoch": 0.8587030716723549,
+      "grad_norm": 3.399120569229126,
+      "learning_rate": 0.0007137656427758818,
+      "loss": 6.8522,
+      "step": 2516
+    },
+    {
+      "epoch": 0.8590443686006826,
+      "grad_norm": 3.6390433311462402,
+      "learning_rate": 0.0007136518771331059,
+      "loss": 6.1696,
+      "step": 2517
+    },
+    {
+      "epoch": 0.8593856655290102,
+      "grad_norm": 9.467489242553711,
+      "learning_rate": 0.00071353811149033,
+      "loss": 6.8595,
+      "step": 2518
+    },
+    {
+      "epoch": 0.8597269624573379,
+      "grad_norm": 3.805907726287842,
+      "learning_rate": 0.000713424345847554,
+      "loss": 6.7477,
+      "step": 2519
+    },
+    {
+      "epoch": 0.8600682593856656,
+      "grad_norm": 3.396374464035034,
+      "learning_rate": 0.0007133105802047781,
+      "loss": 6.3055,
+      "step": 2520
+    },
+    {
+      "epoch": 0.8604095563139932,
+      "grad_norm": 3.43021821975708,
+      "learning_rate": 0.0007131968145620023,
+      "loss": 6.5407,
+      "step": 2521
+    },
+    {
+      "epoch": 0.8607508532423208,
+      "grad_norm": 6.201378345489502,
+      "learning_rate": 0.0007130830489192264,
+      "loss": 5.8928,
+      "step": 2522
+    },
+    {
+      "epoch": 0.8610921501706484,
+      "grad_norm": 3.4402365684509277,
+      "learning_rate": 0.0007129692832764505,
+      "loss": 6.5171,
+      "step": 2523
+    },
+    {
+      "epoch": 0.8614334470989761,
+      "grad_norm": 3.8701541423797607,
+      "learning_rate": 0.0007128555176336746,
+      "loss": 6.4958,
+      "step": 2524
+    },
+    {
+      "epoch": 0.8617747440273038,
+      "grad_norm": 3.2781641483306885,
+      "learning_rate": 0.0007127417519908987,
+      "loss": 6.8188,
+      "step": 2525
+    },
+    {
+      "epoch": 0.8621160409556314,
+      "grad_norm": 3.67041277885437,
+      "learning_rate": 0.0007126279863481228,
+      "loss": 5.8762,
+      "step": 2526
+    },
+    {
+      "epoch": 0.8624573378839591,
+      "grad_norm": 3.6243369579315186,
+      "learning_rate": 0.000712514220705347,
+      "loss": 6.3403,
+      "step": 2527
+    },
+    {
+      "epoch": 0.8627986348122867,
+      "grad_norm": 3.2596120834350586,
+      "learning_rate": 0.0007124004550625711,
+      "loss": 6.812,
+      "step": 2528
+    },
+    {
+      "epoch": 0.8631399317406143,
+      "grad_norm": 4.878437519073486,
+      "learning_rate": 0.0007122866894197952,
+      "loss": 6.3827,
+      "step": 2529
+    },
+    {
+      "epoch": 0.863481228668942,
+      "grad_norm": 2.523153305053711,
+      "learning_rate": 0.0007121729237770194,
+      "loss": 3.6518,
+      "step": 2530
+    },
+    {
+      "epoch": 0.8638225255972696,
+      "grad_norm": 3.9644103050231934,
+      "learning_rate": 0.0007120591581342435,
+      "loss": 5.9803,
+      "step": 2531
+    },
+    {
+      "epoch": 0.8641638225255973,
+      "grad_norm": 3.464620590209961,
+      "learning_rate": 0.0007119453924914676,
+      "loss": 6.6308,
+      "step": 2532
+    },
+    {
+      "epoch": 0.8645051194539249,
+      "grad_norm": 3.587636709213257,
+      "learning_rate": 0.0007118316268486918,
+      "loss": 7.2469,
+      "step": 2533
+    },
+    {
+      "epoch": 0.8648464163822526,
+      "grad_norm": 3.188854217529297,
+      "learning_rate": 0.0007117178612059159,
+      "loss": 6.3745,
+      "step": 2534
+    },
+    {
+      "epoch": 0.8651877133105802,
+      "grad_norm": 3.3936569690704346,
+      "learning_rate": 0.00071160409556314,
+      "loss": 7.0373,
+      "step": 2535
+    },
+    {
+      "epoch": 0.8655290102389078,
+      "grad_norm": 3.3409032821655273,
+      "learning_rate": 0.0007114903299203641,
+      "loss": 6.6337,
+      "step": 2536
+    },
+    {
+      "epoch": 0.8658703071672355,
+      "grad_norm": 5.52333402633667,
+      "learning_rate": 0.0007113765642775882,
+      "loss": 4.6133,
+      "step": 2537
+    },
+    {
+      "epoch": 0.8662116040955632,
+      "grad_norm": 3.3822033405303955,
+      "learning_rate": 0.0007112627986348122,
+      "loss": 6.1598,
+      "step": 2538
+    },
+    {
+      "epoch": 0.8665529010238908,
+      "grad_norm": 3.8111062049865723,
+      "learning_rate": 0.0007111490329920364,
+      "loss": 6.0937,
+      "step": 2539
+    },
+    {
+      "epoch": 0.8668941979522184,
+      "grad_norm": 4.7204909324646,
+      "learning_rate": 0.0007110352673492605,
+      "loss": 5.8796,
+      "step": 2540
+    },
+    {
+      "epoch": 0.8672354948805461,
+      "grad_norm": 3.3463664054870605,
+      "learning_rate": 0.0007109215017064846,
+      "loss": 6.7194,
+      "step": 2541
+    },
+    {
+      "epoch": 0.8675767918088737,
+      "grad_norm": 3.3839924335479736,
+      "learning_rate": 0.0007108077360637087,
+      "loss": 6.8464,
+      "step": 2542
+    },
+    {
+      "epoch": 0.8679180887372013,
+      "grad_norm": 4.8690571784973145,
+      "learning_rate": 0.0007106939704209328,
+      "loss": 5.8016,
+      "step": 2543
+    },
+    {
+      "epoch": 0.868259385665529,
+      "grad_norm": 3.3318896293640137,
+      "learning_rate": 0.000710580204778157,
+      "loss": 6.4526,
+      "step": 2544
+    },
+    {
+      "epoch": 0.8686006825938567,
+      "grad_norm": 3.221559762954712,
+      "learning_rate": 0.0007104664391353811,
+      "loss": 6.6463,
+      "step": 2545
+    },
+    {
+      "epoch": 0.8689419795221843,
+      "grad_norm": 3.1767921447753906,
+      "learning_rate": 0.0007103526734926053,
+      "loss": 6.4654,
+      "step": 2546
+    },
+    {
+      "epoch": 0.869283276450512,
+      "grad_norm": 3.4493680000305176,
+      "learning_rate": 0.0007102389078498294,
+      "loss": 5.9502,
+      "step": 2547
+    },
+    {
+      "epoch": 0.8696245733788396,
+      "grad_norm": 3.9862265586853027,
+      "learning_rate": 0.0007101251422070535,
+      "loss": 6.3229,
+      "step": 2548
+    },
+    {
+      "epoch": 0.8699658703071672,
+      "grad_norm": 3.4914379119873047,
+      "learning_rate": 0.0007100113765642776,
+      "loss": 6.9688,
+      "step": 2549
+    },
+    {
+      "epoch": 0.8703071672354948,
+      "grad_norm": 3.332296371459961,
+      "learning_rate": 0.0007098976109215018,
+      "loss": 6.8452,
+      "step": 2550
+    },
+    {
+      "epoch": 0.8706484641638226,
+      "grad_norm": 4.199538230895996,
+      "learning_rate": 0.0007097838452787259,
+      "loss": 6.881,
+      "step": 2551
+    },
+    {
+      "epoch": 0.8709897610921502,
+      "grad_norm": 3.5314183235168457,
+      "learning_rate": 0.00070967007963595,
+      "loss": 5.8921,
+      "step": 2552
+    },
+    {
+      "epoch": 0.8713310580204778,
+      "grad_norm": 3.459620237350464,
+      "learning_rate": 0.0007095563139931741,
+      "loss": 6.258,
+      "step": 2553
+    },
+    {
+      "epoch": 0.8716723549488055,
+      "grad_norm": 3.3095757961273193,
+      "learning_rate": 0.0007094425483503982,
+      "loss": 6.7562,
+      "step": 2554
+    },
+    {
+      "epoch": 0.8720136518771331,
+      "grad_norm": 3.2606141567230225,
+      "learning_rate": 0.0007093287827076223,
+      "loss": 6.7756,
+      "step": 2555
+    },
+    {
+      "epoch": 0.8723549488054607,
+      "grad_norm": 3.6110377311706543,
+      "learning_rate": 0.0007092150170648465,
+      "loss": 6.3181,
+      "step": 2556
+    },
+    {
+      "epoch": 0.8726962457337883,
+      "grad_norm": 3.273369550704956,
+      "learning_rate": 0.0007091012514220706,
+      "loss": 6.9579,
+      "step": 2557
+    },
+    {
+      "epoch": 0.8730375426621161,
+      "grad_norm": 5.260756492614746,
+      "learning_rate": 0.0007089874857792946,
+      "loss": 5.2098,
+      "step": 2558
+    },
+    {
+      "epoch": 0.8733788395904437,
+      "grad_norm": 3.3652751445770264,
+      "learning_rate": 0.0007088737201365187,
+      "loss": 6.8058,
+      "step": 2559
+    },
+    {
+      "epoch": 0.8737201365187713,
+      "grad_norm": 3.266148805618286,
+      "learning_rate": 0.0007087599544937428,
+      "loss": 7.0574,
+      "step": 2560
+    },
+    {
+      "epoch": 0.874061433447099,
+      "grad_norm": 3.58514666557312,
+      "learning_rate": 0.000708646188850967,
+      "loss": 6.6256,
+      "step": 2561
+    },
+    {
+      "epoch": 0.8744027303754266,
+      "grad_norm": 3.195068597793579,
+      "learning_rate": 0.0007085324232081911,
+      "loss": 6.6367,
+      "step": 2562
+    },
+    {
+      "epoch": 0.8747440273037542,
+      "grad_norm": 4.282649040222168,
+      "learning_rate": 0.0007084186575654153,
+      "loss": 4.9499,
+      "step": 2563
+    },
+    {
+      "epoch": 0.875085324232082,
+      "grad_norm": 3.2996411323547363,
+      "learning_rate": 0.0007083048919226394,
+      "loss": 6.4127,
+      "step": 2564
+    },
+    {
+      "epoch": 0.8754266211604096,
+      "grad_norm": 3.3839046955108643,
+      "learning_rate": 0.0007081911262798635,
+      "loss": 6.6752,
+      "step": 2565
+    },
+    {
+      "epoch": 0.8757679180887372,
+      "grad_norm": 3.211970806121826,
+      "learning_rate": 0.0007080773606370876,
+      "loss": 6.1393,
+      "step": 2566
+    },
+    {
+      "epoch": 0.8761092150170648,
+      "grad_norm": 3.1798558235168457,
+      "learning_rate": 0.0007079635949943118,
+      "loss": 6.6951,
+      "step": 2567
+    },
+    {
+      "epoch": 0.8764505119453925,
+      "grad_norm": 3.2988522052764893,
+      "learning_rate": 0.0007078498293515359,
+      "loss": 6.2898,
+      "step": 2568
+    },
+    {
+      "epoch": 0.8767918088737201,
+      "grad_norm": 3.8183929920196533,
+      "learning_rate": 0.00070773606370876,
+      "loss": 6.2697,
+      "step": 2569
+    },
+    {
+      "epoch": 0.8771331058020477,
+      "grad_norm": 3.168560266494751,
+      "learning_rate": 0.0007076222980659841,
+      "loss": 6.2589,
+      "step": 2570
+    },
+    {
+      "epoch": 0.8774744027303755,
+      "grad_norm": 3.605886459350586,
+      "learning_rate": 0.0007075085324232082,
+      "loss": 5.4431,
+      "step": 2571
+    },
+    {
+      "epoch": 0.8778156996587031,
+      "grad_norm": 3.5225412845611572,
+      "learning_rate": 0.0007073947667804323,
+      "loss": 6.9552,
+      "step": 2572
+    },
+    {
+      "epoch": 0.8781569965870307,
+      "grad_norm": 3.3253746032714844,
+      "learning_rate": 0.0007072810011376565,
+      "loss": 6.5634,
+      "step": 2573
+    },
+    {
+      "epoch": 0.8784982935153584,
+      "grad_norm": 3.3605477809906006,
+      "learning_rate": 0.0007071672354948806,
+      "loss": 6.4666,
+      "step": 2574
+    },
+    {
+      "epoch": 0.878839590443686,
+      "grad_norm": 3.1568098068237305,
+      "learning_rate": 0.0007070534698521047,
+      "loss": 6.6565,
+      "step": 2575
+    },
+    {
+      "epoch": 0.8791808873720136,
+      "grad_norm": 3.7349367141723633,
+      "learning_rate": 0.0007069397042093288,
+      "loss": 6.1776,
+      "step": 2576
+    },
+    {
+      "epoch": 0.8795221843003413,
+      "grad_norm": 3.3942160606384277,
+      "learning_rate": 0.0007068259385665528,
+      "loss": 6.32,
+      "step": 2577
+    },
+    {
+      "epoch": 0.879863481228669,
+      "grad_norm": 3.2961456775665283,
+      "learning_rate": 0.0007067121729237769,
+      "loss": 6.7801,
+      "step": 2578
+    },
+    {
+      "epoch": 0.8802047781569966,
+      "grad_norm": 3.3948254585266113,
+      "learning_rate": 0.0007065984072810011,
+      "loss": 6.0769,
+      "step": 2579
+    },
+    {
+      "epoch": 0.8805460750853242,
+      "grad_norm": 3.3471062183380127,
+      "learning_rate": 0.0007064846416382253,
+      "loss": 5.8324,
+      "step": 2580
+    },
+    {
+      "epoch": 0.8808873720136519,
+      "grad_norm": 3.5792579650878906,
+      "learning_rate": 0.0007063708759954494,
+      "loss": 6.313,
+      "step": 2581
+    },
+    {
+      "epoch": 0.8812286689419795,
+      "grad_norm": 3.430663824081421,
+      "learning_rate": 0.0007062571103526735,
+      "loss": 6.2877,
+      "step": 2582
+    },
+    {
+      "epoch": 0.8815699658703071,
+      "grad_norm": 5.736737251281738,
+      "learning_rate": 0.0007061433447098976,
+      "loss": 6.0753,
+      "step": 2583
+    },
+    {
+      "epoch": 0.8819112627986349,
+      "grad_norm": 3.3799870014190674,
+      "learning_rate": 0.0007060295790671218,
+      "loss": 6.7596,
+      "step": 2584
+    },
+    {
+      "epoch": 0.8822525597269625,
+      "grad_norm": 6.616389751434326,
+      "learning_rate": 0.0007059158134243459,
+      "loss": 5.9919,
+      "step": 2585
+    },
+    {
+      "epoch": 0.8825938566552901,
+      "grad_norm": 3.429597854614258,
+      "learning_rate": 0.00070580204778157,
+      "loss": 6.3736,
+      "step": 2586
+    },
+    {
+      "epoch": 0.8829351535836177,
+      "grad_norm": 3.608384847640991,
+      "learning_rate": 0.0007056882821387941,
+      "loss": 6.3896,
+      "step": 2587
+    },
+    {
+      "epoch": 0.8832764505119454,
+      "grad_norm": 3.358494281768799,
+      "learning_rate": 0.0007055745164960182,
+      "loss": 7.1418,
+      "step": 2588
+    },
+    {
+      "epoch": 0.883617747440273,
+      "grad_norm": 3.164523124694824,
+      "learning_rate": 0.0007054607508532423,
+      "loss": 6.8558,
+      "step": 2589
+    },
+    {
+      "epoch": 0.8839590443686007,
+      "grad_norm": 3.5418014526367188,
+      "learning_rate": 0.0007053469852104665,
+      "loss": 6.2485,
+      "step": 2590
+    },
+    {
+      "epoch": 0.8843003412969284,
+      "grad_norm": 3.181857109069824,
+      "learning_rate": 0.0007052332195676906,
+      "loss": 6.9945,
+      "step": 2591
+    },
+    {
+      "epoch": 0.884641638225256,
+      "grad_norm": 3.445446014404297,
+      "learning_rate": 0.0007051194539249147,
+      "loss": 6.6083,
+      "step": 2592
+    },
+    {
+      "epoch": 0.8849829351535836,
+      "grad_norm": 3.159823417663574,
+      "learning_rate": 0.0007050056882821388,
+      "loss": 6.6446,
+      "step": 2593
+    },
+    {
+      "epoch": 0.8853242320819112,
+      "grad_norm": 5.284470081329346,
+      "learning_rate": 0.0007048919226393629,
+      "loss": 5.6566,
+      "step": 2594
+    },
+    {
+      "epoch": 0.8856655290102389,
+      "grad_norm": 5.081126689910889,
+      "learning_rate": 0.000704778156996587,
+      "loss": 5.952,
+      "step": 2595
+    },
+    {
+      "epoch": 0.8860068259385665,
+      "grad_norm": 3.5319552421569824,
+      "learning_rate": 0.0007046643913538113,
+      "loss": 6.2608,
+      "step": 2596
+    },
+    {
+      "epoch": 0.8863481228668942,
+      "grad_norm": 3.5452115535736084,
+      "learning_rate": 0.0007045506257110353,
+      "loss": 7.0805,
+      "step": 2597
+    },
+    {
+      "epoch": 0.8866894197952219,
+      "grad_norm": 3.5976240634918213,
+      "learning_rate": 0.0007044368600682594,
+      "loss": 6.3286,
+      "step": 2598
+    },
+    {
+      "epoch": 0.8870307167235495,
+      "grad_norm": 3.242631435394287,
+      "learning_rate": 0.0007043230944254835,
+      "loss": 6.8221,
+      "step": 2599
+    },
+    {
+      "epoch": 0.8873720136518771,
+      "grad_norm": 3.2306206226348877,
+      "learning_rate": 0.0007042093287827076,
+      "loss": 6.2416,
+      "step": 2600
+    },
+    {
+      "epoch": 0.8877133105802048,
+      "grad_norm": 3.222269296646118,
+      "learning_rate": 0.0007040955631399318,
+      "loss": 6.9457,
+      "step": 2601
+    },
+    {
+      "epoch": 0.8880546075085324,
+      "grad_norm": 3.364189863204956,
+      "learning_rate": 0.0007039817974971559,
+      "loss": 6.3123,
+      "step": 2602
+    },
+    {
+      "epoch": 0.8883959044368601,
+      "grad_norm": 7.864152908325195,
+      "learning_rate": 0.00070386803185438,
+      "loss": 5.3723,
+      "step": 2603
+    },
+    {
+      "epoch": 0.8887372013651877,
+      "grad_norm": 3.3091676235198975,
+      "learning_rate": 0.0007037542662116041,
+      "loss": 6.5701,
+      "step": 2604
+    },
+    {
+      "epoch": 0.8890784982935154,
+      "grad_norm": 3.4476070404052734,
+      "learning_rate": 0.0007036405005688282,
+      "loss": 6.4836,
+      "step": 2605
+    },
+    {
+      "epoch": 0.889419795221843,
+      "grad_norm": 3.41062593460083,
+      "learning_rate": 0.0007035267349260523,
+      "loss": 6.7782,
+      "step": 2606
+    },
+    {
+      "epoch": 0.8897610921501706,
+      "grad_norm": 3.3153281211853027,
+      "learning_rate": 0.0007034129692832765,
+      "loss": 6.5789,
+      "step": 2607
+    },
+    {
+      "epoch": 0.8901023890784983,
+      "grad_norm": 3.2125444412231445,
+      "learning_rate": 0.0007032992036405006,
+      "loss": 6.57,
+      "step": 2608
+    },
+    {
+      "epoch": 0.8904436860068259,
+      "grad_norm": 3.210056781768799,
+      "learning_rate": 0.0007031854379977247,
+      "loss": 6.4377,
+      "step": 2609
+    },
+    {
+      "epoch": 0.8907849829351536,
+      "grad_norm": 3.1639184951782227,
+      "learning_rate": 0.0007030716723549488,
+      "loss": 6.4598,
+      "step": 2610
+    },
+    {
+      "epoch": 0.8911262798634813,
+      "grad_norm": 3.7884199619293213,
+      "learning_rate": 0.0007029579067121729,
+      "loss": 6.4981,
+      "step": 2611
+    },
+    {
+      "epoch": 0.8914675767918089,
+      "grad_norm": 3.2961177825927734,
+      "learning_rate": 0.000702844141069397,
+      "loss": 6.8189,
+      "step": 2612
+    },
+    {
+      "epoch": 0.8918088737201365,
+      "grad_norm": 3.3489444255828857,
+      "learning_rate": 0.0007027303754266213,
+      "loss": 6.3774,
+      "step": 2613
+    },
+    {
+      "epoch": 0.8921501706484641,
+      "grad_norm": 3.563915967941284,
+      "learning_rate": 0.0007026166097838454,
+      "loss": 6.4961,
+      "step": 2614
+    },
+    {
+      "epoch": 0.8924914675767918,
+      "grad_norm": 3.3189163208007812,
+      "learning_rate": 0.0007025028441410695,
+      "loss": 6.6371,
+      "step": 2615
+    },
+    {
+      "epoch": 0.8928327645051195,
+      "grad_norm": 3.447849750518799,
+      "learning_rate": 0.0007023890784982935,
+      "loss": 5.8723,
+      "step": 2616
+    },
+    {
+      "epoch": 0.8931740614334471,
+      "grad_norm": 3.2559688091278076,
+      "learning_rate": 0.0007022753128555176,
+      "loss": 6.4424,
+      "step": 2617
+    },
+    {
+      "epoch": 0.8935153583617748,
+      "grad_norm": 4.270260334014893,
+      "learning_rate": 0.0007021615472127417,
+      "loss": 5.9736,
+      "step": 2618
+    },
+    {
+      "epoch": 0.8938566552901024,
+      "grad_norm": 3.3759734630584717,
+      "learning_rate": 0.0007020477815699659,
+      "loss": 6.6424,
+      "step": 2619
+    },
+    {
+      "epoch": 0.89419795221843,
+      "grad_norm": 3.839401960372925,
+      "learning_rate": 0.00070193401592719,
+      "loss": 5.8343,
+      "step": 2620
+    },
+    {
+      "epoch": 0.8945392491467576,
+      "grad_norm": 3.4402377605438232,
+      "learning_rate": 0.0007018202502844141,
+      "loss": 5.9067,
+      "step": 2621
+    },
+    {
+      "epoch": 0.8948805460750853,
+      "grad_norm": 3.8046011924743652,
+      "learning_rate": 0.0007017064846416382,
+      "loss": 6.205,
+      "step": 2622
+    },
+    {
+      "epoch": 0.895221843003413,
+      "grad_norm": 3.5362050533294678,
+      "learning_rate": 0.0007015927189988623,
+      "loss": 6.9003,
+      "step": 2623
+    },
+    {
+      "epoch": 0.8955631399317406,
+      "grad_norm": 3.417820930480957,
+      "learning_rate": 0.0007014789533560865,
+      "loss": 6.9618,
+      "step": 2624
+    },
+    {
+      "epoch": 0.8959044368600683,
+      "grad_norm": 3.3132388591766357,
+      "learning_rate": 0.0007013651877133106,
+      "loss": 6.8951,
+      "step": 2625
+    },
+    {
+      "epoch": 0.8962457337883959,
+      "grad_norm": 3.25205659866333,
+      "learning_rate": 0.0007012514220705347,
+      "loss": 6.361,
+      "step": 2626
+    },
+    {
+      "epoch": 0.8965870307167235,
+      "grad_norm": 3.5494885444641113,
+      "learning_rate": 0.0007011376564277588,
+      "loss": 6.2906,
+      "step": 2627
+    },
+    {
+      "epoch": 0.8969283276450511,
+      "grad_norm": 3.837991714477539,
+      "learning_rate": 0.0007010238907849829,
+      "loss": 5.26,
+      "step": 2628
+    },
+    {
+      "epoch": 0.8972696245733789,
+      "grad_norm": 3.9714996814727783,
+      "learning_rate": 0.000700910125142207,
+      "loss": 6.2808,
+      "step": 2629
+    },
+    {
+      "epoch": 0.8976109215017065,
+      "grad_norm": 3.556457042694092,
+      "learning_rate": 0.0007007963594994313,
+      "loss": 6.3822,
+      "step": 2630
+    },
+    {
+      "epoch": 0.8979522184300341,
+      "grad_norm": 3.464637517929077,
+      "learning_rate": 0.0007006825938566554,
+      "loss": 6.9314,
+      "step": 2631
+    },
+    {
+      "epoch": 0.8982935153583618,
+      "grad_norm": 6.831556797027588,
+      "learning_rate": 0.0007005688282138795,
+      "loss": 6.4999,
+      "step": 2632
+    },
+    {
+      "epoch": 0.8986348122866894,
+      "grad_norm": 3.62048602104187,
+      "learning_rate": 0.0007004550625711036,
+      "loss": 6.0907,
+      "step": 2633
+    },
+    {
+      "epoch": 0.898976109215017,
+      "grad_norm": 3.393540620803833,
+      "learning_rate": 0.0007003412969283277,
+      "loss": 6.192,
+      "step": 2634
+    },
+    {
+      "epoch": 0.8993174061433447,
+      "grad_norm": 3.4911301136016846,
+      "learning_rate": 0.0007002275312855518,
+      "loss": 6.8963,
+      "step": 2635
+    },
+    {
+      "epoch": 0.8996587030716724,
+      "grad_norm": 3.324739694595337,
+      "learning_rate": 0.0007001137656427759,
+      "loss": 6.5604,
+      "step": 2636
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.236996650695801,
+      "learning_rate": 0.0007,
+      "loss": 6.8306,
+      "step": 2637
+    },
+    {
+      "epoch": 0.9003412969283277,
+      "grad_norm": 6.527902603149414,
+      "learning_rate": 0.0006998862343572241,
+      "loss": 4.9167,
+      "step": 2638
+    },
+    {
+      "epoch": 0.9006825938566553,
+      "grad_norm": 3.5526578426361084,
+      "learning_rate": 0.0006997724687144482,
+      "loss": 6.7842,
+      "step": 2639
+    },
+    {
+      "epoch": 0.9010238907849829,
+      "grad_norm": 3.3249645233154297,
+      "learning_rate": 0.0006996587030716723,
+      "loss": 6.6712,
+      "step": 2640
+    },
+    {
+      "epoch": 0.9013651877133105,
+      "grad_norm": 3.207120656967163,
+      "learning_rate": 0.0006995449374288964,
+      "loss": 6.8788,
+      "step": 2641
+    },
+    {
+      "epoch": 0.9017064846416383,
+      "grad_norm": 3.128549098968506,
+      "learning_rate": 0.0006994311717861206,
+      "loss": 6.5011,
+      "step": 2642
+    },
+    {
+      "epoch": 0.9020477815699659,
+      "grad_norm": 3.4323368072509766,
+      "learning_rate": 0.0006993174061433447,
+      "loss": 6.3345,
+      "step": 2643
+    },
+    {
+      "epoch": 0.9023890784982935,
+      "grad_norm": 3.2746779918670654,
+      "learning_rate": 0.0006992036405005688,
+      "loss": 6.3755,
+      "step": 2644
+    },
+    {
+      "epoch": 0.9027303754266212,
+      "grad_norm": 8.446745872497559,
+      "learning_rate": 0.0006990898748577929,
+      "loss": 6.7535,
+      "step": 2645
+    },
+    {
+      "epoch": 0.9030716723549488,
+      "grad_norm": 3.5905816555023193,
+      "learning_rate": 0.000698976109215017,
+      "loss": 6.7017,
+      "step": 2646
+    },
+    {
+      "epoch": 0.9034129692832764,
+      "grad_norm": 3.572148084640503,
+      "learning_rate": 0.0006988623435722413,
+      "loss": 6.6064,
+      "step": 2647
+    },
+    {
+      "epoch": 0.903754266211604,
+      "grad_norm": 5.641802787780762,
+      "learning_rate": 0.0006987485779294654,
+      "loss": 5.9658,
+      "step": 2648
+    },
+    {
+      "epoch": 0.9040955631399318,
+      "grad_norm": 3.906459331512451,
+      "learning_rate": 0.0006986348122866895,
+      "loss": 6.1134,
+      "step": 2649
+    },
+    {
+      "epoch": 0.9044368600682594,
+      "grad_norm": 3.3307929039001465,
+      "learning_rate": 0.0006985210466439136,
+      "loss": 6.8376,
+      "step": 2650
+    },
+    {
+      "epoch": 0.904778156996587,
+      "grad_norm": 3.302304267883301,
+      "learning_rate": 0.0006984072810011377,
+      "loss": 6.762,
+      "step": 2651
+    },
+    {
+      "epoch": 0.9051194539249147,
+      "grad_norm": 3.249640941619873,
+      "learning_rate": 0.0006982935153583618,
+      "loss": 6.7481,
+      "step": 2652
+    },
+    {
+      "epoch": 0.9054607508532423,
+      "grad_norm": 5.366897106170654,
+      "learning_rate": 0.000698179749715586,
+      "loss": 6.1722,
+      "step": 2653
+    },
+    {
+      "epoch": 0.9058020477815699,
+      "grad_norm": 3.280503749847412,
+      "learning_rate": 0.0006980659840728101,
+      "loss": 6.5592,
+      "step": 2654
+    },
+    {
+      "epoch": 0.9061433447098977,
+      "grad_norm": 3.4887964725494385,
+      "learning_rate": 0.0006979522184300341,
+      "loss": 6.6677,
+      "step": 2655
+    },
+    {
+      "epoch": 0.9064846416382253,
+      "grad_norm": 3.333489418029785,
+      "learning_rate": 0.0006978384527872582,
+      "loss": 5.8702,
+      "step": 2656
+    },
+    {
+      "epoch": 0.9068259385665529,
+      "grad_norm": 3.1572794914245605,
+      "learning_rate": 0.0006977246871444823,
+      "loss": 6.6554,
+      "step": 2657
+    },
+    {
+      "epoch": 0.9071672354948805,
+      "grad_norm": 3.2989954948425293,
+      "learning_rate": 0.0006976109215017064,
+      "loss": 6.5196,
+      "step": 2658
+    },
+    {
+      "epoch": 0.9075085324232082,
+      "grad_norm": 5.108197212219238,
+      "learning_rate": 0.0006974971558589306,
+      "loss": 5.6353,
+      "step": 2659
+    },
+    {
+      "epoch": 0.9078498293515358,
+      "grad_norm": 3.3741981983184814,
+      "learning_rate": 0.0006973833902161547,
+      "loss": 6.3594,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9081911262798635,
+      "grad_norm": 3.180333137512207,
+      "learning_rate": 0.0006972696245733788,
+      "loss": 6.3406,
+      "step": 2661
+    },
+    {
+      "epoch": 0.9085324232081912,
+      "grad_norm": 3.803738832473755,
+      "learning_rate": 0.0006971558589306029,
+      "loss": 6.3439,
+      "step": 2662
+    },
+    {
+      "epoch": 0.9088737201365188,
+      "grad_norm": 4.112486362457275,
+      "learning_rate": 0.000697042093287827,
+      "loss": 5.6222,
+      "step": 2663
+    },
+    {
+      "epoch": 0.9092150170648464,
+      "grad_norm": 3.6349542140960693,
+      "learning_rate": 0.0006969283276450513,
+      "loss": 6.0037,
+      "step": 2664
+    },
+    {
+      "epoch": 0.909556313993174,
+      "grad_norm": 3.6072964668273926,
+      "learning_rate": 0.0006968145620022754,
+      "loss": 6.6496,
+      "step": 2665
+    },
+    {
+      "epoch": 0.9098976109215017,
+      "grad_norm": 3.841954469680786,
+      "learning_rate": 0.0006967007963594995,
+      "loss": 7.091,
+      "step": 2666
+    },
+    {
+      "epoch": 0.9102389078498293,
+      "grad_norm": 3.223397970199585,
+      "learning_rate": 0.0006965870307167236,
+      "loss": 6.9502,
+      "step": 2667
+    },
+    {
+      "epoch": 0.910580204778157,
+      "grad_norm": 3.3219428062438965,
+      "learning_rate": 0.0006964732650739477,
+      "loss": 6.7914,
+      "step": 2668
+    },
+    {
+      "epoch": 0.9109215017064847,
+      "grad_norm": 3.22790265083313,
+      "learning_rate": 0.0006963594994311718,
+      "loss": 6.8888,
+      "step": 2669
+    },
+    {
+      "epoch": 0.9112627986348123,
+      "grad_norm": 3.58799409866333,
+      "learning_rate": 0.000696245733788396,
+      "loss": 6.5231,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9116040955631399,
+      "grad_norm": 3.4284210205078125,
+      "learning_rate": 0.0006961319681456201,
+      "loss": 6.4324,
+      "step": 2671
+    },
+    {
+      "epoch": 0.9119453924914676,
+      "grad_norm": 3.6052560806274414,
+      "learning_rate": 0.0006960182025028442,
+      "loss": 6.6546,
+      "step": 2672
+    },
+    {
+      "epoch": 0.9122866894197952,
+      "grad_norm": 3.6261966228485107,
+      "learning_rate": 0.0006959044368600683,
+      "loss": 6.6569,
+      "step": 2673
+    },
+    {
+      "epoch": 0.9126279863481229,
+      "grad_norm": 3.32460880279541,
+      "learning_rate": 0.0006957906712172923,
+      "loss": 6.5463,
+      "step": 2674
+    },
+    {
+      "epoch": 0.9129692832764505,
+      "grad_norm": 3.767965078353882,
+      "learning_rate": 0.0006956769055745164,
+      "loss": 6.153,
+      "step": 2675
+    },
+    {
+      "epoch": 0.9133105802047782,
+      "grad_norm": 3.2888388633728027,
+      "learning_rate": 0.0006955631399317406,
+      "loss": 6.5947,
+      "step": 2676
+    },
+    {
+      "epoch": 0.9136518771331058,
+      "grad_norm": 3.235499620437622,
+      "learning_rate": 0.0006954493742889647,
+      "loss": 6.6224,
+      "step": 2677
+    },
+    {
+      "epoch": 0.9139931740614334,
+      "grad_norm": 3.1884522438049316,
+      "learning_rate": 0.0006953356086461888,
+      "loss": 6.8578,
+      "step": 2678
+    },
+    {
+      "epoch": 0.9143344709897611,
+      "grad_norm": 3.4016973972320557,
+      "learning_rate": 0.0006952218430034129,
+      "loss": 6.9575,
+      "step": 2679
+    },
+    {
+      "epoch": 0.9146757679180887,
+      "grad_norm": 3.2565393447875977,
+      "learning_rate": 0.000695108077360637,
+      "loss": 7.086,
+      "step": 2680
+    },
+    {
+      "epoch": 0.9150170648464164,
+      "grad_norm": 3.359285593032837,
+      "learning_rate": 0.0006949943117178612,
+      "loss": 6.695,
+      "step": 2681
+    },
+    {
+      "epoch": 0.9153583617747441,
+      "grad_norm": 3.5385429859161377,
+      "learning_rate": 0.0006948805460750854,
+      "loss": 6.2872,
+      "step": 2682
+    },
+    {
+      "epoch": 0.9156996587030717,
+      "grad_norm": 3.284391164779663,
+      "learning_rate": 0.0006947667804323095,
+      "loss": 6.5209,
+      "step": 2683
+    },
+    {
+      "epoch": 0.9160409556313993,
+      "grad_norm": 3.260636329650879,
+      "learning_rate": 0.0006946530147895336,
+      "loss": 6.4423,
+      "step": 2684
+    },
+    {
+      "epoch": 0.9163822525597269,
+      "grad_norm": 5.266411304473877,
+      "learning_rate": 0.0006945392491467577,
+      "loss": 5.6748,
+      "step": 2685
+    },
+    {
+      "epoch": 0.9167235494880546,
+      "grad_norm": 3.5976195335388184,
+      "learning_rate": 0.0006944254835039818,
+      "loss": 6.3951,
+      "step": 2686
+    },
+    {
+      "epoch": 0.9170648464163823,
+      "grad_norm": 3.503249406814575,
+      "learning_rate": 0.000694311717861206,
+      "loss": 6.8257,
+      "step": 2687
+    },
+    {
+      "epoch": 0.9174061433447099,
+      "grad_norm": 3.3149096965789795,
+      "learning_rate": 0.0006941979522184301,
+      "loss": 6.1159,
+      "step": 2688
+    },
+    {
+      "epoch": 0.9177474402730376,
+      "grad_norm": 3.2986981868743896,
+      "learning_rate": 0.0006940841865756542,
+      "loss": 6.1175,
+      "step": 2689
+    },
+    {
+      "epoch": 0.9180887372013652,
+      "grad_norm": 3.357754707336426,
+      "learning_rate": 0.0006939704209328783,
+      "loss": 6.8085,
+      "step": 2690
+    },
+    {
+      "epoch": 0.9184300341296928,
+      "grad_norm": 3.491800308227539,
+      "learning_rate": 0.0006938566552901024,
+      "loss": 6.4161,
+      "step": 2691
+    },
+    {
+      "epoch": 0.9187713310580204,
+      "grad_norm": 3.2512035369873047,
+      "learning_rate": 0.0006937428896473265,
+      "loss": 6.8517,
+      "step": 2692
+    },
+    {
+      "epoch": 0.9191126279863481,
+      "grad_norm": 3.1842398643493652,
+      "learning_rate": 0.0006936291240045507,
+      "loss": 7.1578,
+      "step": 2693
+    },
+    {
+      "epoch": 0.9194539249146758,
+      "grad_norm": 3.4947540760040283,
+      "learning_rate": 0.0006935153583617747,
+      "loss": 6.0914,
+      "step": 2694
+    },
+    {
+      "epoch": 0.9197952218430034,
+      "grad_norm": 8.686001777648926,
+      "learning_rate": 0.0006934015927189988,
+      "loss": 5.7208,
+      "step": 2695
+    },
+    {
+      "epoch": 0.9201365187713311,
+      "grad_norm": 3.50386643409729,
+      "learning_rate": 0.0006932878270762229,
+      "loss": 6.766,
+      "step": 2696
+    },
+    {
+      "epoch": 0.9204778156996587,
+      "grad_norm": 3.275815486907959,
+      "learning_rate": 0.000693174061433447,
+      "loss": 6.6614,
+      "step": 2697
+    },
+    {
+      "epoch": 0.9208191126279863,
+      "grad_norm": 3.2420060634613037,
+      "learning_rate": 0.0006930602957906712,
+      "loss": 6.4655,
+      "step": 2698
+    },
+    {
+      "epoch": 0.921160409556314,
+      "grad_norm": 3.386936902999878,
+      "learning_rate": 0.0006929465301478954,
+      "loss": 6.318,
+      "step": 2699
+    },
+    {
+      "epoch": 0.9215017064846417,
+      "grad_norm": 5.493778705596924,
+      "learning_rate": 0.0006928327645051195,
+      "loss": 5.9732,
+      "step": 2700
+    },
+    {
+      "epoch": 0.9218430034129693,
+      "grad_norm": 3.2191455364227295,
+      "learning_rate": 0.0006927189988623436,
+      "loss": 6.5306,
+      "step": 2701
+    },
+    {
+      "epoch": 0.922184300341297,
+      "grad_norm": 3.3493263721466064,
+      "learning_rate": 0.0006926052332195677,
+      "loss": 6.4363,
+      "step": 2702
+    },
+    {
+      "epoch": 0.9225255972696246,
+      "grad_norm": 3.3574280738830566,
+      "learning_rate": 0.0006924914675767918,
+      "loss": 6.9716,
+      "step": 2703
+    },
+    {
+      "epoch": 0.9228668941979522,
+      "grad_norm": 3.3387627601623535,
+      "learning_rate": 0.000692377701934016,
+      "loss": 6.9231,
+      "step": 2704
+    },
+    {
+      "epoch": 0.9232081911262798,
+      "grad_norm": 3.2208516597747803,
+      "learning_rate": 0.0006922639362912401,
+      "loss": 6.4118,
+      "step": 2705
+    },
+    {
+      "epoch": 0.9235494880546075,
+      "grad_norm": 3.2908411026000977,
+      "learning_rate": 0.0006921501706484642,
+      "loss": 6.4098,
+      "step": 2706
+    },
+    {
+      "epoch": 0.9238907849829352,
+      "grad_norm": 3.3244807720184326,
+      "learning_rate": 0.0006920364050056883,
+      "loss": 6.8572,
+      "step": 2707
+    },
+    {
+      "epoch": 0.9242320819112628,
+      "grad_norm": 3.2083566188812256,
+      "learning_rate": 0.0006919226393629124,
+      "loss": 6.8654,
+      "step": 2708
+    },
+    {
+      "epoch": 0.9245733788395905,
+      "grad_norm": 3.230299472808838,
+      "learning_rate": 0.0006918088737201365,
+      "loss": 6.6024,
+      "step": 2709
+    },
+    {
+      "epoch": 0.9249146757679181,
+      "grad_norm": 3.518018960952759,
+      "learning_rate": 0.0006916951080773607,
+      "loss": 5.8393,
+      "step": 2710
+    },
+    {
+      "epoch": 0.9252559726962457,
+      "grad_norm": 3.445234775543213,
+      "learning_rate": 0.0006915813424345848,
+      "loss": 6.0757,
+      "step": 2711
+    },
+    {
+      "epoch": 0.9255972696245733,
+      "grad_norm": 3.329686403274536,
+      "learning_rate": 0.000691467576791809,
+      "loss": 6.8527,
+      "step": 2712
+    },
+    {
+      "epoch": 0.9259385665529011,
+      "grad_norm": 3.224597215652466,
+      "learning_rate": 0.0006913538111490329,
+      "loss": 6.5644,
+      "step": 2713
+    },
+    {
+      "epoch": 0.9262798634812287,
+      "grad_norm": 3.4149997234344482,
+      "learning_rate": 0.000691240045506257,
+      "loss": 6.7847,
+      "step": 2714
+    },
+    {
+      "epoch": 0.9266211604095563,
+      "grad_norm": 3.3255600929260254,
+      "learning_rate": 0.0006911262798634812,
+      "loss": 6.5404,
+      "step": 2715
+    },
+    {
+      "epoch": 0.926962457337884,
+      "grad_norm": 5.029857635498047,
+      "learning_rate": 0.0006910125142207054,
+      "loss": 5.6074,
+      "step": 2716
+    },
+    {
+      "epoch": 0.9273037542662116,
+      "grad_norm": 4.0226149559021,
+      "learning_rate": 0.0006908987485779295,
+      "loss": 6.6629,
+      "step": 2717
+    },
+    {
+      "epoch": 0.9276450511945392,
+      "grad_norm": 3.5831539630889893,
+      "learning_rate": 0.0006907849829351536,
+      "loss": 5.8483,
+      "step": 2718
+    },
+    {
+      "epoch": 0.9279863481228668,
+      "grad_norm": 3.3607494831085205,
+      "learning_rate": 0.0006906712172923777,
+      "loss": 6.7091,
+      "step": 2719
+    },
+    {
+      "epoch": 0.9283276450511946,
+      "grad_norm": 5.548701763153076,
+      "learning_rate": 0.0006905574516496018,
+      "loss": 5.6986,
+      "step": 2720
+    },
+    {
+      "epoch": 0.9286689419795222,
+      "grad_norm": 3.317486047744751,
+      "learning_rate": 0.0006904436860068259,
+      "loss": 6.4468,
+      "step": 2721
+    },
+    {
+      "epoch": 0.9290102389078498,
+      "grad_norm": 3.3964345455169678,
+      "learning_rate": 0.0006903299203640501,
+      "loss": 6.7458,
+      "step": 2722
+    },
+    {
+      "epoch": 0.9293515358361775,
+      "grad_norm": 3.141096591949463,
+      "learning_rate": 0.0006902161547212742,
+      "loss": 6.8357,
+      "step": 2723
+    },
+    {
+      "epoch": 0.9296928327645051,
+      "grad_norm": 3.1272401809692383,
+      "learning_rate": 0.0006901023890784983,
+      "loss": 6.2205,
+      "step": 2724
+    },
+    {
+      "epoch": 0.9300341296928327,
+      "grad_norm": 3.82816743850708,
+      "learning_rate": 0.0006899886234357224,
+      "loss": 6.151,
+      "step": 2725
+    },
+    {
+      "epoch": 0.9303754266211605,
+      "grad_norm": 3.0288915634155273,
+      "learning_rate": 0.0006898748577929465,
+      "loss": 6.5263,
+      "step": 2726
+    },
+    {
+      "epoch": 0.9307167235494881,
+      "grad_norm": 3.344312906265259,
+      "learning_rate": 0.0006897610921501707,
+      "loss": 6.6451,
+      "step": 2727
+    },
+    {
+      "epoch": 0.9310580204778157,
+      "grad_norm": 3.386958122253418,
+      "learning_rate": 0.0006896473265073948,
+      "loss": 6.5277,
+      "step": 2728
+    },
+    {
+      "epoch": 0.9313993174061433,
+      "grad_norm": 3.474781036376953,
+      "learning_rate": 0.000689533560864619,
+      "loss": 6.0832,
+      "step": 2729
+    },
+    {
+      "epoch": 0.931740614334471,
+      "grad_norm": 3.3939759731292725,
+      "learning_rate": 0.000689419795221843,
+      "loss": 7.0273,
+      "step": 2730
+    },
+    {
+      "epoch": 0.9320819112627986,
+      "grad_norm": 3.276217460632324,
+      "learning_rate": 0.0006893060295790672,
+      "loss": 5.9832,
+      "step": 2731
+    },
+    {
+      "epoch": 0.9324232081911262,
+      "grad_norm": 3.2596402168273926,
+      "learning_rate": 0.0006891922639362913,
+      "loss": 6.8106,
+      "step": 2732
+    },
+    {
+      "epoch": 0.932764505119454,
+      "grad_norm": 3.262134313583374,
+      "learning_rate": 0.0006890784982935154,
+      "loss": 6.5364,
+      "step": 2733
+    },
+    {
+      "epoch": 0.9331058020477816,
+      "grad_norm": 4.899016380310059,
+      "learning_rate": 0.0006889647326507395,
+      "loss": 6.1792,
+      "step": 2734
+    },
+    {
+      "epoch": 0.9334470989761092,
+      "grad_norm": 4.251522064208984,
+      "learning_rate": 0.0006888509670079636,
+      "loss": 6.3958,
+      "step": 2735
+    },
+    {
+      "epoch": 0.9337883959044369,
+      "grad_norm": 4.013020992279053,
+      "learning_rate": 0.0006887372013651877,
+      "loss": 6.1985,
+      "step": 2736
+    },
+    {
+      "epoch": 0.9341296928327645,
+      "grad_norm": 3.3481905460357666,
+      "learning_rate": 0.0006886234357224118,
+      "loss": 7.0557,
+      "step": 2737
+    },
+    {
+      "epoch": 0.9344709897610921,
+      "grad_norm": 3.201247215270996,
+      "learning_rate": 0.0006885096700796359,
+      "loss": 6.2043,
+      "step": 2738
+    },
+    {
+      "epoch": 0.9348122866894198,
+      "grad_norm": 3.4401636123657227,
+      "learning_rate": 0.0006883959044368601,
+      "loss": 7.097,
+      "step": 2739
+    },
+    {
+      "epoch": 0.9351535836177475,
+      "grad_norm": 3.1851375102996826,
+      "learning_rate": 0.0006882821387940842,
+      "loss": 6.761,
+      "step": 2740
+    },
+    {
+      "epoch": 0.9354948805460751,
+      "grad_norm": 3.280695915222168,
+      "learning_rate": 0.0006881683731513083,
+      "loss": 7.2953,
+      "step": 2741
+    },
+    {
+      "epoch": 0.9358361774744027,
+      "grad_norm": 3.2717418670654297,
+      "learning_rate": 0.0006880546075085324,
+      "loss": 6.71,
+      "step": 2742
+    },
+    {
+      "epoch": 0.9361774744027304,
+      "grad_norm": 3.8960654735565186,
+      "learning_rate": 0.0006879408418657565,
+      "loss": 5.5993,
+      "step": 2743
+    },
+    {
+      "epoch": 0.936518771331058,
+      "grad_norm": 3.261302947998047,
+      "learning_rate": 0.0006878270762229807,
+      "loss": 6.2783,
+      "step": 2744
+    },
+    {
+      "epoch": 0.9368600682593856,
+      "grad_norm": 3.3171653747558594,
+      "learning_rate": 0.0006877133105802048,
+      "loss": 5.9773,
+      "step": 2745
+    },
+    {
+      "epoch": 0.9372013651877134,
+      "grad_norm": 3.2578046321868896,
+      "learning_rate": 0.000687599544937429,
+      "loss": 6.0005,
+      "step": 2746
+    },
+    {
+      "epoch": 0.937542662116041,
+      "grad_norm": 3.587801933288574,
+      "learning_rate": 0.0006874857792946531,
+      "loss": 6.7843,
+      "step": 2747
+    },
+    {
+      "epoch": 0.9378839590443686,
+      "grad_norm": 3.3838870525360107,
+      "learning_rate": 0.0006873720136518772,
+      "loss": 6.8512,
+      "step": 2748
+    },
+    {
+      "epoch": 0.9382252559726962,
+      "grad_norm": 3.3424386978149414,
+      "learning_rate": 0.0006872582480091013,
+      "loss": 6.5035,
+      "step": 2749
+    },
+    {
+      "epoch": 0.9385665529010239,
+      "grad_norm": 4.673618793487549,
+      "learning_rate": 0.0006871444823663255,
+      "loss": 4.9139,
+      "step": 2750
+    },
+    {
+      "epoch": 0.9389078498293515,
+      "grad_norm": 3.2579727172851562,
+      "learning_rate": 0.0006870307167235496,
+      "loss": 6.4092,
+      "step": 2751
+    },
+    {
+      "epoch": 0.9392491467576792,
+      "grad_norm": 3.674421787261963,
+      "learning_rate": 0.0006869169510807736,
+      "loss": 6.4869,
+      "step": 2752
+    },
+    {
+      "epoch": 0.9395904436860069,
+      "grad_norm": 3.2421398162841797,
+      "learning_rate": 0.0006868031854379977,
+      "loss": 6.3196,
+      "step": 2753
+    },
+    {
+      "epoch": 0.9399317406143345,
+      "grad_norm": 3.4086074829101562,
+      "learning_rate": 0.0006866894197952218,
+      "loss": 6.2654,
+      "step": 2754
+    },
+    {
+      "epoch": 0.9402730375426621,
+      "grad_norm": 3.3109261989593506,
+      "learning_rate": 0.0006865756541524459,
+      "loss": 6.3046,
+      "step": 2755
+    },
+    {
+      "epoch": 0.9406143344709897,
+      "grad_norm": 3.2505524158477783,
+      "learning_rate": 0.0006864618885096701,
+      "loss": 6.3068,
+      "step": 2756
+    },
+    {
+      "epoch": 0.9409556313993174,
+      "grad_norm": 3.0877838134765625,
+      "learning_rate": 0.0006863481228668942,
+      "loss": 6.4801,
+      "step": 2757
+    },
+    {
+      "epoch": 0.941296928327645,
+      "grad_norm": 3.232034206390381,
+      "learning_rate": 0.0006862343572241183,
+      "loss": 6.1997,
+      "step": 2758
+    },
+    {
+      "epoch": 0.9416382252559727,
+      "grad_norm": 3.7556653022766113,
+      "learning_rate": 0.0006861205915813424,
+      "loss": 6.0221,
+      "step": 2759
+    },
+    {
+      "epoch": 0.9419795221843004,
+      "grad_norm": 3.408632278442383,
+      "learning_rate": 0.0006860068259385665,
+      "loss": 6.4331,
+      "step": 2760
+    },
+    {
+      "epoch": 0.942320819112628,
+      "grad_norm": 6.045160293579102,
+      "learning_rate": 0.0006858930602957906,
+      "loss": 5.5115,
+      "step": 2761
+    },
+    {
+      "epoch": 0.9426621160409556,
+      "grad_norm": 3.6110312938690186,
+      "learning_rate": 0.0006857792946530148,
+      "loss": 6.957,
+      "step": 2762
+    },
+    {
+      "epoch": 0.9430034129692833,
+      "grad_norm": 3.6176974773406982,
+      "learning_rate": 0.000685665529010239,
+      "loss": 6.2635,
+      "step": 2763
+    },
+    {
+      "epoch": 0.9433447098976109,
+      "grad_norm": 3.2292110919952393,
+      "learning_rate": 0.0006855517633674631,
+      "loss": 6.4325,
+      "step": 2764
+    },
+    {
+      "epoch": 0.9436860068259386,
+      "grad_norm": 3.226062536239624,
+      "learning_rate": 0.0006854379977246872,
+      "loss": 6.1578,
+      "step": 2765
+    },
+    {
+      "epoch": 0.9440273037542662,
+      "grad_norm": 3.0933258533477783,
+      "learning_rate": 0.0006853242320819113,
+      "loss": 6.3193,
+      "step": 2766
+    },
+    {
+      "epoch": 0.9443686006825939,
+      "grad_norm": 3.3063440322875977,
+      "learning_rate": 0.0006852104664391355,
+      "loss": 6.4354,
+      "step": 2767
+    },
+    {
+      "epoch": 0.9447098976109215,
+      "grad_norm": 7.300811290740967,
+      "learning_rate": 0.0006850967007963596,
+      "loss": 5.611,
+      "step": 2768
+    },
+    {
+      "epoch": 0.9450511945392491,
+      "grad_norm": 3.557973623275757,
+      "learning_rate": 0.0006849829351535837,
+      "loss": 6.5173,
+      "step": 2769
+    },
+    {
+      "epoch": 0.9453924914675768,
+      "grad_norm": 2.229107141494751,
+      "learning_rate": 0.0006848691695108078,
+      "loss": 3.489,
+      "step": 2770
+    },
+    {
+      "epoch": 0.9457337883959044,
+      "grad_norm": 3.4767768383026123,
+      "learning_rate": 0.0006847554038680319,
+      "loss": 6.4509,
+      "step": 2771
+    },
+    {
+      "epoch": 0.9460750853242321,
+      "grad_norm": 3.309904098510742,
+      "learning_rate": 0.0006846416382252559,
+      "loss": 6.3209,
+      "step": 2772
+    },
+    {
+      "epoch": 0.9464163822525598,
+      "grad_norm": 3.6992416381835938,
+      "learning_rate": 0.0006845278725824801,
+      "loss": 6.4455,
+      "step": 2773
+    },
+    {
+      "epoch": 0.9467576791808874,
+      "grad_norm": 3.4120285511016846,
+      "learning_rate": 0.0006844141069397042,
+      "loss": 6.5287,
+      "step": 2774
+    },
+    {
+      "epoch": 0.947098976109215,
+      "grad_norm": 3.3059194087982178,
+      "learning_rate": 0.0006843003412969283,
+      "loss": 6.4193,
+      "step": 2775
+    },
+    {
+      "epoch": 0.9474402730375426,
+      "grad_norm": 3.727429151535034,
+      "learning_rate": 0.0006841865756541524,
+      "loss": 6.0036,
+      "step": 2776
+    },
+    {
+      "epoch": 0.9477815699658703,
+      "grad_norm": 5.650639057159424,
+      "learning_rate": 0.0006840728100113765,
+      "loss": 5.4394,
+      "step": 2777
+    },
+    {
+      "epoch": 0.948122866894198,
+      "grad_norm": 3.2986202239990234,
+      "learning_rate": 0.0006839590443686006,
+      "loss": 6.7952,
+      "step": 2778
+    },
+    {
+      "epoch": 0.9484641638225256,
+      "grad_norm": 4.086126804351807,
+      "learning_rate": 0.0006838452787258248,
+      "loss": 4.7384,
+      "step": 2779
+    },
+    {
+      "epoch": 0.9488054607508533,
+      "grad_norm": 3.3148763179779053,
+      "learning_rate": 0.000683731513083049,
+      "loss": 6.6121,
+      "step": 2780
+    },
+    {
+      "epoch": 0.9491467576791809,
+      "grad_norm": 3.115473508834839,
+      "learning_rate": 0.0006836177474402731,
+      "loss": 6.375,
+      "step": 2781
+    },
+    {
+      "epoch": 0.9494880546075085,
+      "grad_norm": 4.739993572235107,
+      "learning_rate": 0.0006835039817974972,
+      "loss": 6.0525,
+      "step": 2782
+    },
+    {
+      "epoch": 0.9498293515358361,
+      "grad_norm": 3.0348520278930664,
+      "learning_rate": 0.0006833902161547213,
+      "loss": 6.6161,
+      "step": 2783
+    },
+    {
+      "epoch": 0.9501706484641638,
+      "grad_norm": 4.478902339935303,
+      "learning_rate": 0.0006832764505119454,
+      "loss": 6.0643,
+      "step": 2784
+    },
+    {
+      "epoch": 0.9505119453924915,
+      "grad_norm": 3.3169867992401123,
+      "learning_rate": 0.0006831626848691696,
+      "loss": 6.4104,
+      "step": 2785
+    },
+    {
+      "epoch": 0.9508532423208191,
+      "grad_norm": 3.3145945072174072,
+      "learning_rate": 0.0006830489192263937,
+      "loss": 6.6644,
+      "step": 2786
+    },
+    {
+      "epoch": 0.9511945392491468,
+      "grad_norm": 3.361513614654541,
+      "learning_rate": 0.0006829351535836178,
+      "loss": 6.4348,
+      "step": 2787
+    },
+    {
+      "epoch": 0.9515358361774744,
+      "grad_norm": 3.221954584121704,
+      "learning_rate": 0.0006828213879408419,
+      "loss": 6.8069,
+      "step": 2788
+    },
+    {
+      "epoch": 0.951877133105802,
+      "grad_norm": 3.3777694702148438,
+      "learning_rate": 0.000682707622298066,
+      "loss": 5.9127,
+      "step": 2789
+    },
+    {
+      "epoch": 0.9522184300341296,
+      "grad_norm": 4.066497325897217,
+      "learning_rate": 0.0006825938566552902,
+      "loss": 5.7392,
+      "step": 2790
+    },
+    {
+      "epoch": 0.9525597269624574,
+      "grad_norm": 3.3121917247772217,
+      "learning_rate": 0.0006824800910125142,
+      "loss": 6.9516,
+      "step": 2791
+    },
+    {
+      "epoch": 0.952901023890785,
+      "grad_norm": 3.3323752880096436,
+      "learning_rate": 0.0006823663253697383,
+      "loss": 6.7219,
+      "step": 2792
+    },
+    {
+      "epoch": 0.9532423208191126,
+      "grad_norm": 3.7000746726989746,
+      "learning_rate": 0.0006822525597269624,
+      "loss": 6.541,
+      "step": 2793
+    },
+    {
+      "epoch": 0.9535836177474403,
+      "grad_norm": 3.4040660858154297,
+      "learning_rate": 0.0006821387940841865,
+      "loss": 6.7015,
+      "step": 2794
+    },
+    {
+      "epoch": 0.9539249146757679,
+      "grad_norm": 3.9156670570373535,
+      "learning_rate": 0.0006820250284414106,
+      "loss": 6.447,
+      "step": 2795
+    },
+    {
+      "epoch": 0.9542662116040955,
+      "grad_norm": 3.4194421768188477,
+      "learning_rate": 0.0006819112627986348,
+      "loss": 6.6974,
+      "step": 2796
+    },
+    {
+      "epoch": 0.9546075085324232,
+      "grad_norm": 4.750198841094971,
+      "learning_rate": 0.000681797497155859,
+      "loss": 6.2898,
+      "step": 2797
+    },
+    {
+      "epoch": 0.9549488054607509,
+      "grad_norm": 3.528273582458496,
+      "learning_rate": 0.0006816837315130831,
+      "loss": 6.2098,
+      "step": 2798
+    },
+    {
+      "epoch": 0.9552901023890785,
+      "grad_norm": 3.286182403564453,
+      "learning_rate": 0.0006815699658703072,
+      "loss": 6.7786,
+      "step": 2799
+    },
+    {
+      "epoch": 0.9556313993174061,
+      "grad_norm": 3.299617052078247,
+      "learning_rate": 0.0006814562002275313,
+      "loss": 6.6818,
+      "step": 2800
+    },
+    {
+      "epoch": 0.9559726962457338,
+      "grad_norm": 3.2190589904785156,
+      "learning_rate": 0.0006813424345847554,
+      "loss": 6.15,
+      "step": 2801
+    },
+    {
+      "epoch": 0.9563139931740614,
+      "grad_norm": 3.2283973693847656,
+      "learning_rate": 0.0006812286689419796,
+      "loss": 5.7856,
+      "step": 2802
+    },
+    {
+      "epoch": 0.956655290102389,
+      "grad_norm": 3.121652364730835,
+      "learning_rate": 0.0006811149032992037,
+      "loss": 6.5968,
+      "step": 2803
+    },
+    {
+      "epoch": 0.9569965870307168,
+      "grad_norm": 3.343250274658203,
+      "learning_rate": 0.0006810011376564278,
+      "loss": 6.9726,
+      "step": 2804
+    },
+    {
+      "epoch": 0.9573378839590444,
+      "grad_norm": 3.670442819595337,
+      "learning_rate": 0.0006808873720136519,
+      "loss": 5.8369,
+      "step": 2805
+    },
+    {
+      "epoch": 0.957679180887372,
+      "grad_norm": 3.4261021614074707,
+      "learning_rate": 0.000680773606370876,
+      "loss": 6.7033,
+      "step": 2806
+    },
+    {
+      "epoch": 0.9580204778156997,
+      "grad_norm": 3.3535799980163574,
+      "learning_rate": 0.0006806598407281002,
+      "loss": 5.2909,
+      "step": 2807
+    },
+    {
+      "epoch": 0.9583617747440273,
+      "grad_norm": 3.462440252304077,
+      "learning_rate": 0.0006805460750853243,
+      "loss": 5.9633,
+      "step": 2808
+    },
+    {
+      "epoch": 0.9587030716723549,
+      "grad_norm": 3.2325925827026367,
+      "learning_rate": 0.0006804323094425484,
+      "loss": 6.4796,
+      "step": 2809
+    },
+    {
+      "epoch": 0.9590443686006825,
+      "grad_norm": 3.415300130844116,
+      "learning_rate": 0.0006803185437997725,
+      "loss": 6.298,
+      "step": 2810
+    },
+    {
+      "epoch": 0.9593856655290103,
+      "grad_norm": 3.2603628635406494,
+      "learning_rate": 0.0006802047781569965,
+      "loss": 6.9542,
+      "step": 2811
+    },
+    {
+      "epoch": 0.9597269624573379,
+      "grad_norm": 3.158876657485962,
+      "learning_rate": 0.0006800910125142206,
+      "loss": 6.7321,
+      "step": 2812
+    },
+    {
+      "epoch": 0.9600682593856655,
+      "grad_norm": 3.137831687927246,
+      "learning_rate": 0.0006799772468714448,
+      "loss": 6.1844,
+      "step": 2813
+    },
+    {
+      "epoch": 0.9604095563139932,
+      "grad_norm": 3.162550687789917,
+      "learning_rate": 0.000679863481228669,
+      "loss": 6.5377,
+      "step": 2814
+    },
+    {
+      "epoch": 0.9607508532423208,
+      "grad_norm": 3.1445960998535156,
+      "learning_rate": 0.0006797497155858931,
+      "loss": 6.4721,
+      "step": 2815
+    },
+    {
+      "epoch": 0.9610921501706484,
+      "grad_norm": 3.4892373085021973,
+      "learning_rate": 0.0006796359499431172,
+      "loss": 6.4251,
+      "step": 2816
+    },
+    {
+      "epoch": 0.9614334470989762,
+      "grad_norm": 3.36893892288208,
+      "learning_rate": 0.0006795221843003413,
+      "loss": 5.9855,
+      "step": 2817
+    },
+    {
+      "epoch": 0.9617747440273038,
+      "grad_norm": 3.3693227767944336,
+      "learning_rate": 0.0006794084186575654,
+      "loss": 6.8197,
+      "step": 2818
+    },
+    {
+      "epoch": 0.9621160409556314,
+      "grad_norm": 3.2168221473693848,
+      "learning_rate": 0.0006792946530147896,
+      "loss": 6.6614,
+      "step": 2819
+    },
+    {
+      "epoch": 0.962457337883959,
+      "grad_norm": 3.355677604675293,
+      "learning_rate": 0.0006791808873720137,
+      "loss": 6.1324,
+      "step": 2820
+    },
+    {
+      "epoch": 0.9627986348122867,
+      "grad_norm": 3.28233003616333,
+      "learning_rate": 0.0006790671217292378,
+      "loss": 6.5953,
+      "step": 2821
+    },
+    {
+      "epoch": 0.9631399317406143,
+      "grad_norm": 3.175426959991455,
+      "learning_rate": 0.0006789533560864619,
+      "loss": 6.4139,
+      "step": 2822
+    },
+    {
+      "epoch": 0.9634812286689419,
+      "grad_norm": 3.2525055408477783,
+      "learning_rate": 0.000678839590443686,
+      "loss": 6.8946,
+      "step": 2823
+    },
+    {
+      "epoch": 0.9638225255972697,
+      "grad_norm": 3.2364585399627686,
+      "learning_rate": 0.0006787258248009101,
+      "loss": 6.5195,
+      "step": 2824
+    },
+    {
+      "epoch": 0.9641638225255973,
+      "grad_norm": 3.8817360401153564,
+      "learning_rate": 0.0006786120591581343,
+      "loss": 5.2455,
+      "step": 2825
+    },
+    {
+      "epoch": 0.9645051194539249,
+      "grad_norm": 3.36077618598938,
+      "learning_rate": 0.0006784982935153584,
+      "loss": 5.7979,
+      "step": 2826
+    },
+    {
+      "epoch": 0.9648464163822525,
+      "grad_norm": 3.421419620513916,
+      "learning_rate": 0.0006783845278725825,
+      "loss": 6.6912,
+      "step": 2827
+    },
+    {
+      "epoch": 0.9651877133105802,
+      "grad_norm": 3.2399160861968994,
+      "learning_rate": 0.0006782707622298066,
+      "loss": 6.7359,
+      "step": 2828
+    },
+    {
+      "epoch": 0.9655290102389078,
+      "grad_norm": 3.5829226970672607,
+      "learning_rate": 0.0006781569965870307,
+      "loss": 6.447,
+      "step": 2829
+    },
+    {
+      "epoch": 0.9658703071672355,
+      "grad_norm": 3.9516849517822266,
+      "learning_rate": 0.0006780432309442548,
+      "loss": 6.1873,
+      "step": 2830
+    },
+    {
+      "epoch": 0.9662116040955632,
+      "grad_norm": 3.371225595474243,
+      "learning_rate": 0.000677929465301479,
+      "loss": 6.8074,
+      "step": 2831
+    },
+    {
+      "epoch": 0.9665529010238908,
+      "grad_norm": 3.1678028106689453,
+      "learning_rate": 0.0006778156996587031,
+      "loss": 6.8476,
+      "step": 2832
+    },
+    {
+      "epoch": 0.9668941979522184,
+      "grad_norm": 3.185497283935547,
+      "learning_rate": 0.0006777019340159272,
+      "loss": 6.7461,
+      "step": 2833
+    },
+    {
+      "epoch": 0.967235494880546,
+      "grad_norm": 6.401412487030029,
+      "learning_rate": 0.0006775881683731513,
+      "loss": 6.6854,
+      "step": 2834
+    },
+    {
+      "epoch": 0.9675767918088737,
+      "grad_norm": 4.533454418182373,
+      "learning_rate": 0.0006774744027303754,
+      "loss": 5.6879,
+      "step": 2835
+    },
+    {
+      "epoch": 0.9679180887372013,
+      "grad_norm": 4.112688064575195,
+      "learning_rate": 0.0006773606370875996,
+      "loss": 6.5312,
+      "step": 2836
+    },
+    {
+      "epoch": 0.968259385665529,
+      "grad_norm": 3.4008233547210693,
+      "learning_rate": 0.0006772468714448237,
+      "loss": 6.5382,
+      "step": 2837
+    },
+    {
+      "epoch": 0.9686006825938567,
+      "grad_norm": 3.319769859313965,
+      "learning_rate": 0.0006771331058020478,
+      "loss": 6.3743,
+      "step": 2838
+    },
+    {
+      "epoch": 0.9689419795221843,
+      "grad_norm": 3.506112813949585,
+      "learning_rate": 0.0006770193401592719,
+      "loss": 5.7598,
+      "step": 2839
+    },
+    {
+      "epoch": 0.9692832764505119,
+      "grad_norm": 3.2987313270568848,
+      "learning_rate": 0.000676905574516496,
+      "loss": 6.7194,
+      "step": 2840
+    },
+    {
+      "epoch": 0.9696245733788396,
+      "grad_norm": 3.148381233215332,
+      "learning_rate": 0.0006767918088737201,
+      "loss": 6.5178,
+      "step": 2841
+    },
+    {
+      "epoch": 0.9699658703071672,
+      "grad_norm": 3.224257230758667,
+      "learning_rate": 0.0006766780432309443,
+      "loss": 6.1293,
+      "step": 2842
+    },
+    {
+      "epoch": 0.9703071672354949,
+      "grad_norm": 3.166309356689453,
+      "learning_rate": 0.0006765642775881684,
+      "loss": 6.7261,
+      "step": 2843
+    },
+    {
+      "epoch": 0.9706484641638226,
+      "grad_norm": 3.201484441757202,
+      "learning_rate": 0.0006764505119453925,
+      "loss": 6.6831,
+      "step": 2844
+    },
+    {
+      "epoch": 0.9709897610921502,
+      "grad_norm": 3.366837501525879,
+      "learning_rate": 0.0006763367463026166,
+      "loss": 6.6644,
+      "step": 2845
+    },
+    {
+      "epoch": 0.9713310580204778,
+      "grad_norm": 3.205770969390869,
+      "learning_rate": 0.0006762229806598407,
+      "loss": 6.787,
+      "step": 2846
+    },
+    {
+      "epoch": 0.9716723549488054,
+      "grad_norm": 3.400615692138672,
+      "learning_rate": 0.000676109215017065,
+      "loss": 6.4538,
+      "step": 2847
+    },
+    {
+      "epoch": 0.9720136518771331,
+      "grad_norm": 3.295478343963623,
+      "learning_rate": 0.0006759954493742891,
+      "loss": 7.0216,
+      "step": 2848
+    },
+    {
+      "epoch": 0.9723549488054608,
+      "grad_norm": 3.9448795318603516,
+      "learning_rate": 0.0006758816837315131,
+      "loss": 6.3134,
+      "step": 2849
+    },
+    {
+      "epoch": 0.9726962457337884,
+      "grad_norm": 3.3810625076293945,
+      "learning_rate": 0.0006757679180887372,
+      "loss": 6.5216,
+      "step": 2850
+    },
+    {
+      "epoch": 0.9730375426621161,
+      "grad_norm": 3.364837169647217,
+      "learning_rate": 0.0006756541524459613,
+      "loss": 6.4476,
+      "step": 2851
+    },
+    {
+      "epoch": 0.9733788395904437,
+      "grad_norm": 3.2336301803588867,
+      "learning_rate": 0.0006755403868031854,
+      "loss": 6.19,
+      "step": 2852
+    },
+    {
+      "epoch": 0.9737201365187713,
+      "grad_norm": 3.081376314163208,
+      "learning_rate": 0.0006754266211604096,
+      "loss": 6.6487,
+      "step": 2853
+    },
+    {
+      "epoch": 0.974061433447099,
+      "grad_norm": 3.2751305103302,
+      "learning_rate": 0.0006753128555176337,
+      "loss": 6.3007,
+      "step": 2854
+    },
+    {
+      "epoch": 0.9744027303754266,
+      "grad_norm": 6.972997665405273,
+      "learning_rate": 0.0006751990898748578,
+      "loss": 5.3055,
+      "step": 2855
+    },
+    {
+      "epoch": 0.9747440273037543,
+      "grad_norm": 3.2816948890686035,
+      "learning_rate": 0.0006750853242320819,
+      "loss": 6.5868,
+      "step": 2856
+    },
+    {
+      "epoch": 0.9750853242320819,
+      "grad_norm": 3.3620402812957764,
+      "learning_rate": 0.000674971558589306,
+      "loss": 6.5817,
+      "step": 2857
+    },
+    {
+      "epoch": 0.9754266211604096,
+      "grad_norm": 3.404201030731201,
+      "learning_rate": 0.0006748577929465301,
+      "loss": 6.3418,
+      "step": 2858
+    },
+    {
+      "epoch": 0.9757679180887372,
+      "grad_norm": 3.4882445335388184,
+      "learning_rate": 0.0006747440273037543,
+      "loss": 5.5313,
+      "step": 2859
+    },
+    {
+      "epoch": 0.9761092150170648,
+      "grad_norm": 3.3639590740203857,
+      "learning_rate": 0.0006746302616609784,
+      "loss": 6.7607,
+      "step": 2860
+    },
+    {
+      "epoch": 0.9764505119453925,
+      "grad_norm": 3.0582730770111084,
+      "learning_rate": 0.0006745164960182025,
+      "loss": 6.4016,
+      "step": 2861
+    },
+    {
+      "epoch": 0.9767918088737202,
+      "grad_norm": 3.409369945526123,
+      "learning_rate": 0.0006744027303754266,
+      "loss": 6.8788,
+      "step": 2862
+    },
+    {
+      "epoch": 0.9771331058020478,
+      "grad_norm": 3.20253849029541,
+      "learning_rate": 0.0006742889647326507,
+      "loss": 6.3911,
+      "step": 2863
+    },
+    {
+      "epoch": 0.9774744027303754,
+      "grad_norm": 3.5735726356506348,
+      "learning_rate": 0.0006741751990898749,
+      "loss": 6.0568,
+      "step": 2864
+    },
+    {
+      "epoch": 0.9778156996587031,
+      "grad_norm": 3.4013800621032715,
+      "learning_rate": 0.0006740614334470991,
+      "loss": 6.2177,
+      "step": 2865
+    },
+    {
+      "epoch": 0.9781569965870307,
+      "grad_norm": 5.85659646987915,
+      "learning_rate": 0.0006739476678043232,
+      "loss": 5.692,
+      "step": 2866
+    },
+    {
+      "epoch": 0.9784982935153583,
+      "grad_norm": 3.534287214279175,
+      "learning_rate": 0.0006738339021615473,
+      "loss": 6.3442,
+      "step": 2867
+    },
+    {
+      "epoch": 0.978839590443686,
+      "grad_norm": 3.368276357650757,
+      "learning_rate": 0.0006737201365187714,
+      "loss": 6.6924,
+      "step": 2868
+    },
+    {
+      "epoch": 0.9791808873720137,
+      "grad_norm": 3.3042497634887695,
+      "learning_rate": 0.0006736063708759954,
+      "loss": 6.2767,
+      "step": 2869
+    },
+    {
+      "epoch": 0.9795221843003413,
+      "grad_norm": 3.2396910190582275,
+      "learning_rate": 0.0006734926052332196,
+      "loss": 6.6282,
+      "step": 2870
+    },
+    {
+      "epoch": 0.979863481228669,
+      "grad_norm": 3.2214739322662354,
+      "learning_rate": 0.0006733788395904437,
+      "loss": 6.5097,
+      "step": 2871
+    },
+    {
+      "epoch": 0.9802047781569966,
+      "grad_norm": 3.2005984783172607,
+      "learning_rate": 0.0006732650739476678,
+      "loss": 6.2032,
+      "step": 2872
+    },
+    {
+      "epoch": 0.9805460750853242,
+      "grad_norm": 4.3526105880737305,
+      "learning_rate": 0.0006731513083048919,
+      "loss": 6.1474,
+      "step": 2873
+    },
+    {
+      "epoch": 0.9808873720136518,
+      "grad_norm": 3.500005006790161,
+      "learning_rate": 0.000673037542662116,
+      "loss": 6.3671,
+      "step": 2874
+    },
+    {
+      "epoch": 0.9812286689419796,
+      "grad_norm": 4.856510162353516,
+      "learning_rate": 0.0006729237770193401,
+      "loss": 5.3342,
+      "step": 2875
+    },
+    {
+      "epoch": 0.9815699658703072,
+      "grad_norm": 3.3631317615509033,
+      "learning_rate": 0.0006728100113765643,
+      "loss": 6.9552,
+      "step": 2876
+    },
+    {
+      "epoch": 0.9819112627986348,
+      "grad_norm": 4.488306999206543,
+      "learning_rate": 0.0006726962457337884,
+      "loss": 5.2524,
+      "step": 2877
+    },
+    {
+      "epoch": 0.9822525597269625,
+      "grad_norm": 3.190113067626953,
+      "learning_rate": 0.0006725824800910125,
+      "loss": 6.5939,
+      "step": 2878
+    },
+    {
+      "epoch": 0.9825938566552901,
+      "grad_norm": 3.2531468868255615,
+      "learning_rate": 0.0006724687144482366,
+      "loss": 6.8033,
+      "step": 2879
+    },
+    {
+      "epoch": 0.9829351535836177,
+      "grad_norm": 3.396819591522217,
+      "learning_rate": 0.0006723549488054607,
+      "loss": 6.5321,
+      "step": 2880
+    },
+    {
+      "epoch": 0.9832764505119453,
+      "grad_norm": 3.1268954277038574,
+      "learning_rate": 0.0006722411831626849,
+      "loss": 6.7329,
+      "step": 2881
+    },
+    {
+      "epoch": 0.9836177474402731,
+      "grad_norm": 3.4589643478393555,
+      "learning_rate": 0.0006721274175199091,
+      "loss": 6.5176,
+      "step": 2882
+    },
+    {
+      "epoch": 0.9839590443686007,
+      "grad_norm": 3.242070198059082,
+      "learning_rate": 0.0006720136518771332,
+      "loss": 6.6396,
+      "step": 2883
+    },
+    {
+      "epoch": 0.9843003412969283,
+      "grad_norm": 3.2771873474121094,
+      "learning_rate": 0.0006718998862343573,
+      "loss": 6.3999,
+      "step": 2884
+    },
+    {
+      "epoch": 0.984641638225256,
+      "grad_norm": 3.497062921524048,
+      "learning_rate": 0.0006717861205915814,
+      "loss": 6.2288,
+      "step": 2885
+    },
+    {
+      "epoch": 0.9849829351535836,
+      "grad_norm": 3.2468268871307373,
+      "learning_rate": 0.0006716723549488055,
+      "loss": 6.7244,
+      "step": 2886
+    },
+    {
+      "epoch": 0.9853242320819112,
+      "grad_norm": 3.181048631668091,
+      "learning_rate": 0.0006715585893060296,
+      "loss": 6.9261,
+      "step": 2887
+    },
+    {
+      "epoch": 0.985665529010239,
+      "grad_norm": 3.220104455947876,
+      "learning_rate": 0.0006714448236632537,
+      "loss": 6.4713,
+      "step": 2888
+    },
+    {
+      "epoch": 0.9860068259385666,
+      "grad_norm": 3.5669116973876953,
+      "learning_rate": 0.0006713310580204778,
+      "loss": 5.612,
+      "step": 2889
+    },
+    {
+      "epoch": 0.9863481228668942,
+      "grad_norm": 3.239811897277832,
+      "learning_rate": 0.0006712172923777019,
+      "loss": 6.8196,
+      "step": 2890
+    },
+    {
+      "epoch": 0.9866894197952218,
+      "grad_norm": 3.4331865310668945,
+      "learning_rate": 0.000671103526734926,
+      "loss": 6.6693,
+      "step": 2891
+    },
+    {
+      "epoch": 0.9870307167235495,
+      "grad_norm": 3.3848814964294434,
+      "learning_rate": 0.0006709897610921501,
+      "loss": 7.0898,
+      "step": 2892
+    },
+    {
+      "epoch": 0.9873720136518771,
+      "grad_norm": 3.2575674057006836,
+      "learning_rate": 0.0006708759954493743,
+      "loss": 6.7033,
+      "step": 2893
+    },
+    {
+      "epoch": 0.9877133105802047,
+      "grad_norm": 3.440713882446289,
+      "learning_rate": 0.0006707622298065984,
+      "loss": 6.4287,
+      "step": 2894
+    },
+    {
+      "epoch": 0.9880546075085325,
+      "grad_norm": 3.2217304706573486,
+      "learning_rate": 0.0006706484641638225,
+      "loss": 6.721,
+      "step": 2895
+    },
+    {
+      "epoch": 0.9883959044368601,
+      "grad_norm": 3.2484257221221924,
+      "learning_rate": 0.0006705346985210466,
+      "loss": 6.8919,
+      "step": 2896
+    },
+    {
+      "epoch": 0.9887372013651877,
+      "grad_norm": 3.206747055053711,
+      "learning_rate": 0.0006704209328782707,
+      "loss": 6.5002,
+      "step": 2897
+    },
+    {
+      "epoch": 0.9890784982935154,
+      "grad_norm": 3.30210280418396,
+      "learning_rate": 0.0006703071672354949,
+      "loss": 6.0187,
+      "step": 2898
+    },
+    {
+      "epoch": 0.989419795221843,
+      "grad_norm": 2.3465044498443604,
+      "learning_rate": 0.0006701934015927191,
+      "loss": 3.2803,
+      "step": 2899
+    },
+    {
+      "epoch": 0.9897610921501706,
+      "grad_norm": 3.455019950866699,
+      "learning_rate": 0.0006700796359499432,
+      "loss": 6.937,
+      "step": 2900
+    },
+    {
+      "epoch": 0.9901023890784983,
+      "grad_norm": 3.3712611198425293,
+      "learning_rate": 0.0006699658703071673,
+      "loss": 6.6826,
+      "step": 2901
+    },
+    {
+      "epoch": 0.990443686006826,
+      "grad_norm": 3.372312068939209,
+      "learning_rate": 0.0006698521046643914,
+      "loss": 6.529,
+      "step": 2902
+    },
+    {
+      "epoch": 0.9907849829351536,
+      "grad_norm": 3.750188112258911,
+      "learning_rate": 0.0006697383390216155,
+      "loss": 6.4456,
+      "step": 2903
+    },
+    {
+      "epoch": 0.9911262798634812,
+      "grad_norm": 3.316756010055542,
+      "learning_rate": 0.0006696245733788396,
+      "loss": 6.311,
+      "step": 2904
+    },
+    {
+      "epoch": 0.9914675767918089,
+      "grad_norm": 3.249242067337036,
+      "learning_rate": 0.0006695108077360638,
+      "loss": 6.9939,
+      "step": 2905
+    },
+    {
+      "epoch": 0.9918088737201365,
+      "grad_norm": 3.5320539474487305,
+      "learning_rate": 0.0006693970420932879,
+      "loss": 6.3807,
+      "step": 2906
+    },
+    {
+      "epoch": 0.9921501706484641,
+      "grad_norm": 3.2891733646392822,
+      "learning_rate": 0.000669283276450512,
+      "loss": 6.5872,
+      "step": 2907
+    },
+    {
+      "epoch": 0.9924914675767919,
+      "grad_norm": 3.310703754425049,
+      "learning_rate": 0.000669169510807736,
+      "loss": 6.7432,
+      "step": 2908
+    },
+    {
+      "epoch": 0.9928327645051195,
+      "grad_norm": 3.4947831630706787,
+      "learning_rate": 0.0006690557451649601,
+      "loss": 6.6943,
+      "step": 2909
+    },
+    {
+      "epoch": 0.9931740614334471,
+      "grad_norm": 6.251656532287598,
+      "learning_rate": 0.0006689419795221842,
+      "loss": 5.9758,
+      "step": 2910
+    },
+    {
+      "epoch": 0.9935153583617747,
+      "grad_norm": 3.236522912979126,
+      "learning_rate": 0.0006688282138794084,
+      "loss": 6.7063,
+      "step": 2911
+    },
+    {
+      "epoch": 0.9938566552901024,
+      "grad_norm": 3.4270472526550293,
+      "learning_rate": 0.0006687144482366325,
+      "loss": 6.9056,
+      "step": 2912
+    },
+    {
+      "epoch": 0.99419795221843,
+      "grad_norm": 3.3885722160339355,
+      "learning_rate": 0.0006686006825938566,
+      "loss": 6.6033,
+      "step": 2913
+    },
+    {
+      "epoch": 0.9945392491467577,
+      "grad_norm": 3.3846211433410645,
+      "learning_rate": 0.0006684869169510807,
+      "loss": 6.8574,
+      "step": 2914
+    },
+    {
+      "epoch": 0.9948805460750854,
+      "grad_norm": 3.4704039096832275,
+      "learning_rate": 0.0006683731513083049,
+      "loss": 6.2599,
+      "step": 2915
+    },
+    {
+      "epoch": 0.995221843003413,
+      "grad_norm": 3.076455593109131,
+      "learning_rate": 0.0006682593856655291,
+      "loss": 6.8536,
+      "step": 2916
+    },
+    {
+      "epoch": 0.9955631399317406,
+      "grad_norm": 3.1758480072021484,
+      "learning_rate": 0.0006681456200227532,
+      "loss": 6.7996,
+      "step": 2917
+    },
+    {
+      "epoch": 0.9959044368600682,
+      "grad_norm": 4.0002851486206055,
+      "learning_rate": 0.0006680318543799773,
+      "loss": 5.2598,
+      "step": 2918
+    },
+    {
+      "epoch": 0.9962457337883959,
+      "grad_norm": 5.545552730560303,
+      "learning_rate": 0.0006679180887372014,
+      "loss": 5.8838,
+      "step": 2919
+    },
+    {
+      "epoch": 0.9965870307167235,
+      "grad_norm": 3.345064163208008,
+      "learning_rate": 0.0006678043230944255,
+      "loss": 6.566,
+      "step": 2920
+    },
+    {
+      "epoch": 0.9969283276450512,
+      "grad_norm": 3.4302754402160645,
+      "learning_rate": 0.0006676905574516496,
+      "loss": 7.0484,
+      "step": 2921
+    },
+    {
+      "epoch": 0.9972696245733789,
+      "grad_norm": 3.445356607437134,
+      "learning_rate": 0.0006675767918088738,
+      "loss": 6.7516,
+      "step": 2922
+    },
+    {
+      "epoch": 0.9976109215017065,
+      "grad_norm": 3.258596181869507,
+      "learning_rate": 0.0006674630261660979,
+      "loss": 6.5921,
+      "step": 2923
+    },
+    {
+      "epoch": 0.9979522184300341,
+      "grad_norm": 3.594090700149536,
+      "learning_rate": 0.000667349260523322,
+      "loss": 6.6817,
+      "step": 2924
+    },
+    {
+      "epoch": 0.9982935153583617,
+      "grad_norm": 4.468551158905029,
+      "learning_rate": 0.0006672354948805461,
+      "loss": 6.0565,
+      "step": 2925
+    },
+    {
+      "epoch": 0.9986348122866894,
+      "grad_norm": 3.330815076828003,
+      "learning_rate": 0.0006671217292377702,
+      "loss": 6.2681,
+      "step": 2926
+    },
+    {
+      "epoch": 0.9989761092150171,
+      "grad_norm": 3.3018605709075928,
+      "learning_rate": 0.0006670079635949942,
+      "loss": 6.4861,
+      "step": 2927
+    },
+    {
+      "epoch": 0.9993174061433447,
+      "grad_norm": 3.5475411415100098,
+      "learning_rate": 0.0006668941979522184,
+      "loss": 6.3887,
+      "step": 2928
+    },
+    {
+      "epoch": 0.9996587030716724,
+      "grad_norm": 3.245288133621216,
+      "learning_rate": 0.0006667804323094425,
+      "loss": 6.7998,
+      "step": 2929
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.472442388534546,
+      "learning_rate": 0.0006666666666666666,
+      "loss": 6.3197,
+      "step": 2930
+    },
+    {
+      "epoch": 1.0003412969283276,
+      "grad_norm": 3.2570908069610596,
+      "learning_rate": 0.0006665529010238907,
+      "loss": 6.4722,
+      "step": 2931
+    },
+    {
+      "epoch": 1.0006825938566553,
+      "grad_norm": 3.1390857696533203,
+      "learning_rate": 0.0006664391353811149,
+      "loss": 6.8016,
+      "step": 2932
+    },
+    {
+      "epoch": 1.0010238907849829,
+      "grad_norm": 3.2926814556121826,
+      "learning_rate": 0.0006663253697383391,
+      "loss": 6.5122,
+      "step": 2933
+    },
+    {
+      "epoch": 1.0013651877133105,
+      "grad_norm": 4.434762477874756,
+      "learning_rate": 0.0006662116040955632,
+      "loss": 4.9442,
+      "step": 2934
+    },
+    {
+      "epoch": 1.0017064846416381,
+      "grad_norm": 3.5702288150787354,
+      "learning_rate": 0.0006660978384527873,
+      "loss": 5.8109,
+      "step": 2935
+    },
+    {
+      "epoch": 1.0020477815699658,
+      "grad_norm": 3.399153709411621,
+      "learning_rate": 0.0006659840728100114,
+      "loss": 7.0274,
+      "step": 2936
+    },
+    {
+      "epoch": 1.0023890784982936,
+      "grad_norm": 3.211782455444336,
+      "learning_rate": 0.0006658703071672355,
+      "loss": 6.4682,
+      "step": 2937
+    },
+    {
+      "epoch": 1.0027303754266212,
+      "grad_norm": 3.3782007694244385,
+      "learning_rate": 0.0006657565415244596,
+      "loss": 5.5324,
+      "step": 2938
+    },
+    {
+      "epoch": 1.0030716723549489,
+      "grad_norm": 3.238039493560791,
+      "learning_rate": 0.0006656427758816838,
+      "loss": 6.6212,
+      "step": 2939
+    },
+    {
+      "epoch": 1.0034129692832765,
+      "grad_norm": 3.5137596130371094,
+      "learning_rate": 0.0006655290102389079,
+      "loss": 6.15,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0037542662116041,
+      "grad_norm": 4.969235897064209,
+      "learning_rate": 0.000665415244596132,
+      "loss": 6.0722,
+      "step": 2941
+    },
+    {
+      "epoch": 1.0040955631399318,
+      "grad_norm": 3.3056838512420654,
+      "learning_rate": 0.0006653014789533561,
+      "loss": 7.2703,
+      "step": 2942
+    },
+    {
+      "epoch": 1.0044368600682594,
+      "grad_norm": 3.27779221534729,
+      "learning_rate": 0.0006651877133105802,
+      "loss": 6.3094,
+      "step": 2943
+    },
+    {
+      "epoch": 1.004778156996587,
+      "grad_norm": 3.4663007259368896,
+      "learning_rate": 0.0006650739476678043,
+      "loss": 6.4624,
+      "step": 2944
+    },
+    {
+      "epoch": 1.0051194539249146,
+      "grad_norm": 3.294342279434204,
+      "learning_rate": 0.0006649601820250285,
+      "loss": 6.1171,
+      "step": 2945
+    },
+    {
+      "epoch": 1.0054607508532423,
+      "grad_norm": 3.324336051940918,
+      "learning_rate": 0.0006648464163822526,
+      "loss": 6.8682,
+      "step": 2946
+    },
+    {
+      "epoch": 1.00580204778157,
+      "grad_norm": 3.631251335144043,
+      "learning_rate": 0.0006647326507394766,
+      "loss": 5.8187,
+      "step": 2947
+    },
+    {
+      "epoch": 1.0061433447098975,
+      "grad_norm": 4.85386848449707,
+      "learning_rate": 0.0006646188850967008,
+      "loss": 5.7936,
+      "step": 2948
+    },
+    {
+      "epoch": 1.0064846416382252,
+      "grad_norm": 3.450941324234009,
+      "learning_rate": 0.0006645051194539249,
+      "loss": 6.8983,
+      "step": 2949
+    },
+    {
+      "epoch": 1.006825938566553,
+      "grad_norm": 3.294748544692993,
+      "learning_rate": 0.000664391353811149,
+      "loss": 5.9347,
+      "step": 2950
+    },
+    {
+      "epoch": 1.0071672354948806,
+      "grad_norm": 3.2384605407714844,
+      "learning_rate": 0.0006642775881683732,
+      "loss": 6.8953,
+      "step": 2951
+    },
+    {
+      "epoch": 1.0075085324232083,
+      "grad_norm": 3.294921875,
+      "learning_rate": 0.0006641638225255973,
+      "loss": 6.7133,
+      "step": 2952
+    },
+    {
+      "epoch": 1.0078498293515359,
+      "grad_norm": 3.22776198387146,
+      "learning_rate": 0.0006640500568828214,
+      "loss": 6.7304,
+      "step": 2953
+    },
+    {
+      "epoch": 1.0081911262798635,
+      "grad_norm": 4.246631145477295,
+      "learning_rate": 0.0006639362912400455,
+      "loss": 4.3664,
+      "step": 2954
+    },
+    {
+      "epoch": 1.0085324232081911,
+      "grad_norm": 3.2894020080566406,
+      "learning_rate": 0.0006638225255972696,
+      "loss": 6.795,
+      "step": 2955
+    },
+    {
+      "epoch": 1.0088737201365188,
+      "grad_norm": 3.3010354042053223,
+      "learning_rate": 0.0006637087599544938,
+      "loss": 6.5039,
+      "step": 2956
+    },
+    {
+      "epoch": 1.0092150170648464,
+      "grad_norm": 3.403155565261841,
+      "learning_rate": 0.0006635949943117179,
+      "loss": 6.5983,
+      "step": 2957
+    },
+    {
+      "epoch": 1.009556313993174,
+      "grad_norm": 3.287907123565674,
+      "learning_rate": 0.000663481228668942,
+      "loss": 5.8215,
+      "step": 2958
+    },
+    {
+      "epoch": 1.0098976109215017,
+      "grad_norm": 3.528432607650757,
+      "learning_rate": 0.0006633674630261661,
+      "loss": 5.889,
+      "step": 2959
+    },
+    {
+      "epoch": 1.0102389078498293,
+      "grad_norm": 3.363466739654541,
+      "learning_rate": 0.0006632536973833902,
+      "loss": 6.8933,
+      "step": 2960
+    },
+    {
+      "epoch": 1.010580204778157,
+      "grad_norm": 3.36714243888855,
+      "learning_rate": 0.0006631399317406143,
+      "loss": 6.4269,
+      "step": 2961
+    },
+    {
+      "epoch": 1.0109215017064845,
+      "grad_norm": 3.2045083045959473,
+      "learning_rate": 0.0006630261660978385,
+      "loss": 6.6594,
+      "step": 2962
+    },
+    {
+      "epoch": 1.0112627986348124,
+      "grad_norm": 3.256004571914673,
+      "learning_rate": 0.0006629124004550626,
+      "loss": 6.5809,
+      "step": 2963
+    },
+    {
+      "epoch": 1.01160409556314,
+      "grad_norm": 3.3133018016815186,
+      "learning_rate": 0.0006627986348122868,
+      "loss": 6.3003,
+      "step": 2964
+    },
+    {
+      "epoch": 1.0119453924914676,
+      "grad_norm": 3.8093559741973877,
+      "learning_rate": 0.0006626848691695109,
+      "loss": 6.2441,
+      "step": 2965
+    },
+    {
+      "epoch": 1.0122866894197953,
+      "grad_norm": 3.227069139480591,
+      "learning_rate": 0.0006625711035267349,
+      "loss": 6.5614,
+      "step": 2966
+    },
+    {
+      "epoch": 1.012627986348123,
+      "grad_norm": 3.215060234069824,
+      "learning_rate": 0.000662457337883959,
+      "loss": 6.2936,
+      "step": 2967
+    },
+    {
+      "epoch": 1.0129692832764505,
+      "grad_norm": 5.511234283447266,
+      "learning_rate": 0.0006623435722411832,
+      "loss": 6.3446,
+      "step": 2968
+    },
+    {
+      "epoch": 1.0133105802047782,
+      "grad_norm": 3.6093709468841553,
+      "learning_rate": 0.0006622298065984073,
+      "loss": 7.026,
+      "step": 2969
+    },
+    {
+      "epoch": 1.0136518771331058,
+      "grad_norm": 9.460335731506348,
+      "learning_rate": 0.0006621160409556314,
+      "loss": 8.476,
+      "step": 2970
+    },
+    {
+      "epoch": 1.0139931740614334,
+      "grad_norm": 3.619230031967163,
+      "learning_rate": 0.0006620022753128555,
+      "loss": 6.4041,
+      "step": 2971
+    },
+    {
+      "epoch": 1.014334470989761,
+      "grad_norm": 3.284637928009033,
+      "learning_rate": 0.0006618885096700796,
+      "loss": 6.6486,
+      "step": 2972
+    },
+    {
+      "epoch": 1.0146757679180887,
+      "grad_norm": 3.4304845333099365,
+      "learning_rate": 0.0006617747440273038,
+      "loss": 5.978,
+      "step": 2973
+    },
+    {
+      "epoch": 1.0150170648464163,
+      "grad_norm": 3.1258368492126465,
+      "learning_rate": 0.0006616609783845279,
+      "loss": 6.6721,
+      "step": 2974
+    },
+    {
+      "epoch": 1.015358361774744,
+      "grad_norm": 3.1004960536956787,
+      "learning_rate": 0.000661547212741752,
+      "loss": 6.6683,
+      "step": 2975
+    },
+    {
+      "epoch": 1.0156996587030718,
+      "grad_norm": 3.346923351287842,
+      "learning_rate": 0.0006614334470989761,
+      "loss": 6.4257,
+      "step": 2976
+    },
+    {
+      "epoch": 1.0160409556313994,
+      "grad_norm": 3.8876302242279053,
+      "learning_rate": 0.0006613196814562002,
+      "loss": 6.0568,
+      "step": 2977
+    },
+    {
+      "epoch": 1.016382252559727,
+      "grad_norm": 3.5111100673675537,
+      "learning_rate": 0.0006612059158134243,
+      "loss": 6.9239,
+      "step": 2978
+    },
+    {
+      "epoch": 1.0167235494880547,
+      "grad_norm": 3.2837634086608887,
+      "learning_rate": 0.0006610921501706485,
+      "loss": 6.5992,
+      "step": 2979
+    },
+    {
+      "epoch": 1.0170648464163823,
+      "grad_norm": 3.1712963581085205,
+      "learning_rate": 0.0006609783845278727,
+      "loss": 6.7537,
+      "step": 2980
+    },
+    {
+      "epoch": 1.01740614334471,
+      "grad_norm": 3.12092924118042,
+      "learning_rate": 0.0006608646188850968,
+      "loss": 6.6126,
+      "step": 2981
+    },
+    {
+      "epoch": 1.0177474402730375,
+      "grad_norm": 3.164799928665161,
+      "learning_rate": 0.0006607508532423209,
+      "loss": 6.2507,
+      "step": 2982
+    },
+    {
+      "epoch": 1.0180887372013652,
+      "grad_norm": 3.2638497352600098,
+      "learning_rate": 0.000660637087599545,
+      "loss": 6.1015,
+      "step": 2983
+    },
+    {
+      "epoch": 1.0184300341296928,
+      "grad_norm": 3.560190200805664,
+      "learning_rate": 0.0006605233219567691,
+      "loss": 6.388,
+      "step": 2984
+    },
+    {
+      "epoch": 1.0187713310580204,
+      "grad_norm": 3.2741174697875977,
+      "learning_rate": 0.0006604095563139933,
+      "loss": 6.9384,
+      "step": 2985
+    },
+    {
+      "epoch": 1.019112627986348,
+      "grad_norm": 6.3302788734436035,
+      "learning_rate": 0.0006602957906712173,
+      "loss": 5.3906,
+      "step": 2986
+    },
+    {
+      "epoch": 1.0194539249146757,
+      "grad_norm": 3.5244359970092773,
+      "learning_rate": 0.0006601820250284414,
+      "loss": 6.3372,
+      "step": 2987
+    },
+    {
+      "epoch": 1.0197952218430033,
+      "grad_norm": 3.4077272415161133,
+      "learning_rate": 0.0006600682593856655,
+      "loss": 6.0502,
+      "step": 2988
+    },
+    {
+      "epoch": 1.0201365187713312,
+      "grad_norm": 3.206965446472168,
+      "learning_rate": 0.0006599544937428896,
+      "loss": 6.3437,
+      "step": 2989
+    },
+    {
+      "epoch": 1.0204778156996588,
+      "grad_norm": 3.215886116027832,
+      "learning_rate": 0.0006598407281001137,
+      "loss": 6.6096,
+      "step": 2990
+    },
+    {
+      "epoch": 1.0208191126279864,
+      "grad_norm": 3.386878728866577,
+      "learning_rate": 0.0006597269624573379,
+      "loss": 6.5003,
+      "step": 2991
+    },
+    {
+      "epoch": 1.021160409556314,
+      "grad_norm": 3.419936180114746,
+      "learning_rate": 0.000659613196814562,
+      "loss": 6.5968,
+      "step": 2992
+    },
+    {
+      "epoch": 1.0215017064846417,
+      "grad_norm": 3.197819471359253,
+      "learning_rate": 0.0006594994311717861,
+      "loss": 6.9043,
+      "step": 2993
+    },
+    {
+      "epoch": 1.0218430034129693,
+      "grad_norm": 3.7721807956695557,
+      "learning_rate": 0.0006593856655290102,
+      "loss": 5.7042,
+      "step": 2994
+    },
+    {
+      "epoch": 1.022184300341297,
+      "grad_norm": 3.4664015769958496,
+      "learning_rate": 0.0006592718998862343,
+      "loss": 6.1596,
+      "step": 2995
+    },
+    {
+      "epoch": 1.0225255972696246,
+      "grad_norm": 3.259398937225342,
+      "learning_rate": 0.0006591581342434585,
+      "loss": 6.705,
+      "step": 2996
+    },
+    {
+      "epoch": 1.0228668941979522,
+      "grad_norm": 3.5420732498168945,
+      "learning_rate": 0.0006590443686006827,
+      "loss": 5.5468,
+      "step": 2997
+    },
+    {
+      "epoch": 1.0232081911262798,
+      "grad_norm": 3.332111120223999,
+      "learning_rate": 0.0006589306029579068,
+      "loss": 6.9041,
+      "step": 2998
+    },
+    {
+      "epoch": 1.0235494880546074,
+      "grad_norm": 3.29581618309021,
+      "learning_rate": 0.0006588168373151309,
+      "loss": 6.0205,
+      "step": 2999
+    },
+    {
+      "epoch": 1.023890784982935,
+      "grad_norm": 3.214153289794922,
+      "learning_rate": 0.000658703071672355,
+      "loss": 6.4507,
+      "step": 3000
+    },
+    {
+      "epoch": 1.0242320819112627,
+      "grad_norm": 3.5588912963867188,
+      "learning_rate": 0.0006585893060295791,
+      "loss": 6.4715,
+      "step": 3001
+    },
+    {
+      "epoch": 1.0245733788395905,
+      "grad_norm": 3.5324950218200684,
+      "learning_rate": 0.0006584755403868033,
+      "loss": 6.2474,
+      "step": 3002
+    },
+    {
+      "epoch": 1.0249146757679182,
+      "grad_norm": 3.397630453109741,
+      "learning_rate": 0.0006583617747440274,
+      "loss": 6.9605,
+      "step": 3003
+    },
+    {
+      "epoch": 1.0252559726962458,
+      "grad_norm": 3.475452423095703,
+      "learning_rate": 0.0006582480091012515,
+      "loss": 6.6207,
+      "step": 3004
+    },
+    {
+      "epoch": 1.0255972696245734,
+      "grad_norm": 3.4406540393829346,
+      "learning_rate": 0.0006581342434584755,
+      "loss": 6.6004,
+      "step": 3005
+    },
+    {
+      "epoch": 1.025938566552901,
+      "grad_norm": 5.332218170166016,
+      "learning_rate": 0.0006580204778156996,
+      "loss": 5.9674,
+      "step": 3006
+    },
+    {
+      "epoch": 1.0262798634812287,
+      "grad_norm": 3.4023447036743164,
+      "learning_rate": 0.0006579067121729237,
+      "loss": 6.3335,
+      "step": 3007
+    },
+    {
+      "epoch": 1.0266211604095563,
+      "grad_norm": 3.25628399848938,
+      "learning_rate": 0.0006577929465301479,
+      "loss": 6.2624,
+      "step": 3008
+    },
+    {
+      "epoch": 1.026962457337884,
+      "grad_norm": 3.3259971141815186,
+      "learning_rate": 0.000657679180887372,
+      "loss": 6.3255,
+      "step": 3009
+    },
+    {
+      "epoch": 1.0273037542662116,
+      "grad_norm": 4.986135482788086,
+      "learning_rate": 0.0006575654152445961,
+      "loss": 5.4787,
+      "step": 3010
+    },
+    {
+      "epoch": 1.0276450511945392,
+      "grad_norm": 3.862529754638672,
+      "learning_rate": 0.0006574516496018202,
+      "loss": 5.2939,
+      "step": 3011
+    },
+    {
+      "epoch": 1.0279863481228668,
+      "grad_norm": 3.2690622806549072,
+      "learning_rate": 0.0006573378839590443,
+      "loss": 6.5039,
+      "step": 3012
+    },
+    {
+      "epoch": 1.0283276450511944,
+      "grad_norm": 3.9144980907440186,
+      "learning_rate": 0.0006572241183162685,
+      "loss": 5.961,
+      "step": 3013
+    },
+    {
+      "epoch": 1.028668941979522,
+      "grad_norm": 3.58562970161438,
+      "learning_rate": 0.0006571103526734927,
+      "loss": 5.8142,
+      "step": 3014
+    },
+    {
+      "epoch": 1.02901023890785,
+      "grad_norm": 3.2888505458831787,
+      "learning_rate": 0.0006569965870307168,
+      "loss": 6.5148,
+      "step": 3015
+    },
+    {
+      "epoch": 1.0293515358361776,
+      "grad_norm": 3.2708396911621094,
+      "learning_rate": 0.0006568828213879409,
+      "loss": 6.4559,
+      "step": 3016
+    },
+    {
+      "epoch": 1.0296928327645052,
+      "grad_norm": 3.5158915519714355,
+      "learning_rate": 0.000656769055745165,
+      "loss": 6.3544,
+      "step": 3017
+    },
+    {
+      "epoch": 1.0300341296928328,
+      "grad_norm": 3.263108968734741,
+      "learning_rate": 0.0006566552901023891,
+      "loss": 6.443,
+      "step": 3018
+    },
+    {
+      "epoch": 1.0303754266211604,
+      "grad_norm": 3.9215052127838135,
+      "learning_rate": 0.0006565415244596133,
+      "loss": 5.9249,
+      "step": 3019
+    },
+    {
+      "epoch": 1.030716723549488,
+      "grad_norm": 3.202061653137207,
+      "learning_rate": 0.0006564277588168374,
+      "loss": 6.5584,
+      "step": 3020
+    },
+    {
+      "epoch": 1.0310580204778157,
+      "grad_norm": 14.988799095153809,
+      "learning_rate": 0.0006563139931740615,
+      "loss": 6.4804,
+      "step": 3021
+    },
+    {
+      "epoch": 1.0313993174061433,
+      "grad_norm": 11.352145195007324,
+      "learning_rate": 0.0006562002275312856,
+      "loss": 7.3238,
+      "step": 3022
+    },
+    {
+      "epoch": 1.031740614334471,
+      "grad_norm": 3.537874221801758,
+      "learning_rate": 0.0006560864618885097,
+      "loss": 7.0763,
+      "step": 3023
+    },
+    {
+      "epoch": 1.0320819112627986,
+      "grad_norm": 3.5525970458984375,
+      "learning_rate": 0.0006559726962457337,
+      "loss": 6.6374,
+      "step": 3024
+    },
+    {
+      "epoch": 1.0324232081911262,
+      "grad_norm": 3.3168790340423584,
+      "learning_rate": 0.0006558589306029579,
+      "loss": 6.5269,
+      "step": 3025
+    },
+    {
+      "epoch": 1.0327645051194538,
+      "grad_norm": 5.2062835693359375,
+      "learning_rate": 0.000655745164960182,
+      "loss": 6.9295,
+      "step": 3026
+    },
+    {
+      "epoch": 1.0331058020477815,
+      "grad_norm": 3.208059310913086,
+      "learning_rate": 0.0006556313993174061,
+      "loss": 7.0108,
+      "step": 3027
+    },
+    {
+      "epoch": 1.0334470989761093,
+      "grad_norm": 3.500671148300171,
+      "learning_rate": 0.0006555176336746302,
+      "loss": 6.1865,
+      "step": 3028
+    },
+    {
+      "epoch": 1.033788395904437,
+      "grad_norm": 5.409757137298584,
+      "learning_rate": 0.0006554038680318543,
+      "loss": 6.2085,
+      "step": 3029
+    },
+    {
+      "epoch": 1.0341296928327646,
+      "grad_norm": 3.258284091949463,
+      "learning_rate": 0.0006552901023890784,
+      "loss": 6.0585,
+      "step": 3030
+    },
+    {
+      "epoch": 1.0344709897610922,
+      "grad_norm": 3.4100899696350098,
+      "learning_rate": 0.0006551763367463027,
+      "loss": 6.069,
+      "step": 3031
+    },
+    {
+      "epoch": 1.0348122866894198,
+      "grad_norm": 3.146129846572876,
+      "learning_rate": 0.0006550625711035268,
+      "loss": 5.9704,
+      "step": 3032
+    },
+    {
+      "epoch": 1.0351535836177475,
+      "grad_norm": 3.3420863151550293,
+      "learning_rate": 0.0006549488054607509,
+      "loss": 7.3534,
+      "step": 3033
+    },
+    {
+      "epoch": 1.035494880546075,
+      "grad_norm": 3.4072580337524414,
+      "learning_rate": 0.000654835039817975,
+      "loss": 6.6606,
+      "step": 3034
+    },
+    {
+      "epoch": 1.0358361774744027,
+      "grad_norm": 3.3108816146850586,
+      "learning_rate": 0.0006547212741751991,
+      "loss": 6.3785,
+      "step": 3035
+    },
+    {
+      "epoch": 1.0361774744027303,
+      "grad_norm": 3.2887344360351562,
+      "learning_rate": 0.0006546075085324233,
+      "loss": 7.0955,
+      "step": 3036
+    },
+    {
+      "epoch": 1.036518771331058,
+      "grad_norm": 6.6958842277526855,
+      "learning_rate": 0.0006544937428896474,
+      "loss": 5.6834,
+      "step": 3037
+    },
+    {
+      "epoch": 1.0368600682593856,
+      "grad_norm": 3.3041765689849854,
+      "learning_rate": 0.0006543799772468715,
+      "loss": 6.5706,
+      "step": 3038
+    },
+    {
+      "epoch": 1.0372013651877132,
+      "grad_norm": 3.6999776363372803,
+      "learning_rate": 0.0006542662116040956,
+      "loss": 6.5919,
+      "step": 3039
+    },
+    {
+      "epoch": 1.0375426621160408,
+      "grad_norm": 3.507978916168213,
+      "learning_rate": 0.0006541524459613197,
+      "loss": 6.4322,
+      "step": 3040
+    },
+    {
+      "epoch": 1.0378839590443687,
+      "grad_norm": 3.151499032974243,
+      "learning_rate": 0.0006540386803185438,
+      "loss": 6.5041,
+      "step": 3041
+    },
+    {
+      "epoch": 1.0382252559726963,
+      "grad_norm": 3.339322805404663,
+      "learning_rate": 0.000653924914675768,
+      "loss": 6.2779,
+      "step": 3042
+    },
+    {
+      "epoch": 1.038566552901024,
+      "grad_norm": 4.641025543212891,
+      "learning_rate": 0.0006538111490329921,
+      "loss": 5.4317,
+      "step": 3043
+    },
+    {
+      "epoch": 1.0389078498293516,
+      "grad_norm": 3.3344995975494385,
+      "learning_rate": 0.0006536973833902161,
+      "loss": 6.5923,
+      "step": 3044
+    },
+    {
+      "epoch": 1.0392491467576792,
+      "grad_norm": 3.2587029933929443,
+      "learning_rate": 0.0006535836177474402,
+      "loss": 6.4993,
+      "step": 3045
+    },
+    {
+      "epoch": 1.0395904436860068,
+      "grad_norm": 4.462654113769531,
+      "learning_rate": 0.0006534698521046643,
+      "loss": 6.0649,
+      "step": 3046
+    },
+    {
+      "epoch": 1.0399317406143345,
+      "grad_norm": 3.293430805206299,
+      "learning_rate": 0.0006533560864618884,
+      "loss": 5.8758,
+      "step": 3047
+    },
+    {
+      "epoch": 1.040273037542662,
+      "grad_norm": 3.2306947708129883,
+      "learning_rate": 0.0006532423208191127,
+      "loss": 6.3128,
+      "step": 3048
+    },
+    {
+      "epoch": 1.0406143344709897,
+      "grad_norm": 3.3286349773406982,
+      "learning_rate": 0.0006531285551763368,
+      "loss": 6.7096,
+      "step": 3049
+    },
+    {
+      "epoch": 1.0409556313993173,
+      "grad_norm": 3.1706297397613525,
+      "learning_rate": 0.0006530147895335609,
+      "loss": 6.7695,
+      "step": 3050
+    },
+    {
+      "epoch": 1.041296928327645,
+      "grad_norm": 3.0948450565338135,
+      "learning_rate": 0.000652901023890785,
+      "loss": 6.4126,
+      "step": 3051
+    },
+    {
+      "epoch": 1.0416382252559726,
+      "grad_norm": 3.1173272132873535,
+      "learning_rate": 0.0006527872582480091,
+      "loss": 6.961,
+      "step": 3052
+    },
+    {
+      "epoch": 1.0419795221843002,
+      "grad_norm": 3.5899710655212402,
+      "learning_rate": 0.0006526734926052332,
+      "loss": 5.9567,
+      "step": 3053
+    },
+    {
+      "epoch": 1.042320819112628,
+      "grad_norm": 3.2167882919311523,
+      "learning_rate": 0.0006525597269624574,
+      "loss": 6.2195,
+      "step": 3054
+    },
+    {
+      "epoch": 1.0426621160409557,
+      "grad_norm": 3.2807223796844482,
+      "learning_rate": 0.0006524459613196815,
+      "loss": 6.4943,
+      "step": 3055
+    },
+    {
+      "epoch": 1.0430034129692833,
+      "grad_norm": 3.2400906085968018,
+      "learning_rate": 0.0006523321956769056,
+      "loss": 6.4222,
+      "step": 3056
+    },
+    {
+      "epoch": 1.043344709897611,
+      "grad_norm": 3.420193672180176,
+      "learning_rate": 0.0006522184300341297,
+      "loss": 6.3227,
+      "step": 3057
+    },
+    {
+      "epoch": 1.0436860068259386,
+      "grad_norm": 3.231858015060425,
+      "learning_rate": 0.0006521046643913538,
+      "loss": 7.0397,
+      "step": 3058
+    },
+    {
+      "epoch": 1.0440273037542662,
+      "grad_norm": 3.291337490081787,
+      "learning_rate": 0.000651990898748578,
+      "loss": 5.907,
+      "step": 3059
+    },
+    {
+      "epoch": 1.0443686006825939,
+      "grad_norm": 3.354321241378784,
+      "learning_rate": 0.0006518771331058021,
+      "loss": 6.532,
+      "step": 3060
+    },
+    {
+      "epoch": 1.0447098976109215,
+      "grad_norm": 3.303192377090454,
+      "learning_rate": 0.0006517633674630262,
+      "loss": 6.2837,
+      "step": 3061
+    },
+    {
+      "epoch": 1.045051194539249,
+      "grad_norm": 3.2552897930145264,
+      "learning_rate": 0.0006516496018202503,
+      "loss": 6.9191,
+      "step": 3062
+    },
+    {
+      "epoch": 1.0453924914675767,
+      "grad_norm": 3.7623484134674072,
+      "learning_rate": 0.0006515358361774743,
+      "loss": 6.0883,
+      "step": 3063
+    },
+    {
+      "epoch": 1.0457337883959044,
+      "grad_norm": 3.314925193786621,
+      "learning_rate": 0.0006514220705346984,
+      "loss": 6.4738,
+      "step": 3064
+    },
+    {
+      "epoch": 1.046075085324232,
+      "grad_norm": 3.2855207920074463,
+      "learning_rate": 0.0006513083048919227,
+      "loss": 5.971,
+      "step": 3065
+    },
+    {
+      "epoch": 1.0464163822525596,
+      "grad_norm": 3.406338930130005,
+      "learning_rate": 0.0006511945392491468,
+      "loss": 6.8355,
+      "step": 3066
+    },
+    {
+      "epoch": 1.0467576791808875,
+      "grad_norm": 3.4004054069519043,
+      "learning_rate": 0.0006510807736063709,
+      "loss": 6.2127,
+      "step": 3067
+    },
+    {
+      "epoch": 1.047098976109215,
+      "grad_norm": 3.392861843109131,
+      "learning_rate": 0.000650967007963595,
+      "loss": 6.3045,
+      "step": 3068
+    },
+    {
+      "epoch": 1.0474402730375427,
+      "grad_norm": 3.3509092330932617,
+      "learning_rate": 0.0006508532423208191,
+      "loss": 6.2746,
+      "step": 3069
+    },
+    {
+      "epoch": 1.0477815699658704,
+      "grad_norm": 3.2432587146759033,
+      "learning_rate": 0.0006507394766780432,
+      "loss": 6.21,
+      "step": 3070
+    },
+    {
+      "epoch": 1.048122866894198,
+      "grad_norm": 8.192042350769043,
+      "learning_rate": 0.0006506257110352674,
+      "loss": 6.7381,
+      "step": 3071
+    },
+    {
+      "epoch": 1.0484641638225256,
+      "grad_norm": 5.284655570983887,
+      "learning_rate": 0.0006505119453924915,
+      "loss": 6.5578,
+      "step": 3072
+    },
+    {
+      "epoch": 1.0488054607508532,
+      "grad_norm": 3.650186538696289,
+      "learning_rate": 0.0006503981797497156,
+      "loss": 6.2315,
+      "step": 3073
+    },
+    {
+      "epoch": 1.0491467576791809,
+      "grad_norm": 3.323975086212158,
+      "learning_rate": 0.0006502844141069397,
+      "loss": 6.7739,
+      "step": 3074
+    },
+    {
+      "epoch": 1.0494880546075085,
+      "grad_norm": 3.2510526180267334,
+      "learning_rate": 0.0006501706484641638,
+      "loss": 6.4504,
+      "step": 3075
+    },
+    {
+      "epoch": 1.0498293515358361,
+      "grad_norm": 7.123908996582031,
+      "learning_rate": 0.000650056882821388,
+      "loss": 7.3726,
+      "step": 3076
+    },
+    {
+      "epoch": 1.0501706484641637,
+      "grad_norm": 3.3244235515594482,
+      "learning_rate": 0.0006499431171786121,
+      "loss": 5.9337,
+      "step": 3077
+    },
+    {
+      "epoch": 1.0505119453924914,
+      "grad_norm": 4.019181728363037,
+      "learning_rate": 0.0006498293515358362,
+      "loss": 5.851,
+      "step": 3078
+    },
+    {
+      "epoch": 1.050853242320819,
+      "grad_norm": 3.2663567066192627,
+      "learning_rate": 0.0006497155858930603,
+      "loss": 6.808,
+      "step": 3079
+    },
+    {
+      "epoch": 1.0511945392491469,
+      "grad_norm": 3.2824909687042236,
+      "learning_rate": 0.0006496018202502844,
+      "loss": 6.8706,
+      "step": 3080
+    },
+    {
+      "epoch": 1.0515358361774745,
+      "grad_norm": 3.404869794845581,
+      "learning_rate": 0.0006494880546075086,
+      "loss": 6.5542,
+      "step": 3081
+    },
+    {
+      "epoch": 1.051877133105802,
+      "grad_norm": 3.04349684715271,
+      "learning_rate": 0.0006493742889647328,
+      "loss": 6.3294,
+      "step": 3082
+    },
+    {
+      "epoch": 1.0522184300341297,
+      "grad_norm": 3.5680627822875977,
+      "learning_rate": 0.0006492605233219568,
+      "loss": 6.0419,
+      "step": 3083
+    },
+    {
+      "epoch": 1.0525597269624574,
+      "grad_norm": 3.081714153289795,
+      "learning_rate": 0.0006491467576791809,
+      "loss": 6.4956,
+      "step": 3084
+    },
+    {
+      "epoch": 1.052901023890785,
+      "grad_norm": 3.4470624923706055,
+      "learning_rate": 0.000649032992036405,
+      "loss": 6.9409,
+      "step": 3085
+    },
+    {
+      "epoch": 1.0532423208191126,
+      "grad_norm": 3.1926770210266113,
+      "learning_rate": 0.0006489192263936291,
+      "loss": 6.461,
+      "step": 3086
+    },
+    {
+      "epoch": 1.0535836177474402,
+      "grad_norm": 10.409677505493164,
+      "learning_rate": 0.0006488054607508532,
+      "loss": 5.6452,
+      "step": 3087
+    },
+    {
+      "epoch": 1.0539249146757679,
+      "grad_norm": 3.4879887104034424,
+      "learning_rate": 0.0006486916951080774,
+      "loss": 6.7937,
+      "step": 3088
+    },
+    {
+      "epoch": 1.0542662116040955,
+      "grad_norm": 3.3193447589874268,
+      "learning_rate": 0.0006485779294653015,
+      "loss": 6.813,
+      "step": 3089
+    },
+    {
+      "epoch": 1.0546075085324231,
+      "grad_norm": 3.5224204063415527,
+      "learning_rate": 0.0006484641638225256,
+      "loss": 6.5331,
+      "step": 3090
+    },
+    {
+      "epoch": 1.0549488054607508,
+      "grad_norm": 3.5668528079986572,
+      "learning_rate": 0.0006483503981797497,
+      "loss": 5.8162,
+      "step": 3091
+    },
+    {
+      "epoch": 1.0552901023890784,
+      "grad_norm": 3.3257477283477783,
+      "learning_rate": 0.0006482366325369738,
+      "loss": 5.6231,
+      "step": 3092
+    },
+    {
+      "epoch": 1.0556313993174062,
+      "grad_norm": 3.734152317047119,
+      "learning_rate": 0.0006481228668941979,
+      "loss": 6.3863,
+      "step": 3093
+    },
+    {
+      "epoch": 1.0559726962457339,
+      "grad_norm": 3.2832605838775635,
+      "learning_rate": 0.0006480091012514221,
+      "loss": 6.9209,
+      "step": 3094
+    },
+    {
+      "epoch": 1.0563139931740615,
+      "grad_norm": 3.4488635063171387,
+      "learning_rate": 0.0006478953356086462,
+      "loss": 6.6284,
+      "step": 3095
+    },
+    {
+      "epoch": 1.0566552901023891,
+      "grad_norm": 3.6133999824523926,
+      "learning_rate": 0.0006477815699658703,
+      "loss": 6.6132,
+      "step": 3096
+    },
+    {
+      "epoch": 1.0569965870307167,
+      "grad_norm": 3.2564053535461426,
+      "learning_rate": 0.0006476678043230944,
+      "loss": 6.4362,
+      "step": 3097
+    },
+    {
+      "epoch": 1.0573378839590444,
+      "grad_norm": 3.2632007598876953,
+      "learning_rate": 0.0006475540386803186,
+      "loss": 6.6581,
+      "step": 3098
+    },
+    {
+      "epoch": 1.057679180887372,
+      "grad_norm": 3.2787528038024902,
+      "learning_rate": 0.0006474402730375428,
+      "loss": 6.1405,
+      "step": 3099
+    },
+    {
+      "epoch": 1.0580204778156996,
+      "grad_norm": 3.1381187438964844,
+      "learning_rate": 0.0006473265073947669,
+      "loss": 6.5018,
+      "step": 3100
+    },
+    {
+      "epoch": 1.0583617747440273,
+      "grad_norm": 6.568216800689697,
+      "learning_rate": 0.000647212741751991,
+      "loss": 5.6483,
+      "step": 3101
+    },
+    {
+      "epoch": 1.058703071672355,
+      "grad_norm": 3.281235694885254,
+      "learning_rate": 0.000647098976109215,
+      "loss": 6.5081,
+      "step": 3102
+    },
+    {
+      "epoch": 1.0590443686006825,
+      "grad_norm": 3.3020272254943848,
+      "learning_rate": 0.0006469852104664391,
+      "loss": 6.0661,
+      "step": 3103
+    },
+    {
+      "epoch": 1.0593856655290101,
+      "grad_norm": 3.2067642211914062,
+      "learning_rate": 0.0006468714448236632,
+      "loss": 6.4818,
+      "step": 3104
+    },
+    {
+      "epoch": 1.0597269624573378,
+      "grad_norm": 8.392196655273438,
+      "learning_rate": 0.0006467576791808874,
+      "loss": 6.1493,
+      "step": 3105
+    },
+    {
+      "epoch": 1.0600682593856656,
+      "grad_norm": 4.222563743591309,
+      "learning_rate": 0.0006466439135381115,
+      "loss": 6.2424,
+      "step": 3106
+    },
+    {
+      "epoch": 1.0604095563139933,
+      "grad_norm": 3.43169903755188,
+      "learning_rate": 0.0006465301478953356,
+      "loss": 6.2598,
+      "step": 3107
+    },
+    {
+      "epoch": 1.0607508532423209,
+      "grad_norm": 3.4820284843444824,
+      "learning_rate": 0.0006464163822525597,
+      "loss": 6.6193,
+      "step": 3108
+    },
+    {
+      "epoch": 1.0610921501706485,
+      "grad_norm": 3.227473497390747,
+      "learning_rate": 0.0006463026166097838,
+      "loss": 6.5877,
+      "step": 3109
+    },
+    {
+      "epoch": 1.0614334470989761,
+      "grad_norm": 3.3051085472106934,
+      "learning_rate": 0.0006461888509670079,
+      "loss": 6.1434,
+      "step": 3110
+    },
+    {
+      "epoch": 1.0617747440273038,
+      "grad_norm": 4.006912708282471,
+      "learning_rate": 0.0006460750853242321,
+      "loss": 6.5416,
+      "step": 3111
+    },
+    {
+      "epoch": 1.0621160409556314,
+      "grad_norm": 3.184049129486084,
+      "learning_rate": 0.0006459613196814562,
+      "loss": 6.7029,
+      "step": 3112
+    },
+    {
+      "epoch": 1.062457337883959,
+      "grad_norm": 3.2582833766937256,
+      "learning_rate": 0.0006458475540386803,
+      "loss": 6.4838,
+      "step": 3113
+    },
+    {
+      "epoch": 1.0627986348122866,
+      "grad_norm": 3.350537061691284,
+      "learning_rate": 0.0006457337883959044,
+      "loss": 6.6411,
+      "step": 3114
+    },
+    {
+      "epoch": 1.0631399317406143,
+      "grad_norm": 4.094681739807129,
+      "learning_rate": 0.0006456200227531286,
+      "loss": 5.957,
+      "step": 3115
+    },
+    {
+      "epoch": 1.063481228668942,
+      "grad_norm": 3.8731682300567627,
+      "learning_rate": 0.0006455062571103528,
+      "loss": 6.3783,
+      "step": 3116
+    },
+    {
+      "epoch": 1.0638225255972695,
+      "grad_norm": 3.384065866470337,
+      "learning_rate": 0.0006453924914675769,
+      "loss": 6.6837,
+      "step": 3117
+    },
+    {
+      "epoch": 1.0641638225255972,
+      "grad_norm": 3.379866361618042,
+      "learning_rate": 0.000645278725824801,
+      "loss": 6.4046,
+      "step": 3118
+    },
+    {
+      "epoch": 1.064505119453925,
+      "grad_norm": 3.2628586292266846,
+      "learning_rate": 0.0006451649601820251,
+      "loss": 6.209,
+      "step": 3119
+    },
+    {
+      "epoch": 1.0648464163822526,
+      "grad_norm": 3.4711813926696777,
+      "learning_rate": 0.0006450511945392492,
+      "loss": 6.1973,
+      "step": 3120
+    },
+    {
+      "epoch": 1.0651877133105803,
+      "grad_norm": 3.7682406902313232,
+      "learning_rate": 0.0006449374288964733,
+      "loss": 6.1473,
+      "step": 3121
+    },
+    {
+      "epoch": 1.065529010238908,
+      "grad_norm": 8.02403450012207,
+      "learning_rate": 0.0006448236632536974,
+      "loss": 5.49,
+      "step": 3122
+    },
+    {
+      "epoch": 1.0658703071672355,
+      "grad_norm": 3.402454376220703,
+      "learning_rate": 0.0006447098976109215,
+      "loss": 6.6287,
+      "step": 3123
+    },
+    {
+      "epoch": 1.0662116040955631,
+      "grad_norm": 4.036313056945801,
+      "learning_rate": 0.0006445961319681456,
+      "loss": 5.7235,
+      "step": 3124
+    },
+    {
+      "epoch": 1.0665529010238908,
+      "grad_norm": 7.888774394989014,
+      "learning_rate": 0.0006444823663253697,
+      "loss": 4.691,
+      "step": 3125
+    },
+    {
+      "epoch": 1.0668941979522184,
+      "grad_norm": 3.3642637729644775,
+      "learning_rate": 0.0006443686006825938,
+      "loss": 6.7421,
+      "step": 3126
+    },
+    {
+      "epoch": 1.067235494880546,
+      "grad_norm": 3.545217990875244,
+      "learning_rate": 0.0006442548350398179,
+      "loss": 5.7046,
+      "step": 3127
+    },
+    {
+      "epoch": 1.0675767918088737,
+      "grad_norm": 3.4235167503356934,
+      "learning_rate": 0.0006441410693970421,
+      "loss": 6.7823,
+      "step": 3128
+    },
+    {
+      "epoch": 1.0679180887372013,
+      "grad_norm": 3.252849578857422,
+      "learning_rate": 0.0006440273037542662,
+      "loss": 7.001,
+      "step": 3129
+    },
+    {
+      "epoch": 1.068259385665529,
+      "grad_norm": 5.600972652435303,
+      "learning_rate": 0.0006439135381114903,
+      "loss": 6.5534,
+      "step": 3130
+    },
+    {
+      "epoch": 1.0686006825938565,
+      "grad_norm": 3.1004841327667236,
+      "learning_rate": 0.0006437997724687144,
+      "loss": 6.7081,
+      "step": 3131
+    },
+    {
+      "epoch": 1.0689419795221844,
+      "grad_norm": 8.813851356506348,
+      "learning_rate": 0.0006436860068259386,
+      "loss": 4.8907,
+      "step": 3132
+    },
+    {
+      "epoch": 1.069283276450512,
+      "grad_norm": 3.3041679859161377,
+      "learning_rate": 0.0006435722411831627,
+      "loss": 6.594,
+      "step": 3133
+    },
+    {
+      "epoch": 1.0696245733788396,
+      "grad_norm": 3.500278949737549,
+      "learning_rate": 0.0006434584755403869,
+      "loss": 6.3211,
+      "step": 3134
+    },
+    {
+      "epoch": 1.0699658703071673,
+      "grad_norm": 3.4029455184936523,
+      "learning_rate": 0.000643344709897611,
+      "loss": 6.3552,
+      "step": 3135
+    },
+    {
+      "epoch": 1.070307167235495,
+      "grad_norm": 3.182002305984497,
+      "learning_rate": 0.0006432309442548351,
+      "loss": 6.8919,
+      "step": 3136
+    },
+    {
+      "epoch": 1.0706484641638225,
+      "grad_norm": 3.758815050125122,
+      "learning_rate": 0.0006431171786120592,
+      "loss": 5.6043,
+      "step": 3137
+    },
+    {
+      "epoch": 1.0709897610921502,
+      "grad_norm": 3.294699192047119,
+      "learning_rate": 0.0006430034129692833,
+      "loss": 6.3708,
+      "step": 3138
+    },
+    {
+      "epoch": 1.0713310580204778,
+      "grad_norm": 3.3556272983551025,
+      "learning_rate": 0.0006428896473265075,
+      "loss": 6.6939,
+      "step": 3139
+    },
+    {
+      "epoch": 1.0716723549488054,
+      "grad_norm": 3.1853625774383545,
+      "learning_rate": 0.0006427758816837316,
+      "loss": 6.1872,
+      "step": 3140
+    },
+    {
+      "epoch": 1.072013651877133,
+      "grad_norm": 3.300252676010132,
+      "learning_rate": 0.0006426621160409556,
+      "loss": 6.771,
+      "step": 3141
+    },
+    {
+      "epoch": 1.0723549488054607,
+      "grad_norm": 3.111130714416504,
+      "learning_rate": 0.0006425483503981797,
+      "loss": 6.3194,
+      "step": 3142
+    },
+    {
+      "epoch": 1.0726962457337883,
+      "grad_norm": 3.808727741241455,
+      "learning_rate": 0.0006424345847554038,
+      "loss": 5.9753,
+      "step": 3143
+    },
+    {
+      "epoch": 1.073037542662116,
+      "grad_norm": 3.226867437362671,
+      "learning_rate": 0.0006423208191126279,
+      "loss": 6.5513,
+      "step": 3144
+    },
+    {
+      "epoch": 1.0733788395904438,
+      "grad_norm": 3.2645046710968018,
+      "learning_rate": 0.0006422070534698521,
+      "loss": 6.7096,
+      "step": 3145
+    },
+    {
+      "epoch": 1.0737201365187714,
+      "grad_norm": 3.1545019149780273,
+      "learning_rate": 0.0006420932878270762,
+      "loss": 6.5052,
+      "step": 3146
+    },
+    {
+      "epoch": 1.074061433447099,
+      "grad_norm": 3.2468883991241455,
+      "learning_rate": 0.0006419795221843003,
+      "loss": 6.8057,
+      "step": 3147
+    },
+    {
+      "epoch": 1.0744027303754267,
+      "grad_norm": 5.185894966125488,
+      "learning_rate": 0.0006418657565415244,
+      "loss": 4.2064,
+      "step": 3148
+    },
+    {
+      "epoch": 1.0747440273037543,
+      "grad_norm": 3.218177318572998,
+      "learning_rate": 0.0006417519908987486,
+      "loss": 6.056,
+      "step": 3149
+    },
+    {
+      "epoch": 1.075085324232082,
+      "grad_norm": 3.383833646774292,
+      "learning_rate": 0.0006416382252559727,
+      "loss": 6.7475,
+      "step": 3150
+    },
+    {
+      "epoch": 1.0754266211604095,
+      "grad_norm": 3.697319507598877,
+      "learning_rate": 0.0006415244596131969,
+      "loss": 6.663,
+      "step": 3151
+    },
+    {
+      "epoch": 1.0757679180887372,
+      "grad_norm": 3.122490644454956,
+      "learning_rate": 0.000641410693970421,
+      "loss": 6.6991,
+      "step": 3152
+    },
+    {
+      "epoch": 1.0761092150170648,
+      "grad_norm": 3.148451089859009,
+      "learning_rate": 0.0006412969283276451,
+      "loss": 6.6486,
+      "step": 3153
+    },
+    {
+      "epoch": 1.0764505119453924,
+      "grad_norm": 3.321945905685425,
+      "learning_rate": 0.0006411831626848692,
+      "loss": 6.1962,
+      "step": 3154
+    },
+    {
+      "epoch": 1.07679180887372,
+      "grad_norm": 3.590642213821411,
+      "learning_rate": 0.0006410693970420933,
+      "loss": 5.5254,
+      "step": 3155
+    },
+    {
+      "epoch": 1.0771331058020477,
+      "grad_norm": 3.300290584564209,
+      "learning_rate": 0.0006409556313993174,
+      "loss": 5.7448,
+      "step": 3156
+    },
+    {
+      "epoch": 1.0774744027303753,
+      "grad_norm": 3.4239137172698975,
+      "learning_rate": 0.0006408418657565416,
+      "loss": 6.2855,
+      "step": 3157
+    },
+    {
+      "epoch": 1.0778156996587032,
+      "grad_norm": 3.3473432064056396,
+      "learning_rate": 0.0006407281001137657,
+      "loss": 6.6531,
+      "step": 3158
+    },
+    {
+      "epoch": 1.0781569965870308,
+      "grad_norm": 3.0467193126678467,
+      "learning_rate": 0.0006406143344709898,
+      "loss": 6.3434,
+      "step": 3159
+    },
+    {
+      "epoch": 1.0784982935153584,
+      "grad_norm": 3.6349399089813232,
+      "learning_rate": 0.0006405005688282139,
+      "loss": 6.3252,
+      "step": 3160
+    },
+    {
+      "epoch": 1.078839590443686,
+      "grad_norm": 3.4417903423309326,
+      "learning_rate": 0.0006403868031854379,
+      "loss": 6.9098,
+      "step": 3161
+    },
+    {
+      "epoch": 1.0791808873720137,
+      "grad_norm": 3.258446455001831,
+      "learning_rate": 0.0006402730375426621,
+      "loss": 6.3267,
+      "step": 3162
+    },
+    {
+      "epoch": 1.0795221843003413,
+      "grad_norm": 3.3519585132598877,
+      "learning_rate": 0.0006401592718998862,
+      "loss": 5.89,
+      "step": 3163
+    },
+    {
+      "epoch": 1.079863481228669,
+      "grad_norm": 3.009504795074463,
+      "learning_rate": 0.0006400455062571103,
+      "loss": 6.2854,
+      "step": 3164
+    },
+    {
+      "epoch": 1.0802047781569966,
+      "grad_norm": 3.346721649169922,
+      "learning_rate": 0.0006399317406143345,
+      "loss": 6.9105,
+      "step": 3165
+    },
+    {
+      "epoch": 1.0805460750853242,
+      "grad_norm": 3.1614081859588623,
+      "learning_rate": 0.0006398179749715586,
+      "loss": 6.4212,
+      "step": 3166
+    },
+    {
+      "epoch": 1.0808873720136518,
+      "grad_norm": 3.280531406402588,
+      "learning_rate": 0.0006397042093287827,
+      "loss": 6.22,
+      "step": 3167
+    },
+    {
+      "epoch": 1.0812286689419794,
+      "grad_norm": 3.3301196098327637,
+      "learning_rate": 0.0006395904436860069,
+      "loss": 6.7717,
+      "step": 3168
+    },
+    {
+      "epoch": 1.081569965870307,
+      "grad_norm": 3.3172807693481445,
+      "learning_rate": 0.000639476678043231,
+      "loss": 6.6153,
+      "step": 3169
+    },
+    {
+      "epoch": 1.0819112627986347,
+      "grad_norm": 3.3986637592315674,
+      "learning_rate": 0.0006393629124004551,
+      "loss": 6.0319,
+      "step": 3170
+    },
+    {
+      "epoch": 1.0822525597269625,
+      "grad_norm": 3.199392318725586,
+      "learning_rate": 0.0006392491467576792,
+      "loss": 6.2352,
+      "step": 3171
+    },
+    {
+      "epoch": 1.0825938566552902,
+      "grad_norm": 5.347891330718994,
+      "learning_rate": 0.0006391353811149033,
+      "loss": 5.6896,
+      "step": 3172
+    },
+    {
+      "epoch": 1.0829351535836178,
+      "grad_norm": 3.7180368900299072,
+      "learning_rate": 0.0006390216154721274,
+      "loss": 6.2303,
+      "step": 3173
+    },
+    {
+      "epoch": 1.0832764505119454,
+      "grad_norm": 3.3773393630981445,
+      "learning_rate": 0.0006389078498293516,
+      "loss": 6.0189,
+      "step": 3174
+    },
+    {
+      "epoch": 1.083617747440273,
+      "grad_norm": 6.927037715911865,
+      "learning_rate": 0.0006387940841865757,
+      "loss": 4.5765,
+      "step": 3175
+    },
+    {
+      "epoch": 1.0839590443686007,
+      "grad_norm": 4.274540901184082,
+      "learning_rate": 0.0006386803185437998,
+      "loss": 5.871,
+      "step": 3176
+    },
+    {
+      "epoch": 1.0843003412969283,
+      "grad_norm": 3.52830171585083,
+      "learning_rate": 0.0006385665529010239,
+      "loss": 6.474,
+      "step": 3177
+    },
+    {
+      "epoch": 1.084641638225256,
+      "grad_norm": 3.9919192790985107,
+      "learning_rate": 0.000638452787258248,
+      "loss": 6.2379,
+      "step": 3178
+    },
+    {
+      "epoch": 1.0849829351535836,
+      "grad_norm": 3.2457633018493652,
+      "learning_rate": 0.0006383390216154722,
+      "loss": 7.1027,
+      "step": 3179
+    },
+    {
+      "epoch": 1.0853242320819112,
+      "grad_norm": 3.2085020542144775,
+      "learning_rate": 0.0006382252559726962,
+      "loss": 6.7974,
+      "step": 3180
+    },
+    {
+      "epoch": 1.0856655290102388,
+      "grad_norm": 3.2093658447265625,
+      "learning_rate": 0.0006381114903299203,
+      "loss": 6.5593,
+      "step": 3181
+    },
+    {
+      "epoch": 1.0860068259385665,
+      "grad_norm": 4.833531856536865,
+      "learning_rate": 0.0006379977246871445,
+      "loss": 6.0521,
+      "step": 3182
+    },
+    {
+      "epoch": 1.086348122866894,
+      "grad_norm": 3.516695499420166,
+      "learning_rate": 0.0006378839590443686,
+      "loss": 6.3321,
+      "step": 3183
+    },
+    {
+      "epoch": 1.086689419795222,
+      "grad_norm": 3.375440835952759,
+      "learning_rate": 0.0006377701934015927,
+      "loss": 6.4715,
+      "step": 3184
+    },
+    {
+      "epoch": 1.0870307167235496,
+      "grad_norm": 3.308617115020752,
+      "learning_rate": 0.0006376564277588169,
+      "loss": 6.5871,
+      "step": 3185
+    },
+    {
+      "epoch": 1.0873720136518772,
+      "grad_norm": 3.2866013050079346,
+      "learning_rate": 0.000637542662116041,
+      "loss": 6.5732,
+      "step": 3186
+    },
+    {
+      "epoch": 1.0877133105802048,
+      "grad_norm": 3.3857619762420654,
+      "learning_rate": 0.0006374288964732651,
+      "loss": 6.3731,
+      "step": 3187
+    },
+    {
+      "epoch": 1.0880546075085324,
+      "grad_norm": 3.51200795173645,
+      "learning_rate": 0.0006373151308304892,
+      "loss": 5.6781,
+      "step": 3188
+    },
+    {
+      "epoch": 1.08839590443686,
+      "grad_norm": 3.297863006591797,
+      "learning_rate": 0.0006372013651877133,
+      "loss": 6.4584,
+      "step": 3189
+    },
+    {
+      "epoch": 1.0887372013651877,
+      "grad_norm": 3.261662244796753,
+      "learning_rate": 0.0006370875995449374,
+      "loss": 6.2231,
+      "step": 3190
+    },
+    {
+      "epoch": 1.0890784982935153,
+      "grad_norm": 3.8961715698242188,
+      "learning_rate": 0.0006369738339021616,
+      "loss": 6.607,
+      "step": 3191
+    },
+    {
+      "epoch": 1.089419795221843,
+      "grad_norm": 3.399348020553589,
+      "learning_rate": 0.0006368600682593857,
+      "loss": 4.4051,
+      "step": 3192
+    },
+    {
+      "epoch": 1.0897610921501706,
+      "grad_norm": 4.216414928436279,
+      "learning_rate": 0.0006367463026166098,
+      "loss": 4.1361,
+      "step": 3193
+    },
+    {
+      "epoch": 1.0901023890784982,
+      "grad_norm": 3.367846727371216,
+      "learning_rate": 0.0006366325369738339,
+      "loss": 6.8591,
+      "step": 3194
+    },
+    {
+      "epoch": 1.0904436860068258,
+      "grad_norm": 3.9558677673339844,
+      "learning_rate": 0.000636518771331058,
+      "loss": 6.0432,
+      "step": 3195
+    },
+    {
+      "epoch": 1.0907849829351535,
+      "grad_norm": 3.5588676929473877,
+      "learning_rate": 0.0006364050056882821,
+      "loss": 6.5362,
+      "step": 3196
+    },
+    {
+      "epoch": 1.0911262798634813,
+      "grad_norm": 4.590184688568115,
+      "learning_rate": 0.0006362912400455064,
+      "loss": 6.3793,
+      "step": 3197
+    },
+    {
+      "epoch": 1.091467576791809,
+      "grad_norm": 3.1657865047454834,
+      "learning_rate": 0.0006361774744027305,
+      "loss": 6.3938,
+      "step": 3198
+    },
+    {
+      "epoch": 1.0918088737201366,
+      "grad_norm": 3.1333377361297607,
+      "learning_rate": 0.0006360637087599546,
+      "loss": 6.9106,
+      "step": 3199
+    },
+    {
+      "epoch": 1.0921501706484642,
+      "grad_norm": 3.446474313735962,
+      "learning_rate": 0.0006359499431171786,
+      "loss": 6.3406,
+      "step": 3200
+    },
+    {
+      "epoch": 1.0924914675767918,
+      "grad_norm": 4.603214263916016,
+      "learning_rate": 0.0006358361774744027,
+      "loss": 5.921,
+      "step": 3201
+    },
+    {
+      "epoch": 1.0928327645051195,
+      "grad_norm": 3.366414785385132,
+      "learning_rate": 0.0006357224118316269,
+      "loss": 6.5776,
+      "step": 3202
+    },
+    {
+      "epoch": 1.093174061433447,
+      "grad_norm": 3.351638078689575,
+      "learning_rate": 0.000635608646188851,
+      "loss": 6.0564,
+      "step": 3203
+    },
+    {
+      "epoch": 1.0935153583617747,
+      "grad_norm": 3.204946517944336,
+      "learning_rate": 0.0006354948805460751,
+      "loss": 6.5415,
+      "step": 3204
+    },
+    {
+      "epoch": 1.0938566552901023,
+      "grad_norm": 3.1665844917297363,
+      "learning_rate": 0.0006353811149032992,
+      "loss": 6.6033,
+      "step": 3205
+    },
+    {
+      "epoch": 1.09419795221843,
+      "grad_norm": 3.453909158706665,
+      "learning_rate": 0.0006352673492605233,
+      "loss": 6.2195,
+      "step": 3206
+    },
+    {
+      "epoch": 1.0945392491467576,
+      "grad_norm": 3.191018581390381,
+      "learning_rate": 0.0006351535836177474,
+      "loss": 6.3949,
+      "step": 3207
+    },
+    {
+      "epoch": 1.0948805460750852,
+      "grad_norm": 3.1959285736083984,
+      "learning_rate": 0.0006350398179749716,
+      "loss": 6.2866,
+      "step": 3208
+    },
+    {
+      "epoch": 1.0952218430034129,
+      "grad_norm": 3.368140459060669,
+      "learning_rate": 0.0006349260523321957,
+      "loss": 6.4439,
+      "step": 3209
+    },
+    {
+      "epoch": 1.0955631399317407,
+      "grad_norm": 3.4131994247436523,
+      "learning_rate": 0.0006348122866894198,
+      "loss": 6.5307,
+      "step": 3210
+    },
+    {
+      "epoch": 1.0959044368600683,
+      "grad_norm": 3.310483455657959,
+      "learning_rate": 0.0006346985210466439,
+      "loss": 6.8359,
+      "step": 3211
+    },
+    {
+      "epoch": 1.096245733788396,
+      "grad_norm": 3.1908204555511475,
+      "learning_rate": 0.000634584755403868,
+      "loss": 6.6964,
+      "step": 3212
+    },
+    {
+      "epoch": 1.0965870307167236,
+      "grad_norm": 3.5243823528289795,
+      "learning_rate": 0.0006344709897610921,
+      "loss": 6.4033,
+      "step": 3213
+    },
+    {
+      "epoch": 1.0969283276450512,
+      "grad_norm": 3.324995756149292,
+      "learning_rate": 0.0006343572241183164,
+      "loss": 6.8275,
+      "step": 3214
+    },
+    {
+      "epoch": 1.0972696245733788,
+      "grad_norm": 3.591324806213379,
+      "learning_rate": 0.0006342434584755405,
+      "loss": 6.5428,
+      "step": 3215
+    },
+    {
+      "epoch": 1.0976109215017065,
+      "grad_norm": 3.3508944511413574,
+      "learning_rate": 0.0006341296928327646,
+      "loss": 6.4207,
+      "step": 3216
+    },
+    {
+      "epoch": 1.097952218430034,
+      "grad_norm": 3.258124589920044,
+      "learning_rate": 0.0006340159271899887,
+      "loss": 6.369,
+      "step": 3217
+    },
+    {
+      "epoch": 1.0982935153583617,
+      "grad_norm": 3.1361072063446045,
+      "learning_rate": 0.0006339021615472128,
+      "loss": 6.3561,
+      "step": 3218
+    },
+    {
+      "epoch": 1.0986348122866894,
+      "grad_norm": 3.4941697120666504,
+      "learning_rate": 0.0006337883959044368,
+      "loss": 5.6931,
+      "step": 3219
+    },
+    {
+      "epoch": 1.098976109215017,
+      "grad_norm": 3.129948854446411,
+      "learning_rate": 0.000633674630261661,
+      "loss": 6.4307,
+      "step": 3220
+    },
+    {
+      "epoch": 1.0993174061433446,
+      "grad_norm": 6.091549396514893,
+      "learning_rate": 0.0006335608646188851,
+      "loss": 5.7017,
+      "step": 3221
+    },
+    {
+      "epoch": 1.0996587030716722,
+      "grad_norm": 3.1629409790039062,
+      "learning_rate": 0.0006334470989761092,
+      "loss": 6.0674,
+      "step": 3222
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 6.926721572875977,
+      "learning_rate": 0.0006333333333333333,
+      "loss": 5.3602,
+      "step": 3223
+    },
+    {
+      "epoch": 1.1003412969283277,
+      "grad_norm": 4.425708770751953,
+      "learning_rate": 0.0006332195676905574,
+      "loss": 6.9397,
+      "step": 3224
+    },
+    {
+      "epoch": 1.1006825938566553,
+      "grad_norm": 3.5748658180236816,
+      "learning_rate": 0.0006331058020477816,
+      "loss": 7.2775,
+      "step": 3225
+    },
+    {
+      "epoch": 1.101023890784983,
+      "grad_norm": 3.3313708305358887,
+      "learning_rate": 0.0006329920364050057,
+      "loss": 6.7344,
+      "step": 3226
+    },
+    {
+      "epoch": 1.1013651877133106,
+      "grad_norm": 4.757328987121582,
+      "learning_rate": 0.0006328782707622298,
+      "loss": 5.769,
+      "step": 3227
+    },
+    {
+      "epoch": 1.1017064846416382,
+      "grad_norm": 3.1376912593841553,
+      "learning_rate": 0.0006327645051194539,
+      "loss": 6.7527,
+      "step": 3228
+    },
+    {
+      "epoch": 1.1020477815699659,
+      "grad_norm": 3.682274341583252,
+      "learning_rate": 0.000632650739476678,
+      "loss": 6.7461,
+      "step": 3229
+    },
+    {
+      "epoch": 1.1023890784982935,
+      "grad_norm": 3.245393991470337,
+      "learning_rate": 0.0006325369738339021,
+      "loss": 6.5636,
+      "step": 3230
+    },
+    {
+      "epoch": 1.1027303754266211,
+      "grad_norm": 3.133185625076294,
+      "learning_rate": 0.0006324232081911264,
+      "loss": 6.2898,
+      "step": 3231
+    },
+    {
+      "epoch": 1.1030716723549487,
+      "grad_norm": 3.1336660385131836,
+      "learning_rate": 0.0006323094425483505,
+      "loss": 6.4333,
+      "step": 3232
+    },
+    {
+      "epoch": 1.1034129692832764,
+      "grad_norm": 3.263469934463501,
+      "learning_rate": 0.0006321956769055746,
+      "loss": 6.4195,
+      "step": 3233
+    },
+    {
+      "epoch": 1.103754266211604,
+      "grad_norm": 3.146401882171631,
+      "learning_rate": 0.0006320819112627987,
+      "loss": 6.7663,
+      "step": 3234
+    },
+    {
+      "epoch": 1.1040955631399316,
+      "grad_norm": 3.2967708110809326,
+      "learning_rate": 0.0006319681456200228,
+      "loss": 6.8334,
+      "step": 3235
+    },
+    {
+      "epoch": 1.1044368600682595,
+      "grad_norm": 3.167576551437378,
+      "learning_rate": 0.0006318543799772469,
+      "loss": 6.619,
+      "step": 3236
+    },
+    {
+      "epoch": 1.104778156996587,
+      "grad_norm": 6.352286338806152,
+      "learning_rate": 0.0006317406143344711,
+      "loss": 6.0053,
+      "step": 3237
+    },
+    {
+      "epoch": 1.1051194539249147,
+      "grad_norm": 3.420991897583008,
+      "learning_rate": 0.0006316268486916951,
+      "loss": 6.2279,
+      "step": 3238
+    },
+    {
+      "epoch": 1.1054607508532424,
+      "grad_norm": 3.3597917556762695,
+      "learning_rate": 0.0006315130830489192,
+      "loss": 6.4583,
+      "step": 3239
+    },
+    {
+      "epoch": 1.10580204778157,
+      "grad_norm": 3.377300262451172,
+      "learning_rate": 0.0006313993174061433,
+      "loss": 6.5939,
+      "step": 3240
+    },
+    {
+      "epoch": 1.1061433447098976,
+      "grad_norm": 3.4640071392059326,
+      "learning_rate": 0.0006312855517633674,
+      "loss": 6.1728,
+      "step": 3241
+    },
+    {
+      "epoch": 1.1064846416382252,
+      "grad_norm": 3.410388708114624,
+      "learning_rate": 0.0006311717861205916,
+      "loss": 6.1942,
+      "step": 3242
+    },
+    {
+      "epoch": 1.1068259385665529,
+      "grad_norm": 3.2869067192077637,
+      "learning_rate": 0.0006310580204778157,
+      "loss": 6.6288,
+      "step": 3243
+    },
+    {
+      "epoch": 1.1071672354948805,
+      "grad_norm": 3.073132038116455,
+      "learning_rate": 0.0006309442548350398,
+      "loss": 6.4948,
+      "step": 3244
+    },
+    {
+      "epoch": 1.1075085324232081,
+      "grad_norm": 3.332573175430298,
+      "learning_rate": 0.0006308304891922639,
+      "loss": 6.2622,
+      "step": 3245
+    },
+    {
+      "epoch": 1.1078498293515358,
+      "grad_norm": 3.1241261959075928,
+      "learning_rate": 0.000630716723549488,
+      "loss": 6.4462,
+      "step": 3246
+    },
+    {
+      "epoch": 1.1081911262798634,
+      "grad_norm": 3.215703248977661,
+      "learning_rate": 0.0006306029579067121,
+      "loss": 6.9699,
+      "step": 3247
+    },
+    {
+      "epoch": 1.108532423208191,
+      "grad_norm": 3.1730904579162598,
+      "learning_rate": 0.0006304891922639364,
+      "loss": 6.5032,
+      "step": 3248
+    },
+    {
+      "epoch": 1.1088737201365189,
+      "grad_norm": 3.286179542541504,
+      "learning_rate": 0.0006303754266211605,
+      "loss": 6.4786,
+      "step": 3249
+    },
+    {
+      "epoch": 1.1092150170648465,
+      "grad_norm": 3.2079620361328125,
+      "learning_rate": 0.0006302616609783846,
+      "loss": 5.9624,
+      "step": 3250
+    },
+    {
+      "epoch": 1.1095563139931741,
+      "grad_norm": 3.4078316688537598,
+      "learning_rate": 0.0006301478953356087,
+      "loss": 5.7281,
+      "step": 3251
+    },
+    {
+      "epoch": 1.1098976109215017,
+      "grad_norm": 3.154865264892578,
+      "learning_rate": 0.0006300341296928328,
+      "loss": 6.2634,
+      "step": 3252
+    },
+    {
+      "epoch": 1.1102389078498294,
+      "grad_norm": 4.15884256362915,
+      "learning_rate": 0.0006299203640500569,
+      "loss": 5.9743,
+      "step": 3253
+    },
+    {
+      "epoch": 1.110580204778157,
+      "grad_norm": 3.244605302810669,
+      "learning_rate": 0.0006298065984072811,
+      "loss": 6.3037,
+      "step": 3254
+    },
+    {
+      "epoch": 1.1109215017064846,
+      "grad_norm": 3.354834794998169,
+      "learning_rate": 0.0006296928327645052,
+      "loss": 6.2177,
+      "step": 3255
+    },
+    {
+      "epoch": 1.1112627986348123,
+      "grad_norm": 3.270688056945801,
+      "learning_rate": 0.0006295790671217293,
+      "loss": 6.2083,
+      "step": 3256
+    },
+    {
+      "epoch": 1.1116040955631399,
+      "grad_norm": 3.2896766662597656,
+      "learning_rate": 0.0006294653014789534,
+      "loss": 6.3809,
+      "step": 3257
+    },
+    {
+      "epoch": 1.1119453924914675,
+      "grad_norm": 3.2658944129943848,
+      "learning_rate": 0.0006293515358361774,
+      "loss": 5.7977,
+      "step": 3258
+    },
+    {
+      "epoch": 1.1122866894197951,
+      "grad_norm": 3.716240644454956,
+      "learning_rate": 0.0006292377701934015,
+      "loss": 6.0917,
+      "step": 3259
+    },
+    {
+      "epoch": 1.1126279863481228,
+      "grad_norm": 3.223571538925171,
+      "learning_rate": 0.0006291240045506257,
+      "loss": 4.5392,
+      "step": 3260
+    },
+    {
+      "epoch": 1.1129692832764504,
+      "grad_norm": 3.6493613719940186,
+      "learning_rate": 0.0006290102389078498,
+      "loss": 6.3741,
+      "step": 3261
+    },
+    {
+      "epoch": 1.1133105802047782,
+      "grad_norm": 3.8748128414154053,
+      "learning_rate": 0.0006288964732650739,
+      "loss": 5.9105,
+      "step": 3262
+    },
+    {
+      "epoch": 1.1136518771331059,
+      "grad_norm": 3.320779323577881,
+      "learning_rate": 0.000628782707622298,
+      "loss": 6.4,
+      "step": 3263
+    },
+    {
+      "epoch": 1.1139931740614335,
+      "grad_norm": 3.4547297954559326,
+      "learning_rate": 0.0006286689419795221,
+      "loss": 6.1842,
+      "step": 3264
+    },
+    {
+      "epoch": 1.1143344709897611,
+      "grad_norm": 3.3176543712615967,
+      "learning_rate": 0.0006285551763367464,
+      "loss": 6.8925,
+      "step": 3265
+    },
+    {
+      "epoch": 1.1146757679180888,
+      "grad_norm": 3.095383882522583,
+      "learning_rate": 0.0006284414106939705,
+      "loss": 6.7177,
+      "step": 3266
+    },
+    {
+      "epoch": 1.1150170648464164,
+      "grad_norm": 3.1322736740112305,
+      "learning_rate": 0.0006283276450511946,
+      "loss": 7.11,
+      "step": 3267
+    },
+    {
+      "epoch": 1.115358361774744,
+      "grad_norm": 3.179048538208008,
+      "learning_rate": 0.0006282138794084187,
+      "loss": 6.6664,
+      "step": 3268
+    },
+    {
+      "epoch": 1.1156996587030716,
+      "grad_norm": 3.2446703910827637,
+      "learning_rate": 0.0006281001137656428,
+      "loss": 6.2555,
+      "step": 3269
+    },
+    {
+      "epoch": 1.1160409556313993,
+      "grad_norm": 3.4410719871520996,
+      "learning_rate": 0.0006279863481228669,
+      "loss": 6.3639,
+      "step": 3270
+    },
+    {
+      "epoch": 1.116382252559727,
+      "grad_norm": 3.151151657104492,
+      "learning_rate": 0.0006278725824800911,
+      "loss": 6.548,
+      "step": 3271
+    },
+    {
+      "epoch": 1.1167235494880545,
+      "grad_norm": 3.394761085510254,
+      "learning_rate": 0.0006277588168373152,
+      "loss": 6.7941,
+      "step": 3272
+    },
+    {
+      "epoch": 1.1170648464163822,
+      "grad_norm": 3.4084866046905518,
+      "learning_rate": 0.0006276450511945393,
+      "loss": 6.0128,
+      "step": 3273
+    },
+    {
+      "epoch": 1.11740614334471,
+      "grad_norm": 3.2266435623168945,
+      "learning_rate": 0.0006275312855517634,
+      "loss": 6.3047,
+      "step": 3274
+    },
+    {
+      "epoch": 1.1177474402730376,
+      "grad_norm": 3.2533907890319824,
+      "learning_rate": 0.0006274175199089875,
+      "loss": 6.4112,
+      "step": 3275
+    },
+    {
+      "epoch": 1.1180887372013653,
+      "grad_norm": 3.1983823776245117,
+      "learning_rate": 0.0006273037542662116,
+      "loss": 6.8299,
+      "step": 3276
+    },
+    {
+      "epoch": 1.1184300341296929,
+      "grad_norm": 3.781130075454712,
+      "learning_rate": 0.0006271899886234357,
+      "loss": 6.1287,
+      "step": 3277
+    },
+    {
+      "epoch": 1.1187713310580205,
+      "grad_norm": 3.3088560104370117,
+      "learning_rate": 0.0006270762229806598,
+      "loss": 6.1003,
+      "step": 3278
+    },
+    {
+      "epoch": 1.1191126279863481,
+      "grad_norm": 3.457223653793335,
+      "learning_rate": 0.0006269624573378839,
+      "loss": 6.1749,
+      "step": 3279
+    },
+    {
+      "epoch": 1.1194539249146758,
+      "grad_norm": 3.4108994007110596,
+      "learning_rate": 0.000626848691695108,
+      "loss": 6.4882,
+      "step": 3280
+    },
+    {
+      "epoch": 1.1197952218430034,
+      "grad_norm": 3.191033363342285,
+      "learning_rate": 0.0006267349260523321,
+      "loss": 6.386,
+      "step": 3281
+    },
+    {
+      "epoch": 1.120136518771331,
+      "grad_norm": 3.2378880977630615,
+      "learning_rate": 0.0006266211604095564,
+      "loss": 6.5682,
+      "step": 3282
+    },
+    {
+      "epoch": 1.1204778156996587,
+      "grad_norm": 3.0198142528533936,
+      "learning_rate": 0.0006265073947667805,
+      "loss": 6.6885,
+      "step": 3283
+    },
+    {
+      "epoch": 1.1208191126279863,
+      "grad_norm": 3.2005209922790527,
+      "learning_rate": 0.0006263936291240046,
+      "loss": 6.7357,
+      "step": 3284
+    },
+    {
+      "epoch": 1.121160409556314,
+      "grad_norm": 3.2650158405303955,
+      "learning_rate": 0.0006262798634812287,
+      "loss": 6.8819,
+      "step": 3285
+    },
+    {
+      "epoch": 1.1215017064846415,
+      "grad_norm": 3.4467949867248535,
+      "learning_rate": 0.0006261660978384528,
+      "loss": 6.327,
+      "step": 3286
+    },
+    {
+      "epoch": 1.1218430034129694,
+      "grad_norm": 3.51582932472229,
+      "learning_rate": 0.0006260523321956769,
+      "loss": 6.4275,
+      "step": 3287
+    },
+    {
+      "epoch": 1.122184300341297,
+      "grad_norm": 3.2341697216033936,
+      "learning_rate": 0.0006259385665529011,
+      "loss": 6.6526,
+      "step": 3288
+    },
+    {
+      "epoch": 1.1225255972696246,
+      "grad_norm": 3.0597870349884033,
+      "learning_rate": 0.0006258248009101252,
+      "loss": 6.3242,
+      "step": 3289
+    },
+    {
+      "epoch": 1.1228668941979523,
+      "grad_norm": 3.1777312755584717,
+      "learning_rate": 0.0006257110352673493,
+      "loss": 6.7306,
+      "step": 3290
+    },
+    {
+      "epoch": 1.12320819112628,
+      "grad_norm": 3.1080517768859863,
+      "learning_rate": 0.0006255972696245734,
+      "loss": 6.5034,
+      "step": 3291
+    },
+    {
+      "epoch": 1.1235494880546075,
+      "grad_norm": 3.166882276535034,
+      "learning_rate": 0.0006254835039817975,
+      "loss": 6.6549,
+      "step": 3292
+    },
+    {
+      "epoch": 1.1238907849829352,
+      "grad_norm": 3.2958264350891113,
+      "learning_rate": 0.0006253697383390216,
+      "loss": 6.3964,
+      "step": 3293
+    },
+    {
+      "epoch": 1.1242320819112628,
+      "grad_norm": 3.2784876823425293,
+      "learning_rate": 0.0006252559726962458,
+      "loss": 6.7125,
+      "step": 3294
+    },
+    {
+      "epoch": 1.1245733788395904,
+      "grad_norm": 3.357529878616333,
+      "learning_rate": 0.0006251422070534699,
+      "loss": 6.2972,
+      "step": 3295
+    },
+    {
+      "epoch": 1.124914675767918,
+      "grad_norm": 3.1187024116516113,
+      "learning_rate": 0.000625028441410694,
+      "loss": 6.4141,
+      "step": 3296
+    },
+    {
+      "epoch": 1.1252559726962457,
+      "grad_norm": 3.192850112915039,
+      "learning_rate": 0.000624914675767918,
+      "loss": 6.4792,
+      "step": 3297
+    },
+    {
+      "epoch": 1.1255972696245733,
+      "grad_norm": 3.124603509902954,
+      "learning_rate": 0.0006248009101251421,
+      "loss": 6.084,
+      "step": 3298
+    },
+    {
+      "epoch": 1.1259385665529011,
+      "grad_norm": 3.316884756088257,
+      "learning_rate": 0.0006246871444823662,
+      "loss": 6.8159,
+      "step": 3299
+    },
+    {
+      "epoch": 1.1262798634812285,
+      "grad_norm": 3.2244949340820312,
+      "learning_rate": 0.0006245733788395905,
+      "loss": 6.0295,
+      "step": 3300
+    },
+    {
+      "epoch": 1.1266211604095564,
+      "grad_norm": 3.214632749557495,
+      "learning_rate": 0.0006244596131968146,
+      "loss": 6.0611,
+      "step": 3301
+    },
+    {
+      "epoch": 1.126962457337884,
+      "grad_norm": 3.4261796474456787,
+      "learning_rate": 0.0006243458475540387,
+      "loss": 6.371,
+      "step": 3302
+    },
+    {
+      "epoch": 1.1273037542662117,
+      "grad_norm": 3.2653138637542725,
+      "learning_rate": 0.0006242320819112628,
+      "loss": 6.4282,
+      "step": 3303
+    },
+    {
+      "epoch": 1.1276450511945393,
+      "grad_norm": 3.3320472240448,
+      "learning_rate": 0.0006241183162684869,
+      "loss": 6.7758,
+      "step": 3304
+    },
+    {
+      "epoch": 1.127986348122867,
+      "grad_norm": 3.2385928630828857,
+      "learning_rate": 0.0006240045506257111,
+      "loss": 6.7638,
+      "step": 3305
+    },
+    {
+      "epoch": 1.1283276450511945,
+      "grad_norm": 3.225954055786133,
+      "learning_rate": 0.0006238907849829352,
+      "loss": 6.3909,
+      "step": 3306
+    },
+    {
+      "epoch": 1.1286689419795222,
+      "grad_norm": 3.2931458950042725,
+      "learning_rate": 0.0006237770193401593,
+      "loss": 6.7229,
+      "step": 3307
+    },
+    {
+      "epoch": 1.1290102389078498,
+      "grad_norm": 3.082092761993408,
+      "learning_rate": 0.0006236632536973834,
+      "loss": 6.673,
+      "step": 3308
+    },
+    {
+      "epoch": 1.1293515358361774,
+      "grad_norm": 4.754388809204102,
+      "learning_rate": 0.0006235494880546075,
+      "loss": 6.1331,
+      "step": 3309
+    },
+    {
+      "epoch": 1.129692832764505,
+      "grad_norm": 3.1302919387817383,
+      "learning_rate": 0.0006234357224118316,
+      "loss": 6.512,
+      "step": 3310
+    },
+    {
+      "epoch": 1.1300341296928327,
+      "grad_norm": 5.564751148223877,
+      "learning_rate": 0.0006233219567690558,
+      "loss": 6.6001,
+      "step": 3311
+    },
+    {
+      "epoch": 1.1303754266211605,
+      "grad_norm": 2.1751673221588135,
+      "learning_rate": 0.0006232081911262799,
+      "loss": 3.1882,
+      "step": 3312
+    },
+    {
+      "epoch": 1.130716723549488,
+      "grad_norm": 3.3056774139404297,
+      "learning_rate": 0.000623094425483504,
+      "loss": 6.0435,
+      "step": 3313
+    },
+    {
+      "epoch": 1.1310580204778158,
+      "grad_norm": 3.5110793113708496,
+      "learning_rate": 0.0006229806598407281,
+      "loss": 6.7698,
+      "step": 3314
+    },
+    {
+      "epoch": 1.1313993174061434,
+      "grad_norm": 3.8492422103881836,
+      "learning_rate": 0.0006228668941979523,
+      "loss": 5.9606,
+      "step": 3315
+    },
+    {
+      "epoch": 1.131740614334471,
+      "grad_norm": 3.3324551582336426,
+      "learning_rate": 0.0006227531285551762,
+      "loss": 6.1918,
+      "step": 3316
+    },
+    {
+      "epoch": 1.1320819112627987,
+      "grad_norm": 3.3533132076263428,
+      "learning_rate": 0.0006226393629124005,
+      "loss": 6.8835,
+      "step": 3317
+    },
+    {
+      "epoch": 1.1324232081911263,
+      "grad_norm": 3.216062068939209,
+      "learning_rate": 0.0006225255972696246,
+      "loss": 6.1811,
+      "step": 3318
+    },
+    {
+      "epoch": 1.132764505119454,
+      "grad_norm": 3.2382802963256836,
+      "learning_rate": 0.0006224118316268487,
+      "loss": 6.3388,
+      "step": 3319
+    },
+    {
+      "epoch": 1.1331058020477816,
+      "grad_norm": 3.712456226348877,
+      "learning_rate": 0.0006222980659840728,
+      "loss": 6.0058,
+      "step": 3320
+    },
+    {
+      "epoch": 1.1334470989761092,
+      "grad_norm": 3.3816704750061035,
+      "learning_rate": 0.0006221843003412969,
+      "loss": 6.3261,
+      "step": 3321
+    },
+    {
+      "epoch": 1.1337883959044368,
+      "grad_norm": 3.733504295349121,
+      "learning_rate": 0.000622070534698521,
+      "loss": 4.9361,
+      "step": 3322
+    },
+    {
+      "epoch": 1.1341296928327644,
+      "grad_norm": 3.436828851699829,
+      "learning_rate": 0.0006219567690557452,
+      "loss": 6.4064,
+      "step": 3323
+    },
+    {
+      "epoch": 1.134470989761092,
+      "grad_norm": 3.211754083633423,
+      "learning_rate": 0.0006218430034129693,
+      "loss": 6.9981,
+      "step": 3324
+    },
+    {
+      "epoch": 1.13481228668942,
+      "grad_norm": 3.2073490619659424,
+      "learning_rate": 0.0006217292377701934,
+      "loss": 6.2856,
+      "step": 3325
+    },
+    {
+      "epoch": 1.1351535836177473,
+      "grad_norm": 3.1387381553649902,
+      "learning_rate": 0.0006216154721274175,
+      "loss": 6.6068,
+      "step": 3326
+    },
+    {
+      "epoch": 1.1354948805460752,
+      "grad_norm": 5.683926582336426,
+      "learning_rate": 0.0006215017064846416,
+      "loss": 6.1509,
+      "step": 3327
+    },
+    {
+      "epoch": 1.1358361774744028,
+      "grad_norm": 3.4648234844207764,
+      "learning_rate": 0.0006213879408418658,
+      "loss": 6.0164,
+      "step": 3328
+    },
+    {
+      "epoch": 1.1361774744027304,
+      "grad_norm": 3.3188838958740234,
+      "learning_rate": 0.0006212741751990899,
+      "loss": 6.6927,
+      "step": 3329
+    },
+    {
+      "epoch": 1.136518771331058,
+      "grad_norm": 5.669882297515869,
+      "learning_rate": 0.000621160409556314,
+      "loss": 5.7178,
+      "step": 3330
+    },
+    {
+      "epoch": 1.1368600682593857,
+      "grad_norm": 3.321120023727417,
+      "learning_rate": 0.0006210466439135381,
+      "loss": 6.3605,
+      "step": 3331
+    },
+    {
+      "epoch": 1.1372013651877133,
+      "grad_norm": 3.19608473777771,
+      "learning_rate": 0.0006209328782707623,
+      "loss": 7.1759,
+      "step": 3332
+    },
+    {
+      "epoch": 1.137542662116041,
+      "grad_norm": 3.3638601303100586,
+      "learning_rate": 0.0006208191126279864,
+      "loss": 6.4738,
+      "step": 3333
+    },
+    {
+      "epoch": 1.1378839590443686,
+      "grad_norm": 3.9922034740448,
+      "learning_rate": 0.0006207053469852106,
+      "loss": 6.4603,
+      "step": 3334
+    },
+    {
+      "epoch": 1.1382252559726962,
+      "grad_norm": 3.5877456665039062,
+      "learning_rate": 0.0006205915813424347,
+      "loss": 5.6385,
+      "step": 3335
+    },
+    {
+      "epoch": 1.1385665529010238,
+      "grad_norm": 3.1927621364593506,
+      "learning_rate": 0.0006204778156996587,
+      "loss": 5.9584,
+      "step": 3336
+    },
+    {
+      "epoch": 1.1389078498293514,
+      "grad_norm": 7.028829574584961,
+      "learning_rate": 0.0006203640500568828,
+      "loss": 4.9702,
+      "step": 3337
+    },
+    {
+      "epoch": 1.1392491467576793,
+      "grad_norm": 3.2814626693725586,
+      "learning_rate": 0.0006202502844141069,
+      "loss": 6.8503,
+      "step": 3338
+    },
+    {
+      "epoch": 1.1395904436860067,
+      "grad_norm": 3.5198097229003906,
+      "learning_rate": 0.000620136518771331,
+      "loss": 6.2619,
+      "step": 3339
+    },
+    {
+      "epoch": 1.1399317406143346,
+      "grad_norm": 3.3513224124908447,
+      "learning_rate": 0.0006200227531285552,
+      "loss": 6.749,
+      "step": 3340
+    },
+    {
+      "epoch": 1.1402730375426622,
+      "grad_norm": 3.736387014389038,
+      "learning_rate": 0.0006199089874857793,
+      "loss": 5.8229,
+      "step": 3341
+    },
+    {
+      "epoch": 1.1406143344709898,
+      "grad_norm": 3.531832456588745,
+      "learning_rate": 0.0006197952218430034,
+      "loss": 5.5888,
+      "step": 3342
+    },
+    {
+      "epoch": 1.1409556313993174,
+      "grad_norm": 3.2506160736083984,
+      "learning_rate": 0.0006196814562002275,
+      "loss": 6.379,
+      "step": 3343
+    },
+    {
+      "epoch": 1.141296928327645,
+      "grad_norm": 5.999919414520264,
+      "learning_rate": 0.0006195676905574516,
+      "loss": 5.3758,
+      "step": 3344
+    },
+    {
+      "epoch": 1.1416382252559727,
+      "grad_norm": 3.500176191329956,
+      "learning_rate": 0.0006194539249146758,
+      "loss": 6.2346,
+      "step": 3345
+    },
+    {
+      "epoch": 1.1419795221843003,
+      "grad_norm": 2.100086212158203,
+      "learning_rate": 0.0006193401592718999,
+      "loss": 3.3155,
+      "step": 3346
+    },
+    {
+      "epoch": 1.142320819112628,
+      "grad_norm": 5.4075798988342285,
+      "learning_rate": 0.000619226393629124,
+      "loss": 4.6973,
+      "step": 3347
+    },
+    {
+      "epoch": 1.1426621160409556,
+      "grad_norm": 3.206171989440918,
+      "learning_rate": 0.0006191126279863481,
+      "loss": 6.4632,
+      "step": 3348
+    },
+    {
+      "epoch": 1.1430034129692832,
+      "grad_norm": 3.123758316040039,
+      "learning_rate": 0.0006189988623435723,
+      "loss": 6.6153,
+      "step": 3349
+    },
+    {
+      "epoch": 1.1433447098976108,
+      "grad_norm": 3.109133243560791,
+      "learning_rate": 0.0006188850967007964,
+      "loss": 6.9189,
+      "step": 3350
+    },
+    {
+      "epoch": 1.1436860068259387,
+      "grad_norm": 3.7639737129211426,
+      "learning_rate": 0.0006187713310580206,
+      "loss": 5.0232,
+      "step": 3351
+    },
+    {
+      "epoch": 1.144027303754266,
+      "grad_norm": 3.2191059589385986,
+      "learning_rate": 0.0006186575654152447,
+      "loss": 6.3187,
+      "step": 3352
+    },
+    {
+      "epoch": 1.144368600682594,
+      "grad_norm": 3.2868123054504395,
+      "learning_rate": 0.0006185437997724688,
+      "loss": 6.3158,
+      "step": 3353
+    },
+    {
+      "epoch": 1.1447098976109216,
+      "grad_norm": 3.318932294845581,
+      "learning_rate": 0.0006184300341296929,
+      "loss": 6.2764,
+      "step": 3354
+    },
+    {
+      "epoch": 1.1450511945392492,
+      "grad_norm": 3.06105899810791,
+      "learning_rate": 0.0006183162684869169,
+      "loss": 6.4749,
+      "step": 3355
+    },
+    {
+      "epoch": 1.1453924914675768,
+      "grad_norm": 3.42996883392334,
+      "learning_rate": 0.000618202502844141,
+      "loss": 6.438,
+      "step": 3356
+    },
+    {
+      "epoch": 1.1457337883959045,
+      "grad_norm": 3.3737375736236572,
+      "learning_rate": 0.0006180887372013652,
+      "loss": 6.6715,
+      "step": 3357
+    },
+    {
+      "epoch": 1.146075085324232,
+      "grad_norm": 3.796318531036377,
+      "learning_rate": 0.0006179749715585893,
+      "loss": 5.5668,
+      "step": 3358
+    },
+    {
+      "epoch": 1.1464163822525597,
+      "grad_norm": 3.3401639461517334,
+      "learning_rate": 0.0006178612059158134,
+      "loss": 6.3617,
+      "step": 3359
+    },
+    {
+      "epoch": 1.1467576791808873,
+      "grad_norm": 3.4352896213531494,
+      "learning_rate": 0.0006177474402730375,
+      "loss": 6.105,
+      "step": 3360
+    },
+    {
+      "epoch": 1.147098976109215,
+      "grad_norm": 3.1488349437713623,
+      "learning_rate": 0.0006176336746302616,
+      "loss": 6.588,
+      "step": 3361
+    },
+    {
+      "epoch": 1.1474402730375426,
+      "grad_norm": 3.108013868331909,
+      "learning_rate": 0.0006175199089874857,
+      "loss": 6.22,
+      "step": 3362
+    },
+    {
+      "epoch": 1.1477815699658702,
+      "grad_norm": 3.5639657974243164,
+      "learning_rate": 0.0006174061433447099,
+      "loss": 6.7161,
+      "step": 3363
+    },
+    {
+      "epoch": 1.148122866894198,
+      "grad_norm": 4.529385566711426,
+      "learning_rate": 0.000617292377701934,
+      "loss": 5.2737,
+      "step": 3364
+    },
+    {
+      "epoch": 1.1484641638225255,
+      "grad_norm": 5.763156414031982,
+      "learning_rate": 0.0006171786120591581,
+      "loss": 5.3901,
+      "step": 3365
+    },
+    {
+      "epoch": 1.1488054607508533,
+      "grad_norm": 3.319201707839966,
+      "learning_rate": 0.0006170648464163823,
+      "loss": 6.2037,
+      "step": 3366
+    },
+    {
+      "epoch": 1.149146757679181,
+      "grad_norm": 3.429962635040283,
+      "learning_rate": 0.0006169510807736064,
+      "loss": 6.1665,
+      "step": 3367
+    },
+    {
+      "epoch": 1.1494880546075086,
+      "grad_norm": 3.73044490814209,
+      "learning_rate": 0.0006168373151308306,
+      "loss": 6.0798,
+      "step": 3368
+    },
+    {
+      "epoch": 1.1498293515358362,
+      "grad_norm": 5.758520126342773,
+      "learning_rate": 0.0006167235494880547,
+      "loss": 5.322,
+      "step": 3369
+    },
+    {
+      "epoch": 1.1501706484641638,
+      "grad_norm": 3.261571168899536,
+      "learning_rate": 0.0006166097838452788,
+      "loss": 6.9592,
+      "step": 3370
+    },
+    {
+      "epoch": 1.1505119453924915,
+      "grad_norm": 3.1305222511291504,
+      "learning_rate": 0.0006164960182025029,
+      "loss": 6.0327,
+      "step": 3371
+    },
+    {
+      "epoch": 1.150853242320819,
+      "grad_norm": 3.285921335220337,
+      "learning_rate": 0.000616382252559727,
+      "loss": 6.7096,
+      "step": 3372
+    },
+    {
+      "epoch": 1.1511945392491467,
+      "grad_norm": 3.030154228210449,
+      "learning_rate": 0.0006162684869169511,
+      "loss": 6.5408,
+      "step": 3373
+    },
+    {
+      "epoch": 1.1515358361774743,
+      "grad_norm": 3.1218512058258057,
+      "learning_rate": 0.0006161547212741753,
+      "loss": 6.5412,
+      "step": 3374
+    },
+    {
+      "epoch": 1.151877133105802,
+      "grad_norm": 3.093693733215332,
+      "learning_rate": 0.0006160409556313993,
+      "loss": 6.7893,
+      "step": 3375
+    },
+    {
+      "epoch": 1.1522184300341296,
+      "grad_norm": 8.688549041748047,
+      "learning_rate": 0.0006159271899886234,
+      "loss": 5.4964,
+      "step": 3376
+    },
+    {
+      "epoch": 1.1525597269624575,
+      "grad_norm": 3.335278272628784,
+      "learning_rate": 0.0006158134243458475,
+      "loss": 6.1672,
+      "step": 3377
+    },
+    {
+      "epoch": 1.1529010238907849,
+      "grad_norm": 3.6452267169952393,
+      "learning_rate": 0.0006156996587030716,
+      "loss": 5.9682,
+      "step": 3378
+    },
+    {
+      "epoch": 1.1532423208191127,
+      "grad_norm": 3.369281530380249,
+      "learning_rate": 0.0006155858930602957,
+      "loss": 6.0847,
+      "step": 3379
+    },
+    {
+      "epoch": 1.1535836177474403,
+      "grad_norm": 3.305046319961548,
+      "learning_rate": 0.0006154721274175199,
+      "loss": 6.8861,
+      "step": 3380
+    },
+    {
+      "epoch": 1.153924914675768,
+      "grad_norm": 4.961038112640381,
+      "learning_rate": 0.000615358361774744,
+      "loss": 4.7492,
+      "step": 3381
+    },
+    {
+      "epoch": 1.1542662116040956,
+      "grad_norm": 3.369922161102295,
+      "learning_rate": 0.0006152445961319682,
+      "loss": 6.3582,
+      "step": 3382
+    },
+    {
+      "epoch": 1.1546075085324232,
+      "grad_norm": 3.3537933826446533,
+      "learning_rate": 0.0006151308304891923,
+      "loss": 6.2107,
+      "step": 3383
+    },
+    {
+      "epoch": 1.1549488054607508,
+      "grad_norm": 3.241588830947876,
+      "learning_rate": 0.0006150170648464164,
+      "loss": 6.0113,
+      "step": 3384
+    },
+    {
+      "epoch": 1.1552901023890785,
+      "grad_norm": 3.1636621952056885,
+      "learning_rate": 0.0006149032992036406,
+      "loss": 6.7975,
+      "step": 3385
+    },
+    {
+      "epoch": 1.155631399317406,
+      "grad_norm": 3.1564855575561523,
+      "learning_rate": 0.0006147895335608647,
+      "loss": 6.2154,
+      "step": 3386
+    },
+    {
+      "epoch": 1.1559726962457337,
+      "grad_norm": 3.18015456199646,
+      "learning_rate": 0.0006146757679180888,
+      "loss": 6.7107,
+      "step": 3387
+    },
+    {
+      "epoch": 1.1563139931740614,
+      "grad_norm": 3.214262008666992,
+      "learning_rate": 0.0006145620022753129,
+      "loss": 6.2703,
+      "step": 3388
+    },
+    {
+      "epoch": 1.156655290102389,
+      "grad_norm": 3.171013593673706,
+      "learning_rate": 0.000614448236632537,
+      "loss": 6.5638,
+      "step": 3389
+    },
+    {
+      "epoch": 1.1569965870307168,
+      "grad_norm": 3.184718608856201,
+      "learning_rate": 0.0006143344709897611,
+      "loss": 6.9765,
+      "step": 3390
+    },
+    {
+      "epoch": 1.1573378839590442,
+      "grad_norm": 3.161717414855957,
+      "learning_rate": 0.0006142207053469853,
+      "loss": 6.8384,
+      "step": 3391
+    },
+    {
+      "epoch": 1.157679180887372,
+      "grad_norm": 5.287087440490723,
+      "learning_rate": 0.0006141069397042094,
+      "loss": 6.4128,
+      "step": 3392
+    },
+    {
+      "epoch": 1.1580204778156997,
+      "grad_norm": 3.629075765609741,
+      "learning_rate": 0.0006139931740614335,
+      "loss": 6.171,
+      "step": 3393
+    },
+    {
+      "epoch": 1.1583617747440274,
+      "grad_norm": 3.260854721069336,
+      "learning_rate": 0.0006138794084186575,
+      "loss": 6.6395,
+      "step": 3394
+    },
+    {
+      "epoch": 1.158703071672355,
+      "grad_norm": 3.2226295471191406,
+      "learning_rate": 0.0006137656427758816,
+      "loss": 6.533,
+      "step": 3395
+    },
+    {
+      "epoch": 1.1590443686006826,
+      "grad_norm": 3.1546380519866943,
+      "learning_rate": 0.0006136518771331057,
+      "loss": 6.5444,
+      "step": 3396
+    },
+    {
+      "epoch": 1.1593856655290102,
+      "grad_norm": 3.138427257537842,
+      "learning_rate": 0.0006135381114903299,
+      "loss": 6.7728,
+      "step": 3397
+    },
+    {
+      "epoch": 1.1597269624573379,
+      "grad_norm": 3.0224416255950928,
+      "learning_rate": 0.000613424345847554,
+      "loss": 6.6283,
+      "step": 3398
+    },
+    {
+      "epoch": 1.1600682593856655,
+      "grad_norm": 3.074084997177124,
+      "learning_rate": 0.0006133105802047782,
+      "loss": 6.5736,
+      "step": 3399
+    },
+    {
+      "epoch": 1.1604095563139931,
+      "grad_norm": 4.4774909019470215,
+      "learning_rate": 0.0006131968145620023,
+      "loss": 5.5653,
+      "step": 3400
+    },
+    {
+      "epoch": 1.1607508532423207,
+      "grad_norm": 3.481534004211426,
+      "learning_rate": 0.0006130830489192264,
+      "loss": 6.0247,
+      "step": 3401
+    },
+    {
+      "epoch": 1.1610921501706484,
+      "grad_norm": 3.528672695159912,
+      "learning_rate": 0.0006129692832764505,
+      "loss": 6.7388,
+      "step": 3402
+    },
+    {
+      "epoch": 1.1614334470989762,
+      "grad_norm": 3.561124324798584,
+      "learning_rate": 0.0006128555176336747,
+      "loss": 6.303,
+      "step": 3403
+    },
+    {
+      "epoch": 1.1617747440273036,
+      "grad_norm": 3.158127546310425,
+      "learning_rate": 0.0006127417519908988,
+      "loss": 6.2118,
+      "step": 3404
+    },
+    {
+      "epoch": 1.1621160409556315,
+      "grad_norm": 3.1194143295288086,
+      "learning_rate": 0.0006126279863481229,
+      "loss": 6.4867,
+      "step": 3405
+    },
+    {
+      "epoch": 1.162457337883959,
+      "grad_norm": 3.337048292160034,
+      "learning_rate": 0.000612514220705347,
+      "loss": 6.7671,
+      "step": 3406
+    },
+    {
+      "epoch": 1.1627986348122867,
+      "grad_norm": 3.0700247287750244,
+      "learning_rate": 0.0006124004550625711,
+      "loss": 6.1681,
+      "step": 3407
+    },
+    {
+      "epoch": 1.1631399317406144,
+      "grad_norm": 3.1669013500213623,
+      "learning_rate": 0.0006122866894197953,
+      "loss": 6.8375,
+      "step": 3408
+    },
+    {
+      "epoch": 1.163481228668942,
+      "grad_norm": 3.3716862201690674,
+      "learning_rate": 0.0006121729237770194,
+      "loss": 6.3905,
+      "step": 3409
+    },
+    {
+      "epoch": 1.1638225255972696,
+      "grad_norm": 7.003607273101807,
+      "learning_rate": 0.0006120591581342435,
+      "loss": 6.1162,
+      "step": 3410
+    },
+    {
+      "epoch": 1.1641638225255972,
+      "grad_norm": 3.2531354427337646,
+      "learning_rate": 0.0006119453924914676,
+      "loss": 6.4429,
+      "step": 3411
+    },
+    {
+      "epoch": 1.1645051194539249,
+      "grad_norm": 3.7305870056152344,
+      "learning_rate": 0.0006118316268486917,
+      "loss": 6.4708,
+      "step": 3412
+    },
+    {
+      "epoch": 1.1648464163822525,
+      "grad_norm": 4.199586868286133,
+      "learning_rate": 0.0006117178612059157,
+      "loss": 6.2342,
+      "step": 3413
+    },
+    {
+      "epoch": 1.1651877133105801,
+      "grad_norm": 3.2192304134368896,
+      "learning_rate": 0.0006116040955631399,
+      "loss": 6.3102,
+      "step": 3414
+    },
+    {
+      "epoch": 1.1655290102389078,
+      "grad_norm": 3.2440237998962402,
+      "learning_rate": 0.000611490329920364,
+      "loss": 7.0148,
+      "step": 3415
+    },
+    {
+      "epoch": 1.1658703071672356,
+      "grad_norm": 3.203922748565674,
+      "learning_rate": 0.0006113765642775882,
+      "loss": 6.0749,
+      "step": 3416
+    },
+    {
+      "epoch": 1.1662116040955632,
+      "grad_norm": 3.4446823596954346,
+      "learning_rate": 0.0006112627986348123,
+      "loss": 6.4998,
+      "step": 3417
+    },
+    {
+      "epoch": 1.1665529010238909,
+      "grad_norm": 3.079411506652832,
+      "learning_rate": 0.0006111490329920364,
+      "loss": 6.3895,
+      "step": 3418
+    },
+    {
+      "epoch": 1.1668941979522185,
+      "grad_norm": 3.134840488433838,
+      "learning_rate": 0.0006110352673492605,
+      "loss": 6.3418,
+      "step": 3419
+    },
+    {
+      "epoch": 1.1672354948805461,
+      "grad_norm": 3.2079269886016846,
+      "learning_rate": 0.0006109215017064847,
+      "loss": 6.4572,
+      "step": 3420
+    },
+    {
+      "epoch": 1.1675767918088737,
+      "grad_norm": 4.834338665008545,
+      "learning_rate": 0.0006108077360637088,
+      "loss": 5.9253,
+      "step": 3421
+    },
+    {
+      "epoch": 1.1679180887372014,
+      "grad_norm": 3.2818708419799805,
+      "learning_rate": 0.0006106939704209329,
+      "loss": 6.789,
+      "step": 3422
+    },
+    {
+      "epoch": 1.168259385665529,
+      "grad_norm": 3.2917566299438477,
+      "learning_rate": 0.000610580204778157,
+      "loss": 6.8878,
+      "step": 3423
+    },
+    {
+      "epoch": 1.1686006825938566,
+      "grad_norm": 3.199425458908081,
+      "learning_rate": 0.0006104664391353811,
+      "loss": 6.4318,
+      "step": 3424
+    },
+    {
+      "epoch": 1.1689419795221843,
+      "grad_norm": 4.149702548980713,
+      "learning_rate": 0.0006103526734926053,
+      "loss": 5.2789,
+      "step": 3425
+    },
+    {
+      "epoch": 1.1692832764505119,
+      "grad_norm": 3.1588354110717773,
+      "learning_rate": 0.0006102389078498294,
+      "loss": 6.5981,
+      "step": 3426
+    },
+    {
+      "epoch": 1.1696245733788395,
+      "grad_norm": 3.282426595687866,
+      "learning_rate": 0.0006101251422070535,
+      "loss": 6.3488,
+      "step": 3427
+    },
+    {
+      "epoch": 1.1699658703071671,
+      "grad_norm": 3.071763515472412,
+      "learning_rate": 0.0006100113765642776,
+      "loss": 6.6647,
+      "step": 3428
+    },
+    {
+      "epoch": 1.170307167235495,
+      "grad_norm": 3.4117684364318848,
+      "learning_rate": 0.0006098976109215017,
+      "loss": 6.057,
+      "step": 3429
+    },
+    {
+      "epoch": 1.1706484641638226,
+      "grad_norm": 3.132619857788086,
+      "learning_rate": 0.0006097838452787258,
+      "loss": 6.3089,
+      "step": 3430
+    },
+    {
+      "epoch": 1.1709897610921502,
+      "grad_norm": 3.3119728565216064,
+      "learning_rate": 0.00060967007963595,
+      "loss": 6.6789,
+      "step": 3431
+    },
+    {
+      "epoch": 1.1713310580204779,
+      "grad_norm": 3.4108917713165283,
+      "learning_rate": 0.0006095563139931742,
+      "loss": 6.664,
+      "step": 3432
+    },
+    {
+      "epoch": 1.1716723549488055,
+      "grad_norm": 3.4007740020751953,
+      "learning_rate": 0.0006094425483503982,
+      "loss": 6.6983,
+      "step": 3433
+    },
+    {
+      "epoch": 1.1720136518771331,
+      "grad_norm": 3.2750303745269775,
+      "learning_rate": 0.0006093287827076223,
+      "loss": 6.7191,
+      "step": 3434
+    },
+    {
+      "epoch": 1.1723549488054608,
+      "grad_norm": 3.2517144680023193,
+      "learning_rate": 0.0006092150170648464,
+      "loss": 6.6333,
+      "step": 3435
+    },
+    {
+      "epoch": 1.1726962457337884,
+      "grad_norm": 3.0657079219818115,
+      "learning_rate": 0.0006091012514220705,
+      "loss": 7.0866,
+      "step": 3436
+    },
+    {
+      "epoch": 1.173037542662116,
+      "grad_norm": 3.8574588298797607,
+      "learning_rate": 0.0006089874857792947,
+      "loss": 6.2078,
+      "step": 3437
+    },
+    {
+      "epoch": 1.1733788395904436,
+      "grad_norm": 3.447795867919922,
+      "learning_rate": 0.0006088737201365188,
+      "loss": 6.5433,
+      "step": 3438
+    },
+    {
+      "epoch": 1.1737201365187713,
+      "grad_norm": 3.2688255310058594,
+      "learning_rate": 0.0006087599544937429,
+      "loss": 6.0859,
+      "step": 3439
+    },
+    {
+      "epoch": 1.174061433447099,
+      "grad_norm": 3.0544850826263428,
+      "learning_rate": 0.000608646188850967,
+      "loss": 6.6904,
+      "step": 3440
+    },
+    {
+      "epoch": 1.1744027303754265,
+      "grad_norm": 3.1975157260894775,
+      "learning_rate": 0.0006085324232081911,
+      "loss": 6.3464,
+      "step": 3441
+    },
+    {
+      "epoch": 1.1747440273037544,
+      "grad_norm": 3.1595888137817383,
+      "learning_rate": 0.0006084186575654152,
+      "loss": 6.721,
+      "step": 3442
+    },
+    {
+      "epoch": 1.175085324232082,
+      "grad_norm": 3.2810568809509277,
+      "learning_rate": 0.0006083048919226394,
+      "loss": 5.9367,
+      "step": 3443
+    },
+    {
+      "epoch": 1.1754266211604096,
+      "grad_norm": 3.1587705612182617,
+      "learning_rate": 0.0006081911262798635,
+      "loss": 6.2515,
+      "step": 3444
+    },
+    {
+      "epoch": 1.1757679180887373,
+      "grad_norm": 3.1385536193847656,
+      "learning_rate": 0.0006080773606370876,
+      "loss": 6.3264,
+      "step": 3445
+    },
+    {
+      "epoch": 1.176109215017065,
+      "grad_norm": 3.262084722518921,
+      "learning_rate": 0.0006079635949943117,
+      "loss": 6.7922,
+      "step": 3446
+    },
+    {
+      "epoch": 1.1764505119453925,
+      "grad_norm": 3.3084287643432617,
+      "learning_rate": 0.0006078498293515358,
+      "loss": 6.5934,
+      "step": 3447
+    },
+    {
+      "epoch": 1.1767918088737201,
+      "grad_norm": 7.798718452453613,
+      "learning_rate": 0.00060773606370876,
+      "loss": 5.7443,
+      "step": 3448
+    },
+    {
+      "epoch": 1.1771331058020478,
+      "grad_norm": 3.4016666412353516,
+      "learning_rate": 0.0006076222980659842,
+      "loss": 7.0521,
+      "step": 3449
+    },
+    {
+      "epoch": 1.1774744027303754,
+      "grad_norm": 3.541292905807495,
+      "learning_rate": 0.0006075085324232083,
+      "loss": 6.5505,
+      "step": 3450
+    },
+    {
+      "epoch": 1.177815699658703,
+      "grad_norm": 3.220148801803589,
+      "learning_rate": 0.0006073947667804324,
+      "loss": 6.8404,
+      "step": 3451
+    },
+    {
+      "epoch": 1.1781569965870307,
+      "grad_norm": 8.531793594360352,
+      "learning_rate": 0.0006072810011376564,
+      "loss": 5.1038,
+      "step": 3452
+    },
+    {
+      "epoch": 1.1784982935153583,
+      "grad_norm": 3.7234575748443604,
+      "learning_rate": 0.0006071672354948805,
+      "loss": 6.5902,
+      "step": 3453
+    },
+    {
+      "epoch": 1.178839590443686,
+      "grad_norm": 3.364846706390381,
+      "learning_rate": 0.0006070534698521047,
+      "loss": 6.1682,
+      "step": 3454
+    },
+    {
+      "epoch": 1.1791808873720138,
+      "grad_norm": 3.291454315185547,
+      "learning_rate": 0.0006069397042093288,
+      "loss": 6.5433,
+      "step": 3455
+    },
+    {
+      "epoch": 1.1795221843003414,
+      "grad_norm": 3.3431570529937744,
+      "learning_rate": 0.0006068259385665529,
+      "loss": 6.5502,
+      "step": 3456
+    },
+    {
+      "epoch": 1.179863481228669,
+      "grad_norm": 5.842983722686768,
+      "learning_rate": 0.000606712172923777,
+      "loss": 5.9161,
+      "step": 3457
+    },
+    {
+      "epoch": 1.1802047781569966,
+      "grad_norm": 3.1439208984375,
+      "learning_rate": 0.0006065984072810011,
+      "loss": 6.7444,
+      "step": 3458
+    },
+    {
+      "epoch": 1.1805460750853243,
+      "grad_norm": 3.290973424911499,
+      "learning_rate": 0.0006064846416382252,
+      "loss": 6.1831,
+      "step": 3459
+    },
+    {
+      "epoch": 1.180887372013652,
+      "grad_norm": 3.0965545177459717,
+      "learning_rate": 0.0006063708759954494,
+      "loss": 6.2298,
+      "step": 3460
+    },
+    {
+      "epoch": 1.1812286689419795,
+      "grad_norm": 3.1501073837280273,
+      "learning_rate": 0.0006062571103526735,
+      "loss": 5.6735,
+      "step": 3461
+    },
+    {
+      "epoch": 1.1815699658703072,
+      "grad_norm": 3.1459507942199707,
+      "learning_rate": 0.0006061433447098976,
+      "loss": 6.6602,
+      "step": 3462
+    },
+    {
+      "epoch": 1.1819112627986348,
+      "grad_norm": 6.302970886230469,
+      "learning_rate": 0.0006060295790671217,
+      "loss": 6.541,
+      "step": 3463
+    },
+    {
+      "epoch": 1.1822525597269624,
+      "grad_norm": 3.364511251449585,
+      "learning_rate": 0.0006059158134243458,
+      "loss": 6.4438,
+      "step": 3464
+    },
+    {
+      "epoch": 1.18259385665529,
+      "grad_norm": 3.289085626602173,
+      "learning_rate": 0.0006058020477815699,
+      "loss": 6.512,
+      "step": 3465
+    },
+    {
+      "epoch": 1.1829351535836177,
+      "grad_norm": 3.2350807189941406,
+      "learning_rate": 0.0006056882821387942,
+      "loss": 6.6411,
+      "step": 3466
+    },
+    {
+      "epoch": 1.1832764505119453,
+      "grad_norm": 3.8296823501586914,
+      "learning_rate": 0.0006055745164960183,
+      "loss": 6.2768,
+      "step": 3467
+    },
+    {
+      "epoch": 1.1836177474402731,
+      "grad_norm": 3.2926394939422607,
+      "learning_rate": 0.0006054607508532424,
+      "loss": 6.6457,
+      "step": 3468
+    },
+    {
+      "epoch": 1.1839590443686008,
+      "grad_norm": 3.056105613708496,
+      "learning_rate": 0.0006053469852104665,
+      "loss": 6.1748,
+      "step": 3469
+    },
+    {
+      "epoch": 1.1843003412969284,
+      "grad_norm": 3.2163851261138916,
+      "learning_rate": 0.0006052332195676906,
+      "loss": 6.1277,
+      "step": 3470
+    },
+    {
+      "epoch": 1.184641638225256,
+      "grad_norm": 3.086276054382324,
+      "learning_rate": 0.0006051194539249148,
+      "loss": 6.2933,
+      "step": 3471
+    },
+    {
+      "epoch": 1.1849829351535837,
+      "grad_norm": 5.672454833984375,
+      "learning_rate": 0.0006050056882821388,
+      "loss": 5.7082,
+      "step": 3472
+    },
+    {
+      "epoch": 1.1853242320819113,
+      "grad_norm": 3.3687098026275635,
+      "learning_rate": 0.0006048919226393629,
+      "loss": 6.9588,
+      "step": 3473
+    },
+    {
+      "epoch": 1.185665529010239,
+      "grad_norm": 3.393932580947876,
+      "learning_rate": 0.000604778156996587,
+      "loss": 6.7387,
+      "step": 3474
+    },
+    {
+      "epoch": 1.1860068259385665,
+      "grad_norm": 3.1261706352233887,
+      "learning_rate": 0.0006046643913538111,
+      "loss": 6.5774,
+      "step": 3475
+    },
+    {
+      "epoch": 1.1863481228668942,
+      "grad_norm": 3.0573582649230957,
+      "learning_rate": 0.0006045506257110352,
+      "loss": 6.3902,
+      "step": 3476
+    },
+    {
+      "epoch": 1.1866894197952218,
+      "grad_norm": 3.2194995880126953,
+      "learning_rate": 0.0006044368600682594,
+      "loss": 6.3269,
+      "step": 3477
+    },
+    {
+      "epoch": 1.1870307167235494,
+      "grad_norm": 3.1429195404052734,
+      "learning_rate": 0.0006043230944254835,
+      "loss": 6.4637,
+      "step": 3478
+    },
+    {
+      "epoch": 1.187372013651877,
+      "grad_norm": 3.0701262950897217,
+      "learning_rate": 0.0006042093287827076,
+      "loss": 6.6339,
+      "step": 3479
+    },
+    {
+      "epoch": 1.1877133105802047,
+      "grad_norm": 3.701120138168335,
+      "learning_rate": 0.0006040955631399317,
+      "loss": 6.075,
+      "step": 3480
+    },
+    {
+      "epoch": 1.1880546075085325,
+      "grad_norm": 3.3148467540740967,
+      "learning_rate": 0.0006039817974971558,
+      "loss": 6.1615,
+      "step": 3481
+    },
+    {
+      "epoch": 1.1883959044368602,
+      "grad_norm": 3.282397985458374,
+      "learning_rate": 0.0006038680318543799,
+      "loss": 6.6828,
+      "step": 3482
+    },
+    {
+      "epoch": 1.1887372013651878,
+      "grad_norm": 3.1746835708618164,
+      "learning_rate": 0.0006037542662116042,
+      "loss": 6.7092,
+      "step": 3483
+    },
+    {
+      "epoch": 1.1890784982935154,
+      "grad_norm": 3.044534206390381,
+      "learning_rate": 0.0006036405005688283,
+      "loss": 5.9796,
+      "step": 3484
+    },
+    {
+      "epoch": 1.189419795221843,
+      "grad_norm": 3.210864543914795,
+      "learning_rate": 0.0006035267349260524,
+      "loss": 6.522,
+      "step": 3485
+    },
+    {
+      "epoch": 1.1897610921501707,
+      "grad_norm": 3.328636646270752,
+      "learning_rate": 0.0006034129692832765,
+      "loss": 6.472,
+      "step": 3486
+    },
+    {
+      "epoch": 1.1901023890784983,
+      "grad_norm": 3.153980255126953,
+      "learning_rate": 0.0006032992036405006,
+      "loss": 7.1446,
+      "step": 3487
+    },
+    {
+      "epoch": 1.190443686006826,
+      "grad_norm": 3.198655128479004,
+      "learning_rate": 0.0006031854379977248,
+      "loss": 6.6485,
+      "step": 3488
+    },
+    {
+      "epoch": 1.1907849829351536,
+      "grad_norm": 3.2300875186920166,
+      "learning_rate": 0.0006030716723549489,
+      "loss": 6.5407,
+      "step": 3489
+    },
+    {
+      "epoch": 1.1911262798634812,
+      "grad_norm": 3.580975294113159,
+      "learning_rate": 0.000602957906712173,
+      "loss": 6.4401,
+      "step": 3490
+    },
+    {
+      "epoch": 1.1914675767918088,
+      "grad_norm": 3.525164842605591,
+      "learning_rate": 0.000602844141069397,
+      "loss": 6.486,
+      "step": 3491
+    },
+    {
+      "epoch": 1.1918088737201364,
+      "grad_norm": 3.407437562942505,
+      "learning_rate": 0.0006027303754266211,
+      "loss": 6.3749,
+      "step": 3492
+    },
+    {
+      "epoch": 1.192150170648464,
+      "grad_norm": 3.6547207832336426,
+      "learning_rate": 0.0006026166097838452,
+      "loss": 5.8585,
+      "step": 3493
+    },
+    {
+      "epoch": 1.192491467576792,
+      "grad_norm": 3.2446658611297607,
+      "learning_rate": 0.0006025028441410694,
+      "loss": 6.5614,
+      "step": 3494
+    },
+    {
+      "epoch": 1.1928327645051195,
+      "grad_norm": 3.150815963745117,
+      "learning_rate": 0.0006023890784982935,
+      "loss": 6.0707,
+      "step": 3495
+    },
+    {
+      "epoch": 1.1931740614334472,
+      "grad_norm": 5.976104736328125,
+      "learning_rate": 0.0006022753128555176,
+      "loss": 6.0327,
+      "step": 3496
+    },
+    {
+      "epoch": 1.1935153583617748,
+      "grad_norm": 3.18243408203125,
+      "learning_rate": 0.0006021615472127417,
+      "loss": 6.7489,
+      "step": 3497
+    },
+    {
+      "epoch": 1.1938566552901024,
+      "grad_norm": 3.4427740573883057,
+      "learning_rate": 0.0006020477815699658,
+      "loss": 5.9822,
+      "step": 3498
+    },
+    {
+      "epoch": 1.19419795221843,
+      "grad_norm": 3.2316274642944336,
+      "learning_rate": 0.00060193401592719,
+      "loss": 6.4285,
+      "step": 3499
+    },
+    {
+      "epoch": 1.1945392491467577,
+      "grad_norm": 3.2206499576568604,
+      "learning_rate": 0.0006018202502844142,
+      "loss": 6.5806,
+      "step": 3500
+    },
+    {
+      "epoch": 1.1948805460750853,
+      "grad_norm": 3.1856689453125,
+      "learning_rate": 0.0006017064846416383,
+      "loss": 6.5702,
+      "step": 3501
+    },
+    {
+      "epoch": 1.195221843003413,
+      "grad_norm": 3.3193821907043457,
+      "learning_rate": 0.0006015927189988624,
+      "loss": 6.6808,
+      "step": 3502
+    },
+    {
+      "epoch": 1.1955631399317406,
+      "grad_norm": 4.13479471206665,
+      "learning_rate": 0.0006014789533560865,
+      "loss": 5.7823,
+      "step": 3503
+    },
+    {
+      "epoch": 1.1959044368600682,
+      "grad_norm": 4.107855319976807,
+      "learning_rate": 0.0006013651877133106,
+      "loss": 5.8399,
+      "step": 3504
+    },
+    {
+      "epoch": 1.1962457337883958,
+      "grad_norm": 3.282972574234009,
+      "learning_rate": 0.0006012514220705347,
+      "loss": 6.0828,
+      "step": 3505
+    },
+    {
+      "epoch": 1.1965870307167235,
+      "grad_norm": 3.2156879901885986,
+      "learning_rate": 0.0006011376564277589,
+      "loss": 6.3889,
+      "step": 3506
+    },
+    {
+      "epoch": 1.1969283276450513,
+      "grad_norm": 3.4601330757141113,
+      "learning_rate": 0.000601023890784983,
+      "loss": 5.7612,
+      "step": 3507
+    },
+    {
+      "epoch": 1.197269624573379,
+      "grad_norm": 3.1823012828826904,
+      "learning_rate": 0.0006009101251422071,
+      "loss": 6.3664,
+      "step": 3508
+    },
+    {
+      "epoch": 1.1976109215017066,
+      "grad_norm": 3.098069906234741,
+      "learning_rate": 0.0006007963594994312,
+      "loss": 6.0548,
+      "step": 3509
+    },
+    {
+      "epoch": 1.1979522184300342,
+      "grad_norm": 3.121912956237793,
+      "learning_rate": 0.0006006825938566553,
+      "loss": 6.4154,
+      "step": 3510
+    },
+    {
+      "epoch": 1.1982935153583618,
+      "grad_norm": 3.1986167430877686,
+      "learning_rate": 0.0006005688282138794,
+      "loss": 6.5749,
+      "step": 3511
+    },
+    {
+      "epoch": 1.1986348122866894,
+      "grad_norm": 3.2935080528259277,
+      "learning_rate": 0.0006004550625711035,
+      "loss": 6.0803,
+      "step": 3512
+    },
+    {
+      "epoch": 1.198976109215017,
+      "grad_norm": 4.030786991119385,
+      "learning_rate": 0.0006003412969283276,
+      "loss": 6.0646,
+      "step": 3513
+    },
+    {
+      "epoch": 1.1993174061433447,
+      "grad_norm": 3.3741588592529297,
+      "learning_rate": 0.0006002275312855517,
+      "loss": 6.4553,
+      "step": 3514
+    },
+    {
+      "epoch": 1.1996587030716723,
+      "grad_norm": 3.256930112838745,
+      "learning_rate": 0.0006001137656427758,
+      "loss": 6.6533,
+      "step": 3515
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.3063952922821045,
+      "learning_rate": 0.0006,
+      "loss": 6.369,
+      "step": 3516
+    },
+    {
+      "epoch": 1.2003412969283276,
+      "grad_norm": 3.2215917110443115,
+      "learning_rate": 0.0005998862343572242,
+      "loss": 6.4497,
+      "step": 3517
+    },
+    {
+      "epoch": 1.2006825938566552,
+      "grad_norm": 4.139420032501221,
+      "learning_rate": 0.0005997724687144483,
+      "loss": 6.2639,
+      "step": 3518
+    },
+    {
+      "epoch": 1.2010238907849828,
+      "grad_norm": 3.161328077316284,
+      "learning_rate": 0.0005996587030716724,
+      "loss": 6.4553,
+      "step": 3519
+    },
+    {
+      "epoch": 1.2013651877133107,
+      "grad_norm": 3.1102607250213623,
+      "learning_rate": 0.0005995449374288965,
+      "loss": 6.6922,
+      "step": 3520
+    },
+    {
+      "epoch": 1.2017064846416383,
+      "grad_norm": 3.2522287368774414,
+      "learning_rate": 0.0005994311717861206,
+      "loss": 6.8216,
+      "step": 3521
+    },
+    {
+      "epoch": 1.202047781569966,
+      "grad_norm": 3.175919771194458,
+      "learning_rate": 0.0005993174061433447,
+      "loss": 6.2899,
+      "step": 3522
+    },
+    {
+      "epoch": 1.2023890784982936,
+      "grad_norm": 3.2448034286499023,
+      "learning_rate": 0.0005992036405005689,
+      "loss": 6.7938,
+      "step": 3523
+    },
+    {
+      "epoch": 1.2027303754266212,
+      "grad_norm": 3.362255811691284,
+      "learning_rate": 0.000599089874857793,
+      "loss": 6.4267,
+      "step": 3524
+    },
+    {
+      "epoch": 1.2030716723549488,
+      "grad_norm": 3.1305243968963623,
+      "learning_rate": 0.0005989761092150171,
+      "loss": 6.7481,
+      "step": 3525
+    },
+    {
+      "epoch": 1.2034129692832765,
+      "grad_norm": 3.1050209999084473,
+      "learning_rate": 0.0005988623435722412,
+      "loss": 6.5467,
+      "step": 3526
+    },
+    {
+      "epoch": 1.203754266211604,
+      "grad_norm": 3.1616594791412354,
+      "learning_rate": 0.0005987485779294653,
+      "loss": 6.7965,
+      "step": 3527
+    },
+    {
+      "epoch": 1.2040955631399317,
+      "grad_norm": 3.2334542274475098,
+      "learning_rate": 0.0005986348122866895,
+      "loss": 6.6105,
+      "step": 3528
+    },
+    {
+      "epoch": 1.2044368600682593,
+      "grad_norm": 4.107026100158691,
+      "learning_rate": 0.0005985210466439136,
+      "loss": 6.1539,
+      "step": 3529
+    },
+    {
+      "epoch": 1.204778156996587,
+      "grad_norm": 3.437368392944336,
+      "learning_rate": 0.0005984072810011376,
+      "loss": 6.7131,
+      "step": 3530
+    },
+    {
+      "epoch": 1.2051194539249146,
+      "grad_norm": 3.2274441719055176,
+      "learning_rate": 0.0005982935153583617,
+      "loss": 6.4707,
+      "step": 3531
+    },
+    {
+      "epoch": 1.2054607508532422,
+      "grad_norm": 3.5051920413970947,
+      "learning_rate": 0.0005981797497155858,
+      "loss": 6.4562,
+      "step": 3532
+    },
+    {
+      "epoch": 1.20580204778157,
+      "grad_norm": 3.2286267280578613,
+      "learning_rate": 0.00059806598407281,
+      "loss": 6.3122,
+      "step": 3533
+    },
+    {
+      "epoch": 1.2061433447098977,
+      "grad_norm": 3.3295505046844482,
+      "learning_rate": 0.0005979522184300342,
+      "loss": 6.7408,
+      "step": 3534
+    },
+    {
+      "epoch": 1.2064846416382253,
+      "grad_norm": 3.2205469608306885,
+      "learning_rate": 0.0005978384527872583,
+      "loss": 6.4856,
+      "step": 3535
+    },
+    {
+      "epoch": 1.206825938566553,
+      "grad_norm": 3.2709555625915527,
+      "learning_rate": 0.0005977246871444824,
+      "loss": 6.2622,
+      "step": 3536
+    },
+    {
+      "epoch": 1.2071672354948806,
+      "grad_norm": 3.1822428703308105,
+      "learning_rate": 0.0005976109215017065,
+      "loss": 6.2327,
+      "step": 3537
+    },
+    {
+      "epoch": 1.2075085324232082,
+      "grad_norm": 3.1136908531188965,
+      "learning_rate": 0.0005974971558589306,
+      "loss": 6.6299,
+      "step": 3538
+    },
+    {
+      "epoch": 1.2078498293515358,
+      "grad_norm": 3.7378649711608887,
+      "learning_rate": 0.0005973833902161547,
+      "loss": 5.8328,
+      "step": 3539
+    },
+    {
+      "epoch": 1.2081911262798635,
+      "grad_norm": 3.2474365234375,
+      "learning_rate": 0.0005972696245733789,
+      "loss": 6.47,
+      "step": 3540
+    },
+    {
+      "epoch": 1.208532423208191,
+      "grad_norm": 3.283991575241089,
+      "learning_rate": 0.000597155858930603,
+      "loss": 6.2577,
+      "step": 3541
+    },
+    {
+      "epoch": 1.2088737201365187,
+      "grad_norm": 3.1167514324188232,
+      "learning_rate": 0.0005970420932878271,
+      "loss": 6.4208,
+      "step": 3542
+    },
+    {
+      "epoch": 1.2092150170648464,
+      "grad_norm": 3.0923006534576416,
+      "learning_rate": 0.0005969283276450512,
+      "loss": 6.7528,
+      "step": 3543
+    },
+    {
+      "epoch": 1.209556313993174,
+      "grad_norm": 3.308056354522705,
+      "learning_rate": 0.0005968145620022753,
+      "loss": 7.0382,
+      "step": 3544
+    },
+    {
+      "epoch": 1.2098976109215016,
+      "grad_norm": 3.772918701171875,
+      "learning_rate": 0.0005967007963594994,
+      "loss": 6.3507,
+      "step": 3545
+    },
+    {
+      "epoch": 1.2102389078498295,
+      "grad_norm": 3.130896806716919,
+      "learning_rate": 0.0005965870307167236,
+      "loss": 6.9891,
+      "step": 3546
+    },
+    {
+      "epoch": 1.210580204778157,
+      "grad_norm": 6.454501152038574,
+      "learning_rate": 0.0005964732650739477,
+      "loss": 6.0827,
+      "step": 3547
+    },
+    {
+      "epoch": 1.2109215017064847,
+      "grad_norm": 3.259948968887329,
+      "learning_rate": 0.0005963594994311718,
+      "loss": 5.4018,
+      "step": 3548
+    },
+    {
+      "epoch": 1.2112627986348123,
+      "grad_norm": 3.4372003078460693,
+      "learning_rate": 0.000596245733788396,
+      "loss": 6.5116,
+      "step": 3549
+    },
+    {
+      "epoch": 1.21160409556314,
+      "grad_norm": 3.244441509246826,
+      "learning_rate": 0.00059613196814562,
+      "loss": 6.7611,
+      "step": 3550
+    },
+    {
+      "epoch": 1.2119453924914676,
+      "grad_norm": 3.4917261600494385,
+      "learning_rate": 0.0005960182025028442,
+      "loss": 5.7466,
+      "step": 3551
+    },
+    {
+      "epoch": 1.2122866894197952,
+      "grad_norm": 3.247199058532715,
+      "learning_rate": 0.0005959044368600683,
+      "loss": 6.3,
+      "step": 3552
+    },
+    {
+      "epoch": 1.2126279863481229,
+      "grad_norm": 3.103398323059082,
+      "learning_rate": 0.0005957906712172924,
+      "loss": 6.9337,
+      "step": 3553
+    },
+    {
+      "epoch": 1.2129692832764505,
+      "grad_norm": 3.120478868484497,
+      "learning_rate": 0.0005956769055745165,
+      "loss": 6.3149,
+      "step": 3554
+    },
+    {
+      "epoch": 1.213310580204778,
+      "grad_norm": 3.190842866897583,
+      "learning_rate": 0.0005955631399317406,
+      "loss": 6.078,
+      "step": 3555
+    },
+    {
+      "epoch": 1.2136518771331057,
+      "grad_norm": 3.1261842250823975,
+      "learning_rate": 0.0005954493742889647,
+      "loss": 6.4843,
+      "step": 3556
+    },
+    {
+      "epoch": 1.2139931740614334,
+      "grad_norm": 3.333153486251831,
+      "learning_rate": 0.0005953356086461889,
+      "loss": 6.1083,
+      "step": 3557
+    },
+    {
+      "epoch": 1.214334470989761,
+      "grad_norm": 3.2019073963165283,
+      "learning_rate": 0.000595221843003413,
+      "loss": 6.4916,
+      "step": 3558
+    },
+    {
+      "epoch": 1.2146757679180888,
+      "grad_norm": 3.232032060623169,
+      "learning_rate": 0.0005951080773606371,
+      "loss": 6.5154,
+      "step": 3559
+    },
+    {
+      "epoch": 1.2150170648464165,
+      "grad_norm": 3.299344301223755,
+      "learning_rate": 0.0005949943117178612,
+      "loss": 6.6801,
+      "step": 3560
+    },
+    {
+      "epoch": 1.215358361774744,
+      "grad_norm": 3.3514556884765625,
+      "learning_rate": 0.0005948805460750853,
+      "loss": 6.742,
+      "step": 3561
+    },
+    {
+      "epoch": 1.2156996587030717,
+      "grad_norm": 3.3210854530334473,
+      "learning_rate": 0.0005947667804323094,
+      "loss": 5.9867,
+      "step": 3562
+    },
+    {
+      "epoch": 1.2160409556313994,
+      "grad_norm": 3.2020204067230225,
+      "learning_rate": 0.0005946530147895336,
+      "loss": 6.715,
+      "step": 3563
+    },
+    {
+      "epoch": 1.216382252559727,
+      "grad_norm": 3.123915433883667,
+      "learning_rate": 0.0005945392491467577,
+      "loss": 6.6287,
+      "step": 3564
+    },
+    {
+      "epoch": 1.2167235494880546,
+      "grad_norm": 3.2689836025238037,
+      "learning_rate": 0.0005944254835039818,
+      "loss": 6.6374,
+      "step": 3565
+    },
+    {
+      "epoch": 1.2170648464163822,
+      "grad_norm": 4.918799877166748,
+      "learning_rate": 0.000594311717861206,
+      "loss": 5.6419,
+      "step": 3566
+    },
+    {
+      "epoch": 1.2174061433447099,
+      "grad_norm": 3.0726449489593506,
+      "learning_rate": 0.0005941979522184301,
+      "loss": 6.2995,
+      "step": 3567
+    },
+    {
+      "epoch": 1.2177474402730375,
+      "grad_norm": 3.210508108139038,
+      "learning_rate": 0.0005940841865756542,
+      "loss": 6.058,
+      "step": 3568
+    },
+    {
+      "epoch": 1.2180887372013651,
+      "grad_norm": 3.5676910877227783,
+      "learning_rate": 0.0005939704209328783,
+      "loss": 5.5099,
+      "step": 3569
+    },
+    {
+      "epoch": 1.2184300341296928,
+      "grad_norm": 3.3140625953674316,
+      "learning_rate": 0.0005938566552901024,
+      "loss": 6.7096,
+      "step": 3570
+    },
+    {
+      "epoch": 1.2187713310580204,
+      "grad_norm": 3.6696548461914062,
+      "learning_rate": 0.0005937428896473265,
+      "loss": 4.3158,
+      "step": 3571
+    },
+    {
+      "epoch": 1.2191126279863482,
+      "grad_norm": 3.3248839378356934,
+      "learning_rate": 0.0005936291240045506,
+      "loss": 6.9934,
+      "step": 3572
+    },
+    {
+      "epoch": 1.2194539249146759,
+      "grad_norm": 3.5815327167510986,
+      "learning_rate": 0.0005935153583617747,
+      "loss": 5.5062,
+      "step": 3573
+    },
+    {
+      "epoch": 1.2197952218430035,
+      "grad_norm": 3.4461398124694824,
+      "learning_rate": 0.0005934015927189989,
+      "loss": 6.7242,
+      "step": 3574
+    },
+    {
+      "epoch": 1.2201365187713311,
+      "grad_norm": 3.182681083679199,
+      "learning_rate": 0.000593287827076223,
+      "loss": 6.8025,
+      "step": 3575
+    },
+    {
+      "epoch": 1.2204778156996587,
+      "grad_norm": 3.6863200664520264,
+      "learning_rate": 0.0005931740614334471,
+      "loss": 6.4275,
+      "step": 3576
+    },
+    {
+      "epoch": 1.2208191126279864,
+      "grad_norm": 3.2513556480407715,
+      "learning_rate": 0.0005930602957906712,
+      "loss": 6.3153,
+      "step": 3577
+    },
+    {
+      "epoch": 1.221160409556314,
+      "grad_norm": 5.446213722229004,
+      "learning_rate": 0.0005929465301478953,
+      "loss": 6.1899,
+      "step": 3578
+    },
+    {
+      "epoch": 1.2215017064846416,
+      "grad_norm": 5.742640972137451,
+      "learning_rate": 0.0005928327645051194,
+      "loss": 4.5945,
+      "step": 3579
+    },
+    {
+      "epoch": 1.2218430034129693,
+      "grad_norm": 3.372920036315918,
+      "learning_rate": 0.0005927189988623436,
+      "loss": 6.4181,
+      "step": 3580
+    },
+    {
+      "epoch": 1.2221843003412969,
+      "grad_norm": 3.323343515396118,
+      "learning_rate": 0.0005926052332195677,
+      "loss": 6.5921,
+      "step": 3581
+    },
+    {
+      "epoch": 1.2225255972696245,
+      "grad_norm": 3.2678415775299072,
+      "learning_rate": 0.0005924914675767918,
+      "loss": 6.544,
+      "step": 3582
+    },
+    {
+      "epoch": 1.2228668941979521,
+      "grad_norm": 3.0852653980255127,
+      "learning_rate": 0.000592377701934016,
+      "loss": 6.5581,
+      "step": 3583
+    },
+    {
+      "epoch": 1.2232081911262798,
+      "grad_norm": 3.8021903038024902,
+      "learning_rate": 0.0005922639362912401,
+      "loss": 5.3767,
+      "step": 3584
+    },
+    {
+      "epoch": 1.2235494880546076,
+      "grad_norm": 3.143950939178467,
+      "learning_rate": 0.0005921501706484642,
+      "loss": 6.1757,
+      "step": 3585
+    },
+    {
+      "epoch": 1.2238907849829352,
+      "grad_norm": 3.0818896293640137,
+      "learning_rate": 0.0005920364050056884,
+      "loss": 6.837,
+      "step": 3586
+    },
+    {
+      "epoch": 1.2242320819112629,
+      "grad_norm": 2.336493730545044,
+      "learning_rate": 0.0005919226393629125,
+      "loss": 3.1151,
+      "step": 3587
+    },
+    {
+      "epoch": 1.2245733788395905,
+      "grad_norm": 3.16595721244812,
+      "learning_rate": 0.0005918088737201365,
+      "loss": 6.2877,
+      "step": 3588
+    },
+    {
+      "epoch": 1.2249146757679181,
+      "grad_norm": 4.046875,
+      "learning_rate": 0.0005916951080773606,
+      "loss": 5.9708,
+      "step": 3589
+    },
+    {
+      "epoch": 1.2252559726962458,
+      "grad_norm": 3.3334009647369385,
+      "learning_rate": 0.0005915813424345847,
+      "loss": 6.8327,
+      "step": 3590
+    },
+    {
+      "epoch": 1.2255972696245734,
+      "grad_norm": 3.452604055404663,
+      "learning_rate": 0.0005914675767918088,
+      "loss": 6.5917,
+      "step": 3591
+    },
+    {
+      "epoch": 1.225938566552901,
+      "grad_norm": 3.072746753692627,
+      "learning_rate": 0.000591353811149033,
+      "loss": 6.3403,
+      "step": 3592
+    },
+    {
+      "epoch": 1.2262798634812286,
+      "grad_norm": 3.248479127883911,
+      "learning_rate": 0.0005912400455062571,
+      "loss": 6.3434,
+      "step": 3593
+    },
+    {
+      "epoch": 1.2266211604095563,
+      "grad_norm": 3.0548229217529297,
+      "learning_rate": 0.0005911262798634812,
+      "loss": 6.6718,
+      "step": 3594
+    },
+    {
+      "epoch": 1.226962457337884,
+      "grad_norm": 3.1804940700531006,
+      "learning_rate": 0.0005910125142207053,
+      "loss": 6.3622,
+      "step": 3595
+    },
+    {
+      "epoch": 1.2273037542662115,
+      "grad_norm": 3.511061668395996,
+      "learning_rate": 0.0005908987485779294,
+      "loss": 6.1706,
+      "step": 3596
+    },
+    {
+      "epoch": 1.2276450511945391,
+      "grad_norm": 3.2657320499420166,
+      "learning_rate": 0.0005907849829351536,
+      "loss": 6.4201,
+      "step": 3597
+    },
+    {
+      "epoch": 1.227986348122867,
+      "grad_norm": 3.191793918609619,
+      "learning_rate": 0.0005906712172923777,
+      "loss": 6.2123,
+      "step": 3598
+    },
+    {
+      "epoch": 1.2283276450511946,
+      "grad_norm": 3.1244120597839355,
+      "learning_rate": 0.0005905574516496019,
+      "loss": 6.2795,
+      "step": 3599
+    },
+    {
+      "epoch": 1.2286689419795223,
+      "grad_norm": 3.2375831604003906,
+      "learning_rate": 0.000590443686006826,
+      "loss": 6.175,
+      "step": 3600
+    },
+    {
+      "epoch": 1.2290102389078499,
+      "grad_norm": 3.0911331176757812,
+      "learning_rate": 0.0005903299203640501,
+      "loss": 5.3192,
+      "step": 3601
+    },
+    {
+      "epoch": 1.2293515358361775,
+      "grad_norm": 3.3348145484924316,
+      "learning_rate": 0.0005902161547212742,
+      "loss": 6.614,
+      "step": 3602
+    },
+    {
+      "epoch": 1.2296928327645051,
+      "grad_norm": 3.407654047012329,
+      "learning_rate": 0.0005901023890784984,
+      "loss": 5.4858,
+      "step": 3603
+    },
+    {
+      "epoch": 1.2300341296928328,
+      "grad_norm": 3.8060050010681152,
+      "learning_rate": 0.0005899886234357225,
+      "loss": 6.4772,
+      "step": 3604
+    },
+    {
+      "epoch": 1.2303754266211604,
+      "grad_norm": 3.2755227088928223,
+      "learning_rate": 0.0005898748577929466,
+      "loss": 6.4768,
+      "step": 3605
+    },
+    {
+      "epoch": 1.230716723549488,
+      "grad_norm": 3.256535530090332,
+      "learning_rate": 0.0005897610921501707,
+      "loss": 6.8184,
+      "step": 3606
+    },
+    {
+      "epoch": 1.2310580204778157,
+      "grad_norm": 3.2337422370910645,
+      "learning_rate": 0.0005896473265073948,
+      "loss": 6.2524,
+      "step": 3607
+    },
+    {
+      "epoch": 1.2313993174061433,
+      "grad_norm": 3.262669801712036,
+      "learning_rate": 0.0005895335608646188,
+      "loss": 5.987,
+      "step": 3608
+    },
+    {
+      "epoch": 1.231740614334471,
+      "grad_norm": 3.1079490184783936,
+      "learning_rate": 0.000589419795221843,
+      "loss": 6.9693,
+      "step": 3609
+    },
+    {
+      "epoch": 1.2320819112627985,
+      "grad_norm": 3.214324951171875,
+      "learning_rate": 0.0005893060295790671,
+      "loss": 6.4031,
+      "step": 3610
+    },
+    {
+      "epoch": 1.2324232081911264,
+      "grad_norm": 3.1826834678649902,
+      "learning_rate": 0.0005891922639362912,
+      "loss": 6.5096,
+      "step": 3611
+    },
+    {
+      "epoch": 1.232764505119454,
+      "grad_norm": 3.2287216186523438,
+      "learning_rate": 0.0005890784982935153,
+      "loss": 6.587,
+      "step": 3612
+    },
+    {
+      "epoch": 1.2331058020477816,
+      "grad_norm": 3.313389778137207,
+      "learning_rate": 0.0005889647326507394,
+      "loss": 6.2603,
+      "step": 3613
+    },
+    {
+      "epoch": 1.2334470989761093,
+      "grad_norm": 3.2733147144317627,
+      "learning_rate": 0.0005888509670079636,
+      "loss": 6.7555,
+      "step": 3614
+    },
+    {
+      "epoch": 1.233788395904437,
+      "grad_norm": 3.265892505645752,
+      "learning_rate": 0.0005887372013651877,
+      "loss": 6.1515,
+      "step": 3615
+    },
+    {
+      "epoch": 1.2341296928327645,
+      "grad_norm": 3.340308904647827,
+      "learning_rate": 0.0005886234357224119,
+      "loss": 6.3667,
+      "step": 3616
+    },
+    {
+      "epoch": 1.2344709897610922,
+      "grad_norm": 3.3036296367645264,
+      "learning_rate": 0.000588509670079636,
+      "loss": 7.1784,
+      "step": 3617
+    },
+    {
+      "epoch": 1.2348122866894198,
+      "grad_norm": 3.338773250579834,
+      "learning_rate": 0.0005883959044368601,
+      "loss": 6.5319,
+      "step": 3618
+    },
+    {
+      "epoch": 1.2351535836177474,
+      "grad_norm": 3.15885329246521,
+      "learning_rate": 0.0005882821387940842,
+      "loss": 6.8377,
+      "step": 3619
+    },
+    {
+      "epoch": 1.235494880546075,
+      "grad_norm": 3.140784978866577,
+      "learning_rate": 0.0005881683731513084,
+      "loss": 6.2892,
+      "step": 3620
+    },
+    {
+      "epoch": 1.2358361774744027,
+      "grad_norm": 3.2073025703430176,
+      "learning_rate": 0.0005880546075085325,
+      "loss": 5.9276,
+      "step": 3621
+    },
+    {
+      "epoch": 1.2361774744027303,
+      "grad_norm": 4.180164813995361,
+      "learning_rate": 0.0005879408418657566,
+      "loss": 5.3202,
+      "step": 3622
+    },
+    {
+      "epoch": 1.236518771331058,
+      "grad_norm": 3.397432327270508,
+      "learning_rate": 0.0005878270762229807,
+      "loss": 6.0951,
+      "step": 3623
+    },
+    {
+      "epoch": 1.2368600682593858,
+      "grad_norm": 3.3566372394561768,
+      "learning_rate": 0.0005877133105802048,
+      "loss": 6.4745,
+      "step": 3624
+    },
+    {
+      "epoch": 1.2372013651877134,
+      "grad_norm": 3.236814260482788,
+      "learning_rate": 0.0005875995449374289,
+      "loss": 6.5627,
+      "step": 3625
+    },
+    {
+      "epoch": 1.237542662116041,
+      "grad_norm": 3.453336238861084,
+      "learning_rate": 0.0005874857792946531,
+      "loss": 5.8774,
+      "step": 3626
+    },
+    {
+      "epoch": 1.2378839590443687,
+      "grad_norm": 6.298210620880127,
+      "learning_rate": 0.0005873720136518771,
+      "loss": 5.7175,
+      "step": 3627
+    },
+    {
+      "epoch": 1.2382252559726963,
+      "grad_norm": 3.396085023880005,
+      "learning_rate": 0.0005872582480091012,
+      "loss": 6.5425,
+      "step": 3628
+    },
+    {
+      "epoch": 1.238566552901024,
+      "grad_norm": 3.876892566680908,
+      "learning_rate": 0.0005871444823663253,
+      "loss": 6.0154,
+      "step": 3629
+    },
+    {
+      "epoch": 1.2389078498293515,
+      "grad_norm": 3.221203088760376,
+      "learning_rate": 0.0005870307167235494,
+      "loss": 6.6467,
+      "step": 3630
+    },
+    {
+      "epoch": 1.2392491467576792,
+      "grad_norm": 4.505368232727051,
+      "learning_rate": 0.0005869169510807735,
+      "loss": 5.1206,
+      "step": 3631
+    },
+    {
+      "epoch": 1.2395904436860068,
+      "grad_norm": 3.1172850131988525,
+      "learning_rate": 0.0005868031854379977,
+      "loss": 6.6021,
+      "step": 3632
+    },
+    {
+      "epoch": 1.2399317406143344,
+      "grad_norm": 3.2955751419067383,
+      "learning_rate": 0.0005866894197952219,
+      "loss": 5.3394,
+      "step": 3633
+    },
+    {
+      "epoch": 1.240273037542662,
+      "grad_norm": 3.096254348754883,
+      "learning_rate": 0.000586575654152446,
+      "loss": 6.3385,
+      "step": 3634
+    },
+    {
+      "epoch": 1.2406143344709897,
+      "grad_norm": 3.1540637016296387,
+      "learning_rate": 0.0005864618885096701,
+      "loss": 6.8789,
+      "step": 3635
+    },
+    {
+      "epoch": 1.2409556313993173,
+      "grad_norm": 3.2090256214141846,
+      "learning_rate": 0.0005863481228668942,
+      "loss": 5.8644,
+      "step": 3636
+    },
+    {
+      "epoch": 1.2412969283276452,
+      "grad_norm": 3.128225564956665,
+      "learning_rate": 0.0005862343572241184,
+      "loss": 6.6566,
+      "step": 3637
+    },
+    {
+      "epoch": 1.2416382252559728,
+      "grad_norm": 4.46166467666626,
+      "learning_rate": 0.0005861205915813425,
+      "loss": 5.8012,
+      "step": 3638
+    },
+    {
+      "epoch": 1.2419795221843004,
+      "grad_norm": 3.2098820209503174,
+      "learning_rate": 0.0005860068259385666,
+      "loss": 6.0215,
+      "step": 3639
+    },
+    {
+      "epoch": 1.242320819112628,
+      "grad_norm": 4.331230163574219,
+      "learning_rate": 0.0005858930602957907,
+      "loss": 5.8321,
+      "step": 3640
+    },
+    {
+      "epoch": 1.2426621160409557,
+      "grad_norm": 3.4035027027130127,
+      "learning_rate": 0.0005857792946530148,
+      "loss": 6.608,
+      "step": 3641
+    },
+    {
+      "epoch": 1.2430034129692833,
+      "grad_norm": 3.085808515548706,
+      "learning_rate": 0.0005856655290102389,
+      "loss": 6.7047,
+      "step": 3642
+    },
+    {
+      "epoch": 1.243344709897611,
+      "grad_norm": 3.15273380279541,
+      "learning_rate": 0.0005855517633674631,
+      "loss": 5.9884,
+      "step": 3643
+    },
+    {
+      "epoch": 1.2436860068259386,
+      "grad_norm": 3.1481151580810547,
+      "learning_rate": 0.0005854379977246872,
+      "loss": 6.578,
+      "step": 3644
+    },
+    {
+      "epoch": 1.2440273037542662,
+      "grad_norm": 3.1449875831604004,
+      "learning_rate": 0.0005853242320819113,
+      "loss": 6.3897,
+      "step": 3645
+    },
+    {
+      "epoch": 1.2443686006825938,
+      "grad_norm": 3.1563587188720703,
+      "learning_rate": 0.0005852104664391354,
+      "loss": 6.6151,
+      "step": 3646
+    },
+    {
+      "epoch": 1.2447098976109214,
+      "grad_norm": 4.4367146492004395,
+      "learning_rate": 0.0005850967007963594,
+      "loss": 5.5777,
+      "step": 3647
+    },
+    {
+      "epoch": 1.245051194539249,
+      "grad_norm": 3.3431198596954346,
+      "learning_rate": 0.0005849829351535835,
+      "loss": 6.4409,
+      "step": 3648
+    },
+    {
+      "epoch": 1.2453924914675767,
+      "grad_norm": 4.078390121459961,
+      "learning_rate": 0.0005848691695108077,
+      "loss": 5.2748,
+      "step": 3649
+    },
+    {
+      "epoch": 1.2457337883959045,
+      "grad_norm": 3.3056936264038086,
+      "learning_rate": 0.0005847554038680319,
+      "loss": 6.1259,
+      "step": 3650
+    },
+    {
+      "epoch": 1.2460750853242322,
+      "grad_norm": 3.204592227935791,
+      "learning_rate": 0.000584641638225256,
+      "loss": 6.6076,
+      "step": 3651
+    },
+    {
+      "epoch": 1.2464163822525598,
+      "grad_norm": 3.014092445373535,
+      "learning_rate": 0.0005845278725824801,
+      "loss": 6.6634,
+      "step": 3652
+    },
+    {
+      "epoch": 1.2467576791808874,
+      "grad_norm": 2.9963581562042236,
+      "learning_rate": 0.0005844141069397042,
+      "loss": 6.0894,
+      "step": 3653
+    },
+    {
+      "epoch": 1.247098976109215,
+      "grad_norm": 3.0778584480285645,
+      "learning_rate": 0.0005843003412969284,
+      "loss": 6.6684,
+      "step": 3654
+    },
+    {
+      "epoch": 1.2474402730375427,
+      "grad_norm": 3.172811985015869,
+      "learning_rate": 0.0005841865756541525,
+      "loss": 6.6487,
+      "step": 3655
+    },
+    {
+      "epoch": 1.2477815699658703,
+      "grad_norm": 3.1092119216918945,
+      "learning_rate": 0.0005840728100113766,
+      "loss": 6.0495,
+      "step": 3656
+    },
+    {
+      "epoch": 1.248122866894198,
+      "grad_norm": 3.1255362033843994,
+      "learning_rate": 0.0005839590443686007,
+      "loss": 6.6304,
+      "step": 3657
+    },
+    {
+      "epoch": 1.2484641638225256,
+      "grad_norm": 3.0759193897247314,
+      "learning_rate": 0.0005838452787258248,
+      "loss": 6.9172,
+      "step": 3658
+    },
+    {
+      "epoch": 1.2488054607508532,
+      "grad_norm": 3.2935664653778076,
+      "learning_rate": 0.0005837315130830489,
+      "loss": 5.9662,
+      "step": 3659
+    },
+    {
+      "epoch": 1.2491467576791808,
+      "grad_norm": 3.370425224304199,
+      "learning_rate": 0.0005836177474402731,
+      "loss": 6.481,
+      "step": 3660
+    },
+    {
+      "epoch": 1.2494880546075084,
+      "grad_norm": 3.1100013256073,
+      "learning_rate": 0.0005835039817974972,
+      "loss": 6.6346,
+      "step": 3661
+    },
+    {
+      "epoch": 1.249829351535836,
+      "grad_norm": 3.2249128818511963,
+      "learning_rate": 0.0005833902161547213,
+      "loss": 6.2269,
+      "step": 3662
+    },
+    {
+      "epoch": 1.250170648464164,
+      "grad_norm": 4.628625869750977,
+      "learning_rate": 0.0005832764505119454,
+      "loss": 5.2683,
+      "step": 3663
+    },
+    {
+      "epoch": 1.2505119453924913,
+      "grad_norm": 3.3335788249969482,
+      "learning_rate": 0.0005831626848691695,
+      "loss": 5.914,
+      "step": 3664
+    },
+    {
+      "epoch": 1.2508532423208192,
+      "grad_norm": 3.5215747356414795,
+      "learning_rate": 0.0005830489192263936,
+      "loss": 5.747,
+      "step": 3665
+    },
+    {
+      "epoch": 1.2511945392491468,
+      "grad_norm": 3.2354788780212402,
+      "learning_rate": 0.0005829351535836177,
+      "loss": 5.9926,
+      "step": 3666
+    },
+    {
+      "epoch": 1.2515358361774744,
+      "grad_norm": 3.289290428161621,
+      "learning_rate": 0.0005828213879408419,
+      "loss": 6.9403,
+      "step": 3667
+    },
+    {
+      "epoch": 1.251877133105802,
+      "grad_norm": 3.2592785358428955,
+      "learning_rate": 0.000582707622298066,
+      "loss": 6.102,
+      "step": 3668
+    },
+    {
+      "epoch": 1.2522184300341297,
+      "grad_norm": 3.279924154281616,
+      "learning_rate": 0.0005825938566552901,
+      "loss": 6.5536,
+      "step": 3669
+    },
+    {
+      "epoch": 1.2525597269624573,
+      "grad_norm": 6.693105697631836,
+      "learning_rate": 0.0005824800910125142,
+      "loss": 4.9568,
+      "step": 3670
+    },
+    {
+      "epoch": 1.252901023890785,
+      "grad_norm": 3.1594643592834473,
+      "learning_rate": 0.0005823663253697383,
+      "loss": 6.6421,
+      "step": 3671
+    },
+    {
+      "epoch": 1.2532423208191126,
+      "grad_norm": 3.6135575771331787,
+      "learning_rate": 0.0005822525597269625,
+      "loss": 5.984,
+      "step": 3672
+    },
+    {
+      "epoch": 1.2535836177474402,
+      "grad_norm": 3.2949254512786865,
+      "learning_rate": 0.0005821387940841866,
+      "loss": 6.6492,
+      "step": 3673
+    },
+    {
+      "epoch": 1.253924914675768,
+      "grad_norm": 3.300008773803711,
+      "learning_rate": 0.0005820250284414107,
+      "loss": 5.9884,
+      "step": 3674
+    },
+    {
+      "epoch": 1.2542662116040955,
+      "grad_norm": 3.126598596572876,
+      "learning_rate": 0.0005819112627986348,
+      "loss": 6.3651,
+      "step": 3675
+    },
+    {
+      "epoch": 1.2546075085324233,
+      "grad_norm": 3.004107713699341,
+      "learning_rate": 0.0005817974971558589,
+      "loss": 6.3782,
+      "step": 3676
+    },
+    {
+      "epoch": 1.2549488054607507,
+      "grad_norm": 3.3004016876220703,
+      "learning_rate": 0.0005816837315130831,
+      "loss": 5.9222,
+      "step": 3677
+    },
+    {
+      "epoch": 1.2552901023890786,
+      "grad_norm": 3.169092893600464,
+      "learning_rate": 0.0005815699658703072,
+      "loss": 6.5686,
+      "step": 3678
+    },
+    {
+      "epoch": 1.2556313993174062,
+      "grad_norm": 4.297203063964844,
+      "learning_rate": 0.0005814562002275313,
+      "loss": 5.7196,
+      "step": 3679
+    },
+    {
+      "epoch": 1.2559726962457338,
+      "grad_norm": 3.2692508697509766,
+      "learning_rate": 0.0005813424345847554,
+      "loss": 6.7681,
+      "step": 3680
+    },
+    {
+      "epoch": 1.2563139931740614,
+      "grad_norm": 3.4113283157348633,
+      "learning_rate": 0.0005812286689419795,
+      "loss": 5.9537,
+      "step": 3681
+    },
+    {
+      "epoch": 1.256655290102389,
+      "grad_norm": 3.1349599361419678,
+      "learning_rate": 0.0005811149032992036,
+      "loss": 6.4065,
+      "step": 3682
+    },
+    {
+      "epoch": 1.2569965870307167,
+      "grad_norm": 3.8762474060058594,
+      "learning_rate": 0.0005810011376564279,
+      "loss": 6.0212,
+      "step": 3683
+    },
+    {
+      "epoch": 1.2573378839590443,
+      "grad_norm": 3.2077221870422363,
+      "learning_rate": 0.000580887372013652,
+      "loss": 6.3899,
+      "step": 3684
+    },
+    {
+      "epoch": 1.257679180887372,
+      "grad_norm": 3.3499104976654053,
+      "learning_rate": 0.0005807736063708761,
+      "loss": 6.1954,
+      "step": 3685
+    },
+    {
+      "epoch": 1.2580204778156996,
+      "grad_norm": 3.0911295413970947,
+      "learning_rate": 0.0005806598407281001,
+      "loss": 6.2744,
+      "step": 3686
+    },
+    {
+      "epoch": 1.2583617747440274,
+      "grad_norm": 3.294782876968384,
+      "learning_rate": 0.0005805460750853242,
+      "loss": 6.4801,
+      "step": 3687
+    },
+    {
+      "epoch": 1.2587030716723548,
+      "grad_norm": 3.7116618156433105,
+      "learning_rate": 0.0005804323094425483,
+      "loss": 5.8219,
+      "step": 3688
+    },
+    {
+      "epoch": 1.2590443686006827,
+      "grad_norm": 3.1158978939056396,
+      "learning_rate": 0.0005803185437997725,
+      "loss": 6.1431,
+      "step": 3689
+    },
+    {
+      "epoch": 1.25938566552901,
+      "grad_norm": 3.4460155963897705,
+      "learning_rate": 0.0005802047781569966,
+      "loss": 6.9438,
+      "step": 3690
+    },
+    {
+      "epoch": 1.259726962457338,
+      "grad_norm": 3.296086311340332,
+      "learning_rate": 0.0005800910125142207,
+      "loss": 6.5641,
+      "step": 3691
+    },
+    {
+      "epoch": 1.2600682593856656,
+      "grad_norm": 3.484635591506958,
+      "learning_rate": 0.0005799772468714448,
+      "loss": 6.6866,
+      "step": 3692
+    },
+    {
+      "epoch": 1.2604095563139932,
+      "grad_norm": 3.001859664916992,
+      "learning_rate": 0.0005798634812286689,
+      "loss": 6.845,
+      "step": 3693
+    },
+    {
+      "epoch": 1.2607508532423208,
+      "grad_norm": 3.264299154281616,
+      "learning_rate": 0.0005797497155858931,
+      "loss": 6.5779,
+      "step": 3694
+    },
+    {
+      "epoch": 1.2610921501706485,
+      "grad_norm": 4.64964485168457,
+      "learning_rate": 0.0005796359499431172,
+      "loss": 6.1898,
+      "step": 3695
+    },
+    {
+      "epoch": 1.261433447098976,
+      "grad_norm": 3.335503339767456,
+      "learning_rate": 0.0005795221843003413,
+      "loss": 6.6275,
+      "step": 3696
+    },
+    {
+      "epoch": 1.2617747440273037,
+      "grad_norm": 3.441882610321045,
+      "learning_rate": 0.0005794084186575654,
+      "loss": 6.1924,
+      "step": 3697
+    },
+    {
+      "epoch": 1.2621160409556313,
+      "grad_norm": 3.124157428741455,
+      "learning_rate": 0.0005792946530147895,
+      "loss": 6.4121,
+      "step": 3698
+    },
+    {
+      "epoch": 1.262457337883959,
+      "grad_norm": 3.6481199264526367,
+      "learning_rate": 0.0005791808873720136,
+      "loss": 6.0989,
+      "step": 3699
+    },
+    {
+      "epoch": 1.2627986348122868,
+      "grad_norm": 3.1325788497924805,
+      "learning_rate": 0.0005790671217292379,
+      "loss": 6.6209,
+      "step": 3700
+    },
+    {
+      "epoch": 1.2631399317406142,
+      "grad_norm": 3.0426456928253174,
+      "learning_rate": 0.000578953356086462,
+      "loss": 6.2306,
+      "step": 3701
+    },
+    {
+      "epoch": 1.263481228668942,
+      "grad_norm": 3.1229195594787598,
+      "learning_rate": 0.0005788395904436861,
+      "loss": 6.9213,
+      "step": 3702
+    },
+    {
+      "epoch": 1.2638225255972695,
+      "grad_norm": 3.100487232208252,
+      "learning_rate": 0.0005787258248009102,
+      "loss": 6.5218,
+      "step": 3703
+    },
+    {
+      "epoch": 1.2641638225255973,
+      "grad_norm": 3.2305166721343994,
+      "learning_rate": 0.0005786120591581343,
+      "loss": 6.4707,
+      "step": 3704
+    },
+    {
+      "epoch": 1.264505119453925,
+      "grad_norm": 3.1007251739501953,
+      "learning_rate": 0.0005784982935153583,
+      "loss": 7.0472,
+      "step": 3705
+    },
+    {
+      "epoch": 1.2648464163822526,
+      "grad_norm": 3.452636480331421,
+      "learning_rate": 0.0005783845278725825,
+      "loss": 6.0918,
+      "step": 3706
+    },
+    {
+      "epoch": 1.2651877133105802,
+      "grad_norm": 3.7483577728271484,
+      "learning_rate": 0.0005782707622298066,
+      "loss": 5.3098,
+      "step": 3707
+    },
+    {
+      "epoch": 1.2655290102389078,
+      "grad_norm": 3.2747714519500732,
+      "learning_rate": 0.0005781569965870307,
+      "loss": 6.4773,
+      "step": 3708
+    },
+    {
+      "epoch": 1.2658703071672355,
+      "grad_norm": 3.334418773651123,
+      "learning_rate": 0.0005780432309442548,
+      "loss": 5.9476,
+      "step": 3709
+    },
+    {
+      "epoch": 1.266211604095563,
+      "grad_norm": 3.388519287109375,
+      "learning_rate": 0.0005779294653014789,
+      "loss": 6.7832,
+      "step": 3710
+    },
+    {
+      "epoch": 1.2665529010238907,
+      "grad_norm": 3.2145025730133057,
+      "learning_rate": 0.000577815699658703,
+      "loss": 7.1023,
+      "step": 3711
+    },
+    {
+      "epoch": 1.2668941979522184,
+      "grad_norm": 3.1737163066864014,
+      "learning_rate": 0.0005777019340159272,
+      "loss": 6.2078,
+      "step": 3712
+    },
+    {
+      "epoch": 1.2672354948805462,
+      "grad_norm": 3.514313220977783,
+      "learning_rate": 0.0005775881683731513,
+      "loss": 5.9245,
+      "step": 3713
+    },
+    {
+      "epoch": 1.2675767918088736,
+      "grad_norm": 3.119048595428467,
+      "learning_rate": 0.0005774744027303754,
+      "loss": 6.1917,
+      "step": 3714
+    },
+    {
+      "epoch": 1.2679180887372015,
+      "grad_norm": 3.2024073600769043,
+      "learning_rate": 0.0005773606370875995,
+      "loss": 6.2773,
+      "step": 3715
+    },
+    {
+      "epoch": 1.268259385665529,
+      "grad_norm": 3.101339817047119,
+      "learning_rate": 0.0005772468714448236,
+      "loss": 6.6976,
+      "step": 3716
+    },
+    {
+      "epoch": 1.2686006825938567,
+      "grad_norm": 4.5886616706848145,
+      "learning_rate": 0.0005771331058020479,
+      "loss": 5.4188,
+      "step": 3717
+    },
+    {
+      "epoch": 1.2689419795221843,
+      "grad_norm": 3.2115495204925537,
+      "learning_rate": 0.000577019340159272,
+      "loss": 6.3656,
+      "step": 3718
+    },
+    {
+      "epoch": 1.269283276450512,
+      "grad_norm": 3.105013608932495,
+      "learning_rate": 0.0005769055745164961,
+      "loss": 6.04,
+      "step": 3719
+    },
+    {
+      "epoch": 1.2696245733788396,
+      "grad_norm": 3.715817451477051,
+      "learning_rate": 0.0005767918088737202,
+      "loss": 6.0363,
+      "step": 3720
+    },
+    {
+      "epoch": 1.2699658703071672,
+      "grad_norm": 3.0130434036254883,
+      "learning_rate": 0.0005766780432309443,
+      "loss": 6.3197,
+      "step": 3721
+    },
+    {
+      "epoch": 1.2703071672354949,
+      "grad_norm": 3.1294775009155273,
+      "learning_rate": 0.0005765642775881684,
+      "loss": 6.4423,
+      "step": 3722
+    },
+    {
+      "epoch": 1.2706484641638225,
+      "grad_norm": 3.1962530612945557,
+      "learning_rate": 0.0005764505119453926,
+      "loss": 6.3955,
+      "step": 3723
+    },
+    {
+      "epoch": 1.2709897610921501,
+      "grad_norm": 3.1761527061462402,
+      "learning_rate": 0.0005763367463026167,
+      "loss": 6.3929,
+      "step": 3724
+    },
+    {
+      "epoch": 1.2713310580204777,
+      "grad_norm": 3.09688138961792,
+      "learning_rate": 0.0005762229806598407,
+      "loss": 6.5535,
+      "step": 3725
+    },
+    {
+      "epoch": 1.2716723549488056,
+      "grad_norm": 3.14316463470459,
+      "learning_rate": 0.0005761092150170648,
+      "loss": 6.3527,
+      "step": 3726
+    },
+    {
+      "epoch": 1.272013651877133,
+      "grad_norm": 2.774230480194092,
+      "learning_rate": 0.0005759954493742889,
+      "loss": 3.1619,
+      "step": 3727
+    },
+    {
+      "epoch": 1.2723549488054609,
+      "grad_norm": 3.723099946975708,
+      "learning_rate": 0.000575881683731513,
+      "loss": 6.4016,
+      "step": 3728
+    },
+    {
+      "epoch": 1.2726962457337885,
+      "grad_norm": 3.691892385482788,
+      "learning_rate": 0.0005757679180887372,
+      "loss": 5.958,
+      "step": 3729
+    },
+    {
+      "epoch": 1.273037542662116,
+      "grad_norm": 3.364720582962036,
+      "learning_rate": 0.0005756541524459613,
+      "loss": 6.9903,
+      "step": 3730
+    },
+    {
+      "epoch": 1.2733788395904437,
+      "grad_norm": 3.500404119491577,
+      "learning_rate": 0.0005755403868031854,
+      "loss": 6.1624,
+      "step": 3731
+    },
+    {
+      "epoch": 1.2737201365187714,
+      "grad_norm": 4.1718597412109375,
+      "learning_rate": 0.0005754266211604095,
+      "loss": 6.2387,
+      "step": 3732
+    },
+    {
+      "epoch": 1.274061433447099,
+      "grad_norm": 3.1191086769104004,
+      "learning_rate": 0.0005753128555176336,
+      "loss": 7.2727,
+      "step": 3733
+    },
+    {
+      "epoch": 1.2744027303754266,
+      "grad_norm": 3.2956604957580566,
+      "learning_rate": 0.0005751990898748578,
+      "loss": 6.6461,
+      "step": 3734
+    },
+    {
+      "epoch": 1.2747440273037542,
+      "grad_norm": 3.2914106845855713,
+      "learning_rate": 0.000575085324232082,
+      "loss": 6.8773,
+      "step": 3735
+    },
+    {
+      "epoch": 1.2750853242320819,
+      "grad_norm": 3.343107223510742,
+      "learning_rate": 0.0005749715585893061,
+      "loss": 6.4053,
+      "step": 3736
+    },
+    {
+      "epoch": 1.2754266211604095,
+      "grad_norm": 3.184845447540283,
+      "learning_rate": 0.0005748577929465302,
+      "loss": 5.8288,
+      "step": 3737
+    },
+    {
+      "epoch": 1.2757679180887371,
+      "grad_norm": 3.4929275512695312,
+      "learning_rate": 0.0005747440273037543,
+      "loss": 6.2784,
+      "step": 3738
+    },
+    {
+      "epoch": 1.276109215017065,
+      "grad_norm": 3.1286535263061523,
+      "learning_rate": 0.0005746302616609784,
+      "loss": 6.8593,
+      "step": 3739
+    },
+    {
+      "epoch": 1.2764505119453924,
+      "grad_norm": 3.4388904571533203,
+      "learning_rate": 0.0005745164960182026,
+      "loss": 5.3746,
+      "step": 3740
+    },
+    {
+      "epoch": 1.2767918088737202,
+      "grad_norm": 3.239522933959961,
+      "learning_rate": 0.0005744027303754267,
+      "loss": 5.9408,
+      "step": 3741
+    },
+    {
+      "epoch": 1.2771331058020479,
+      "grad_norm": 3.316533088684082,
+      "learning_rate": 0.0005742889647326508,
+      "loss": 6.9465,
+      "step": 3742
+    },
+    {
+      "epoch": 1.2774744027303755,
+      "grad_norm": 3.470165967941284,
+      "learning_rate": 0.0005741751990898749,
+      "loss": 5.991,
+      "step": 3743
+    },
+    {
+      "epoch": 1.2778156996587031,
+      "grad_norm": 3.250302314758301,
+      "learning_rate": 0.0005740614334470989,
+      "loss": 6.1206,
+      "step": 3744
+    },
+    {
+      "epoch": 1.2781569965870307,
+      "grad_norm": 3.4087443351745605,
+      "learning_rate": 0.000573947667804323,
+      "loss": 6.4992,
+      "step": 3745
+    },
+    {
+      "epoch": 1.2784982935153584,
+      "grad_norm": 3.2577805519104004,
+      "learning_rate": 0.0005738339021615472,
+      "loss": 6.3207,
+      "step": 3746
+    },
+    {
+      "epoch": 1.278839590443686,
+      "grad_norm": 4.152851104736328,
+      "learning_rate": 0.0005737201365187713,
+      "loss": 4.8242,
+      "step": 3747
+    },
+    {
+      "epoch": 1.2791808873720136,
+      "grad_norm": 3.181995391845703,
+      "learning_rate": 0.0005736063708759954,
+      "loss": 6.548,
+      "step": 3748
+    },
+    {
+      "epoch": 1.2795221843003413,
+      "grad_norm": 3.4868032932281494,
+      "learning_rate": 0.0005734926052332195,
+      "loss": 6.2143,
+      "step": 3749
+    },
+    {
+      "epoch": 1.2798634812286689,
+      "grad_norm": 3.12666392326355,
+      "learning_rate": 0.0005733788395904436,
+      "loss": 6.9013,
+      "step": 3750
+    },
+    {
+      "epoch": 1.2802047781569965,
+      "grad_norm": 3.1087594032287598,
+      "learning_rate": 0.0005732650739476678,
+      "loss": 6.8698,
+      "step": 3751
+    },
+    {
+      "epoch": 1.2805460750853244,
+      "grad_norm": 3.2161049842834473,
+      "learning_rate": 0.000573151308304892,
+      "loss": 6.4872,
+      "step": 3752
+    },
+    {
+      "epoch": 1.2808873720136518,
+      "grad_norm": 3.140566110610962,
+      "learning_rate": 0.0005730375426621161,
+      "loss": 6.7521,
+      "step": 3753
+    },
+    {
+      "epoch": 1.2812286689419796,
+      "grad_norm": 3.375011920928955,
+      "learning_rate": 0.0005729237770193402,
+      "loss": 6.361,
+      "step": 3754
+    },
+    {
+      "epoch": 1.2815699658703072,
+      "grad_norm": 3.8893067836761475,
+      "learning_rate": 0.0005728100113765643,
+      "loss": 6.0318,
+      "step": 3755
+    },
+    {
+      "epoch": 1.2819112627986349,
+      "grad_norm": 3.1433284282684326,
+      "learning_rate": 0.0005726962457337884,
+      "loss": 6.0137,
+      "step": 3756
+    },
+    {
+      "epoch": 1.2822525597269625,
+      "grad_norm": 3.2328553199768066,
+      "learning_rate": 0.0005725824800910126,
+      "loss": 6.3901,
+      "step": 3757
+    },
+    {
+      "epoch": 1.2825938566552901,
+      "grad_norm": 3.0348122119903564,
+      "learning_rate": 0.0005724687144482367,
+      "loss": 6.2821,
+      "step": 3758
+    },
+    {
+      "epoch": 1.2829351535836178,
+      "grad_norm": 3.163795232772827,
+      "learning_rate": 0.0005723549488054608,
+      "loss": 6.2773,
+      "step": 3759
+    },
+    {
+      "epoch": 1.2832764505119454,
+      "grad_norm": 3.1185197830200195,
+      "learning_rate": 0.0005722411831626849,
+      "loss": 6.575,
+      "step": 3760
+    },
+    {
+      "epoch": 1.283617747440273,
+      "grad_norm": 4.635683536529541,
+      "learning_rate": 0.000572127417519909,
+      "loss": 5.0669,
+      "step": 3761
+    },
+    {
+      "epoch": 1.2839590443686006,
+      "grad_norm": 5.593306541442871,
+      "learning_rate": 0.0005720136518771331,
+      "loss": 5.0333,
+      "step": 3762
+    },
+    {
+      "epoch": 1.2843003412969283,
+      "grad_norm": 3.199862241744995,
+      "learning_rate": 0.0005718998862343572,
+      "loss": 6.5938,
+      "step": 3763
+    },
+    {
+      "epoch": 1.284641638225256,
+      "grad_norm": 3.554161787033081,
+      "learning_rate": 0.0005717861205915813,
+      "loss": 6.2362,
+      "step": 3764
+    },
+    {
+      "epoch": 1.2849829351535837,
+      "grad_norm": 3.397326946258545,
+      "learning_rate": 0.0005716723549488054,
+      "loss": 6.0153,
+      "step": 3765
+    },
+    {
+      "epoch": 1.2853242320819112,
+      "grad_norm": 3.148334264755249,
+      "learning_rate": 0.0005715585893060295,
+      "loss": 6.2978,
+      "step": 3766
+    },
+    {
+      "epoch": 1.285665529010239,
+      "grad_norm": 4.568434715270996,
+      "learning_rate": 0.0005714448236632536,
+      "loss": 5.762,
+      "step": 3767
+    },
+    {
+      "epoch": 1.2860068259385666,
+      "grad_norm": 3.284611225128174,
+      "learning_rate": 0.0005713310580204778,
+      "loss": 6.6268,
+      "step": 3768
+    },
+    {
+      "epoch": 1.2863481228668943,
+      "grad_norm": 14.985135078430176,
+      "learning_rate": 0.000571217292377702,
+      "loss": 4.1334,
+      "step": 3769
+    },
+    {
+      "epoch": 1.286689419795222,
+      "grad_norm": 3.237342119216919,
+      "learning_rate": 0.0005711035267349261,
+      "loss": 6.3474,
+      "step": 3770
+    },
+    {
+      "epoch": 1.2870307167235495,
+      "grad_norm": 3.287044048309326,
+      "learning_rate": 0.0005709897610921502,
+      "loss": 6.2234,
+      "step": 3771
+    },
+    {
+      "epoch": 1.2873720136518771,
+      "grad_norm": 3.0932626724243164,
+      "learning_rate": 0.0005708759954493743,
+      "loss": 6.8984,
+      "step": 3772
+    },
+    {
+      "epoch": 1.2877133105802048,
+      "grad_norm": 4.495637893676758,
+      "learning_rate": 0.0005707622298065984,
+      "loss": 5.0791,
+      "step": 3773
+    },
+    {
+      "epoch": 1.2880546075085324,
+      "grad_norm": 3.20120906829834,
+      "learning_rate": 0.0005706484641638225,
+      "loss": 6.4873,
+      "step": 3774
+    },
+    {
+      "epoch": 1.28839590443686,
+      "grad_norm": 3.146425485610962,
+      "learning_rate": 0.0005705346985210467,
+      "loss": 6.2309,
+      "step": 3775
+    },
+    {
+      "epoch": 1.2887372013651877,
+      "grad_norm": 3.152387857437134,
+      "learning_rate": 0.0005704209328782708,
+      "loss": 6.3663,
+      "step": 3776
+    },
+    {
+      "epoch": 1.2890784982935153,
+      "grad_norm": 3.333812952041626,
+      "learning_rate": 0.0005703071672354949,
+      "loss": 6.1972,
+      "step": 3777
+    },
+    {
+      "epoch": 1.2894197952218431,
+      "grad_norm": 4.746973037719727,
+      "learning_rate": 0.000570193401592719,
+      "loss": 4.9203,
+      "step": 3778
+    },
+    {
+      "epoch": 1.2897610921501705,
+      "grad_norm": 5.360435962677002,
+      "learning_rate": 0.0005700796359499431,
+      "loss": 5.874,
+      "step": 3779
+    },
+    {
+      "epoch": 1.2901023890784984,
+      "grad_norm": 3.3603954315185547,
+      "learning_rate": 0.0005699658703071673,
+      "loss": 6.2611,
+      "step": 3780
+    },
+    {
+      "epoch": 1.290443686006826,
+      "grad_norm": 3.251047134399414,
+      "learning_rate": 0.0005698521046643914,
+      "loss": 5.6341,
+      "step": 3781
+    },
+    {
+      "epoch": 1.2907849829351536,
+      "grad_norm": 3.316932201385498,
+      "learning_rate": 0.0005697383390216155,
+      "loss": 6.9426,
+      "step": 3782
+    },
+    {
+      "epoch": 1.2911262798634813,
+      "grad_norm": 5.679823398590088,
+      "learning_rate": 0.0005696245733788395,
+      "loss": 5.4432,
+      "step": 3783
+    },
+    {
+      "epoch": 1.291467576791809,
+      "grad_norm": 3.251844644546509,
+      "learning_rate": 0.0005695108077360637,
+      "loss": 6.9916,
+      "step": 3784
+    },
+    {
+      "epoch": 1.2918088737201365,
+      "grad_norm": 3.1990809440612793,
+      "learning_rate": 0.0005693970420932878,
+      "loss": 6.6435,
+      "step": 3785
+    },
+    {
+      "epoch": 1.2921501706484642,
+      "grad_norm": 3.5617237091064453,
+      "learning_rate": 0.000569283276450512,
+      "loss": 5.8256,
+      "step": 3786
+    },
+    {
+      "epoch": 1.2924914675767918,
+      "grad_norm": 3.2828528881073,
+      "learning_rate": 0.0005691695108077361,
+      "loss": 3.0921,
+      "step": 3787
+    },
+    {
+      "epoch": 1.2928327645051194,
+      "grad_norm": 4.945265769958496,
+      "learning_rate": 0.0005690557451649602,
+      "loss": 5.9462,
+      "step": 3788
+    },
+    {
+      "epoch": 1.293174061433447,
+      "grad_norm": 4.702214241027832,
+      "learning_rate": 0.0005689419795221843,
+      "loss": 5.9888,
+      "step": 3789
+    },
+    {
+      "epoch": 1.2935153583617747,
+      "grad_norm": 3.3534915447235107,
+      "learning_rate": 0.0005688282138794084,
+      "loss": 6.2144,
+      "step": 3790
+    },
+    {
+      "epoch": 1.2938566552901025,
+      "grad_norm": 3.387922525405884,
+      "learning_rate": 0.0005687144482366325,
+      "loss": 5.9508,
+      "step": 3791
+    },
+    {
+      "epoch": 1.29419795221843,
+      "grad_norm": 3.280163049697876,
+      "learning_rate": 0.0005686006825938567,
+      "loss": 7.1366,
+      "step": 3792
+    },
+    {
+      "epoch": 1.2945392491467578,
+      "grad_norm": 3.0818960666656494,
+      "learning_rate": 0.0005684869169510808,
+      "loss": 6.249,
+      "step": 3793
+    },
+    {
+      "epoch": 1.2948805460750854,
+      "grad_norm": 3.1594650745391846,
+      "learning_rate": 0.0005683731513083049,
+      "loss": 6.8152,
+      "step": 3794
+    },
+    {
+      "epoch": 1.295221843003413,
+      "grad_norm": 3.0736000537872314,
+      "learning_rate": 0.000568259385665529,
+      "loss": 6.0188,
+      "step": 3795
+    },
+    {
+      "epoch": 1.2955631399317407,
+      "grad_norm": 3.239778757095337,
+      "learning_rate": 0.0005681456200227531,
+      "loss": 6.2566,
+      "step": 3796
+    },
+    {
+      "epoch": 1.2959044368600683,
+      "grad_norm": 3.039710521697998,
+      "learning_rate": 0.0005680318543799773,
+      "loss": 6.3336,
+      "step": 3797
+    },
+    {
+      "epoch": 1.296245733788396,
+      "grad_norm": 3.1235196590423584,
+      "learning_rate": 0.0005679180887372014,
+      "loss": 6.0182,
+      "step": 3798
+    },
+    {
+      "epoch": 1.2965870307167235,
+      "grad_norm": 3.3898894786834717,
+      "learning_rate": 0.0005678043230944255,
+      "loss": 5.9916,
+      "step": 3799
+    },
+    {
+      "epoch": 1.2969283276450512,
+      "grad_norm": 3.1873528957366943,
+      "learning_rate": 0.0005676905574516497,
+      "loss": 6.7079,
+      "step": 3800
+    },
+    {
+      "epoch": 1.2972696245733788,
+      "grad_norm": 3.133732557296753,
+      "learning_rate": 0.0005675767918088738,
+      "loss": 6.0902,
+      "step": 3801
+    },
+    {
+      "epoch": 1.2976109215017064,
+      "grad_norm": 3.195161819458008,
+      "learning_rate": 0.0005674630261660978,
+      "loss": 6.3961,
+      "step": 3802
+    },
+    {
+      "epoch": 1.297952218430034,
+      "grad_norm": 3.204354763031006,
+      "learning_rate": 0.000567349260523322,
+      "loss": 5.9973,
+      "step": 3803
+    },
+    {
+      "epoch": 1.298293515358362,
+      "grad_norm": 3.1573190689086914,
+      "learning_rate": 0.0005672354948805461,
+      "loss": 6.5401,
+      "step": 3804
+    },
+    {
+      "epoch": 1.2986348122866893,
+      "grad_norm": 3.1424853801727295,
+      "learning_rate": 0.0005671217292377702,
+      "loss": 6.5217,
+      "step": 3805
+    },
+    {
+      "epoch": 1.2989761092150172,
+      "grad_norm": 3.2337796688079834,
+      "learning_rate": 0.0005670079635949943,
+      "loss": 6.818,
+      "step": 3806
+    },
+    {
+      "epoch": 1.2993174061433448,
+      "grad_norm": 3.1818370819091797,
+      "learning_rate": 0.0005668941979522184,
+      "loss": 6.4562,
+      "step": 3807
+    },
+    {
+      "epoch": 1.2996587030716724,
+      "grad_norm": 3.1999707221984863,
+      "learning_rate": 0.0005667804323094425,
+      "loss": 6.2445,
+      "step": 3808
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.0630509853363037,
+      "learning_rate": 0.0005666666666666667,
+      "loss": 6.6849,
+      "step": 3809
+    },
+    {
+      "epoch": 1.3003412969283277,
+      "grad_norm": 3.165708541870117,
+      "learning_rate": 0.0005665529010238908,
+      "loss": 6.9432,
+      "step": 3810
+    },
+    {
+      "epoch": 1.3006825938566553,
+      "grad_norm": 3.325191020965576,
+      "learning_rate": 0.0005664391353811149,
+      "loss": 6.0563,
+      "step": 3811
+    },
+    {
+      "epoch": 1.301023890784983,
+      "grad_norm": 3.1484081745147705,
+      "learning_rate": 0.000566325369738339,
+      "loss": 6.8012,
+      "step": 3812
+    },
+    {
+      "epoch": 1.3013651877133106,
+      "grad_norm": 3.0226058959960938,
+      "learning_rate": 0.0005662116040955631,
+      "loss": 6.4212,
+      "step": 3813
+    },
+    {
+      "epoch": 1.3017064846416382,
+      "grad_norm": 3.7095937728881836,
+      "learning_rate": 0.0005660978384527872,
+      "loss": 5.5472,
+      "step": 3814
+    },
+    {
+      "epoch": 1.3020477815699658,
+      "grad_norm": 8.578839302062988,
+      "learning_rate": 0.0005659840728100114,
+      "loss": 6.6667,
+      "step": 3815
+    },
+    {
+      "epoch": 1.3023890784982934,
+      "grad_norm": 3.352297067642212,
+      "learning_rate": 0.0005658703071672356,
+      "loss": 6.7404,
+      "step": 3816
+    },
+    {
+      "epoch": 1.3027303754266213,
+      "grad_norm": 3.2766904830932617,
+      "learning_rate": 0.0005657565415244597,
+      "loss": 5.9404,
+      "step": 3817
+    },
+    {
+      "epoch": 1.3030716723549487,
+      "grad_norm": 3.1568057537078857,
+      "learning_rate": 0.0005656427758816838,
+      "loss": 6.3997,
+      "step": 3818
+    },
+    {
+      "epoch": 1.3034129692832765,
+      "grad_norm": 3.079411029815674,
+      "learning_rate": 0.0005655290102389079,
+      "loss": 6.5721,
+      "step": 3819
+    },
+    {
+      "epoch": 1.3037542662116042,
+      "grad_norm": 3.156961679458618,
+      "learning_rate": 0.0005654152445961321,
+      "loss": 6.3826,
+      "step": 3820
+    },
+    {
+      "epoch": 1.3040955631399318,
+      "grad_norm": 3.102477788925171,
+      "learning_rate": 0.0005653014789533562,
+      "loss": 6.4791,
+      "step": 3821
+    },
+    {
+      "epoch": 1.3044368600682594,
+      "grad_norm": 3.097460985183716,
+      "learning_rate": 0.0005651877133105802,
+      "loss": 6.6575,
+      "step": 3822
+    },
+    {
+      "epoch": 1.304778156996587,
+      "grad_norm": 3.1785974502563477,
+      "learning_rate": 0.0005650739476678043,
+      "loss": 6.6686,
+      "step": 3823
+    },
+    {
+      "epoch": 1.3051194539249147,
+      "grad_norm": 3.1115939617156982,
+      "learning_rate": 0.0005649601820250284,
+      "loss": 6.6035,
+      "step": 3824
+    },
+    {
+      "epoch": 1.3054607508532423,
+      "grad_norm": 3.2212889194488525,
+      "learning_rate": 0.0005648464163822525,
+      "loss": 6.1973,
+      "step": 3825
+    },
+    {
+      "epoch": 1.30580204778157,
+      "grad_norm": 3.3271737098693848,
+      "learning_rate": 0.0005647326507394767,
+      "loss": 5.9092,
+      "step": 3826
+    },
+    {
+      "epoch": 1.3061433447098976,
+      "grad_norm": 3.7578651905059814,
+      "learning_rate": 0.0005646188850967008,
+      "loss": 5.8015,
+      "step": 3827
+    },
+    {
+      "epoch": 1.3064846416382252,
+      "grad_norm": 3.2217273712158203,
+      "learning_rate": 0.0005645051194539249,
+      "loss": 6.6484,
+      "step": 3828
+    },
+    {
+      "epoch": 1.3068259385665528,
+      "grad_norm": 3.182534694671631,
+      "learning_rate": 0.000564391353811149,
+      "loss": 6.494,
+      "step": 3829
+    },
+    {
+      "epoch": 1.3071672354948807,
+      "grad_norm": 3.2115378379821777,
+      "learning_rate": 0.0005642775881683731,
+      "loss": 6.4137,
+      "step": 3830
+    },
+    {
+      "epoch": 1.307508532423208,
+      "grad_norm": 3.060603618621826,
+      "learning_rate": 0.0005641638225255972,
+      "loss": 6.485,
+      "step": 3831
+    },
+    {
+      "epoch": 1.307849829351536,
+      "grad_norm": 3.5397446155548096,
+      "learning_rate": 0.0005640500568828214,
+      "loss": 6.3132,
+      "step": 3832
+    },
+    {
+      "epoch": 1.3081911262798636,
+      "grad_norm": 3.190903663635254,
+      "learning_rate": 0.0005639362912400456,
+      "loss": 5.9605,
+      "step": 3833
+    },
+    {
+      "epoch": 1.3085324232081912,
+      "grad_norm": 6.479193210601807,
+      "learning_rate": 0.0005638225255972697,
+      "loss": 5.4088,
+      "step": 3834
+    },
+    {
+      "epoch": 1.3088737201365188,
+      "grad_norm": 3.177168846130371,
+      "learning_rate": 0.0005637087599544938,
+      "loss": 6.6025,
+      "step": 3835
+    },
+    {
+      "epoch": 1.3092150170648464,
+      "grad_norm": 3.4449946880340576,
+      "learning_rate": 0.0005635949943117179,
+      "loss": 5.3262,
+      "step": 3836
+    },
+    {
+      "epoch": 1.309556313993174,
+      "grad_norm": 3.2236359119415283,
+      "learning_rate": 0.0005634812286689421,
+      "loss": 6.2613,
+      "step": 3837
+    },
+    {
+      "epoch": 1.3098976109215017,
+      "grad_norm": 3.60327410697937,
+      "learning_rate": 0.0005633674630261662,
+      "loss": 5.2672,
+      "step": 3838
+    },
+    {
+      "epoch": 1.3102389078498293,
+      "grad_norm": 2.984083890914917,
+      "learning_rate": 0.0005632536973833903,
+      "loss": 6.2554,
+      "step": 3839
+    },
+    {
+      "epoch": 1.310580204778157,
+      "grad_norm": 3.2155168056488037,
+      "learning_rate": 0.0005631399317406144,
+      "loss": 6.54,
+      "step": 3840
+    },
+    {
+      "epoch": 1.3109215017064846,
+      "grad_norm": 3.312378168106079,
+      "learning_rate": 0.0005630261660978384,
+      "loss": 6.4537,
+      "step": 3841
+    },
+    {
+      "epoch": 1.3112627986348122,
+      "grad_norm": 3.047915458679199,
+      "learning_rate": 0.0005629124004550625,
+      "loss": 6.4776,
+      "step": 3842
+    },
+    {
+      "epoch": 1.31160409556314,
+      "grad_norm": 3.2649431228637695,
+      "learning_rate": 0.0005627986348122867,
+      "loss": 6.1978,
+      "step": 3843
+    },
+    {
+      "epoch": 1.3119453924914675,
+      "grad_norm": 3.579108238220215,
+      "learning_rate": 0.0005626848691695108,
+      "loss": 6.1248,
+      "step": 3844
+    },
+    {
+      "epoch": 1.3122866894197953,
+      "grad_norm": 5.931406497955322,
+      "learning_rate": 0.0005625711035267349,
+      "loss": 5.7591,
+      "step": 3845
+    },
+    {
+      "epoch": 1.312627986348123,
+      "grad_norm": 3.2167887687683105,
+      "learning_rate": 0.000562457337883959,
+      "loss": 6.8081,
+      "step": 3846
+    },
+    {
+      "epoch": 1.3129692832764506,
+      "grad_norm": 3.454554557800293,
+      "learning_rate": 0.0005623435722411831,
+      "loss": 6.5649,
+      "step": 3847
+    },
+    {
+      "epoch": 1.3133105802047782,
+      "grad_norm": 7.931427478790283,
+      "learning_rate": 0.0005622298065984072,
+      "loss": 4.5475,
+      "step": 3848
+    },
+    {
+      "epoch": 1.3136518771331058,
+      "grad_norm": 3.3232662677764893,
+      "learning_rate": 0.0005621160409556314,
+      "loss": 6.3771,
+      "step": 3849
+    },
+    {
+      "epoch": 1.3139931740614335,
+      "grad_norm": 3.6749861240386963,
+      "learning_rate": 0.0005620022753128556,
+      "loss": 6.2799,
+      "step": 3850
+    },
+    {
+      "epoch": 1.314334470989761,
+      "grad_norm": 3.1684978008270264,
+      "learning_rate": 0.0005618885096700797,
+      "loss": 6.4631,
+      "step": 3851
+    },
+    {
+      "epoch": 1.3146757679180887,
+      "grad_norm": 3.638622283935547,
+      "learning_rate": 0.0005617747440273038,
+      "loss": 5.7704,
+      "step": 3852
+    },
+    {
+      "epoch": 1.3150170648464163,
+      "grad_norm": 3.2697503566741943,
+      "learning_rate": 0.0005616609783845279,
+      "loss": 5.7991,
+      "step": 3853
+    },
+    {
+      "epoch": 1.315358361774744,
+      "grad_norm": 3.5689799785614014,
+      "learning_rate": 0.000561547212741752,
+      "loss": 5.2592,
+      "step": 3854
+    },
+    {
+      "epoch": 1.3156996587030716,
+      "grad_norm": 3.1521520614624023,
+      "learning_rate": 0.0005614334470989762,
+      "loss": 6.3141,
+      "step": 3855
+    },
+    {
+      "epoch": 1.3160409556313994,
+      "grad_norm": 3.137810468673706,
+      "learning_rate": 0.0005613196814562003,
+      "loss": 6.8293,
+      "step": 3856
+    },
+    {
+      "epoch": 1.3163822525597269,
+      "grad_norm": 3.1368062496185303,
+      "learning_rate": 0.0005612059158134244,
+      "loss": 6.7367,
+      "step": 3857
+    },
+    {
+      "epoch": 1.3167235494880547,
+      "grad_norm": 3.516813278198242,
+      "learning_rate": 0.0005610921501706485,
+      "loss": 6.4741,
+      "step": 3858
+    },
+    {
+      "epoch": 1.3170648464163823,
+      "grad_norm": 5.221506118774414,
+      "learning_rate": 0.0005609783845278726,
+      "loss": 5.8297,
+      "step": 3859
+    },
+    {
+      "epoch": 1.31740614334471,
+      "grad_norm": 3.2400710582733154,
+      "learning_rate": 0.0005608646188850968,
+      "loss": 7.1234,
+      "step": 3860
+    },
+    {
+      "epoch": 1.3177474402730376,
+      "grad_norm": 3.2599074840545654,
+      "learning_rate": 0.0005607508532423208,
+      "loss": 6.3569,
+      "step": 3861
+    },
+    {
+      "epoch": 1.3180887372013652,
+      "grad_norm": 3.250030994415283,
+      "learning_rate": 0.0005606370875995449,
+      "loss": 7.1309,
+      "step": 3862
+    },
+    {
+      "epoch": 1.3184300341296928,
+      "grad_norm": 3.246612071990967,
+      "learning_rate": 0.000560523321956769,
+      "loss": 6.5661,
+      "step": 3863
+    },
+    {
+      "epoch": 1.3187713310580205,
+      "grad_norm": 3.082751989364624,
+      "learning_rate": 0.0005604095563139931,
+      "loss": 6.5423,
+      "step": 3864
+    },
+    {
+      "epoch": 1.319112627986348,
+      "grad_norm": 3.3138341903686523,
+      "learning_rate": 0.0005602957906712172,
+      "loss": 6.3662,
+      "step": 3865
+    },
+    {
+      "epoch": 1.3194539249146757,
+      "grad_norm": 5.791090488433838,
+      "learning_rate": 0.0005601820250284414,
+      "loss": 5.5428,
+      "step": 3866
+    },
+    {
+      "epoch": 1.3197952218430034,
+      "grad_norm": 3.186373710632324,
+      "learning_rate": 0.0005600682593856656,
+      "loss": 6.7147,
+      "step": 3867
+    },
+    {
+      "epoch": 1.320136518771331,
+      "grad_norm": 3.214484930038452,
+      "learning_rate": 0.0005599544937428897,
+      "loss": 6.3017,
+      "step": 3868
+    },
+    {
+      "epoch": 1.3204778156996588,
+      "grad_norm": 3.255305051803589,
+      "learning_rate": 0.0005598407281001138,
+      "loss": 6.3534,
+      "step": 3869
+    },
+    {
+      "epoch": 1.3208191126279862,
+      "grad_norm": 3.318035125732422,
+      "learning_rate": 0.0005597269624573379,
+      "loss": 6.9723,
+      "step": 3870
+    },
+    {
+      "epoch": 1.321160409556314,
+      "grad_norm": 3.18615984916687,
+      "learning_rate": 0.000559613196814562,
+      "loss": 6.6132,
+      "step": 3871
+    },
+    {
+      "epoch": 1.3215017064846417,
+      "grad_norm": 3.249497652053833,
+      "learning_rate": 0.0005594994311717862,
+      "loss": 6.7708,
+      "step": 3872
+    },
+    {
+      "epoch": 1.3218430034129693,
+      "grad_norm": 4.298493385314941,
+      "learning_rate": 0.0005593856655290103,
+      "loss": 6.0935,
+      "step": 3873
+    },
+    {
+      "epoch": 1.322184300341297,
+      "grad_norm": 3.8184046745300293,
+      "learning_rate": 0.0005592718998862344,
+      "loss": 5.617,
+      "step": 3874
+    },
+    {
+      "epoch": 1.3225255972696246,
+      "grad_norm": 3.3545162677764893,
+      "learning_rate": 0.0005591581342434585,
+      "loss": 6.2788,
+      "step": 3875
+    },
+    {
+      "epoch": 1.3228668941979522,
+      "grad_norm": 3.2084295749664307,
+      "learning_rate": 0.0005590443686006826,
+      "loss": 6.5599,
+      "step": 3876
+    },
+    {
+      "epoch": 1.3232081911262799,
+      "grad_norm": 3.0472261905670166,
+      "learning_rate": 0.0005589306029579067,
+      "loss": 6.0838,
+      "step": 3877
+    },
+    {
+      "epoch": 1.3235494880546075,
+      "grad_norm": 3.785743236541748,
+      "learning_rate": 0.0005588168373151309,
+      "loss": 5.699,
+      "step": 3878
+    },
+    {
+      "epoch": 1.323890784982935,
+      "grad_norm": 3.114495038986206,
+      "learning_rate": 0.000558703071672355,
+      "loss": 6.5545,
+      "step": 3879
+    },
+    {
+      "epoch": 1.3242320819112627,
+      "grad_norm": 4.037391662597656,
+      "learning_rate": 0.000558589306029579,
+      "loss": 5.8293,
+      "step": 3880
+    },
+    {
+      "epoch": 1.3245733788395904,
+      "grad_norm": 3.1285529136657715,
+      "learning_rate": 0.0005584755403868031,
+      "loss": 6.5584,
+      "step": 3881
+    },
+    {
+      "epoch": 1.3249146757679182,
+      "grad_norm": 3.292494058609009,
+      "learning_rate": 0.0005583617747440272,
+      "loss": 6.8052,
+      "step": 3882
+    },
+    {
+      "epoch": 1.3252559726962456,
+      "grad_norm": 5.072427272796631,
+      "learning_rate": 0.0005582480091012514,
+      "loss": 5.6256,
+      "step": 3883
+    },
+    {
+      "epoch": 1.3255972696245735,
+      "grad_norm": 3.1769468784332275,
+      "learning_rate": 0.0005581342434584756,
+      "loss": 6.5001,
+      "step": 3884
+    },
+    {
+      "epoch": 1.325938566552901,
+      "grad_norm": 6.59473180770874,
+      "learning_rate": 0.0005580204778156997,
+      "loss": 6.2227,
+      "step": 3885
+    },
+    {
+      "epoch": 1.3262798634812287,
+      "grad_norm": 3.345031261444092,
+      "learning_rate": 0.0005579067121729238,
+      "loss": 6.3236,
+      "step": 3886
+    },
+    {
+      "epoch": 1.3266211604095564,
+      "grad_norm": 3.1531882286071777,
+      "learning_rate": 0.0005577929465301479,
+      "loss": 6.8264,
+      "step": 3887
+    },
+    {
+      "epoch": 1.326962457337884,
+      "grad_norm": 3.17917537689209,
+      "learning_rate": 0.000557679180887372,
+      "loss": 6.032,
+      "step": 3888
+    },
+    {
+      "epoch": 1.3273037542662116,
+      "grad_norm": 3.783651351928711,
+      "learning_rate": 0.0005575654152445962,
+      "loss": 5.9765,
+      "step": 3889
+    },
+    {
+      "epoch": 1.3276450511945392,
+      "grad_norm": 3.515580892562866,
+      "learning_rate": 0.0005574516496018203,
+      "loss": 5.7462,
+      "step": 3890
+    },
+    {
+      "epoch": 1.3279863481228669,
+      "grad_norm": 2.9463393688201904,
+      "learning_rate": 0.0005573378839590444,
+      "loss": 6.4177,
+      "step": 3891
+    },
+    {
+      "epoch": 1.3283276450511945,
+      "grad_norm": 3.109710216522217,
+      "learning_rate": 0.0005572241183162685,
+      "loss": 6.3956,
+      "step": 3892
+    },
+    {
+      "epoch": 1.3286689419795221,
+      "grad_norm": 3.2260215282440186,
+      "learning_rate": 0.0005571103526734926,
+      "loss": 6.6147,
+      "step": 3893
+    },
+    {
+      "epoch": 1.3290102389078498,
+      "grad_norm": 3.0810201168060303,
+      "learning_rate": 0.0005569965870307167,
+      "loss": 6.1781,
+      "step": 3894
+    },
+    {
+      "epoch": 1.3293515358361776,
+      "grad_norm": 3.121894598007202,
+      "learning_rate": 0.0005568828213879409,
+      "loss": 6.7156,
+      "step": 3895
+    },
+    {
+      "epoch": 1.329692832764505,
+      "grad_norm": 3.2574121952056885,
+      "learning_rate": 0.000556769055745165,
+      "loss": 6.3006,
+      "step": 3896
+    },
+    {
+      "epoch": 1.3300341296928329,
+      "grad_norm": 3.250694990158081,
+      "learning_rate": 0.0005566552901023891,
+      "loss": 6.9496,
+      "step": 3897
+    },
+    {
+      "epoch": 1.3303754266211605,
+      "grad_norm": 3.090780258178711,
+      "learning_rate": 0.0005565415244596132,
+      "loss": 6.5721,
+      "step": 3898
+    },
+    {
+      "epoch": 1.3307167235494881,
+      "grad_norm": 3.135728120803833,
+      "learning_rate": 0.0005564277588168373,
+      "loss": 7.0426,
+      "step": 3899
+    },
+    {
+      "epoch": 1.3310580204778157,
+      "grad_norm": 3.099902391433716,
+      "learning_rate": 0.0005563139931740613,
+      "loss": 6.4846,
+      "step": 3900
+    },
+    {
+      "epoch": 1.3313993174061434,
+      "grad_norm": 3.4714698791503906,
+      "learning_rate": 0.0005562002275312856,
+      "loss": 6.1539,
+      "step": 3901
+    },
+    {
+      "epoch": 1.331740614334471,
+      "grad_norm": 3.3556177616119385,
+      "learning_rate": 0.0005560864618885097,
+      "loss": 6.3849,
+      "step": 3902
+    },
+    {
+      "epoch": 1.3320819112627986,
+      "grad_norm": 3.5183370113372803,
+      "learning_rate": 0.0005559726962457338,
+      "loss": 6.1542,
+      "step": 3903
+    },
+    {
+      "epoch": 1.3324232081911263,
+      "grad_norm": 3.0659613609313965,
+      "learning_rate": 0.0005558589306029579,
+      "loss": 6.262,
+      "step": 3904
+    },
+    {
+      "epoch": 1.3327645051194539,
+      "grad_norm": 3.1507010459899902,
+      "learning_rate": 0.000555745164960182,
+      "loss": 6.7591,
+      "step": 3905
+    },
+    {
+      "epoch": 1.3331058020477815,
+      "grad_norm": 3.488461971282959,
+      "learning_rate": 0.0005556313993174062,
+      "loss": 6.2227,
+      "step": 3906
+    },
+    {
+      "epoch": 1.3334470989761091,
+      "grad_norm": 3.122450828552246,
+      "learning_rate": 0.0005555176336746303,
+      "loss": 6.2148,
+      "step": 3907
+    },
+    {
+      "epoch": 1.333788395904437,
+      "grad_norm": 3.227902412414551,
+      "learning_rate": 0.0005554038680318544,
+      "loss": 6.7408,
+      "step": 3908
+    },
+    {
+      "epoch": 1.3341296928327644,
+      "grad_norm": 3.1958107948303223,
+      "learning_rate": 0.0005552901023890785,
+      "loss": 6.4788,
+      "step": 3909
+    },
+    {
+      "epoch": 1.3344709897610922,
+      "grad_norm": 3.2183525562286377,
+      "learning_rate": 0.0005551763367463026,
+      "loss": 6.5644,
+      "step": 3910
+    },
+    {
+      "epoch": 1.3348122866894199,
+      "grad_norm": 3.3385112285614014,
+      "learning_rate": 0.0005550625711035267,
+      "loss": 6.0782,
+      "step": 3911
+    },
+    {
+      "epoch": 1.3351535836177475,
+      "grad_norm": 4.089425086975098,
+      "learning_rate": 0.0005549488054607509,
+      "loss": 5.6889,
+      "step": 3912
+    },
+    {
+      "epoch": 1.3354948805460751,
+      "grad_norm": 3.097321033477783,
+      "learning_rate": 0.000554835039817975,
+      "loss": 6.336,
+      "step": 3913
+    },
+    {
+      "epoch": 1.3358361774744028,
+      "grad_norm": 5.853010654449463,
+      "learning_rate": 0.0005547212741751991,
+      "loss": 4.3251,
+      "step": 3914
+    },
+    {
+      "epoch": 1.3361774744027304,
+      "grad_norm": 3.582655191421509,
+      "learning_rate": 0.0005546075085324232,
+      "loss": 5.5901,
+      "step": 3915
+    },
+    {
+      "epoch": 1.336518771331058,
+      "grad_norm": 3.233471155166626,
+      "learning_rate": 0.0005544937428896473,
+      "loss": 6.5578,
+      "step": 3916
+    },
+    {
+      "epoch": 1.3368600682593856,
+      "grad_norm": 3.3025286197662354,
+      "learning_rate": 0.0005543799772468715,
+      "loss": 6.0288,
+      "step": 3917
+    },
+    {
+      "epoch": 1.3372013651877133,
+      "grad_norm": 3.4599759578704834,
+      "learning_rate": 0.0005542662116040957,
+      "loss": 6.4112,
+      "step": 3918
+    },
+    {
+      "epoch": 1.337542662116041,
+      "grad_norm": 3.0111966133117676,
+      "learning_rate": 0.0005541524459613197,
+      "loss": 6.3284,
+      "step": 3919
+    },
+    {
+      "epoch": 1.3378839590443685,
+      "grad_norm": 4.267691135406494,
+      "learning_rate": 0.0005540386803185438,
+      "loss": 6.3674,
+      "step": 3920
+    },
+    {
+      "epoch": 1.3382252559726964,
+      "grad_norm": 3.2080326080322266,
+      "learning_rate": 0.0005539249146757679,
+      "loss": 6.5919,
+      "step": 3921
+    },
+    {
+      "epoch": 1.3385665529010238,
+      "grad_norm": 3.23378849029541,
+      "learning_rate": 0.000553811149032992,
+      "loss": 6.1681,
+      "step": 3922
+    },
+    {
+      "epoch": 1.3389078498293516,
+      "grad_norm": 3.0529868602752686,
+      "learning_rate": 0.0005536973833902162,
+      "loss": 6.6733,
+      "step": 3923
+    },
+    {
+      "epoch": 1.3392491467576793,
+      "grad_norm": 4.308516502380371,
+      "learning_rate": 0.0005535836177474403,
+      "loss": 5.9069,
+      "step": 3924
+    },
+    {
+      "epoch": 1.3395904436860069,
+      "grad_norm": 6.184211254119873,
+      "learning_rate": 0.0005534698521046644,
+      "loss": 5.3009,
+      "step": 3925
+    },
+    {
+      "epoch": 1.3399317406143345,
+      "grad_norm": 3.4031319618225098,
+      "learning_rate": 0.0005533560864618885,
+      "loss": 6.694,
+      "step": 3926
+    },
+    {
+      "epoch": 1.3402730375426621,
+      "grad_norm": 3.15828275680542,
+      "learning_rate": 0.0005532423208191126,
+      "loss": 6.8263,
+      "step": 3927
+    },
+    {
+      "epoch": 1.3406143344709898,
+      "grad_norm": 3.1163811683654785,
+      "learning_rate": 0.0005531285551763367,
+      "loss": 6.3808,
+      "step": 3928
+    },
+    {
+      "epoch": 1.3409556313993174,
+      "grad_norm": 3.1072165966033936,
+      "learning_rate": 0.0005530147895335609,
+      "loss": 6.4251,
+      "step": 3929
+    },
+    {
+      "epoch": 1.341296928327645,
+      "grad_norm": 3.2711892127990723,
+      "learning_rate": 0.000552901023890785,
+      "loss": 6.1466,
+      "step": 3930
+    },
+    {
+      "epoch": 1.3416382252559726,
+      "grad_norm": 3.1679635047912598,
+      "learning_rate": 0.0005527872582480091,
+      "loss": 6.3643,
+      "step": 3931
+    },
+    {
+      "epoch": 1.3419795221843003,
+      "grad_norm": 3.1142637729644775,
+      "learning_rate": 0.0005526734926052332,
+      "loss": 6.8845,
+      "step": 3932
+    },
+    {
+      "epoch": 1.342320819112628,
+      "grad_norm": 3.082751989364624,
+      "learning_rate": 0.0005525597269624573,
+      "loss": 6.1942,
+      "step": 3933
+    },
+    {
+      "epoch": 1.3426621160409558,
+      "grad_norm": 3.139395236968994,
+      "learning_rate": 0.0005524459613196815,
+      "loss": 6.6958,
+      "step": 3934
+    },
+    {
+      "epoch": 1.3430034129692832,
+      "grad_norm": 3.3839261531829834,
+      "learning_rate": 0.0005523321956769057,
+      "loss": 5.6164,
+      "step": 3935
+    },
+    {
+      "epoch": 1.343344709897611,
+      "grad_norm": 3.2381277084350586,
+      "learning_rate": 0.0005522184300341298,
+      "loss": 6.8543,
+      "step": 3936
+    },
+    {
+      "epoch": 1.3436860068259386,
+      "grad_norm": 3.236882209777832,
+      "learning_rate": 0.0005521046643913539,
+      "loss": 6.2873,
+      "step": 3937
+    },
+    {
+      "epoch": 1.3440273037542663,
+      "grad_norm": 3.085549831390381,
+      "learning_rate": 0.0005519908987485779,
+      "loss": 6.6844,
+      "step": 3938
+    },
+    {
+      "epoch": 1.344368600682594,
+      "grad_norm": 3.2600576877593994,
+      "learning_rate": 0.000551877133105802,
+      "loss": 6.4399,
+      "step": 3939
+    },
+    {
+      "epoch": 1.3447098976109215,
+      "grad_norm": 3.097806215286255,
+      "learning_rate": 0.0005517633674630261,
+      "loss": 6.825,
+      "step": 3940
+    },
+    {
+      "epoch": 1.3450511945392492,
+      "grad_norm": 3.1946775913238525,
+      "learning_rate": 0.0005516496018202503,
+      "loss": 6.6171,
+      "step": 3941
+    },
+    {
+      "epoch": 1.3453924914675768,
+      "grad_norm": 3.3030920028686523,
+      "learning_rate": 0.0005515358361774744,
+      "loss": 6.3477,
+      "step": 3942
+    },
+    {
+      "epoch": 1.3457337883959044,
+      "grad_norm": 3.1355345249176025,
+      "learning_rate": 0.0005514220705346985,
+      "loss": 6.8327,
+      "step": 3943
+    },
+    {
+      "epoch": 1.346075085324232,
+      "grad_norm": 3.1044278144836426,
+      "learning_rate": 0.0005513083048919226,
+      "loss": 6.3775,
+      "step": 3944
+    },
+    {
+      "epoch": 1.3464163822525597,
+      "grad_norm": 3.12194561958313,
+      "learning_rate": 0.0005511945392491467,
+      "loss": 6.2563,
+      "step": 3945
+    },
+    {
+      "epoch": 1.3467576791808873,
+      "grad_norm": 3.1160433292388916,
+      "learning_rate": 0.0005510807736063709,
+      "loss": 6.0824,
+      "step": 3946
+    },
+    {
+      "epoch": 1.3470989761092151,
+      "grad_norm": 3.5935933589935303,
+      "learning_rate": 0.000550967007963595,
+      "loss": 3.7834,
+      "step": 3947
+    },
+    {
+      "epoch": 1.3474402730375425,
+      "grad_norm": 3.4486262798309326,
+      "learning_rate": 0.0005508532423208191,
+      "loss": 6.0705,
+      "step": 3948
+    },
+    {
+      "epoch": 1.3477815699658704,
+      "grad_norm": 3.192941427230835,
+      "learning_rate": 0.0005507394766780432,
+      "loss": 5.8078,
+      "step": 3949
+    },
+    {
+      "epoch": 1.348122866894198,
+      "grad_norm": 8.985308647155762,
+      "learning_rate": 0.0005506257110352673,
+      "loss": 5.2075,
+      "step": 3950
+    },
+    {
+      "epoch": 1.3484641638225257,
+      "grad_norm": 4.667975902557373,
+      "learning_rate": 0.0005505119453924915,
+      "loss": 6.0374,
+      "step": 3951
+    },
+    {
+      "epoch": 1.3488054607508533,
+      "grad_norm": 3.4033608436584473,
+      "learning_rate": 0.0005503981797497157,
+      "loss": 6.1368,
+      "step": 3952
+    },
+    {
+      "epoch": 1.349146757679181,
+      "grad_norm": 3.455477237701416,
+      "learning_rate": 0.0005502844141069398,
+      "loss": 6.2675,
+      "step": 3953
+    },
+    {
+      "epoch": 1.3494880546075085,
+      "grad_norm": 3.2633755207061768,
+      "learning_rate": 0.0005501706484641639,
+      "loss": 6.7096,
+      "step": 3954
+    },
+    {
+      "epoch": 1.3498293515358362,
+      "grad_norm": 4.205435752868652,
+      "learning_rate": 0.000550056882821388,
+      "loss": 5.0339,
+      "step": 3955
+    },
+    {
+      "epoch": 1.3501706484641638,
+      "grad_norm": 3.1582601070404053,
+      "learning_rate": 0.0005499431171786121,
+      "loss": 7.1195,
+      "step": 3956
+    },
+    {
+      "epoch": 1.3505119453924914,
+      "grad_norm": 3.342383623123169,
+      "learning_rate": 0.0005498293515358362,
+      "loss": 5.5482,
+      "step": 3957
+    },
+    {
+      "epoch": 1.350853242320819,
+      "grad_norm": 3.2618634700775146,
+      "learning_rate": 0.0005497155858930603,
+      "loss": 6.5156,
+      "step": 3958
+    },
+    {
+      "epoch": 1.3511945392491467,
+      "grad_norm": 3.0584144592285156,
+      "learning_rate": 0.0005496018202502844,
+      "loss": 6.5958,
+      "step": 3959
+    },
+    {
+      "epoch": 1.3515358361774745,
+      "grad_norm": 3.212893009185791,
+      "learning_rate": 0.0005494880546075085,
+      "loss": 6.7449,
+      "step": 3960
+    },
+    {
+      "epoch": 1.351877133105802,
+      "grad_norm": 3.140817403793335,
+      "learning_rate": 0.0005493742889647326,
+      "loss": 6.134,
+      "step": 3961
+    },
+    {
+      "epoch": 1.3522184300341298,
+      "grad_norm": 3.27764630317688,
+      "learning_rate": 0.0005492605233219567,
+      "loss": 5.9015,
+      "step": 3962
+    },
+    {
+      "epoch": 1.3525597269624574,
+      "grad_norm": 3.1260316371917725,
+      "learning_rate": 0.0005491467576791809,
+      "loss": 6.9168,
+      "step": 3963
+    },
+    {
+      "epoch": 1.352901023890785,
+      "grad_norm": 4.186235427856445,
+      "learning_rate": 0.000549032992036405,
+      "loss": 5.7906,
+      "step": 3964
+    },
+    {
+      "epoch": 1.3532423208191127,
+      "grad_norm": 3.5168609619140625,
+      "learning_rate": 0.0005489192263936291,
+      "loss": 6.4136,
+      "step": 3965
+    },
+    {
+      "epoch": 1.3535836177474403,
+      "grad_norm": 3.287663698196411,
+      "learning_rate": 0.0005488054607508532,
+      "loss": 6.072,
+      "step": 3966
+    },
+    {
+      "epoch": 1.353924914675768,
+      "grad_norm": 3.4218294620513916,
+      "learning_rate": 0.0005486916951080773,
+      "loss": 5.65,
+      "step": 3967
+    },
+    {
+      "epoch": 1.3542662116040955,
+      "grad_norm": 3.3212225437164307,
+      "learning_rate": 0.0005485779294653015,
+      "loss": 6.4668,
+      "step": 3968
+    },
+    {
+      "epoch": 1.3546075085324232,
+      "grad_norm": 3.127525568008423,
+      "learning_rate": 0.0005484641638225257,
+      "loss": 6.4388,
+      "step": 3969
+    },
+    {
+      "epoch": 1.3549488054607508,
+      "grad_norm": 3.166355848312378,
+      "learning_rate": 0.0005483503981797498,
+      "loss": 6.3717,
+      "step": 3970
+    },
+    {
+      "epoch": 1.3552901023890784,
+      "grad_norm": 3.090057134628296,
+      "learning_rate": 0.0005482366325369739,
+      "loss": 5.9435,
+      "step": 3971
+    },
+    {
+      "epoch": 1.355631399317406,
+      "grad_norm": 3.0684850215911865,
+      "learning_rate": 0.000548122866894198,
+      "loss": 7.0271,
+      "step": 3972
+    },
+    {
+      "epoch": 1.355972696245734,
+      "grad_norm": 3.6041576862335205,
+      "learning_rate": 0.0005480091012514221,
+      "loss": 6.0391,
+      "step": 3973
+    },
+    {
+      "epoch": 1.3563139931740613,
+      "grad_norm": 5.369800567626953,
+      "learning_rate": 0.0005478953356086462,
+      "loss": 6.1618,
+      "step": 3974
+    },
+    {
+      "epoch": 1.3566552901023892,
+      "grad_norm": 3.1870033740997314,
+      "learning_rate": 0.0005477815699658704,
+      "loss": 6.782,
+      "step": 3975
+    },
+    {
+      "epoch": 1.3569965870307168,
+      "grad_norm": 3.168116331100464,
+      "learning_rate": 0.0005476678043230945,
+      "loss": 6.3682,
+      "step": 3976
+    },
+    {
+      "epoch": 1.3573378839590444,
+      "grad_norm": 3.2741923332214355,
+      "learning_rate": 0.0005475540386803185,
+      "loss": 6.0431,
+      "step": 3977
+    },
+    {
+      "epoch": 1.357679180887372,
+      "grad_norm": 3.190312147140503,
+      "learning_rate": 0.0005474402730375426,
+      "loss": 6.2224,
+      "step": 3978
+    },
+    {
+      "epoch": 1.3580204778156997,
+      "grad_norm": 3.3220231533050537,
+      "learning_rate": 0.0005473265073947667,
+      "loss": 6.61,
+      "step": 3979
+    },
+    {
+      "epoch": 1.3583617747440273,
+      "grad_norm": 3.0520358085632324,
+      "learning_rate": 0.0005472127417519908,
+      "loss": 6.6153,
+      "step": 3980
+    },
+    {
+      "epoch": 1.358703071672355,
+      "grad_norm": 3.942625045776367,
+      "learning_rate": 0.000547098976109215,
+      "loss": 4.871,
+      "step": 3981
+    },
+    {
+      "epoch": 1.3590443686006826,
+      "grad_norm": 3.148261785507202,
+      "learning_rate": 0.0005469852104664391,
+      "loss": 6.5981,
+      "step": 3982
+    },
+    {
+      "epoch": 1.3593856655290102,
+      "grad_norm": 3.1184489727020264,
+      "learning_rate": 0.0005468714448236632,
+      "loss": 6.0445,
+      "step": 3983
+    },
+    {
+      "epoch": 1.3597269624573378,
+      "grad_norm": 3.2603631019592285,
+      "learning_rate": 0.0005467576791808873,
+      "loss": 6.847,
+      "step": 3984
+    },
+    {
+      "epoch": 1.3600682593856654,
+      "grad_norm": 3.4152493476867676,
+      "learning_rate": 0.0005466439135381115,
+      "loss": 5.5912,
+      "step": 3985
+    },
+    {
+      "epoch": 1.3604095563139933,
+      "grad_norm": 3.3140194416046143,
+      "learning_rate": 0.0005465301478953357,
+      "loss": 6.3923,
+      "step": 3986
+    },
+    {
+      "epoch": 1.3607508532423207,
+      "grad_norm": 3.526538610458374,
+      "learning_rate": 0.0005464163822525598,
+      "loss": 5.259,
+      "step": 3987
+    },
+    {
+      "epoch": 1.3610921501706486,
+      "grad_norm": 3.1631855964660645,
+      "learning_rate": 0.0005463026166097839,
+      "loss": 6.5663,
+      "step": 3988
+    },
+    {
+      "epoch": 1.3614334470989762,
+      "grad_norm": 3.11613392829895,
+      "learning_rate": 0.000546188850967008,
+      "loss": 6.3917,
+      "step": 3989
+    },
+    {
+      "epoch": 1.3617747440273038,
+      "grad_norm": 3.1377980709075928,
+      "learning_rate": 0.0005460750853242321,
+      "loss": 6.2215,
+      "step": 3990
+    },
+    {
+      "epoch": 1.3621160409556314,
+      "grad_norm": 2.9950690269470215,
+      "learning_rate": 0.0005459613196814562,
+      "loss": 6.8778,
+      "step": 3991
+    },
+    {
+      "epoch": 1.362457337883959,
+      "grad_norm": 3.1760122776031494,
+      "learning_rate": 0.0005458475540386804,
+      "loss": 6.2066,
+      "step": 3992
+    },
+    {
+      "epoch": 1.3627986348122867,
+      "grad_norm": 3.045168399810791,
+      "learning_rate": 0.0005457337883959045,
+      "loss": 6.3082,
+      "step": 3993
+    },
+    {
+      "epoch": 1.3631399317406143,
+      "grad_norm": 3.054022789001465,
+      "learning_rate": 0.0005456200227531286,
+      "loss": 6.0026,
+      "step": 3994
+    },
+    {
+      "epoch": 1.363481228668942,
+      "grad_norm": 3.007483959197998,
+      "learning_rate": 0.0005455062571103527,
+      "loss": 6.052,
+      "step": 3995
+    },
+    {
+      "epoch": 1.3638225255972696,
+      "grad_norm": 3.831319570541382,
+      "learning_rate": 0.0005453924914675768,
+      "loss": 5.1814,
+      "step": 3996
+    },
+    {
+      "epoch": 1.3641638225255972,
+      "grad_norm": 3.610045909881592,
+      "learning_rate": 0.0005452787258248008,
+      "loss": 6.2811,
+      "step": 3997
+    },
+    {
+      "epoch": 1.3645051194539248,
+      "grad_norm": 3.2371370792388916,
+      "learning_rate": 0.000545164960182025,
+      "loss": 6.8801,
+      "step": 3998
+    },
+    {
+      "epoch": 1.3648464163822527,
+      "grad_norm": 3.0516843795776367,
+      "learning_rate": 0.0005450511945392491,
+      "loss": 6.3365,
+      "step": 3999
+    },
+    {
+      "epoch": 1.36518771331058,
+      "grad_norm": 3.2358148097991943,
+      "learning_rate": 0.0005449374288964732,
+      "loss": 6.708,
+      "step": 4000
+    },
+    {
+      "epoch": 1.365529010238908,
+      "grad_norm": 3.3484854698181152,
+      "learning_rate": 0.0005448236632536974,
+      "loss": 5.9241,
+      "step": 4001
+    },
+    {
+      "epoch": 1.3658703071672356,
+      "grad_norm": 3.215952157974243,
+      "learning_rate": 0.0005447098976109215,
+      "loss": 6.2136,
+      "step": 4002
+    },
+    {
+      "epoch": 1.3662116040955632,
+      "grad_norm": 3.1921606063842773,
+      "learning_rate": 0.0005445961319681456,
+      "loss": 6.4693,
+      "step": 4003
+    },
+    {
+      "epoch": 1.3665529010238908,
+      "grad_norm": 3.2230448722839355,
+      "learning_rate": 0.0005444823663253698,
+      "loss": 6.3591,
+      "step": 4004
+    },
+    {
+      "epoch": 1.3668941979522184,
+      "grad_norm": 3.30001163482666,
+      "learning_rate": 0.0005443686006825939,
+      "loss": 7.2393,
+      "step": 4005
+    },
+    {
+      "epoch": 1.367235494880546,
+      "grad_norm": 3.1565656661987305,
+      "learning_rate": 0.000544254835039818,
+      "loss": 6.7251,
+      "step": 4006
+    },
+    {
+      "epoch": 1.3675767918088737,
+      "grad_norm": 2.9923689365386963,
+      "learning_rate": 0.0005441410693970421,
+      "loss": 6.4636,
+      "step": 4007
+    },
+    {
+      "epoch": 1.3679180887372013,
+      "grad_norm": 3.189656972885132,
+      "learning_rate": 0.0005440273037542662,
+      "loss": 6.586,
+      "step": 4008
+    },
+    {
+      "epoch": 1.368259385665529,
+      "grad_norm": 4.6809234619140625,
+      "learning_rate": 0.0005439135381114904,
+      "loss": 5.4901,
+      "step": 4009
+    },
+    {
+      "epoch": 1.3686006825938566,
+      "grad_norm": 3.192366600036621,
+      "learning_rate": 0.0005437997724687145,
+      "loss": 6.0856,
+      "step": 4010
+    },
+    {
+      "epoch": 1.3689419795221842,
+      "grad_norm": 3.0873725414276123,
+      "learning_rate": 0.0005436860068259386,
+      "loss": 6.8565,
+      "step": 4011
+    },
+    {
+      "epoch": 1.369283276450512,
+      "grad_norm": 3.1170427799224854,
+      "learning_rate": 0.0005435722411831627,
+      "loss": 6.371,
+      "step": 4012
+    },
+    {
+      "epoch": 1.3696245733788395,
+      "grad_norm": 3.2047770023345947,
+      "learning_rate": 0.0005434584755403868,
+      "loss": 6.8304,
+      "step": 4013
+    },
+    {
+      "epoch": 1.3699658703071673,
+      "grad_norm": 3.132357120513916,
+      "learning_rate": 0.0005433447098976109,
+      "loss": 6.4195,
+      "step": 4014
+    },
+    {
+      "epoch": 1.370307167235495,
+      "grad_norm": 3.173907518386841,
+      "learning_rate": 0.0005432309442548351,
+      "loss": 6.4185,
+      "step": 4015
+    },
+    {
+      "epoch": 1.3706484641638226,
+      "grad_norm": 5.250078201293945,
+      "learning_rate": 0.0005431171786120591,
+      "loss": 5.0679,
+      "step": 4016
+    },
+    {
+      "epoch": 1.3709897610921502,
+      "grad_norm": 3.5316922664642334,
+      "learning_rate": 0.0005430034129692832,
+      "loss": 6.0273,
+      "step": 4017
+    },
+    {
+      "epoch": 1.3713310580204778,
+      "grad_norm": 3.2254114151000977,
+      "learning_rate": 0.0005428896473265074,
+      "loss": 6.6384,
+      "step": 4018
+    },
+    {
+      "epoch": 1.3716723549488055,
+      "grad_norm": 3.242680788040161,
+      "learning_rate": 0.0005427758816837315,
+      "loss": 6.5512,
+      "step": 4019
+    },
+    {
+      "epoch": 1.372013651877133,
+      "grad_norm": 3.2910420894622803,
+      "learning_rate": 0.0005426621160409556,
+      "loss": 6.7562,
+      "step": 4020
+    },
+    {
+      "epoch": 1.3723549488054607,
+      "grad_norm": 3.184915065765381,
+      "learning_rate": 0.0005425483503981798,
+      "loss": 6.5688,
+      "step": 4021
+    },
+    {
+      "epoch": 1.3726962457337883,
+      "grad_norm": 3.098815679550171,
+      "learning_rate": 0.0005424345847554039,
+      "loss": 6.6371,
+      "step": 4022
+    },
+    {
+      "epoch": 1.373037542662116,
+      "grad_norm": 3.283712863922119,
+      "learning_rate": 0.000542320819112628,
+      "loss": 6.0917,
+      "step": 4023
+    },
+    {
+      "epoch": 1.3733788395904436,
+      "grad_norm": 3.190316677093506,
+      "learning_rate": 0.0005422070534698521,
+      "loss": 6.1911,
+      "step": 4024
+    },
+    {
+      "epoch": 1.3737201365187715,
+      "grad_norm": 3.122403621673584,
+      "learning_rate": 0.0005420932878270762,
+      "loss": 6.3405,
+      "step": 4025
+    },
+    {
+      "epoch": 1.3740614334470989,
+      "grad_norm": 3.1362032890319824,
+      "learning_rate": 0.0005419795221843004,
+      "loss": 6.6867,
+      "step": 4026
+    },
+    {
+      "epoch": 1.3744027303754267,
+      "grad_norm": 3.2245562076568604,
+      "learning_rate": 0.0005418657565415245,
+      "loss": 6.2402,
+      "step": 4027
+    },
+    {
+      "epoch": 1.3747440273037543,
+      "grad_norm": 3.087644338607788,
+      "learning_rate": 0.0005417519908987486,
+      "loss": 6.16,
+      "step": 4028
+    },
+    {
+      "epoch": 1.375085324232082,
+      "grad_norm": 3.1691482067108154,
+      "learning_rate": 0.0005416382252559727,
+      "loss": 6.5541,
+      "step": 4029
+    },
+    {
+      "epoch": 1.3754266211604096,
+      "grad_norm": 3.053196668624878,
+      "learning_rate": 0.0005415244596131968,
+      "loss": 6.7324,
+      "step": 4030
+    },
+    {
+      "epoch": 1.3757679180887372,
+      "grad_norm": 4.789836406707764,
+      "learning_rate": 0.0005414106939704209,
+      "loss": 5.2476,
+      "step": 4031
+    },
+    {
+      "epoch": 1.3761092150170648,
+      "grad_norm": 3.1950435638427734,
+      "learning_rate": 0.0005412969283276451,
+      "loss": 6.7076,
+      "step": 4032
+    },
+    {
+      "epoch": 1.3764505119453925,
+      "grad_norm": 3.2419111728668213,
+      "learning_rate": 0.0005411831626848693,
+      "loss": 6.1823,
+      "step": 4033
+    },
+    {
+      "epoch": 1.37679180887372,
+      "grad_norm": 7.173503398895264,
+      "learning_rate": 0.0005410693970420934,
+      "loss": 4.7265,
+      "step": 4034
+    },
+    {
+      "epoch": 1.3771331058020477,
+      "grad_norm": 3.251122236251831,
+      "learning_rate": 0.0005409556313993175,
+      "loss": 6.432,
+      "step": 4035
+    },
+    {
+      "epoch": 1.3774744027303754,
+      "grad_norm": 3.205535888671875,
+      "learning_rate": 0.0005408418657565415,
+      "loss": 6.6821,
+      "step": 4036
+    },
+    {
+      "epoch": 1.377815699658703,
+      "grad_norm": 3.3682541847229004,
+      "learning_rate": 0.0005407281001137656,
+      "loss": 6.1631,
+      "step": 4037
+    },
+    {
+      "epoch": 1.3781569965870308,
+      "grad_norm": 3.0801639556884766,
+      "learning_rate": 0.0005406143344709898,
+      "loss": 6.1151,
+      "step": 4038
+    },
+    {
+      "epoch": 1.3784982935153582,
+      "grad_norm": 3.063805341720581,
+      "learning_rate": 0.0005405005688282139,
+      "loss": 6.5842,
+      "step": 4039
+    },
+    {
+      "epoch": 1.378839590443686,
+      "grad_norm": 2.9392287731170654,
+      "learning_rate": 0.000540386803185438,
+      "loss": 6.3864,
+      "step": 4040
+    },
+    {
+      "epoch": 1.3791808873720137,
+      "grad_norm": 3.018649101257324,
+      "learning_rate": 0.0005402730375426621,
+      "loss": 6.542,
+      "step": 4041
+    },
+    {
+      "epoch": 1.3795221843003413,
+      "grad_norm": 4.661662578582764,
+      "learning_rate": 0.0005401592718998862,
+      "loss": 5.7853,
+      "step": 4042
+    },
+    {
+      "epoch": 1.379863481228669,
+      "grad_norm": 3.264328956604004,
+      "learning_rate": 0.0005400455062571103,
+      "loss": 6.3248,
+      "step": 4043
+    },
+    {
+      "epoch": 1.3802047781569966,
+      "grad_norm": 3.175499200820923,
+      "learning_rate": 0.0005399317406143345,
+      "loss": 6.5147,
+      "step": 4044
+    },
+    {
+      "epoch": 1.3805460750853242,
+      "grad_norm": 3.1350579261779785,
+      "learning_rate": 0.0005398179749715586,
+      "loss": 6.9947,
+      "step": 4045
+    },
+    {
+      "epoch": 1.3808873720136519,
+      "grad_norm": 3.96781063079834,
+      "learning_rate": 0.0005397042093287827,
+      "loss": 5.874,
+      "step": 4046
+    },
+    {
+      "epoch": 1.3812286689419795,
+      "grad_norm": 3.2071373462677,
+      "learning_rate": 0.0005395904436860068,
+      "loss": 6.668,
+      "step": 4047
+    },
+    {
+      "epoch": 1.3815699658703071,
+      "grad_norm": 3.1385679244995117,
+      "learning_rate": 0.0005394766780432309,
+      "loss": 6.7825,
+      "step": 4048
+    },
+    {
+      "epoch": 1.3819112627986347,
+      "grad_norm": 3.573509931564331,
+      "learning_rate": 0.0005393629124004551,
+      "loss": 5.5573,
+      "step": 4049
+    },
+    {
+      "epoch": 1.3822525597269624,
+      "grad_norm": 3.066441059112549,
+      "learning_rate": 0.0005392491467576793,
+      "loss": 6.6349,
+      "step": 4050
+    },
+    {
+      "epoch": 1.3825938566552902,
+      "grad_norm": 2.976698160171509,
+      "learning_rate": 0.0005391353811149034,
+      "loss": 6.6275,
+      "step": 4051
+    },
+    {
+      "epoch": 1.3829351535836176,
+      "grad_norm": 3.050534963607788,
+      "learning_rate": 0.0005390216154721275,
+      "loss": 6.8268,
+      "step": 4052
+    },
+    {
+      "epoch": 1.3832764505119455,
+      "grad_norm": 3.335890769958496,
+      "learning_rate": 0.0005389078498293516,
+      "loss": 6.1537,
+      "step": 4053
+    },
+    {
+      "epoch": 1.383617747440273,
+      "grad_norm": 3.080980062484741,
+      "learning_rate": 0.0005387940841865757,
+      "loss": 6.7967,
+      "step": 4054
+    },
+    {
+      "epoch": 1.3839590443686007,
+      "grad_norm": 3.0135996341705322,
+      "learning_rate": 0.0005386803185437998,
+      "loss": 7.1126,
+      "step": 4055
+    },
+    {
+      "epoch": 1.3843003412969284,
+      "grad_norm": 3.079207420349121,
+      "learning_rate": 0.0005385665529010239,
+      "loss": 6.7536,
+      "step": 4056
+    },
+    {
+      "epoch": 1.384641638225256,
+      "grad_norm": 3.1604158878326416,
+      "learning_rate": 0.000538452787258248,
+      "loss": 6.2925,
+      "step": 4057
+    },
+    {
+      "epoch": 1.3849829351535836,
+      "grad_norm": 3.069831371307373,
+      "learning_rate": 0.0005383390216154721,
+      "loss": 6.494,
+      "step": 4058
+    },
+    {
+      "epoch": 1.3853242320819112,
+      "grad_norm": 3.445303201675415,
+      "learning_rate": 0.0005382252559726962,
+      "loss": 6.0185,
+      "step": 4059
+    },
+    {
+      "epoch": 1.3856655290102389,
+      "grad_norm": 3.107506513595581,
+      "learning_rate": 0.0005381114903299203,
+      "loss": 6.3259,
+      "step": 4060
+    },
+    {
+      "epoch": 1.3860068259385665,
+      "grad_norm": 3.251415729522705,
+      "learning_rate": 0.0005379977246871445,
+      "loss": 6.1274,
+      "step": 4061
+    },
+    {
+      "epoch": 1.3863481228668941,
+      "grad_norm": 3.4389657974243164,
+      "learning_rate": 0.0005378839590443686,
+      "loss": 5.9044,
+      "step": 4062
+    },
+    {
+      "epoch": 1.3866894197952218,
+      "grad_norm": 2.960808515548706,
+      "learning_rate": 0.0005377701934015927,
+      "loss": 6.5314,
+      "step": 4063
+    },
+    {
+      "epoch": 1.3870307167235496,
+      "grad_norm": 3.1747775077819824,
+      "learning_rate": 0.0005376564277588168,
+      "loss": 6.3545,
+      "step": 4064
+    },
+    {
+      "epoch": 1.387372013651877,
+      "grad_norm": 3.289687156677246,
+      "learning_rate": 0.0005375426621160409,
+      "loss": 6.4242,
+      "step": 4065
+    },
+    {
+      "epoch": 1.3877133105802049,
+      "grad_norm": 3.1048786640167236,
+      "learning_rate": 0.0005374288964732651,
+      "loss": 6.6646,
+      "step": 4066
+    },
+    {
+      "epoch": 1.3880546075085325,
+      "grad_norm": 3.1970558166503906,
+      "learning_rate": 0.0005373151308304893,
+      "loss": 6.5832,
+      "step": 4067
+    },
+    {
+      "epoch": 1.3883959044368601,
+      "grad_norm": 4.1045637130737305,
+      "learning_rate": 0.0005372013651877134,
+      "loss": 5.232,
+      "step": 4068
+    },
+    {
+      "epoch": 1.3887372013651877,
+      "grad_norm": 3.392815589904785,
+      "learning_rate": 0.0005370875995449375,
+      "loss": 5.7788,
+      "step": 4069
+    },
+    {
+      "epoch": 1.3890784982935154,
+      "grad_norm": 3.2663235664367676,
+      "learning_rate": 0.0005369738339021616,
+      "loss": 6.5488,
+      "step": 4070
+    },
+    {
+      "epoch": 1.389419795221843,
+      "grad_norm": 4.128077030181885,
+      "learning_rate": 0.0005368600682593857,
+      "loss": 5.8966,
+      "step": 4071
+    },
+    {
+      "epoch": 1.3897610921501706,
+      "grad_norm": 3.2271320819854736,
+      "learning_rate": 0.0005367463026166099,
+      "loss": 6.8354,
+      "step": 4072
+    },
+    {
+      "epoch": 1.3901023890784983,
+      "grad_norm": 3.334517478942871,
+      "learning_rate": 0.000536632536973834,
+      "loss": 4.8601,
+      "step": 4073
+    },
+    {
+      "epoch": 1.3904436860068259,
+      "grad_norm": 3.136608362197876,
+      "learning_rate": 0.0005365187713310581,
+      "loss": 6.6863,
+      "step": 4074
+    },
+    {
+      "epoch": 1.3907849829351535,
+      "grad_norm": 3.104353904724121,
+      "learning_rate": 0.0005364050056882821,
+      "loss": 6.5178,
+      "step": 4075
+    },
+    {
+      "epoch": 1.3911262798634811,
+      "grad_norm": 3.1452524662017822,
+      "learning_rate": 0.0005362912400455062,
+      "loss": 6.2725,
+      "step": 4076
+    },
+    {
+      "epoch": 1.391467576791809,
+      "grad_norm": 3.041658639907837,
+      "learning_rate": 0.0005361774744027303,
+      "loss": 6.231,
+      "step": 4077
+    },
+    {
+      "epoch": 1.3918088737201364,
+      "grad_norm": 3.1051037311553955,
+      "learning_rate": 0.0005360637087599545,
+      "loss": 6.0589,
+      "step": 4078
+    },
+    {
+      "epoch": 1.3921501706484642,
+      "grad_norm": 3.173142433166504,
+      "learning_rate": 0.0005359499431171786,
+      "loss": 6.2059,
+      "step": 4079
+    },
+    {
+      "epoch": 1.3924914675767919,
+      "grad_norm": 3.0535824298858643,
+      "learning_rate": 0.0005358361774744027,
+      "loss": 6.2619,
+      "step": 4080
+    },
+    {
+      "epoch": 1.3928327645051195,
+      "grad_norm": 3.0652310848236084,
+      "learning_rate": 0.0005357224118316268,
+      "loss": 6.6454,
+      "step": 4081
+    },
+    {
+      "epoch": 1.3931740614334471,
+      "grad_norm": 3.147127151489258,
+      "learning_rate": 0.0005356086461888509,
+      "loss": 6.4706,
+      "step": 4082
+    },
+    {
+      "epoch": 1.3935153583617748,
+      "grad_norm": 3.9887466430664062,
+      "learning_rate": 0.000535494880546075,
+      "loss": 5.9159,
+      "step": 4083
+    },
+    {
+      "epoch": 1.3938566552901024,
+      "grad_norm": 3.139221429824829,
+      "learning_rate": 0.0005353811149032993,
+      "loss": 6.4576,
+      "step": 4084
+    },
+    {
+      "epoch": 1.39419795221843,
+      "grad_norm": 3.5257785320281982,
+      "learning_rate": 0.0005352673492605234,
+      "loss": 5.7365,
+      "step": 4085
+    },
+    {
+      "epoch": 1.3945392491467576,
+      "grad_norm": 3.062695026397705,
+      "learning_rate": 0.0005351535836177475,
+      "loss": 6.4737,
+      "step": 4086
+    },
+    {
+      "epoch": 1.3948805460750853,
+      "grad_norm": 3.23443603515625,
+      "learning_rate": 0.0005350398179749716,
+      "loss": 5.8095,
+      "step": 4087
+    },
+    {
+      "epoch": 1.395221843003413,
+      "grad_norm": 3.249667167663574,
+      "learning_rate": 0.0005349260523321957,
+      "loss": 6.3928,
+      "step": 4088
+    },
+    {
+      "epoch": 1.3955631399317405,
+      "grad_norm": 3.0842514038085938,
+      "learning_rate": 0.0005348122866894199,
+      "loss": 6.2861,
+      "step": 4089
+    },
+    {
+      "epoch": 1.3959044368600684,
+      "grad_norm": 3.2530109882354736,
+      "learning_rate": 0.000534698521046644,
+      "loss": 5.9651,
+      "step": 4090
+    },
+    {
+      "epoch": 1.3962457337883958,
+      "grad_norm": 4.654758930206299,
+      "learning_rate": 0.0005345847554038681,
+      "loss": 5.5245,
+      "step": 4091
+    },
+    {
+      "epoch": 1.3965870307167236,
+      "grad_norm": 3.233858108520508,
+      "learning_rate": 0.0005344709897610922,
+      "loss": 6.6915,
+      "step": 4092
+    },
+    {
+      "epoch": 1.3969283276450513,
+      "grad_norm": 3.159165382385254,
+      "learning_rate": 0.0005343572241183163,
+      "loss": 6.3194,
+      "step": 4093
+    },
+    {
+      "epoch": 1.3972696245733789,
+      "grad_norm": 3.0182223320007324,
+      "learning_rate": 0.0005342434584755403,
+      "loss": 6.0745,
+      "step": 4094
+    },
+    {
+      "epoch": 1.3976109215017065,
+      "grad_norm": 3.3083250522613525,
+      "learning_rate": 0.0005341296928327645,
+      "loss": 5.7519,
+      "step": 4095
+    },
+    {
+      "epoch": 1.3979522184300341,
+      "grad_norm": 3.4524002075195312,
+      "learning_rate": 0.0005340159271899886,
+      "loss": 5.9577,
+      "step": 4096
+    },
+    {
+      "epoch": 1.3982935153583618,
+      "grad_norm": 3.095432996749878,
+      "learning_rate": 0.0005339021615472127,
+      "loss": 6.4214,
+      "step": 4097
+    },
+    {
+      "epoch": 1.3986348122866894,
+      "grad_norm": 3.4075958728790283,
+      "learning_rate": 0.0005337883959044368,
+      "loss": 5.6942,
+      "step": 4098
+    },
+    {
+      "epoch": 1.398976109215017,
+      "grad_norm": 4.778470993041992,
+      "learning_rate": 0.0005336746302616609,
+      "loss": 5.1441,
+      "step": 4099
+    },
+    {
+      "epoch": 1.3993174061433447,
+      "grad_norm": 4.083403587341309,
+      "learning_rate": 0.000533560864618885,
+      "loss": 5.8321,
+      "step": 4100
+    },
+    {
+      "epoch": 1.3996587030716723,
+      "grad_norm": 3.242581367492676,
+      "learning_rate": 0.0005334470989761093,
+      "loss": 6.1174,
+      "step": 4101
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.0662283897399902,
+      "learning_rate": 0.0005333333333333334,
+      "loss": 6.7307,
+      "step": 4102
+    },
+    {
+      "epoch": 1.4003412969283278,
+      "grad_norm": 3.119428873062134,
+      "learning_rate": 0.0005332195676905575,
+      "loss": 6.5688,
+      "step": 4103
+    },
+    {
+      "epoch": 1.4006825938566552,
+      "grad_norm": 3.014010190963745,
+      "learning_rate": 0.0005331058020477816,
+      "loss": 6.5248,
+      "step": 4104
+    },
+    {
+      "epoch": 1.401023890784983,
+      "grad_norm": 3.5993144512176514,
+      "learning_rate": 0.0005329920364050057,
+      "loss": 5.533,
+      "step": 4105
+    },
+    {
+      "epoch": 1.4013651877133106,
+      "grad_norm": 3.0127501487731934,
+      "learning_rate": 0.0005328782707622299,
+      "loss": 6.3895,
+      "step": 4106
+    },
+    {
+      "epoch": 1.4017064846416383,
+      "grad_norm": 3.770829916000366,
+      "learning_rate": 0.000532764505119454,
+      "loss": 5.6792,
+      "step": 4107
+    },
+    {
+      "epoch": 1.402047781569966,
+      "grad_norm": 3.10686993598938,
+      "learning_rate": 0.0005326507394766781,
+      "loss": 6.5603,
+      "step": 4108
+    },
+    {
+      "epoch": 1.4023890784982935,
+      "grad_norm": 3.173125982284546,
+      "learning_rate": 0.0005325369738339022,
+      "loss": 5.9965,
+      "step": 4109
+    },
+    {
+      "epoch": 1.4027303754266212,
+      "grad_norm": 3.075287103652954,
+      "learning_rate": 0.0005324232081911263,
+      "loss": 6.0827,
+      "step": 4110
+    },
+    {
+      "epoch": 1.4030716723549488,
+      "grad_norm": 3.0550026893615723,
+      "learning_rate": 0.0005323094425483504,
+      "loss": 6.2533,
+      "step": 4111
+    },
+    {
+      "epoch": 1.4034129692832764,
+      "grad_norm": 4.698912620544434,
+      "learning_rate": 0.0005321956769055746,
+      "loss": 4.8645,
+      "step": 4112
+    },
+    {
+      "epoch": 1.403754266211604,
+      "grad_norm": 3.12372088432312,
+      "learning_rate": 0.0005320819112627987,
+      "loss": 6.4869,
+      "step": 4113
+    },
+    {
+      "epoch": 1.4040955631399317,
+      "grad_norm": 3.2088253498077393,
+      "learning_rate": 0.0005319681456200227,
+      "loss": 6.1782,
+      "step": 4114
+    },
+    {
+      "epoch": 1.4044368600682593,
+      "grad_norm": 3.946972370147705,
+      "learning_rate": 0.0005318543799772468,
+      "loss": 5.4858,
+      "step": 4115
+    },
+    {
+      "epoch": 1.4047781569965871,
+      "grad_norm": 3.758061408996582,
+      "learning_rate": 0.0005317406143344709,
+      "loss": 5.5092,
+      "step": 4116
+    },
+    {
+      "epoch": 1.4051194539249146,
+      "grad_norm": 3.1338324546813965,
+      "learning_rate": 0.000531626848691695,
+      "loss": 6.4477,
+      "step": 4117
+    },
+    {
+      "epoch": 1.4054607508532424,
+      "grad_norm": 3.124582529067993,
+      "learning_rate": 0.0005315130830489193,
+      "loss": 5.2994,
+      "step": 4118
+    },
+    {
+      "epoch": 1.40580204778157,
+      "grad_norm": 3.2185561656951904,
+      "learning_rate": 0.0005313993174061434,
+      "loss": 6.29,
+      "step": 4119
+    },
+    {
+      "epoch": 1.4061433447098977,
+      "grad_norm": 3.439194440841675,
+      "learning_rate": 0.0005312855517633675,
+      "loss": 5.8294,
+      "step": 4120
+    },
+    {
+      "epoch": 1.4064846416382253,
+      "grad_norm": 3.0680735111236572,
+      "learning_rate": 0.0005311717861205916,
+      "loss": 6.3791,
+      "step": 4121
+    },
+    {
+      "epoch": 1.406825938566553,
+      "grad_norm": 3.082590341567993,
+      "learning_rate": 0.0005310580204778157,
+      "loss": 6.5443,
+      "step": 4122
+    },
+    {
+      "epoch": 1.4071672354948805,
+      "grad_norm": 3.0105934143066406,
+      "learning_rate": 0.0005309442548350398,
+      "loss": 6.4141,
+      "step": 4123
+    },
+    {
+      "epoch": 1.4075085324232082,
+      "grad_norm": 3.0490424633026123,
+      "learning_rate": 0.000530830489192264,
+      "loss": 6.5894,
+      "step": 4124
+    },
+    {
+      "epoch": 1.4078498293515358,
+      "grad_norm": 3.1783745288848877,
+      "learning_rate": 0.0005307167235494881,
+      "loss": 6.3531,
+      "step": 4125
+    },
+    {
+      "epoch": 1.4081911262798634,
+      "grad_norm": 3.387235641479492,
+      "learning_rate": 0.0005306029579067122,
+      "loss": 6.4013,
+      "step": 4126
+    },
+    {
+      "epoch": 1.408532423208191,
+      "grad_norm": 3.4885988235473633,
+      "learning_rate": 0.0005304891922639363,
+      "loss": 6.4457,
+      "step": 4127
+    },
+    {
+      "epoch": 1.4088737201365187,
+      "grad_norm": 3.323430061340332,
+      "learning_rate": 0.0005303754266211604,
+      "loss": 6.1437,
+      "step": 4128
+    },
+    {
+      "epoch": 1.4092150170648465,
+      "grad_norm": 5.294995307922363,
+      "learning_rate": 0.0005302616609783846,
+      "loss": 5.1343,
+      "step": 4129
+    },
+    {
+      "epoch": 1.409556313993174,
+      "grad_norm": 3.099463701248169,
+      "learning_rate": 0.0005301478953356087,
+      "loss": 6.3936,
+      "step": 4130
+    },
+    {
+      "epoch": 1.4098976109215018,
+      "grad_norm": 3.1338493824005127,
+      "learning_rate": 0.0005300341296928328,
+      "loss": 6.8859,
+      "step": 4131
+    },
+    {
+      "epoch": 1.4102389078498294,
+      "grad_norm": 3.034945011138916,
+      "learning_rate": 0.0005299203640500569,
+      "loss": 6.1265,
+      "step": 4132
+    },
+    {
+      "epoch": 1.410580204778157,
+      "grad_norm": 3.033184051513672,
+      "learning_rate": 0.0005298065984072809,
+      "loss": 6.3814,
+      "step": 4133
+    },
+    {
+      "epoch": 1.4109215017064847,
+      "grad_norm": 2.992227554321289,
+      "learning_rate": 0.000529692832764505,
+      "loss": 6.4773,
+      "step": 4134
+    },
+    {
+      "epoch": 1.4112627986348123,
+      "grad_norm": 3.4906859397888184,
+      "learning_rate": 0.0005295790671217293,
+      "loss": 5.8243,
+      "step": 4135
+    },
+    {
+      "epoch": 1.41160409556314,
+      "grad_norm": 3.140052318572998,
+      "learning_rate": 0.0005294653014789534,
+      "loss": 6.3378,
+      "step": 4136
+    },
+    {
+      "epoch": 1.4119453924914676,
+      "grad_norm": 4.089855670928955,
+      "learning_rate": 0.0005293515358361775,
+      "loss": 6.1393,
+      "step": 4137
+    },
+    {
+      "epoch": 1.4122866894197952,
+      "grad_norm": 3.1651060581207275,
+      "learning_rate": 0.0005292377701934016,
+      "loss": 6.5695,
+      "step": 4138
+    },
+    {
+      "epoch": 1.4126279863481228,
+      "grad_norm": 3.1677775382995605,
+      "learning_rate": 0.0005291240045506257,
+      "loss": 6.1942,
+      "step": 4139
+    },
+    {
+      "epoch": 1.4129692832764504,
+      "grad_norm": 3.1922414302825928,
+      "learning_rate": 0.0005290102389078498,
+      "loss": 6.1814,
+      "step": 4140
+    },
+    {
+      "epoch": 1.413310580204778,
+      "grad_norm": 3.341269016265869,
+      "learning_rate": 0.000528896473265074,
+      "loss": 6.1189,
+      "step": 4141
+    },
+    {
+      "epoch": 1.413651877133106,
+      "grad_norm": 3.1737465858459473,
+      "learning_rate": 0.0005287827076222981,
+      "loss": 6.324,
+      "step": 4142
+    },
+    {
+      "epoch": 1.4139931740614333,
+      "grad_norm": 3.0902111530303955,
+      "learning_rate": 0.0005286689419795222,
+      "loss": 6.0682,
+      "step": 4143
+    },
+    {
+      "epoch": 1.4143344709897612,
+      "grad_norm": 3.198076009750366,
+      "learning_rate": 0.0005285551763367463,
+      "loss": 6.6576,
+      "step": 4144
+    },
+    {
+      "epoch": 1.4146757679180888,
+      "grad_norm": 3.349595785140991,
+      "learning_rate": 0.0005284414106939704,
+      "loss": 5.8567,
+      "step": 4145
+    },
+    {
+      "epoch": 1.4150170648464164,
+      "grad_norm": 3.0596208572387695,
+      "learning_rate": 0.0005283276450511945,
+      "loss": 6.6378,
+      "step": 4146
+    },
+    {
+      "epoch": 1.415358361774744,
+      "grad_norm": 3.092496633529663,
+      "learning_rate": 0.0005282138794084187,
+      "loss": 6.8541,
+      "step": 4147
+    },
+    {
+      "epoch": 1.4156996587030717,
+      "grad_norm": 3.092621088027954,
+      "learning_rate": 0.0005281001137656428,
+      "loss": 6.2547,
+      "step": 4148
+    },
+    {
+      "epoch": 1.4160409556313993,
+      "grad_norm": 3.077143907546997,
+      "learning_rate": 0.0005279863481228669,
+      "loss": 6.9397,
+      "step": 4149
+    },
+    {
+      "epoch": 1.416382252559727,
+      "grad_norm": 3.1491472721099854,
+      "learning_rate": 0.000527872582480091,
+      "loss": 6.3133,
+      "step": 4150
+    },
+    {
+      "epoch": 1.4167235494880546,
+      "grad_norm": 3.1234889030456543,
+      "learning_rate": 0.0005277588168373152,
+      "loss": 6.5563,
+      "step": 4151
+    },
+    {
+      "epoch": 1.4170648464163822,
+      "grad_norm": 3.086249828338623,
+      "learning_rate": 0.0005276450511945393,
+      "loss": 6.5361,
+      "step": 4152
+    },
+    {
+      "epoch": 1.4174061433447098,
+      "grad_norm": 3.6380252838134766,
+      "learning_rate": 0.0005275312855517634,
+      "loss": 6.0914,
+      "step": 4153
+    },
+    {
+      "epoch": 1.4177474402730375,
+      "grad_norm": 4.482765197753906,
+      "learning_rate": 0.0005274175199089875,
+      "loss": 6.6235,
+      "step": 4154
+    },
+    {
+      "epoch": 1.4180887372013653,
+      "grad_norm": 3.4601078033447266,
+      "learning_rate": 0.0005273037542662116,
+      "loss": 6.413,
+      "step": 4155
+    },
+    {
+      "epoch": 1.4184300341296927,
+      "grad_norm": 3.319906234741211,
+      "learning_rate": 0.0005271899886234357,
+      "loss": 6.8646,
+      "step": 4156
+    },
+    {
+      "epoch": 1.4187713310580206,
+      "grad_norm": 3.219625473022461,
+      "learning_rate": 0.0005270762229806598,
+      "loss": 6.1368,
+      "step": 4157
+    },
+    {
+      "epoch": 1.4191126279863482,
+      "grad_norm": 3.033139228820801,
+      "learning_rate": 0.000526962457337884,
+      "loss": 6.9841,
+      "step": 4158
+    },
+    {
+      "epoch": 1.4194539249146758,
+      "grad_norm": 3.063420534133911,
+      "learning_rate": 0.0005268486916951081,
+      "loss": 6.2634,
+      "step": 4159
+    },
+    {
+      "epoch": 1.4197952218430034,
+      "grad_norm": 6.298565864562988,
+      "learning_rate": 0.0005267349260523322,
+      "loss": 5.7468,
+      "step": 4160
+    },
+    {
+      "epoch": 1.420136518771331,
+      "grad_norm": 3.0811851024627686,
+      "learning_rate": 0.0005266211604095563,
+      "loss": 6.4445,
+      "step": 4161
+    },
+    {
+      "epoch": 1.4204778156996587,
+      "grad_norm": 3.210052967071533,
+      "learning_rate": 0.0005265073947667804,
+      "loss": 6.1846,
+      "step": 4162
+    },
+    {
+      "epoch": 1.4208191126279863,
+      "grad_norm": 3.0224571228027344,
+      "learning_rate": 0.0005263936291240045,
+      "loss": 6.5175,
+      "step": 4163
+    },
+    {
+      "epoch": 1.421160409556314,
+      "grad_norm": 3.2952568531036377,
+      "learning_rate": 0.0005262798634812287,
+      "loss": 6.3716,
+      "step": 4164
+    },
+    {
+      "epoch": 1.4215017064846416,
+      "grad_norm": 3.3877005577087402,
+      "learning_rate": 0.0005261660978384528,
+      "loss": 6.2331,
+      "step": 4165
+    },
+    {
+      "epoch": 1.4218430034129692,
+      "grad_norm": 3.012554168701172,
+      "learning_rate": 0.0005260523321956769,
+      "loss": 6.0445,
+      "step": 4166
+    },
+    {
+      "epoch": 1.4221843003412968,
+      "grad_norm": 3.1453211307525635,
+      "learning_rate": 0.000525938566552901,
+      "loss": 6.6895,
+      "step": 4167
+    },
+    {
+      "epoch": 1.4225255972696247,
+      "grad_norm": 3.146012783050537,
+      "learning_rate": 0.0005258248009101252,
+      "loss": 6.581,
+      "step": 4168
+    },
+    {
+      "epoch": 1.422866894197952,
+      "grad_norm": 3.0287742614746094,
+      "learning_rate": 0.0005257110352673494,
+      "loss": 6.2703,
+      "step": 4169
+    },
+    {
+      "epoch": 1.42320819112628,
+      "grad_norm": 3.0194754600524902,
+      "learning_rate": 0.0005255972696245735,
+      "loss": 6.4683,
+      "step": 4170
+    },
+    {
+      "epoch": 1.4235494880546076,
+      "grad_norm": 2.9725260734558105,
+      "learning_rate": 0.0005254835039817976,
+      "loss": 6.3189,
+      "step": 4171
+    },
+    {
+      "epoch": 1.4238907849829352,
+      "grad_norm": 3.101661443710327,
+      "learning_rate": 0.0005253697383390216,
+      "loss": 6.1797,
+      "step": 4172
+    },
+    {
+      "epoch": 1.4242320819112628,
+      "grad_norm": 3.5125575065612793,
+      "learning_rate": 0.0005252559726962457,
+      "loss": 6.1972,
+      "step": 4173
+    },
+    {
+      "epoch": 1.4245733788395905,
+      "grad_norm": 3.2034642696380615,
+      "learning_rate": 0.0005251422070534698,
+      "loss": 6.4367,
+      "step": 4174
+    },
+    {
+      "epoch": 1.424914675767918,
+      "grad_norm": 3.173069715499878,
+      "learning_rate": 0.000525028441410694,
+      "loss": 6.1895,
+      "step": 4175
+    },
+    {
+      "epoch": 1.4252559726962457,
+      "grad_norm": 3.3157718181610107,
+      "learning_rate": 0.0005249146757679181,
+      "loss": 5.2822,
+      "step": 4176
+    },
+    {
+      "epoch": 1.4255972696245733,
+      "grad_norm": 3.15852952003479,
+      "learning_rate": 0.0005248009101251422,
+      "loss": 6.4959,
+      "step": 4177
+    },
+    {
+      "epoch": 1.425938566552901,
+      "grad_norm": 2.8751838207244873,
+      "learning_rate": 0.0005246871444823663,
+      "loss": 4.8944,
+      "step": 4178
+    },
+    {
+      "epoch": 1.4262798634812286,
+      "grad_norm": 3.131141185760498,
+      "learning_rate": 0.0005245733788395904,
+      "loss": 6.5915,
+      "step": 4179
+    },
+    {
+      "epoch": 1.4266211604095562,
+      "grad_norm": 3.1316215991973877,
+      "learning_rate": 0.0005244596131968145,
+      "loss": 5.7723,
+      "step": 4180
+    },
+    {
+      "epoch": 1.426962457337884,
+      "grad_norm": 4.314980506896973,
+      "learning_rate": 0.0005243458475540387,
+      "loss": 5.9181,
+      "step": 4181
+    },
+    {
+      "epoch": 1.4273037542662115,
+      "grad_norm": 3.1601321697235107,
+      "learning_rate": 0.0005242320819112628,
+      "loss": 6.1492,
+      "step": 4182
+    },
+    {
+      "epoch": 1.4276450511945393,
+      "grad_norm": 3.1197149753570557,
+      "learning_rate": 0.0005241183162684869,
+      "loss": 6.6196,
+      "step": 4183
+    },
+    {
+      "epoch": 1.427986348122867,
+      "grad_norm": 3.109541893005371,
+      "learning_rate": 0.000524004550625711,
+      "loss": 6.5163,
+      "step": 4184
+    },
+    {
+      "epoch": 1.4283276450511946,
+      "grad_norm": 3.131387233734131,
+      "learning_rate": 0.0005238907849829352,
+      "loss": 6.6058,
+      "step": 4185
+    },
+    {
+      "epoch": 1.4286689419795222,
+      "grad_norm": 4.881186485290527,
+      "learning_rate": 0.0005237770193401593,
+      "loss": 6.105,
+      "step": 4186
+    },
+    {
+      "epoch": 1.4290102389078498,
+      "grad_norm": 3.3822786808013916,
+      "learning_rate": 0.0005236632536973835,
+      "loss": 6.0654,
+      "step": 4187
+    },
+    {
+      "epoch": 1.4293515358361775,
+      "grad_norm": 3.2397496700286865,
+      "learning_rate": 0.0005235494880546076,
+      "loss": 6.7056,
+      "step": 4188
+    },
+    {
+      "epoch": 1.429692832764505,
+      "grad_norm": 3.1808483600616455,
+      "learning_rate": 0.0005234357224118317,
+      "loss": 6.3953,
+      "step": 4189
+    },
+    {
+      "epoch": 1.4300341296928327,
+      "grad_norm": 3.1761183738708496,
+      "learning_rate": 0.0005233219567690558,
+      "loss": 6.3699,
+      "step": 4190
+    },
+    {
+      "epoch": 1.4303754266211604,
+      "grad_norm": 3.019099712371826,
+      "learning_rate": 0.0005232081911262798,
+      "loss": 6.266,
+      "step": 4191
+    },
+    {
+      "epoch": 1.430716723549488,
+      "grad_norm": 3.0634679794311523,
+      "learning_rate": 0.000523094425483504,
+      "loss": 6.5473,
+      "step": 4192
+    },
+    {
+      "epoch": 1.4310580204778156,
+      "grad_norm": 3.0792815685272217,
+      "learning_rate": 0.0005229806598407281,
+      "loss": 6.3998,
+      "step": 4193
+    },
+    {
+      "epoch": 1.4313993174061435,
+      "grad_norm": 3.048142433166504,
+      "learning_rate": 0.0005228668941979522,
+      "loss": 6.0785,
+      "step": 4194
+    },
+    {
+      "epoch": 1.4317406143344709,
+      "grad_norm": 3.5605955123901367,
+      "learning_rate": 0.0005227531285551763,
+      "loss": 6.2169,
+      "step": 4195
+    },
+    {
+      "epoch": 1.4320819112627987,
+      "grad_norm": 3.1062545776367188,
+      "learning_rate": 0.0005226393629124004,
+      "loss": 6.2116,
+      "step": 4196
+    },
+    {
+      "epoch": 1.4324232081911263,
+      "grad_norm": 3.5202982425689697,
+      "learning_rate": 0.0005225255972696245,
+      "loss": 6.6248,
+      "step": 4197
+    },
+    {
+      "epoch": 1.432764505119454,
+      "grad_norm": 4.942377090454102,
+      "learning_rate": 0.0005224118316268487,
+      "loss": 4.8293,
+      "step": 4198
+    },
+    {
+      "epoch": 1.4331058020477816,
+      "grad_norm": 3.237367868423462,
+      "learning_rate": 0.0005222980659840728,
+      "loss": 6.9193,
+      "step": 4199
+    },
+    {
+      "epoch": 1.4334470989761092,
+      "grad_norm": 3.33652400970459,
+      "learning_rate": 0.0005221843003412969,
+      "loss": 5.7304,
+      "step": 4200
+    },
+    {
+      "epoch": 1.4337883959044369,
+      "grad_norm": 4.465225696563721,
+      "learning_rate": 0.000522070534698521,
+      "loss": 5.5101,
+      "step": 4201
+    },
+    {
+      "epoch": 1.4341296928327645,
+      "grad_norm": 3.111236572265625,
+      "learning_rate": 0.0005219567690557452,
+      "loss": 6.8656,
+      "step": 4202
+    },
+    {
+      "epoch": 1.434470989761092,
+      "grad_norm": 3.2016916275024414,
+      "learning_rate": 0.0005218430034129693,
+      "loss": 6.7554,
+      "step": 4203
+    },
+    {
+      "epoch": 1.4348122866894197,
+      "grad_norm": 3.0454368591308594,
+      "learning_rate": 0.0005217292377701935,
+      "loss": 6.2635,
+      "step": 4204
+    },
+    {
+      "epoch": 1.4351535836177474,
+      "grad_norm": 3.07362699508667,
+      "learning_rate": 0.0005216154721274176,
+      "loss": 5.867,
+      "step": 4205
+    },
+    {
+      "epoch": 1.435494880546075,
+      "grad_norm": 3.0416789054870605,
+      "learning_rate": 0.0005215017064846417,
+      "loss": 6.4948,
+      "step": 4206
+    },
+    {
+      "epoch": 1.4358361774744028,
+      "grad_norm": 3.082453966140747,
+      "learning_rate": 0.0005213879408418658,
+      "loss": 6.4393,
+      "step": 4207
+    },
+    {
+      "epoch": 1.4361774744027302,
+      "grad_norm": 4.051252365112305,
+      "learning_rate": 0.0005212741751990899,
+      "loss": 5.6161,
+      "step": 4208
+    },
+    {
+      "epoch": 1.436518771331058,
+      "grad_norm": 3.2135181427001953,
+      "learning_rate": 0.0005211604095563141,
+      "loss": 6.2619,
+      "step": 4209
+    },
+    {
+      "epoch": 1.4368600682593857,
+      "grad_norm": 4.387260437011719,
+      "learning_rate": 0.0005210466439135382,
+      "loss": 5.5295,
+      "step": 4210
+    },
+    {
+      "epoch": 1.4372013651877134,
+      "grad_norm": 3.195601463317871,
+      "learning_rate": 0.0005209328782707622,
+      "loss": 6.3014,
+      "step": 4211
+    },
+    {
+      "epoch": 1.437542662116041,
+      "grad_norm": 3.236250877380371,
+      "learning_rate": 0.0005208191126279863,
+      "loss": 6.6634,
+      "step": 4212
+    },
+    {
+      "epoch": 1.4378839590443686,
+      "grad_norm": 3.2227210998535156,
+      "learning_rate": 0.0005207053469852104,
+      "loss": 6.0613,
+      "step": 4213
+    },
+    {
+      "epoch": 1.4382252559726962,
+      "grad_norm": 4.377707004547119,
+      "learning_rate": 0.0005205915813424345,
+      "loss": 4.7205,
+      "step": 4214
+    },
+    {
+      "epoch": 1.4385665529010239,
+      "grad_norm": 3.1697402000427246,
+      "learning_rate": 0.0005204778156996587,
+      "loss": 6.6822,
+      "step": 4215
+    },
+    {
+      "epoch": 1.4389078498293515,
+      "grad_norm": 3.424391746520996,
+      "learning_rate": 0.0005203640500568828,
+      "loss": 6.3304,
+      "step": 4216
+    },
+    {
+      "epoch": 1.4392491467576791,
+      "grad_norm": 3.166321039199829,
+      "learning_rate": 0.0005202502844141069,
+      "loss": 6.5789,
+      "step": 4217
+    },
+    {
+      "epoch": 1.4395904436860067,
+      "grad_norm": 3.058227062225342,
+      "learning_rate": 0.000520136518771331,
+      "loss": 6.3097,
+      "step": 4218
+    },
+    {
+      "epoch": 1.4399317406143344,
+      "grad_norm": 3.0803213119506836,
+      "learning_rate": 0.0005200227531285552,
+      "loss": 6.195,
+      "step": 4219
+    },
+    {
+      "epoch": 1.4402730375426622,
+      "grad_norm": 3.0819287300109863,
+      "learning_rate": 0.0005199089874857793,
+      "loss": 6.3407,
+      "step": 4220
+    },
+    {
+      "epoch": 1.4406143344709896,
+      "grad_norm": 3.058324098587036,
+      "learning_rate": 0.0005197952218430035,
+      "loss": 6.1638,
+      "step": 4221
+    },
+    {
+      "epoch": 1.4409556313993175,
+      "grad_norm": 3.040989398956299,
+      "learning_rate": 0.0005196814562002276,
+      "loss": 6.2791,
+      "step": 4222
+    },
+    {
+      "epoch": 1.441296928327645,
+      "grad_norm": 3.0801877975463867,
+      "learning_rate": 0.0005195676905574517,
+      "loss": 5.9373,
+      "step": 4223
+    },
+    {
+      "epoch": 1.4416382252559727,
+      "grad_norm": 3.13028883934021,
+      "learning_rate": 0.0005194539249146758,
+      "loss": 5.9899,
+      "step": 4224
+    },
+    {
+      "epoch": 1.4419795221843004,
+      "grad_norm": 3.0730881690979004,
+      "learning_rate": 0.0005193401592718999,
+      "loss": 6.0214,
+      "step": 4225
+    },
+    {
+      "epoch": 1.442320819112628,
+      "grad_norm": 3.320467233657837,
+      "learning_rate": 0.000519226393629124,
+      "loss": 6.4929,
+      "step": 4226
+    },
+    {
+      "epoch": 1.4426621160409556,
+      "grad_norm": 3.381704330444336,
+      "learning_rate": 0.0005191126279863482,
+      "loss": 5.9312,
+      "step": 4227
+    },
+    {
+      "epoch": 1.4430034129692833,
+      "grad_norm": 3.1299855709075928,
+      "learning_rate": 0.0005189988623435723,
+      "loss": 6.446,
+      "step": 4228
+    },
+    {
+      "epoch": 1.4433447098976109,
+      "grad_norm": 3.0571303367614746,
+      "learning_rate": 0.0005188850967007964,
+      "loss": 6.3526,
+      "step": 4229
+    },
+    {
+      "epoch": 1.4436860068259385,
+      "grad_norm": 3.1851227283477783,
+      "learning_rate": 0.0005187713310580204,
+      "loss": 6.7403,
+      "step": 4230
+    },
+    {
+      "epoch": 1.4440273037542661,
+      "grad_norm": 3.1886208057403564,
+      "learning_rate": 0.0005186575654152445,
+      "loss": 6.0976,
+      "step": 4231
+    },
+    {
+      "epoch": 1.4443686006825938,
+      "grad_norm": 3.0047943592071533,
+      "learning_rate": 0.0005185437997724687,
+      "loss": 6.605,
+      "step": 4232
+    },
+    {
+      "epoch": 1.4447098976109216,
+      "grad_norm": 3.09120512008667,
+      "learning_rate": 0.0005184300341296928,
+      "loss": 6.6422,
+      "step": 4233
+    },
+    {
+      "epoch": 1.445051194539249,
+      "grad_norm": 3.228661060333252,
+      "learning_rate": 0.000518316268486917,
+      "loss": 6.1845,
+      "step": 4234
+    },
+    {
+      "epoch": 1.4453924914675769,
+      "grad_norm": 3.080799102783203,
+      "learning_rate": 0.000518202502844141,
+      "loss": 6.8722,
+      "step": 4235
+    },
+    {
+      "epoch": 1.4457337883959045,
+      "grad_norm": 3.116177558898926,
+      "learning_rate": 0.0005180887372013652,
+      "loss": 6.7637,
+      "step": 4236
+    },
+    {
+      "epoch": 1.4460750853242321,
+      "grad_norm": 3.183600664138794,
+      "learning_rate": 0.0005179749715585893,
+      "loss": 5.8418,
+      "step": 4237
+    },
+    {
+      "epoch": 1.4464163822525598,
+      "grad_norm": 3.1986725330352783,
+      "learning_rate": 0.0005178612059158135,
+      "loss": 5.7188,
+      "step": 4238
+    },
+    {
+      "epoch": 1.4467576791808874,
+      "grad_norm": 3.249572277069092,
+      "learning_rate": 0.0005177474402730376,
+      "loss": 5.8931,
+      "step": 4239
+    },
+    {
+      "epoch": 1.447098976109215,
+      "grad_norm": 3.0727765560150146,
+      "learning_rate": 0.0005176336746302617,
+      "loss": 6.6941,
+      "step": 4240
+    },
+    {
+      "epoch": 1.4474402730375426,
+      "grad_norm": 2.9626481533050537,
+      "learning_rate": 0.0005175199089874858,
+      "loss": 5.5147,
+      "step": 4241
+    },
+    {
+      "epoch": 1.4477815699658703,
+      "grad_norm": 2.4201412200927734,
+      "learning_rate": 0.0005174061433447099,
+      "loss": 3.1755,
+      "step": 4242
+    },
+    {
+      "epoch": 1.448122866894198,
+      "grad_norm": 3.06439471244812,
+      "learning_rate": 0.000517292377701934,
+      "loss": 6.101,
+      "step": 4243
+    },
+    {
+      "epoch": 1.4484641638225255,
+      "grad_norm": 3.1331796646118164,
+      "learning_rate": 0.0005171786120591582,
+      "loss": 6.3183,
+      "step": 4244
+    },
+    {
+      "epoch": 1.4488054607508531,
+      "grad_norm": 3.0797486305236816,
+      "learning_rate": 0.0005170648464163823,
+      "loss": 6.3455,
+      "step": 4245
+    },
+    {
+      "epoch": 1.449146757679181,
+      "grad_norm": 3.1885828971862793,
+      "learning_rate": 0.0005169510807736064,
+      "loss": 6.6001,
+      "step": 4246
+    },
+    {
+      "epoch": 1.4494880546075084,
+      "grad_norm": 4.609078884124756,
+      "learning_rate": 0.0005168373151308305,
+      "loss": 4.6183,
+      "step": 4247
+    },
+    {
+      "epoch": 1.4498293515358363,
+      "grad_norm": 3.357383966445923,
+      "learning_rate": 0.0005167235494880546,
+      "loss": 6.4766,
+      "step": 4248
+    },
+    {
+      "epoch": 1.4501706484641639,
+      "grad_norm": 3.172637701034546,
+      "learning_rate": 0.0005166097838452787,
+      "loss": 6.3503,
+      "step": 4249
+    },
+    {
+      "epoch": 1.4505119453924915,
+      "grad_norm": 3.2504348754882812,
+      "learning_rate": 0.0005164960182025028,
+      "loss": 6.4823,
+      "step": 4250
+    },
+    {
+      "epoch": 1.4508532423208191,
+      "grad_norm": 3.318784475326538,
+      "learning_rate": 0.000516382252559727,
+      "loss": 5.2902,
+      "step": 4251
+    },
+    {
+      "epoch": 1.4511945392491468,
+      "grad_norm": 3.166903257369995,
+      "learning_rate": 0.000516268486916951,
+      "loss": 7.0076,
+      "step": 4252
+    },
+    {
+      "epoch": 1.4515358361774744,
+      "grad_norm": 3.29073166847229,
+      "learning_rate": 0.0005161547212741752,
+      "loss": 5.809,
+      "step": 4253
+    },
+    {
+      "epoch": 1.451877133105802,
+      "grad_norm": 3.2139933109283447,
+      "learning_rate": 0.0005160409556313993,
+      "loss": 6.6659,
+      "step": 4254
+    },
+    {
+      "epoch": 1.4522184300341296,
+      "grad_norm": 2.99873685836792,
+      "learning_rate": 0.0005159271899886235,
+      "loss": 6.3525,
+      "step": 4255
+    },
+    {
+      "epoch": 1.4525597269624573,
+      "grad_norm": 3.2787253856658936,
+      "learning_rate": 0.0005158134243458476,
+      "loss": 5.645,
+      "step": 4256
+    },
+    {
+      "epoch": 1.452901023890785,
+      "grad_norm": 3.140230178833008,
+      "learning_rate": 0.0005156996587030717,
+      "loss": 6.1587,
+      "step": 4257
+    },
+    {
+      "epoch": 1.4532423208191125,
+      "grad_norm": 4.204946994781494,
+      "learning_rate": 0.0005155858930602958,
+      "loss": 5.4848,
+      "step": 4258
+    },
+    {
+      "epoch": 1.4535836177474404,
+      "grad_norm": 3.3353021144866943,
+      "learning_rate": 0.0005154721274175199,
+      "loss": 6.1801,
+      "step": 4259
+    },
+    {
+      "epoch": 1.4539249146757678,
+      "grad_norm": 3.392322540283203,
+      "learning_rate": 0.000515358361774744,
+      "loss": 6.8188,
+      "step": 4260
+    },
+    {
+      "epoch": 1.4542662116040956,
+      "grad_norm": 3.1940486431121826,
+      "learning_rate": 0.0005152445961319682,
+      "loss": 7.0783,
+      "step": 4261
+    },
+    {
+      "epoch": 1.4546075085324233,
+      "grad_norm": 3.1675353050231934,
+      "learning_rate": 0.0005151308304891923,
+      "loss": 6.6465,
+      "step": 4262
+    },
+    {
+      "epoch": 1.454948805460751,
+      "grad_norm": 3.268160581588745,
+      "learning_rate": 0.0005150170648464164,
+      "loss": 6.48,
+      "step": 4263
+    },
+    {
+      "epoch": 1.4552901023890785,
+      "grad_norm": 3.4621543884277344,
+      "learning_rate": 0.0005149032992036405,
+      "loss": 6.2458,
+      "step": 4264
+    },
+    {
+      "epoch": 1.4556313993174061,
+      "grad_norm": 3.037792682647705,
+      "learning_rate": 0.0005147895335608646,
+      "loss": 6.4756,
+      "step": 4265
+    },
+    {
+      "epoch": 1.4559726962457338,
+      "grad_norm": 3.1558916568756104,
+      "learning_rate": 0.0005146757679180887,
+      "loss": 6.99,
+      "step": 4266
+    },
+    {
+      "epoch": 1.4563139931740614,
+      "grad_norm": 3.1241016387939453,
+      "learning_rate": 0.000514562002275313,
+      "loss": 6.639,
+      "step": 4267
+    },
+    {
+      "epoch": 1.456655290102389,
+      "grad_norm": 3.117509603500366,
+      "learning_rate": 0.0005144482366325371,
+      "loss": 6.2899,
+      "step": 4268
+    },
+    {
+      "epoch": 1.4569965870307167,
+      "grad_norm": 3.071044683456421,
+      "learning_rate": 0.000514334470989761,
+      "loss": 6.1585,
+      "step": 4269
+    },
+    {
+      "epoch": 1.4573378839590443,
+      "grad_norm": 3.1628692150115967,
+      "learning_rate": 0.0005142207053469852,
+      "loss": 6.4385,
+      "step": 4270
+    },
+    {
+      "epoch": 1.457679180887372,
+      "grad_norm": 5.46381950378418,
+      "learning_rate": 0.0005141069397042093,
+      "loss": 4.9057,
+      "step": 4271
+    },
+    {
+      "epoch": 1.4580204778156998,
+      "grad_norm": 3.0911929607391357,
+      "learning_rate": 0.0005139931740614334,
+      "loss": 6.5145,
+      "step": 4272
+    },
+    {
+      "epoch": 1.4583617747440272,
+      "grad_norm": 3.1896402835845947,
+      "learning_rate": 0.0005138794084186576,
+      "loss": 5.9364,
+      "step": 4273
+    },
+    {
+      "epoch": 1.458703071672355,
+      "grad_norm": 3.1629385948181152,
+      "learning_rate": 0.0005137656427758817,
+      "loss": 5.7098,
+      "step": 4274
+    },
+    {
+      "epoch": 1.4590443686006827,
+      "grad_norm": 3.908473014831543,
+      "learning_rate": 0.0005136518771331058,
+      "loss": 6.1301,
+      "step": 4275
+    },
+    {
+      "epoch": 1.4593856655290103,
+      "grad_norm": 3.4416491985321045,
+      "learning_rate": 0.0005135381114903299,
+      "loss": 5.7183,
+      "step": 4276
+    },
+    {
+      "epoch": 1.459726962457338,
+      "grad_norm": 3.129523992538452,
+      "learning_rate": 0.000513424345847554,
+      "loss": 6.4804,
+      "step": 4277
+    },
+    {
+      "epoch": 1.4600682593856655,
+      "grad_norm": 3.913393020629883,
+      "learning_rate": 0.0005133105802047782,
+      "loss": 5.0513,
+      "step": 4278
+    },
+    {
+      "epoch": 1.4604095563139932,
+      "grad_norm": 3.2042019367218018,
+      "learning_rate": 0.0005131968145620023,
+      "loss": 5.7857,
+      "step": 4279
+    },
+    {
+      "epoch": 1.4607508532423208,
+      "grad_norm": 3.098621368408203,
+      "learning_rate": 0.0005130830489192264,
+      "loss": 6.7836,
+      "step": 4280
+    },
+    {
+      "epoch": 1.4610921501706484,
+      "grad_norm": 3.1765682697296143,
+      "learning_rate": 0.0005129692832764505,
+      "loss": 6.9352,
+      "step": 4281
+    },
+    {
+      "epoch": 1.461433447098976,
+      "grad_norm": 3.0995981693267822,
+      "learning_rate": 0.0005128555176336746,
+      "loss": 6.4874,
+      "step": 4282
+    },
+    {
+      "epoch": 1.4617747440273037,
+      "grad_norm": 3.3691446781158447,
+      "learning_rate": 0.0005127417519908987,
+      "loss": 6.0037,
+      "step": 4283
+    },
+    {
+      "epoch": 1.4621160409556313,
+      "grad_norm": 4.908443450927734,
+      "learning_rate": 0.000512627986348123,
+      "loss": 4.7213,
+      "step": 4284
+    },
+    {
+      "epoch": 1.4624573378839592,
+      "grad_norm": 3.1471049785614014,
+      "learning_rate": 0.0005125142207053471,
+      "loss": 6.9427,
+      "step": 4285
+    },
+    {
+      "epoch": 1.4627986348122866,
+      "grad_norm": 3.119262456893921,
+      "learning_rate": 0.0005124004550625712,
+      "loss": 6.4854,
+      "step": 4286
+    },
+    {
+      "epoch": 1.4631399317406144,
+      "grad_norm": 3.1892127990722656,
+      "learning_rate": 0.0005122866894197953,
+      "loss": 6.2065,
+      "step": 4287
+    },
+    {
+      "epoch": 1.463481228668942,
+      "grad_norm": 3.0533523559570312,
+      "learning_rate": 0.0005121729237770194,
+      "loss": 6.6329,
+      "step": 4288
+    },
+    {
+      "epoch": 1.4638225255972697,
+      "grad_norm": 4.774298191070557,
+      "learning_rate": 0.0005120591581342434,
+      "loss": 5.365,
+      "step": 4289
+    },
+    {
+      "epoch": 1.4641638225255973,
+      "grad_norm": 4.030971050262451,
+      "learning_rate": 0.0005119453924914676,
+      "loss": 5.6187,
+      "step": 4290
+    },
+    {
+      "epoch": 1.464505119453925,
+      "grad_norm": 3.2482876777648926,
+      "learning_rate": 0.0005118316268486917,
+      "loss": 6.1619,
+      "step": 4291
+    },
+    {
+      "epoch": 1.4648464163822525,
+      "grad_norm": 2.9875893592834473,
+      "learning_rate": 0.0005117178612059158,
+      "loss": 6.4415,
+      "step": 4292
+    },
+    {
+      "epoch": 1.4651877133105802,
+      "grad_norm": 4.620938301086426,
+      "learning_rate": 0.0005116040955631399,
+      "loss": 1.5726,
+      "step": 4293
+    },
+    {
+      "epoch": 1.4655290102389078,
+      "grad_norm": 3.08268141746521,
+      "learning_rate": 0.000511490329920364,
+      "loss": 6.0083,
+      "step": 4294
+    },
+    {
+      "epoch": 1.4658703071672354,
+      "grad_norm": 3.3414804935455322,
+      "learning_rate": 0.0005113765642775882,
+      "loss": 6.0082,
+      "step": 4295
+    },
+    {
+      "epoch": 1.466211604095563,
+      "grad_norm": 4.3554792404174805,
+      "learning_rate": 0.0005112627986348123,
+      "loss": 4.8411,
+      "step": 4296
+    },
+    {
+      "epoch": 1.4665529010238907,
+      "grad_norm": 3.1311569213867188,
+      "learning_rate": 0.0005111490329920364,
+      "loss": 6.3444,
+      "step": 4297
+    },
+    {
+      "epoch": 1.4668941979522185,
+      "grad_norm": 3.0714824199676514,
+      "learning_rate": 0.0005110352673492605,
+      "loss": 6.2691,
+      "step": 4298
+    },
+    {
+      "epoch": 1.467235494880546,
+      "grad_norm": 2.927114963531494,
+      "learning_rate": 0.0005109215017064846,
+      "loss": 6.2727,
+      "step": 4299
+    },
+    {
+      "epoch": 1.4675767918088738,
+      "grad_norm": 2.885935068130493,
+      "learning_rate": 0.0005108077360637087,
+      "loss": 6.4839,
+      "step": 4300
+    },
+    {
+      "epoch": 1.4679180887372014,
+      "grad_norm": 2.985340118408203,
+      "learning_rate": 0.000510693970420933,
+      "loss": 6.5858,
+      "step": 4301
+    },
+    {
+      "epoch": 1.468259385665529,
+      "grad_norm": 3.0189003944396973,
+      "learning_rate": 0.0005105802047781571,
+      "loss": 6.4866,
+      "step": 4302
+    },
+    {
+      "epoch": 1.4686006825938567,
+      "grad_norm": 3.324467658996582,
+      "learning_rate": 0.0005104664391353812,
+      "loss": 6.1826,
+      "step": 4303
+    },
+    {
+      "epoch": 1.4689419795221843,
+      "grad_norm": 3.1696572303771973,
+      "learning_rate": 0.0005103526734926053,
+      "loss": 6.7885,
+      "step": 4304
+    },
+    {
+      "epoch": 1.469283276450512,
+      "grad_norm": 3.074852705001831,
+      "learning_rate": 0.0005102389078498294,
+      "loss": 6.6288,
+      "step": 4305
+    },
+    {
+      "epoch": 1.4696245733788396,
+      "grad_norm": 2.911682367324829,
+      "learning_rate": 0.0005101251422070535,
+      "loss": 6.4831,
+      "step": 4306
+    },
+    {
+      "epoch": 1.4699658703071672,
+      "grad_norm": 3.0853192806243896,
+      "learning_rate": 0.0005100113765642777,
+      "loss": 6.6247,
+      "step": 4307
+    },
+    {
+      "epoch": 1.4703071672354948,
+      "grad_norm": 3.0632660388946533,
+      "learning_rate": 0.0005098976109215017,
+      "loss": 6.4772,
+      "step": 4308
+    },
+    {
+      "epoch": 1.4706484641638224,
+      "grad_norm": 3.191326379776001,
+      "learning_rate": 0.0005097838452787258,
+      "loss": 6.1043,
+      "step": 4309
+    },
+    {
+      "epoch": 1.47098976109215,
+      "grad_norm": 2.7078959941864014,
+      "learning_rate": 0.0005096700796359499,
+      "loss": 3.3701,
+      "step": 4310
+    },
+    {
+      "epoch": 1.471331058020478,
+      "grad_norm": 3.1352057456970215,
+      "learning_rate": 0.000509556313993174,
+      "loss": 6.237,
+      "step": 4311
+    },
+    {
+      "epoch": 1.4716723549488053,
+      "grad_norm": 3.192700147628784,
+      "learning_rate": 0.0005094425483503981,
+      "loss": 6.6129,
+      "step": 4312
+    },
+    {
+      "epoch": 1.4720136518771332,
+      "grad_norm": 2.962541103363037,
+      "learning_rate": 0.0005093287827076223,
+      "loss": 5.8705,
+      "step": 4313
+    },
+    {
+      "epoch": 1.4723549488054608,
+      "grad_norm": 3.1734907627105713,
+      "learning_rate": 0.0005092150170648464,
+      "loss": 5.8751,
+      "step": 4314
+    },
+    {
+      "epoch": 1.4726962457337884,
+      "grad_norm": 5.9300456047058105,
+      "learning_rate": 0.0005091012514220705,
+      "loss": 5.5791,
+      "step": 4315
+    },
+    {
+      "epoch": 1.473037542662116,
+      "grad_norm": 2.9836061000823975,
+      "learning_rate": 0.0005089874857792946,
+      "loss": 6.6064,
+      "step": 4316
+    },
+    {
+      "epoch": 1.4733788395904437,
+      "grad_norm": 3.1486117839813232,
+      "learning_rate": 0.0005088737201365187,
+      "loss": 6.6439,
+      "step": 4317
+    },
+    {
+      "epoch": 1.4737201365187713,
+      "grad_norm": 3.243680000305176,
+      "learning_rate": 0.000508759954493743,
+      "loss": 6.4473,
+      "step": 4318
+    },
+    {
+      "epoch": 1.474061433447099,
+      "grad_norm": 3.0161967277526855,
+      "learning_rate": 0.0005086461888509671,
+      "loss": 6.6844,
+      "step": 4319
+    },
+    {
+      "epoch": 1.4744027303754266,
+      "grad_norm": 3.0081169605255127,
+      "learning_rate": 0.0005085324232081912,
+      "loss": 6.537,
+      "step": 4320
+    },
+    {
+      "epoch": 1.4747440273037542,
+      "grad_norm": 2.991122245788574,
+      "learning_rate": 0.0005084186575654153,
+      "loss": 6.3377,
+      "step": 4321
+    },
+    {
+      "epoch": 1.4750853242320818,
+      "grad_norm": 3.0482964515686035,
+      "learning_rate": 0.0005083048919226394,
+      "loss": 6.6401,
+      "step": 4322
+    },
+    {
+      "epoch": 1.4754266211604095,
+      "grad_norm": 4.201125621795654,
+      "learning_rate": 0.0005081911262798635,
+      "loss": 6.4634,
+      "step": 4323
+    },
+    {
+      "epoch": 1.4757679180887373,
+      "grad_norm": 3.031925916671753,
+      "learning_rate": 0.0005080773606370877,
+      "loss": 6.3946,
+      "step": 4324
+    },
+    {
+      "epoch": 1.4761092150170647,
+      "grad_norm": 3.21101975440979,
+      "learning_rate": 0.0005079635949943118,
+      "loss": 6.258,
+      "step": 4325
+    },
+    {
+      "epoch": 1.4764505119453926,
+      "grad_norm": 3.451096534729004,
+      "learning_rate": 0.0005078498293515359,
+      "loss": 5.4511,
+      "step": 4326
+    },
+    {
+      "epoch": 1.4767918088737202,
+      "grad_norm": 3.0925445556640625,
+      "learning_rate": 0.0005077360637087599,
+      "loss": 6.7531,
+      "step": 4327
+    },
+    {
+      "epoch": 1.4771331058020478,
+      "grad_norm": 3.206573963165283,
+      "learning_rate": 0.000507622298065984,
+      "loss": 6.8944,
+      "step": 4328
+    },
+    {
+      "epoch": 1.4774744027303754,
+      "grad_norm": 3.1508631706237793,
+      "learning_rate": 0.0005075085324232081,
+      "loss": 6.8273,
+      "step": 4329
+    },
+    {
+      "epoch": 1.477815699658703,
+      "grad_norm": 3.3948967456817627,
+      "learning_rate": 0.0005073947667804323,
+      "loss": 5.824,
+      "step": 4330
+    },
+    {
+      "epoch": 1.4781569965870307,
+      "grad_norm": 3.199467182159424,
+      "learning_rate": 0.0005072810011376564,
+      "loss": 6.1532,
+      "step": 4331
+    },
+    {
+      "epoch": 1.4784982935153583,
+      "grad_norm": 3.0080020427703857,
+      "learning_rate": 0.0005071672354948805,
+      "loss": 6.5633,
+      "step": 4332
+    },
+    {
+      "epoch": 1.478839590443686,
+      "grad_norm": 3.230149030685425,
+      "learning_rate": 0.0005070534698521046,
+      "loss": 6.5214,
+      "step": 4333
+    },
+    {
+      "epoch": 1.4791808873720136,
+      "grad_norm": 3.4948723316192627,
+      "learning_rate": 0.0005069397042093287,
+      "loss": 5.9058,
+      "step": 4334
+    },
+    {
+      "epoch": 1.4795221843003412,
+      "grad_norm": 3.171333074569702,
+      "learning_rate": 0.000506825938566553,
+      "loss": 6.203,
+      "step": 4335
+    },
+    {
+      "epoch": 1.4798634812286688,
+      "grad_norm": 3.0840699672698975,
+      "learning_rate": 0.0005067121729237771,
+      "loss": 6.1753,
+      "step": 4336
+    },
+    {
+      "epoch": 1.4802047781569967,
+      "grad_norm": 3.062875747680664,
+      "learning_rate": 0.0005065984072810012,
+      "loss": 6.574,
+      "step": 4337
+    },
+    {
+      "epoch": 1.480546075085324,
+      "grad_norm": 5.447630882263184,
+      "learning_rate": 0.0005064846416382253,
+      "loss": 4.2755,
+      "step": 4338
+    },
+    {
+      "epoch": 1.480887372013652,
+      "grad_norm": 3.1661298274993896,
+      "learning_rate": 0.0005063708759954494,
+      "loss": 6.4025,
+      "step": 4339
+    },
+    {
+      "epoch": 1.4812286689419796,
+      "grad_norm": 3.3090450763702393,
+      "learning_rate": 0.0005062571103526735,
+      "loss": 6.0994,
+      "step": 4340
+    },
+    {
+      "epoch": 1.4815699658703072,
+      "grad_norm": 3.69329571723938,
+      "learning_rate": 0.0005061433447098977,
+      "loss": 4.7961,
+      "step": 4341
+    },
+    {
+      "epoch": 1.4819112627986348,
+      "grad_norm": 3.3056480884552,
+      "learning_rate": 0.0005060295790671218,
+      "loss": 6.4119,
+      "step": 4342
+    },
+    {
+      "epoch": 1.4822525597269625,
+      "grad_norm": 10.146489143371582,
+      "learning_rate": 0.0005059158134243459,
+      "loss": 6.9681,
+      "step": 4343
+    },
+    {
+      "epoch": 1.48259385665529,
+      "grad_norm": 3.1964523792266846,
+      "learning_rate": 0.00050580204778157,
+      "loss": 6.9197,
+      "step": 4344
+    },
+    {
+      "epoch": 1.4829351535836177,
+      "grad_norm": 3.3806843757629395,
+      "learning_rate": 0.0005056882821387941,
+      "loss": 6.3696,
+      "step": 4345
+    },
+    {
+      "epoch": 1.4832764505119453,
+      "grad_norm": 3.172544002532959,
+      "learning_rate": 0.0005055745164960182,
+      "loss": 5.6425,
+      "step": 4346
+    },
+    {
+      "epoch": 1.483617747440273,
+      "grad_norm": 3.494082450866699,
+      "learning_rate": 0.0005054607508532423,
+      "loss": 6.5567,
+      "step": 4347
+    },
+    {
+      "epoch": 1.4839590443686006,
+      "grad_norm": 3.1376161575317383,
+      "learning_rate": 0.0005053469852104664,
+      "loss": 6.1425,
+      "step": 4348
+    },
+    {
+      "epoch": 1.4843003412969282,
+      "grad_norm": 4.658734321594238,
+      "learning_rate": 0.0005052332195676905,
+      "loss": 4.829,
+      "step": 4349
+    },
+    {
+      "epoch": 1.484641638225256,
+      "grad_norm": 3.0457332134246826,
+      "learning_rate": 0.0005051194539249146,
+      "loss": 6.6027,
+      "step": 4350
+    },
+    {
+      "epoch": 1.4849829351535835,
+      "grad_norm": 3.252115249633789,
+      "learning_rate": 0.0005050056882821387,
+      "loss": 6.4487,
+      "step": 4351
+    },
+    {
+      "epoch": 1.4853242320819113,
+      "grad_norm": 3.3878161907196045,
+      "learning_rate": 0.0005048919226393628,
+      "loss": 5.874,
+      "step": 4352
+    },
+    {
+      "epoch": 1.485665529010239,
+      "grad_norm": 3.082244873046875,
+      "learning_rate": 0.0005047781569965871,
+      "loss": 6.3979,
+      "step": 4353
+    },
+    {
+      "epoch": 1.4860068259385666,
+      "grad_norm": 3.0858993530273438,
+      "learning_rate": 0.0005046643913538112,
+      "loss": 6.2685,
+      "step": 4354
+    },
+    {
+      "epoch": 1.4863481228668942,
+      "grad_norm": 3.532917022705078,
+      "learning_rate": 0.0005045506257110353,
+      "loss": 6.3543,
+      "step": 4355
+    },
+    {
+      "epoch": 1.4866894197952218,
+      "grad_norm": 3.1342530250549316,
+      "learning_rate": 0.0005044368600682594,
+      "loss": 5.8409,
+      "step": 4356
+    },
+    {
+      "epoch": 1.4870307167235495,
+      "grad_norm": 3.64797306060791,
+      "learning_rate": 0.0005043230944254835,
+      "loss": 6.3942,
+      "step": 4357
+    },
+    {
+      "epoch": 1.487372013651877,
+      "grad_norm": 4.409688472747803,
+      "learning_rate": 0.0005042093287827077,
+      "loss": 5.6323,
+      "step": 4358
+    },
+    {
+      "epoch": 1.4877133105802047,
+      "grad_norm": 3.2418742179870605,
+      "learning_rate": 0.0005040955631399318,
+      "loss": 6.3286,
+      "step": 4359
+    },
+    {
+      "epoch": 1.4880546075085324,
+      "grad_norm": 3.112736225128174,
+      "learning_rate": 0.0005039817974971559,
+      "loss": 6.2084,
+      "step": 4360
+    },
+    {
+      "epoch": 1.4883959044368602,
+      "grad_norm": 3.1486265659332275,
+      "learning_rate": 0.00050386803185438,
+      "loss": 6.2417,
+      "step": 4361
+    },
+    {
+      "epoch": 1.4887372013651876,
+      "grad_norm": 5.536955833435059,
+      "learning_rate": 0.0005037542662116041,
+      "loss": 5.6501,
+      "step": 4362
+    },
+    {
+      "epoch": 1.4890784982935155,
+      "grad_norm": 3.1049556732177734,
+      "learning_rate": 0.0005036405005688282,
+      "loss": 6.3247,
+      "step": 4363
+    },
+    {
+      "epoch": 1.4894197952218429,
+      "grad_norm": 3.001415252685547,
+      "learning_rate": 0.0005035267349260524,
+      "loss": 6.7637,
+      "step": 4364
+    },
+    {
+      "epoch": 1.4897610921501707,
+      "grad_norm": 2.8671603202819824,
+      "learning_rate": 0.0005034129692832765,
+      "loss": 6.3853,
+      "step": 4365
+    },
+    {
+      "epoch": 1.4901023890784983,
+      "grad_norm": 4.22214937210083,
+      "learning_rate": 0.0005032992036405005,
+      "loss": 6.064,
+      "step": 4366
+    },
+    {
+      "epoch": 1.490443686006826,
+      "grad_norm": 3.110260009765625,
+      "learning_rate": 0.0005031854379977246,
+      "loss": 6.2449,
+      "step": 4367
+    },
+    {
+      "epoch": 1.4907849829351536,
+      "grad_norm": 3.0448741912841797,
+      "learning_rate": 0.0005030716723549487,
+      "loss": 6.6455,
+      "step": 4368
+    },
+    {
+      "epoch": 1.4911262798634812,
+      "grad_norm": 3.040241003036499,
+      "learning_rate": 0.0005029579067121728,
+      "loss": 6.2379,
+      "step": 4369
+    },
+    {
+      "epoch": 1.4914675767918089,
+      "grad_norm": 3.210784912109375,
+      "learning_rate": 0.0005028441410693971,
+      "loss": 5.9151,
+      "step": 4370
+    },
+    {
+      "epoch": 1.4918088737201365,
+      "grad_norm": 4.361735820770264,
+      "learning_rate": 0.0005027303754266212,
+      "loss": 5.6326,
+      "step": 4371
+    },
+    {
+      "epoch": 1.4921501706484641,
+      "grad_norm": 3.1003541946411133,
+      "learning_rate": 0.0005026166097838453,
+      "loss": 6.7568,
+      "step": 4372
+    },
+    {
+      "epoch": 1.4924914675767917,
+      "grad_norm": 4.7033281326293945,
+      "learning_rate": 0.0005025028441410694,
+      "loss": 5.7138,
+      "step": 4373
+    },
+    {
+      "epoch": 1.4928327645051196,
+      "grad_norm": 3.1924164295196533,
+      "learning_rate": 0.0005023890784982935,
+      "loss": 6.2701,
+      "step": 4374
+    },
+    {
+      "epoch": 1.493174061433447,
+      "grad_norm": 3.1967172622680664,
+      "learning_rate": 0.0005022753128555177,
+      "loss": 6.2449,
+      "step": 4375
+    },
+    {
+      "epoch": 1.4935153583617748,
+      "grad_norm": 3.1058359146118164,
+      "learning_rate": 0.0005021615472127418,
+      "loss": 6.567,
+      "step": 4376
+    },
+    {
+      "epoch": 1.4938566552901023,
+      "grad_norm": 3.0760414600372314,
+      "learning_rate": 0.0005020477815699659,
+      "loss": 6.221,
+      "step": 4377
+    },
+    {
+      "epoch": 1.49419795221843,
+      "grad_norm": 3.031233549118042,
+      "learning_rate": 0.00050193401592719,
+      "loss": 6.3645,
+      "step": 4378
+    },
+    {
+      "epoch": 1.4945392491467577,
+      "grad_norm": 2.9455747604370117,
+      "learning_rate": 0.0005018202502844141,
+      "loss": 6.3219,
+      "step": 4379
+    },
+    {
+      "epoch": 1.4948805460750854,
+      "grad_norm": 3.113830804824829,
+      "learning_rate": 0.0005017064846416382,
+      "loss": 6.1877,
+      "step": 4380
+    },
+    {
+      "epoch": 1.495221843003413,
+      "grad_norm": 2.8928475379943848,
+      "learning_rate": 0.0005015927189988624,
+      "loss": 6.0893,
+      "step": 4381
+    },
+    {
+      "epoch": 1.4955631399317406,
+      "grad_norm": 5.041830539703369,
+      "learning_rate": 0.0005014789533560865,
+      "loss": 5.6682,
+      "step": 4382
+    },
+    {
+      "epoch": 1.4959044368600682,
+      "grad_norm": 3.120551109313965,
+      "learning_rate": 0.0005013651877133106,
+      "loss": 6.5289,
+      "step": 4383
+    },
+    {
+      "epoch": 1.4962457337883959,
+      "grad_norm": 3.156862735748291,
+      "learning_rate": 0.0005012514220705347,
+      "loss": 6.1597,
+      "step": 4384
+    },
+    {
+      "epoch": 1.4965870307167235,
+      "grad_norm": 2.876394510269165,
+      "learning_rate": 0.0005011376564277589,
+      "loss": 3.1941,
+      "step": 4385
+    },
+    {
+      "epoch": 1.4969283276450511,
+      "grad_norm": 3.063058376312256,
+      "learning_rate": 0.0005010238907849828,
+      "loss": 6.0953,
+      "step": 4386
+    },
+    {
+      "epoch": 1.497269624573379,
+      "grad_norm": 3.1170380115509033,
+      "learning_rate": 0.0005009101251422071,
+      "loss": 6.3125,
+      "step": 4387
+    },
+    {
+      "epoch": 1.4976109215017064,
+      "grad_norm": 5.3596062660217285,
+      "learning_rate": 0.0005007963594994312,
+      "loss": 4.3308,
+      "step": 4388
+    },
+    {
+      "epoch": 1.4979522184300342,
+      "grad_norm": 3.159137487411499,
+      "learning_rate": 0.0005006825938566553,
+      "loss": 6.5011,
+      "step": 4389
+    },
+    {
+      "epoch": 1.4982935153583616,
+      "grad_norm": 3.204058885574341,
+      "learning_rate": 0.0005005688282138794,
+      "loss": 6.6059,
+      "step": 4390
+    },
+    {
+      "epoch": 1.4986348122866895,
+      "grad_norm": 3.0355639457702637,
+      "learning_rate": 0.0005004550625711035,
+      "loss": 6.5694,
+      "step": 4391
+    },
+    {
+      "epoch": 1.4989761092150171,
+      "grad_norm": 3.0069313049316406,
+      "learning_rate": 0.0005003412969283276,
+      "loss": 6.1367,
+      "step": 4392
+    },
+    {
+      "epoch": 1.4993174061433447,
+      "grad_norm": 3.0549659729003906,
+      "learning_rate": 0.0005002275312855518,
+      "loss": 6.1733,
+      "step": 4393
+    },
+    {
+      "epoch": 1.4996587030716724,
+      "grad_norm": 3.0711121559143066,
+      "learning_rate": 0.0005001137656427759,
+      "loss": 6.0055,
+      "step": 4394
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 3.120171546936035,
+      "learning_rate": 0.0005,
+      "loss": 6.6731,
+      "step": 4395
+    },
+    {
+      "epoch": 1.5003412969283276,
+      "grad_norm": 3.0403029918670654,
+      "learning_rate": 0.0004998862343572241,
+      "loss": 6.1493,
+      "step": 4396
+    },
+    {
+      "epoch": 1.5006825938566553,
+      "grad_norm": 3.202802896499634,
+      "learning_rate": 0.0004997724687144482,
+      "loss": 6.1068,
+      "step": 4397
+    },
+    {
+      "epoch": 1.5010238907849829,
+      "grad_norm": 3.0703487396240234,
+      "learning_rate": 0.0004996587030716724,
+      "loss": 6.4405,
+      "step": 4398
+    },
+    {
+      "epoch": 1.5013651877133105,
+      "grad_norm": 3.1371710300445557,
+      "learning_rate": 0.0004995449374288965,
+      "loss": 6.5393,
+      "step": 4399
+    },
+    {
+      "epoch": 1.5017064846416384,
+      "grad_norm": 3.1856324672698975,
+      "learning_rate": 0.0004994311717861205,
+      "loss": 6.4713,
+      "step": 4400
+    },
+    {
+      "epoch": 1.5020477815699658,
+      "grad_norm": 3.43066143989563,
+      "learning_rate": 0.0004993174061433447,
+      "loss": 5.8309,
+      "step": 4401
+    },
+    {
+      "epoch": 1.5023890784982936,
+      "grad_norm": 3.037707567214966,
+      "learning_rate": 0.0004992036405005689,
+      "loss": 6.4918,
+      "step": 4402
+    },
+    {
+      "epoch": 1.502730375426621,
+      "grad_norm": 3.1267921924591064,
+      "learning_rate": 0.000499089874857793,
+      "loss": 6.2709,
+      "step": 4403
+    },
+    {
+      "epoch": 1.5030716723549489,
+      "grad_norm": 3.2483456134796143,
+      "learning_rate": 0.0004989761092150171,
+      "loss": 6.1242,
+      "step": 4404
+    },
+    {
+      "epoch": 1.5034129692832765,
+      "grad_norm": 3.104841947555542,
+      "learning_rate": 0.0004988623435722412,
+      "loss": 6.3647,
+      "step": 4405
+    },
+    {
+      "epoch": 1.5037542662116041,
+      "grad_norm": 3.147634267807007,
+      "learning_rate": 0.0004987485779294653,
+      "loss": 6.5279,
+      "step": 4406
+    },
+    {
+      "epoch": 1.5040955631399318,
+      "grad_norm": 3.1218855381011963,
+      "learning_rate": 0.0004986348122866895,
+      "loss": 6.1442,
+      "step": 4407
+    },
+    {
+      "epoch": 1.5044368600682594,
+      "grad_norm": 3.1170077323913574,
+      "learning_rate": 0.0004985210466439136,
+      "loss": 6.1911,
+      "step": 4408
+    },
+    {
+      "epoch": 1.504778156996587,
+      "grad_norm": 4.791984558105469,
+      "learning_rate": 0.0004984072810011377,
+      "loss": 5.5688,
+      "step": 4409
+    },
+    {
+      "epoch": 1.5051194539249146,
+      "grad_norm": 2.986837863922119,
+      "learning_rate": 0.0004982935153583618,
+      "loss": 6.2765,
+      "step": 4410
+    },
+    {
+      "epoch": 1.5054607508532423,
+      "grad_norm": 3.148416757583618,
+      "learning_rate": 0.0004981797497155859,
+      "loss": 6.2501,
+      "step": 4411
+    },
+    {
+      "epoch": 1.50580204778157,
+      "grad_norm": 3.0591113567352295,
+      "learning_rate": 0.00049806598407281,
+      "loss": 6.6508,
+      "step": 4412
+    },
+    {
+      "epoch": 1.5061433447098977,
+      "grad_norm": 3.104835271835327,
+      "learning_rate": 0.0004979522184300341,
+      "loss": 6.0843,
+      "step": 4413
+    },
+    {
+      "epoch": 1.5064846416382252,
+      "grad_norm": 3.147568464279175,
+      "learning_rate": 0.0004978384527872582,
+      "loss": 5.7625,
+      "step": 4414
+    },
+    {
+      "epoch": 1.506825938566553,
+      "grad_norm": 3.17414927482605,
+      "learning_rate": 0.0004977246871444823,
+      "loss": 5.7501,
+      "step": 4415
+    },
+    {
+      "epoch": 1.5071672354948804,
+      "grad_norm": 10.699484825134277,
+      "learning_rate": 0.0004976109215017065,
+      "loss": 5.8049,
+      "step": 4416
+    },
+    {
+      "epoch": 1.5075085324232083,
+      "grad_norm": 3.1057991981506348,
+      "learning_rate": 0.0004974971558589306,
+      "loss": 6.3801,
+      "step": 4417
+    },
+    {
+      "epoch": 1.5078498293515359,
+      "grad_norm": 3.1610498428344727,
+      "learning_rate": 0.0004973833902161547,
+      "loss": 6.239,
+      "step": 4418
+    },
+    {
+      "epoch": 1.5081911262798635,
+      "grad_norm": 3.110896348953247,
+      "learning_rate": 0.0004972696245733789,
+      "loss": 6.927,
+      "step": 4419
+    },
+    {
+      "epoch": 1.5085324232081911,
+      "grad_norm": 3.309873580932617,
+      "learning_rate": 0.000497155858930603,
+      "loss": 6.2068,
+      "step": 4420
+    },
+    {
+      "epoch": 1.5088737201365188,
+      "grad_norm": 2.9430341720581055,
+      "learning_rate": 0.0004970420932878271,
+      "loss": 6.6838,
+      "step": 4421
+    },
+    {
+      "epoch": 1.5092150170648464,
+      "grad_norm": 3.1343164443969727,
+      "learning_rate": 0.0004969283276450512,
+      "loss": 6.2543,
+      "step": 4422
+    },
+    {
+      "epoch": 1.509556313993174,
+      "grad_norm": 3.106687068939209,
+      "learning_rate": 0.0004968145620022753,
+      "loss": 6.4328,
+      "step": 4423
+    },
+    {
+      "epoch": 1.5098976109215017,
+      "grad_norm": 3.152050018310547,
+      "learning_rate": 0.0004967007963594995,
+      "loss": 5.92,
+      "step": 4424
+    },
+    {
+      "epoch": 1.5102389078498293,
+      "grad_norm": 3.016350507736206,
+      "learning_rate": 0.0004965870307167236,
+      "loss": 6.5876,
+      "step": 4425
+    },
+    {
+      "epoch": 1.5105802047781571,
+      "grad_norm": 3.782809257507324,
+      "learning_rate": 0.0004964732650739477,
+      "loss": 5.1762,
+      "step": 4426
+    },
+    {
+      "epoch": 1.5109215017064845,
+      "grad_norm": 3.22169828414917,
+      "learning_rate": 0.0004963594994311718,
+      "loss": 5.9286,
+      "step": 4427
+    },
+    {
+      "epoch": 1.5112627986348124,
+      "grad_norm": 3.2454440593719482,
+      "learning_rate": 0.0004962457337883959,
+      "loss": 6.5536,
+      "step": 4428
+    },
+    {
+      "epoch": 1.5116040955631398,
+      "grad_norm": 3.292848587036133,
+      "learning_rate": 0.00049613196814562,
+      "loss": 5.9237,
+      "step": 4429
+    },
+    {
+      "epoch": 1.5119453924914676,
+      "grad_norm": 3.0619893074035645,
+      "learning_rate": 0.0004960182025028441,
+      "loss": 6.0451,
+      "step": 4430
+    },
+    {
+      "epoch": 1.5122866894197953,
+      "grad_norm": 3.2105019092559814,
+      "learning_rate": 0.0004959044368600682,
+      "loss": 6.5059,
+      "step": 4431
+    },
+    {
+      "epoch": 1.512627986348123,
+      "grad_norm": 3.131542921066284,
+      "learning_rate": 0.0004957906712172923,
+      "loss": 6.6094,
+      "step": 4432
+    },
+    {
+      "epoch": 1.5129692832764505,
+      "grad_norm": 3.2492835521698,
+      "learning_rate": 0.0004956769055745165,
+      "loss": 6.0207,
+      "step": 4433
+    },
+    {
+      "epoch": 1.5133105802047782,
+      "grad_norm": 3.1075937747955322,
+      "learning_rate": 0.0004955631399317406,
+      "loss": 6.6052,
+      "step": 4434
+    },
+    {
+      "epoch": 1.5136518771331058,
+      "grad_norm": 3.096921443939209,
+      "learning_rate": 0.0004954493742889648,
+      "loss": 6.0225,
+      "step": 4435
+    },
+    {
+      "epoch": 1.5139931740614334,
+      "grad_norm": 3.1452181339263916,
+      "learning_rate": 0.0004953356086461889,
+      "loss": 5.4943,
+      "step": 4436
+    },
+    {
+      "epoch": 1.514334470989761,
+      "grad_norm": 3.2415482997894287,
+      "learning_rate": 0.000495221843003413,
+      "loss": 6.7057,
+      "step": 4437
+    },
+    {
+      "epoch": 1.5146757679180887,
+      "grad_norm": 3.020129442214966,
+      "learning_rate": 0.0004951080773606372,
+      "loss": 6.2966,
+      "step": 4438
+    },
+    {
+      "epoch": 1.5150170648464165,
+      "grad_norm": 3.014113426208496,
+      "learning_rate": 0.0004949943117178612,
+      "loss": 5.4332,
+      "step": 4439
+    },
+    {
+      "epoch": 1.515358361774744,
+      "grad_norm": 3.206845760345459,
+      "learning_rate": 0.0004948805460750853,
+      "loss": 5.9552,
+      "step": 4440
+    },
+    {
+      "epoch": 1.5156996587030718,
+      "grad_norm": 3.4275362491607666,
+      "learning_rate": 0.0004947667804323095,
+      "loss": 5.7084,
+      "step": 4441
+    },
+    {
+      "epoch": 1.5160409556313992,
+      "grad_norm": 3.051011085510254,
+      "learning_rate": 0.0004946530147895336,
+      "loss": 6.4459,
+      "step": 4442
+    },
+    {
+      "epoch": 1.516382252559727,
+      "grad_norm": 4.869009017944336,
+      "learning_rate": 0.0004945392491467577,
+      "loss": 5.4903,
+      "step": 4443
+    },
+    {
+      "epoch": 1.5167235494880547,
+      "grad_norm": 3.0409438610076904,
+      "learning_rate": 0.0004944254835039818,
+      "loss": 6.5647,
+      "step": 4444
+    },
+    {
+      "epoch": 1.5170648464163823,
+      "grad_norm": 3.4340035915374756,
+      "learning_rate": 0.0004943117178612059,
+      "loss": 6.1847,
+      "step": 4445
+    },
+    {
+      "epoch": 1.51740614334471,
+      "grad_norm": 3.105072021484375,
+      "learning_rate": 0.00049419795221843,
+      "loss": 6.2396,
+      "step": 4446
+    },
+    {
+      "epoch": 1.5177474402730375,
+      "grad_norm": 7.7960638999938965,
+      "learning_rate": 0.0004940841865756542,
+      "loss": 5.791,
+      "step": 4447
+    },
+    {
+      "epoch": 1.5180887372013652,
+      "grad_norm": 3.257929801940918,
+      "learning_rate": 0.0004939704209328783,
+      "loss": 6.3392,
+      "step": 4448
+    },
+    {
+      "epoch": 1.5184300341296928,
+      "grad_norm": 3.071336269378662,
+      "learning_rate": 0.0004938566552901023,
+      "loss": 6.1078,
+      "step": 4449
+    },
+    {
+      "epoch": 1.5187713310580204,
+      "grad_norm": 3.080676317214966,
+      "learning_rate": 0.0004937428896473265,
+      "loss": 6.3788,
+      "step": 4450
+    },
+    {
+      "epoch": 1.519112627986348,
+      "grad_norm": 3.079050064086914,
+      "learning_rate": 0.0004936291240045506,
+      "loss": 6.1288,
+      "step": 4451
+    },
+    {
+      "epoch": 1.519453924914676,
+      "grad_norm": 3.3807246685028076,
+      "learning_rate": 0.0004935153583617748,
+      "loss": 6.4449,
+      "step": 4452
+    },
+    {
+      "epoch": 1.5197952218430033,
+      "grad_norm": 4.4379730224609375,
+      "learning_rate": 0.0004934015927189989,
+      "loss": 6.0084,
+      "step": 4453
+    },
+    {
+      "epoch": 1.5201365187713312,
+      "grad_norm": 5.487371444702148,
+      "learning_rate": 0.000493287827076223,
+      "loss": 6.276,
+      "step": 4454
+    },
+    {
+      "epoch": 1.5204778156996586,
+      "grad_norm": 3.2696280479431152,
+      "learning_rate": 0.0004931740614334471,
+      "loss": 6.928,
+      "step": 4455
+    },
+    {
+      "epoch": 1.5208191126279864,
+      "grad_norm": 3.299424648284912,
+      "learning_rate": 0.0004930602957906713,
+      "loss": 5.6853,
+      "step": 4456
+    },
+    {
+      "epoch": 1.521160409556314,
+      "grad_norm": 3.520862340927124,
+      "learning_rate": 0.0004929465301478954,
+      "loss": 6.0821,
+      "step": 4457
+    },
+    {
+      "epoch": 1.5215017064846417,
+      "grad_norm": 3.142890691757202,
+      "learning_rate": 0.0004928327645051195,
+      "loss": 6.0354,
+      "step": 4458
+    },
+    {
+      "epoch": 1.5218430034129693,
+      "grad_norm": 2.9746880531311035,
+      "learning_rate": 0.0004927189988623436,
+      "loss": 6.4439,
+      "step": 4459
+    },
+    {
+      "epoch": 1.522184300341297,
+      "grad_norm": 4.289761066436768,
+      "learning_rate": 0.0004926052332195677,
+      "loss": 5.2062,
+      "step": 4460
+    },
+    {
+      "epoch": 1.5225255972696246,
+      "grad_norm": 3.730030059814453,
+      "learning_rate": 0.0004924914675767918,
+      "loss": 4.8441,
+      "step": 4461
+    },
+    {
+      "epoch": 1.5228668941979522,
+      "grad_norm": 3.0463922023773193,
+      "learning_rate": 0.0004923777019340159,
+      "loss": 6.1014,
+      "step": 4462
+    },
+    {
+      "epoch": 1.5232081911262798,
+      "grad_norm": 3.1336638927459717,
+      "learning_rate": 0.00049226393629124,
+      "loss": 6.3418,
+      "step": 4463
+    },
+    {
+      "epoch": 1.5235494880546074,
+      "grad_norm": 3.0351781845092773,
+      "learning_rate": 0.0004921501706484642,
+      "loss": 6.3558,
+      "step": 4464
+    },
+    {
+      "epoch": 1.5238907849829353,
+      "grad_norm": 3.130140542984009,
+      "learning_rate": 0.0004920364050056883,
+      "loss": 6.5153,
+      "step": 4465
+    },
+    {
+      "epoch": 1.5242320819112627,
+      "grad_norm": 3.104055404663086,
+      "learning_rate": 0.0004919226393629124,
+      "loss": 6.4679,
+      "step": 4466
+    },
+    {
+      "epoch": 1.5245733788395905,
+      "grad_norm": 4.129725933074951,
+      "learning_rate": 0.0004918088737201365,
+      "loss": 5.9675,
+      "step": 4467
+    },
+    {
+      "epoch": 1.524914675767918,
+      "grad_norm": 3.040032386779785,
+      "learning_rate": 0.0004916951080773606,
+      "loss": 6.476,
+      "step": 4468
+    },
+    {
+      "epoch": 1.5252559726962458,
+      "grad_norm": 2.9849636554718018,
+      "learning_rate": 0.0004915813424345848,
+      "loss": 6.3094,
+      "step": 4469
+    },
+    {
+      "epoch": 1.5255972696245734,
+      "grad_norm": 3.130922794342041,
+      "learning_rate": 0.0004914675767918089,
+      "loss": 6.2869,
+      "step": 4470
+    },
+    {
+      "epoch": 1.525938566552901,
+      "grad_norm": 3.0397064685821533,
+      "learning_rate": 0.000491353811149033,
+      "loss": 6.6056,
+      "step": 4471
+    },
+    {
+      "epoch": 1.5262798634812287,
+      "grad_norm": 3.0539159774780273,
+      "learning_rate": 0.0004912400455062571,
+      "loss": 6.4248,
+      "step": 4472
+    },
+    {
+      "epoch": 1.5266211604095563,
+      "grad_norm": 2.9415955543518066,
+      "learning_rate": 0.0004911262798634813,
+      "loss": 6.1774,
+      "step": 4473
+    },
+    {
+      "epoch": 1.526962457337884,
+      "grad_norm": 3.0912396907806396,
+      "learning_rate": 0.0004910125142207054,
+      "loss": 6.3592,
+      "step": 4474
+    },
+    {
+      "epoch": 1.5273037542662116,
+      "grad_norm": 3.2052664756774902,
+      "learning_rate": 0.0004908987485779295,
+      "loss": 5.9968,
+      "step": 4475
+    },
+    {
+      "epoch": 1.5276450511945392,
+      "grad_norm": 3.073906421661377,
+      "learning_rate": 0.0004907849829351536,
+      "loss": 6.3548,
+      "step": 4476
+    },
+    {
+      "epoch": 1.5279863481228668,
+      "grad_norm": 3.0415360927581787,
+      "learning_rate": 0.0004906712172923777,
+      "loss": 5.5661,
+      "step": 4477
+    },
+    {
+      "epoch": 1.5283276450511947,
+      "grad_norm": 2.9644320011138916,
+      "learning_rate": 0.0004905574516496018,
+      "loss": 5.781,
+      "step": 4478
+    },
+    {
+      "epoch": 1.528668941979522,
+      "grad_norm": 3.074263095855713,
+      "learning_rate": 0.0004904436860068259,
+      "loss": 6.4673,
+      "step": 4479
+    },
+    {
+      "epoch": 1.52901023890785,
+      "grad_norm": 3.1107773780822754,
+      "learning_rate": 0.00049032992036405,
+      "loss": 6.2733,
+      "step": 4480
+    },
+    {
+      "epoch": 1.5293515358361773,
+      "grad_norm": 3.262098550796509,
+      "learning_rate": 0.0004902161547212742,
+      "loss": 5.7654,
+      "step": 4481
+    },
+    {
+      "epoch": 1.5296928327645052,
+      "grad_norm": 7.093457221984863,
+      "learning_rate": 0.0004901023890784983,
+      "loss": 5.9941,
+      "step": 4482
+    },
+    {
+      "epoch": 1.5300341296928328,
+      "grad_norm": 3.19046950340271,
+      "learning_rate": 0.0004899886234357224,
+      "loss": 6.786,
+      "step": 4483
+    },
+    {
+      "epoch": 1.5303754266211604,
+      "grad_norm": 3.732973337173462,
+      "learning_rate": 0.0004898748577929465,
+      "loss": 5.1743,
+      "step": 4484
+    },
+    {
+      "epoch": 1.530716723549488,
+      "grad_norm": 3.1745400428771973,
+      "learning_rate": 0.0004897610921501706,
+      "loss": 5.9867,
+      "step": 4485
+    },
+    {
+      "epoch": 1.5310580204778157,
+      "grad_norm": 3.155008554458618,
+      "learning_rate": 0.0004896473265073948,
+      "loss": 6.6222,
+      "step": 4486
+    },
+    {
+      "epoch": 1.5313993174061433,
+      "grad_norm": 4.09003210067749,
+      "learning_rate": 0.000489533560864619,
+      "loss": 5.6898,
+      "step": 4487
+    },
+    {
+      "epoch": 1.531740614334471,
+      "grad_norm": 4.12489128112793,
+      "learning_rate": 0.000489419795221843,
+      "loss": 4.9424,
+      "step": 4488
+    },
+    {
+      "epoch": 1.5320819112627988,
+      "grad_norm": 3.1861050128936768,
+      "learning_rate": 0.0004893060295790671,
+      "loss": 6.032,
+      "step": 4489
+    },
+    {
+      "epoch": 1.5324232081911262,
+      "grad_norm": 3.163435935974121,
+      "learning_rate": 0.0004891922639362913,
+      "loss": 6.3353,
+      "step": 4490
+    },
+    {
+      "epoch": 1.532764505119454,
+      "grad_norm": 2.9932167530059814,
+      "learning_rate": 0.0004890784982935154,
+      "loss": 6.5483,
+      "step": 4491
+    },
+    {
+      "epoch": 1.5331058020477815,
+      "grad_norm": 3.199826955795288,
+      "learning_rate": 0.0004889647326507395,
+      "loss": 6.1149,
+      "step": 4492
+    },
+    {
+      "epoch": 1.5334470989761093,
+      "grad_norm": 3.0402212142944336,
+      "learning_rate": 0.0004888509670079636,
+      "loss": 6.484,
+      "step": 4493
+    },
+    {
+      "epoch": 1.5337883959044367,
+      "grad_norm": 3.0578010082244873,
+      "learning_rate": 0.0004887372013651877,
+      "loss": 6.7166,
+      "step": 4494
+    },
+    {
+      "epoch": 1.5341296928327646,
+      "grad_norm": 3.0253403186798096,
+      "learning_rate": 0.0004886234357224118,
+      "loss": 6.8642,
+      "step": 4495
+    },
+    {
+      "epoch": 1.5344709897610922,
+      "grad_norm": 3.225531578063965,
+      "learning_rate": 0.000488509670079636,
+      "loss": 5.9513,
+      "step": 4496
+    },
+    {
+      "epoch": 1.5348122866894198,
+      "grad_norm": 3.0968496799468994,
+      "learning_rate": 0.0004883959044368601,
+      "loss": 6.327,
+      "step": 4497
+    },
+    {
+      "epoch": 1.5351535836177475,
+      "grad_norm": 3.0139734745025635,
+      "learning_rate": 0.0004882821387940841,
+      "loss": 6.8722,
+      "step": 4498
+    },
+    {
+      "epoch": 1.535494880546075,
+      "grad_norm": 3.061629056930542,
+      "learning_rate": 0.0004881683731513083,
+      "loss": 6.5616,
+      "step": 4499
+    },
+    {
+      "epoch": 1.5358361774744027,
+      "grad_norm": 3.062117338180542,
+      "learning_rate": 0.00048805460750853244,
+      "loss": 6.8498,
+      "step": 4500
+    },
+    {
+      "epoch": 1.5361774744027303,
+      "grad_norm": 3.0551111698150635,
+      "learning_rate": 0.00048794084186575654,
+      "loss": 6.2062,
+      "step": 4501
+    },
+    {
+      "epoch": 1.5365187713310582,
+      "grad_norm": 3.084839105606079,
+      "learning_rate": 0.00048782707622298065,
+      "loss": 6.1916,
+      "step": 4502
+    },
+    {
+      "epoch": 1.5368600682593856,
+      "grad_norm": 2.976888418197632,
+      "learning_rate": 0.0004877133105802048,
+      "loss": 6.6639,
+      "step": 4503
+    },
+    {
+      "epoch": 1.5372013651877134,
+      "grad_norm": 2.9720981121063232,
+      "learning_rate": 0.0004875995449374289,
+      "loss": 6.4217,
+      "step": 4504
+    },
+    {
+      "epoch": 1.5375426621160408,
+      "grad_norm": 3.941185712814331,
+      "learning_rate": 0.000487485779294653,
+      "loss": 5.9576,
+      "step": 4505
+    },
+    {
+      "epoch": 1.5378839590443687,
+      "grad_norm": 3.0946285724639893,
+      "learning_rate": 0.0004873720136518772,
+      "loss": 6.3165,
+      "step": 4506
+    },
+    {
+      "epoch": 1.538225255972696,
+      "grad_norm": 3.0190694332122803,
+      "learning_rate": 0.00048725824800910123,
+      "loss": 6.4502,
+      "step": 4507
+    },
+    {
+      "epoch": 1.538566552901024,
+      "grad_norm": 3.0796146392822266,
+      "learning_rate": 0.00048714448236632533,
+      "loss": 6.1499,
+      "step": 4508
+    },
+    {
+      "epoch": 1.5389078498293516,
+      "grad_norm": 4.050348281860352,
+      "learning_rate": 0.0004870307167235495,
+      "loss": 6.0962,
+      "step": 4509
+    },
+    {
+      "epoch": 1.5392491467576792,
+      "grad_norm": 3.1888248920440674,
+      "learning_rate": 0.0004869169510807736,
+      "loss": 6.0087,
+      "step": 4510
+    },
+    {
+      "epoch": 1.5395904436860068,
+      "grad_norm": 2.9527971744537354,
+      "learning_rate": 0.00048680318543799776,
+      "loss": 6.3688,
+      "step": 4511
+    },
+    {
+      "epoch": 1.5399317406143345,
+      "grad_norm": 3.0263445377349854,
+      "learning_rate": 0.00048668941979522186,
+      "loss": 5.7253,
+      "step": 4512
+    },
+    {
+      "epoch": 1.540273037542662,
+      "grad_norm": 3.0024619102478027,
+      "learning_rate": 0.00048657565415244597,
+      "loss": 6.0979,
+      "step": 4513
+    },
+    {
+      "epoch": 1.5406143344709897,
+      "grad_norm": 3.0450499057769775,
+      "learning_rate": 0.0004864618885096701,
+      "loss": 6.0442,
+      "step": 4514
+    },
+    {
+      "epoch": 1.5409556313993176,
+      "grad_norm": 3.137784242630005,
+      "learning_rate": 0.00048634812286689423,
+      "loss": 6.4424,
+      "step": 4515
+    },
+    {
+      "epoch": 1.541296928327645,
+      "grad_norm": 3.5526490211486816,
+      "learning_rate": 0.00048623435722411833,
+      "loss": 6.2877,
+      "step": 4516
+    },
+    {
+      "epoch": 1.5416382252559728,
+      "grad_norm": 3.030540704727173,
+      "learning_rate": 0.00048612059158134244,
+      "loss": 6.2586,
+      "step": 4517
+    },
+    {
+      "epoch": 1.5419795221843002,
+      "grad_norm": 3.2171597480773926,
+      "learning_rate": 0.00048600682593856654,
+      "loss": 6.0998,
+      "step": 4518
+    },
+    {
+      "epoch": 1.542320819112628,
+      "grad_norm": 2.999274730682373,
+      "learning_rate": 0.00048589306029579065,
+      "loss": 6.0791,
+      "step": 4519
+    },
+    {
+      "epoch": 1.5426621160409555,
+      "grad_norm": 3.330794334411621,
+      "learning_rate": 0.0004857792946530148,
+      "loss": 5.9229,
+      "step": 4520
+    },
+    {
+      "epoch": 1.5430034129692833,
+      "grad_norm": 3.1574463844299316,
+      "learning_rate": 0.0004856655290102389,
+      "loss": 5.7773,
+      "step": 4521
+    },
+    {
+      "epoch": 1.543344709897611,
+      "grad_norm": 3.089190721511841,
+      "learning_rate": 0.000485551763367463,
+      "loss": 6.2909,
+      "step": 4522
+    },
+    {
+      "epoch": 1.5436860068259386,
+      "grad_norm": 2.9726176261901855,
+      "learning_rate": 0.0004854379977246872,
+      "loss": 5.816,
+      "step": 4523
+    },
+    {
+      "epoch": 1.5440273037542662,
+      "grad_norm": 3.124121904373169,
+      "learning_rate": 0.0004853242320819113,
+      "loss": 6.2375,
+      "step": 4524
+    },
+    {
+      "epoch": 1.5443686006825939,
+      "grad_norm": 3.3395678997039795,
+      "learning_rate": 0.0004852104664391354,
+      "loss": 6.0664,
+      "step": 4525
+    },
+    {
+      "epoch": 1.5447098976109215,
+      "grad_norm": 3.0564520359039307,
+      "learning_rate": 0.00048509670079635955,
+      "loss": 6.4603,
+      "step": 4526
+    },
+    {
+      "epoch": 1.545051194539249,
+      "grad_norm": 3.5565192699432373,
+      "learning_rate": 0.0004849829351535836,
+      "loss": 6.0142,
+      "step": 4527
+    },
+    {
+      "epoch": 1.545392491467577,
+      "grad_norm": 3.0865042209625244,
+      "learning_rate": 0.0004848691695108077,
+      "loss": 6.3314,
+      "step": 4528
+    },
+    {
+      "epoch": 1.5457337883959044,
+      "grad_norm": 4.123645782470703,
+      "learning_rate": 0.00048475540386803186,
+      "loss": 5.3,
+      "step": 4529
+    },
+    {
+      "epoch": 1.5460750853242322,
+      "grad_norm": 3.4055612087249756,
+      "learning_rate": 0.00048464163822525597,
+      "loss": 5.816,
+      "step": 4530
+    },
+    {
+      "epoch": 1.5464163822525596,
+      "grad_norm": 3.080965042114258,
+      "learning_rate": 0.0004845278725824801,
+      "loss": 6.9478,
+      "step": 4531
+    },
+    {
+      "epoch": 1.5467576791808875,
+      "grad_norm": 3.1426844596862793,
+      "learning_rate": 0.00048441410693970423,
+      "loss": 5.8243,
+      "step": 4532
+    },
+    {
+      "epoch": 1.5470989761092149,
+      "grad_norm": 3.0503451824188232,
+      "learning_rate": 0.00048430034129692834,
+      "loss": 6.595,
+      "step": 4533
+    },
+    {
+      "epoch": 1.5474402730375427,
+      "grad_norm": 2.9815714359283447,
+      "learning_rate": 0.0004841865756541525,
+      "loss": 6.7113,
+      "step": 4534
+    },
+    {
+      "epoch": 1.5477815699658704,
+      "grad_norm": 2.955928325653076,
+      "learning_rate": 0.0004840728100113766,
+      "loss": 6.4205,
+      "step": 4535
+    },
+    {
+      "epoch": 1.548122866894198,
+      "grad_norm": 3.0034239292144775,
+      "learning_rate": 0.00048395904436860065,
+      "loss": 6.4585,
+      "step": 4536
+    },
+    {
+      "epoch": 1.5484641638225256,
+      "grad_norm": 3.0914156436920166,
+      "learning_rate": 0.0004838452787258248,
+      "loss": 6.3329,
+      "step": 4537
+    },
+    {
+      "epoch": 1.5488054607508532,
+      "grad_norm": 3.089557409286499,
+      "learning_rate": 0.0004837315130830489,
+      "loss": 6.2133,
+      "step": 4538
+    },
+    {
+      "epoch": 1.5491467576791809,
+      "grad_norm": 3.246049642562866,
+      "learning_rate": 0.000483617747440273,
+      "loss": 6.5331,
+      "step": 4539
+    },
+    {
+      "epoch": 1.5494880546075085,
+      "grad_norm": 3.0475661754608154,
+      "learning_rate": 0.0004835039817974972,
+      "loss": 6.365,
+      "step": 4540
+    },
+    {
+      "epoch": 1.5498293515358363,
+      "grad_norm": 3.186208486557007,
+      "learning_rate": 0.0004833902161547213,
+      "loss": 6.1348,
+      "step": 4541
+    },
+    {
+      "epoch": 1.5501706484641637,
+      "grad_norm": 3.0234334468841553,
+      "learning_rate": 0.0004832764505119454,
+      "loss": 6.4846,
+      "step": 4542
+    },
+    {
+      "epoch": 1.5505119453924916,
+      "grad_norm": 7.129222869873047,
+      "learning_rate": 0.00048316268486916955,
+      "loss": 5.4173,
+      "step": 4543
+    },
+    {
+      "epoch": 1.550853242320819,
+      "grad_norm": 3.194537401199341,
+      "learning_rate": 0.00048304891922639365,
+      "loss": 6.3047,
+      "step": 4544
+    },
+    {
+      "epoch": 1.5511945392491469,
+      "grad_norm": 3.1084678173065186,
+      "learning_rate": 0.00048293515358361776,
+      "loss": 6.288,
+      "step": 4545
+    },
+    {
+      "epoch": 1.5515358361774743,
+      "grad_norm": 3.1638569831848145,
+      "learning_rate": 0.00048282138794084186,
+      "loss": 6.6077,
+      "step": 4546
+    },
+    {
+      "epoch": 1.551877133105802,
+      "grad_norm": 3.195992946624756,
+      "learning_rate": 0.00048270762229806597,
+      "loss": 6.6632,
+      "step": 4547
+    },
+    {
+      "epoch": 1.5522184300341297,
+      "grad_norm": 3.146247148513794,
+      "learning_rate": 0.00048259385665529007,
+      "loss": 6.5658,
+      "step": 4548
+    },
+    {
+      "epoch": 1.5525597269624574,
+      "grad_norm": 3.0369887351989746,
+      "learning_rate": 0.00048248009101251423,
+      "loss": 6.1692,
+      "step": 4549
+    },
+    {
+      "epoch": 1.552901023890785,
+      "grad_norm": 3.0904510021209717,
+      "learning_rate": 0.00048236632536973834,
+      "loss": 6.1782,
+      "step": 4550
+    },
+    {
+      "epoch": 1.5532423208191126,
+      "grad_norm": 3.021652936935425,
+      "learning_rate": 0.0004822525597269625,
+      "loss": 6.3469,
+      "step": 4551
+    },
+    {
+      "epoch": 1.5535836177474402,
+      "grad_norm": 3.161728858947754,
+      "learning_rate": 0.0004821387940841866,
+      "loss": 6.258,
+      "step": 4552
+    },
+    {
+      "epoch": 1.5539249146757679,
+      "grad_norm": 3.3773646354675293,
+      "learning_rate": 0.0004820250284414107,
+      "loss": 5.8451,
+      "step": 4553
+    },
+    {
+      "epoch": 1.5542662116040957,
+      "grad_norm": 3.128512144088745,
+      "learning_rate": 0.00048191126279863486,
+      "loss": 6.5491,
+      "step": 4554
+    },
+    {
+      "epoch": 1.5546075085324231,
+      "grad_norm": 3.0670344829559326,
+      "learning_rate": 0.00048179749715585897,
+      "loss": 7.086,
+      "step": 4555
+    },
+    {
+      "epoch": 1.554948805460751,
+      "grad_norm": 3.074653387069702,
+      "learning_rate": 0.000481683731513083,
+      "loss": 6.4795,
+      "step": 4556
+    },
+    {
+      "epoch": 1.5552901023890784,
+      "grad_norm": 3.1203551292419434,
+      "learning_rate": 0.0004815699658703072,
+      "loss": 6.3373,
+      "step": 4557
+    },
+    {
+      "epoch": 1.5556313993174062,
+      "grad_norm": 3.1555371284484863,
+      "learning_rate": 0.0004814562002275313,
+      "loss": 6.8241,
+      "step": 4558
+    },
+    {
+      "epoch": 1.5559726962457336,
+      "grad_norm": 3.1487646102905273,
+      "learning_rate": 0.0004813424345847554,
+      "loss": 6.4579,
+      "step": 4559
+    },
+    {
+      "epoch": 1.5563139931740615,
+      "grad_norm": 3.830792188644409,
+      "learning_rate": 0.00048122866894197955,
+      "loss": 6.1349,
+      "step": 4560
+    },
+    {
+      "epoch": 1.5566552901023891,
+      "grad_norm": 3.0813772678375244,
+      "learning_rate": 0.00048111490329920365,
+      "loss": 6.9713,
+      "step": 4561
+    },
+    {
+      "epoch": 1.5569965870307167,
+      "grad_norm": 3.344524621963501,
+      "learning_rate": 0.00048100113765642776,
+      "loss": 5.4983,
+      "step": 4562
+    },
+    {
+      "epoch": 1.5573378839590444,
+      "grad_norm": 3.2821524143218994,
+      "learning_rate": 0.0004808873720136519,
+      "loss": 6.4667,
+      "step": 4563
+    },
+    {
+      "epoch": 1.557679180887372,
+      "grad_norm": 3.1274075508117676,
+      "learning_rate": 0.000480773606370876,
+      "loss": 6.7131,
+      "step": 4564
+    },
+    {
+      "epoch": 1.5580204778156996,
+      "grad_norm": 2.94832444190979,
+      "learning_rate": 0.00048065984072810013,
+      "loss": 6.3848,
+      "step": 4565
+    },
+    {
+      "epoch": 1.5583617747440273,
+      "grad_norm": 3.0074613094329834,
+      "learning_rate": 0.00048054607508532423,
+      "loss": 6.2491,
+      "step": 4566
+    },
+    {
+      "epoch": 1.5587030716723551,
+      "grad_norm": 3.0959746837615967,
+      "learning_rate": 0.00048043230944254834,
+      "loss": 5.8382,
+      "step": 4567
+    },
+    {
+      "epoch": 1.5590443686006825,
+      "grad_norm": 3.2001736164093018,
+      "learning_rate": 0.00048031854379977244,
+      "loss": 6.5361,
+      "step": 4568
+    },
+    {
+      "epoch": 1.5593856655290104,
+      "grad_norm": 3.8832218647003174,
+      "learning_rate": 0.0004802047781569966,
+      "loss": 5.3809,
+      "step": 4569
+    },
+    {
+      "epoch": 1.5597269624573378,
+      "grad_norm": 3.133195638656616,
+      "learning_rate": 0.0004800910125142207,
+      "loss": 6.6976,
+      "step": 4570
+    },
+    {
+      "epoch": 1.5600682593856656,
+      "grad_norm": 3.2037956714630127,
+      "learning_rate": 0.00047997724687144487,
+      "loss": 5.5344,
+      "step": 4571
+    },
+    {
+      "epoch": 1.560409556313993,
+      "grad_norm": 3.157449722290039,
+      "learning_rate": 0.00047986348122866897,
+      "loss": 6.2814,
+      "step": 4572
+    },
+    {
+      "epoch": 1.5607508532423209,
+      "grad_norm": 3.062404155731201,
+      "learning_rate": 0.0004797497155858931,
+      "loss": 6.316,
+      "step": 4573
+    },
+    {
+      "epoch": 1.5610921501706485,
+      "grad_norm": 3.134361982345581,
+      "learning_rate": 0.00047963594994311723,
+      "loss": 6.5262,
+      "step": 4574
+    },
+    {
+      "epoch": 1.5614334470989761,
+      "grad_norm": 3.2360646724700928,
+      "learning_rate": 0.0004795221843003413,
+      "loss": 6.2296,
+      "step": 4575
+    },
+    {
+      "epoch": 1.5617747440273038,
+      "grad_norm": 3.012258529663086,
+      "learning_rate": 0.0004794084186575654,
+      "loss": 6.7743,
+      "step": 4576
+    },
+    {
+      "epoch": 1.5621160409556314,
+      "grad_norm": 3.5233166217803955,
+      "learning_rate": 0.00047929465301478955,
+      "loss": 5.9291,
+      "step": 4577
+    },
+    {
+      "epoch": 1.562457337883959,
+      "grad_norm": 4.878261566162109,
+      "learning_rate": 0.00047918088737201365,
+      "loss": 5.7263,
+      "step": 4578
+    },
+    {
+      "epoch": 1.5627986348122866,
+      "grad_norm": 3.228557825088501,
+      "learning_rate": 0.00047906712172923776,
+      "loss": 6.0776,
+      "step": 4579
+    },
+    {
+      "epoch": 1.5631399317406145,
+      "grad_norm": 3.1920437812805176,
+      "learning_rate": 0.0004789533560864619,
+      "loss": 6.7448,
+      "step": 4580
+    },
+    {
+      "epoch": 1.563481228668942,
+      "grad_norm": 3.1803622245788574,
+      "learning_rate": 0.000478839590443686,
+      "loss": 6.0921,
+      "step": 4581
+    },
+    {
+      "epoch": 1.5638225255972698,
+      "grad_norm": 5.843010902404785,
+      "learning_rate": 0.00047872582480091013,
+      "loss": 5.7007,
+      "step": 4582
+    },
+    {
+      "epoch": 1.5641638225255972,
+      "grad_norm": 3.6518936157226562,
+      "learning_rate": 0.0004786120591581343,
+      "loss": 6.3155,
+      "step": 4583
+    },
+    {
+      "epoch": 1.564505119453925,
+      "grad_norm": 3.08497953414917,
+      "learning_rate": 0.0004784982935153584,
+      "loss": 6.524,
+      "step": 4584
+    },
+    {
+      "epoch": 1.5648464163822524,
+      "grad_norm": 2.1860828399658203,
+      "learning_rate": 0.00047838452787258244,
+      "loss": 3.6702,
+      "step": 4585
+    },
+    {
+      "epoch": 1.5651877133105803,
+      "grad_norm": 3.152881145477295,
+      "learning_rate": 0.0004782707622298066,
+      "loss": 5.9726,
+      "step": 4586
+    },
+    {
+      "epoch": 1.565529010238908,
+      "grad_norm": 3.119457721710205,
+      "learning_rate": 0.0004781569965870307,
+      "loss": 5.8977,
+      "step": 4587
+    },
+    {
+      "epoch": 1.5658703071672355,
+      "grad_norm": 3.2837250232696533,
+      "learning_rate": 0.0004780432309442548,
+      "loss": 6.7914,
+      "step": 4588
+    },
+    {
+      "epoch": 1.5662116040955631,
+      "grad_norm": 3.0338499546051025,
+      "learning_rate": 0.00047792946530147897,
+      "loss": 6.6875,
+      "step": 4589
+    },
+    {
+      "epoch": 1.5665529010238908,
+      "grad_norm": 3.1285126209259033,
+      "learning_rate": 0.0004778156996587031,
+      "loss": 6.115,
+      "step": 4590
+    },
+    {
+      "epoch": 1.5668941979522184,
+      "grad_norm": 3.540517568588257,
+      "learning_rate": 0.00047770193401592724,
+      "loss": 5.9633,
+      "step": 4591
+    },
+    {
+      "epoch": 1.567235494880546,
+      "grad_norm": 3.079838991165161,
+      "learning_rate": 0.00047758816837315134,
+      "loss": 5.9609,
+      "step": 4592
+    },
+    {
+      "epoch": 1.5675767918088739,
+      "grad_norm": 3.100437879562378,
+      "learning_rate": 0.00047747440273037545,
+      "loss": 5.9144,
+      "step": 4593
+    },
+    {
+      "epoch": 1.5679180887372013,
+      "grad_norm": 3.044776439666748,
+      "learning_rate": 0.0004773606370875996,
+      "loss": 6.1149,
+      "step": 4594
+    },
+    {
+      "epoch": 1.5682593856655291,
+      "grad_norm": 3.157855749130249,
+      "learning_rate": 0.00047724687144482366,
+      "loss": 6.7326,
+      "step": 4595
+    },
+    {
+      "epoch": 1.5686006825938565,
+      "grad_norm": 6.666663646697998,
+      "learning_rate": 0.00047713310580204776,
+      "loss": 5.6815,
+      "step": 4596
+    },
+    {
+      "epoch": 1.5689419795221844,
+      "grad_norm": 3.1114113330841064,
+      "learning_rate": 0.0004770193401592719,
+      "loss": 5.6961,
+      "step": 4597
+    },
+    {
+      "epoch": 1.5692832764505118,
+      "grad_norm": 3.6068615913391113,
+      "learning_rate": 0.000476905574516496,
+      "loss": 4.9314,
+      "step": 4598
+    },
+    {
+      "epoch": 1.5696245733788396,
+      "grad_norm": 3.157696008682251,
+      "learning_rate": 0.00047679180887372013,
+      "loss": 6.3051,
+      "step": 4599
+    },
+    {
+      "epoch": 1.5699658703071673,
+      "grad_norm": 3.262202262878418,
+      "learning_rate": 0.0004766780432309443,
+      "loss": 6.2528,
+      "step": 4600
+    },
+    {
+      "epoch": 1.570307167235495,
+      "grad_norm": 3.1908304691314697,
+      "learning_rate": 0.0004765642775881684,
+      "loss": 6.1413,
+      "step": 4601
+    },
+    {
+      "epoch": 1.5706484641638225,
+      "grad_norm": 3.041872978210449,
+      "learning_rate": 0.0004764505119453925,
+      "loss": 6.4687,
+      "step": 4602
+    },
+    {
+      "epoch": 1.5709897610921502,
+      "grad_norm": 3.907120943069458,
+      "learning_rate": 0.00047633674630261666,
+      "loss": 5.4988,
+      "step": 4603
+    },
+    {
+      "epoch": 1.5713310580204778,
+      "grad_norm": 3.080669403076172,
+      "learning_rate": 0.00047622298065984076,
+      "loss": 6.2327,
+      "step": 4604
+    },
+    {
+      "epoch": 1.5716723549488054,
+      "grad_norm": 3.0005602836608887,
+      "learning_rate": 0.0004761092150170648,
+      "loss": 6.2388,
+      "step": 4605
+    },
+    {
+      "epoch": 1.5720136518771333,
+      "grad_norm": 3.291757345199585,
+      "learning_rate": 0.00047599544937428897,
+      "loss": 6.102,
+      "step": 4606
+    },
+    {
+      "epoch": 1.5723549488054607,
+      "grad_norm": 3.188426971435547,
+      "learning_rate": 0.0004758816837315131,
+      "loss": 6.7411,
+      "step": 4607
+    },
+    {
+      "epoch": 1.5726962457337885,
+      "grad_norm": 3.0288374423980713,
+      "learning_rate": 0.0004757679180887372,
+      "loss": 5.8925,
+      "step": 4608
+    },
+    {
+      "epoch": 1.573037542662116,
+      "grad_norm": 3.1033778190612793,
+      "learning_rate": 0.00047565415244596134,
+      "loss": 6.0647,
+      "step": 4609
+    },
+    {
+      "epoch": 1.5733788395904438,
+      "grad_norm": 3.0476388931274414,
+      "learning_rate": 0.00047554038680318545,
+      "loss": 6.4688,
+      "step": 4610
+    },
+    {
+      "epoch": 1.5737201365187712,
+      "grad_norm": 4.029483795166016,
+      "learning_rate": 0.00047542662116040955,
+      "loss": 5.5514,
+      "step": 4611
+    },
+    {
+      "epoch": 1.574061433447099,
+      "grad_norm": 3.1188085079193115,
+      "learning_rate": 0.0004753128555176337,
+      "loss": 6.6279,
+      "step": 4612
+    },
+    {
+      "epoch": 1.5744027303754267,
+      "grad_norm": 3.5747570991516113,
+      "learning_rate": 0.0004751990898748578,
+      "loss": 6.019,
+      "step": 4613
+    },
+    {
+      "epoch": 1.5747440273037543,
+      "grad_norm": 3.1913697719573975,
+      "learning_rate": 0.0004750853242320819,
+      "loss": 5.9978,
+      "step": 4614
+    },
+    {
+      "epoch": 1.575085324232082,
+      "grad_norm": 3.071608543395996,
+      "learning_rate": 0.000474971558589306,
+      "loss": 6.2824,
+      "step": 4615
+    },
+    {
+      "epoch": 1.5754266211604095,
+      "grad_norm": 3.110900402069092,
+      "learning_rate": 0.00047485779294653013,
+      "loss": 6.441,
+      "step": 4616
+    },
+    {
+      "epoch": 1.5757679180887372,
+      "grad_norm": 3.037877321243286,
+      "learning_rate": 0.0004747440273037543,
+      "loss": 6.5409,
+      "step": 4617
+    },
+    {
+      "epoch": 1.5761092150170648,
+      "grad_norm": 3.036573886871338,
+      "learning_rate": 0.0004746302616609784,
+      "loss": 6.5253,
+      "step": 4618
+    },
+    {
+      "epoch": 1.5764505119453927,
+      "grad_norm": 3.087519645690918,
+      "learning_rate": 0.0004745164960182025,
+      "loss": 6.2424,
+      "step": 4619
+    },
+    {
+      "epoch": 1.57679180887372,
+      "grad_norm": 2.925489902496338,
+      "learning_rate": 0.00047440273037542666,
+      "loss": 6.7424,
+      "step": 4620
+    },
+    {
+      "epoch": 1.577133105802048,
+      "grad_norm": 2.996990442276001,
+      "learning_rate": 0.00047428896473265076,
+      "loss": 6.6232,
+      "step": 4621
+    },
+    {
+      "epoch": 1.5774744027303753,
+      "grad_norm": 3.041271924972534,
+      "learning_rate": 0.00047417519908987487,
+      "loss": 6.5346,
+      "step": 4622
+    },
+    {
+      "epoch": 1.5778156996587032,
+      "grad_norm": 3.0396969318389893,
+      "learning_rate": 0.00047406143344709903,
+      "loss": 6.3334,
+      "step": 4623
+    },
+    {
+      "epoch": 1.5781569965870306,
+      "grad_norm": 2.9244205951690674,
+      "learning_rate": 0.0004739476678043231,
+      "loss": 6.4611,
+      "step": 4624
+    },
+    {
+      "epoch": 1.5784982935153584,
+      "grad_norm": 3.119096517562866,
+      "learning_rate": 0.0004738339021615472,
+      "loss": 6.2721,
+      "step": 4625
+    },
+    {
+      "epoch": 1.578839590443686,
+      "grad_norm": 3.0966577529907227,
+      "learning_rate": 0.00047372013651877134,
+      "loss": 6.0143,
+      "step": 4626
+    },
+    {
+      "epoch": 1.5791808873720137,
+      "grad_norm": 2.9753429889678955,
+      "learning_rate": 0.00047360637087599545,
+      "loss": 6.1939,
+      "step": 4627
+    },
+    {
+      "epoch": 1.5795221843003413,
+      "grad_norm": 3.2640163898468018,
+      "learning_rate": 0.00047349260523321955,
+      "loss": 6.7567,
+      "step": 4628
+    },
+    {
+      "epoch": 1.579863481228669,
+      "grad_norm": 3.492380380630493,
+      "learning_rate": 0.0004733788395904437,
+      "loss": 5.9374,
+      "step": 4629
+    },
+    {
+      "epoch": 1.5802047781569966,
+      "grad_norm": 3.118227481842041,
+      "learning_rate": 0.0004732650739476678,
+      "loss": 6.7009,
+      "step": 4630
+    },
+    {
+      "epoch": 1.5805460750853242,
+      "grad_norm": 3.0515055656433105,
+      "learning_rate": 0.0004731513083048919,
+      "loss": 6.1687,
+      "step": 4631
+    },
+    {
+      "epoch": 1.580887372013652,
+      "grad_norm": 2.957733154296875,
+      "learning_rate": 0.0004730375426621161,
+      "loss": 6.4641,
+      "step": 4632
+    },
+    {
+      "epoch": 1.5812286689419794,
+      "grad_norm": 3.1707279682159424,
+      "learning_rate": 0.0004729237770193402,
+      "loss": 5.8587,
+      "step": 4633
+    },
+    {
+      "epoch": 1.5815699658703073,
+      "grad_norm": 3.1099092960357666,
+      "learning_rate": 0.0004728100113765643,
+      "loss": 6.0575,
+      "step": 4634
+    },
+    {
+      "epoch": 1.5819112627986347,
+      "grad_norm": 2.9342041015625,
+      "learning_rate": 0.0004726962457337884,
+      "loss": 5.9659,
+      "step": 4635
+    },
+    {
+      "epoch": 1.5822525597269625,
+      "grad_norm": 3.138521909713745,
+      "learning_rate": 0.0004725824800910125,
+      "loss": 5.3131,
+      "step": 4636
+    },
+    {
+      "epoch": 1.58259385665529,
+      "grad_norm": 3.0649149417877197,
+      "learning_rate": 0.00047246871444823666,
+      "loss": 6.2565,
+      "step": 4637
+    },
+    {
+      "epoch": 1.5829351535836178,
+      "grad_norm": 5.445117950439453,
+      "learning_rate": 0.00047235494880546076,
+      "loss": 5.9185,
+      "step": 4638
+    },
+    {
+      "epoch": 1.5832764505119454,
+      "grad_norm": 3.1243233680725098,
+      "learning_rate": 0.00047224118316268487,
+      "loss": 6.2452,
+      "step": 4639
+    },
+    {
+      "epoch": 1.583617747440273,
+      "grad_norm": 3.1046507358551025,
+      "learning_rate": 0.00047212741751990903,
+      "loss": 6.2636,
+      "step": 4640
+    },
+    {
+      "epoch": 1.5839590443686007,
+      "grad_norm": 3.108381509780884,
+      "learning_rate": 0.00047201365187713313,
+      "loss": 6.7736,
+      "step": 4641
+    },
+    {
+      "epoch": 1.5843003412969283,
+      "grad_norm": 3.1656174659729004,
+      "learning_rate": 0.00047189988623435724,
+      "loss": 5.8736,
+      "step": 4642
+    },
+    {
+      "epoch": 1.584641638225256,
+      "grad_norm": 2.9965546131134033,
+      "learning_rate": 0.00047178612059158134,
+      "loss": 5.8275,
+      "step": 4643
+    },
+    {
+      "epoch": 1.5849829351535836,
+      "grad_norm": 2.997084617614746,
+      "learning_rate": 0.00047167235494880545,
+      "loss": 6.993,
+      "step": 4644
+    },
+    {
+      "epoch": 1.5853242320819114,
+      "grad_norm": 2.2224056720733643,
+      "learning_rate": 0.00047155858930602955,
+      "loss": 2.9852,
+      "step": 4645
+    },
+    {
+      "epoch": 1.5856655290102388,
+      "grad_norm": 3.1612727642059326,
+      "learning_rate": 0.0004714448236632537,
+      "loss": 6.3212,
+      "step": 4646
+    },
+    {
+      "epoch": 1.5860068259385667,
+      "grad_norm": 3.2868475914001465,
+      "learning_rate": 0.0004713310580204778,
+      "loss": 6.2314,
+      "step": 4647
+    },
+    {
+      "epoch": 1.586348122866894,
+      "grad_norm": 3.0663857460021973,
+      "learning_rate": 0.0004712172923777019,
+      "loss": 6.4269,
+      "step": 4648
+    },
+    {
+      "epoch": 1.586689419795222,
+      "grad_norm": 3.672306537628174,
+      "learning_rate": 0.0004711035267349261,
+      "loss": 5.4267,
+      "step": 4649
+    },
+    {
+      "epoch": 1.5870307167235493,
+      "grad_norm": 3.1437530517578125,
+      "learning_rate": 0.0004709897610921502,
+      "loss": 6.4256,
+      "step": 4650
+    },
+    {
+      "epoch": 1.5873720136518772,
+      "grad_norm": 3.077855348587036,
+      "learning_rate": 0.0004708759954493743,
+      "loss": 6.1129,
+      "step": 4651
+    },
+    {
+      "epoch": 1.5877133105802048,
+      "grad_norm": 3.188419818878174,
+      "learning_rate": 0.00047076222980659845,
+      "loss": 6.0257,
+      "step": 4652
+    },
+    {
+      "epoch": 1.5880546075085324,
+      "grad_norm": 2.986910581588745,
+      "learning_rate": 0.0004706484641638225,
+      "loss": 6.3658,
+      "step": 4653
+    },
+    {
+      "epoch": 1.58839590443686,
+      "grad_norm": 3.301227569580078,
+      "learning_rate": 0.00047053469852104666,
+      "loss": 5.64,
+      "step": 4654
+    },
+    {
+      "epoch": 1.5887372013651877,
+      "grad_norm": 4.034798622131348,
+      "learning_rate": 0.00047042093287827077,
+      "loss": 5.0341,
+      "step": 4655
+    },
+    {
+      "epoch": 1.5890784982935153,
+      "grad_norm": 3.077500104904175,
+      "learning_rate": 0.00047030716723549487,
+      "loss": 5.9843,
+      "step": 4656
+    },
+    {
+      "epoch": 1.589419795221843,
+      "grad_norm": 2.99000883102417,
+      "learning_rate": 0.00047019340159271903,
+      "loss": 6.4999,
+      "step": 4657
+    },
+    {
+      "epoch": 1.5897610921501708,
+      "grad_norm": 3.0318803787231445,
+      "learning_rate": 0.00047007963594994313,
+      "loss": 6.5713,
+      "step": 4658
+    },
+    {
+      "epoch": 1.5901023890784982,
+      "grad_norm": 3.398522138595581,
+      "learning_rate": 0.00046996587030716724,
+      "loss": 5.365,
+      "step": 4659
+    },
+    {
+      "epoch": 1.590443686006826,
+      "grad_norm": 2.9953789710998535,
+      "learning_rate": 0.0004698521046643914,
+      "loss": 6.1898,
+      "step": 4660
+    },
+    {
+      "epoch": 1.5907849829351535,
+      "grad_norm": 3.6404740810394287,
+      "learning_rate": 0.0004697383390216155,
+      "loss": 6.2839,
+      "step": 4661
+    },
+    {
+      "epoch": 1.5911262798634813,
+      "grad_norm": 3.0940699577331543,
+      "learning_rate": 0.0004696245733788396,
+      "loss": 6.2371,
+      "step": 4662
+    },
+    {
+      "epoch": 1.5914675767918087,
+      "grad_norm": 2.9704160690307617,
+      "learning_rate": 0.0004695108077360637,
+      "loss": 6.0635,
+      "step": 4663
+    },
+    {
+      "epoch": 1.5918088737201366,
+      "grad_norm": 3.2446553707122803,
+      "learning_rate": 0.0004693970420932878,
+      "loss": 6.2583,
+      "step": 4664
+    },
+    {
+      "epoch": 1.5921501706484642,
+      "grad_norm": 2.948814868927002,
+      "learning_rate": 0.0004692832764505119,
+      "loss": 6.6138,
+      "step": 4665
+    },
+    {
+      "epoch": 1.5924914675767918,
+      "grad_norm": 2.9930472373962402,
+      "learning_rate": 0.0004691695108077361,
+      "loss": 6.747,
+      "step": 4666
+    },
+    {
+      "epoch": 1.5928327645051195,
+      "grad_norm": 3.368851661682129,
+      "learning_rate": 0.0004690557451649602,
+      "loss": 5.8773,
+      "step": 4667
+    },
+    {
+      "epoch": 1.593174061433447,
+      "grad_norm": 2.977428674697876,
+      "learning_rate": 0.0004689419795221843,
+      "loss": 6.556,
+      "step": 4668
+    },
+    {
+      "epoch": 1.5935153583617747,
+      "grad_norm": 3.041341543197632,
+      "learning_rate": 0.00046882821387940845,
+      "loss": 6.6892,
+      "step": 4669
+    },
+    {
+      "epoch": 1.5938566552901023,
+      "grad_norm": 2.9362432956695557,
+      "learning_rate": 0.00046871444823663256,
+      "loss": 6.8953,
+      "step": 4670
+    },
+    {
+      "epoch": 1.5941979522184302,
+      "grad_norm": 3.3086955547332764,
+      "learning_rate": 0.00046860068259385666,
+      "loss": 6.2347,
+      "step": 4671
+    },
+    {
+      "epoch": 1.5945392491467576,
+      "grad_norm": 3.477236270904541,
+      "learning_rate": 0.0004684869169510808,
+      "loss": 2.9186,
+      "step": 4672
+    },
+    {
+      "epoch": 1.5948805460750854,
+      "grad_norm": 3.360569477081299,
+      "learning_rate": 0.00046837315130830487,
+      "loss": 5.8384,
+      "step": 4673
+    },
+    {
+      "epoch": 1.5952218430034129,
+      "grad_norm": 3.140916585922241,
+      "learning_rate": 0.000468259385665529,
+      "loss": 6.0714,
+      "step": 4674
+    },
+    {
+      "epoch": 1.5955631399317407,
+      "grad_norm": 3.2916316986083984,
+      "learning_rate": 0.00046814562002275314,
+      "loss": 5.5684,
+      "step": 4675
+    },
+    {
+      "epoch": 1.595904436860068,
+      "grad_norm": 3.8546245098114014,
+      "learning_rate": 0.00046803185437997724,
+      "loss": 6.0605,
+      "step": 4676
+    },
+    {
+      "epoch": 1.596245733788396,
+      "grad_norm": 3.0560851097106934,
+      "learning_rate": 0.0004679180887372014,
+      "loss": 6.554,
+      "step": 4677
+    },
+    {
+      "epoch": 1.5965870307167236,
+      "grad_norm": 2.926163911819458,
+      "learning_rate": 0.0004678043230944255,
+      "loss": 6.2362,
+      "step": 4678
+    },
+    {
+      "epoch": 1.5969283276450512,
+      "grad_norm": 3.3139231204986572,
+      "learning_rate": 0.0004676905574516496,
+      "loss": 6.3501,
+      "step": 4679
+    },
+    {
+      "epoch": 1.5972696245733788,
+      "grad_norm": 3.2012598514556885,
+      "learning_rate": 0.00046757679180887377,
+      "loss": 5.9128,
+      "step": 4680
+    },
+    {
+      "epoch": 1.5976109215017065,
+      "grad_norm": 3.0795986652374268,
+      "learning_rate": 0.0004674630261660979,
+      "loss": 5.7596,
+      "step": 4681
+    },
+    {
+      "epoch": 1.597952218430034,
+      "grad_norm": 3.0973923206329346,
+      "learning_rate": 0.0004673492605233219,
+      "loss": 6.176,
+      "step": 4682
+    },
+    {
+      "epoch": 1.5982935153583617,
+      "grad_norm": 2.9738688468933105,
+      "learning_rate": 0.0004672354948805461,
+      "loss": 6.2366,
+      "step": 4683
+    },
+    {
+      "epoch": 1.5986348122866896,
+      "grad_norm": 3.048534870147705,
+      "learning_rate": 0.0004671217292377702,
+      "loss": 6.2017,
+      "step": 4684
+    },
+    {
+      "epoch": 1.598976109215017,
+      "grad_norm": 3.043328285217285,
+      "learning_rate": 0.0004670079635949943,
+      "loss": 6.1733,
+      "step": 4685
+    },
+    {
+      "epoch": 1.5993174061433448,
+      "grad_norm": 3.086865186691284,
+      "learning_rate": 0.00046689419795221845,
+      "loss": 6.1099,
+      "step": 4686
+    },
+    {
+      "epoch": 1.5996587030716722,
+      "grad_norm": 3.1289331912994385,
+      "learning_rate": 0.00046678043230944256,
+      "loss": 5.693,
+      "step": 4687
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 3.2024731636047363,
+      "learning_rate": 0.00046666666666666666,
+      "loss": 6.616,
+      "step": 4688
+    },
+    {
+      "epoch": 1.6003412969283275,
+      "grad_norm": 3.0767719745635986,
+      "learning_rate": 0.0004665529010238908,
+      "loss": 5.9802,
+      "step": 4689
+    },
+    {
+      "epoch": 1.6006825938566553,
+      "grad_norm": 3.016394853591919,
+      "learning_rate": 0.0004664391353811149,
+      "loss": 6.0161,
+      "step": 4690
+    },
+    {
+      "epoch": 1.601023890784983,
+      "grad_norm": 2.9882636070251465,
+      "learning_rate": 0.00046632536973833903,
+      "loss": 6.7943,
+      "step": 4691
+    },
+    {
+      "epoch": 1.6013651877133106,
+      "grad_norm": 3.6328351497650146,
+      "learning_rate": 0.00046621160409556314,
+      "loss": 5.8045,
+      "step": 4692
+    },
+    {
+      "epoch": 1.6017064846416382,
+      "grad_norm": 3.0648207664489746,
+      "learning_rate": 0.00046609783845278724,
+      "loss": 6.569,
+      "step": 4693
+    },
+    {
+      "epoch": 1.6020477815699659,
+      "grad_norm": 3.3391435146331787,
+      "learning_rate": 0.00046598407281001135,
+      "loss": 6.1539,
+      "step": 4694
+    },
+    {
+      "epoch": 1.6023890784982935,
+      "grad_norm": 2.938784122467041,
+      "learning_rate": 0.0004658703071672355,
+      "loss": 6.4672,
+      "step": 4695
+    },
+    {
+      "epoch": 1.6027303754266211,
+      "grad_norm": 3.095799684524536,
+      "learning_rate": 0.0004657565415244596,
+      "loss": 6.4475,
+      "step": 4696
+    },
+    {
+      "epoch": 1.603071672354949,
+      "grad_norm": 3.8702211380004883,
+      "learning_rate": 0.00046564277588168377,
+      "loss": 5.1767,
+      "step": 4697
+    },
+    {
+      "epoch": 1.6034129692832764,
+      "grad_norm": 3.0789310932159424,
+      "learning_rate": 0.0004655290102389079,
+      "loss": 5.2598,
+      "step": 4698
+    },
+    {
+      "epoch": 1.6037542662116042,
+      "grad_norm": 3.129061698913574,
+      "learning_rate": 0.000465415244596132,
+      "loss": 6.486,
+      "step": 4699
+    },
+    {
+      "epoch": 1.6040955631399316,
+      "grad_norm": 3.2048168182373047,
+      "learning_rate": 0.00046530147895335614,
+      "loss": 6.495,
+      "step": 4700
+    },
+    {
+      "epoch": 1.6044368600682595,
+      "grad_norm": 3.1725878715515137,
+      "learning_rate": 0.00046518771331058024,
+      "loss": 6.2983,
+      "step": 4701
+    },
+    {
+      "epoch": 1.6047781569965869,
+      "grad_norm": 3.1969106197357178,
+      "learning_rate": 0.0004650739476678043,
+      "loss": 5.2289,
+      "step": 4702
+    },
+    {
+      "epoch": 1.6051194539249147,
+      "grad_norm": 3.340092658996582,
+      "learning_rate": 0.00046496018202502845,
+      "loss": 6.19,
+      "step": 4703
+    },
+    {
+      "epoch": 1.6054607508532424,
+      "grad_norm": 3.10861873626709,
+      "learning_rate": 0.00046484641638225256,
+      "loss": 6.4409,
+      "step": 4704
+    },
+    {
+      "epoch": 1.60580204778157,
+      "grad_norm": 2.947946310043335,
+      "learning_rate": 0.00046473265073947666,
+      "loss": 5.9036,
+      "step": 4705
+    },
+    {
+      "epoch": 1.6061433447098976,
+      "grad_norm": 3.063727855682373,
+      "learning_rate": 0.0004646188850967008,
+      "loss": 6.3689,
+      "step": 4706
+    },
+    {
+      "epoch": 1.6064846416382252,
+      "grad_norm": 3.1022796630859375,
+      "learning_rate": 0.0004645051194539249,
+      "loss": 6.3524,
+      "step": 4707
+    },
+    {
+      "epoch": 1.6068259385665529,
+      "grad_norm": 3.0189907550811768,
+      "learning_rate": 0.00046439135381114903,
+      "loss": 6.7521,
+      "step": 4708
+    },
+    {
+      "epoch": 1.6071672354948805,
+      "grad_norm": 3.067929744720459,
+      "learning_rate": 0.0004642775881683732,
+      "loss": 6.0322,
+      "step": 4709
+    },
+    {
+      "epoch": 1.6075085324232083,
+      "grad_norm": 3.403313159942627,
+      "learning_rate": 0.0004641638225255973,
+      "loss": 5.9158,
+      "step": 4710
+    },
+    {
+      "epoch": 1.6078498293515358,
+      "grad_norm": 3.100106954574585,
+      "learning_rate": 0.0004640500568828214,
+      "loss": 6.0732,
+      "step": 4711
+    },
+    {
+      "epoch": 1.6081911262798636,
+      "grad_norm": 3.0627291202545166,
+      "learning_rate": 0.0004639362912400455,
+      "loss": 5.8361,
+      "step": 4712
+    },
+    {
+      "epoch": 1.608532423208191,
+      "grad_norm": 3.6144957542419434,
+      "learning_rate": 0.0004638225255972696,
+      "loss": 6.1397,
+      "step": 4713
+    },
+    {
+      "epoch": 1.6088737201365189,
+      "grad_norm": 3.0293452739715576,
+      "learning_rate": 0.0004637087599544937,
+      "loss": 6.5087,
+      "step": 4714
+    },
+    {
+      "epoch": 1.6092150170648463,
+      "grad_norm": 3.0504560470581055,
+      "learning_rate": 0.0004635949943117179,
+      "loss": 6.5581,
+      "step": 4715
+    },
+    {
+      "epoch": 1.6095563139931741,
+      "grad_norm": 3.043027877807617,
+      "learning_rate": 0.000463481228668942,
+      "loss": 6.8183,
+      "step": 4716
+    },
+    {
+      "epoch": 1.6098976109215017,
+      "grad_norm": 3.135340452194214,
+      "learning_rate": 0.00046336746302616614,
+      "loss": 6.3905,
+      "step": 4717
+    },
+    {
+      "epoch": 1.6102389078498294,
+      "grad_norm": 2.9933533668518066,
+      "learning_rate": 0.00046325369738339024,
+      "loss": 6.2646,
+      "step": 4718
+    },
+    {
+      "epoch": 1.610580204778157,
+      "grad_norm": 3.0164849758148193,
+      "learning_rate": 0.00046313993174061435,
+      "loss": 6.5204,
+      "step": 4719
+    },
+    {
+      "epoch": 1.6109215017064846,
+      "grad_norm": 2.9858627319335938,
+      "learning_rate": 0.0004630261660978385,
+      "loss": 6.2363,
+      "step": 4720
+    },
+    {
+      "epoch": 1.6112627986348123,
+      "grad_norm": 2.998555898666382,
+      "learning_rate": 0.00046291240045506256,
+      "loss": 6.4151,
+      "step": 4721
+    },
+    {
+      "epoch": 1.6116040955631399,
+      "grad_norm": 3.497677803039551,
+      "learning_rate": 0.00046279863481228666,
+      "loss": 5.135,
+      "step": 4722
+    },
+    {
+      "epoch": 1.6119453924914677,
+      "grad_norm": 3.07749080657959,
+      "learning_rate": 0.0004626848691695108,
+      "loss": 6.402,
+      "step": 4723
+    },
+    {
+      "epoch": 1.6122866894197951,
+      "grad_norm": 3.127281665802002,
+      "learning_rate": 0.00046257110352673493,
+      "loss": 6.8232,
+      "step": 4724
+    },
+    {
+      "epoch": 1.612627986348123,
+      "grad_norm": 3.1097822189331055,
+      "learning_rate": 0.00046245733788395903,
+      "loss": 6.0242,
+      "step": 4725
+    },
+    {
+      "epoch": 1.6129692832764504,
+      "grad_norm": 3.2166764736175537,
+      "learning_rate": 0.0004623435722411832,
+      "loss": 5.9335,
+      "step": 4726
+    },
+    {
+      "epoch": 1.6133105802047782,
+      "grad_norm": 3.05297589302063,
+      "learning_rate": 0.0004622298065984073,
+      "loss": 6.4292,
+      "step": 4727
+    },
+    {
+      "epoch": 1.6136518771331056,
+      "grad_norm": 2.0636985301971436,
+      "learning_rate": 0.0004621160409556314,
+      "loss": 3.1146,
+      "step": 4728
+    },
+    {
+      "epoch": 1.6139931740614335,
+      "grad_norm": 3.0197324752807617,
+      "learning_rate": 0.00046200227531285556,
+      "loss": 6.0745,
+      "step": 4729
+    },
+    {
+      "epoch": 1.6143344709897611,
+      "grad_norm": 3.0842418670654297,
+      "learning_rate": 0.00046188850967007967,
+      "loss": 6.6684,
+      "step": 4730
+    },
+    {
+      "epoch": 1.6146757679180888,
+      "grad_norm": 3.0386857986450195,
+      "learning_rate": 0.0004617747440273037,
+      "loss": 6.3415,
+      "step": 4731
+    },
+    {
+      "epoch": 1.6150170648464164,
+      "grad_norm": 3.1578781604766846,
+      "learning_rate": 0.0004616609783845279,
+      "loss": 5.7987,
+      "step": 4732
+    },
+    {
+      "epoch": 1.615358361774744,
+      "grad_norm": 3.1602065563201904,
+      "learning_rate": 0.000461547212741752,
+      "loss": 6.5547,
+      "step": 4733
+    },
+    {
+      "epoch": 1.6156996587030716,
+      "grad_norm": 3.2017452716827393,
+      "learning_rate": 0.0004614334470989761,
+      "loss": 6.0994,
+      "step": 4734
+    },
+    {
+      "epoch": 1.6160409556313993,
+      "grad_norm": 3.114725112915039,
+      "learning_rate": 0.00046131968145620024,
+      "loss": 6.5211,
+      "step": 4735
+    },
+    {
+      "epoch": 1.6163822525597271,
+      "grad_norm": 3.1060032844543457,
+      "learning_rate": 0.00046120591581342435,
+      "loss": 6.3797,
+      "step": 4736
+    },
+    {
+      "epoch": 1.6167235494880545,
+      "grad_norm": 3.1103594303131104,
+      "learning_rate": 0.0004610921501706485,
+      "loss": 6.0067,
+      "step": 4737
+    },
+    {
+      "epoch": 1.6170648464163824,
+      "grad_norm": 3.124356985092163,
+      "learning_rate": 0.0004609783845278726,
+      "loss": 6.7605,
+      "step": 4738
+    },
+    {
+      "epoch": 1.6174061433447098,
+      "grad_norm": 3.092719793319702,
+      "learning_rate": 0.0004608646188850967,
+      "loss": 5.9352,
+      "step": 4739
+    },
+    {
+      "epoch": 1.6177474402730376,
+      "grad_norm": 2.993748903274536,
+      "learning_rate": 0.0004607508532423209,
+      "loss": 6.4874,
+      "step": 4740
+    },
+    {
+      "epoch": 1.618088737201365,
+      "grad_norm": 3.0820672512054443,
+      "learning_rate": 0.00046063708759954493,
+      "loss": 5.4731,
+      "step": 4741
+    },
+    {
+      "epoch": 1.6184300341296929,
+      "grad_norm": 3.193246841430664,
+      "learning_rate": 0.00046052332195676903,
+      "loss": 6.4882,
+      "step": 4742
+    },
+    {
+      "epoch": 1.6187713310580205,
+      "grad_norm": 3.0461337566375732,
+      "learning_rate": 0.0004604095563139932,
+      "loss": 6.7046,
+      "step": 4743
+    },
+    {
+      "epoch": 1.6191126279863481,
+      "grad_norm": 3.1627285480499268,
+      "learning_rate": 0.0004602957906712173,
+      "loss": 5.9241,
+      "step": 4744
+    },
+    {
+      "epoch": 1.6194539249146758,
+      "grad_norm": 3.004056215286255,
+      "learning_rate": 0.0004601820250284414,
+      "loss": 6.182,
+      "step": 4745
+    },
+    {
+      "epoch": 1.6197952218430034,
+      "grad_norm": 3.3949027061462402,
+      "learning_rate": 0.00046006825938566556,
+      "loss": 6.1612,
+      "step": 4746
+    },
+    {
+      "epoch": 1.620136518771331,
+      "grad_norm": 3.4996936321258545,
+      "learning_rate": 0.00045995449374288967,
+      "loss": 5.6995,
+      "step": 4747
+    },
+    {
+      "epoch": 1.6204778156996587,
+      "grad_norm": 3.5690553188323975,
+      "learning_rate": 0.00045984072810011377,
+      "loss": 6.0195,
+      "step": 4748
+    },
+    {
+      "epoch": 1.6208191126279865,
+      "grad_norm": 3.138127326965332,
+      "learning_rate": 0.00045972696245733793,
+      "loss": 6.9745,
+      "step": 4749
+    },
+    {
+      "epoch": 1.621160409556314,
+      "grad_norm": 3.223238468170166,
+      "learning_rate": 0.000459613196814562,
+      "loss": 6.5723,
+      "step": 4750
+    },
+    {
+      "epoch": 1.6215017064846418,
+      "grad_norm": 3.028900146484375,
+      "learning_rate": 0.0004594994311717861,
+      "loss": 6.8207,
+      "step": 4751
+    },
+    {
+      "epoch": 1.6218430034129692,
+      "grad_norm": 3.1101443767547607,
+      "learning_rate": 0.00045938566552901025,
+      "loss": 6.2446,
+      "step": 4752
+    },
+    {
+      "epoch": 1.622184300341297,
+      "grad_norm": 3.129108190536499,
+      "learning_rate": 0.00045927189988623435,
+      "loss": 6.6041,
+      "step": 4753
+    },
+    {
+      "epoch": 1.6225255972696244,
+      "grad_norm": 2.9134464263916016,
+      "learning_rate": 0.00045915813424345846,
+      "loss": 6.674,
+      "step": 4754
+    },
+    {
+      "epoch": 1.6228668941979523,
+      "grad_norm": 2.890514373779297,
+      "learning_rate": 0.0004590443686006826,
+      "loss": 6.4995,
+      "step": 4755
+    },
+    {
+      "epoch": 1.62320819112628,
+      "grad_norm": 3.06087589263916,
+      "learning_rate": 0.0004589306029579067,
+      "loss": 5.8313,
+      "step": 4756
+    },
+    {
+      "epoch": 1.6235494880546075,
+      "grad_norm": 2.98832631111145,
+      "learning_rate": 0.0004588168373151309,
+      "loss": 6.4396,
+      "step": 4757
+    },
+    {
+      "epoch": 1.6238907849829352,
+      "grad_norm": 2.935410976409912,
+      "learning_rate": 0.000458703071672355,
+      "loss": 6.4128,
+      "step": 4758
+    },
+    {
+      "epoch": 1.6242320819112628,
+      "grad_norm": 3.138000965118408,
+      "learning_rate": 0.0004585893060295791,
+      "loss": 6.956,
+      "step": 4759
+    },
+    {
+      "epoch": 1.6245733788395904,
+      "grad_norm": 2.011180877685547,
+      "learning_rate": 0.0004584755403868032,
+      "loss": 3.2119,
+      "step": 4760
+    },
+    {
+      "epoch": 1.624914675767918,
+      "grad_norm": 3.1553590297698975,
+      "learning_rate": 0.0004583617747440273,
+      "loss": 6.419,
+      "step": 4761
+    },
+    {
+      "epoch": 1.6252559726962459,
+      "grad_norm": 3.162964344024658,
+      "learning_rate": 0.0004582480091012514,
+      "loss": 6.3694,
+      "step": 4762
+    },
+    {
+      "epoch": 1.6255972696245733,
+      "grad_norm": 3.1152918338775635,
+      "learning_rate": 0.00045813424345847556,
+      "loss": 6.0361,
+      "step": 4763
+    },
+    {
+      "epoch": 1.6259385665529011,
+      "grad_norm": 3.273135185241699,
+      "learning_rate": 0.00045802047781569967,
+      "loss": 6.0278,
+      "step": 4764
+    },
+    {
+      "epoch": 1.6262798634812285,
+      "grad_norm": 4.751234531402588,
+      "learning_rate": 0.00045790671217292377,
+      "loss": 5.5492,
+      "step": 4765
+    },
+    {
+      "epoch": 1.6266211604095564,
+      "grad_norm": 3.12723445892334,
+      "learning_rate": 0.00045779294653014793,
+      "loss": 6.8186,
+      "step": 4766
+    },
+    {
+      "epoch": 1.6269624573378838,
+      "grad_norm": 3.2533812522888184,
+      "learning_rate": 0.00045767918088737204,
+      "loss": 6.105,
+      "step": 4767
+    },
+    {
+      "epoch": 1.6273037542662117,
+      "grad_norm": 3.344517946243286,
+      "learning_rate": 0.00045756541524459614,
+      "loss": 5.5507,
+      "step": 4768
+    },
+    {
+      "epoch": 1.6276450511945393,
+      "grad_norm": 3.1536638736724854,
+      "learning_rate": 0.0004574516496018203,
+      "loss": 5.9071,
+      "step": 4769
+    },
+    {
+      "epoch": 1.627986348122867,
+      "grad_norm": 3.0342910289764404,
+      "learning_rate": 0.00045733788395904435,
+      "loss": 6.3558,
+      "step": 4770
+    },
+    {
+      "epoch": 1.6283276450511945,
+      "grad_norm": 2.96907377243042,
+      "learning_rate": 0.00045722411831626846,
+      "loss": 6.4653,
+      "step": 4771
+    },
+    {
+      "epoch": 1.6286689419795222,
+      "grad_norm": 3.0146501064300537,
+      "learning_rate": 0.0004571103526734926,
+      "loss": 6.3785,
+      "step": 4772
+    },
+    {
+      "epoch": 1.6290102389078498,
+      "grad_norm": 3.0192205905914307,
+      "learning_rate": 0.0004569965870307167,
+      "loss": 6.1357,
+      "step": 4773
+    },
+    {
+      "epoch": 1.6293515358361774,
+      "grad_norm": 2.924898862838745,
+      "learning_rate": 0.0004568828213879408,
+      "loss": 6.2558,
+      "step": 4774
+    },
+    {
+      "epoch": 1.6296928327645053,
+      "grad_norm": 2.9523444175720215,
+      "learning_rate": 0.000456769055745165,
+      "loss": 6.6745,
+      "step": 4775
+    },
+    {
+      "epoch": 1.6300341296928327,
+      "grad_norm": 3.029269218444824,
+      "learning_rate": 0.0004566552901023891,
+      "loss": 6.6144,
+      "step": 4776
+    },
+    {
+      "epoch": 1.6303754266211605,
+      "grad_norm": 3.0683581829071045,
+      "learning_rate": 0.00045654152445961325,
+      "loss": 6.2914,
+      "step": 4777
+    },
+    {
+      "epoch": 1.630716723549488,
+      "grad_norm": 3.100022554397583,
+      "learning_rate": 0.00045642775881683735,
+      "loss": 6.7963,
+      "step": 4778
+    },
+    {
+      "epoch": 1.6310580204778158,
+      "grad_norm": 2.997330904006958,
+      "learning_rate": 0.00045631399317406146,
+      "loss": 6.5692,
+      "step": 4779
+    },
+    {
+      "epoch": 1.6313993174061432,
+      "grad_norm": 3.10273814201355,
+      "learning_rate": 0.00045620022753128556,
+      "loss": 6.137,
+      "step": 4780
+    },
+    {
+      "epoch": 1.631740614334471,
+      "grad_norm": 2.9892499446868896,
+      "learning_rate": 0.00045608646188850967,
+      "loss": 6.4949,
+      "step": 4781
+    },
+    {
+      "epoch": 1.6320819112627987,
+      "grad_norm": 3.031235933303833,
+      "learning_rate": 0.0004559726962457338,
+      "loss": 6.3293,
+      "step": 4782
+    },
+    {
+      "epoch": 1.6324232081911263,
+      "grad_norm": 3.0643863677978516,
+      "learning_rate": 0.00045585893060295793,
+      "loss": 6.279,
+      "step": 4783
+    },
+    {
+      "epoch": 1.632764505119454,
+      "grad_norm": 3.4579365253448486,
+      "learning_rate": 0.00045574516496018204,
+      "loss": 6.1474,
+      "step": 4784
+    },
+    {
+      "epoch": 1.6331058020477816,
+      "grad_norm": 3.1474268436431885,
+      "learning_rate": 0.00045563139931740614,
+      "loss": 6.3421,
+      "step": 4785
+    },
+    {
+      "epoch": 1.6334470989761092,
+      "grad_norm": 3.405183792114258,
+      "learning_rate": 0.0004555176336746303,
+      "loss": 6.1607,
+      "step": 4786
+    },
+    {
+      "epoch": 1.6337883959044368,
+      "grad_norm": 3.0848171710968018,
+      "learning_rate": 0.0004554038680318544,
+      "loss": 6.4013,
+      "step": 4787
+    },
+    {
+      "epoch": 1.6341296928327647,
+      "grad_norm": 3.092343807220459,
+      "learning_rate": 0.0004552901023890785,
+      "loss": 6.4724,
+      "step": 4788
+    },
+    {
+      "epoch": 1.634470989761092,
+      "grad_norm": 3.5067667961120605,
+      "learning_rate": 0.0004551763367463026,
+      "loss": 5.3636,
+      "step": 4789
+    },
+    {
+      "epoch": 1.63481228668942,
+      "grad_norm": 3.0137641429901123,
+      "learning_rate": 0.0004550625711035267,
+      "loss": 6.6243,
+      "step": 4790
+    },
+    {
+      "epoch": 1.6351535836177473,
+      "grad_norm": 3.554887533187866,
+      "learning_rate": 0.0004549488054607508,
+      "loss": 5.7127,
+      "step": 4791
+    },
+    {
+      "epoch": 1.6354948805460752,
+      "grad_norm": 3.113088607788086,
+      "learning_rate": 0.000454835039817975,
+      "loss": 6.3377,
+      "step": 4792
+    },
+    {
+      "epoch": 1.6358361774744026,
+      "grad_norm": 3.3043792247772217,
+      "learning_rate": 0.0004547212741751991,
+      "loss": 6.3576,
+      "step": 4793
+    },
+    {
+      "epoch": 1.6361774744027304,
+      "grad_norm": 3.1431515216827393,
+      "learning_rate": 0.0004546075085324232,
+      "loss": 6.6704,
+      "step": 4794
+    },
+    {
+      "epoch": 1.636518771331058,
+      "grad_norm": 2.899198293685913,
+      "learning_rate": 0.00045449374288964735,
+      "loss": 6.5668,
+      "step": 4795
+    },
+    {
+      "epoch": 1.6368600682593857,
+      "grad_norm": 2.860257625579834,
+      "learning_rate": 0.00045437997724687146,
+      "loss": 6.3462,
+      "step": 4796
+    },
+    {
+      "epoch": 1.6372013651877133,
+      "grad_norm": 2.903857946395874,
+      "learning_rate": 0.0004542662116040956,
+      "loss": 6.5885,
+      "step": 4797
+    },
+    {
+      "epoch": 1.637542662116041,
+      "grad_norm": 3.1243457794189453,
+      "learning_rate": 0.0004541524459613197,
+      "loss": 6.5971,
+      "step": 4798
+    },
+    {
+      "epoch": 1.6378839590443686,
+      "grad_norm": 2.9572298526763916,
+      "learning_rate": 0.0004540386803185438,
+      "loss": 6.5093,
+      "step": 4799
+    },
+    {
+      "epoch": 1.6382252559726962,
+      "grad_norm": 2.8882224559783936,
+      "learning_rate": 0.00045392491467576793,
+      "loss": 6.5098,
+      "step": 4800
+    },
+    {
+      "epoch": 1.638566552901024,
+      "grad_norm": 4.230573654174805,
+      "learning_rate": 0.00045381114903299204,
+      "loss": 5.5162,
+      "step": 4801
+    },
+    {
+      "epoch": 1.6389078498293514,
+      "grad_norm": 3.288973093032837,
+      "learning_rate": 0.00045369738339021614,
+      "loss": 6.3791,
+      "step": 4802
+    },
+    {
+      "epoch": 1.6392491467576793,
+      "grad_norm": 3.1963343620300293,
+      "learning_rate": 0.0004535836177474403,
+      "loss": 6.6689,
+      "step": 4803
+    },
+    {
+      "epoch": 1.6395904436860067,
+      "grad_norm": 3.1104302406311035,
+      "learning_rate": 0.0004534698521046644,
+      "loss": 6.6137,
+      "step": 4804
+    },
+    {
+      "epoch": 1.6399317406143346,
+      "grad_norm": 3.651779890060425,
+      "learning_rate": 0.0004533560864618885,
+      "loss": 5.9103,
+      "step": 4805
+    },
+    {
+      "epoch": 1.640273037542662,
+      "grad_norm": 5.380152702331543,
+      "learning_rate": 0.00045324232081911267,
+      "loss": 5.7795,
+      "step": 4806
+    },
+    {
+      "epoch": 1.6406143344709898,
+      "grad_norm": 3.1054184436798096,
+      "learning_rate": 0.0004531285551763368,
+      "loss": 6.1345,
+      "step": 4807
+    },
+    {
+      "epoch": 1.6409556313993174,
+      "grad_norm": 3.150416851043701,
+      "learning_rate": 0.0004530147895335609,
+      "loss": 6.2737,
+      "step": 4808
+    },
+    {
+      "epoch": 1.641296928327645,
+      "grad_norm": 6.0928053855896,
+      "learning_rate": 0.000452901023890785,
+      "loss": 5.2429,
+      "step": 4809
+    },
+    {
+      "epoch": 1.6416382252559727,
+      "grad_norm": 3.073267698287964,
+      "learning_rate": 0.0004527872582480091,
+      "loss": 6.2179,
+      "step": 4810
+    },
+    {
+      "epoch": 1.6419795221843003,
+      "grad_norm": 3.0769455432891846,
+      "learning_rate": 0.0004526734926052332,
+      "loss": 6.1769,
+      "step": 4811
+    },
+    {
+      "epoch": 1.642320819112628,
+      "grad_norm": 4.3502888679504395,
+      "learning_rate": 0.00045255972696245736,
+      "loss": 5.1454,
+      "step": 4812
+    },
+    {
+      "epoch": 1.6426621160409556,
+      "grad_norm": 3.0830323696136475,
+      "learning_rate": 0.00045244596131968146,
+      "loss": 6.0589,
+      "step": 4813
+    },
+    {
+      "epoch": 1.6430034129692834,
+      "grad_norm": 3.1683781147003174,
+      "learning_rate": 0.00045233219567690557,
+      "loss": 6.5236,
+      "step": 4814
+    },
+    {
+      "epoch": 1.6433447098976108,
+      "grad_norm": 3.1316497325897217,
+      "learning_rate": 0.0004522184300341297,
+      "loss": 6.4113,
+      "step": 4815
+    },
+    {
+      "epoch": 1.6436860068259387,
+      "grad_norm": 3.0804522037506104,
+      "learning_rate": 0.00045210466439135383,
+      "loss": 5.9039,
+      "step": 4816
+    },
+    {
+      "epoch": 1.644027303754266,
+      "grad_norm": 3.0023155212402344,
+      "learning_rate": 0.00045199089874857793,
+      "loss": 6.1391,
+      "step": 4817
+    },
+    {
+      "epoch": 1.644368600682594,
+      "grad_norm": 2.9633994102478027,
+      "learning_rate": 0.00045187713310580204,
+      "loss": 6.148,
+      "step": 4818
+    },
+    {
+      "epoch": 1.6447098976109213,
+      "grad_norm": 2.8795106410980225,
+      "learning_rate": 0.00045176336746302614,
+      "loss": 6.4872,
+      "step": 4819
+    },
+    {
+      "epoch": 1.6450511945392492,
+      "grad_norm": 6.828871250152588,
+      "learning_rate": 0.0004516496018202503,
+      "loss": 5.7615,
+      "step": 4820
+    },
+    {
+      "epoch": 1.6453924914675768,
+      "grad_norm": 3.061558485031128,
+      "learning_rate": 0.0004515358361774744,
+      "loss": 6.7772,
+      "step": 4821
+    },
+    {
+      "epoch": 1.6457337883959045,
+      "grad_norm": 3.115145444869995,
+      "learning_rate": 0.0004514220705346985,
+      "loss": 6.3194,
+      "step": 4822
+    },
+    {
+      "epoch": 1.646075085324232,
+      "grad_norm": 2.98173451423645,
+      "learning_rate": 0.00045130830489192267,
+      "loss": 6.3607,
+      "step": 4823
+    },
+    {
+      "epoch": 1.6464163822525597,
+      "grad_norm": 3.000434637069702,
+      "learning_rate": 0.0004511945392491468,
+      "loss": 6.7843,
+      "step": 4824
+    },
+    {
+      "epoch": 1.6467576791808873,
+      "grad_norm": 2.882544755935669,
+      "learning_rate": 0.0004510807736063709,
+      "loss": 6.2495,
+      "step": 4825
+    },
+    {
+      "epoch": 1.647098976109215,
+      "grad_norm": 3.010756492614746,
+      "learning_rate": 0.00045096700796359504,
+      "loss": 6.589,
+      "step": 4826
+    },
+    {
+      "epoch": 1.6474402730375428,
+      "grad_norm": 2.991612434387207,
+      "learning_rate": 0.00045085324232081915,
+      "loss": 5.9436,
+      "step": 4827
+    },
+    {
+      "epoch": 1.6477815699658702,
+      "grad_norm": 3.1836135387420654,
+      "learning_rate": 0.0004507394766780432,
+      "loss": 6.5523,
+      "step": 4828
+    },
+    {
+      "epoch": 1.648122866894198,
+      "grad_norm": 3.123823404312134,
+      "learning_rate": 0.00045062571103526736,
+      "loss": 6.2423,
+      "step": 4829
+    },
+    {
+      "epoch": 1.6484641638225255,
+      "grad_norm": 3.108682632446289,
+      "learning_rate": 0.00045051194539249146,
+      "loss": 5.803,
+      "step": 4830
+    },
+    {
+      "epoch": 1.6488054607508533,
+      "grad_norm": 2.941983938217163,
+      "learning_rate": 0.00045039817974971557,
+      "loss": 6.7919,
+      "step": 4831
+    },
+    {
+      "epoch": 1.6491467576791807,
+      "grad_norm": 3.0571041107177734,
+      "learning_rate": 0.0004502844141069397,
+      "loss": 6.3158,
+      "step": 4832
+    },
+    {
+      "epoch": 1.6494880546075086,
+      "grad_norm": 3.1755189895629883,
+      "learning_rate": 0.00045017064846416383,
+      "loss": 6.4594,
+      "step": 4833
+    },
+    {
+      "epoch": 1.6498293515358362,
+      "grad_norm": 3.0600199699401855,
+      "learning_rate": 0.00045005688282138794,
+      "loss": 5.9204,
+      "step": 4834
+    },
+    {
+      "epoch": 1.6501706484641638,
+      "grad_norm": 3.043494939804077,
+      "learning_rate": 0.0004499431171786121,
+      "loss": 6.3821,
+      "step": 4835
+    },
+    {
+      "epoch": 1.6505119453924915,
+      "grad_norm": 3.1068077087402344,
+      "learning_rate": 0.0004498293515358362,
+      "loss": 6.4803,
+      "step": 4836
+    },
+    {
+      "epoch": 1.650853242320819,
+      "grad_norm": 2.9238088130950928,
+      "learning_rate": 0.0004497155858930603,
+      "loss": 5.7569,
+      "step": 4837
+    },
+    {
+      "epoch": 1.6511945392491467,
+      "grad_norm": 3.37758207321167,
+      "learning_rate": 0.0004496018202502844,
+      "loss": 6.4423,
+      "step": 4838
+    },
+    {
+      "epoch": 1.6515358361774743,
+      "grad_norm": 2.9938437938690186,
+      "learning_rate": 0.0004494880546075085,
+      "loss": 6.492,
+      "step": 4839
+    },
+    {
+      "epoch": 1.6518771331058022,
+      "grad_norm": 2.9494314193725586,
+      "learning_rate": 0.0004493742889647327,
+      "loss": 5.878,
+      "step": 4840
+    },
+    {
+      "epoch": 1.6522184300341296,
+      "grad_norm": 3.7142817974090576,
+      "learning_rate": 0.0004492605233219568,
+      "loss": 5.7933,
+      "step": 4841
+    },
+    {
+      "epoch": 1.6525597269624575,
+      "grad_norm": 3.0025503635406494,
+      "learning_rate": 0.0004491467576791809,
+      "loss": 6.2064,
+      "step": 4842
+    },
+    {
+      "epoch": 1.6529010238907849,
+      "grad_norm": 3.326956272125244,
+      "learning_rate": 0.00044903299203640504,
+      "loss": 5.7267,
+      "step": 4843
+    },
+    {
+      "epoch": 1.6532423208191127,
+      "grad_norm": 2.990328311920166,
+      "learning_rate": 0.00044891922639362915,
+      "loss": 5.7901,
+      "step": 4844
+    },
+    {
+      "epoch": 1.6535836177474401,
+      "grad_norm": 3.0785820484161377,
+      "learning_rate": 0.00044880546075085325,
+      "loss": 5.9214,
+      "step": 4845
+    },
+    {
+      "epoch": 1.653924914675768,
+      "grad_norm": 3.112694025039673,
+      "learning_rate": 0.0004486916951080774,
+      "loss": 6.3741,
+      "step": 4846
+    },
+    {
+      "epoch": 1.6542662116040956,
+      "grad_norm": 3.4003303050994873,
+      "learning_rate": 0.0004485779294653015,
+      "loss": 5.736,
+      "step": 4847
+    },
+    {
+      "epoch": 1.6546075085324232,
+      "grad_norm": 2.963510274887085,
+      "learning_rate": 0.00044846416382252557,
+      "loss": 6.0387,
+      "step": 4848
+    },
+    {
+      "epoch": 1.6549488054607508,
+      "grad_norm": 3.8840463161468506,
+      "learning_rate": 0.0004483503981797497,
+      "loss": 5.371,
+      "step": 4849
+    },
+    {
+      "epoch": 1.6552901023890785,
+      "grad_norm": 3.072937488555908,
+      "learning_rate": 0.00044823663253697383,
+      "loss": 6.6929,
+      "step": 4850
+    },
+    {
+      "epoch": 1.655631399317406,
+      "grad_norm": 3.033498764038086,
+      "learning_rate": 0.00044812286689419794,
+      "loss": 5.939,
+      "step": 4851
+    },
+    {
+      "epoch": 1.6559726962457337,
+      "grad_norm": 3.0038321018218994,
+      "learning_rate": 0.0004480091012514221,
+      "loss": 6.4066,
+      "step": 4852
+    },
+    {
+      "epoch": 1.6563139931740616,
+      "grad_norm": 3.0644583702087402,
+      "learning_rate": 0.0004478953356086462,
+      "loss": 6.2456,
+      "step": 4853
+    },
+    {
+      "epoch": 1.656655290102389,
+      "grad_norm": 3.0209004878997803,
+      "learning_rate": 0.0004477815699658703,
+      "loss": 6.2575,
+      "step": 4854
+    },
+    {
+      "epoch": 1.6569965870307168,
+      "grad_norm": 2.9037437438964844,
+      "learning_rate": 0.00044766780432309446,
+      "loss": 6.1034,
+      "step": 4855
+    },
+    {
+      "epoch": 1.6573378839590442,
+      "grad_norm": 2.919391393661499,
+      "learning_rate": 0.00044755403868031857,
+      "loss": 6.4839,
+      "step": 4856
+    },
+    {
+      "epoch": 1.657679180887372,
+      "grad_norm": 2.8986079692840576,
+      "learning_rate": 0.0004474402730375426,
+      "loss": 6.5156,
+      "step": 4857
+    },
+    {
+      "epoch": 1.6580204778156995,
+      "grad_norm": 3.268554210662842,
+      "learning_rate": 0.0004473265073947668,
+      "loss": 6.1534,
+      "step": 4858
+    },
+    {
+      "epoch": 1.6583617747440274,
+      "grad_norm": 3.4434092044830322,
+      "learning_rate": 0.0004472127417519909,
+      "loss": 5.8377,
+      "step": 4859
+    },
+    {
+      "epoch": 1.658703071672355,
+      "grad_norm": 2.988135814666748,
+      "learning_rate": 0.00044709897610921504,
+      "loss": 6.4622,
+      "step": 4860
+    },
+    {
+      "epoch": 1.6590443686006826,
+      "grad_norm": 3.0992002487182617,
+      "learning_rate": 0.00044698521046643915,
+      "loss": 6.4363,
+      "step": 4861
+    },
+    {
+      "epoch": 1.6593856655290102,
+      "grad_norm": 3.1471357345581055,
+      "learning_rate": 0.00044687144482366325,
+      "loss": 6.0289,
+      "step": 4862
+    },
+    {
+      "epoch": 1.6597269624573379,
+      "grad_norm": 3.0062060356140137,
+      "learning_rate": 0.0004467576791808874,
+      "loss": 6.4248,
+      "step": 4863
+    },
+    {
+      "epoch": 1.6600682593856655,
+      "grad_norm": 3.003093957901001,
+      "learning_rate": 0.0004466439135381115,
+      "loss": 6.4863,
+      "step": 4864
+    },
+    {
+      "epoch": 1.6604095563139931,
+      "grad_norm": 3.077218770980835,
+      "learning_rate": 0.0004465301478953356,
+      "loss": 6.3079,
+      "step": 4865
+    },
+    {
+      "epoch": 1.660750853242321,
+      "grad_norm": 3.4413108825683594,
+      "learning_rate": 0.0004464163822525598,
+      "loss": 5.9743,
+      "step": 4866
+    },
+    {
+      "epoch": 1.6610921501706484,
+      "grad_norm": 3.0126073360443115,
+      "learning_rate": 0.00044630261660978383,
+      "loss": 6.3671,
+      "step": 4867
+    },
+    {
+      "epoch": 1.6614334470989762,
+      "grad_norm": 3.0863595008850098,
+      "learning_rate": 0.00044618885096700794,
+      "loss": 6.6083,
+      "step": 4868
+    },
+    {
+      "epoch": 1.6617747440273036,
+      "grad_norm": 3.1746888160705566,
+      "learning_rate": 0.0004460750853242321,
+      "loss": 6.5011,
+      "step": 4869
+    },
+    {
+      "epoch": 1.6621160409556315,
+      "grad_norm": 3.1492929458618164,
+      "learning_rate": 0.0004459613196814562,
+      "loss": 6.4877,
+      "step": 4870
+    },
+    {
+      "epoch": 1.6624573378839589,
+      "grad_norm": 3.422447443008423,
+      "learning_rate": 0.0004458475540386803,
+      "loss": 5.9579,
+      "step": 4871
+    },
+    {
+      "epoch": 1.6627986348122867,
+      "grad_norm": 3.057371139526367,
+      "learning_rate": 0.00044573378839590447,
+      "loss": 6.0468,
+      "step": 4872
+    },
+    {
+      "epoch": 1.6631399317406144,
+      "grad_norm": 4.449178695678711,
+      "learning_rate": 0.00044562002275312857,
+      "loss": 5.8584,
+      "step": 4873
+    },
+    {
+      "epoch": 1.663481228668942,
+      "grad_norm": 3.1592650413513184,
+      "learning_rate": 0.0004455062571103527,
+      "loss": 6.3337,
+      "step": 4874
+    },
+    {
+      "epoch": 1.6638225255972696,
+      "grad_norm": 3.210031509399414,
+      "learning_rate": 0.00044539249146757683,
+      "loss": 6.4076,
+      "step": 4875
+    },
+    {
+      "epoch": 1.6641638225255972,
+      "grad_norm": 3.0083580017089844,
+      "learning_rate": 0.00044527872582480094,
+      "loss": 6.2168,
+      "step": 4876
+    },
+    {
+      "epoch": 1.6645051194539249,
+      "grad_norm": 2.0590007305145264,
+      "learning_rate": 0.000445164960182025,
+      "loss": 3.1449,
+      "step": 4877
+    },
+    {
+      "epoch": 1.6648464163822525,
+      "grad_norm": 3.020071506500244,
+      "learning_rate": 0.00044505119453924915,
+      "loss": 6.0866,
+      "step": 4878
+    },
+    {
+      "epoch": 1.6651877133105804,
+      "grad_norm": 4.307253360748291,
+      "learning_rate": 0.00044493742889647325,
+      "loss": 6.2752,
+      "step": 4879
+    },
+    {
+      "epoch": 1.6655290102389078,
+      "grad_norm": 3.5636332035064697,
+      "learning_rate": 0.00044482366325369736,
+      "loss": 4.8149,
+      "step": 4880
+    },
+    {
+      "epoch": 1.6658703071672356,
+      "grad_norm": 3.1198694705963135,
+      "learning_rate": 0.0004447098976109215,
+      "loss": 6.368,
+      "step": 4881
+    },
+    {
+      "epoch": 1.666211604095563,
+      "grad_norm": 3.0112080574035645,
+      "learning_rate": 0.0004445961319681456,
+      "loss": 6.3972,
+      "step": 4882
+    },
+    {
+      "epoch": 1.6665529010238909,
+      "grad_norm": 3.0651512145996094,
+      "learning_rate": 0.0004444823663253698,
+      "loss": 4.845,
+      "step": 4883
+    },
+    {
+      "epoch": 1.6668941979522183,
+      "grad_norm": 3.187875270843506,
+      "learning_rate": 0.0004443686006825939,
+      "loss": 6.1866,
+      "step": 4884
+    },
+    {
+      "epoch": 1.6672354948805461,
+      "grad_norm": 3.1536598205566406,
+      "learning_rate": 0.000444254835039818,
+      "loss": 6.3055,
+      "step": 4885
+    },
+    {
+      "epoch": 1.6675767918088737,
+      "grad_norm": 3.3544106483459473,
+      "learning_rate": 0.00044414106939704215,
+      "loss": 6.1421,
+      "step": 4886
+    },
+    {
+      "epoch": 1.6679180887372014,
+      "grad_norm": 3.0838112831115723,
+      "learning_rate": 0.0004440273037542662,
+      "loss": 6.1572,
+      "step": 4887
+    },
+    {
+      "epoch": 1.668259385665529,
+      "grad_norm": 2.9401392936706543,
+      "learning_rate": 0.0004439135381114903,
+      "loss": 6.5586,
+      "step": 4888
+    },
+    {
+      "epoch": 1.6686006825938566,
+      "grad_norm": 3.06618595123291,
+      "learning_rate": 0.00044379977246871447,
+      "loss": 6.0765,
+      "step": 4889
+    },
+    {
+      "epoch": 1.6689419795221843,
+      "grad_norm": 2.9306881427764893,
+      "learning_rate": 0.00044368600682593857,
+      "loss": 6.2415,
+      "step": 4890
+    },
+    {
+      "epoch": 1.6692832764505119,
+      "grad_norm": 3.1481986045837402,
+      "learning_rate": 0.0004435722411831627,
+      "loss": 5.8938,
+      "step": 4891
+    },
+    {
+      "epoch": 1.6696245733788397,
+      "grad_norm": 3.0431253910064697,
+      "learning_rate": 0.00044345847554038684,
+      "loss": 6.1017,
+      "step": 4892
+    },
+    {
+      "epoch": 1.6699658703071671,
+      "grad_norm": 2.9972074031829834,
+      "learning_rate": 0.00044334470989761094,
+      "loss": 6.3889,
+      "step": 4893
+    },
+    {
+      "epoch": 1.670307167235495,
+      "grad_norm": 3.1371490955352783,
+      "learning_rate": 0.00044323094425483504,
+      "loss": 6.6684,
+      "step": 4894
+    },
+    {
+      "epoch": 1.6706484641638224,
+      "grad_norm": 3.0418312549591064,
+      "learning_rate": 0.0004431171786120592,
+      "loss": 6.478,
+      "step": 4895
+    },
+    {
+      "epoch": 1.6709897610921502,
+      "grad_norm": 3.074183464050293,
+      "learning_rate": 0.00044300341296928325,
+      "loss": 6.388,
+      "step": 4896
+    },
+    {
+      "epoch": 1.6713310580204777,
+      "grad_norm": 3.0589749813079834,
+      "learning_rate": 0.00044288964732650736,
+      "loss": 6.2,
+      "step": 4897
+    },
+    {
+      "epoch": 1.6716723549488055,
+      "grad_norm": 3.161790609359741,
+      "learning_rate": 0.0004427758816837315,
+      "loss": 5.6493,
+      "step": 4898
+    },
+    {
+      "epoch": 1.6720136518771331,
+      "grad_norm": 2.9835970401763916,
+      "learning_rate": 0.0004426621160409556,
+      "loss": 6.1388,
+      "step": 4899
+    },
+    {
+      "epoch": 1.6723549488054608,
+      "grad_norm": 3.0649311542510986,
+      "learning_rate": 0.00044254835039817973,
+      "loss": 5.9025,
+      "step": 4900
+    },
+    {
+      "epoch": 1.6726962457337884,
+      "grad_norm": 3.1138432025909424,
+      "learning_rate": 0.0004424345847554039,
+      "loss": 6.4551,
+      "step": 4901
+    },
+    {
+      "epoch": 1.673037542662116,
+      "grad_norm": 3.0755131244659424,
+      "learning_rate": 0.000442320819112628,
+      "loss": 6.5686,
+      "step": 4902
+    },
+    {
+      "epoch": 1.6733788395904436,
+      "grad_norm": 3.0821003913879395,
+      "learning_rate": 0.00044220705346985215,
+      "loss": 6.092,
+      "step": 4903
+    },
+    {
+      "epoch": 1.6737201365187713,
+      "grad_norm": 2.937201499938965,
+      "learning_rate": 0.00044209328782707626,
+      "loss": 6.2059,
+      "step": 4904
+    },
+    {
+      "epoch": 1.6740614334470991,
+      "grad_norm": 2.9311842918395996,
+      "learning_rate": 0.00044197952218430036,
+      "loss": 6.4499,
+      "step": 4905
+    },
+    {
+      "epoch": 1.6744027303754265,
+      "grad_norm": 3.036937713623047,
+      "learning_rate": 0.00044186575654152447,
+      "loss": 6.1801,
+      "step": 4906
+    },
+    {
+      "epoch": 1.6747440273037544,
+      "grad_norm": 3.5501644611358643,
+      "learning_rate": 0.00044175199089874857,
+      "loss": 5.8053,
+      "step": 4907
+    },
+    {
+      "epoch": 1.6750853242320818,
+      "grad_norm": 3.023075580596924,
+      "learning_rate": 0.0004416382252559727,
+      "loss": 6.1232,
+      "step": 4908
+    },
+    {
+      "epoch": 1.6754266211604096,
+      "grad_norm": 3.7344937324523926,
+      "learning_rate": 0.00044152445961319684,
+      "loss": 4.8391,
+      "step": 4909
+    },
+    {
+      "epoch": 1.675767918088737,
+      "grad_norm": 4.388406276702881,
+      "learning_rate": 0.00044141069397042094,
+      "loss": 6.2587,
+      "step": 4910
+    },
+    {
+      "epoch": 1.676109215017065,
+      "grad_norm": 3.0161640644073486,
+      "learning_rate": 0.00044129692832764505,
+      "loss": 6.2966,
+      "step": 4911
+    },
+    {
+      "epoch": 1.6764505119453925,
+      "grad_norm": 3.0810742378234863,
+      "learning_rate": 0.0004411831626848692,
+      "loss": 6.915,
+      "step": 4912
+    },
+    {
+      "epoch": 1.6767918088737201,
+      "grad_norm": 3.1341419219970703,
+      "learning_rate": 0.0004410693970420933,
+      "loss": 7.0646,
+      "step": 4913
+    },
+    {
+      "epoch": 1.6771331058020478,
+      "grad_norm": 3.8616952896118164,
+      "learning_rate": 0.0004409556313993174,
+      "loss": 5.5981,
+      "step": 4914
+    },
+    {
+      "epoch": 1.6774744027303754,
+      "grad_norm": 3.4659619331359863,
+      "learning_rate": 0.0004408418657565416,
+      "loss": 5.83,
+      "step": 4915
+    },
+    {
+      "epoch": 1.677815699658703,
+      "grad_norm": 3.06744647026062,
+      "learning_rate": 0.0004407281001137656,
+      "loss": 6.2469,
+      "step": 4916
+    },
+    {
+      "epoch": 1.6781569965870307,
+      "grad_norm": 3.057793140411377,
+      "learning_rate": 0.00044061433447098973,
+      "loss": 6.788,
+      "step": 4917
+    },
+    {
+      "epoch": 1.6784982935153585,
+      "grad_norm": 2.971437931060791,
+      "learning_rate": 0.0004405005688282139,
+      "loss": 6.2344,
+      "step": 4918
+    },
+    {
+      "epoch": 1.678839590443686,
+      "grad_norm": 3.025042772293091,
+      "learning_rate": 0.000440386803185438,
+      "loss": 6.3374,
+      "step": 4919
+    },
+    {
+      "epoch": 1.6791808873720138,
+      "grad_norm": 3.0731565952301025,
+      "learning_rate": 0.0004402730375426621,
+      "loss": 5.7545,
+      "step": 4920
+    },
+    {
+      "epoch": 1.6795221843003412,
+      "grad_norm": 2.9983479976654053,
+      "learning_rate": 0.00044015927189988626,
+      "loss": 6.1115,
+      "step": 4921
+    },
+    {
+      "epoch": 1.679863481228669,
+      "grad_norm": 2.97145414352417,
+      "learning_rate": 0.00044004550625711036,
+      "loss": 6.2915,
+      "step": 4922
+    },
+    {
+      "epoch": 1.6802047781569964,
+      "grad_norm": 3.1883833408355713,
+      "learning_rate": 0.0004399317406143345,
+      "loss": 5.9769,
+      "step": 4923
+    },
+    {
+      "epoch": 1.6805460750853243,
+      "grad_norm": 2.9268434047698975,
+      "learning_rate": 0.0004398179749715586,
+      "loss": 6.5054,
+      "step": 4924
+    },
+    {
+      "epoch": 1.680887372013652,
+      "grad_norm": 3.26137375831604,
+      "learning_rate": 0.0004397042093287827,
+      "loss": 6.3084,
+      "step": 4925
+    },
+    {
+      "epoch": 1.6812286689419795,
+      "grad_norm": 3.009206533432007,
+      "learning_rate": 0.00043959044368600684,
+      "loss": 6.0177,
+      "step": 4926
+    },
+    {
+      "epoch": 1.6815699658703072,
+      "grad_norm": 3.147836685180664,
+      "learning_rate": 0.00043947667804323094,
+      "loss": 5.1989,
+      "step": 4927
+    },
+    {
+      "epoch": 1.6819112627986348,
+      "grad_norm": 3.239095687866211,
+      "learning_rate": 0.00043936291240045505,
+      "loss": 6.1885,
+      "step": 4928
+    },
+    {
+      "epoch": 1.6822525597269624,
+      "grad_norm": 3.0361344814300537,
+      "learning_rate": 0.0004392491467576792,
+      "loss": 6.1484,
+      "step": 4929
+    },
+    {
+      "epoch": 1.68259385665529,
+      "grad_norm": 2.998265504837036,
+      "learning_rate": 0.0004391353811149033,
+      "loss": 6.7742,
+      "step": 4930
+    },
+    {
+      "epoch": 1.682935153583618,
+      "grad_norm": 3.0824317932128906,
+      "learning_rate": 0.0004390216154721274,
+      "loss": 6.5443,
+      "step": 4931
+    },
+    {
+      "epoch": 1.6832764505119453,
+      "grad_norm": 5.664287090301514,
+      "learning_rate": 0.0004389078498293516,
+      "loss": 5.2706,
+      "step": 4932
+    },
+    {
+      "epoch": 1.6836177474402731,
+      "grad_norm": 2.9945015907287598,
+      "learning_rate": 0.0004387940841865757,
+      "loss": 6.2064,
+      "step": 4933
+    },
+    {
+      "epoch": 1.6839590443686006,
+      "grad_norm": 3.081329584121704,
+      "learning_rate": 0.0004386803185437998,
+      "loss": 6.5694,
+      "step": 4934
+    },
+    {
+      "epoch": 1.6843003412969284,
+      "grad_norm": 3.1098530292510986,
+      "learning_rate": 0.0004385665529010239,
+      "loss": 5.8361,
+      "step": 4935
+    },
+    {
+      "epoch": 1.6846416382252558,
+      "grad_norm": 3.166102170944214,
+      "learning_rate": 0.000438452787258248,
+      "loss": 6.0032,
+      "step": 4936
+    },
+    {
+      "epoch": 1.6849829351535837,
+      "grad_norm": 3.0542194843292236,
+      "learning_rate": 0.0004383390216154721,
+      "loss": 5.7641,
+      "step": 4937
+    },
+    {
+      "epoch": 1.6853242320819113,
+      "grad_norm": 2.8499386310577393,
+      "learning_rate": 0.00043822525597269626,
+      "loss": 6.0653,
+      "step": 4938
+    },
+    {
+      "epoch": 1.685665529010239,
+      "grad_norm": 3.0446012020111084,
+      "learning_rate": 0.00043811149032992036,
+      "loss": 6.3366,
+      "step": 4939
+    },
+    {
+      "epoch": 1.6860068259385665,
+      "grad_norm": 3.0523667335510254,
+      "learning_rate": 0.00043799772468714447,
+      "loss": 6.0754,
+      "step": 4940
+    },
+    {
+      "epoch": 1.6863481228668942,
+      "grad_norm": 3.245394229888916,
+      "learning_rate": 0.00043788395904436863,
+      "loss": 6.1414,
+      "step": 4941
+    },
+    {
+      "epoch": 1.6866894197952218,
+      "grad_norm": 3.0902605056762695,
+      "learning_rate": 0.00043777019340159273,
+      "loss": 6.5985,
+      "step": 4942
+    },
+    {
+      "epoch": 1.6870307167235494,
+      "grad_norm": 2.9489684104919434,
+      "learning_rate": 0.0004376564277588169,
+      "loss": 6.3939,
+      "step": 4943
+    },
+    {
+      "epoch": 1.6873720136518773,
+      "grad_norm": 3.0497970581054688,
+      "learning_rate": 0.000437542662116041,
+      "loss": 6.8189,
+      "step": 4944
+    },
+    {
+      "epoch": 1.6877133105802047,
+      "grad_norm": 3.0870232582092285,
+      "learning_rate": 0.00043742889647326505,
+      "loss": 5.96,
+      "step": 4945
+    },
+    {
+      "epoch": 1.6880546075085325,
+      "grad_norm": 3.0094494819641113,
+      "learning_rate": 0.0004373151308304892,
+      "loss": 6.2432,
+      "step": 4946
+    },
+    {
+      "epoch": 1.68839590443686,
+      "grad_norm": 3.130962371826172,
+      "learning_rate": 0.0004372013651877133,
+      "loss": 6.3456,
+      "step": 4947
+    },
+    {
+      "epoch": 1.6887372013651878,
+      "grad_norm": 2.9619531631469727,
+      "learning_rate": 0.0004370875995449374,
+      "loss": 6.666,
+      "step": 4948
+    },
+    {
+      "epoch": 1.6890784982935152,
+      "grad_norm": 3.149620532989502,
+      "learning_rate": 0.0004369738339021616,
+      "loss": 6.0721,
+      "step": 4949
+    },
+    {
+      "epoch": 1.689419795221843,
+      "grad_norm": 2.937761068344116,
+      "learning_rate": 0.0004368600682593857,
+      "loss": 6.7564,
+      "step": 4950
+    },
+    {
+      "epoch": 1.6897610921501707,
+      "grad_norm": 3.5034937858581543,
+      "learning_rate": 0.0004367463026166098,
+      "loss": 5.4092,
+      "step": 4951
+    },
+    {
+      "epoch": 1.6901023890784983,
+      "grad_norm": 3.117332935333252,
+      "learning_rate": 0.00043663253697383394,
+      "loss": 6.3505,
+      "step": 4952
+    },
+    {
+      "epoch": 1.690443686006826,
+      "grad_norm": 5.035308361053467,
+      "learning_rate": 0.00043651877133105805,
+      "loss": 5.6475,
+      "step": 4953
+    },
+    {
+      "epoch": 1.6907849829351536,
+      "grad_norm": 3.0669705867767334,
+      "learning_rate": 0.00043640500568828215,
+      "loss": 5.8391,
+      "step": 4954
+    },
+    {
+      "epoch": 1.6911262798634812,
+      "grad_norm": 3.2862114906311035,
+      "learning_rate": 0.00043629124004550626,
+      "loss": 5.9049,
+      "step": 4955
+    },
+    {
+      "epoch": 1.6914675767918088,
+      "grad_norm": 3.088000774383545,
+      "learning_rate": 0.00043617747440273036,
+      "loss": 6.5768,
+      "step": 4956
+    },
+    {
+      "epoch": 1.6918088737201367,
+      "grad_norm": 3.001810073852539,
+      "learning_rate": 0.00043606370875995447,
+      "loss": 6.2097,
+      "step": 4957
+    },
+    {
+      "epoch": 1.692150170648464,
+      "grad_norm": 3.081796407699585,
+      "learning_rate": 0.00043594994311717863,
+      "loss": 6.8779,
+      "step": 4958
+    },
+    {
+      "epoch": 1.692491467576792,
+      "grad_norm": 3.0735950469970703,
+      "learning_rate": 0.00043583617747440273,
+      "loss": 5.9983,
+      "step": 4959
+    },
+    {
+      "epoch": 1.6928327645051193,
+      "grad_norm": 3.0525989532470703,
+      "learning_rate": 0.00043572241183162684,
+      "loss": 6.4789,
+      "step": 4960
+    },
+    {
+      "epoch": 1.6931740614334472,
+      "grad_norm": 2.9428293704986572,
+      "learning_rate": 0.000435608646188851,
+      "loss": 6.3712,
+      "step": 4961
+    },
+    {
+      "epoch": 1.6935153583617746,
+      "grad_norm": 3.002599000930786,
+      "learning_rate": 0.0004354948805460751,
+      "loss": 6.2377,
+      "step": 4962
+    },
+    {
+      "epoch": 1.6938566552901024,
+      "grad_norm": 3.1271250247955322,
+      "learning_rate": 0.00043538111490329926,
+      "loss": 6.357,
+      "step": 4963
+    },
+    {
+      "epoch": 1.69419795221843,
+      "grad_norm": 2.927199125289917,
+      "learning_rate": 0.0004352673492605233,
+      "loss": 6.3556,
+      "step": 4964
+    },
+    {
+      "epoch": 1.6945392491467577,
+      "grad_norm": 3.364306926727295,
+      "learning_rate": 0.0004351535836177474,
+      "loss": 6.1131,
+      "step": 4965
+    },
+    {
+      "epoch": 1.6948805460750853,
+      "grad_norm": 3.0801026821136475,
+      "learning_rate": 0.0004350398179749716,
+      "loss": 5.7936,
+      "step": 4966
+    },
+    {
+      "epoch": 1.695221843003413,
+      "grad_norm": 4.118381500244141,
+      "learning_rate": 0.0004349260523321957,
+      "loss": 4.5499,
+      "step": 4967
+    },
+    {
+      "epoch": 1.6955631399317406,
+      "grad_norm": 3.0644776821136475,
+      "learning_rate": 0.0004348122866894198,
+      "loss": 6.8683,
+      "step": 4968
+    },
+    {
+      "epoch": 1.6959044368600682,
+      "grad_norm": 3.154467821121216,
+      "learning_rate": 0.00043469852104664395,
+      "loss": 6.1359,
+      "step": 4969
+    },
+    {
+      "epoch": 1.696245733788396,
+      "grad_norm": 2.9244272708892822,
+      "learning_rate": 0.00043458475540386805,
+      "loss": 5.8604,
+      "step": 4970
+    },
+    {
+      "epoch": 1.6965870307167235,
+      "grad_norm": 3.170546054840088,
+      "learning_rate": 0.00043447098976109216,
+      "loss": 6.1517,
+      "step": 4971
+    },
+    {
+      "epoch": 1.6969283276450513,
+      "grad_norm": 3.1615684032440186,
+      "learning_rate": 0.0004343572241183163,
+      "loss": 6.2409,
+      "step": 4972
+    },
+    {
+      "epoch": 1.6972696245733787,
+      "grad_norm": 2.9471771717071533,
+      "learning_rate": 0.0004342434584755404,
+      "loss": 5.7621,
+      "step": 4973
+    },
+    {
+      "epoch": 1.6976109215017066,
+      "grad_norm": 3.0422165393829346,
+      "learning_rate": 0.00043412969283276447,
+      "loss": 6.4401,
+      "step": 4974
+    },
+    {
+      "epoch": 1.697952218430034,
+      "grad_norm": 3.3858096599578857,
+      "learning_rate": 0.00043401592718998863,
+      "loss": 5.916,
+      "step": 4975
+    },
+    {
+      "epoch": 1.6982935153583618,
+      "grad_norm": 3.054570198059082,
+      "learning_rate": 0.00043390216154721273,
+      "loss": 6.6209,
+      "step": 4976
+    },
+    {
+      "epoch": 1.6986348122866894,
+      "grad_norm": 3.7693064212799072,
+      "learning_rate": 0.00043378839590443684,
+      "loss": 5.5803,
+      "step": 4977
+    },
+    {
+      "epoch": 1.698976109215017,
+      "grad_norm": 3.033956289291382,
+      "learning_rate": 0.000433674630261661,
+      "loss": 6.0595,
+      "step": 4978
+    },
+    {
+      "epoch": 1.6993174061433447,
+      "grad_norm": 6.4478912353515625,
+      "learning_rate": 0.0004335608646188851,
+      "loss": 5.3277,
+      "step": 4979
+    },
+    {
+      "epoch": 1.6996587030716723,
+      "grad_norm": 3.0002944469451904,
+      "learning_rate": 0.0004334470989761092,
+      "loss": 6.3766,
+      "step": 4980
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 3.1205992698669434,
+      "learning_rate": 0.00043333333333333337,
+      "loss": 6.3214,
+      "step": 4981
+    },
+    {
+      "epoch": 1.7003412969283276,
+      "grad_norm": 3.3028411865234375,
+      "learning_rate": 0.00043321956769055747,
+      "loss": 6.2618,
+      "step": 4982
+    },
+    {
+      "epoch": 1.7006825938566554,
+      "grad_norm": 3.0650041103363037,
+      "learning_rate": 0.00043310580204778163,
+      "loss": 6.48,
+      "step": 4983
+    },
+    {
+      "epoch": 1.7010238907849828,
+      "grad_norm": 2.9815592765808105,
+      "learning_rate": 0.0004329920364050057,
+      "loss": 5.3141,
+      "step": 4984
+    },
+    {
+      "epoch": 1.7013651877133107,
+      "grad_norm": 3.0242958068847656,
+      "learning_rate": 0.0004328782707622298,
+      "loss": 6.1838,
+      "step": 4985
+    },
+    {
+      "epoch": 1.701706484641638,
+      "grad_norm": 4.361161708831787,
+      "learning_rate": 0.00043276450511945395,
+      "loss": 4.7251,
+      "step": 4986
+    },
+    {
+      "epoch": 1.702047781569966,
+      "grad_norm": 2.9456946849823,
+      "learning_rate": 0.00043265073947667805,
+      "loss": 6.3513,
+      "step": 4987
+    },
+    {
+      "epoch": 1.7023890784982934,
+      "grad_norm": 3.1762263774871826,
+      "learning_rate": 0.00043253697383390216,
+      "loss": 6.6752,
+      "step": 4988
+    },
+    {
+      "epoch": 1.7027303754266212,
+      "grad_norm": 3.128908634185791,
+      "learning_rate": 0.0004324232081911263,
+      "loss": 6.6874,
+      "step": 4989
+    },
+    {
+      "epoch": 1.7030716723549488,
+      "grad_norm": 2.998319625854492,
+      "learning_rate": 0.0004323094425483504,
+      "loss": 5.9913,
+      "step": 4990
+    },
+    {
+      "epoch": 1.7034129692832765,
+      "grad_norm": 3.161729097366333,
+      "learning_rate": 0.0004321956769055745,
+      "loss": 6.0972,
+      "step": 4991
+    },
+    {
+      "epoch": 1.703754266211604,
+      "grad_norm": 3.145906686782837,
+      "learning_rate": 0.0004320819112627987,
+      "loss": 6.0995,
+      "step": 4992
+    },
+    {
+      "epoch": 1.7040955631399317,
+      "grad_norm": 2.9762144088745117,
+      "learning_rate": 0.00043196814562002274,
+      "loss": 6.3909,
+      "step": 4993
+    },
+    {
+      "epoch": 1.7044368600682593,
+      "grad_norm": 3.2298426628112793,
+      "learning_rate": 0.00043185437997724684,
+      "loss": 5.4184,
+      "step": 4994
+    },
+    {
+      "epoch": 1.704778156996587,
+      "grad_norm": 3.0982184410095215,
+      "learning_rate": 0.000431740614334471,
+      "loss": 5.8238,
+      "step": 4995
+    },
+    {
+      "epoch": 1.7051194539249148,
+      "grad_norm": 2.9920384883880615,
+      "learning_rate": 0.0004316268486916951,
+      "loss": 5.8196,
+      "step": 4996
+    },
+    {
+      "epoch": 1.7054607508532422,
+      "grad_norm": 4.23321533203125,
+      "learning_rate": 0.0004315130830489192,
+      "loss": 5.5014,
+      "step": 4997
+    },
+    {
+      "epoch": 1.70580204778157,
+      "grad_norm": 3.3732895851135254,
+      "learning_rate": 0.00043139931740614337,
+      "loss": 5.9258,
+      "step": 4998
+    },
+    {
+      "epoch": 1.7061433447098975,
+      "grad_norm": 3.0920844078063965,
+      "learning_rate": 0.0004312855517633675,
+      "loss": 6.3851,
+      "step": 4999
+    },
+    {
+      "epoch": 1.7064846416382253,
+      "grad_norm": 3.1090736389160156,
+      "learning_rate": 0.0004311717861205916,
+      "loss": 6.5109,
+      "step": 5000
+    },
+    {
+      "epoch": 1.7068259385665527,
+      "grad_norm": 2.9566802978515625,
+      "learning_rate": 0.00043105802047781574,
+      "loss": 6.5596,
+      "step": 5001
+    },
+    {
+      "epoch": 1.7071672354948806,
+      "grad_norm": 2.936339855194092,
+      "learning_rate": 0.00043094425483503984,
+      "loss": 6.7089,
+      "step": 5002
+    },
+    {
+      "epoch": 1.7075085324232082,
+      "grad_norm": 3.0246903896331787,
+      "learning_rate": 0.0004308304891922639,
+      "loss": 6.5555,
+      "step": 5003
+    },
+    {
+      "epoch": 1.7078498293515358,
+      "grad_norm": 3.0471384525299072,
+      "learning_rate": 0.00043071672354948805,
+      "loss": 6.3993,
+      "step": 5004
+    },
+    {
+      "epoch": 1.7081911262798635,
+      "grad_norm": 3.12800931930542,
+      "learning_rate": 0.00043060295790671216,
+      "loss": 6.0564,
+      "step": 5005
+    },
+    {
+      "epoch": 1.708532423208191,
+      "grad_norm": 6.632503032684326,
+      "learning_rate": 0.0004304891922639363,
+      "loss": 5.0925,
+      "step": 5006
+    },
+    {
+      "epoch": 1.7088737201365187,
+      "grad_norm": 3.128831624984741,
+      "learning_rate": 0.0004303754266211604,
+      "loss": 5.274,
+      "step": 5007
+    },
+    {
+      "epoch": 1.7092150170648464,
+      "grad_norm": 3.365473508834839,
+      "learning_rate": 0.0004302616609783845,
+      "loss": 6.5734,
+      "step": 5008
+    },
+    {
+      "epoch": 1.7095563139931742,
+      "grad_norm": 3.8749191761016846,
+      "learning_rate": 0.0004301478953356087,
+      "loss": 6.8053,
+      "step": 5009
+    },
+    {
+      "epoch": 1.7098976109215016,
+      "grad_norm": 3.1307623386383057,
+      "learning_rate": 0.0004300341296928328,
+      "loss": 6.4477,
+      "step": 5010
+    },
+    {
+      "epoch": 1.7102389078498295,
+      "grad_norm": 2.960883140563965,
+      "learning_rate": 0.0004299203640500569,
+      "loss": 6.4975,
+      "step": 5011
+    },
+    {
+      "epoch": 1.7105802047781569,
+      "grad_norm": 2.9841647148132324,
+      "learning_rate": 0.00042980659840728105,
+      "loss": 6.5142,
+      "step": 5012
+    },
+    {
+      "epoch": 1.7109215017064847,
+      "grad_norm": 3.0024847984313965,
+      "learning_rate": 0.0004296928327645051,
+      "loss": 6.8747,
+      "step": 5013
+    },
+    {
+      "epoch": 1.7112627986348121,
+      "grad_norm": 3.0802016258239746,
+      "learning_rate": 0.0004295790671217292,
+      "loss": 5.6291,
+      "step": 5014
+    },
+    {
+      "epoch": 1.71160409556314,
+      "grad_norm": 3.068802833557129,
+      "learning_rate": 0.00042946530147895337,
+      "loss": 5.808,
+      "step": 5015
+    },
+    {
+      "epoch": 1.7119453924914676,
+      "grad_norm": 2.9802236557006836,
+      "learning_rate": 0.0004293515358361775,
+      "loss": 6.4699,
+      "step": 5016
+    },
+    {
+      "epoch": 1.7122866894197952,
+      "grad_norm": 2.9703474044799805,
+      "learning_rate": 0.0004292377701934016,
+      "loss": 5.8941,
+      "step": 5017
+    },
+    {
+      "epoch": 1.7126279863481229,
+      "grad_norm": 2.9278199672698975,
+      "learning_rate": 0.00042912400455062574,
+      "loss": 6.2988,
+      "step": 5018
+    },
+    {
+      "epoch": 1.7129692832764505,
+      "grad_norm": 3.066082239151001,
+      "learning_rate": 0.00042901023890784984,
+      "loss": 6.5867,
+      "step": 5019
+    },
+    {
+      "epoch": 1.713310580204778,
+      "grad_norm": 3.058166980743408,
+      "learning_rate": 0.00042889647326507395,
+      "loss": 6.0648,
+      "step": 5020
+    },
+    {
+      "epoch": 1.7136518771331057,
+      "grad_norm": 3.1347908973693848,
+      "learning_rate": 0.0004287827076222981,
+      "loss": 5.8548,
+      "step": 5021
+    },
+    {
+      "epoch": 1.7139931740614336,
+      "grad_norm": 3.1643781661987305,
+      "learning_rate": 0.0004286689419795222,
+      "loss": 5.9265,
+      "step": 5022
+    },
+    {
+      "epoch": 1.714334470989761,
+      "grad_norm": 3.030947685241699,
+      "learning_rate": 0.00042855517633674626,
+      "loss": 5.989,
+      "step": 5023
+    },
+    {
+      "epoch": 1.7146757679180888,
+      "grad_norm": 3.111829996109009,
+      "learning_rate": 0.0004284414106939704,
+      "loss": 6.5268,
+      "step": 5024
+    },
+    {
+      "epoch": 1.7150170648464163,
+      "grad_norm": 2.8761911392211914,
+      "learning_rate": 0.00042832764505119453,
+      "loss": 6.5559,
+      "step": 5025
+    },
+    {
+      "epoch": 1.715358361774744,
+      "grad_norm": 5.675846099853516,
+      "learning_rate": 0.0004282138794084187,
+      "loss": 5.3305,
+      "step": 5026
+    },
+    {
+      "epoch": 1.7156996587030715,
+      "grad_norm": 3.131277322769165,
+      "learning_rate": 0.0004281001137656428,
+      "loss": 6.8141,
+      "step": 5027
+    },
+    {
+      "epoch": 1.7160409556313994,
+      "grad_norm": 3.332944631576538,
+      "learning_rate": 0.0004279863481228669,
+      "loss": 6.1245,
+      "step": 5028
+    },
+    {
+      "epoch": 1.716382252559727,
+      "grad_norm": 3.1765148639678955,
+      "learning_rate": 0.00042787258248009106,
+      "loss": 5.8214,
+      "step": 5029
+    },
+    {
+      "epoch": 1.7167235494880546,
+      "grad_norm": 3.1236507892608643,
+      "learning_rate": 0.00042775881683731516,
+      "loss": 5.903,
+      "step": 5030
+    },
+    {
+      "epoch": 1.7170648464163822,
+      "grad_norm": 3.129188060760498,
+      "learning_rate": 0.00042764505119453927,
+      "loss": 6.1272,
+      "step": 5031
+    },
+    {
+      "epoch": 1.7174061433447099,
+      "grad_norm": 3.017763614654541,
+      "learning_rate": 0.00042753128555176337,
+      "loss": 6.2087,
+      "step": 5032
+    },
+    {
+      "epoch": 1.7177474402730375,
+      "grad_norm": 3.383964776992798,
+      "learning_rate": 0.0004274175199089875,
+      "loss": 5.143,
+      "step": 5033
+    },
+    {
+      "epoch": 1.7180887372013651,
+      "grad_norm": 3.011268377304077,
+      "learning_rate": 0.0004273037542662116,
+      "loss": 6.7169,
+      "step": 5034
+    },
+    {
+      "epoch": 1.718430034129693,
+      "grad_norm": 3.053954601287842,
+      "learning_rate": 0.00042718998862343574,
+      "loss": 5.9654,
+      "step": 5035
+    },
+    {
+      "epoch": 1.7187713310580204,
+      "grad_norm": 3.0908567905426025,
+      "learning_rate": 0.00042707622298065984,
+      "loss": 6.1534,
+      "step": 5036
+    },
+    {
+      "epoch": 1.7191126279863482,
+      "grad_norm": 3.233099937438965,
+      "learning_rate": 0.00042696245733788395,
+      "loss": 5.985,
+      "step": 5037
+    },
+    {
+      "epoch": 1.7194539249146756,
+      "grad_norm": 2.9370503425598145,
+      "learning_rate": 0.0004268486916951081,
+      "loss": 6.4747,
+      "step": 5038
+    },
+    {
+      "epoch": 1.7197952218430035,
+      "grad_norm": 3.1536028385162354,
+      "learning_rate": 0.0004267349260523322,
+      "loss": 5.9485,
+      "step": 5039
+    },
+    {
+      "epoch": 1.7201365187713311,
+      "grad_norm": 3.166968584060669,
+      "learning_rate": 0.0004266211604095563,
+      "loss": 6.7526,
+      "step": 5040
+    },
+    {
+      "epoch": 1.7204778156996587,
+      "grad_norm": 3.0940346717834473,
+      "learning_rate": 0.0004265073947667805,
+      "loss": 6.4837,
+      "step": 5041
+    },
+    {
+      "epoch": 1.7208191126279864,
+      "grad_norm": 2.989959955215454,
+      "learning_rate": 0.00042639362912400453,
+      "loss": 6.5663,
+      "step": 5042
+    },
+    {
+      "epoch": 1.721160409556314,
+      "grad_norm": 2.976271152496338,
+      "learning_rate": 0.00042627986348122863,
+      "loss": 6.0443,
+      "step": 5043
+    },
+    {
+      "epoch": 1.7215017064846416,
+      "grad_norm": 3.0593347549438477,
+      "learning_rate": 0.0004261660978384528,
+      "loss": 5.6778,
+      "step": 5044
+    },
+    {
+      "epoch": 1.7218430034129693,
+      "grad_norm": 3.020432710647583,
+      "learning_rate": 0.0004260523321956769,
+      "loss": 6.2861,
+      "step": 5045
+    },
+    {
+      "epoch": 1.7221843003412969,
+      "grad_norm": 3.6655890941619873,
+      "learning_rate": 0.00042593856655290106,
+      "loss": 5.8077,
+      "step": 5046
+    },
+    {
+      "epoch": 1.7225255972696245,
+      "grad_norm": 2.9208641052246094,
+      "learning_rate": 0.00042582480091012516,
+      "loss": 6.405,
+      "step": 5047
+    },
+    {
+      "epoch": 1.7228668941979524,
+      "grad_norm": 2.9895660877227783,
+      "learning_rate": 0.00042571103526734927,
+      "loss": 6.1637,
+      "step": 5048
+    },
+    {
+      "epoch": 1.7232081911262798,
+      "grad_norm": 3.0086591243743896,
+      "learning_rate": 0.0004255972696245734,
+      "loss": 6.4718,
+      "step": 5049
+    },
+    {
+      "epoch": 1.7235494880546076,
+      "grad_norm": 3.0350987911224365,
+      "learning_rate": 0.00042548350398179753,
+      "loss": 6.4154,
+      "step": 5050
+    },
+    {
+      "epoch": 1.723890784982935,
+      "grad_norm": 3.0023648738861084,
+      "learning_rate": 0.00042536973833902164,
+      "loss": 6.7531,
+      "step": 5051
+    },
+    {
+      "epoch": 1.7242320819112629,
+      "grad_norm": 2.9815728664398193,
+      "learning_rate": 0.00042525597269624574,
+      "loss": 5.6956,
+      "step": 5052
+    },
+    {
+      "epoch": 1.7245733788395905,
+      "grad_norm": 3.007316827774048,
+      "learning_rate": 0.00042514220705346985,
+      "loss": 6.5796,
+      "step": 5053
+    },
+    {
+      "epoch": 1.7249146757679181,
+      "grad_norm": 3.032801866531372,
+      "learning_rate": 0.00042502844141069395,
+      "loss": 6.1046,
+      "step": 5054
+    },
+    {
+      "epoch": 1.7252559726962458,
+      "grad_norm": 2.990694284439087,
+      "learning_rate": 0.0004249146757679181,
+      "loss": 6.188,
+      "step": 5055
+    },
+    {
+      "epoch": 1.7255972696245734,
+      "grad_norm": 3.120084285736084,
+      "learning_rate": 0.0004248009101251422,
+      "loss": 6.7109,
+      "step": 5056
+    },
+    {
+      "epoch": 1.725938566552901,
+      "grad_norm": 3.544955253601074,
+      "learning_rate": 0.0004246871444823663,
+      "loss": 5.6435,
+      "step": 5057
+    },
+    {
+      "epoch": 1.7262798634812286,
+      "grad_norm": 2.977841377258301,
+      "learning_rate": 0.0004245733788395905,
+      "loss": 5.9886,
+      "step": 5058
+    },
+    {
+      "epoch": 1.7266211604095563,
+      "grad_norm": 3.0229032039642334,
+      "learning_rate": 0.0004244596131968146,
+      "loss": 6.4989,
+      "step": 5059
+    },
+    {
+      "epoch": 1.726962457337884,
+      "grad_norm": 2.9560835361480713,
+      "learning_rate": 0.0004243458475540387,
+      "loss": 6.0739,
+      "step": 5060
+    },
+    {
+      "epoch": 1.7273037542662117,
+      "grad_norm": 2.8804562091827393,
+      "learning_rate": 0.00042423208191126285,
+      "loss": 6.0115,
+      "step": 5061
+    },
+    {
+      "epoch": 1.7276450511945391,
+      "grad_norm": 2.9489169120788574,
+      "learning_rate": 0.0004241183162684869,
+      "loss": 6.7899,
+      "step": 5062
+    },
+    {
+      "epoch": 1.727986348122867,
+      "grad_norm": 3.0079617500305176,
+      "learning_rate": 0.000424004550625711,
+      "loss": 6.6014,
+      "step": 5063
+    },
+    {
+      "epoch": 1.7283276450511944,
+      "grad_norm": 3.1168062686920166,
+      "learning_rate": 0.00042389078498293516,
+      "loss": 6.5893,
+      "step": 5064
+    },
+    {
+      "epoch": 1.7286689419795223,
+      "grad_norm": 2.953608512878418,
+      "learning_rate": 0.00042377701934015927,
+      "loss": 5.8904,
+      "step": 5065
+    },
+    {
+      "epoch": 1.7290102389078499,
+      "grad_norm": 3.050062417984009,
+      "learning_rate": 0.0004236632536973834,
+      "loss": 6.3134,
+      "step": 5066
+    },
+    {
+      "epoch": 1.7293515358361775,
+      "grad_norm": 2.984161138534546,
+      "learning_rate": 0.00042354948805460753,
+      "loss": 6.3411,
+      "step": 5067
+    },
+    {
+      "epoch": 1.7296928327645051,
+      "grad_norm": 3.0105361938476562,
+      "learning_rate": 0.00042343572241183164,
+      "loss": 6.1606,
+      "step": 5068
+    },
+    {
+      "epoch": 1.7300341296928328,
+      "grad_norm": 2.9933488368988037,
+      "learning_rate": 0.0004233219567690558,
+      "loss": 6.2474,
+      "step": 5069
+    },
+    {
+      "epoch": 1.7303754266211604,
+      "grad_norm": 3.0552947521209717,
+      "learning_rate": 0.0004232081911262799,
+      "loss": 5.894,
+      "step": 5070
+    },
+    {
+      "epoch": 1.730716723549488,
+      "grad_norm": 3.3107733726501465,
+      "learning_rate": 0.00042309442548350395,
+      "loss": 6.2136,
+      "step": 5071
+    },
+    {
+      "epoch": 1.7310580204778157,
+      "grad_norm": 3.0709095001220703,
+      "learning_rate": 0.0004229806598407281,
+      "loss": 5.9675,
+      "step": 5072
+    },
+    {
+      "epoch": 1.7313993174061433,
+      "grad_norm": 3.0665359497070312,
+      "learning_rate": 0.0004228668941979522,
+      "loss": 6.2627,
+      "step": 5073
+    },
+    {
+      "epoch": 1.7317406143344711,
+      "grad_norm": 3.0029146671295166,
+      "learning_rate": 0.0004227531285551763,
+      "loss": 6.1561,
+      "step": 5074
+    },
+    {
+      "epoch": 1.7320819112627985,
+      "grad_norm": 3.0567500591278076,
+      "learning_rate": 0.0004226393629124005,
+      "loss": 6.0348,
+      "step": 5075
+    },
+    {
+      "epoch": 1.7324232081911264,
+      "grad_norm": 3.08443546295166,
+      "learning_rate": 0.0004225255972696246,
+      "loss": 5.8998,
+      "step": 5076
+    },
+    {
+      "epoch": 1.7327645051194538,
+      "grad_norm": 2.951045274734497,
+      "learning_rate": 0.0004224118316268487,
+      "loss": 6.1719,
+      "step": 5077
+    },
+    {
+      "epoch": 1.7331058020477816,
+      "grad_norm": 3.0656325817108154,
+      "learning_rate": 0.00042229806598407285,
+      "loss": 6.3616,
+      "step": 5078
+    },
+    {
+      "epoch": 1.7334470989761093,
+      "grad_norm": 2.9396920204162598,
+      "learning_rate": 0.00042218430034129695,
+      "loss": 6.2838,
+      "step": 5079
+    },
+    {
+      "epoch": 1.733788395904437,
+      "grad_norm": 2.96988844871521,
+      "learning_rate": 0.00042207053469852106,
+      "loss": 6.6509,
+      "step": 5080
+    },
+    {
+      "epoch": 1.7341296928327645,
+      "grad_norm": 2.8381896018981934,
+      "learning_rate": 0.00042195676905574516,
+      "loss": 6.1965,
+      "step": 5081
+    },
+    {
+      "epoch": 1.7344709897610922,
+      "grad_norm": 6.883829116821289,
+      "learning_rate": 0.00042184300341296927,
+      "loss": 5.7458,
+      "step": 5082
+    },
+    {
+      "epoch": 1.7348122866894198,
+      "grad_norm": 3.1585566997528076,
+      "learning_rate": 0.0004217292377701934,
+      "loss": 6.4391,
+      "step": 5083
+    },
+    {
+      "epoch": 1.7351535836177474,
+      "grad_norm": 3.090667486190796,
+      "learning_rate": 0.00042161547212741753,
+      "loss": 6.7133,
+      "step": 5084
+    },
+    {
+      "epoch": 1.735494880546075,
+      "grad_norm": 3.033228874206543,
+      "learning_rate": 0.00042150170648464164,
+      "loss": 6.2394,
+      "step": 5085
+    },
+    {
+      "epoch": 1.7358361774744027,
+      "grad_norm": 4.285531520843506,
+      "learning_rate": 0.00042138794084186574,
+      "loss": 4.8546,
+      "step": 5086
+    },
+    {
+      "epoch": 1.7361774744027305,
+      "grad_norm": 3.2763266563415527,
+      "learning_rate": 0.0004212741751990899,
+      "loss": 6.2258,
+      "step": 5087
+    },
+    {
+      "epoch": 1.736518771331058,
+      "grad_norm": 3.614485025405884,
+      "learning_rate": 0.000421160409556314,
+      "loss": 5.6639,
+      "step": 5088
+    },
+    {
+      "epoch": 1.7368600682593858,
+      "grad_norm": 3.088728666305542,
+      "learning_rate": 0.00042104664391353817,
+      "loss": 6.7367,
+      "step": 5089
+    },
+    {
+      "epoch": 1.7372013651877132,
+      "grad_norm": 3.1327497959136963,
+      "learning_rate": 0.00042093287827076227,
+      "loss": 5.9986,
+      "step": 5090
+    },
+    {
+      "epoch": 1.737542662116041,
+      "grad_norm": 2.9615418910980225,
+      "learning_rate": 0.0004208191126279863,
+      "loss": 6.1185,
+      "step": 5091
+    },
+    {
+      "epoch": 1.7378839590443687,
+      "grad_norm": 3.1406466960906982,
+      "learning_rate": 0.0004207053469852105,
+      "loss": 6.3329,
+      "step": 5092
+    },
+    {
+      "epoch": 1.7382252559726963,
+      "grad_norm": 2.9437146186828613,
+      "learning_rate": 0.0004205915813424346,
+      "loss": 6.3977,
+      "step": 5093
+    },
+    {
+      "epoch": 1.738566552901024,
+      "grad_norm": 2.894519090652466,
+      "learning_rate": 0.0004204778156996587,
+      "loss": 6.7851,
+      "step": 5094
+    },
+    {
+      "epoch": 1.7389078498293515,
+      "grad_norm": 3.03589129447937,
+      "learning_rate": 0.00042036405005688285,
+      "loss": 5.6625,
+      "step": 5095
+    },
+    {
+      "epoch": 1.7392491467576792,
+      "grad_norm": 3.0257980823516846,
+      "learning_rate": 0.00042025028441410695,
+      "loss": 5.6283,
+      "step": 5096
+    },
+    {
+      "epoch": 1.7395904436860068,
+      "grad_norm": 2.9698126316070557,
+      "learning_rate": 0.00042013651877133106,
+      "loss": 5.9327,
+      "step": 5097
+    },
+    {
+      "epoch": 1.7399317406143344,
+      "grad_norm": 3.094011068344116,
+      "learning_rate": 0.0004200227531285552,
+      "loss": 6.1025,
+      "step": 5098
+    },
+    {
+      "epoch": 1.740273037542662,
+      "grad_norm": 2.933558702468872,
+      "learning_rate": 0.0004199089874857793,
+      "loss": 6.3226,
+      "step": 5099
+    },
+    {
+      "epoch": 1.74061433447099,
+      "grad_norm": 3.014883518218994,
+      "learning_rate": 0.0004197952218430034,
+      "loss": 6.1689,
+      "step": 5100
+    },
+    {
+      "epoch": 1.7409556313993173,
+      "grad_norm": 2.9969851970672607,
+      "learning_rate": 0.00041968145620022753,
+      "loss": 6.4516,
+      "step": 5101
+    },
+    {
+      "epoch": 1.7412969283276452,
+      "grad_norm": 2.958338737487793,
+      "learning_rate": 0.00041956769055745164,
+      "loss": 6.3338,
+      "step": 5102
+    },
+    {
+      "epoch": 1.7416382252559726,
+      "grad_norm": 2.871307373046875,
+      "learning_rate": 0.00041945392491467574,
+      "loss": 6.3378,
+      "step": 5103
+    },
+    {
+      "epoch": 1.7419795221843004,
+      "grad_norm": 3.485704183578491,
+      "learning_rate": 0.0004193401592718999,
+      "loss": 5.556,
+      "step": 5104
+    },
+    {
+      "epoch": 1.742320819112628,
+      "grad_norm": 3.0032105445861816,
+      "learning_rate": 0.000419226393629124,
+      "loss": 5.7468,
+      "step": 5105
+    },
+    {
+      "epoch": 1.7426621160409557,
+      "grad_norm": 3.2163028717041016,
+      "learning_rate": 0.0004191126279863481,
+      "loss": 6.1866,
+      "step": 5106
+    },
+    {
+      "epoch": 1.7430034129692833,
+      "grad_norm": 3.722259044647217,
+      "learning_rate": 0.00041899886234357227,
+      "loss": 5.2801,
+      "step": 5107
+    },
+    {
+      "epoch": 1.743344709897611,
+      "grad_norm": 3.0470032691955566,
+      "learning_rate": 0.0004188850967007964,
+      "loss": 6.1884,
+      "step": 5108
+    },
+    {
+      "epoch": 1.7436860068259386,
+      "grad_norm": 3.0684938430786133,
+      "learning_rate": 0.00041877133105802054,
+      "loss": 6.9454,
+      "step": 5109
+    },
+    {
+      "epoch": 1.7440273037542662,
+      "grad_norm": 2.902766227722168,
+      "learning_rate": 0.0004186575654152446,
+      "loss": 5.6917,
+      "step": 5110
+    },
+    {
+      "epoch": 1.7443686006825938,
+      "grad_norm": 2.978318929672241,
+      "learning_rate": 0.0004185437997724687,
+      "loss": 5.6608,
+      "step": 5111
+    },
+    {
+      "epoch": 1.7447098976109214,
+      "grad_norm": 2.9816012382507324,
+      "learning_rate": 0.00041843003412969285,
+      "loss": 6.0245,
+      "step": 5112
+    },
+    {
+      "epoch": 1.7450511945392493,
+      "grad_norm": 3.0680909156799316,
+      "learning_rate": 0.00041831626848691695,
+      "loss": 5.9496,
+      "step": 5113
+    },
+    {
+      "epoch": 1.7453924914675767,
+      "grad_norm": 2.8896589279174805,
+      "learning_rate": 0.00041820250284414106,
+      "loss": 6.1912,
+      "step": 5114
+    },
+    {
+      "epoch": 1.7457337883959045,
+      "grad_norm": 2.889788866043091,
+      "learning_rate": 0.0004180887372013652,
+      "loss": 6.7443,
+      "step": 5115
+    },
+    {
+      "epoch": 1.746075085324232,
+      "grad_norm": 3.0151519775390625,
+      "learning_rate": 0.0004179749715585893,
+      "loss": 6.5864,
+      "step": 5116
+    },
+    {
+      "epoch": 1.7464163822525598,
+      "grad_norm": 2.864457368850708,
+      "learning_rate": 0.00041786120591581343,
+      "loss": 6.5104,
+      "step": 5117
+    },
+    {
+      "epoch": 1.7467576791808874,
+      "grad_norm": 3.0471596717834473,
+      "learning_rate": 0.0004177474402730376,
+      "loss": 6.4869,
+      "step": 5118
+    },
+    {
+      "epoch": 1.747098976109215,
+      "grad_norm": 3.1271932125091553,
+      "learning_rate": 0.0004176336746302617,
+      "loss": 6.1027,
+      "step": 5119
+    },
+    {
+      "epoch": 1.7474402730375427,
+      "grad_norm": 3.2092442512512207,
+      "learning_rate": 0.00041751990898748574,
+      "loss": 6.081,
+      "step": 5120
+    },
+    {
+      "epoch": 1.7477815699658703,
+      "grad_norm": 3.0421433448791504,
+      "learning_rate": 0.0004174061433447099,
+      "loss": 6.3332,
+      "step": 5121
+    },
+    {
+      "epoch": 1.748122866894198,
+      "grad_norm": 2.9517619609832764,
+      "learning_rate": 0.000417292377701934,
+      "loss": 6.3718,
+      "step": 5122
+    },
+    {
+      "epoch": 1.7484641638225256,
+      "grad_norm": 2.967085838317871,
+      "learning_rate": 0.0004171786120591581,
+      "loss": 6.2602,
+      "step": 5123
+    },
+    {
+      "epoch": 1.7488054607508532,
+      "grad_norm": 5.240909576416016,
+      "learning_rate": 0.00041706484641638227,
+      "loss": 5.1849,
+      "step": 5124
+    },
+    {
+      "epoch": 1.7491467576791808,
+      "grad_norm": 3.141845464706421,
+      "learning_rate": 0.0004169510807736064,
+      "loss": 5.9711,
+      "step": 5125
+    },
+    {
+      "epoch": 1.7494880546075087,
+      "grad_norm": 3.0010454654693604,
+      "learning_rate": 0.0004168373151308305,
+      "loss": 6.1771,
+      "step": 5126
+    },
+    {
+      "epoch": 1.749829351535836,
+      "grad_norm": 2.5118589401245117,
+      "learning_rate": 0.00041672354948805464,
+      "loss": 3.2816,
+      "step": 5127
+    },
+    {
+      "epoch": 1.750170648464164,
+      "grad_norm": 3.0626370906829834,
+      "learning_rate": 0.00041660978384527875,
+      "loss": 6.3235,
+      "step": 5128
+    },
+    {
+      "epoch": 1.7505119453924913,
+      "grad_norm": 3.541957378387451,
+      "learning_rate": 0.0004164960182025029,
+      "loss": 5.6468,
+      "step": 5129
+    },
+    {
+      "epoch": 1.7508532423208192,
+      "grad_norm": 3.089536428451538,
+      "learning_rate": 0.00041638225255972696,
+      "loss": 6.5741,
+      "step": 5130
+    },
+    {
+      "epoch": 1.7511945392491468,
+      "grad_norm": 3.0071637630462646,
+      "learning_rate": 0.00041626848691695106,
+      "loss": 6.1721,
+      "step": 5131
+    },
+    {
+      "epoch": 1.7515358361774744,
+      "grad_norm": 3.3689115047454834,
+      "learning_rate": 0.0004161547212741752,
+      "loss": 6.32,
+      "step": 5132
+    },
+    {
+      "epoch": 1.751877133105802,
+      "grad_norm": 3.6956567764282227,
+      "learning_rate": 0.0004160409556313993,
+      "loss": 5.7046,
+      "step": 5133
+    },
+    {
+      "epoch": 1.7522184300341297,
+      "grad_norm": 4.705052375793457,
+      "learning_rate": 0.00041592718998862343,
+      "loss": 5.2505,
+      "step": 5134
+    },
+    {
+      "epoch": 1.7525597269624573,
+      "grad_norm": 2.986630916595459,
+      "learning_rate": 0.0004158134243458476,
+      "loss": 6.1452,
+      "step": 5135
+    },
+    {
+      "epoch": 1.752901023890785,
+      "grad_norm": 3.066112995147705,
+      "learning_rate": 0.0004156996587030717,
+      "loss": 6.1217,
+      "step": 5136
+    },
+    {
+      "epoch": 1.7532423208191126,
+      "grad_norm": 3.2455801963806152,
+      "learning_rate": 0.0004155858930602958,
+      "loss": 6.1903,
+      "step": 5137
+    },
+    {
+      "epoch": 1.7535836177474402,
+      "grad_norm": 4.036913871765137,
+      "learning_rate": 0.00041547212741751996,
+      "loss": 5.0287,
+      "step": 5138
+    },
+    {
+      "epoch": 1.753924914675768,
+      "grad_norm": 2.882148265838623,
+      "learning_rate": 0.000415358361774744,
+      "loss": 6.9066,
+      "step": 5139
+    },
+    {
+      "epoch": 1.7542662116040955,
+      "grad_norm": 3.251460552215576,
+      "learning_rate": 0.0004152445961319681,
+      "loss": 5.8373,
+      "step": 5140
+    },
+    {
+      "epoch": 1.7546075085324233,
+      "grad_norm": 3.242882013320923,
+      "learning_rate": 0.00041513083048919227,
+      "loss": 5.5887,
+      "step": 5141
+    },
+    {
+      "epoch": 1.7549488054607507,
+      "grad_norm": 2.918837070465088,
+      "learning_rate": 0.0004150170648464164,
+      "loss": 6.3827,
+      "step": 5142
+    },
+    {
+      "epoch": 1.7552901023890786,
+      "grad_norm": 5.232894420623779,
+      "learning_rate": 0.0004149032992036405,
+      "loss": 4.7531,
+      "step": 5143
+    },
+    {
+      "epoch": 1.7556313993174062,
+      "grad_norm": 2.9766147136688232,
+      "learning_rate": 0.00041478953356086464,
+      "loss": 6.5025,
+      "step": 5144
+    },
+    {
+      "epoch": 1.7559726962457338,
+      "grad_norm": 3.3592119216918945,
+      "learning_rate": 0.00041467576791808875,
+      "loss": 5.1757,
+      "step": 5145
+    },
+    {
+      "epoch": 1.7563139931740614,
+      "grad_norm": 3.114332437515259,
+      "learning_rate": 0.00041456200227531285,
+      "loss": 5.9303,
+      "step": 5146
+    },
+    {
+      "epoch": 1.756655290102389,
+      "grad_norm": 3.0244557857513428,
+      "learning_rate": 0.000414448236632537,
+      "loss": 5.9333,
+      "step": 5147
+    },
+    {
+      "epoch": 1.7569965870307167,
+      "grad_norm": 3.0684051513671875,
+      "learning_rate": 0.0004143344709897611,
+      "loss": 5.7864,
+      "step": 5148
+    },
+    {
+      "epoch": 1.7573378839590443,
+      "grad_norm": 2.931272029876709,
+      "learning_rate": 0.0004142207053469852,
+      "loss": 6.2102,
+      "step": 5149
+    },
+    {
+      "epoch": 1.757679180887372,
+      "grad_norm": 3.0313282012939453,
+      "learning_rate": 0.0004141069397042093,
+      "loss": 6.2017,
+      "step": 5150
+    },
+    {
+      "epoch": 1.7580204778156996,
+      "grad_norm": 3.021868944168091,
+      "learning_rate": 0.00041399317406143343,
+      "loss": 6.8173,
+      "step": 5151
+    },
+    {
+      "epoch": 1.7583617747440274,
+      "grad_norm": 2.8583309650421143,
+      "learning_rate": 0.0004138794084186576,
+      "loss": 6.4639,
+      "step": 5152
+    },
+    {
+      "epoch": 1.7587030716723548,
+      "grad_norm": 5.182238578796387,
+      "learning_rate": 0.0004137656427758817,
+      "loss": 5.7794,
+      "step": 5153
+    },
+    {
+      "epoch": 1.7590443686006827,
+      "grad_norm": 2.988149881362915,
+      "learning_rate": 0.0004136518771331058,
+      "loss": 6.5781,
+      "step": 5154
+    },
+    {
+      "epoch": 1.75938566552901,
+      "grad_norm": 3.7683768272399902,
+      "learning_rate": 0.00041353811149032996,
+      "loss": 6.1431,
+      "step": 5155
+    },
+    {
+      "epoch": 1.759726962457338,
+      "grad_norm": 3.9146881103515625,
+      "learning_rate": 0.00041342434584755406,
+      "loss": 5.6584,
+      "step": 5156
+    },
+    {
+      "epoch": 1.7600682593856656,
+      "grad_norm": 2.989680290222168,
+      "learning_rate": 0.00041331058020477817,
+      "loss": 6.8671,
+      "step": 5157
+    },
+    {
+      "epoch": 1.7604095563139932,
+      "grad_norm": 3.8746161460876465,
+      "learning_rate": 0.00041319681456200233,
+      "loss": 5.0076,
+      "step": 5158
+    },
+    {
+      "epoch": 1.7607508532423208,
+      "grad_norm": 3.014831066131592,
+      "learning_rate": 0.0004130830489192264,
+      "loss": 6.3627,
+      "step": 5159
+    },
+    {
+      "epoch": 1.7610921501706485,
+      "grad_norm": 2.946559429168701,
+      "learning_rate": 0.0004129692832764505,
+      "loss": 5.7226,
+      "step": 5160
+    },
+    {
+      "epoch": 1.761433447098976,
+      "grad_norm": 2.9957525730133057,
+      "learning_rate": 0.00041285551763367464,
+      "loss": 6.5074,
+      "step": 5161
+    },
+    {
+      "epoch": 1.7617747440273037,
+      "grad_norm": 2.896390438079834,
+      "learning_rate": 0.00041274175199089875,
+      "loss": 6.9841,
+      "step": 5162
+    },
+    {
+      "epoch": 1.7621160409556313,
+      "grad_norm": 3.291266918182373,
+      "learning_rate": 0.00041262798634812285,
+      "loss": 6.0102,
+      "step": 5163
+    },
+    {
+      "epoch": 1.762457337883959,
+      "grad_norm": 3.1446962356567383,
+      "learning_rate": 0.000412514220705347,
+      "loss": 5.8353,
+      "step": 5164
+    },
+    {
+      "epoch": 1.7627986348122868,
+      "grad_norm": 2.8148372173309326,
+      "learning_rate": 0.0004124004550625711,
+      "loss": 6.1815,
+      "step": 5165
+    },
+    {
+      "epoch": 1.7631399317406142,
+      "grad_norm": 3.0298731327056885,
+      "learning_rate": 0.0004122866894197952,
+      "loss": 6.2174,
+      "step": 5166
+    },
+    {
+      "epoch": 1.763481228668942,
+      "grad_norm": 3.0104305744171143,
+      "learning_rate": 0.0004121729237770194,
+      "loss": 6.3785,
+      "step": 5167
+    },
+    {
+      "epoch": 1.7638225255972695,
+      "grad_norm": 2.9257543087005615,
+      "learning_rate": 0.0004120591581342435,
+      "loss": 6.0503,
+      "step": 5168
+    },
+    {
+      "epoch": 1.7641638225255973,
+      "grad_norm": 3.0766642093658447,
+      "learning_rate": 0.00041194539249146754,
+      "loss": 5.7167,
+      "step": 5169
+    },
+    {
+      "epoch": 1.764505119453925,
+      "grad_norm": 2.8836543560028076,
+      "learning_rate": 0.0004118316268486917,
+      "loss": 6.0921,
+      "step": 5170
+    },
+    {
+      "epoch": 1.7648464163822526,
+      "grad_norm": 3.0110116004943848,
+      "learning_rate": 0.0004117178612059158,
+      "loss": 6.3009,
+      "step": 5171
+    },
+    {
+      "epoch": 1.7651877133105802,
+      "grad_norm": 2.9918863773345947,
+      "learning_rate": 0.00041160409556313996,
+      "loss": 6.0492,
+      "step": 5172
+    },
+    {
+      "epoch": 1.7655290102389078,
+      "grad_norm": 4.458995342254639,
+      "learning_rate": 0.00041149032992036406,
+      "loss": 5.4651,
+      "step": 5173
+    },
+    {
+      "epoch": 1.7658703071672355,
+      "grad_norm": 3.1331934928894043,
+      "learning_rate": 0.00041137656427758817,
+      "loss": 5.8811,
+      "step": 5174
+    },
+    {
+      "epoch": 1.766211604095563,
+      "grad_norm": 2.99288010597229,
+      "learning_rate": 0.00041126279863481233,
+      "loss": 6.1899,
+      "step": 5175
+    },
+    {
+      "epoch": 1.7665529010238907,
+      "grad_norm": 6.400101661682129,
+      "learning_rate": 0.00041114903299203643,
+      "loss": 4.2093,
+      "step": 5176
+    },
+    {
+      "epoch": 1.7668941979522184,
+      "grad_norm": 3.309239387512207,
+      "learning_rate": 0.00041103526734926054,
+      "loss": 5.9248,
+      "step": 5177
+    },
+    {
+      "epoch": 1.7672354948805462,
+      "grad_norm": 2.9934983253479004,
+      "learning_rate": 0.00041092150170648464,
+      "loss": 6.4463,
+      "step": 5178
+    },
+    {
+      "epoch": 1.7675767918088736,
+      "grad_norm": 3.0152664184570312,
+      "learning_rate": 0.00041080773606370875,
+      "loss": 6.4454,
+      "step": 5179
+    },
+    {
+      "epoch": 1.7679180887372015,
+      "grad_norm": 3.141155958175659,
+      "learning_rate": 0.00041069397042093285,
+      "loss": 5.738,
+      "step": 5180
+    },
+    {
+      "epoch": 1.7682593856655289,
+      "grad_norm": 3.011068344116211,
+      "learning_rate": 0.000410580204778157,
+      "loss": 6.8537,
+      "step": 5181
+    },
+    {
+      "epoch": 1.7686006825938567,
+      "grad_norm": 7.368420600891113,
+      "learning_rate": 0.0004104664391353811,
+      "loss": 5.6683,
+      "step": 5182
+    },
+    {
+      "epoch": 1.7689419795221843,
+      "grad_norm": 3.043335199356079,
+      "learning_rate": 0.0004103526734926052,
+      "loss": 6.6035,
+      "step": 5183
+    },
+    {
+      "epoch": 1.769283276450512,
+      "grad_norm": 3.1109790802001953,
+      "learning_rate": 0.0004102389078498294,
+      "loss": 6.0835,
+      "step": 5184
+    },
+    {
+      "epoch": 1.7696245733788396,
+      "grad_norm": 3.0300333499908447,
+      "learning_rate": 0.0004101251422070535,
+      "loss": 6.4015,
+      "step": 5185
+    },
+    {
+      "epoch": 1.7699658703071672,
+      "grad_norm": 2.9411182403564453,
+      "learning_rate": 0.0004100113765642776,
+      "loss": 6.0647,
+      "step": 5186
+    },
+    {
+      "epoch": 1.7703071672354949,
+      "grad_norm": 3.1921133995056152,
+      "learning_rate": 0.00040989761092150175,
+      "loss": 5.8967,
+      "step": 5187
+    },
+    {
+      "epoch": 1.7706484641638225,
+      "grad_norm": 2.9762604236602783,
+      "learning_rate": 0.0004097838452787258,
+      "loss": 6.4729,
+      "step": 5188
+    },
+    {
+      "epoch": 1.7709897610921501,
+      "grad_norm": 2.857520341873169,
+      "learning_rate": 0.0004096700796359499,
+      "loss": 6.0116,
+      "step": 5189
+    },
+    {
+      "epoch": 1.7713310580204777,
+      "grad_norm": 2.9309916496276855,
+      "learning_rate": 0.00040955631399317407,
+      "loss": 6.5562,
+      "step": 5190
+    },
+    {
+      "epoch": 1.7716723549488056,
+      "grad_norm": 3.0163867473602295,
+      "learning_rate": 0.00040944254835039817,
+      "loss": 6.6054,
+      "step": 5191
+    },
+    {
+      "epoch": 1.772013651877133,
+      "grad_norm": 2.942237377166748,
+      "learning_rate": 0.00040932878270762233,
+      "loss": 6.6192,
+      "step": 5192
+    },
+    {
+      "epoch": 1.7723549488054609,
+      "grad_norm": 2.938898801803589,
+      "learning_rate": 0.00040921501706484643,
+      "loss": 6.3657,
+      "step": 5193
+    },
+    {
+      "epoch": 1.7726962457337883,
+      "grad_norm": 3.5897812843322754,
+      "learning_rate": 0.00040910125142207054,
+      "loss": 5.9367,
+      "step": 5194
+    },
+    {
+      "epoch": 1.773037542662116,
+      "grad_norm": 2.926736831665039,
+      "learning_rate": 0.0004089874857792947,
+      "loss": 6.6937,
+      "step": 5195
+    },
+    {
+      "epoch": 1.7733788395904437,
+      "grad_norm": 3.1813864707946777,
+      "learning_rate": 0.0004088737201365188,
+      "loss": 5.4131,
+      "step": 5196
+    },
+    {
+      "epoch": 1.7737201365187714,
+      "grad_norm": 3.0285255908966064,
+      "learning_rate": 0.0004087599544937429,
+      "loss": 6.0017,
+      "step": 5197
+    },
+    {
+      "epoch": 1.774061433447099,
+      "grad_norm": 3.036341428756714,
+      "learning_rate": 0.000408646188850967,
+      "loss": 6.6581,
+      "step": 5198
+    },
+    {
+      "epoch": 1.7744027303754266,
+      "grad_norm": 3.0285723209381104,
+      "learning_rate": 0.0004085324232081911,
+      "loss": 6.2925,
+      "step": 5199
+    },
+    {
+      "epoch": 1.7747440273037542,
+      "grad_norm": 2.9925947189331055,
+      "learning_rate": 0.0004084186575654152,
+      "loss": 6.1965,
+      "step": 5200
+    },
+    {
+      "epoch": 1.7750853242320819,
+      "grad_norm": 3.050577163696289,
+      "learning_rate": 0.0004083048919226394,
+      "loss": 6.3542,
+      "step": 5201
+    },
+    {
+      "epoch": 1.7754266211604095,
+      "grad_norm": 2.938793897628784,
+      "learning_rate": 0.0004081911262798635,
+      "loss": 6.4952,
+      "step": 5202
+    },
+    {
+      "epoch": 1.7757679180887371,
+      "grad_norm": 3.007162570953369,
+      "learning_rate": 0.0004080773606370876,
+      "loss": 5.9792,
+      "step": 5203
+    },
+    {
+      "epoch": 1.776109215017065,
+      "grad_norm": 2.8875231742858887,
+      "learning_rate": 0.00040796359499431175,
+      "loss": 6.2729,
+      "step": 5204
+    },
+    {
+      "epoch": 1.7764505119453924,
+      "grad_norm": 4.119450092315674,
+      "learning_rate": 0.00040784982935153586,
+      "loss": 4.7705,
+      "step": 5205
+    },
+    {
+      "epoch": 1.7767918088737202,
+      "grad_norm": 3.139040231704712,
+      "learning_rate": 0.00040773606370875996,
+      "loss": 5.8157,
+      "step": 5206
+    },
+    {
+      "epoch": 1.7771331058020476,
+      "grad_norm": 2.9930758476257324,
+      "learning_rate": 0.00040762229806598407,
+      "loss": 6.2695,
+      "step": 5207
+    },
+    {
+      "epoch": 1.7774744027303755,
+      "grad_norm": 2.9738657474517822,
+      "learning_rate": 0.00040750853242320817,
+      "loss": 6.24,
+      "step": 5208
+    },
+    {
+      "epoch": 1.7778156996587031,
+      "grad_norm": 4.560755252838135,
+      "learning_rate": 0.0004073947667804323,
+      "loss": 4.1842,
+      "step": 5209
+    },
+    {
+      "epoch": 1.7781569965870307,
+      "grad_norm": 3.2815661430358887,
+      "learning_rate": 0.00040728100113765644,
+      "loss": 5.3795,
+      "step": 5210
+    },
+    {
+      "epoch": 1.7784982935153584,
+      "grad_norm": 3.145738124847412,
+      "learning_rate": 0.00040716723549488054,
+      "loss": 5.5073,
+      "step": 5211
+    },
+    {
+      "epoch": 1.778839590443686,
+      "grad_norm": 3.058608293533325,
+      "learning_rate": 0.0004070534698521047,
+      "loss": 5.9812,
+      "step": 5212
+    },
+    {
+      "epoch": 1.7791808873720136,
+      "grad_norm": 2.994455337524414,
+      "learning_rate": 0.0004069397042093288,
+      "loss": 6.39,
+      "step": 5213
+    },
+    {
+      "epoch": 1.7795221843003413,
+      "grad_norm": 2.96401309967041,
+      "learning_rate": 0.0004068259385665529,
+      "loss": 6.4591,
+      "step": 5214
+    },
+    {
+      "epoch": 1.7798634812286689,
+      "grad_norm": 3.074064016342163,
+      "learning_rate": 0.00040671217292377707,
+      "loss": 5.9995,
+      "step": 5215
+    },
+    {
+      "epoch": 1.7802047781569965,
+      "grad_norm": 2.9683525562286377,
+      "learning_rate": 0.0004065984072810012,
+      "loss": 6.4996,
+      "step": 5216
+    },
+    {
+      "epoch": 1.7805460750853244,
+      "grad_norm": 2.9950125217437744,
+      "learning_rate": 0.0004064846416382252,
+      "loss": 6.6542,
+      "step": 5217
+    },
+    {
+      "epoch": 1.7808873720136518,
+      "grad_norm": 2.9647035598754883,
+      "learning_rate": 0.0004063708759954494,
+      "loss": 5.9651,
+      "step": 5218
+    },
+    {
+      "epoch": 1.7812286689419796,
+      "grad_norm": 9.434081077575684,
+      "learning_rate": 0.0004062571103526735,
+      "loss": 5.0819,
+      "step": 5219
+    },
+    {
+      "epoch": 1.781569965870307,
+      "grad_norm": 3.1312179565429688,
+      "learning_rate": 0.0004061433447098976,
+      "loss": 6.6445,
+      "step": 5220
+    },
+    {
+      "epoch": 1.7819112627986349,
+      "grad_norm": 3.0493083000183105,
+      "learning_rate": 0.00040602957906712175,
+      "loss": 6.2367,
+      "step": 5221
+    },
+    {
+      "epoch": 1.7822525597269625,
+      "grad_norm": 3.070904493331909,
+      "learning_rate": 0.00040591581342434586,
+      "loss": 6.6448,
+      "step": 5222
+    },
+    {
+      "epoch": 1.7825938566552901,
+      "grad_norm": 3.064748525619507,
+      "learning_rate": 0.00040580204778156996,
+      "loss": 6.255,
+      "step": 5223
+    },
+    {
+      "epoch": 1.7829351535836178,
+      "grad_norm": 2.90639328956604,
+      "learning_rate": 0.0004056882821387941,
+      "loss": 6.2975,
+      "step": 5224
+    },
+    {
+      "epoch": 1.7832764505119454,
+      "grad_norm": 2.9274797439575195,
+      "learning_rate": 0.0004055745164960182,
+      "loss": 6.3858,
+      "step": 5225
+    },
+    {
+      "epoch": 1.783617747440273,
+      "grad_norm": 3.3938796520233154,
+      "learning_rate": 0.00040546075085324233,
+      "loss": 5.6207,
+      "step": 5226
+    },
+    {
+      "epoch": 1.7839590443686006,
+      "grad_norm": 2.930269956588745,
+      "learning_rate": 0.00040534698521046644,
+      "loss": 6.4902,
+      "step": 5227
+    },
+    {
+      "epoch": 1.7843003412969285,
+      "grad_norm": 2.949573040008545,
+      "learning_rate": 0.00040523321956769054,
+      "loss": 6.3274,
+      "step": 5228
+    },
+    {
+      "epoch": 1.784641638225256,
+      "grad_norm": 3.765118360519409,
+      "learning_rate": 0.00040511945392491465,
+      "loss": 5.895,
+      "step": 5229
+    },
+    {
+      "epoch": 1.7849829351535837,
+      "grad_norm": 3.8920938968658447,
+      "learning_rate": 0.0004050056882821388,
+      "loss": 5.751,
+      "step": 5230
+    },
+    {
+      "epoch": 1.7853242320819112,
+      "grad_norm": 3.290802240371704,
+      "learning_rate": 0.0004048919226393629,
+      "loss": 5.801,
+      "step": 5231
+    },
+    {
+      "epoch": 1.785665529010239,
+      "grad_norm": 3.1709654331207275,
+      "learning_rate": 0.00040477815699658707,
+      "loss": 5.6403,
+      "step": 5232
+    },
+    {
+      "epoch": 1.7860068259385664,
+      "grad_norm": 3.0955452919006348,
+      "learning_rate": 0.0004046643913538112,
+      "loss": 6.5735,
+      "step": 5233
+    },
+    {
+      "epoch": 1.7863481228668943,
+      "grad_norm": 2.9965522289276123,
+      "learning_rate": 0.0004045506257110353,
+      "loss": 6.1418,
+      "step": 5234
+    },
+    {
+      "epoch": 1.786689419795222,
+      "grad_norm": 2.980431079864502,
+      "learning_rate": 0.00040443686006825944,
+      "loss": 5.5827,
+      "step": 5235
+    },
+    {
+      "epoch": 1.7870307167235495,
+      "grad_norm": 2.8499867916107178,
+      "learning_rate": 0.00040432309442548354,
+      "loss": 5.8907,
+      "step": 5236
+    },
+    {
+      "epoch": 1.7873720136518771,
+      "grad_norm": 2.8207998275756836,
+      "learning_rate": 0.0004042093287827076,
+      "loss": 6.0233,
+      "step": 5237
+    },
+    {
+      "epoch": 1.7877133105802048,
+      "grad_norm": 3.0672411918640137,
+      "learning_rate": 0.00040409556313993175,
+      "loss": 5.8684,
+      "step": 5238
+    },
+    {
+      "epoch": 1.7880546075085324,
+      "grad_norm": 3.0053303241729736,
+      "learning_rate": 0.00040398179749715586,
+      "loss": 6.1898,
+      "step": 5239
+    },
+    {
+      "epoch": 1.78839590443686,
+      "grad_norm": 8.463170051574707,
+      "learning_rate": 0.00040386803185437996,
+      "loss": 5.5515,
+      "step": 5240
+    },
+    {
+      "epoch": 1.7887372013651879,
+      "grad_norm": 9.868966102600098,
+      "learning_rate": 0.0004037542662116041,
+      "loss": 5.336,
+      "step": 5241
+    },
+    {
+      "epoch": 1.7890784982935153,
+      "grad_norm": 3.046468496322632,
+      "learning_rate": 0.00040364050056882823,
+      "loss": 6.4945,
+      "step": 5242
+    },
+    {
+      "epoch": 1.7894197952218431,
+      "grad_norm": 3.114753007888794,
+      "learning_rate": 0.00040352673492605233,
+      "loss": 6.645,
+      "step": 5243
+    },
+    {
+      "epoch": 1.7897610921501705,
+      "grad_norm": 4.43574857711792,
+      "learning_rate": 0.0004034129692832765,
+      "loss": 6.1538,
+      "step": 5244
+    },
+    {
+      "epoch": 1.7901023890784984,
+      "grad_norm": 3.092228889465332,
+      "learning_rate": 0.0004032992036405006,
+      "loss": 6.236,
+      "step": 5245
+    },
+    {
+      "epoch": 1.7904436860068258,
+      "grad_norm": 3.4146082401275635,
+      "learning_rate": 0.00040318543799772465,
+      "loss": 6.3327,
+      "step": 5246
+    },
+    {
+      "epoch": 1.7907849829351536,
+      "grad_norm": 3.112100124359131,
+      "learning_rate": 0.0004030716723549488,
+      "loss": 6.1249,
+      "step": 5247
+    },
+    {
+      "epoch": 1.7911262798634813,
+      "grad_norm": 3.0014235973358154,
+      "learning_rate": 0.0004029579067121729,
+      "loss": 6.4472,
+      "step": 5248
+    },
+    {
+      "epoch": 1.791467576791809,
+      "grad_norm": 3.3161520957946777,
+      "learning_rate": 0.000402844141069397,
+      "loss": 5.6153,
+      "step": 5249
+    },
+    {
+      "epoch": 1.7918088737201365,
+      "grad_norm": 2.991649866104126,
+      "learning_rate": 0.0004027303754266212,
+      "loss": 6.4251,
+      "step": 5250
+    },
+    {
+      "epoch": 1.7921501706484642,
+      "grad_norm": 2.9196534156799316,
+      "learning_rate": 0.0004026166097838453,
+      "loss": 6.4241,
+      "step": 5251
+    },
+    {
+      "epoch": 1.7924914675767918,
+      "grad_norm": 4.535854339599609,
+      "learning_rate": 0.00040250284414106944,
+      "loss": 4.7127,
+      "step": 5252
+    },
+    {
+      "epoch": 1.7928327645051194,
+      "grad_norm": 2.958326578140259,
+      "learning_rate": 0.00040238907849829354,
+      "loss": 6.3131,
+      "step": 5253
+    },
+    {
+      "epoch": 1.7931740614334473,
+      "grad_norm": 3.1165990829467773,
+      "learning_rate": 0.00040227531285551765,
+      "loss": 5.3994,
+      "step": 5254
+    },
+    {
+      "epoch": 1.7935153583617747,
+      "grad_norm": 3.234656572341919,
+      "learning_rate": 0.0004021615472127418,
+      "loss": 6.8472,
+      "step": 5255
+    },
+    {
+      "epoch": 1.7938566552901025,
+      "grad_norm": 3.0498690605163574,
+      "learning_rate": 0.00040204778156996586,
+      "loss": 6.2153,
+      "step": 5256
+    },
+    {
+      "epoch": 1.79419795221843,
+      "grad_norm": 2.997565507888794,
+      "learning_rate": 0.00040193401592718996,
+      "loss": 6.3802,
+      "step": 5257
+    },
+    {
+      "epoch": 1.7945392491467578,
+      "grad_norm": 2.9837565422058105,
+      "learning_rate": 0.0004018202502844141,
+      "loss": 5.8155,
+      "step": 5258
+    },
+    {
+      "epoch": 1.7948805460750852,
+      "grad_norm": 2.9351367950439453,
+      "learning_rate": 0.00040170648464163823,
+      "loss": 6.6041,
+      "step": 5259
+    },
+    {
+      "epoch": 1.795221843003413,
+      "grad_norm": 2.891256332397461,
+      "learning_rate": 0.00040159271899886233,
+      "loss": 5.978,
+      "step": 5260
+    },
+    {
+      "epoch": 1.7955631399317407,
+      "grad_norm": 2.9434101581573486,
+      "learning_rate": 0.0004014789533560865,
+      "loss": 6.2294,
+      "step": 5261
+    },
+    {
+      "epoch": 1.7959044368600683,
+      "grad_norm": 3.0695149898529053,
+      "learning_rate": 0.0004013651877133106,
+      "loss": 6.538,
+      "step": 5262
+    },
+    {
+      "epoch": 1.796245733788396,
+      "grad_norm": 3.029841661453247,
+      "learning_rate": 0.0004012514220705347,
+      "loss": 5.6806,
+      "step": 5263
+    },
+    {
+      "epoch": 1.7965870307167235,
+      "grad_norm": 2.961660623550415,
+      "learning_rate": 0.00040113765642775886,
+      "loss": 6.4616,
+      "step": 5264
+    },
+    {
+      "epoch": 1.7969283276450512,
+      "grad_norm": 4.678473472595215,
+      "learning_rate": 0.00040102389078498297,
+      "loss": 5.3174,
+      "step": 5265
+    },
+    {
+      "epoch": 1.7972696245733788,
+      "grad_norm": 3.08498215675354,
+      "learning_rate": 0.000400910125142207,
+      "loss": 5.7922,
+      "step": 5266
+    },
+    {
+      "epoch": 1.7976109215017066,
+      "grad_norm": 2.980827808380127,
+      "learning_rate": 0.0004007963594994312,
+      "loss": 6.1678,
+      "step": 5267
+    },
+    {
+      "epoch": 1.797952218430034,
+      "grad_norm": 2.927297592163086,
+      "learning_rate": 0.0004006825938566553,
+      "loss": 6.192,
+      "step": 5268
+    },
+    {
+      "epoch": 1.798293515358362,
+      "grad_norm": 3.01383638381958,
+      "learning_rate": 0.0004005688282138794,
+      "loss": 6.2782,
+      "step": 5269
+    },
+    {
+      "epoch": 1.7986348122866893,
+      "grad_norm": 3.114114999771118,
+      "learning_rate": 0.00040045506257110355,
+      "loss": 5.8744,
+      "step": 5270
+    },
+    {
+      "epoch": 1.7989761092150172,
+      "grad_norm": 3.1586034297943115,
+      "learning_rate": 0.00040034129692832765,
+      "loss": 6.3056,
+      "step": 5271
+    },
+    {
+      "epoch": 1.7993174061433446,
+      "grad_norm": 2.895045518875122,
+      "learning_rate": 0.0004002275312855518,
+      "loss": 5.8402,
+      "step": 5272
+    },
+    {
+      "epoch": 1.7996587030716724,
+      "grad_norm": 3.156750202178955,
+      "learning_rate": 0.0004001137656427759,
+      "loss": 5.9734,
+      "step": 5273
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 3.116347312927246,
+      "learning_rate": 0.0004,
+      "loss": 5.6309,
+      "step": 5274
+    },
+    {
+      "epoch": 1.8003412969283277,
+      "grad_norm": 3.3794455528259277,
+      "learning_rate": 0.0003998862343572241,
+      "loss": 6.1649,
+      "step": 5275
+    },
+    {
+      "epoch": 1.8006825938566553,
+      "grad_norm": 3.00872540473938,
+      "learning_rate": 0.00039977246871444823,
+      "loss": 6.449,
+      "step": 5276
+    },
+    {
+      "epoch": 1.801023890784983,
+      "grad_norm": 3.6723685264587402,
+      "learning_rate": 0.00039965870307167233,
+      "loss": 6.0114,
+      "step": 5277
+    },
+    {
+      "epoch": 1.8013651877133106,
+      "grad_norm": 2.9524502754211426,
+      "learning_rate": 0.0003995449374288965,
+      "loss": 6.1426,
+      "step": 5278
+    },
+    {
+      "epoch": 1.8017064846416382,
+      "grad_norm": 2.987380266189575,
+      "learning_rate": 0.0003994311717861206,
+      "loss": 5.964,
+      "step": 5279
+    },
+    {
+      "epoch": 1.802047781569966,
+      "grad_norm": 3.3028364181518555,
+      "learning_rate": 0.0003993174061433447,
+      "loss": 5.3033,
+      "step": 5280
+    },
+    {
+      "epoch": 1.8023890784982934,
+      "grad_norm": 2.9730100631713867,
+      "learning_rate": 0.00039920364050056886,
+      "loss": 5.9817,
+      "step": 5281
+    },
+    {
+      "epoch": 1.8027303754266213,
+      "grad_norm": 3.1205883026123047,
+      "learning_rate": 0.00039908987485779297,
+      "loss": 5.8908,
+      "step": 5282
+    },
+    {
+      "epoch": 1.8030716723549487,
+      "grad_norm": 2.9819066524505615,
+      "learning_rate": 0.00039897610921501707,
+      "loss": 5.9874,
+      "step": 5283
+    },
+    {
+      "epoch": 1.8034129692832765,
+      "grad_norm": 3.0491702556610107,
+      "learning_rate": 0.00039886234357224123,
+      "loss": 6.1682,
+      "step": 5284
+    },
+    {
+      "epoch": 1.803754266211604,
+      "grad_norm": 3.0884945392608643,
+      "learning_rate": 0.0003987485779294653,
+      "loss": 6.6682,
+      "step": 5285
+    },
+    {
+      "epoch": 1.8040955631399318,
+      "grad_norm": 3.4136314392089844,
+      "learning_rate": 0.0003986348122866894,
+      "loss": 5.6285,
+      "step": 5286
+    },
+    {
+      "epoch": 1.8044368600682594,
+      "grad_norm": 3.038418769836426,
+      "learning_rate": 0.00039852104664391355,
+      "loss": 6.8089,
+      "step": 5287
+    },
+    {
+      "epoch": 1.804778156996587,
+      "grad_norm": 2.9346542358398438,
+      "learning_rate": 0.00039840728100113765,
+      "loss": 5.9561,
+      "step": 5288
+    },
+    {
+      "epoch": 1.8051194539249147,
+      "grad_norm": 4.106077671051025,
+      "learning_rate": 0.00039829351535836176,
+      "loss": 5.2572,
+      "step": 5289
+    },
+    {
+      "epoch": 1.8054607508532423,
+      "grad_norm": 3.0919151306152344,
+      "learning_rate": 0.0003981797497155859,
+      "loss": 6.5056,
+      "step": 5290
+    },
+    {
+      "epoch": 1.80580204778157,
+      "grad_norm": 3.0660130977630615,
+      "learning_rate": 0.00039806598407281,
+      "loss": 6.0935,
+      "step": 5291
+    },
+    {
+      "epoch": 1.8061433447098976,
+      "grad_norm": 3.0275156497955322,
+      "learning_rate": 0.0003979522184300341,
+      "loss": 4.5541,
+      "step": 5292
+    },
+    {
+      "epoch": 1.8064846416382254,
+      "grad_norm": 3.370018243789673,
+      "learning_rate": 0.0003978384527872583,
+      "loss": 5.0421,
+      "step": 5293
+    },
+    {
+      "epoch": 1.8068259385665528,
+      "grad_norm": 3.184339761734009,
+      "learning_rate": 0.0003977246871444824,
+      "loss": 5.1131,
+      "step": 5294
+    },
+    {
+      "epoch": 1.8071672354948807,
+      "grad_norm": 3.00337553024292,
+      "learning_rate": 0.0003976109215017065,
+      "loss": 6.6479,
+      "step": 5295
+    },
+    {
+      "epoch": 1.807508532423208,
+      "grad_norm": 2.92510724067688,
+      "learning_rate": 0.0003974971558589306,
+      "loss": 6.097,
+      "step": 5296
+    },
+    {
+      "epoch": 1.807849829351536,
+      "grad_norm": 3.030064105987549,
+      "learning_rate": 0.0003973833902161547,
+      "loss": 6.3918,
+      "step": 5297
+    },
+    {
+      "epoch": 1.8081911262798633,
+      "grad_norm": 3.019908905029297,
+      "learning_rate": 0.00039726962457337886,
+      "loss": 5.8597,
+      "step": 5298
+    },
+    {
+      "epoch": 1.8085324232081912,
+      "grad_norm": 2.938002109527588,
+      "learning_rate": 0.00039715585893060297,
+      "loss": 6.1059,
+      "step": 5299
+    },
+    {
+      "epoch": 1.8088737201365188,
+      "grad_norm": 3.3193299770355225,
+      "learning_rate": 0.0003970420932878271,
+      "loss": 5.3933,
+      "step": 5300
+    },
+    {
+      "epoch": 1.8092150170648464,
+      "grad_norm": 3.112321376800537,
+      "learning_rate": 0.00039692832764505123,
+      "loss": 6.1252,
+      "step": 5301
+    },
+    {
+      "epoch": 1.809556313993174,
+      "grad_norm": 2.9885871410369873,
+      "learning_rate": 0.00039681456200227534,
+      "loss": 6.2435,
+      "step": 5302
+    },
+    {
+      "epoch": 1.8098976109215017,
+      "grad_norm": 3.8620083332061768,
+      "learning_rate": 0.00039670079635949944,
+      "loss": 5.2043,
+      "step": 5303
+    },
+    {
+      "epoch": 1.8102389078498293,
+      "grad_norm": 3.0883119106292725,
+      "learning_rate": 0.0003965870307167236,
+      "loss": 6.0623,
+      "step": 5304
+    },
+    {
+      "epoch": 1.810580204778157,
+      "grad_norm": 2.965674877166748,
+      "learning_rate": 0.00039647326507394765,
+      "loss": 6.138,
+      "step": 5305
+    },
+    {
+      "epoch": 1.8109215017064848,
+      "grad_norm": 2.917496681213379,
+      "learning_rate": 0.00039635949943117176,
+      "loss": 6.5031,
+      "step": 5306
+    },
+    {
+      "epoch": 1.8112627986348122,
+      "grad_norm": 2.951918363571167,
+      "learning_rate": 0.0003962457337883959,
+      "loss": 6.0293,
+      "step": 5307
+    },
+    {
+      "epoch": 1.81160409556314,
+      "grad_norm": 4.126885890960693,
+      "learning_rate": 0.00039613196814562,
+      "loss": 5.7647,
+      "step": 5308
+    },
+    {
+      "epoch": 1.8119453924914675,
+      "grad_norm": 2.963454484939575,
+      "learning_rate": 0.0003960182025028441,
+      "loss": 6.4464,
+      "step": 5309
+    },
+    {
+      "epoch": 1.8122866894197953,
+      "grad_norm": 3.006791591644287,
+      "learning_rate": 0.0003959044368600683,
+      "loss": 6.4746,
+      "step": 5310
+    },
+    {
+      "epoch": 1.8126279863481227,
+      "grad_norm": 2.443580150604248,
+      "learning_rate": 0.0003957906712172924,
+      "loss": 3.2005,
+      "step": 5311
+    },
+    {
+      "epoch": 1.8129692832764506,
+      "grad_norm": 3.0745327472686768,
+      "learning_rate": 0.0003956769055745165,
+      "loss": 6.3638,
+      "step": 5312
+    },
+    {
+      "epoch": 1.8133105802047782,
+      "grad_norm": 3.1432762145996094,
+      "learning_rate": 0.00039556313993174065,
+      "loss": 5.9445,
+      "step": 5313
+    },
+    {
+      "epoch": 1.8136518771331058,
+      "grad_norm": 3.154367685317993,
+      "learning_rate": 0.0003954493742889647,
+      "loss": 6.4095,
+      "step": 5314
+    },
+    {
+      "epoch": 1.8139931740614335,
+      "grad_norm": 3.050010919570923,
+      "learning_rate": 0.00039533560864618886,
+      "loss": 6.4283,
+      "step": 5315
+    },
+    {
+      "epoch": 1.814334470989761,
+      "grad_norm": 2.967775344848633,
+      "learning_rate": 0.00039522184300341297,
+      "loss": 6.1403,
+      "step": 5316
+    },
+    {
+      "epoch": 1.8146757679180887,
+      "grad_norm": 3.014551877975464,
+      "learning_rate": 0.0003951080773606371,
+      "loss": 6.4165,
+      "step": 5317
+    },
+    {
+      "epoch": 1.8150170648464163,
+      "grad_norm": 3.006186008453369,
+      "learning_rate": 0.00039499431171786123,
+      "loss": 7.0178,
+      "step": 5318
+    },
+    {
+      "epoch": 1.8153583617747442,
+      "grad_norm": 2.8938446044921875,
+      "learning_rate": 0.00039488054607508534,
+      "loss": 6.7276,
+      "step": 5319
+    },
+    {
+      "epoch": 1.8156996587030716,
+      "grad_norm": 2.9930148124694824,
+      "learning_rate": 0.00039476678043230944,
+      "loss": 6.5492,
+      "step": 5320
+    },
+    {
+      "epoch": 1.8160409556313994,
+      "grad_norm": 3.3085694313049316,
+      "learning_rate": 0.0003946530147895336,
+      "loss": 6.1508,
+      "step": 5321
+    },
+    {
+      "epoch": 1.8163822525597269,
+      "grad_norm": 3.0276403427124023,
+      "learning_rate": 0.0003945392491467577,
+      "loss": 5.6613,
+      "step": 5322
+    },
+    {
+      "epoch": 1.8167235494880547,
+      "grad_norm": 2.933673620223999,
+      "learning_rate": 0.0003944254835039818,
+      "loss": 6.7415,
+      "step": 5323
+    },
+    {
+      "epoch": 1.817064846416382,
+      "grad_norm": 2.8729095458984375,
+      "learning_rate": 0.0003943117178612059,
+      "loss": 5.8731,
+      "step": 5324
+    },
+    {
+      "epoch": 1.81740614334471,
+      "grad_norm": 2.9451277256011963,
+      "learning_rate": 0.00039419795221843,
+      "loss": 5.8406,
+      "step": 5325
+    },
+    {
+      "epoch": 1.8177474402730376,
+      "grad_norm": 2.9216501712799072,
+      "learning_rate": 0.0003940841865756541,
+      "loss": 5.6633,
+      "step": 5326
+    },
+    {
+      "epoch": 1.8180887372013652,
+      "grad_norm": 3.156517744064331,
+      "learning_rate": 0.0003939704209328783,
+      "loss": 6.3426,
+      "step": 5327
+    },
+    {
+      "epoch": 1.8184300341296928,
+      "grad_norm": 3.5125882625579834,
+      "learning_rate": 0.0003938566552901024,
+      "loss": 5.8847,
+      "step": 5328
+    },
+    {
+      "epoch": 1.8187713310580205,
+      "grad_norm": 3.0242371559143066,
+      "learning_rate": 0.0003937428896473265,
+      "loss": 6.1378,
+      "step": 5329
+    },
+    {
+      "epoch": 1.819112627986348,
+      "grad_norm": 3.081592559814453,
+      "learning_rate": 0.00039362912400455065,
+      "loss": 6.0888,
+      "step": 5330
+    },
+    {
+      "epoch": 1.8194539249146757,
+      "grad_norm": 2.9294071197509766,
+      "learning_rate": 0.00039351535836177476,
+      "loss": 6.4891,
+      "step": 5331
+    },
+    {
+      "epoch": 1.8197952218430036,
+      "grad_norm": 2.9522759914398193,
+      "learning_rate": 0.00039340159271899886,
+      "loss": 6.6007,
+      "step": 5332
+    },
+    {
+      "epoch": 1.820136518771331,
+      "grad_norm": 2.9654457569122314,
+      "learning_rate": 0.000393287827076223,
+      "loss": 6.6926,
+      "step": 5333
+    },
+    {
+      "epoch": 1.8204778156996588,
+      "grad_norm": 3.4901723861694336,
+      "learning_rate": 0.0003931740614334471,
+      "loss": 4.3765,
+      "step": 5334
+    },
+    {
+      "epoch": 1.8208191126279862,
+      "grad_norm": 4.748300552368164,
+      "learning_rate": 0.00039306029579067123,
+      "loss": 5.744,
+      "step": 5335
+    },
+    {
+      "epoch": 1.821160409556314,
+      "grad_norm": 2.887054204940796,
+      "learning_rate": 0.00039294653014789534,
+      "loss": 6.1389,
+      "step": 5336
+    },
+    {
+      "epoch": 1.8215017064846415,
+      "grad_norm": 2.9811062812805176,
+      "learning_rate": 0.00039283276450511944,
+      "loss": 6.3022,
+      "step": 5337
+    },
+    {
+      "epoch": 1.8218430034129693,
+      "grad_norm": 3.0545904636383057,
+      "learning_rate": 0.0003927189988623436,
+      "loss": 6.1094,
+      "step": 5338
+    },
+    {
+      "epoch": 1.822184300341297,
+      "grad_norm": 3.0074126720428467,
+      "learning_rate": 0.0003926052332195677,
+      "loss": 6.274,
+      "step": 5339
+    },
+    {
+      "epoch": 1.8225255972696246,
+      "grad_norm": 2.941384792327881,
+      "learning_rate": 0.0003924914675767918,
+      "loss": 6.2696,
+      "step": 5340
+    },
+    {
+      "epoch": 1.8228668941979522,
+      "grad_norm": 2.8732473850250244,
+      "learning_rate": 0.00039237770193401597,
+      "loss": 6.7092,
+      "step": 5341
+    },
+    {
+      "epoch": 1.8232081911262799,
+      "grad_norm": 2.9521396160125732,
+      "learning_rate": 0.0003922639362912401,
+      "loss": 6.506,
+      "step": 5342
+    },
+    {
+      "epoch": 1.8235494880546075,
+      "grad_norm": 2.961531162261963,
+      "learning_rate": 0.0003921501706484642,
+      "loss": 6.3482,
+      "step": 5343
+    },
+    {
+      "epoch": 1.823890784982935,
+      "grad_norm": 3.0271778106689453,
+      "learning_rate": 0.0003920364050056883,
+      "loss": 5.76,
+      "step": 5344
+    },
+    {
+      "epoch": 1.824232081911263,
+      "grad_norm": 2.9559972286224365,
+      "learning_rate": 0.0003919226393629124,
+      "loss": 6.5723,
+      "step": 5345
+    },
+    {
+      "epoch": 1.8245733788395904,
+      "grad_norm": 2.9276723861694336,
+      "learning_rate": 0.0003918088737201365,
+      "loss": 6.3815,
+      "step": 5346
+    },
+    {
+      "epoch": 1.8249146757679182,
+      "grad_norm": 2.9702205657958984,
+      "learning_rate": 0.00039169510807736066,
+      "loss": 6.6194,
+      "step": 5347
+    },
+    {
+      "epoch": 1.8252559726962456,
+      "grad_norm": 3.0158731937408447,
+      "learning_rate": 0.00039158134243458476,
+      "loss": 6.287,
+      "step": 5348
+    },
+    {
+      "epoch": 1.8255972696245735,
+      "grad_norm": 3.1500322818756104,
+      "learning_rate": 0.00039146757679180887,
+      "loss": 6.433,
+      "step": 5349
+    },
+    {
+      "epoch": 1.8259385665529009,
+      "grad_norm": 3.0445759296417236,
+      "learning_rate": 0.000391353811149033,
+      "loss": 6.2129,
+      "step": 5350
+    },
+    {
+      "epoch": 1.8262798634812287,
+      "grad_norm": 3.0558390617370605,
+      "learning_rate": 0.00039124004550625713,
+      "loss": 6.182,
+      "step": 5351
+    },
+    {
+      "epoch": 1.8266211604095564,
+      "grad_norm": 3.010730504989624,
+      "learning_rate": 0.00039112627986348123,
+      "loss": 5.7881,
+      "step": 5352
+    },
+    {
+      "epoch": 1.826962457337884,
+      "grad_norm": 2.9294490814208984,
+      "learning_rate": 0.00039101251422070534,
+      "loss": 6.0147,
+      "step": 5353
+    },
+    {
+      "epoch": 1.8273037542662116,
+      "grad_norm": 2.883665084838867,
+      "learning_rate": 0.00039089874857792944,
+      "loss": 6.3523,
+      "step": 5354
+    },
+    {
+      "epoch": 1.8276450511945392,
+      "grad_norm": 3.021610975265503,
+      "learning_rate": 0.0003907849829351536,
+      "loss": 6.6719,
+      "step": 5355
+    },
+    {
+      "epoch": 1.8279863481228669,
+      "grad_norm": 3.135248899459839,
+      "learning_rate": 0.0003906712172923777,
+      "loss": 6.1267,
+      "step": 5356
+    },
+    {
+      "epoch": 1.8283276450511945,
+      "grad_norm": 3.0037925243377686,
+      "learning_rate": 0.0003905574516496018,
+      "loss": 5.9688,
+      "step": 5357
+    },
+    {
+      "epoch": 1.8286689419795223,
+      "grad_norm": 3.0769035816192627,
+      "learning_rate": 0.00039044368600682597,
+      "loss": 5.994,
+      "step": 5358
+    },
+    {
+      "epoch": 1.8290102389078498,
+      "grad_norm": 2.995614767074585,
+      "learning_rate": 0.0003903299203640501,
+      "loss": 6.2174,
+      "step": 5359
+    },
+    {
+      "epoch": 1.8293515358361776,
+      "grad_norm": 3.0436437129974365,
+      "learning_rate": 0.0003902161547212742,
+      "loss": 5.6113,
+      "step": 5360
+    },
+    {
+      "epoch": 1.829692832764505,
+      "grad_norm": 2.8753719329833984,
+      "learning_rate": 0.00039010238907849834,
+      "loss": 5.9463,
+      "step": 5361
+    },
+    {
+      "epoch": 1.8300341296928329,
+      "grad_norm": 2.950981378555298,
+      "learning_rate": 0.00038998862343572245,
+      "loss": 5.9652,
+      "step": 5362
+    },
+    {
+      "epoch": 1.8303754266211603,
+      "grad_norm": 3.0115854740142822,
+      "learning_rate": 0.0003898748577929465,
+      "loss": 6.1722,
+      "step": 5363
+    },
+    {
+      "epoch": 1.8307167235494881,
+      "grad_norm": 3.0039737224578857,
+      "learning_rate": 0.00038976109215017066,
+      "loss": 6.8049,
+      "step": 5364
+    },
+    {
+      "epoch": 1.8310580204778157,
+      "grad_norm": 2.8962740898132324,
+      "learning_rate": 0.00038964732650739476,
+      "loss": 6.1094,
+      "step": 5365
+    },
+    {
+      "epoch": 1.8313993174061434,
+      "grad_norm": 3.1070163249969482,
+      "learning_rate": 0.00038953356086461887,
+      "loss": 5.8425,
+      "step": 5366
+    },
+    {
+      "epoch": 1.831740614334471,
+      "grad_norm": 4.08699893951416,
+      "learning_rate": 0.000389419795221843,
+      "loss": 4.5633,
+      "step": 5367
+    },
+    {
+      "epoch": 1.8320819112627986,
+      "grad_norm": 3.1629393100738525,
+      "learning_rate": 0.00038930602957906713,
+      "loss": 6.2469,
+      "step": 5368
+    },
+    {
+      "epoch": 1.8324232081911263,
+      "grad_norm": 3.210838556289673,
+      "learning_rate": 0.00038919226393629124,
+      "loss": 6.6676,
+      "step": 5369
+    },
+    {
+      "epoch": 1.8327645051194539,
+      "grad_norm": 3.3137311935424805,
+      "learning_rate": 0.0003890784982935154,
+      "loss": 5.3108,
+      "step": 5370
+    },
+    {
+      "epoch": 1.8331058020477817,
+      "grad_norm": 2.9676313400268555,
+      "learning_rate": 0.0003889647326507395,
+      "loss": 6.4932,
+      "step": 5371
+    },
+    {
+      "epoch": 1.8334470989761091,
+      "grad_norm": 2.9483461380004883,
+      "learning_rate": 0.0003888509670079636,
+      "loss": 6.2493,
+      "step": 5372
+    },
+    {
+      "epoch": 1.833788395904437,
+      "grad_norm": 3.206331491470337,
+      "learning_rate": 0.0003887372013651877,
+      "loss": 5.3646,
+      "step": 5373
+    },
+    {
+      "epoch": 1.8341296928327644,
+      "grad_norm": 3.0024166107177734,
+      "learning_rate": 0.0003886234357224118,
+      "loss": 6.3351,
+      "step": 5374
+    },
+    {
+      "epoch": 1.8344709897610922,
+      "grad_norm": 3.38862943649292,
+      "learning_rate": 0.0003885096700796359,
+      "loss": 6.2358,
+      "step": 5375
+    },
+    {
+      "epoch": 1.8348122866894196,
+      "grad_norm": 2.8507230281829834,
+      "learning_rate": 0.0003883959044368601,
+      "loss": 5.6975,
+      "step": 5376
+    },
+    {
+      "epoch": 1.8351535836177475,
+      "grad_norm": 2.947876214981079,
+      "learning_rate": 0.0003882821387940842,
+      "loss": 6.2397,
+      "step": 5377
+    },
+    {
+      "epoch": 1.8354948805460751,
+      "grad_norm": 3.339482307434082,
+      "learning_rate": 0.00038816837315130834,
+      "loss": 5.5517,
+      "step": 5378
+    },
+    {
+      "epoch": 1.8358361774744028,
+      "grad_norm": 2.9442296028137207,
+      "learning_rate": 0.00038805460750853245,
+      "loss": 6.5012,
+      "step": 5379
+    },
+    {
+      "epoch": 1.8361774744027304,
+      "grad_norm": 3.01493239402771,
+      "learning_rate": 0.00038794084186575655,
+      "loss": 6.1444,
+      "step": 5380
+    },
+    {
+      "epoch": 1.836518771331058,
+      "grad_norm": 3.0220446586608887,
+      "learning_rate": 0.0003878270762229807,
+      "loss": 6.8928,
+      "step": 5381
+    },
+    {
+      "epoch": 1.8368600682593856,
+      "grad_norm": 2.9548652172088623,
+      "learning_rate": 0.00038771331058020476,
+      "loss": 6.0756,
+      "step": 5382
+    },
+    {
+      "epoch": 1.8372013651877133,
+      "grad_norm": 3.209747314453125,
+      "learning_rate": 0.00038759954493742887,
+      "loss": 5.5828,
+      "step": 5383
+    },
+    {
+      "epoch": 1.8375426621160411,
+      "grad_norm": 3.493947744369507,
+      "learning_rate": 0.000387485779294653,
+      "loss": 6.0699,
+      "step": 5384
+    },
+    {
+      "epoch": 1.8378839590443685,
+      "grad_norm": 3.8855106830596924,
+      "learning_rate": 0.00038737201365187713,
+      "loss": 4.7189,
+      "step": 5385
+    },
+    {
+      "epoch": 1.8382252559726964,
+      "grad_norm": 3.2830464839935303,
+      "learning_rate": 0.00038725824800910124,
+      "loss": 5.7651,
+      "step": 5386
+    },
+    {
+      "epoch": 1.8385665529010238,
+      "grad_norm": 3.040435314178467,
+      "learning_rate": 0.0003871444823663254,
+      "loss": 5.9345,
+      "step": 5387
+    },
+    {
+      "epoch": 1.8389078498293516,
+      "grad_norm": 3.311544179916382,
+      "learning_rate": 0.0003870307167235495,
+      "loss": 5.9783,
+      "step": 5388
+    },
+    {
+      "epoch": 1.839249146757679,
+      "grad_norm": 2.9888603687286377,
+      "learning_rate": 0.0003869169510807736,
+      "loss": 5.666,
+      "step": 5389
+    },
+    {
+      "epoch": 1.8395904436860069,
+      "grad_norm": 3.0224196910858154,
+      "learning_rate": 0.00038680318543799776,
+      "loss": 6.6377,
+      "step": 5390
+    },
+    {
+      "epoch": 1.8399317406143345,
+      "grad_norm": 3.0179591178894043,
+      "learning_rate": 0.00038668941979522187,
+      "loss": 6.0367,
+      "step": 5391
+    },
+    {
+      "epoch": 1.8402730375426621,
+      "grad_norm": 2.922994375228882,
+      "learning_rate": 0.0003865756541524459,
+      "loss": 6.0782,
+      "step": 5392
+    },
+    {
+      "epoch": 1.8406143344709898,
+      "grad_norm": 2.8612077236175537,
+      "learning_rate": 0.0003864618885096701,
+      "loss": 6.0352,
+      "step": 5393
+    },
+    {
+      "epoch": 1.8409556313993174,
+      "grad_norm": 2.957252025604248,
+      "learning_rate": 0.0003863481228668942,
+      "loss": 6.5599,
+      "step": 5394
+    },
+    {
+      "epoch": 1.841296928327645,
+      "grad_norm": 3.019355535507202,
+      "learning_rate": 0.0003862343572241183,
+      "loss": 5.697,
+      "step": 5395
+    },
+    {
+      "epoch": 1.8416382252559726,
+      "grad_norm": 2.8055245876312256,
+      "learning_rate": 0.00038612059158134245,
+      "loss": 6.1868,
+      "step": 5396
+    },
+    {
+      "epoch": 1.8419795221843005,
+      "grad_norm": 3.0640079975128174,
+      "learning_rate": 0.00038600682593856655,
+      "loss": 6.252,
+      "step": 5397
+    },
+    {
+      "epoch": 1.842320819112628,
+      "grad_norm": 2.855372428894043,
+      "learning_rate": 0.0003858930602957907,
+      "loss": 6.6329,
+      "step": 5398
+    },
+    {
+      "epoch": 1.8426621160409558,
+      "grad_norm": 2.9066994190216064,
+      "learning_rate": 0.0003857792946530148,
+      "loss": 6.5626,
+      "step": 5399
+    },
+    {
+      "epoch": 1.8430034129692832,
+      "grad_norm": 2.871976852416992,
+      "learning_rate": 0.0003856655290102389,
+      "loss": 6.0375,
+      "step": 5400
+    },
+    {
+      "epoch": 1.843344709897611,
+      "grad_norm": 3.041769504547119,
+      "learning_rate": 0.0003855517633674631,
+      "loss": 6.551,
+      "step": 5401
+    },
+    {
+      "epoch": 1.8436860068259384,
+      "grad_norm": 3.0764667987823486,
+      "learning_rate": 0.00038543799772468713,
+      "loss": 6.187,
+      "step": 5402
+    },
+    {
+      "epoch": 1.8440273037542663,
+      "grad_norm": 3.1484851837158203,
+      "learning_rate": 0.00038532423208191124,
+      "loss": 6.0588,
+      "step": 5403
+    },
+    {
+      "epoch": 1.844368600682594,
+      "grad_norm": 3.08003306388855,
+      "learning_rate": 0.0003852104664391354,
+      "loss": 6.6638,
+      "step": 5404
+    },
+    {
+      "epoch": 1.8447098976109215,
+      "grad_norm": 2.9924423694610596,
+      "learning_rate": 0.0003850967007963595,
+      "loss": 6.4312,
+      "step": 5405
+    },
+    {
+      "epoch": 1.8450511945392492,
+      "grad_norm": 3.137577533721924,
+      "learning_rate": 0.0003849829351535836,
+      "loss": 6.4421,
+      "step": 5406
+    },
+    {
+      "epoch": 1.8453924914675768,
+      "grad_norm": 2.974426507949829,
+      "learning_rate": 0.00038486916951080777,
+      "loss": 6.4856,
+      "step": 5407
+    },
+    {
+      "epoch": 1.8457337883959044,
+      "grad_norm": 3.020400047302246,
+      "learning_rate": 0.00038475540386803187,
+      "loss": 6.534,
+      "step": 5408
+    },
+    {
+      "epoch": 1.846075085324232,
+      "grad_norm": 2.938917636871338,
+      "learning_rate": 0.000384641638225256,
+      "loss": 6.6811,
+      "step": 5409
+    },
+    {
+      "epoch": 1.8464163822525599,
+      "grad_norm": 3.0424704551696777,
+      "learning_rate": 0.00038452787258248013,
+      "loss": 6.4537,
+      "step": 5410
+    },
+    {
+      "epoch": 1.8467576791808873,
+      "grad_norm": 3.186312198638916,
+      "learning_rate": 0.00038441410693970424,
+      "loss": 6.5699,
+      "step": 5411
+    },
+    {
+      "epoch": 1.8470989761092151,
+      "grad_norm": 2.9400906562805176,
+      "learning_rate": 0.0003843003412969283,
+      "loss": 6.4694,
+      "step": 5412
+    },
+    {
+      "epoch": 1.8474402730375425,
+      "grad_norm": 3.567128896713257,
+      "learning_rate": 0.00038418657565415245,
+      "loss": 4.6422,
+      "step": 5413
+    },
+    {
+      "epoch": 1.8477815699658704,
+      "grad_norm": 5.326208591461182,
+      "learning_rate": 0.00038407281001137655,
+      "loss": 5.3053,
+      "step": 5414
+    },
+    {
+      "epoch": 1.8481228668941978,
+      "grad_norm": 3.0531771183013916,
+      "learning_rate": 0.00038395904436860066,
+      "loss": 5.9597,
+      "step": 5415
+    },
+    {
+      "epoch": 1.8484641638225257,
+      "grad_norm": 3.1060829162597656,
+      "learning_rate": 0.0003838452787258248,
+      "loss": 6.5521,
+      "step": 5416
+    },
+    {
+      "epoch": 1.8488054607508533,
+      "grad_norm": 3.0353734493255615,
+      "learning_rate": 0.0003837315130830489,
+      "loss": 6.4911,
+      "step": 5417
+    },
+    {
+      "epoch": 1.849146757679181,
+      "grad_norm": 2.9575839042663574,
+      "learning_rate": 0.0003836177474402731,
+      "loss": 6.3016,
+      "step": 5418
+    },
+    {
+      "epoch": 1.8494880546075085,
+      "grad_norm": 2.9795007705688477,
+      "learning_rate": 0.0003835039817974972,
+      "loss": 6.3147,
+      "step": 5419
+    },
+    {
+      "epoch": 1.8498293515358362,
+      "grad_norm": 2.8920490741729736,
+      "learning_rate": 0.0003833902161547213,
+      "loss": 6.7194,
+      "step": 5420
+    },
+    {
+      "epoch": 1.8501706484641638,
+      "grad_norm": 3.0831782817840576,
+      "learning_rate": 0.0003832764505119454,
+      "loss": 6.2686,
+      "step": 5421
+    },
+    {
+      "epoch": 1.8505119453924914,
+      "grad_norm": 2.8378100395202637,
+      "learning_rate": 0.0003831626848691695,
+      "loss": 6.5928,
+      "step": 5422
+    },
+    {
+      "epoch": 1.8508532423208193,
+      "grad_norm": 2.8427324295043945,
+      "learning_rate": 0.0003830489192263936,
+      "loss": 6.1053,
+      "step": 5423
+    },
+    {
+      "epoch": 1.8511945392491467,
+      "grad_norm": 3.0172784328460693,
+      "learning_rate": 0.00038293515358361777,
+      "loss": 6.1388,
+      "step": 5424
+    },
+    {
+      "epoch": 1.8515358361774745,
+      "grad_norm": 2.933791160583496,
+      "learning_rate": 0.00038282138794084187,
+      "loss": 6.4189,
+      "step": 5425
+    },
+    {
+      "epoch": 1.851877133105802,
+      "grad_norm": 2.942861557006836,
+      "learning_rate": 0.000382707622298066,
+      "loss": 6.5474,
+      "step": 5426
+    },
+    {
+      "epoch": 1.8522184300341298,
+      "grad_norm": 6.984040260314941,
+      "learning_rate": 0.00038259385665529014,
+      "loss": 6.0704,
+      "step": 5427
+    },
+    {
+      "epoch": 1.8525597269624572,
+      "grad_norm": 3.091132640838623,
+      "learning_rate": 0.00038248009101251424,
+      "loss": 6.6446,
+      "step": 5428
+    },
+    {
+      "epoch": 1.852901023890785,
+      "grad_norm": 3.019805908203125,
+      "learning_rate": 0.00038236632536973835,
+      "loss": 5.9389,
+      "step": 5429
+    },
+    {
+      "epoch": 1.8532423208191127,
+      "grad_norm": 3.0177457332611084,
+      "learning_rate": 0.0003822525597269625,
+      "loss": 6.3112,
+      "step": 5430
+    },
+    {
+      "epoch": 1.8535836177474403,
+      "grad_norm": 2.8653550148010254,
+      "learning_rate": 0.00038213879408418656,
+      "loss": 6.0127,
+      "step": 5431
+    },
+    {
+      "epoch": 1.853924914675768,
+      "grad_norm": 2.9369118213653564,
+      "learning_rate": 0.00038202502844141066,
+      "loss": 5.6815,
+      "step": 5432
+    },
+    {
+      "epoch": 1.8542662116040955,
+      "grad_norm": 2.954627513885498,
+      "learning_rate": 0.0003819112627986348,
+      "loss": 6.5244,
+      "step": 5433
+    },
+    {
+      "epoch": 1.8546075085324232,
+      "grad_norm": 2.8261237144470215,
+      "learning_rate": 0.0003817974971558589,
+      "loss": 6.4313,
+      "step": 5434
+    },
+    {
+      "epoch": 1.8549488054607508,
+      "grad_norm": 3.0321621894836426,
+      "learning_rate": 0.00038168373151308303,
+      "loss": 6.5262,
+      "step": 5435
+    },
+    {
+      "epoch": 1.8552901023890787,
+      "grad_norm": 3.0571656227111816,
+      "learning_rate": 0.0003815699658703072,
+      "loss": 6.4915,
+      "step": 5436
+    },
+    {
+      "epoch": 1.855631399317406,
+      "grad_norm": 2.8418610095977783,
+      "learning_rate": 0.0003814562002275313,
+      "loss": 6.0628,
+      "step": 5437
+    },
+    {
+      "epoch": 1.855972696245734,
+      "grad_norm": 3.0574285984039307,
+      "learning_rate": 0.00038134243458475545,
+      "loss": 6.2824,
+      "step": 5438
+    },
+    {
+      "epoch": 1.8563139931740613,
+      "grad_norm": 2.951812267303467,
+      "learning_rate": 0.00038122866894197956,
+      "loss": 6.4227,
+      "step": 5439
+    },
+    {
+      "epoch": 1.8566552901023892,
+      "grad_norm": 2.9646008014678955,
+      "learning_rate": 0.00038111490329920366,
+      "loss": 6.4111,
+      "step": 5440
+    },
+    {
+      "epoch": 1.8569965870307166,
+      "grad_norm": 2.9364631175994873,
+      "learning_rate": 0.00038100113765642777,
+      "loss": 6.3866,
+      "step": 5441
+    },
+    {
+      "epoch": 1.8573378839590444,
+      "grad_norm": 4.254305839538574,
+      "learning_rate": 0.00038088737201365187,
+      "loss": 5.721,
+      "step": 5442
+    },
+    {
+      "epoch": 1.857679180887372,
+      "grad_norm": 2.9221060276031494,
+      "learning_rate": 0.000380773606370876,
+      "loss": 5.6808,
+      "step": 5443
+    },
+    {
+      "epoch": 1.8580204778156997,
+      "grad_norm": 2.9992268085479736,
+      "learning_rate": 0.00038065984072810014,
+      "loss": 5.8649,
+      "step": 5444
+    },
+    {
+      "epoch": 1.8583617747440273,
+      "grad_norm": 2.9220197200775146,
+      "learning_rate": 0.00038054607508532424,
+      "loss": 5.8822,
+      "step": 5445
+    },
+    {
+      "epoch": 1.858703071672355,
+      "grad_norm": 2.9059829711914062,
+      "learning_rate": 0.00038043230944254835,
+      "loss": 6.4213,
+      "step": 5446
+    },
+    {
+      "epoch": 1.8590443686006826,
+      "grad_norm": 3.104856252670288,
+      "learning_rate": 0.0003803185437997725,
+      "loss": 5.7524,
+      "step": 5447
+    },
+    {
+      "epoch": 1.8593856655290102,
+      "grad_norm": 2.9057981967926025,
+      "learning_rate": 0.0003802047781569966,
+      "loss": 6.5828,
+      "step": 5448
+    },
+    {
+      "epoch": 1.859726962457338,
+      "grad_norm": 2.952005624771118,
+      "learning_rate": 0.0003800910125142207,
+      "loss": 6.5241,
+      "step": 5449
+    },
+    {
+      "epoch": 1.8600682593856654,
+      "grad_norm": 2.8778185844421387,
+      "learning_rate": 0.0003799772468714448,
+      "loss": 6.7302,
+      "step": 5450
+    },
+    {
+      "epoch": 1.8604095563139933,
+      "grad_norm": 2.990785837173462,
+      "learning_rate": 0.0003798634812286689,
+      "loss": 5.8828,
+      "step": 5451
+    },
+    {
+      "epoch": 1.8607508532423207,
+      "grad_norm": 2.9240405559539795,
+      "learning_rate": 0.00037974971558589303,
+      "loss": 6.4997,
+      "step": 5452
+    },
+    {
+      "epoch": 1.8610921501706486,
+      "grad_norm": 3.0843167304992676,
+      "learning_rate": 0.0003796359499431172,
+      "loss": 5.7164,
+      "step": 5453
+    },
+    {
+      "epoch": 1.861433447098976,
+      "grad_norm": 2.9468958377838135,
+      "learning_rate": 0.0003795221843003413,
+      "loss": 6.3823,
+      "step": 5454
+    },
+    {
+      "epoch": 1.8617747440273038,
+      "grad_norm": 2.890228271484375,
+      "learning_rate": 0.0003794084186575654,
+      "loss": 6.1479,
+      "step": 5455
+    },
+    {
+      "epoch": 1.8621160409556314,
+      "grad_norm": 7.07792329788208,
+      "learning_rate": 0.00037929465301478956,
+      "loss": 4.19,
+      "step": 5456
+    },
+    {
+      "epoch": 1.862457337883959,
+      "grad_norm": 2.9772236347198486,
+      "learning_rate": 0.00037918088737201366,
+      "loss": 6.5797,
+      "step": 5457
+    },
+    {
+      "epoch": 1.8627986348122867,
+      "grad_norm": 3.0900766849517822,
+      "learning_rate": 0.0003790671217292378,
+      "loss": 5.6404,
+      "step": 5458
+    },
+    {
+      "epoch": 1.8631399317406143,
+      "grad_norm": 3.135882616043091,
+      "learning_rate": 0.00037895335608646193,
+      "loss": 6.2669,
+      "step": 5459
+    },
+    {
+      "epoch": 1.863481228668942,
+      "grad_norm": 3.0472939014434814,
+      "learning_rate": 0.000378839590443686,
+      "loss": 5.6298,
+      "step": 5460
+    },
+    {
+      "epoch": 1.8638225255972696,
+      "grad_norm": 3.0007853507995605,
+      "learning_rate": 0.00037872582480091014,
+      "loss": 5.4573,
+      "step": 5461
+    },
+    {
+      "epoch": 1.8641638225255974,
+      "grad_norm": 2.992143154144287,
+      "learning_rate": 0.00037861205915813424,
+      "loss": 6.2351,
+      "step": 5462
+    },
+    {
+      "epoch": 1.8645051194539248,
+      "grad_norm": 2.8720524311065674,
+      "learning_rate": 0.00037849829351535835,
+      "loss": 5.7451,
+      "step": 5463
+    },
+    {
+      "epoch": 1.8648464163822527,
+      "grad_norm": 3.13032603263855,
+      "learning_rate": 0.0003783845278725825,
+      "loss": 5.7272,
+      "step": 5464
+    },
+    {
+      "epoch": 1.86518771331058,
+      "grad_norm": 3.009556293487549,
+      "learning_rate": 0.0003782707622298066,
+      "loss": 5.9469,
+      "step": 5465
+    },
+    {
+      "epoch": 1.865529010238908,
+      "grad_norm": 2.857898712158203,
+      "learning_rate": 0.0003781569965870307,
+      "loss": 5.6815,
+      "step": 5466
+    },
+    {
+      "epoch": 1.8658703071672353,
+      "grad_norm": 2.863910675048828,
+      "learning_rate": 0.0003780432309442549,
+      "loss": 6.0975,
+      "step": 5467
+    },
+    {
+      "epoch": 1.8662116040955632,
+      "grad_norm": 2.9575035572052,
+      "learning_rate": 0.000377929465301479,
+      "loss": 6.5311,
+      "step": 5468
+    },
+    {
+      "epoch": 1.8665529010238908,
+      "grad_norm": 3.4162585735321045,
+      "learning_rate": 0.0003778156996587031,
+      "loss": 5.9707,
+      "step": 5469
+    },
+    {
+      "epoch": 1.8668941979522184,
+      "grad_norm": 3.0944206714630127,
+      "learning_rate": 0.0003777019340159272,
+      "loss": 6.2304,
+      "step": 5470
+    },
+    {
+      "epoch": 1.867235494880546,
+      "grad_norm": 3.3516929149627686,
+      "learning_rate": 0.0003775881683731513,
+      "loss": 5.7523,
+      "step": 5471
+    },
+    {
+      "epoch": 1.8675767918088737,
+      "grad_norm": 2.970231294631958,
+      "learning_rate": 0.0003774744027303754,
+      "loss": 5.5873,
+      "step": 5472
+    },
+    {
+      "epoch": 1.8679180887372013,
+      "grad_norm": 3.073521137237549,
+      "learning_rate": 0.00037736063708759956,
+      "loss": 6.0575,
+      "step": 5473
+    },
+    {
+      "epoch": 1.868259385665529,
+      "grad_norm": 2.949779748916626,
+      "learning_rate": 0.00037724687144482366,
+      "loss": 6.1106,
+      "step": 5474
+    },
+    {
+      "epoch": 1.8686006825938568,
+      "grad_norm": 2.9631776809692383,
+      "learning_rate": 0.00037713310580204777,
+      "loss": 5.8668,
+      "step": 5475
+    },
+    {
+      "epoch": 1.8689419795221842,
+      "grad_norm": 2.9993419647216797,
+      "learning_rate": 0.00037701934015927193,
+      "loss": 6.5943,
+      "step": 5476
+    },
+    {
+      "epoch": 1.869283276450512,
+      "grad_norm": 2.911229133605957,
+      "learning_rate": 0.00037690557451649603,
+      "loss": 6.1717,
+      "step": 5477
+    },
+    {
+      "epoch": 1.8696245733788395,
+      "grad_norm": 3.7616472244262695,
+      "learning_rate": 0.0003767918088737202,
+      "loss": 4.3073,
+      "step": 5478
+    },
+    {
+      "epoch": 1.8699658703071673,
+      "grad_norm": 3.202282190322876,
+      "learning_rate": 0.0003766780432309443,
+      "loss": 5.8356,
+      "step": 5479
+    },
+    {
+      "epoch": 1.8703071672354947,
+      "grad_norm": 3.0891096591949463,
+      "learning_rate": 0.00037656427758816835,
+      "loss": 4.9737,
+      "step": 5480
+    },
+    {
+      "epoch": 1.8706484641638226,
+      "grad_norm": 1.8888578414916992,
+      "learning_rate": 0.0003764505119453925,
+      "loss": 2.898,
+      "step": 5481
+    },
+    {
+      "epoch": 1.8709897610921502,
+      "grad_norm": 3.4496963024139404,
+      "learning_rate": 0.0003763367463026166,
+      "loss": 5.7924,
+      "step": 5482
+    },
+    {
+      "epoch": 1.8713310580204778,
+      "grad_norm": 3.0166397094726562,
+      "learning_rate": 0.0003762229806598407,
+      "loss": 6.2951,
+      "step": 5483
+    },
+    {
+      "epoch": 1.8716723549488055,
+      "grad_norm": 3.0307724475860596,
+      "learning_rate": 0.0003761092150170649,
+      "loss": 6.5138,
+      "step": 5484
+    },
+    {
+      "epoch": 1.872013651877133,
+      "grad_norm": 2.994663953781128,
+      "learning_rate": 0.000375995449374289,
+      "loss": 6.3472,
+      "step": 5485
+    },
+    {
+      "epoch": 1.8723549488054607,
+      "grad_norm": 2.9022762775421143,
+      "learning_rate": 0.0003758816837315131,
+      "loss": 5.8626,
+      "step": 5486
+    },
+    {
+      "epoch": 1.8726962457337883,
+      "grad_norm": 2.9295828342437744,
+      "learning_rate": 0.00037576791808873725,
+      "loss": 5.9134,
+      "step": 5487
+    },
+    {
+      "epoch": 1.8730375426621162,
+      "grad_norm": 2.866713047027588,
+      "learning_rate": 0.00037565415244596135,
+      "loss": 6.403,
+      "step": 5488
+    },
+    {
+      "epoch": 1.8733788395904436,
+      "grad_norm": 2.9316976070404053,
+      "learning_rate": 0.0003755403868031854,
+      "loss": 5.9694,
+      "step": 5489
+    },
+    {
+      "epoch": 1.8737201365187715,
+      "grad_norm": 3.494288682937622,
+      "learning_rate": 0.00037542662116040956,
+      "loss": 6.0734,
+      "step": 5490
+    },
+    {
+      "epoch": 1.8740614334470989,
+      "grad_norm": 7.175894260406494,
+      "learning_rate": 0.00037531285551763367,
+      "loss": 5.6944,
+      "step": 5491
+    },
+    {
+      "epoch": 1.8744027303754267,
+      "grad_norm": 3.8673219680786133,
+      "learning_rate": 0.00037519908987485777,
+      "loss": 6.0276,
+      "step": 5492
+    },
+    {
+      "epoch": 1.8747440273037541,
+      "grad_norm": 3.381589412689209,
+      "learning_rate": 0.00037508532423208193,
+      "loss": 5.8487,
+      "step": 5493
+    },
+    {
+      "epoch": 1.875085324232082,
+      "grad_norm": 3.1894888877868652,
+      "learning_rate": 0.00037497155858930603,
+      "loss": 6.3183,
+      "step": 5494
+    },
+    {
+      "epoch": 1.8754266211604096,
+      "grad_norm": 3.0183558464050293,
+      "learning_rate": 0.00037485779294653014,
+      "loss": 6.1961,
+      "step": 5495
+    },
+    {
+      "epoch": 1.8757679180887372,
+      "grad_norm": 2.9099154472351074,
+      "learning_rate": 0.0003747440273037543,
+      "loss": 6.3843,
+      "step": 5496
+    },
+    {
+      "epoch": 1.8761092150170648,
+      "grad_norm": 3.4501168727874756,
+      "learning_rate": 0.0003746302616609784,
+      "loss": 5.7031,
+      "step": 5497
+    },
+    {
+      "epoch": 1.8764505119453925,
+      "grad_norm": 4.108460426330566,
+      "learning_rate": 0.0003745164960182025,
+      "loss": 5.5165,
+      "step": 5498
+    },
+    {
+      "epoch": 1.87679180887372,
+      "grad_norm": 2.9657464027404785,
+      "learning_rate": 0.0003744027303754266,
+      "loss": 6.6402,
+      "step": 5499
+    },
+    {
+      "epoch": 1.8771331058020477,
+      "grad_norm": 2.9928600788116455,
+      "learning_rate": 0.0003742889647326507,
+      "loss": 5.5103,
+      "step": 5500
+    },
+    {
+      "epoch": 1.8774744027303756,
+      "grad_norm": 2.88197922706604,
+      "learning_rate": 0.0003741751990898749,
+      "loss": 6.0788,
+      "step": 5501
+    },
+    {
+      "epoch": 1.877815699658703,
+      "grad_norm": 2.8990108966827393,
+      "learning_rate": 0.000374061433447099,
+      "loss": 5.9397,
+      "step": 5502
+    },
+    {
+      "epoch": 1.8781569965870308,
+      "grad_norm": 6.239475250244141,
+      "learning_rate": 0.0003739476678043231,
+      "loss": 4.31,
+      "step": 5503
+    },
+    {
+      "epoch": 1.8784982935153582,
+      "grad_norm": 2.9518520832061768,
+      "learning_rate": 0.00037383390216154725,
+      "loss": 6.4256,
+      "step": 5504
+    },
+    {
+      "epoch": 1.878839590443686,
+      "grad_norm": 3.101850748062134,
+      "learning_rate": 0.00037372013651877135,
+      "loss": 6.2821,
+      "step": 5505
+    },
+    {
+      "epoch": 1.8791808873720135,
+      "grad_norm": 2.954547643661499,
+      "learning_rate": 0.00037360637087599546,
+      "loss": 5.9507,
+      "step": 5506
+    },
+    {
+      "epoch": 1.8795221843003413,
+      "grad_norm": 3.1224567890167236,
+      "learning_rate": 0.0003734926052332196,
+      "loss": 6.0534,
+      "step": 5507
+    },
+    {
+      "epoch": 1.879863481228669,
+      "grad_norm": 2.9866816997528076,
+      "learning_rate": 0.0003733788395904437,
+      "loss": 6.8862,
+      "step": 5508
+    },
+    {
+      "epoch": 1.8802047781569966,
+      "grad_norm": 2.98016095161438,
+      "learning_rate": 0.00037326507394766777,
+      "loss": 6.7283,
+      "step": 5509
+    },
+    {
+      "epoch": 1.8805460750853242,
+      "grad_norm": 2.978584051132202,
+      "learning_rate": 0.00037315130830489193,
+      "loss": 5.4551,
+      "step": 5510
+    },
+    {
+      "epoch": 1.8808873720136519,
+      "grad_norm": 3.2457802295684814,
+      "learning_rate": 0.00037303754266211603,
+      "loss": 6.5083,
+      "step": 5511
+    },
+    {
+      "epoch": 1.8812286689419795,
+      "grad_norm": 2.834044933319092,
+      "learning_rate": 0.00037292377701934014,
+      "loss": 6.3825,
+      "step": 5512
+    },
+    {
+      "epoch": 1.8815699658703071,
+      "grad_norm": 3.000852346420288,
+      "learning_rate": 0.0003728100113765643,
+      "loss": 5.9047,
+      "step": 5513
+    },
+    {
+      "epoch": 1.881911262798635,
+      "grad_norm": 3.954674482345581,
+      "learning_rate": 0.0003726962457337884,
+      "loss": 5.7054,
+      "step": 5514
+    },
+    {
+      "epoch": 1.8822525597269624,
+      "grad_norm": 3.103377103805542,
+      "learning_rate": 0.0003725824800910125,
+      "loss": 6.532,
+      "step": 5515
+    },
+    {
+      "epoch": 1.8825938566552902,
+      "grad_norm": 3.011568546295166,
+      "learning_rate": 0.00037246871444823667,
+      "loss": 6.4145,
+      "step": 5516
+    },
+    {
+      "epoch": 1.8829351535836176,
+      "grad_norm": 2.94783091545105,
+      "learning_rate": 0.0003723549488054608,
+      "loss": 5.7669,
+      "step": 5517
+    },
+    {
+      "epoch": 1.8832764505119455,
+      "grad_norm": 3.0503170490264893,
+      "learning_rate": 0.0003722411831626849,
+      "loss": 6.1408,
+      "step": 5518
+    },
+    {
+      "epoch": 1.8836177474402729,
+      "grad_norm": 2.829469919204712,
+      "learning_rate": 0.000372127417519909,
+      "loss": 6.3711,
+      "step": 5519
+    },
+    {
+      "epoch": 1.8839590443686007,
+      "grad_norm": 2.903998374938965,
+      "learning_rate": 0.0003720136518771331,
+      "loss": 6.3651,
+      "step": 5520
+    },
+    {
+      "epoch": 1.8843003412969284,
+      "grad_norm": 3.0705928802490234,
+      "learning_rate": 0.00037189988623435725,
+      "loss": 5.9707,
+      "step": 5521
+    },
+    {
+      "epoch": 1.884641638225256,
+      "grad_norm": 3.0179994106292725,
+      "learning_rate": 0.00037178612059158135,
+      "loss": 5.813,
+      "step": 5522
+    },
+    {
+      "epoch": 1.8849829351535836,
+      "grad_norm": 3.0558056831359863,
+      "learning_rate": 0.00037167235494880546,
+      "loss": 5.9914,
+      "step": 5523
+    },
+    {
+      "epoch": 1.8853242320819112,
+      "grad_norm": 2.9075541496276855,
+      "learning_rate": 0.0003715585893060296,
+      "loss": 6.2369,
+      "step": 5524
+    },
+    {
+      "epoch": 1.8856655290102389,
+      "grad_norm": 2.8797600269317627,
+      "learning_rate": 0.0003714448236632537,
+      "loss": 6.2321,
+      "step": 5525
+    },
+    {
+      "epoch": 1.8860068259385665,
+      "grad_norm": 2.88400936126709,
+      "learning_rate": 0.0003713310580204778,
+      "loss": 5.2605,
+      "step": 5526
+    },
+    {
+      "epoch": 1.8863481228668944,
+      "grad_norm": 3.1179773807525635,
+      "learning_rate": 0.000371217292377702,
+      "loss": 6.1447,
+      "step": 5527
+    },
+    {
+      "epoch": 1.8866894197952218,
+      "grad_norm": 2.8629674911499023,
+      "learning_rate": 0.00037110352673492604,
+      "loss": 6.4715,
+      "step": 5528
+    },
+    {
+      "epoch": 1.8870307167235496,
+      "grad_norm": 2.896420478820801,
+      "learning_rate": 0.00037098976109215014,
+      "loss": 6.3566,
+      "step": 5529
+    },
+    {
+      "epoch": 1.887372013651877,
+      "grad_norm": 2.9856109619140625,
+      "learning_rate": 0.0003708759954493743,
+      "loss": 6.3204,
+      "step": 5530
+    },
+    {
+      "epoch": 1.8877133105802049,
+      "grad_norm": 3.28257417678833,
+      "learning_rate": 0.0003707622298065984,
+      "loss": 5.3285,
+      "step": 5531
+    },
+    {
+      "epoch": 1.8880546075085323,
+      "grad_norm": 2.9387850761413574,
+      "learning_rate": 0.0003706484641638225,
+      "loss": 6.0632,
+      "step": 5532
+    },
+    {
+      "epoch": 1.8883959044368601,
+      "grad_norm": 2.9324870109558105,
+      "learning_rate": 0.00037053469852104667,
+      "loss": 5.8282,
+      "step": 5533
+    },
+    {
+      "epoch": 1.8887372013651877,
+      "grad_norm": 2.979710340499878,
+      "learning_rate": 0.0003704209328782708,
+      "loss": 6.2292,
+      "step": 5534
+    },
+    {
+      "epoch": 1.8890784982935154,
+      "grad_norm": 2.9082748889923096,
+      "learning_rate": 0.0003703071672354949,
+      "loss": 6.0311,
+      "step": 5535
+    },
+    {
+      "epoch": 1.889419795221843,
+      "grad_norm": 2.8582496643066406,
+      "learning_rate": 0.00037019340159271904,
+      "loss": 6.5768,
+      "step": 5536
+    },
+    {
+      "epoch": 1.8897610921501706,
+      "grad_norm": 3.1130855083465576,
+      "learning_rate": 0.00037007963594994314,
+      "loss": 5.4812,
+      "step": 5537
+    },
+    {
+      "epoch": 1.8901023890784983,
+      "grad_norm": 3.0239624977111816,
+      "learning_rate": 0.0003699658703071672,
+      "loss": 6.754,
+      "step": 5538
+    },
+    {
+      "epoch": 1.8904436860068259,
+      "grad_norm": 2.8626439571380615,
+      "learning_rate": 0.00036985210466439135,
+      "loss": 6.0451,
+      "step": 5539
+    },
+    {
+      "epoch": 1.8907849829351537,
+      "grad_norm": 3.52713942527771,
+      "learning_rate": 0.00036973833902161546,
+      "loss": 5.5374,
+      "step": 5540
+    },
+    {
+      "epoch": 1.8911262798634811,
+      "grad_norm": 3.0411159992218018,
+      "learning_rate": 0.0003696245733788396,
+      "loss": 6.253,
+      "step": 5541
+    },
+    {
+      "epoch": 1.891467576791809,
+      "grad_norm": 3.1052937507629395,
+      "learning_rate": 0.0003695108077360637,
+      "loss": 5.3534,
+      "step": 5542
+    },
+    {
+      "epoch": 1.8918088737201364,
+      "grad_norm": 3.007514715194702,
+      "learning_rate": 0.0003693970420932878,
+      "loss": 5.9601,
+      "step": 5543
+    },
+    {
+      "epoch": 1.8921501706484642,
+      "grad_norm": 3.2602505683898926,
+      "learning_rate": 0.000369283276450512,
+      "loss": 5.6877,
+      "step": 5544
+    },
+    {
+      "epoch": 1.8924914675767917,
+      "grad_norm": 3.1978507041931152,
+      "learning_rate": 0.0003691695108077361,
+      "loss": 5.6045,
+      "step": 5545
+    },
+    {
+      "epoch": 1.8928327645051195,
+      "grad_norm": 2.927302360534668,
+      "learning_rate": 0.0003690557451649602,
+      "loss": 5.7842,
+      "step": 5546
+    },
+    {
+      "epoch": 1.8931740614334471,
+      "grad_norm": 3.1250858306884766,
+      "learning_rate": 0.00036894197952218435,
+      "loss": 5.8389,
+      "step": 5547
+    },
+    {
+      "epoch": 1.8935153583617748,
+      "grad_norm": 3.004615545272827,
+      "learning_rate": 0.0003688282138794084,
+      "loss": 6.1601,
+      "step": 5548
+    },
+    {
+      "epoch": 1.8938566552901024,
+      "grad_norm": 3.0711987018585205,
+      "learning_rate": 0.0003687144482366325,
+      "loss": 6.316,
+      "step": 5549
+    },
+    {
+      "epoch": 1.89419795221843,
+      "grad_norm": 3.4832606315612793,
+      "learning_rate": 0.00036860068259385667,
+      "loss": 5.1936,
+      "step": 5550
+    },
+    {
+      "epoch": 1.8945392491467576,
+      "grad_norm": 3.176182508468628,
+      "learning_rate": 0.0003684869169510808,
+      "loss": 5.5715,
+      "step": 5551
+    },
+    {
+      "epoch": 1.8948805460750853,
+      "grad_norm": 2.930396318435669,
+      "learning_rate": 0.0003683731513083049,
+      "loss": 6.3699,
+      "step": 5552
+    },
+    {
+      "epoch": 1.8952218430034131,
+      "grad_norm": 2.8957557678222656,
+      "learning_rate": 0.00036825938566552904,
+      "loss": 5.121,
+      "step": 5553
+    },
+    {
+      "epoch": 1.8955631399317405,
+      "grad_norm": 2.885211229324341,
+      "learning_rate": 0.00036814562002275314,
+      "loss": 6.0966,
+      "step": 5554
+    },
+    {
+      "epoch": 1.8959044368600684,
+      "grad_norm": 3.3072848320007324,
+      "learning_rate": 0.00036803185437997725,
+      "loss": 5.6025,
+      "step": 5555
+    },
+    {
+      "epoch": 1.8962457337883958,
+      "grad_norm": 3.156798839569092,
+      "learning_rate": 0.0003679180887372014,
+      "loss": 6.0641,
+      "step": 5556
+    },
+    {
+      "epoch": 1.8965870307167236,
+      "grad_norm": 11.321859359741211,
+      "learning_rate": 0.00036780432309442546,
+      "loss": 5.4411,
+      "step": 5557
+    },
+    {
+      "epoch": 1.896928327645051,
+      "grad_norm": 2.977379322052002,
+      "learning_rate": 0.00036769055745164956,
+      "loss": 6.6542,
+      "step": 5558
+    },
+    {
+      "epoch": 1.8972696245733789,
+      "grad_norm": 3.0141522884368896,
+      "learning_rate": 0.0003675767918088737,
+      "loss": 6.0875,
+      "step": 5559
+    },
+    {
+      "epoch": 1.8976109215017065,
+      "grad_norm": 2.9410688877105713,
+      "learning_rate": 0.00036746302616609783,
+      "loss": 6.4585,
+      "step": 5560
+    },
+    {
+      "epoch": 1.8979522184300341,
+      "grad_norm": 3.0146484375,
+      "learning_rate": 0.000367349260523322,
+      "loss": 6.1579,
+      "step": 5561
+    },
+    {
+      "epoch": 1.8982935153583618,
+      "grad_norm": 2.873905897140503,
+      "learning_rate": 0.0003672354948805461,
+      "loss": 5.932,
+      "step": 5562
+    },
+    {
+      "epoch": 1.8986348122866894,
+      "grad_norm": 2.8637454509735107,
+      "learning_rate": 0.0003671217292377702,
+      "loss": 6.2974,
+      "step": 5563
+    },
+    {
+      "epoch": 1.898976109215017,
+      "grad_norm": 3.0655555725097656,
+      "learning_rate": 0.00036700796359499436,
+      "loss": 5.8897,
+      "step": 5564
+    },
+    {
+      "epoch": 1.8993174061433447,
+      "grad_norm": 2.8714025020599365,
+      "learning_rate": 0.00036689419795221846,
+      "loss": 6.1978,
+      "step": 5565
+    },
+    {
+      "epoch": 1.8996587030716725,
+      "grad_norm": 2.971374034881592,
+      "learning_rate": 0.00036678043230944257,
+      "loss": 5.2991,
+      "step": 5566
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.9192891120910645,
+      "learning_rate": 0.00036666666666666667,
+      "loss": 5.9472,
+      "step": 5567
+    },
+    {
+      "epoch": 1.9003412969283278,
+      "grad_norm": 3.0181756019592285,
+      "learning_rate": 0.0003665529010238908,
+      "loss": 6.4597,
+      "step": 5568
+    },
+    {
+      "epoch": 1.9006825938566552,
+      "grad_norm": 2.9252567291259766,
+      "learning_rate": 0.0003664391353811149,
+      "loss": 6.3444,
+      "step": 5569
+    },
+    {
+      "epoch": 1.901023890784983,
+      "grad_norm": 3.165065050125122,
+      "learning_rate": 0.00036632536973833904,
+      "loss": 5.9354,
+      "step": 5570
+    },
+    {
+      "epoch": 1.9013651877133104,
+      "grad_norm": 2.935777425765991,
+      "learning_rate": 0.00036621160409556314,
+      "loss": 6.4579,
+      "step": 5571
+    },
+    {
+      "epoch": 1.9017064846416383,
+      "grad_norm": 2.785609722137451,
+      "learning_rate": 0.00036609783845278725,
+      "loss": 5.7246,
+      "step": 5572
+    },
+    {
+      "epoch": 1.902047781569966,
+      "grad_norm": 2.8585057258605957,
+      "learning_rate": 0.0003659840728100114,
+      "loss": 6.8029,
+      "step": 5573
+    },
+    {
+      "epoch": 1.9023890784982935,
+      "grad_norm": 2.8563308715820312,
+      "learning_rate": 0.0003658703071672355,
+      "loss": 6.3342,
+      "step": 5574
+    },
+    {
+      "epoch": 1.9027303754266212,
+      "grad_norm": 9.50164794921875,
+      "learning_rate": 0.0003657565415244596,
+      "loss": 5.5615,
+      "step": 5575
+    },
+    {
+      "epoch": 1.9030716723549488,
+      "grad_norm": 2.983273983001709,
+      "learning_rate": 0.0003656427758816838,
+      "loss": 6.0595,
+      "step": 5576
+    },
+    {
+      "epoch": 1.9034129692832764,
+      "grad_norm": 2.895350217819214,
+      "learning_rate": 0.00036552901023890783,
+      "loss": 6.2199,
+      "step": 5577
+    },
+    {
+      "epoch": 1.903754266211604,
+      "grad_norm": 3.4141218662261963,
+      "learning_rate": 0.00036541524459613193,
+      "loss": 6.1079,
+      "step": 5578
+    },
+    {
+      "epoch": 1.904095563139932,
+      "grad_norm": 3.1629059314727783,
+      "learning_rate": 0.0003653014789533561,
+      "loss": 5.895,
+      "step": 5579
+    },
+    {
+      "epoch": 1.9044368600682593,
+      "grad_norm": 2.89243483543396,
+      "learning_rate": 0.0003651877133105802,
+      "loss": 5.857,
+      "step": 5580
+    },
+    {
+      "epoch": 1.9047781569965871,
+      "grad_norm": 2.929072618484497,
+      "learning_rate": 0.0003650739476678043,
+      "loss": 6.2954,
+      "step": 5581
+    },
+    {
+      "epoch": 1.9051194539249146,
+      "grad_norm": 3.005039691925049,
+      "learning_rate": 0.00036496018202502846,
+      "loss": 6.2573,
+      "step": 5582
+    },
+    {
+      "epoch": 1.9054607508532424,
+      "grad_norm": 3.169196367263794,
+      "learning_rate": 0.00036484641638225257,
+      "loss": 5.6108,
+      "step": 5583
+    },
+    {
+      "epoch": 1.9058020477815698,
+      "grad_norm": 2.895752429962158,
+      "learning_rate": 0.0003647326507394767,
+      "loss": 6.3867,
+      "step": 5584
+    },
+    {
+      "epoch": 1.9061433447098977,
+      "grad_norm": 3.001361131668091,
+      "learning_rate": 0.00036461888509670083,
+      "loss": 6.3686,
+      "step": 5585
+    },
+    {
+      "epoch": 1.9064846416382253,
+      "grad_norm": 2.938120126724243,
+      "learning_rate": 0.00036450511945392494,
+      "loss": 6.4083,
+      "step": 5586
+    },
+    {
+      "epoch": 1.906825938566553,
+      "grad_norm": 2.791254758834839,
+      "learning_rate": 0.00036439135381114904,
+      "loss": 6.2613,
+      "step": 5587
+    },
+    {
+      "epoch": 1.9071672354948805,
+      "grad_norm": 3.121710777282715,
+      "learning_rate": 0.00036427758816837315,
+      "loss": 5.9979,
+      "step": 5588
+    },
+    {
+      "epoch": 1.9075085324232082,
+      "grad_norm": 3.0421764850616455,
+      "learning_rate": 0.00036416382252559725,
+      "loss": 5.7314,
+      "step": 5589
+    },
+    {
+      "epoch": 1.9078498293515358,
+      "grad_norm": 3.0027949810028076,
+      "learning_rate": 0.0003640500568828214,
+      "loss": 6.3889,
+      "step": 5590
+    },
+    {
+      "epoch": 1.9081911262798634,
+      "grad_norm": 2.965477466583252,
+      "learning_rate": 0.0003639362912400455,
+      "loss": 6.4909,
+      "step": 5591
+    },
+    {
+      "epoch": 1.9085324232081913,
+      "grad_norm": 3.089905023574829,
+      "learning_rate": 0.0003638225255972696,
+      "loss": 6.4991,
+      "step": 5592
+    },
+    {
+      "epoch": 1.9088737201365187,
+      "grad_norm": 2.832148551940918,
+      "learning_rate": 0.0003637087599544938,
+      "loss": 6.2179,
+      "step": 5593
+    },
+    {
+      "epoch": 1.9092150170648465,
+      "grad_norm": 2.9781718254089355,
+      "learning_rate": 0.0003635949943117179,
+      "loss": 6.366,
+      "step": 5594
+    },
+    {
+      "epoch": 1.909556313993174,
+      "grad_norm": 2.9966466426849365,
+      "learning_rate": 0.000363481228668942,
+      "loss": 6.1874,
+      "step": 5595
+    },
+    {
+      "epoch": 1.9098976109215018,
+      "grad_norm": 2.9594128131866455,
+      "learning_rate": 0.0003633674630261661,
+      "loss": 6.2093,
+      "step": 5596
+    },
+    {
+      "epoch": 1.9102389078498292,
+      "grad_norm": 2.8735949993133545,
+      "learning_rate": 0.0003632536973833902,
+      "loss": 6.6883,
+      "step": 5597
+    },
+    {
+      "epoch": 1.910580204778157,
+      "grad_norm": 3.143655300140381,
+      "learning_rate": 0.0003631399317406143,
+      "loss": 5.5111,
+      "step": 5598
+    },
+    {
+      "epoch": 1.9109215017064847,
+      "grad_norm": 2.9463629722595215,
+      "learning_rate": 0.00036302616609783846,
+      "loss": 6.3481,
+      "step": 5599
+    },
+    {
+      "epoch": 1.9112627986348123,
+      "grad_norm": 3.0451953411102295,
+      "learning_rate": 0.00036291240045506257,
+      "loss": 6.3788,
+      "step": 5600
+    },
+    {
+      "epoch": 1.91160409556314,
+      "grad_norm": 2.9045491218566895,
+      "learning_rate": 0.00036279863481228667,
+      "loss": 6.6246,
+      "step": 5601
+    },
+    {
+      "epoch": 1.9119453924914676,
+      "grad_norm": 2.970602035522461,
+      "learning_rate": 0.00036268486916951083,
+      "loss": 6.1013,
+      "step": 5602
+    },
+    {
+      "epoch": 1.9122866894197952,
+      "grad_norm": 2.88826847076416,
+      "learning_rate": 0.00036257110352673494,
+      "loss": 6.2451,
+      "step": 5603
+    },
+    {
+      "epoch": 1.9126279863481228,
+      "grad_norm": 2.8405919075012207,
+      "learning_rate": 0.0003624573378839591,
+      "loss": 6.4149,
+      "step": 5604
+    },
+    {
+      "epoch": 1.9129692832764507,
+      "grad_norm": 2.8814759254455566,
+      "learning_rate": 0.0003623435722411832,
+      "loss": 5.613,
+      "step": 5605
+    },
+    {
+      "epoch": 1.913310580204778,
+      "grad_norm": 2.8033180236816406,
+      "learning_rate": 0.00036222980659840725,
+      "loss": 5.9609,
+      "step": 5606
+    },
+    {
+      "epoch": 1.913651877133106,
+      "grad_norm": 2.9600167274475098,
+      "learning_rate": 0.0003621160409556314,
+      "loss": 6.6377,
+      "step": 5607
+    },
+    {
+      "epoch": 1.9139931740614333,
+      "grad_norm": 2.917205572128296,
+      "learning_rate": 0.0003620022753128555,
+      "loss": 5.7594,
+      "step": 5608
+    },
+    {
+      "epoch": 1.9143344709897612,
+      "grad_norm": 2.8245956897735596,
+      "learning_rate": 0.0003618885096700796,
+      "loss": 6.1144,
+      "step": 5609
+    },
+    {
+      "epoch": 1.9146757679180886,
+      "grad_norm": 6.755703449249268,
+      "learning_rate": 0.0003617747440273038,
+      "loss": 5.8463,
+      "step": 5610
+    },
+    {
+      "epoch": 1.9150170648464164,
+      "grad_norm": 2.968301773071289,
+      "learning_rate": 0.0003616609783845279,
+      "loss": 6.0244,
+      "step": 5611
+    },
+    {
+      "epoch": 1.915358361774744,
+      "grad_norm": 3.3348543643951416,
+      "learning_rate": 0.000361547212741752,
+      "loss": 4.6684,
+      "step": 5612
+    },
+    {
+      "epoch": 1.9156996587030717,
+      "grad_norm": 3.070195436477661,
+      "learning_rate": 0.00036143344709897615,
+      "loss": 5.955,
+      "step": 5613
+    },
+    {
+      "epoch": 1.9160409556313993,
+      "grad_norm": 3.008603572845459,
+      "learning_rate": 0.00036131968145620025,
+      "loss": 6.47,
+      "step": 5614
+    },
+    {
+      "epoch": 1.916382252559727,
+      "grad_norm": 3.069282293319702,
+      "learning_rate": 0.00036120591581342436,
+      "loss": 5.9866,
+      "step": 5615
+    },
+    {
+      "epoch": 1.9167235494880546,
+      "grad_norm": 2.977428913116455,
+      "learning_rate": 0.00036109215017064846,
+      "loss": 6.5193,
+      "step": 5616
+    },
+    {
+      "epoch": 1.9170648464163822,
+      "grad_norm": 2.9662117958068848,
+      "learning_rate": 0.00036097838452787257,
+      "loss": 5.8879,
+      "step": 5617
+    },
+    {
+      "epoch": 1.91740614334471,
+      "grad_norm": 3.1188912391662598,
+      "learning_rate": 0.0003608646188850967,
+      "loss": 5.7213,
+      "step": 5618
+    },
+    {
+      "epoch": 1.9177474402730375,
+      "grad_norm": 2.914436101913452,
+      "learning_rate": 0.00036075085324232083,
+      "loss": 6.7126,
+      "step": 5619
+    },
+    {
+      "epoch": 1.9180887372013653,
+      "grad_norm": 2.956573247909546,
+      "learning_rate": 0.00036063708759954494,
+      "loss": 6.3765,
+      "step": 5620
+    },
+    {
+      "epoch": 1.9184300341296927,
+      "grad_norm": 2.9372453689575195,
+      "learning_rate": 0.00036052332195676904,
+      "loss": 6.046,
+      "step": 5621
+    },
+    {
+      "epoch": 1.9187713310580206,
+      "grad_norm": 2.8803412914276123,
+      "learning_rate": 0.0003604095563139932,
+      "loss": 7.05,
+      "step": 5622
+    },
+    {
+      "epoch": 1.919112627986348,
+      "grad_norm": 3.022796392440796,
+      "learning_rate": 0.0003602957906712173,
+      "loss": 6.3435,
+      "step": 5623
+    },
+    {
+      "epoch": 1.9194539249146758,
+      "grad_norm": 2.9055187702178955,
+      "learning_rate": 0.00036018202502844147,
+      "loss": 6.021,
+      "step": 5624
+    },
+    {
+      "epoch": 1.9197952218430034,
+      "grad_norm": 3.274238348007202,
+      "learning_rate": 0.00036006825938566557,
+      "loss": 5.819,
+      "step": 5625
+    },
+    {
+      "epoch": 1.920136518771331,
+      "grad_norm": 3.3062515258789062,
+      "learning_rate": 0.0003599544937428896,
+      "loss": 6.3418,
+      "step": 5626
+    },
+    {
+      "epoch": 1.9204778156996587,
+      "grad_norm": 3.065157890319824,
+      "learning_rate": 0.0003598407281001138,
+      "loss": 6.9273,
+      "step": 5627
+    },
+    {
+      "epoch": 1.9208191126279863,
+      "grad_norm": 3.0245931148529053,
+      "learning_rate": 0.0003597269624573379,
+      "loss": 6.6343,
+      "step": 5628
+    },
+    {
+      "epoch": 1.921160409556314,
+      "grad_norm": 3.0174522399902344,
+      "learning_rate": 0.000359613196814562,
+      "loss": 5.9226,
+      "step": 5629
+    },
+    {
+      "epoch": 1.9215017064846416,
+      "grad_norm": 2.8802099227905273,
+      "learning_rate": 0.00035949943117178615,
+      "loss": 6.4793,
+      "step": 5630
+    },
+    {
+      "epoch": 1.9218430034129694,
+      "grad_norm": 3.4380247592926025,
+      "learning_rate": 0.00035938566552901025,
+      "loss": 5.8155,
+      "step": 5631
+    },
+    {
+      "epoch": 1.9221843003412968,
+      "grad_norm": 2.8854217529296875,
+      "learning_rate": 0.00035927189988623436,
+      "loss": 6.4598,
+      "step": 5632
+    },
+    {
+      "epoch": 1.9225255972696247,
+      "grad_norm": 3.8663971424102783,
+      "learning_rate": 0.0003591581342434585,
+      "loss": 4.1946,
+      "step": 5633
+    },
+    {
+      "epoch": 1.922866894197952,
+      "grad_norm": 2.801706314086914,
+      "learning_rate": 0.0003590443686006826,
+      "loss": 4.3066,
+      "step": 5634
+    },
+    {
+      "epoch": 1.92320819112628,
+      "grad_norm": 3.0731019973754883,
+      "learning_rate": 0.0003589306029579067,
+      "loss": 6.1641,
+      "step": 5635
+    },
+    {
+      "epoch": 1.9235494880546073,
+      "grad_norm": 3.997176170349121,
+      "learning_rate": 0.00035881683731513083,
+      "loss": 5.2667,
+      "step": 5636
+    },
+    {
+      "epoch": 1.9238907849829352,
+      "grad_norm": 3.5250935554504395,
+      "learning_rate": 0.00035870307167235494,
+      "loss": 3.3671,
+      "step": 5637
+    },
+    {
+      "epoch": 1.9242320819112628,
+      "grad_norm": 3.1355345249176025,
+      "learning_rate": 0.00035858930602957904,
+      "loss": 5.3227,
+      "step": 5638
+    },
+    {
+      "epoch": 1.9245733788395905,
+      "grad_norm": 3.095639228820801,
+      "learning_rate": 0.0003584755403868032,
+      "loss": 6.4683,
+      "step": 5639
+    },
+    {
+      "epoch": 1.924914675767918,
+      "grad_norm": 3.36194109916687,
+      "learning_rate": 0.0003583617747440273,
+      "loss": 6.0146,
+      "step": 5640
+    },
+    {
+      "epoch": 1.9252559726962457,
+      "grad_norm": 3.3805904388427734,
+      "learning_rate": 0.0003582480091012514,
+      "loss": 5.8779,
+      "step": 5641
+    },
+    {
+      "epoch": 1.9255972696245733,
+      "grad_norm": 3.0124683380126953,
+      "learning_rate": 0.00035813424345847557,
+      "loss": 6.5908,
+      "step": 5642
+    },
+    {
+      "epoch": 1.925938566552901,
+      "grad_norm": 3.2803475856781006,
+      "learning_rate": 0.0003580204778156997,
+      "loss": 5.7909,
+      "step": 5643
+    },
+    {
+      "epoch": 1.9262798634812288,
+      "grad_norm": 4.804317951202393,
+      "learning_rate": 0.00035790671217292384,
+      "loss": 5.5523,
+      "step": 5644
+    },
+    {
+      "epoch": 1.9266211604095562,
+      "grad_norm": 2.983696460723877,
+      "learning_rate": 0.0003577929465301479,
+      "loss": 6.0915,
+      "step": 5645
+    },
+    {
+      "epoch": 1.926962457337884,
+      "grad_norm": 2.929771900177002,
+      "learning_rate": 0.000357679180887372,
+      "loss": 5.4666,
+      "step": 5646
+    },
+    {
+      "epoch": 1.9273037542662115,
+      "grad_norm": 4.5054731369018555,
+      "learning_rate": 0.00035756541524459615,
+      "loss": 5.3908,
+      "step": 5647
+    },
+    {
+      "epoch": 1.9276450511945393,
+      "grad_norm": 2.8603098392486572,
+      "learning_rate": 0.00035745164960182026,
+      "loss": 6.3809,
+      "step": 5648
+    },
+    {
+      "epoch": 1.9279863481228667,
+      "grad_norm": 2.934943675994873,
+      "learning_rate": 0.00035733788395904436,
+      "loss": 5.8931,
+      "step": 5649
+    },
+    {
+      "epoch": 1.9283276450511946,
+      "grad_norm": 3.016292095184326,
+      "learning_rate": 0.0003572241183162685,
+      "loss": 5.6018,
+      "step": 5650
+    },
+    {
+      "epoch": 1.9286689419795222,
+      "grad_norm": 3.0841245651245117,
+      "learning_rate": 0.0003571103526734926,
+      "loss": 5.5162,
+      "step": 5651
+    },
+    {
+      "epoch": 1.9290102389078498,
+      "grad_norm": 2.870736837387085,
+      "learning_rate": 0.00035699658703071673,
+      "loss": 6.1772,
+      "step": 5652
+    },
+    {
+      "epoch": 1.9293515358361775,
+      "grad_norm": 6.509230136871338,
+      "learning_rate": 0.0003568828213879409,
+      "loss": 5.6765,
+      "step": 5653
+    },
+    {
+      "epoch": 1.929692832764505,
+      "grad_norm": 2.972872257232666,
+      "learning_rate": 0.000356769055745165,
+      "loss": 6.6489,
+      "step": 5654
+    },
+    {
+      "epoch": 1.9300341296928327,
+      "grad_norm": 3.467580556869507,
+      "learning_rate": 0.00035665529010238904,
+      "loss": 5.2504,
+      "step": 5655
+    },
+    {
+      "epoch": 1.9303754266211604,
+      "grad_norm": 3.250774621963501,
+      "learning_rate": 0.0003565415244596132,
+      "loss": 5.4942,
+      "step": 5656
+    },
+    {
+      "epoch": 1.9307167235494882,
+      "grad_norm": 2.9641740322113037,
+      "learning_rate": 0.0003564277588168373,
+      "loss": 6.0985,
+      "step": 5657
+    },
+    {
+      "epoch": 1.9310580204778156,
+      "grad_norm": 2.939337730407715,
+      "learning_rate": 0.0003563139931740614,
+      "loss": 5.0818,
+      "step": 5658
+    },
+    {
+      "epoch": 1.9313993174061435,
+      "grad_norm": 3.126194715499878,
+      "learning_rate": 0.00035620022753128557,
+      "loss": 6.1819,
+      "step": 5659
+    },
+    {
+      "epoch": 1.9317406143344709,
+      "grad_norm": 5.255848407745361,
+      "learning_rate": 0.0003560864618885097,
+      "loss": 5.0863,
+      "step": 5660
+    },
+    {
+      "epoch": 1.9320819112627987,
+      "grad_norm": 3.117389678955078,
+      "learning_rate": 0.0003559726962457338,
+      "loss": 6.127,
+      "step": 5661
+    },
+    {
+      "epoch": 1.9324232081911261,
+      "grad_norm": 2.9117963314056396,
+      "learning_rate": 0.00035585893060295794,
+      "loss": 6.263,
+      "step": 5662
+    },
+    {
+      "epoch": 1.932764505119454,
+      "grad_norm": 3.053349733352661,
+      "learning_rate": 0.00035574516496018205,
+      "loss": 6.0983,
+      "step": 5663
+    },
+    {
+      "epoch": 1.9331058020477816,
+      "grad_norm": 3.072718381881714,
+      "learning_rate": 0.0003556313993174061,
+      "loss": 5.3886,
+      "step": 5664
+    },
+    {
+      "epoch": 1.9334470989761092,
+      "grad_norm": 10.232709884643555,
+      "learning_rate": 0.00035551763367463026,
+      "loss": 6.7737,
+      "step": 5665
+    },
+    {
+      "epoch": 1.9337883959044369,
+      "grad_norm": 2.86478328704834,
+      "learning_rate": 0.00035540386803185436,
+      "loss": 5.8319,
+      "step": 5666
+    },
+    {
+      "epoch": 1.9341296928327645,
+      "grad_norm": 2.977330446243286,
+      "learning_rate": 0.0003552901023890785,
+      "loss": 6.3019,
+      "step": 5667
+    },
+    {
+      "epoch": 1.934470989761092,
+      "grad_norm": 3.6822073459625244,
+      "learning_rate": 0.0003551763367463026,
+      "loss": 3.6634,
+      "step": 5668
+    },
+    {
+      "epoch": 1.9348122866894197,
+      "grad_norm": 2.952073097229004,
+      "learning_rate": 0.00035506257110352673,
+      "loss": 6.0564,
+      "step": 5669
+    },
+    {
+      "epoch": 1.9351535836177476,
+      "grad_norm": 3.094076156616211,
+      "learning_rate": 0.0003549488054607509,
+      "loss": 6.3278,
+      "step": 5670
+    },
+    {
+      "epoch": 1.935494880546075,
+      "grad_norm": 3.062957763671875,
+      "learning_rate": 0.000354835039817975,
+      "loss": 6.9028,
+      "step": 5671
+    },
+    {
+      "epoch": 1.9358361774744028,
+      "grad_norm": 3.019155502319336,
+      "learning_rate": 0.0003547212741751991,
+      "loss": 5.549,
+      "step": 5672
+    },
+    {
+      "epoch": 1.9361774744027302,
+      "grad_norm": 2.855597734451294,
+      "learning_rate": 0.00035460750853242326,
+      "loss": 6.4601,
+      "step": 5673
+    },
+    {
+      "epoch": 1.936518771331058,
+      "grad_norm": 2.83564829826355,
+      "learning_rate": 0.0003544937428896473,
+      "loss": 6.4995,
+      "step": 5674
+    },
+    {
+      "epoch": 1.9368600682593855,
+      "grad_norm": 2.9659066200256348,
+      "learning_rate": 0.0003543799772468714,
+      "loss": 6.5664,
+      "step": 5675
+    },
+    {
+      "epoch": 1.9372013651877134,
+      "grad_norm": 2.9743475914001465,
+      "learning_rate": 0.0003542662116040956,
+      "loss": 7.1904,
+      "step": 5676
+    },
+    {
+      "epoch": 1.937542662116041,
+      "grad_norm": 8.094204902648926,
+      "learning_rate": 0.0003541524459613197,
+      "loss": 5.402,
+      "step": 5677
+    },
+    {
+      "epoch": 1.9378839590443686,
+      "grad_norm": 3.036323070526123,
+      "learning_rate": 0.0003540386803185438,
+      "loss": 6.4258,
+      "step": 5678
+    },
+    {
+      "epoch": 1.9382252559726962,
+      "grad_norm": 3.9837539196014404,
+      "learning_rate": 0.00035392491467576794,
+      "loss": 4.7752,
+      "step": 5679
+    },
+    {
+      "epoch": 1.9385665529010239,
+      "grad_norm": 3.116345167160034,
+      "learning_rate": 0.00035381114903299205,
+      "loss": 6.3365,
+      "step": 5680
+    },
+    {
+      "epoch": 1.9389078498293515,
+      "grad_norm": 2.885256290435791,
+      "learning_rate": 0.00035369738339021615,
+      "loss": 6.3768,
+      "step": 5681
+    },
+    {
+      "epoch": 1.9392491467576791,
+      "grad_norm": 2.9523653984069824,
+      "learning_rate": 0.0003535836177474403,
+      "loss": 6.459,
+      "step": 5682
+    },
+    {
+      "epoch": 1.939590443686007,
+      "grad_norm": 3.4799153804779053,
+      "learning_rate": 0.0003534698521046644,
+      "loss": 5.524,
+      "step": 5683
+    },
+    {
+      "epoch": 1.9399317406143344,
+      "grad_norm": 2.8675577640533447,
+      "learning_rate": 0.00035335608646188847,
+      "loss": 6.0246,
+      "step": 5684
+    },
+    {
+      "epoch": 1.9402730375426622,
+      "grad_norm": 2.8474411964416504,
+      "learning_rate": 0.0003532423208191126,
+      "loss": 6.6248,
+      "step": 5685
+    },
+    {
+      "epoch": 1.9406143344709896,
+      "grad_norm": 2.8228769302368164,
+      "learning_rate": 0.00035312855517633673,
+      "loss": 5.826,
+      "step": 5686
+    },
+    {
+      "epoch": 1.9409556313993175,
+      "grad_norm": 2.9075095653533936,
+      "learning_rate": 0.0003530147895335609,
+      "loss": 6.5578,
+      "step": 5687
+    },
+    {
+      "epoch": 1.9412969283276449,
+      "grad_norm": 2.9296059608459473,
+      "learning_rate": 0.000352901023890785,
+      "loss": 6.1996,
+      "step": 5688
+    },
+    {
+      "epoch": 1.9416382252559727,
+      "grad_norm": 3.073141098022461,
+      "learning_rate": 0.0003527872582480091,
+      "loss": 5.9504,
+      "step": 5689
+    },
+    {
+      "epoch": 1.9419795221843004,
+      "grad_norm": 2.9059715270996094,
+      "learning_rate": 0.00035267349260523326,
+      "loss": 6.5117,
+      "step": 5690
+    },
+    {
+      "epoch": 1.942320819112628,
+      "grad_norm": 2.9183788299560547,
+      "learning_rate": 0.00035255972696245736,
+      "loss": 6.7984,
+      "step": 5691
+    },
+    {
+      "epoch": 1.9426621160409556,
+      "grad_norm": 4.353701591491699,
+      "learning_rate": 0.00035244596131968147,
+      "loss": 4.3046,
+      "step": 5692
+    },
+    {
+      "epoch": 1.9430034129692833,
+      "grad_norm": 2.997781753540039,
+      "learning_rate": 0.00035233219567690563,
+      "loss": 6.7823,
+      "step": 5693
+    },
+    {
+      "epoch": 1.9433447098976109,
+      "grad_norm": 3.029731512069702,
+      "learning_rate": 0.0003522184300341297,
+      "loss": 6.3343,
+      "step": 5694
+    },
+    {
+      "epoch": 1.9436860068259385,
+      "grad_norm": 3.035346746444702,
+      "learning_rate": 0.0003521046643913538,
+      "loss": 6.6839,
+      "step": 5695
+    },
+    {
+      "epoch": 1.9440273037542664,
+      "grad_norm": 2.9383420944213867,
+      "learning_rate": 0.00035199089874857794,
+      "loss": 5.8788,
+      "step": 5696
+    },
+    {
+      "epoch": 1.9443686006825938,
+      "grad_norm": 2.8888626098632812,
+      "learning_rate": 0.00035187713310580205,
+      "loss": 6.6803,
+      "step": 5697
+    },
+    {
+      "epoch": 1.9447098976109216,
+      "grad_norm": 2.8905954360961914,
+      "learning_rate": 0.00035176336746302615,
+      "loss": 6.7251,
+      "step": 5698
+    },
+    {
+      "epoch": 1.945051194539249,
+      "grad_norm": 2.848459482192993,
+      "learning_rate": 0.0003516496018202503,
+      "loss": 6.3719,
+      "step": 5699
+    },
+    {
+      "epoch": 1.9453924914675769,
+      "grad_norm": 6.845625877380371,
+      "learning_rate": 0.0003515358361774744,
+      "loss": 5.3028,
+      "step": 5700
+    },
+    {
+      "epoch": 1.9457337883959043,
+      "grad_norm": 2.9232285022735596,
+      "learning_rate": 0.0003514220705346985,
+      "loss": 6.7182,
+      "step": 5701
+    },
+    {
+      "epoch": 1.9460750853242321,
+      "grad_norm": 2.9307868480682373,
+      "learning_rate": 0.0003513083048919227,
+      "loss": 5.891,
+      "step": 5702
+    },
+    {
+      "epoch": 1.9464163822525598,
+      "grad_norm": 4.067267894744873,
+      "learning_rate": 0.00035119453924914673,
+      "loss": 5.0473,
+      "step": 5703
+    },
+    {
+      "epoch": 1.9467576791808874,
+      "grad_norm": 3.438600778579712,
+      "learning_rate": 0.00035108077360637084,
+      "loss": 4.1299,
+      "step": 5704
+    },
+    {
+      "epoch": 1.947098976109215,
+      "grad_norm": 2.9375200271606445,
+      "learning_rate": 0.000350967007963595,
+      "loss": 6.3161,
+      "step": 5705
+    },
+    {
+      "epoch": 1.9474402730375426,
+      "grad_norm": 3.0774381160736084,
+      "learning_rate": 0.0003508532423208191,
+      "loss": 6.826,
+      "step": 5706
+    },
+    {
+      "epoch": 1.9477815699658703,
+      "grad_norm": 2.943984270095825,
+      "learning_rate": 0.00035073947667804326,
+      "loss": 6.4722,
+      "step": 5707
+    },
+    {
+      "epoch": 1.948122866894198,
+      "grad_norm": 2.850332021713257,
+      "learning_rate": 0.00035062571103526737,
+      "loss": 6.3759,
+      "step": 5708
+    },
+    {
+      "epoch": 1.9484641638225257,
+      "grad_norm": 2.926992177963257,
+      "learning_rate": 0.00035051194539249147,
+      "loss": 5.968,
+      "step": 5709
+    },
+    {
+      "epoch": 1.9488054607508531,
+      "grad_norm": 2.8328535556793213,
+      "learning_rate": 0.00035039817974971563,
+      "loss": 6.1558,
+      "step": 5710
+    },
+    {
+      "epoch": 1.949146757679181,
+      "grad_norm": 2.906438112258911,
+      "learning_rate": 0.00035028441410693973,
+      "loss": 6.7331,
+      "step": 5711
+    },
+    {
+      "epoch": 1.9494880546075084,
+      "grad_norm": 2.8392958641052246,
+      "learning_rate": 0.00035017064846416384,
+      "loss": 6.4798,
+      "step": 5712
+    },
+    {
+      "epoch": 1.9498293515358363,
+      "grad_norm": 3.147331476211548,
+      "learning_rate": 0.00035005688282138794,
+      "loss": 5.3819,
+      "step": 5713
+    },
+    {
+      "epoch": 1.9501706484641637,
+      "grad_norm": 2.8738808631896973,
+      "learning_rate": 0.00034994311717861205,
+      "loss": 6.0735,
+      "step": 5714
+    },
+    {
+      "epoch": 1.9505119453924915,
+      "grad_norm": 2.8813462257385254,
+      "learning_rate": 0.00034982935153583615,
+      "loss": 6.228,
+      "step": 5715
+    },
+    {
+      "epoch": 1.9508532423208191,
+      "grad_norm": 2.9783530235290527,
+      "learning_rate": 0.0003497155858930603,
+      "loss": 6.3557,
+      "step": 5716
+    },
+    {
+      "epoch": 1.9511945392491468,
+      "grad_norm": 2.8971831798553467,
+      "learning_rate": 0.0003496018202502844,
+      "loss": 6.5061,
+      "step": 5717
+    },
+    {
+      "epoch": 1.9515358361774744,
+      "grad_norm": 7.339874267578125,
+      "learning_rate": 0.0003494880546075085,
+      "loss": 5.3234,
+      "step": 5718
+    },
+    {
+      "epoch": 1.951877133105802,
+      "grad_norm": 2.992372751235962,
+      "learning_rate": 0.0003493742889647327,
+      "loss": 5.4487,
+      "step": 5719
+    },
+    {
+      "epoch": 1.9522184300341296,
+      "grad_norm": 2.892254590988159,
+      "learning_rate": 0.0003492605233219568,
+      "loss": 6.1741,
+      "step": 5720
+    },
+    {
+      "epoch": 1.9525597269624573,
+      "grad_norm": 2.97224497795105,
+      "learning_rate": 0.0003491467576791809,
+      "loss": 6.3184,
+      "step": 5721
+    },
+    {
+      "epoch": 1.9529010238907851,
+      "grad_norm": 2.979079008102417,
+      "learning_rate": 0.00034903299203640505,
+      "loss": 6.1467,
+      "step": 5722
+    },
+    {
+      "epoch": 1.9532423208191125,
+      "grad_norm": 3.279611825942993,
+      "learning_rate": 0.0003489192263936291,
+      "loss": 5.636,
+      "step": 5723
+    },
+    {
+      "epoch": 1.9535836177474404,
+      "grad_norm": 2.8635432720184326,
+      "learning_rate": 0.0003488054607508532,
+      "loss": 6.1327,
+      "step": 5724
+    },
+    {
+      "epoch": 1.9539249146757678,
+      "grad_norm": 2.876227855682373,
+      "learning_rate": 0.00034869169510807737,
+      "loss": 5.9195,
+      "step": 5725
+    },
+    {
+      "epoch": 1.9542662116040956,
+      "grad_norm": 2.923100233078003,
+      "learning_rate": 0.00034857792946530147,
+      "loss": 5.9977,
+      "step": 5726
+    },
+    {
+      "epoch": 1.954607508532423,
+      "grad_norm": 2.884539842605591,
+      "learning_rate": 0.00034846416382252563,
+      "loss": 6.6596,
+      "step": 5727
+    },
+    {
+      "epoch": 1.954948805460751,
+      "grad_norm": 2.869551420211792,
+      "learning_rate": 0.00034835039817974973,
+      "loss": 6.7263,
+      "step": 5728
+    },
+    {
+      "epoch": 1.9552901023890785,
+      "grad_norm": 3.0249316692352295,
+      "learning_rate": 0.00034823663253697384,
+      "loss": 5.9006,
+      "step": 5729
+    },
+    {
+      "epoch": 1.9556313993174061,
+      "grad_norm": 3.008315086364746,
+      "learning_rate": 0.000348122866894198,
+      "loss": 6.1326,
+      "step": 5730
+    },
+    {
+      "epoch": 1.9559726962457338,
+      "grad_norm": 2.924003839492798,
+      "learning_rate": 0.0003480091012514221,
+      "loss": 5.8715,
+      "step": 5731
+    },
+    {
+      "epoch": 1.9563139931740614,
+      "grad_norm": 2.9903616905212402,
+      "learning_rate": 0.00034789533560864615,
+      "loss": 6.8732,
+      "step": 5732
+    },
+    {
+      "epoch": 1.956655290102389,
+      "grad_norm": 3.197800636291504,
+      "learning_rate": 0.0003477815699658703,
+      "loss": 5.9274,
+      "step": 5733
+    },
+    {
+      "epoch": 1.9569965870307167,
+      "grad_norm": 2.8411238193511963,
+      "learning_rate": 0.0003476678043230944,
+      "loss": 6.1661,
+      "step": 5734
+    },
+    {
+      "epoch": 1.9573378839590445,
+      "grad_norm": 2.9080448150634766,
+      "learning_rate": 0.0003475540386803185,
+      "loss": 6.2348,
+      "step": 5735
+    },
+    {
+      "epoch": 1.957679180887372,
+      "grad_norm": 2.778029441833496,
+      "learning_rate": 0.0003474402730375427,
+      "loss": 6.2891,
+      "step": 5736
+    },
+    {
+      "epoch": 1.9580204778156998,
+      "grad_norm": 2.954080581665039,
+      "learning_rate": 0.0003473265073947668,
+      "loss": 6.1235,
+      "step": 5737
+    },
+    {
+      "epoch": 1.9583617747440272,
+      "grad_norm": 2.9157159328460693,
+      "learning_rate": 0.0003472127417519909,
+      "loss": 6.2918,
+      "step": 5738
+    },
+    {
+      "epoch": 1.958703071672355,
+      "grad_norm": 2.926003932952881,
+      "learning_rate": 0.00034709897610921505,
+      "loss": 6.6191,
+      "step": 5739
+    },
+    {
+      "epoch": 1.9590443686006824,
+      "grad_norm": 3.026291847229004,
+      "learning_rate": 0.00034698521046643916,
+      "loss": 6.0096,
+      "step": 5740
+    },
+    {
+      "epoch": 1.9593856655290103,
+      "grad_norm": 2.8232579231262207,
+      "learning_rate": 0.00034687144482366326,
+      "loss": 6.4494,
+      "step": 5741
+    },
+    {
+      "epoch": 1.959726962457338,
+      "grad_norm": 3.837078332901001,
+      "learning_rate": 0.00034675767918088737,
+      "loss": 5.45,
+      "step": 5742
+    },
+    {
+      "epoch": 1.9600682593856655,
+      "grad_norm": 2.8543741703033447,
+      "learning_rate": 0.00034664391353811147,
+      "loss": 6.1748,
+      "step": 5743
+    },
+    {
+      "epoch": 1.9604095563139932,
+      "grad_norm": 3.0786311626434326,
+      "learning_rate": 0.0003465301478953356,
+      "loss": 5.7486,
+      "step": 5744
+    },
+    {
+      "epoch": 1.9607508532423208,
+      "grad_norm": 3.161501407623291,
+      "learning_rate": 0.00034641638225255974,
+      "loss": 5.8587,
+      "step": 5745
+    },
+    {
+      "epoch": 1.9610921501706484,
+      "grad_norm": 2.9017693996429443,
+      "learning_rate": 0.00034630261660978384,
+      "loss": 6.2358,
+      "step": 5746
+    },
+    {
+      "epoch": 1.961433447098976,
+      "grad_norm": 2.9906439781188965,
+      "learning_rate": 0.000346188850967008,
+      "loss": 6.5953,
+      "step": 5747
+    },
+    {
+      "epoch": 1.961774744027304,
+      "grad_norm": 2.825509786605835,
+      "learning_rate": 0.0003460750853242321,
+      "loss": 6.4981,
+      "step": 5748
+    },
+    {
+      "epoch": 1.9621160409556313,
+      "grad_norm": 4.668010711669922,
+      "learning_rate": 0.0003459613196814562,
+      "loss": 4.1262,
+      "step": 5749
+    },
+    {
+      "epoch": 1.9624573378839592,
+      "grad_norm": 2.996762990951538,
+      "learning_rate": 0.00034584755403868037,
+      "loss": 5.5406,
+      "step": 5750
+    },
+    {
+      "epoch": 1.9627986348122866,
+      "grad_norm": 3.0203421115875244,
+      "learning_rate": 0.0003457337883959045,
+      "loss": 5.8989,
+      "step": 5751
+    },
+    {
+      "epoch": 1.9631399317406144,
+      "grad_norm": 2.8792803287506104,
+      "learning_rate": 0.0003456200227531285,
+      "loss": 6.4116,
+      "step": 5752
+    },
+    {
+      "epoch": 1.9634812286689418,
+      "grad_norm": 2.9204788208007812,
+      "learning_rate": 0.0003455062571103527,
+      "loss": 5.7728,
+      "step": 5753
+    },
+    {
+      "epoch": 1.9638225255972697,
+      "grad_norm": 2.9870834350585938,
+      "learning_rate": 0.0003453924914675768,
+      "loss": 6.179,
+      "step": 5754
+    },
+    {
+      "epoch": 1.9641638225255973,
+      "grad_norm": 3.1838748455047607,
+      "learning_rate": 0.0003452787258248009,
+      "loss": 5.5057,
+      "step": 5755
+    },
+    {
+      "epoch": 1.964505119453925,
+      "grad_norm": 2.916292667388916,
+      "learning_rate": 0.00034516496018202505,
+      "loss": 6.7139,
+      "step": 5756
+    },
+    {
+      "epoch": 1.9648464163822525,
+      "grad_norm": 2.948239326477051,
+      "learning_rate": 0.00034505119453924916,
+      "loss": 6.5786,
+      "step": 5757
+    },
+    {
+      "epoch": 1.9651877133105802,
+      "grad_norm": 2.867830276489258,
+      "learning_rate": 0.00034493742889647326,
+      "loss": 6.4021,
+      "step": 5758
+    },
+    {
+      "epoch": 1.9655290102389078,
+      "grad_norm": 3.0150563716888428,
+      "learning_rate": 0.0003448236632536974,
+      "loss": 5.8162,
+      "step": 5759
+    },
+    {
+      "epoch": 1.9658703071672354,
+      "grad_norm": 3.1503114700317383,
+      "learning_rate": 0.0003447098976109215,
+      "loss": 5.9391,
+      "step": 5760
+    },
+    {
+      "epoch": 1.9662116040955633,
+      "grad_norm": 2.8918957710266113,
+      "learning_rate": 0.00034459613196814563,
+      "loss": 6.6574,
+      "step": 5761
+    },
+    {
+      "epoch": 1.9665529010238907,
+      "grad_norm": 3.020156145095825,
+      "learning_rate": 0.00034448236632536974,
+      "loss": 5.4734,
+      "step": 5762
+    },
+    {
+      "epoch": 1.9668941979522185,
+      "grad_norm": 3.0351099967956543,
+      "learning_rate": 0.00034436860068259384,
+      "loss": 5.4351,
+      "step": 5763
+    },
+    {
+      "epoch": 1.967235494880546,
+      "grad_norm": 2.8954880237579346,
+      "learning_rate": 0.00034425483503981795,
+      "loss": 6.2349,
+      "step": 5764
+    },
+    {
+      "epoch": 1.9675767918088738,
+      "grad_norm": 2.7660810947418213,
+      "learning_rate": 0.0003441410693970421,
+      "loss": 6.1523,
+      "step": 5765
+    },
+    {
+      "epoch": 1.9679180887372012,
+      "grad_norm": 11.300498962402344,
+      "learning_rate": 0.0003440273037542662,
+      "loss": 3.3175,
+      "step": 5766
+    },
+    {
+      "epoch": 1.968259385665529,
+      "grad_norm": 9.648717880249023,
+      "learning_rate": 0.00034391353811149037,
+      "loss": 4.9295,
+      "step": 5767
+    },
+    {
+      "epoch": 1.9686006825938567,
+      "grad_norm": 3.16329288482666,
+      "learning_rate": 0.0003437997724687145,
+      "loss": 5.9512,
+      "step": 5768
+    },
+    {
+      "epoch": 1.9689419795221843,
+      "grad_norm": 3.4901375770568848,
+      "learning_rate": 0.0003436860068259386,
+      "loss": 5.3588,
+      "step": 5769
+    },
+    {
+      "epoch": 1.969283276450512,
+      "grad_norm": 3.6523478031158447,
+      "learning_rate": 0.00034357224118316274,
+      "loss": 3.9288,
+      "step": 5770
+    },
+    {
+      "epoch": 1.9696245733788396,
+      "grad_norm": 3.0849642753601074,
+      "learning_rate": 0.0003434584755403868,
+      "loss": 6.282,
+      "step": 5771
+    },
+    {
+      "epoch": 1.9699658703071672,
+      "grad_norm": 3.0182032585144043,
+      "learning_rate": 0.0003433447098976109,
+      "loss": 6.6789,
+      "step": 5772
+    },
+    {
+      "epoch": 1.9703071672354948,
+      "grad_norm": 2.9407904148101807,
+      "learning_rate": 0.00034323094425483505,
+      "loss": 6.3049,
+      "step": 5773
+    },
+    {
+      "epoch": 1.9706484641638227,
+      "grad_norm": 2.94191837310791,
+      "learning_rate": 0.00034311717861205916,
+      "loss": 6.1876,
+      "step": 5774
+    },
+    {
+      "epoch": 1.97098976109215,
+      "grad_norm": 2.8799357414245605,
+      "learning_rate": 0.00034300341296928326,
+      "loss": 6.0039,
+      "step": 5775
+    },
+    {
+      "epoch": 1.971331058020478,
+      "grad_norm": 3.216027021408081,
+      "learning_rate": 0.0003428896473265074,
+      "loss": 5.3073,
+      "step": 5776
+    },
+    {
+      "epoch": 1.9716723549488053,
+      "grad_norm": 2.9986681938171387,
+      "learning_rate": 0.00034277588168373153,
+      "loss": 5.4347,
+      "step": 5777
+    },
+    {
+      "epoch": 1.9720136518771332,
+      "grad_norm": 2.878371477127075,
+      "learning_rate": 0.00034266211604095563,
+      "loss": 6.3747,
+      "step": 5778
+    },
+    {
+      "epoch": 1.9723549488054608,
+      "grad_norm": 2.826388359069824,
+      "learning_rate": 0.0003425483503981798,
+      "loss": 6.0383,
+      "step": 5779
+    },
+    {
+      "epoch": 1.9726962457337884,
+      "grad_norm": 2.887346029281616,
+      "learning_rate": 0.0003424345847554039,
+      "loss": 5.9541,
+      "step": 5780
+    },
+    {
+      "epoch": 1.973037542662116,
+      "grad_norm": 3.0003409385681152,
+      "learning_rate": 0.00034232081911262795,
+      "loss": 5.4532,
+      "step": 5781
+    },
+    {
+      "epoch": 1.9733788395904437,
+      "grad_norm": 2.9644722938537598,
+      "learning_rate": 0.0003422070534698521,
+      "loss": 6.1134,
+      "step": 5782
+    },
+    {
+      "epoch": 1.9737201365187713,
+      "grad_norm": 3.0157651901245117,
+      "learning_rate": 0.0003420932878270762,
+      "loss": 6.7906,
+      "step": 5783
+    },
+    {
+      "epoch": 1.974061433447099,
+      "grad_norm": 3.0431642532348633,
+      "learning_rate": 0.0003419795221843003,
+      "loss": 6.5403,
+      "step": 5784
+    },
+    {
+      "epoch": 1.9744027303754266,
+      "grad_norm": 4.550543785095215,
+      "learning_rate": 0.0003418657565415245,
+      "loss": 5.8783,
+      "step": 5785
+    },
+    {
+      "epoch": 1.9747440273037542,
+      "grad_norm": 3.095123052597046,
+      "learning_rate": 0.0003417519908987486,
+      "loss": 5.6745,
+      "step": 5786
+    },
+    {
+      "epoch": 1.975085324232082,
+      "grad_norm": 2.822967290878296,
+      "learning_rate": 0.0003416382252559727,
+      "loss": 6.32,
+      "step": 5787
+    },
+    {
+      "epoch": 1.9754266211604095,
+      "grad_norm": 3.229649066925049,
+      "learning_rate": 0.00034152445961319684,
+      "loss": 6.0626,
+      "step": 5788
+    },
+    {
+      "epoch": 1.9757679180887373,
+      "grad_norm": 2.855510950088501,
+      "learning_rate": 0.00034141069397042095,
+      "loss": 5.828,
+      "step": 5789
+    },
+    {
+      "epoch": 1.9761092150170647,
+      "grad_norm": 2.838651657104492,
+      "learning_rate": 0.0003412969283276451,
+      "loss": 6.1334,
+      "step": 5790
+    },
+    {
+      "epoch": 1.9764505119453926,
+      "grad_norm": 2.9031753540039062,
+      "learning_rate": 0.00034118316268486916,
+      "loss": 6.1161,
+      "step": 5791
+    },
+    {
+      "epoch": 1.9767918088737202,
+      "grad_norm": 3.6981256008148193,
+      "learning_rate": 0.00034106939704209326,
+      "loss": 5.6065,
+      "step": 5792
+    },
+    {
+      "epoch": 1.9771331058020478,
+      "grad_norm": 2.8721656799316406,
+      "learning_rate": 0.0003409556313993174,
+      "loss": 6.2977,
+      "step": 5793
+    },
+    {
+      "epoch": 1.9774744027303754,
+      "grad_norm": 2.9534177780151367,
+      "learning_rate": 0.00034084186575654153,
+      "loss": 5.5651,
+      "step": 5794
+    },
+    {
+      "epoch": 1.977815699658703,
+      "grad_norm": 3.0801563262939453,
+      "learning_rate": 0.00034072810011376563,
+      "loss": 5.7325,
+      "step": 5795
+    },
+    {
+      "epoch": 1.9781569965870307,
+      "grad_norm": 3.880596399307251,
+      "learning_rate": 0.0003406143344709898,
+      "loss": 4.5394,
+      "step": 5796
+    },
+    {
+      "epoch": 1.9784982935153583,
+      "grad_norm": 2.8708319664001465,
+      "learning_rate": 0.0003405005688282139,
+      "loss": 6.3677,
+      "step": 5797
+    },
+    {
+      "epoch": 1.978839590443686,
+      "grad_norm": 2.896517753601074,
+      "learning_rate": 0.000340386803185438,
+      "loss": 6.0279,
+      "step": 5798
+    },
+    {
+      "epoch": 1.9791808873720136,
+      "grad_norm": 2.8943135738372803,
+      "learning_rate": 0.00034027303754266216,
+      "loss": 6.1411,
+      "step": 5799
+    },
+    {
+      "epoch": 1.9795221843003414,
+      "grad_norm": 2.9464333057403564,
+      "learning_rate": 0.00034015927189988627,
+      "loss": 6.6108,
+      "step": 5800
+    },
+    {
+      "epoch": 1.9798634812286688,
+      "grad_norm": 2.76023268699646,
+      "learning_rate": 0.0003400455062571103,
+      "loss": 6.2494,
+      "step": 5801
+    },
+    {
+      "epoch": 1.9802047781569967,
+      "grad_norm": 2.8224380016326904,
+      "learning_rate": 0.0003399317406143345,
+      "loss": 6.3798,
+      "step": 5802
+    },
+    {
+      "epoch": 1.980546075085324,
+      "grad_norm": 6.953861713409424,
+      "learning_rate": 0.0003398179749715586,
+      "loss": 6.2183,
+      "step": 5803
+    },
+    {
+      "epoch": 1.980887372013652,
+      "grad_norm": 3.372380256652832,
+      "learning_rate": 0.0003397042093287827,
+      "loss": 5.2131,
+      "step": 5804
+    },
+    {
+      "epoch": 1.9812286689419796,
+      "grad_norm": 2.964792490005493,
+      "learning_rate": 0.00033959044368600685,
+      "loss": 6.8906,
+      "step": 5805
+    },
+    {
+      "epoch": 1.9815699658703072,
+      "grad_norm": 9.055648803710938,
+      "learning_rate": 0.00033947667804323095,
+      "loss": 7.0398,
+      "step": 5806
+    },
+    {
+      "epoch": 1.9819112627986348,
+      "grad_norm": 3.136854648590088,
+      "learning_rate": 0.00033936291240045506,
+      "loss": 5.7341,
+      "step": 5807
+    },
+    {
+      "epoch": 1.9822525597269625,
+      "grad_norm": 3.117483139038086,
+      "learning_rate": 0.0003392491467576792,
+      "loss": 6.1711,
+      "step": 5808
+    },
+    {
+      "epoch": 1.98259385665529,
+      "grad_norm": 3.032315492630005,
+      "learning_rate": 0.0003391353811149033,
+      "loss": 5.4116,
+      "step": 5809
+    },
+    {
+      "epoch": 1.9829351535836177,
+      "grad_norm": 2.858267068862915,
+      "learning_rate": 0.0003390216154721274,
+      "loss": 6.2049,
+      "step": 5810
+    },
+    {
+      "epoch": 1.9832764505119453,
+      "grad_norm": 2.9900565147399902,
+      "learning_rate": 0.00033890784982935153,
+      "loss": 5.5319,
+      "step": 5811
+    },
+    {
+      "epoch": 1.983617747440273,
+      "grad_norm": 4.313586711883545,
+      "learning_rate": 0.00033879408418657563,
+      "loss": 4.6686,
+      "step": 5812
+    },
+    {
+      "epoch": 1.9839590443686008,
+      "grad_norm": 2.9994773864746094,
+      "learning_rate": 0.0003386803185437998,
+      "loss": 5.9989,
+      "step": 5813
+    },
+    {
+      "epoch": 1.9843003412969282,
+      "grad_norm": 2.9691319465637207,
+      "learning_rate": 0.0003385665529010239,
+      "loss": 6.2986,
+      "step": 5814
+    },
+    {
+      "epoch": 1.984641638225256,
+      "grad_norm": 2.8935835361480713,
+      "learning_rate": 0.000338452787258248,
+      "loss": 5.6747,
+      "step": 5815
+    },
+    {
+      "epoch": 1.9849829351535835,
+      "grad_norm": 3.180413007736206,
+      "learning_rate": 0.00033833902161547216,
+      "loss": 5.2211,
+      "step": 5816
+    },
+    {
+      "epoch": 1.9853242320819113,
+      "grad_norm": 3.1071324348449707,
+      "learning_rate": 0.00033822525597269627,
+      "loss": 6.0189,
+      "step": 5817
+    },
+    {
+      "epoch": 1.985665529010239,
+      "grad_norm": 3.498764753341675,
+      "learning_rate": 0.00033811149032992037,
+      "loss": 5.6862,
+      "step": 5818
+    },
+    {
+      "epoch": 1.9860068259385666,
+      "grad_norm": 2.8438425064086914,
+      "learning_rate": 0.00033799772468714453,
+      "loss": 6.1487,
+      "step": 5819
+    },
+    {
+      "epoch": 1.9863481228668942,
+      "grad_norm": 2.9216058254241943,
+      "learning_rate": 0.0003378839590443686,
+      "loss": 6.1612,
+      "step": 5820
+    },
+    {
+      "epoch": 1.9866894197952218,
+      "grad_norm": 2.917191743850708,
+      "learning_rate": 0.0003377701934015927,
+      "loss": 6.3926,
+      "step": 5821
+    },
+    {
+      "epoch": 1.9870307167235495,
+      "grad_norm": 2.862936496734619,
+      "learning_rate": 0.00033765642775881685,
+      "loss": 6.1182,
+      "step": 5822
+    },
+    {
+      "epoch": 1.987372013651877,
+      "grad_norm": 3.223823308944702,
+      "learning_rate": 0.00033754266211604095,
+      "loss": 5.9686,
+      "step": 5823
+    },
+    {
+      "epoch": 1.9877133105802047,
+      "grad_norm": 3.052494764328003,
+      "learning_rate": 0.00033742889647326506,
+      "loss": 5.3687,
+      "step": 5824
+    },
+    {
+      "epoch": 1.9880546075085324,
+      "grad_norm": 3.0097758769989014,
+      "learning_rate": 0.0003373151308304892,
+      "loss": 6.1109,
+      "step": 5825
+    },
+    {
+      "epoch": 1.9883959044368602,
+      "grad_norm": 2.950054168701172,
+      "learning_rate": 0.0003372013651877133,
+      "loss": 6.4818,
+      "step": 5826
+    },
+    {
+      "epoch": 1.9887372013651876,
+      "grad_norm": 2.8829801082611084,
+      "learning_rate": 0.0003370875995449374,
+      "loss": 6.2982,
+      "step": 5827
+    },
+    {
+      "epoch": 1.9890784982935155,
+      "grad_norm": 2.908803701400757,
+      "learning_rate": 0.0003369738339021616,
+      "loss": 6.0992,
+      "step": 5828
+    },
+    {
+      "epoch": 1.9894197952218429,
+      "grad_norm": 3.0301074981689453,
+      "learning_rate": 0.0003368600682593857,
+      "loss": 6.7445,
+      "step": 5829
+    },
+    {
+      "epoch": 1.9897610921501707,
+      "grad_norm": 2.9564156532287598,
+      "learning_rate": 0.0003367463026166098,
+      "loss": 6.5264,
+      "step": 5830
+    },
+    {
+      "epoch": 1.9901023890784983,
+      "grad_norm": 2.951019763946533,
+      "learning_rate": 0.0003366325369738339,
+      "loss": 5.8757,
+      "step": 5831
+    },
+    {
+      "epoch": 1.990443686006826,
+      "grad_norm": 4.174768924713135,
+      "learning_rate": 0.000336518771331058,
+      "loss": 4.409,
+      "step": 5832
+    },
+    {
+      "epoch": 1.9907849829351536,
+      "grad_norm": 6.287966251373291,
+      "learning_rate": 0.00033640500568828216,
+      "loss": 5.3929,
+      "step": 5833
+    },
+    {
+      "epoch": 1.9911262798634812,
+      "grad_norm": 2.996116876602173,
+      "learning_rate": 0.00033629124004550627,
+      "loss": 6.4399,
+      "step": 5834
+    },
+    {
+      "epoch": 1.9914675767918089,
+      "grad_norm": 3.3284964561462402,
+      "learning_rate": 0.0003361774744027304,
+      "loss": 5.0079,
+      "step": 5835
+    },
+    {
+      "epoch": 1.9918088737201365,
+      "grad_norm": 2.8886935710906982,
+      "learning_rate": 0.00033606370875995453,
+      "loss": 6.3832,
+      "step": 5836
+    },
+    {
+      "epoch": 1.9921501706484641,
+      "grad_norm": 2.967557907104492,
+      "learning_rate": 0.00033594994311717864,
+      "loss": 6.3409,
+      "step": 5837
+    },
+    {
+      "epoch": 1.9924914675767917,
+      "grad_norm": 2.8980183601379395,
+      "learning_rate": 0.00033583617747440274,
+      "loss": 6.3379,
+      "step": 5838
+    },
+    {
+      "epoch": 1.9928327645051196,
+      "grad_norm": 3.0234856605529785,
+      "learning_rate": 0.00033572241183162685,
+      "loss": 6.155,
+      "step": 5839
+    },
+    {
+      "epoch": 1.993174061433447,
+      "grad_norm": 2.7687742710113525,
+      "learning_rate": 0.00033560864618885095,
+      "loss": 6.3193,
+      "step": 5840
+    },
+    {
+      "epoch": 1.9935153583617748,
+      "grad_norm": 2.851742744445801,
+      "learning_rate": 0.00033549488054607506,
+      "loss": 6.1613,
+      "step": 5841
+    },
+    {
+      "epoch": 1.9938566552901023,
+      "grad_norm": 2.8196914196014404,
+      "learning_rate": 0.0003353811149032992,
+      "loss": 6.0988,
+      "step": 5842
+    },
+    {
+      "epoch": 1.99419795221843,
+      "grad_norm": 2.847468614578247,
+      "learning_rate": 0.0003352673492605233,
+      "loss": 6.8367,
+      "step": 5843
+    },
+    {
+      "epoch": 1.9945392491467577,
+      "grad_norm": 2.88217830657959,
+      "learning_rate": 0.0003351535836177474,
+      "loss": 6.3181,
+      "step": 5844
+    },
+    {
+      "epoch": 1.9948805460750854,
+      "grad_norm": 2.8297886848449707,
+      "learning_rate": 0.0003350398179749716,
+      "loss": 5.9361,
+      "step": 5845
+    },
+    {
+      "epoch": 1.995221843003413,
+      "grad_norm": 2.9879589080810547,
+      "learning_rate": 0.0003349260523321957,
+      "loss": 6.2692,
+      "step": 5846
+    },
+    {
+      "epoch": 1.9955631399317406,
+      "grad_norm": 2.923250198364258,
+      "learning_rate": 0.0003348122866894198,
+      "loss": 6.4919,
+      "step": 5847
+    },
+    {
+      "epoch": 1.9959044368600682,
+      "grad_norm": 2.852890968322754,
+      "learning_rate": 0.00033469852104664395,
+      "loss": 5.8794,
+      "step": 5848
+    },
+    {
+      "epoch": 1.9962457337883959,
+      "grad_norm": 4.5131001472473145,
+      "learning_rate": 0.000334584755403868,
+      "loss": 4.9364,
+      "step": 5849
+    },
+    {
+      "epoch": 1.9965870307167235,
+      "grad_norm": 4.635778903961182,
+      "learning_rate": 0.0003344709897610921,
+      "loss": 5.363,
+      "step": 5850
+    },
+    {
+      "epoch": 1.9969283276450511,
+      "grad_norm": 2.963787317276001,
+      "learning_rate": 0.00033435722411831627,
+      "loss": 6.3649,
+      "step": 5851
+    },
+    {
+      "epoch": 1.997269624573379,
+      "grad_norm": 2.9836738109588623,
+      "learning_rate": 0.0003342434584755404,
+      "loss": 6.69,
+      "step": 5852
+    },
+    {
+      "epoch": 1.9976109215017064,
+      "grad_norm": 2.863739252090454,
+      "learning_rate": 0.00033412969283276453,
+      "loss": 6.4131,
+      "step": 5853
+    },
+    {
+      "epoch": 1.9979522184300342,
+      "grad_norm": 2.899862289428711,
+      "learning_rate": 0.00033401592718998864,
+      "loss": 6.0362,
+      "step": 5854
+    },
+    {
+      "epoch": 1.9982935153583616,
+      "grad_norm": 2.9389188289642334,
+      "learning_rate": 0.00033390216154721274,
+      "loss": 6.5771,
+      "step": 5855
+    },
+    {
+      "epoch": 1.9986348122866895,
+      "grad_norm": 5.9408183097839355,
+      "learning_rate": 0.0003337883959044369,
+      "loss": 5.5061,
+      "step": 5856
+    },
+    {
+      "epoch": 1.9989761092150171,
+      "grad_norm": 2.952606201171875,
+      "learning_rate": 0.000333674630261661,
+      "loss": 6.0148,
+      "step": 5857
+    },
+    {
+      "epoch": 1.9993174061433447,
+      "grad_norm": 2.8606393337249756,
+      "learning_rate": 0.0003335608646188851,
+      "loss": 6.0906,
+      "step": 5858
+    },
+    {
+      "epoch": 1.9996587030716724,
+      "grad_norm": 2.852064371109009,
+      "learning_rate": 0.0003334470989761092,
+      "loss": 6.4203,
+      "step": 5859
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.8471901416778564,
+      "learning_rate": 0.0003333333333333333,
+      "loss": 6.2409,
+      "step": 5860
+    },
+    {
+      "epoch": 2.000341296928328,
+      "grad_norm": 2.941194772720337,
+      "learning_rate": 0.00033321956769055743,
+      "loss": 6.3285,
+      "step": 5861
+    },
+    {
+      "epoch": 2.0006825938566553,
+      "grad_norm": 3.083709955215454,
+      "learning_rate": 0.0003331058020477816,
+      "loss": 6.2642,
+      "step": 5862
+    },
+    {
+      "epoch": 2.001023890784983,
+      "grad_norm": 3.8229711055755615,
+      "learning_rate": 0.0003329920364050057,
+      "loss": 3.7616,
+      "step": 5863
+    },
+    {
+      "epoch": 2.0013651877133105,
+      "grad_norm": 2.923698902130127,
+      "learning_rate": 0.0003328782707622298,
+      "loss": 5.66,
+      "step": 5864
+    },
+    {
+      "epoch": 2.0017064846416384,
+      "grad_norm": 3.1256601810455322,
+      "learning_rate": 0.00033276450511945396,
+      "loss": 5.7579,
+      "step": 5865
+    },
+    {
+      "epoch": 2.0020477815699658,
+      "grad_norm": 3.4216883182525635,
+      "learning_rate": 0.00033265073947667806,
+      "loss": 5.9409,
+      "step": 5866
+    },
+    {
+      "epoch": 2.0023890784982936,
+      "grad_norm": 3.03593373298645,
+      "learning_rate": 0.00033253697383390217,
+      "loss": 5.8932,
+      "step": 5867
+    },
+    {
+      "epoch": 2.002730375426621,
+      "grad_norm": 2.919698715209961,
+      "learning_rate": 0.0003324232081911263,
+      "loss": 5.8919,
+      "step": 5868
+    },
+    {
+      "epoch": 2.003071672354949,
+      "grad_norm": 2.8888025283813477,
+      "learning_rate": 0.0003323094425483504,
+      "loss": 5.7327,
+      "step": 5869
+    },
+    {
+      "epoch": 2.0034129692832763,
+      "grad_norm": 2.8589115142822266,
+      "learning_rate": 0.0003321956769055745,
+      "loss": 5.6514,
+      "step": 5870
+    },
+    {
+      "epoch": 2.003754266211604,
+      "grad_norm": 2.8871238231658936,
+      "learning_rate": 0.00033208191126279864,
+      "loss": 5.9583,
+      "step": 5871
+    },
+    {
+      "epoch": 2.0040955631399315,
+      "grad_norm": 2.953996181488037,
+      "learning_rate": 0.00033196814562002274,
+      "loss": 5.6646,
+      "step": 5872
+    },
+    {
+      "epoch": 2.0044368600682594,
+      "grad_norm": 2.8244423866271973,
+      "learning_rate": 0.0003318543799772469,
+      "loss": 6.174,
+      "step": 5873
+    },
+    {
+      "epoch": 2.0047781569965872,
+      "grad_norm": 2.8501133918762207,
+      "learning_rate": 0.000331740614334471,
+      "loss": 5.4038,
+      "step": 5874
+    },
+    {
+      "epoch": 2.0051194539249146,
+      "grad_norm": 2.8104753494262695,
+      "learning_rate": 0.0003316268486916951,
+      "loss": 6.5321,
+      "step": 5875
+    },
+    {
+      "epoch": 2.0054607508532425,
+      "grad_norm": 2.9065990447998047,
+      "learning_rate": 0.00033151308304891927,
+      "loss": 6.6124,
+      "step": 5876
+    },
+    {
+      "epoch": 2.00580204778157,
+      "grad_norm": 2.9225258827209473,
+      "learning_rate": 0.0003313993174061434,
+      "loss": 6.2582,
+      "step": 5877
+    },
+    {
+      "epoch": 2.0061433447098977,
+      "grad_norm": 2.842495918273926,
+      "learning_rate": 0.00033128555176336743,
+      "loss": 6.2283,
+      "step": 5878
+    },
+    {
+      "epoch": 2.006484641638225,
+      "grad_norm": 2.993636131286621,
+      "learning_rate": 0.0003311717861205916,
+      "loss": 6.1056,
+      "step": 5879
+    },
+    {
+      "epoch": 2.006825938566553,
+      "grad_norm": 2.954638957977295,
+      "learning_rate": 0.0003310580204778157,
+      "loss": 6.1397,
+      "step": 5880
+    },
+    {
+      "epoch": 2.0071672354948804,
+      "grad_norm": 2.83888840675354,
+      "learning_rate": 0.0003309442548350398,
+      "loss": 6.3731,
+      "step": 5881
+    },
+    {
+      "epoch": 2.0075085324232083,
+      "grad_norm": 2.950007438659668,
+      "learning_rate": 0.00033083048919226396,
+      "loss": 5.9267,
+      "step": 5882
+    },
+    {
+      "epoch": 2.0078498293515357,
+      "grad_norm": 2.8402795791625977,
+      "learning_rate": 0.00033071672354948806,
+      "loss": 6.5702,
+      "step": 5883
+    },
+    {
+      "epoch": 2.0081911262798635,
+      "grad_norm": 2.827852964401245,
+      "learning_rate": 0.00033060295790671217,
+      "loss": 6.7476,
+      "step": 5884
+    },
+    {
+      "epoch": 2.008532423208191,
+      "grad_norm": 2.843194007873535,
+      "learning_rate": 0.0003304891922639363,
+      "loss": 5.988,
+      "step": 5885
+    },
+    {
+      "epoch": 2.0088737201365188,
+      "grad_norm": 2.8419055938720703,
+      "learning_rate": 0.00033037542662116043,
+      "loss": 5.5696,
+      "step": 5886
+    },
+    {
+      "epoch": 2.0092150170648466,
+      "grad_norm": 2.968064308166504,
+      "learning_rate": 0.00033026166097838454,
+      "loss": 6.4374,
+      "step": 5887
+    },
+    {
+      "epoch": 2.009556313993174,
+      "grad_norm": 3.4870381355285645,
+      "learning_rate": 0.00033014789533560864,
+      "loss": 4.7893,
+      "step": 5888
+    },
+    {
+      "epoch": 2.009897610921502,
+      "grad_norm": 6.496885299682617,
+      "learning_rate": 0.00033003412969283275,
+      "loss": 4.7551,
+      "step": 5889
+    },
+    {
+      "epoch": 2.0102389078498293,
+      "grad_norm": 2.9047958850860596,
+      "learning_rate": 0.00032992036405005685,
+      "loss": 5.9317,
+      "step": 5890
+    },
+    {
+      "epoch": 2.010580204778157,
+      "grad_norm": 3.021362781524658,
+      "learning_rate": 0.000329806598407281,
+      "loss": 5.6877,
+      "step": 5891
+    },
+    {
+      "epoch": 2.0109215017064845,
+      "grad_norm": 3.03979754447937,
+      "learning_rate": 0.0003296928327645051,
+      "loss": 5.5397,
+      "step": 5892
+    },
+    {
+      "epoch": 2.0112627986348124,
+      "grad_norm": 3.0143215656280518,
+      "learning_rate": 0.0003295790671217293,
+      "loss": 6.1574,
+      "step": 5893
+    },
+    {
+      "epoch": 2.01160409556314,
+      "grad_norm": 2.9029927253723145,
+      "learning_rate": 0.0003294653014789534,
+      "loss": 6.3871,
+      "step": 5894
+    },
+    {
+      "epoch": 2.0119453924914676,
+      "grad_norm": 2.924941062927246,
+      "learning_rate": 0.0003293515358361775,
+      "loss": 5.6172,
+      "step": 5895
+    },
+    {
+      "epoch": 2.012286689419795,
+      "grad_norm": 2.9214353561401367,
+      "learning_rate": 0.00032923777019340164,
+      "loss": 6.5576,
+      "step": 5896
+    },
+    {
+      "epoch": 2.012627986348123,
+      "grad_norm": 2.9179751873016357,
+      "learning_rate": 0.00032912400455062575,
+      "loss": 6.3363,
+      "step": 5897
+    },
+    {
+      "epoch": 2.0129692832764503,
+      "grad_norm": 2.8579938411712646,
+      "learning_rate": 0.0003290102389078498,
+      "loss": 6.3428,
+      "step": 5898
+    },
+    {
+      "epoch": 2.013310580204778,
+      "grad_norm": 3.0229973793029785,
+      "learning_rate": 0.00032889647326507396,
+      "loss": 6.2619,
+      "step": 5899
+    },
+    {
+      "epoch": 2.013651877133106,
+      "grad_norm": 4.09330940246582,
+      "learning_rate": 0.00032878270762229806,
+      "loss": 4.4237,
+      "step": 5900
+    },
+    {
+      "epoch": 2.0139931740614334,
+      "grad_norm": 3.2335963249206543,
+      "learning_rate": 0.00032866894197952217,
+      "loss": 5.7218,
+      "step": 5901
+    },
+    {
+      "epoch": 2.0143344709897613,
+      "grad_norm": 3.3706767559051514,
+      "learning_rate": 0.0003285551763367463,
+      "loss": 5.6349,
+      "step": 5902
+    },
+    {
+      "epoch": 2.0146757679180887,
+      "grad_norm": 3.0019240379333496,
+      "learning_rate": 0.00032844141069397043,
+      "loss": 5.9354,
+      "step": 5903
+    },
+    {
+      "epoch": 2.0150170648464165,
+      "grad_norm": 2.8847649097442627,
+      "learning_rate": 0.00032832764505119454,
+      "loss": 6.571,
+      "step": 5904
+    },
+    {
+      "epoch": 2.015358361774744,
+      "grad_norm": 2.9488613605499268,
+      "learning_rate": 0.0003282138794084187,
+      "loss": 6.7779,
+      "step": 5905
+    },
+    {
+      "epoch": 2.0156996587030718,
+      "grad_norm": 2.824702739715576,
+      "learning_rate": 0.0003281001137656428,
+      "loss": 6.7856,
+      "step": 5906
+    },
+    {
+      "epoch": 2.016040955631399,
+      "grad_norm": 2.872587203979492,
+      "learning_rate": 0.00032798634812286685,
+      "loss": 6.1544,
+      "step": 5907
+    },
+    {
+      "epoch": 2.016382252559727,
+      "grad_norm": 3.073383092880249,
+      "learning_rate": 0.000327872582480091,
+      "loss": 6.0585,
+      "step": 5908
+    },
+    {
+      "epoch": 2.0167235494880544,
+      "grad_norm": 3.2500085830688477,
+      "learning_rate": 0.0003277588168373151,
+      "loss": 4.9903,
+      "step": 5909
+    },
+    {
+      "epoch": 2.0170648464163823,
+      "grad_norm": 3.061755895614624,
+      "learning_rate": 0.0003276450511945392,
+      "loss": 5.9263,
+      "step": 5910
+    },
+    {
+      "epoch": 2.0174061433447097,
+      "grad_norm": 3.07401442527771,
+      "learning_rate": 0.0003275312855517634,
+      "loss": 6.5287,
+      "step": 5911
+    },
+    {
+      "epoch": 2.0177474402730375,
+      "grad_norm": 2.897965908050537,
+      "learning_rate": 0.0003274175199089875,
+      "loss": 6.4134,
+      "step": 5912
+    },
+    {
+      "epoch": 2.0180887372013654,
+      "grad_norm": 3.0315048694610596,
+      "learning_rate": 0.00032730375426621164,
+      "loss": 5.2327,
+      "step": 5913
+    },
+    {
+      "epoch": 2.018430034129693,
+      "grad_norm": 2.9312515258789062,
+      "learning_rate": 0.00032718998862343575,
+      "loss": 5.4876,
+      "step": 5914
+    },
+    {
+      "epoch": 2.0187713310580206,
+      "grad_norm": 2.948835849761963,
+      "learning_rate": 0.00032707622298065985,
+      "loss": 6.1955,
+      "step": 5915
+    },
+    {
+      "epoch": 2.019112627986348,
+      "grad_norm": 2.9583792686462402,
+      "learning_rate": 0.000326962457337884,
+      "loss": 4.9185,
+      "step": 5916
+    },
+    {
+      "epoch": 2.019453924914676,
+      "grad_norm": 2.9608967304229736,
+      "learning_rate": 0.00032684869169510806,
+      "loss": 6.5537,
+      "step": 5917
+    },
+    {
+      "epoch": 2.0197952218430033,
+      "grad_norm": 2.8537795543670654,
+      "learning_rate": 0.00032673492605233217,
+      "loss": 6.518,
+      "step": 5918
+    },
+    {
+      "epoch": 2.020136518771331,
+      "grad_norm": 3.467559576034546,
+      "learning_rate": 0.0003266211604095563,
+      "loss": 5.4909,
+      "step": 5919
+    },
+    {
+      "epoch": 2.0204778156996586,
+      "grad_norm": 2.9610671997070312,
+      "learning_rate": 0.00032650739476678043,
+      "loss": 5.8352,
+      "step": 5920
+    },
+    {
+      "epoch": 2.0208191126279864,
+      "grad_norm": 2.9346728324890137,
+      "learning_rate": 0.00032639362912400454,
+      "loss": 5.7961,
+      "step": 5921
+    },
+    {
+      "epoch": 2.021160409556314,
+      "grad_norm": 3.0085277557373047,
+      "learning_rate": 0.0003262798634812287,
+      "loss": 5.5517,
+      "step": 5922
+    },
+    {
+      "epoch": 2.0215017064846417,
+      "grad_norm": 2.949371337890625,
+      "learning_rate": 0.0003261660978384528,
+      "loss": 6.364,
+      "step": 5923
+    },
+    {
+      "epoch": 2.021843003412969,
+      "grad_norm": 4.0263752937316895,
+      "learning_rate": 0.0003260523321956769,
+      "loss": 3.3965,
+      "step": 5924
+    },
+    {
+      "epoch": 2.022184300341297,
+      "grad_norm": 3.15136456489563,
+      "learning_rate": 0.00032593856655290107,
+      "loss": 5.744,
+      "step": 5925
+    },
+    {
+      "epoch": 2.0225255972696248,
+      "grad_norm": 2.9317257404327393,
+      "learning_rate": 0.00032582480091012517,
+      "loss": 6.2774,
+      "step": 5926
+    },
+    {
+      "epoch": 2.022866894197952,
+      "grad_norm": 2.901693344116211,
+      "learning_rate": 0.0003257110352673492,
+      "loss": 5.9736,
+      "step": 5927
+    },
+    {
+      "epoch": 2.02320819112628,
+      "grad_norm": 2.804474353790283,
+      "learning_rate": 0.0003255972696245734,
+      "loss": 6.2335,
+      "step": 5928
+    },
+    {
+      "epoch": 2.0235494880546074,
+      "grad_norm": 2.836078405380249,
+      "learning_rate": 0.0003254835039817975,
+      "loss": 6.5145,
+      "step": 5929
+    },
+    {
+      "epoch": 2.0238907849829353,
+      "grad_norm": 2.8914759159088135,
+      "learning_rate": 0.0003253697383390216,
+      "loss": 5.9104,
+      "step": 5930
+    },
+    {
+      "epoch": 2.0242320819112627,
+      "grad_norm": 3.001803398132324,
+      "learning_rate": 0.00032525597269624575,
+      "loss": 5.733,
+      "step": 5931
+    },
+    {
+      "epoch": 2.0245733788395905,
+      "grad_norm": 3.05202054977417,
+      "learning_rate": 0.00032514220705346985,
+      "loss": 5.7444,
+      "step": 5932
+    },
+    {
+      "epoch": 2.024914675767918,
+      "grad_norm": 2.8296566009521484,
+      "learning_rate": 0.000325028441410694,
+      "loss": 6.3294,
+      "step": 5933
+    },
+    {
+      "epoch": 2.025255972696246,
+      "grad_norm": 2.734096050262451,
+      "learning_rate": 0.0003249146757679181,
+      "loss": 6.2183,
+      "step": 5934
+    },
+    {
+      "epoch": 2.025597269624573,
+      "grad_norm": 2.8434300422668457,
+      "learning_rate": 0.0003248009101251422,
+      "loss": 5.6148,
+      "step": 5935
+    },
+    {
+      "epoch": 2.025938566552901,
+      "grad_norm": 3.7026021480560303,
+      "learning_rate": 0.0003246871444823664,
+      "loss": 5.4888,
+      "step": 5936
+    },
+    {
+      "epoch": 2.0262798634812285,
+      "grad_norm": 2.7599542140960693,
+      "learning_rate": 0.00032457337883959043,
+      "loss": 5.6935,
+      "step": 5937
+    },
+    {
+      "epoch": 2.0266211604095563,
+      "grad_norm": 4.715770244598389,
+      "learning_rate": 0.00032445961319681454,
+      "loss": 5.3177,
+      "step": 5938
+    },
+    {
+      "epoch": 2.026962457337884,
+      "grad_norm": 2.897782802581787,
+      "learning_rate": 0.0003243458475540387,
+      "loss": 5.7596,
+      "step": 5939
+    },
+    {
+      "epoch": 2.0273037542662116,
+      "grad_norm": 2.934124708175659,
+      "learning_rate": 0.0003242320819112628,
+      "loss": 6.1473,
+      "step": 5940
+    },
+    {
+      "epoch": 2.0276450511945394,
+      "grad_norm": 2.9181840419769287,
+      "learning_rate": 0.0003241183162684869,
+      "loss": 5.8151,
+      "step": 5941
+    },
+    {
+      "epoch": 2.027986348122867,
+      "grad_norm": 2.937746286392212,
+      "learning_rate": 0.00032400455062571107,
+      "loss": 5.9434,
+      "step": 5942
+    },
+    {
+      "epoch": 2.0283276450511947,
+      "grad_norm": 3.7255730628967285,
+      "learning_rate": 0.00032389078498293517,
+      "loss": 5.0394,
+      "step": 5943
+    },
+    {
+      "epoch": 2.028668941979522,
+      "grad_norm": 2.898202896118164,
+      "learning_rate": 0.0003237770193401593,
+      "loss": 5.7591,
+      "step": 5944
+    },
+    {
+      "epoch": 2.02901023890785,
+      "grad_norm": 3.671320915222168,
+      "learning_rate": 0.00032366325369738344,
+      "loss": 6.2918,
+      "step": 5945
+    },
+    {
+      "epoch": 2.0293515358361773,
+      "grad_norm": 2.9321327209472656,
+      "learning_rate": 0.0003235494880546075,
+      "loss": 6.0174,
+      "step": 5946
+    },
+    {
+      "epoch": 2.029692832764505,
+      "grad_norm": 2.786679744720459,
+      "learning_rate": 0.0003234357224118316,
+      "loss": 6.4303,
+      "step": 5947
+    },
+    {
+      "epoch": 2.0300341296928326,
+      "grad_norm": 2.862600088119507,
+      "learning_rate": 0.00032332195676905575,
+      "loss": 6.0229,
+      "step": 5948
+    },
+    {
+      "epoch": 2.0303754266211604,
+      "grad_norm": 2.9102251529693604,
+      "learning_rate": 0.00032320819112627985,
+      "loss": 5.6793,
+      "step": 5949
+    },
+    {
+      "epoch": 2.030716723549488,
+      "grad_norm": 2.8339219093322754,
+      "learning_rate": 0.00032309442548350396,
+      "loss": 6.1311,
+      "step": 5950
+    },
+    {
+      "epoch": 2.0310580204778157,
+      "grad_norm": 2.7801735401153564,
+      "learning_rate": 0.0003229806598407281,
+      "loss": 6.1714,
+      "step": 5951
+    },
+    {
+      "epoch": 2.0313993174061435,
+      "grad_norm": 2.792600631713867,
+      "learning_rate": 0.0003228668941979522,
+      "loss": 6.4515,
+      "step": 5952
+    },
+    {
+      "epoch": 2.031740614334471,
+      "grad_norm": 2.876671075820923,
+      "learning_rate": 0.0003227531285551764,
+      "loss": 6.3113,
+      "step": 5953
+    },
+    {
+      "epoch": 2.032081911262799,
+      "grad_norm": 2.9454519748687744,
+      "learning_rate": 0.0003226393629124005,
+      "loss": 6.5142,
+      "step": 5954
+    },
+    {
+      "epoch": 2.032423208191126,
+      "grad_norm": 2.833214044570923,
+      "learning_rate": 0.0003225255972696246,
+      "loss": 6.0907,
+      "step": 5955
+    },
+    {
+      "epoch": 2.032764505119454,
+      "grad_norm": 4.851685047149658,
+      "learning_rate": 0.0003224118316268487,
+      "loss": 5.22,
+      "step": 5956
+    },
+    {
+      "epoch": 2.0331058020477815,
+      "grad_norm": 3.4849393367767334,
+      "learning_rate": 0.0003222980659840728,
+      "loss": 5.0735,
+      "step": 5957
+    },
+    {
+      "epoch": 2.0334470989761093,
+      "grad_norm": 3.021411180496216,
+      "learning_rate": 0.0003221843003412969,
+      "loss": 6.2273,
+      "step": 5958
+    },
+    {
+      "epoch": 2.0337883959044367,
+      "grad_norm": 2.9004580974578857,
+      "learning_rate": 0.00032207053469852107,
+      "loss": 6.6881,
+      "step": 5959
+    },
+    {
+      "epoch": 2.0341296928327646,
+      "grad_norm": 2.9186768531799316,
+      "learning_rate": 0.00032195676905574517,
+      "loss": 5.6578,
+      "step": 5960
+    },
+    {
+      "epoch": 2.034470989761092,
+      "grad_norm": 2.9466514587402344,
+      "learning_rate": 0.0003218430034129693,
+      "loss": 6.1247,
+      "step": 5961
+    },
+    {
+      "epoch": 2.03481228668942,
+      "grad_norm": 2.9205827713012695,
+      "learning_rate": 0.00032172923777019344,
+      "loss": 5.8445,
+      "step": 5962
+    },
+    {
+      "epoch": 2.0351535836177472,
+      "grad_norm": 2.9399664402008057,
+      "learning_rate": 0.00032161547212741754,
+      "loss": 6.3061,
+      "step": 5963
+    },
+    {
+      "epoch": 2.035494880546075,
+      "grad_norm": 2.7923221588134766,
+      "learning_rate": 0.00032150170648464165,
+      "loss": 6.9318,
+      "step": 5964
+    },
+    {
+      "epoch": 2.035836177474403,
+      "grad_norm": 2.841721773147583,
+      "learning_rate": 0.0003213879408418658,
+      "loss": 6.2328,
+      "step": 5965
+    },
+    {
+      "epoch": 2.0361774744027303,
+      "grad_norm": 2.805899143218994,
+      "learning_rate": 0.00032127417519908986,
+      "loss": 5.0354,
+      "step": 5966
+    },
+    {
+      "epoch": 2.036518771331058,
+      "grad_norm": 2.8029048442840576,
+      "learning_rate": 0.00032116040955631396,
+      "loss": 6.3708,
+      "step": 5967
+    },
+    {
+      "epoch": 2.0368600682593856,
+      "grad_norm": 2.7926340103149414,
+      "learning_rate": 0.0003210466439135381,
+      "loss": 5.9584,
+      "step": 5968
+    },
+    {
+      "epoch": 2.0372013651877134,
+      "grad_norm": 2.774442672729492,
+      "learning_rate": 0.0003209328782707622,
+      "loss": 6.4848,
+      "step": 5969
+    },
+    {
+      "epoch": 2.037542662116041,
+      "grad_norm": 4.830069065093994,
+      "learning_rate": 0.00032081911262798633,
+      "loss": 5.8823,
+      "step": 5970
+    },
+    {
+      "epoch": 2.0378839590443687,
+      "grad_norm": 2.8296518325805664,
+      "learning_rate": 0.0003207053469852105,
+      "loss": 5.8528,
+      "step": 5971
+    },
+    {
+      "epoch": 2.038225255972696,
+      "grad_norm": 2.9755330085754395,
+      "learning_rate": 0.0003205915813424346,
+      "loss": 6.5889,
+      "step": 5972
+    },
+    {
+      "epoch": 2.038566552901024,
+      "grad_norm": 3.333840847015381,
+      "learning_rate": 0.0003204778156996587,
+      "loss": 5.3275,
+      "step": 5973
+    },
+    {
+      "epoch": 2.0389078498293514,
+      "grad_norm": 2.9385743141174316,
+      "learning_rate": 0.00032036405005688286,
+      "loss": 6.0905,
+      "step": 5974
+    },
+    {
+      "epoch": 2.039249146757679,
+      "grad_norm": 4.615128040313721,
+      "learning_rate": 0.00032025028441410696,
+      "loss": 5.379,
+      "step": 5975
+    },
+    {
+      "epoch": 2.0395904436860066,
+      "grad_norm": 2.938904047012329,
+      "learning_rate": 0.00032013651877133107,
+      "loss": 6.5808,
+      "step": 5976
+    },
+    {
+      "epoch": 2.0399317406143345,
+      "grad_norm": 2.9147865772247314,
+      "learning_rate": 0.00032002275312855517,
+      "loss": 6.7389,
+      "step": 5977
+    },
+    {
+      "epoch": 2.0402730375426623,
+      "grad_norm": 2.8320443630218506,
+      "learning_rate": 0.0003199089874857793,
+      "loss": 5.7683,
+      "step": 5978
+    },
+    {
+      "epoch": 2.0406143344709897,
+      "grad_norm": 3.8267099857330322,
+      "learning_rate": 0.00031979522184300344,
+      "loss": 4.3207,
+      "step": 5979
+    },
+    {
+      "epoch": 2.0409556313993176,
+      "grad_norm": 3.4486052989959717,
+      "learning_rate": 0.00031968145620022754,
+      "loss": 6.0473,
+      "step": 5980
+    },
+    {
+      "epoch": 2.041296928327645,
+      "grad_norm": 3.4830849170684814,
+      "learning_rate": 0.00031956769055745165,
+      "loss": 5.3791,
+      "step": 5981
+    },
+    {
+      "epoch": 2.041638225255973,
+      "grad_norm": 2.9314374923706055,
+      "learning_rate": 0.0003194539249146758,
+      "loss": 6.1228,
+      "step": 5982
+    },
+    {
+      "epoch": 2.0419795221843002,
+      "grad_norm": 3.42529034614563,
+      "learning_rate": 0.0003193401592718999,
+      "loss": 5.5453,
+      "step": 5983
+    },
+    {
+      "epoch": 2.042320819112628,
+      "grad_norm": 2.903790235519409,
+      "learning_rate": 0.000319226393629124,
+      "loss": 5.9758,
+      "step": 5984
+    },
+    {
+      "epoch": 2.0426621160409555,
+      "grad_norm": 3.009942054748535,
+      "learning_rate": 0.0003191126279863481,
+      "loss": 6.5988,
+      "step": 5985
+    },
+    {
+      "epoch": 2.0430034129692833,
+      "grad_norm": 2.9778008460998535,
+      "learning_rate": 0.0003189988623435722,
+      "loss": 6.0532,
+      "step": 5986
+    },
+    {
+      "epoch": 2.0433447098976107,
+      "grad_norm": 2.8316783905029297,
+      "learning_rate": 0.00031888509670079633,
+      "loss": 6.3487,
+      "step": 5987
+    },
+    {
+      "epoch": 2.0436860068259386,
+      "grad_norm": 2.837663173675537,
+      "learning_rate": 0.0003187713310580205,
+      "loss": 6.4316,
+      "step": 5988
+    },
+    {
+      "epoch": 2.044027303754266,
+      "grad_norm": 2.863551616668701,
+      "learning_rate": 0.0003186575654152446,
+      "loss": 5.7195,
+      "step": 5989
+    },
+    {
+      "epoch": 2.044368600682594,
+      "grad_norm": 2.814051389694214,
+      "learning_rate": 0.0003185437997724687,
+      "loss": 6.0491,
+      "step": 5990
+    },
+    {
+      "epoch": 2.0447098976109217,
+      "grad_norm": 2.8193931579589844,
+      "learning_rate": 0.00031843003412969286,
+      "loss": 6.4645,
+      "step": 5991
+    },
+    {
+      "epoch": 2.045051194539249,
+      "grad_norm": 2.9294307231903076,
+      "learning_rate": 0.00031831626848691696,
+      "loss": 5.8573,
+      "step": 5992
+    },
+    {
+      "epoch": 2.045392491467577,
+      "grad_norm": 2.906787395477295,
+      "learning_rate": 0.00031820250284414107,
+      "loss": 6.3565,
+      "step": 5993
+    },
+    {
+      "epoch": 2.0457337883959044,
+      "grad_norm": 3.021745443344116,
+      "learning_rate": 0.00031808873720136523,
+      "loss": 5.4691,
+      "step": 5994
+    },
+    {
+      "epoch": 2.046075085324232,
+      "grad_norm": 2.877673387527466,
+      "learning_rate": 0.0003179749715585893,
+      "loss": 5.5119,
+      "step": 5995
+    },
+    {
+      "epoch": 2.0464163822525596,
+      "grad_norm": 3.2106802463531494,
+      "learning_rate": 0.00031786120591581344,
+      "loss": 5.6146,
+      "step": 5996
+    },
+    {
+      "epoch": 2.0467576791808875,
+      "grad_norm": 2.90977144241333,
+      "learning_rate": 0.00031774744027303754,
+      "loss": 6.2368,
+      "step": 5997
+    },
+    {
+      "epoch": 2.047098976109215,
+      "grad_norm": 2.888395309448242,
+      "learning_rate": 0.00031763367463026165,
+      "loss": 6.3792,
+      "step": 5998
+    },
+    {
+      "epoch": 2.0474402730375427,
+      "grad_norm": 2.8118643760681152,
+      "learning_rate": 0.0003175199089874858,
+      "loss": 6.0576,
+      "step": 5999
+    },
+    {
+      "epoch": 2.04778156996587,
+      "grad_norm": 2.768461227416992,
+      "learning_rate": 0.0003174061433447099,
+      "loss": 6.3125,
+      "step": 6000
+    },
+    {
+      "epoch": 2.048122866894198,
+      "grad_norm": 2.7696850299835205,
+      "learning_rate": 0.000317292377701934,
+      "loss": 6.3775,
+      "step": 6001
+    },
+    {
+      "epoch": 2.0484641638225254,
+      "grad_norm": 2.7902398109436035,
+      "learning_rate": 0.0003171786120591582,
+      "loss": 6.0764,
+      "step": 6002
+    },
+    {
+      "epoch": 2.0488054607508532,
+      "grad_norm": 2.9154062271118164,
+      "learning_rate": 0.0003170648464163823,
+      "loss": 6.4892,
+      "step": 6003
+    },
+    {
+      "epoch": 2.049146757679181,
+      "grad_norm": 2.908606767654419,
+      "learning_rate": 0.0003169510807736064,
+      "loss": 6.5462,
+      "step": 6004
+    },
+    {
+      "epoch": 2.0494880546075085,
+      "grad_norm": 2.8876357078552246,
+      "learning_rate": 0.0003168373151308305,
+      "loss": 6.7508,
+      "step": 6005
+    },
+    {
+      "epoch": 2.0498293515358363,
+      "grad_norm": 2.836838960647583,
+      "learning_rate": 0.0003167235494880546,
+      "loss": 6.5657,
+      "step": 6006
+    },
+    {
+      "epoch": 2.0501706484641637,
+      "grad_norm": 2.8034183979034424,
+      "learning_rate": 0.0003166097838452787,
+      "loss": 6.3752,
+      "step": 6007
+    },
+    {
+      "epoch": 2.0505119453924916,
+      "grad_norm": 2.8830206394195557,
+      "learning_rate": 0.00031649601820250286,
+      "loss": 6.294,
+      "step": 6008
+    },
+    {
+      "epoch": 2.050853242320819,
+      "grad_norm": 2.811297655105591,
+      "learning_rate": 0.00031638225255972696,
+      "loss": 6.1428,
+      "step": 6009
+    },
+    {
+      "epoch": 2.051194539249147,
+      "grad_norm": 2.8806638717651367,
+      "learning_rate": 0.00031626848691695107,
+      "loss": 6.192,
+      "step": 6010
+    },
+    {
+      "epoch": 2.0515358361774743,
+      "grad_norm": 2.9297285079956055,
+      "learning_rate": 0.00031615472127417523,
+      "loss": 6.2655,
+      "step": 6011
+    },
+    {
+      "epoch": 2.051877133105802,
+      "grad_norm": 3.0114855766296387,
+      "learning_rate": 0.00031604095563139933,
+      "loss": 5.8824,
+      "step": 6012
+    },
+    {
+      "epoch": 2.0522184300341295,
+      "grad_norm": 3.1099727153778076,
+      "learning_rate": 0.00031592718998862344,
+      "loss": 5.416,
+      "step": 6013
+    },
+    {
+      "epoch": 2.0525597269624574,
+      "grad_norm": 2.8879382610321045,
+      "learning_rate": 0.00031581342434584754,
+      "loss": 5.822,
+      "step": 6014
+    },
+    {
+      "epoch": 2.0529010238907848,
+      "grad_norm": 2.9049463272094727,
+      "learning_rate": 0.00031569965870307165,
+      "loss": 5.4447,
+      "step": 6015
+    },
+    {
+      "epoch": 2.0532423208191126,
+      "grad_norm": 2.8603172302246094,
+      "learning_rate": 0.0003155858930602958,
+      "loss": 6.007,
+      "step": 6016
+    },
+    {
+      "epoch": 2.0535836177474405,
+      "grad_norm": 3.1179702281951904,
+      "learning_rate": 0.0003154721274175199,
+      "loss": 6.4848,
+      "step": 6017
+    },
+    {
+      "epoch": 2.053924914675768,
+      "grad_norm": 3.008000612258911,
+      "learning_rate": 0.000315358361774744,
+      "loss": 6.0294,
+      "step": 6018
+    },
+    {
+      "epoch": 2.0542662116040957,
+      "grad_norm": 2.8747920989990234,
+      "learning_rate": 0.0003152445961319682,
+      "loss": 6.0033,
+      "step": 6019
+    },
+    {
+      "epoch": 2.054607508532423,
+      "grad_norm": 2.8617520332336426,
+      "learning_rate": 0.0003151308304891923,
+      "loss": 6.4272,
+      "step": 6020
+    },
+    {
+      "epoch": 2.054948805460751,
+      "grad_norm": 2.861337900161743,
+      "learning_rate": 0.0003150170648464164,
+      "loss": 6.1443,
+      "step": 6021
+    },
+    {
+      "epoch": 2.0552901023890784,
+      "grad_norm": 2.852907180786133,
+      "learning_rate": 0.00031490329920364055,
+      "loss": 5.7435,
+      "step": 6022
+    },
+    {
+      "epoch": 2.0556313993174062,
+      "grad_norm": 3.9914369583129883,
+      "learning_rate": 0.00031478953356086465,
+      "loss": 5.5631,
+      "step": 6023
+    },
+    {
+      "epoch": 2.0559726962457336,
+      "grad_norm": 3.1209657192230225,
+      "learning_rate": 0.0003146757679180887,
+      "loss": 5.8574,
+      "step": 6024
+    },
+    {
+      "epoch": 2.0563139931740615,
+      "grad_norm": 2.9207725524902344,
+      "learning_rate": 0.00031456200227531286,
+      "loss": 6.2952,
+      "step": 6025
+    },
+    {
+      "epoch": 2.056655290102389,
+      "grad_norm": 3.0054728984832764,
+      "learning_rate": 0.00031444823663253697,
+      "loss": 5.3445,
+      "step": 6026
+    },
+    {
+      "epoch": 2.0569965870307167,
+      "grad_norm": 3.291370391845703,
+      "learning_rate": 0.00031433447098976107,
+      "loss": 4.9664,
+      "step": 6027
+    },
+    {
+      "epoch": 2.057337883959044,
+      "grad_norm": 2.8859901428222656,
+      "learning_rate": 0.00031422070534698523,
+      "loss": 5.9449,
+      "step": 6028
+    },
+    {
+      "epoch": 2.057679180887372,
+      "grad_norm": 3.0395874977111816,
+      "learning_rate": 0.00031410693970420933,
+      "loss": 6.0376,
+      "step": 6029
+    },
+    {
+      "epoch": 2.0580204778157,
+      "grad_norm": 2.9008989334106445,
+      "learning_rate": 0.00031399317406143344,
+      "loss": 6.4105,
+      "step": 6030
+    },
+    {
+      "epoch": 2.0583617747440273,
+      "grad_norm": 2.9411637783050537,
+      "learning_rate": 0.0003138794084186576,
+      "loss": 5.6366,
+      "step": 6031
+    },
+    {
+      "epoch": 2.058703071672355,
+      "grad_norm": 2.858750104904175,
+      "learning_rate": 0.0003137656427758817,
+      "loss": 6.5325,
+      "step": 6032
+    },
+    {
+      "epoch": 2.0590443686006825,
+      "grad_norm": 3.046865701675415,
+      "learning_rate": 0.0003136518771331058,
+      "loss": 6.317,
+      "step": 6033
+    },
+    {
+      "epoch": 2.0593856655290104,
+      "grad_norm": 2.996433734893799,
+      "learning_rate": 0.0003135381114903299,
+      "loss": 6.144,
+      "step": 6034
+    },
+    {
+      "epoch": 2.0597269624573378,
+      "grad_norm": 2.8240692615509033,
+      "learning_rate": 0.000313424345847554,
+      "loss": 6.3732,
+      "step": 6035
+    },
+    {
+      "epoch": 2.0600682593856656,
+      "grad_norm": 2.915379047393799,
+      "learning_rate": 0.0003133105802047782,
+      "loss": 6.1081,
+      "step": 6036
+    },
+    {
+      "epoch": 2.060409556313993,
+      "grad_norm": 6.734116077423096,
+      "learning_rate": 0.0003131968145620023,
+      "loss": 4.9619,
+      "step": 6037
+    },
+    {
+      "epoch": 2.060750853242321,
+      "grad_norm": 2.909330129623413,
+      "learning_rate": 0.0003130830489192264,
+      "loss": 6.0949,
+      "step": 6038
+    },
+    {
+      "epoch": 2.0610921501706483,
+      "grad_norm": 2.775883197784424,
+      "learning_rate": 0.00031296928327645055,
+      "loss": 6.0901,
+      "step": 6039
+    },
+    {
+      "epoch": 2.061433447098976,
+      "grad_norm": 3.041524648666382,
+      "learning_rate": 0.00031285551763367465,
+      "loss": 6.8417,
+      "step": 6040
+    },
+    {
+      "epoch": 2.0617747440273035,
+      "grad_norm": 2.883164882659912,
+      "learning_rate": 0.00031274175199089876,
+      "loss": 6.3803,
+      "step": 6041
+    },
+    {
+      "epoch": 2.0621160409556314,
+      "grad_norm": 3.1183626651763916,
+      "learning_rate": 0.0003126279863481229,
+      "loss": 6.1928,
+      "step": 6042
+    },
+    {
+      "epoch": 2.0624573378839592,
+      "grad_norm": 2.8333499431610107,
+      "learning_rate": 0.000312514220705347,
+      "loss": 6.2385,
+      "step": 6043
+    },
+    {
+      "epoch": 2.0627986348122866,
+      "grad_norm": 2.8665261268615723,
+      "learning_rate": 0.00031240045506257107,
+      "loss": 6.8755,
+      "step": 6044
+    },
+    {
+      "epoch": 2.0631399317406145,
+      "grad_norm": 4.331704139709473,
+      "learning_rate": 0.00031228668941979523,
+      "loss": 4.8018,
+      "step": 6045
+    },
+    {
+      "epoch": 2.063481228668942,
+      "grad_norm": 2.9394123554229736,
+      "learning_rate": 0.00031217292377701934,
+      "loss": 6.2594,
+      "step": 6046
+    },
+    {
+      "epoch": 2.0638225255972698,
+      "grad_norm": 2.87528395652771,
+      "learning_rate": 0.00031205915813424344,
+      "loss": 6.0032,
+      "step": 6047
+    },
+    {
+      "epoch": 2.064163822525597,
+      "grad_norm": 2.8144853115081787,
+      "learning_rate": 0.0003119453924914676,
+      "loss": 6.0189,
+      "step": 6048
+    },
+    {
+      "epoch": 2.064505119453925,
+      "grad_norm": 4.7918243408203125,
+      "learning_rate": 0.0003118316268486917,
+      "loss": 4.9166,
+      "step": 6049
+    },
+    {
+      "epoch": 2.0648464163822524,
+      "grad_norm": 2.0117385387420654,
+      "learning_rate": 0.0003117178612059158,
+      "loss": 3.1662,
+      "step": 6050
+    },
+    {
+      "epoch": 2.0651877133105803,
+      "grad_norm": 2.930635690689087,
+      "learning_rate": 0.00031160409556313997,
+      "loss": 6.7214,
+      "step": 6051
+    },
+    {
+      "epoch": 2.0655290102389077,
+      "grad_norm": 2.864907741546631,
+      "learning_rate": 0.0003114903299203641,
+      "loss": 6.3194,
+      "step": 6052
+    },
+    {
+      "epoch": 2.0658703071672355,
+      "grad_norm": 2.9267427921295166,
+      "learning_rate": 0.0003113765642775881,
+      "loss": 6.393,
+      "step": 6053
+    },
+    {
+      "epoch": 2.066211604095563,
+      "grad_norm": 2.9477977752685547,
+      "learning_rate": 0.0003112627986348123,
+      "loss": 6.4259,
+      "step": 6054
+    },
+    {
+      "epoch": 2.0665529010238908,
+      "grad_norm": 5.001644611358643,
+      "learning_rate": 0.0003111490329920364,
+      "loss": 4.6699,
+      "step": 6055
+    },
+    {
+      "epoch": 2.0668941979522186,
+      "grad_norm": 2.8814961910247803,
+      "learning_rate": 0.0003110352673492605,
+      "loss": 6.1531,
+      "step": 6056
+    },
+    {
+      "epoch": 2.067235494880546,
+      "grad_norm": 2.9188053607940674,
+      "learning_rate": 0.00031092150170648465,
+      "loss": 5.8409,
+      "step": 6057
+    },
+    {
+      "epoch": 2.067576791808874,
+      "grad_norm": 2.8715567588806152,
+      "learning_rate": 0.00031080773606370876,
+      "loss": 5.4793,
+      "step": 6058
+    },
+    {
+      "epoch": 2.0679180887372013,
+      "grad_norm": 9.393294334411621,
+      "learning_rate": 0.0003106939704209329,
+      "loss": 4.9354,
+      "step": 6059
+    },
+    {
+      "epoch": 2.068259385665529,
+      "grad_norm": 2.8406503200531006,
+      "learning_rate": 0.000310580204778157,
+      "loss": 5.92,
+      "step": 6060
+    },
+    {
+      "epoch": 2.0686006825938565,
+      "grad_norm": 3.1945927143096924,
+      "learning_rate": 0.0003104664391353811,
+      "loss": 5.5444,
+      "step": 6061
+    },
+    {
+      "epoch": 2.0689419795221844,
+      "grad_norm": 2.9627017974853516,
+      "learning_rate": 0.0003103526734926053,
+      "loss": 6.2811,
+      "step": 6062
+    },
+    {
+      "epoch": 2.069283276450512,
+      "grad_norm": 2.85520076751709,
+      "learning_rate": 0.00031023890784982934,
+      "loss": 6.3404,
+      "step": 6063
+    },
+    {
+      "epoch": 2.0696245733788396,
+      "grad_norm": 3.261565685272217,
+      "learning_rate": 0.00031012514220705344,
+      "loss": 5.8875,
+      "step": 6064
+    },
+    {
+      "epoch": 2.069965870307167,
+      "grad_norm": 2.8483152389526367,
+      "learning_rate": 0.0003100113765642776,
+      "loss": 6.4443,
+      "step": 6065
+    },
+    {
+      "epoch": 2.070307167235495,
+      "grad_norm": 2.9961583614349365,
+      "learning_rate": 0.0003098976109215017,
+      "loss": 5.5248,
+      "step": 6066
+    },
+    {
+      "epoch": 2.0706484641638223,
+      "grad_norm": 2.947376251220703,
+      "learning_rate": 0.0003097838452787258,
+      "loss": 5.3489,
+      "step": 6067
+    },
+    {
+      "epoch": 2.07098976109215,
+      "grad_norm": 2.8761508464813232,
+      "learning_rate": 0.00030967007963594997,
+      "loss": 5.4662,
+      "step": 6068
+    },
+    {
+      "epoch": 2.071331058020478,
+      "grad_norm": 2.8893637657165527,
+      "learning_rate": 0.0003095563139931741,
+      "loss": 5.9248,
+      "step": 6069
+    },
+    {
+      "epoch": 2.0716723549488054,
+      "grad_norm": 2.9129505157470703,
+      "learning_rate": 0.0003094425483503982,
+      "loss": 6.376,
+      "step": 6070
+    },
+    {
+      "epoch": 2.0720136518771333,
+      "grad_norm": 4.613974094390869,
+      "learning_rate": 0.00030932878270762234,
+      "loss": 5.5792,
+      "step": 6071
+    },
+    {
+      "epoch": 2.0723549488054607,
+      "grad_norm": 2.938722610473633,
+      "learning_rate": 0.00030921501706484644,
+      "loss": 6.8475,
+      "step": 6072
+    },
+    {
+      "epoch": 2.0726962457337885,
+      "grad_norm": 2.8759310245513916,
+      "learning_rate": 0.0003091012514220705,
+      "loss": 6.688,
+      "step": 6073
+    },
+    {
+      "epoch": 2.073037542662116,
+      "grad_norm": 2.9209375381469727,
+      "learning_rate": 0.00030898748577929465,
+      "loss": 6.3908,
+      "step": 6074
+    },
+    {
+      "epoch": 2.073378839590444,
+      "grad_norm": 2.9436604976654053,
+      "learning_rate": 0.00030887372013651876,
+      "loss": 6.0059,
+      "step": 6075
+    },
+    {
+      "epoch": 2.073720136518771,
+      "grad_norm": 3.093762159347534,
+      "learning_rate": 0.00030875995449374286,
+      "loss": 5.5346,
+      "step": 6076
+    },
+    {
+      "epoch": 2.074061433447099,
+      "grad_norm": 2.913107395172119,
+      "learning_rate": 0.000308646188850967,
+      "loss": 6.3079,
+      "step": 6077
+    },
+    {
+      "epoch": 2.0744027303754264,
+      "grad_norm": 2.8740787506103516,
+      "learning_rate": 0.00030853242320819113,
+      "loss": 5.9121,
+      "step": 6078
+    },
+    {
+      "epoch": 2.0747440273037543,
+      "grad_norm": 3.232553243637085,
+      "learning_rate": 0.0003084186575654153,
+      "loss": 5.6589,
+      "step": 6079
+    },
+    {
+      "epoch": 2.0750853242320817,
+      "grad_norm": 2.818389654159546,
+      "learning_rate": 0.0003083048919226394,
+      "loss": 6.1773,
+      "step": 6080
+    },
+    {
+      "epoch": 2.0754266211604095,
+      "grad_norm": 2.880063533782959,
+      "learning_rate": 0.0003081911262798635,
+      "loss": 6.005,
+      "step": 6081
+    },
+    {
+      "epoch": 2.0757679180887374,
+      "grad_norm": 3.2499842643737793,
+      "learning_rate": 0.00030807736063708766,
+      "loss": 6.1325,
+      "step": 6082
+    },
+    {
+      "epoch": 2.076109215017065,
+      "grad_norm": 2.889770746231079,
+      "learning_rate": 0.0003079635949943117,
+      "loss": 5.9898,
+      "step": 6083
+    },
+    {
+      "epoch": 2.0764505119453927,
+      "grad_norm": 2.825451612472534,
+      "learning_rate": 0.0003078498293515358,
+      "loss": 6.3039,
+      "step": 6084
+    },
+    {
+      "epoch": 2.07679180887372,
+      "grad_norm": 3.2816312313079834,
+      "learning_rate": 0.00030773606370875997,
+      "loss": 4.5501,
+      "step": 6085
+    },
+    {
+      "epoch": 2.077133105802048,
+      "grad_norm": 2.836883068084717,
+      "learning_rate": 0.0003076222980659841,
+      "loss": 6.3832,
+      "step": 6086
+    },
+    {
+      "epoch": 2.0774744027303753,
+      "grad_norm": 2.905423879623413,
+      "learning_rate": 0.0003075085324232082,
+      "loss": 6.6088,
+      "step": 6087
+    },
+    {
+      "epoch": 2.077815699658703,
+      "grad_norm": 3.0025651454925537,
+      "learning_rate": 0.00030739476678043234,
+      "loss": 6.6776,
+      "step": 6088
+    },
+    {
+      "epoch": 2.0781569965870306,
+      "grad_norm": 2.935211181640625,
+      "learning_rate": 0.00030728100113765644,
+      "loss": 5.9203,
+      "step": 6089
+    },
+    {
+      "epoch": 2.0784982935153584,
+      "grad_norm": 3.124129295349121,
+      "learning_rate": 0.00030716723549488055,
+      "loss": 5.9244,
+      "step": 6090
+    },
+    {
+      "epoch": 2.078839590443686,
+      "grad_norm": 2.805335283279419,
+      "learning_rate": 0.0003070534698521047,
+      "loss": 6.3963,
+      "step": 6091
+    },
+    {
+      "epoch": 2.0791808873720137,
+      "grad_norm": 2.806159496307373,
+      "learning_rate": 0.00030693970420932876,
+      "loss": 6.1115,
+      "step": 6092
+    },
+    {
+      "epoch": 2.079522184300341,
+      "grad_norm": 2.8614501953125,
+      "learning_rate": 0.00030682593856655286,
+      "loss": 6.4517,
+      "step": 6093
+    },
+    {
+      "epoch": 2.079863481228669,
+      "grad_norm": 2.736070394515991,
+      "learning_rate": 0.000306712172923777,
+      "loss": 6.4008,
+      "step": 6094
+    },
+    {
+      "epoch": 2.080204778156997,
+      "grad_norm": 2.861682176589966,
+      "learning_rate": 0.00030659840728100113,
+      "loss": 6.0243,
+      "step": 6095
+    },
+    {
+      "epoch": 2.080546075085324,
+      "grad_norm": 2.79305100440979,
+      "learning_rate": 0.00030648464163822523,
+      "loss": 5.9656,
+      "step": 6096
+    },
+    {
+      "epoch": 2.080887372013652,
+      "grad_norm": 2.781764030456543,
+      "learning_rate": 0.0003063708759954494,
+      "loss": 5.8456,
+      "step": 6097
+    },
+    {
+      "epoch": 2.0812286689419794,
+      "grad_norm": 3.013824224472046,
+      "learning_rate": 0.0003062571103526735,
+      "loss": 6.0931,
+      "step": 6098
+    },
+    {
+      "epoch": 2.0815699658703073,
+      "grad_norm": 2.9550583362579346,
+      "learning_rate": 0.00030614334470989766,
+      "loss": 6.5948,
+      "step": 6099
+    },
+    {
+      "epoch": 2.0819112627986347,
+      "grad_norm": 2.903958320617676,
+      "learning_rate": 0.00030602957906712176,
+      "loss": 6.3639,
+      "step": 6100
+    },
+    {
+      "epoch": 2.0822525597269625,
+      "grad_norm": 2.8269143104553223,
+      "learning_rate": 0.00030591581342434587,
+      "loss": 6.0978,
+      "step": 6101
+    },
+    {
+      "epoch": 2.08259385665529,
+      "grad_norm": 2.9017698764801025,
+      "learning_rate": 0.00030580204778156997,
+      "loss": 6.2451,
+      "step": 6102
+    },
+    {
+      "epoch": 2.082935153583618,
+      "grad_norm": 2.8492634296417236,
+      "learning_rate": 0.0003056882821387941,
+      "loss": 6.0594,
+      "step": 6103
+    },
+    {
+      "epoch": 2.083276450511945,
+      "grad_norm": 2.8176167011260986,
+      "learning_rate": 0.0003055745164960182,
+      "loss": 6.4956,
+      "step": 6104
+    },
+    {
+      "epoch": 2.083617747440273,
+      "grad_norm": 2.8858425617218018,
+      "learning_rate": 0.00030546075085324234,
+      "loss": 6.1588,
+      "step": 6105
+    },
+    {
+      "epoch": 2.0839590443686005,
+      "grad_norm": 2.932605266571045,
+      "learning_rate": 0.00030534698521046645,
+      "loss": 5.9801,
+      "step": 6106
+    },
+    {
+      "epoch": 2.0843003412969283,
+      "grad_norm": 2.9948959350585938,
+      "learning_rate": 0.00030523321956769055,
+      "loss": 6.1619,
+      "step": 6107
+    },
+    {
+      "epoch": 2.084641638225256,
+      "grad_norm": 2.8777294158935547,
+      "learning_rate": 0.0003051194539249147,
+      "loss": 5.9433,
+      "step": 6108
+    },
+    {
+      "epoch": 2.0849829351535836,
+      "grad_norm": 4.565830707550049,
+      "learning_rate": 0.0003050056882821388,
+      "loss": 5.8684,
+      "step": 6109
+    },
+    {
+      "epoch": 2.0853242320819114,
+      "grad_norm": 2.942716360092163,
+      "learning_rate": 0.0003048919226393629,
+      "loss": 6.3863,
+      "step": 6110
+    },
+    {
+      "epoch": 2.085665529010239,
+      "grad_norm": 3.1933271884918213,
+      "learning_rate": 0.0003047781569965871,
+      "loss": 4.8312,
+      "step": 6111
+    },
+    {
+      "epoch": 2.0860068259385667,
+      "grad_norm": 2.8330047130584717,
+      "learning_rate": 0.00030466439135381113,
+      "loss": 6.6125,
+      "step": 6112
+    },
+    {
+      "epoch": 2.086348122866894,
+      "grad_norm": 2.899695873260498,
+      "learning_rate": 0.00030455062571103523,
+      "loss": 5.9607,
+      "step": 6113
+    },
+    {
+      "epoch": 2.086689419795222,
+      "grad_norm": 2.945712089538574,
+      "learning_rate": 0.0003044368600682594,
+      "loss": 6.2117,
+      "step": 6114
+    },
+    {
+      "epoch": 2.0870307167235493,
+      "grad_norm": 3.072864532470703,
+      "learning_rate": 0.0003043230944254835,
+      "loss": 5.743,
+      "step": 6115
+    },
+    {
+      "epoch": 2.087372013651877,
+      "grad_norm": 2.916473865509033,
+      "learning_rate": 0.0003042093287827076,
+      "loss": 6.4488,
+      "step": 6116
+    },
+    {
+      "epoch": 2.0877133105802046,
+      "grad_norm": 2.8467938899993896,
+      "learning_rate": 0.00030409556313993176,
+      "loss": 6.1196,
+      "step": 6117
+    },
+    {
+      "epoch": 2.0880546075085324,
+      "grad_norm": 2.8041112422943115,
+      "learning_rate": 0.00030398179749715587,
+      "loss": 6.3867,
+      "step": 6118
+    },
+    {
+      "epoch": 2.08839590443686,
+      "grad_norm": 2.6718533039093018,
+      "learning_rate": 0.00030386803185438,
+      "loss": 6.1606,
+      "step": 6119
+    },
+    {
+      "epoch": 2.0887372013651877,
+      "grad_norm": 3.134387254714966,
+      "learning_rate": 0.00030375426621160413,
+      "loss": 5.3715,
+      "step": 6120
+    },
+    {
+      "epoch": 2.0890784982935156,
+      "grad_norm": 2.8396830558776855,
+      "learning_rate": 0.0003036405005688282,
+      "loss": 6.6257,
+      "step": 6121
+    },
+    {
+      "epoch": 2.089419795221843,
+      "grad_norm": 2.89739990234375,
+      "learning_rate": 0.00030352673492605234,
+      "loss": 6.4547,
+      "step": 6122
+    },
+    {
+      "epoch": 2.089761092150171,
+      "grad_norm": 2.870203971862793,
+      "learning_rate": 0.00030341296928327645,
+      "loss": 5.6783,
+      "step": 6123
+    },
+    {
+      "epoch": 2.090102389078498,
+      "grad_norm": 6.795978546142578,
+      "learning_rate": 0.00030329920364050055,
+      "loss": 4.8163,
+      "step": 6124
+    },
+    {
+      "epoch": 2.090443686006826,
+      "grad_norm": 2.864149570465088,
+      "learning_rate": 0.0003031854379977247,
+      "loss": 6.3274,
+      "step": 6125
+    },
+    {
+      "epoch": 2.0907849829351535,
+      "grad_norm": 2.8654708862304688,
+      "learning_rate": 0.0003030716723549488,
+      "loss": 6.3438,
+      "step": 6126
+    },
+    {
+      "epoch": 2.0911262798634813,
+      "grad_norm": 2.834479331970215,
+      "learning_rate": 0.0003029579067121729,
+      "loss": 6.1642,
+      "step": 6127
+    },
+    {
+      "epoch": 2.0914675767918087,
+      "grad_norm": 2.9898204803466797,
+      "learning_rate": 0.0003028441410693971,
+      "loss": 6.4757,
+      "step": 6128
+    },
+    {
+      "epoch": 2.0918088737201366,
+      "grad_norm": 2.8628549575805664,
+      "learning_rate": 0.0003027303754266212,
+      "loss": 6.5145,
+      "step": 6129
+    },
+    {
+      "epoch": 2.092150170648464,
+      "grad_norm": 2.945585012435913,
+      "learning_rate": 0.0003026166097838453,
+      "loss": 6.4056,
+      "step": 6130
+    },
+    {
+      "epoch": 2.092491467576792,
+      "grad_norm": 2.85774564743042,
+      "learning_rate": 0.0003025028441410694,
+      "loss": 5.6863,
+      "step": 6131
+    },
+    {
+      "epoch": 2.0928327645051192,
+      "grad_norm": 3.0979244709014893,
+      "learning_rate": 0.0003023890784982935,
+      "loss": 6.1059,
+      "step": 6132
+    },
+    {
+      "epoch": 2.093174061433447,
+      "grad_norm": 2.8489980697631836,
+      "learning_rate": 0.0003022753128555176,
+      "loss": 5.884,
+      "step": 6133
+    },
+    {
+      "epoch": 2.093515358361775,
+      "grad_norm": 2.7917068004608154,
+      "learning_rate": 0.00030216154721274176,
+      "loss": 5.8435,
+      "step": 6134
+    },
+    {
+      "epoch": 2.0938566552901023,
+      "grad_norm": 3.5278565883636475,
+      "learning_rate": 0.00030204778156996587,
+      "loss": 3.6562,
+      "step": 6135
+    },
+    {
+      "epoch": 2.09419795221843,
+      "grad_norm": 2.9529824256896973,
+      "learning_rate": 0.00030193401592718997,
+      "loss": 5.6781,
+      "step": 6136
+    },
+    {
+      "epoch": 2.0945392491467576,
+      "grad_norm": 7.044441223144531,
+      "learning_rate": 0.00030182025028441413,
+      "loss": 6.3167,
+      "step": 6137
+    },
+    {
+      "epoch": 2.0948805460750854,
+      "grad_norm": 3.1856296062469482,
+      "learning_rate": 0.00030170648464163824,
+      "loss": 4.6697,
+      "step": 6138
+    },
+    {
+      "epoch": 2.095221843003413,
+      "grad_norm": 2.9588096141815186,
+      "learning_rate": 0.0003015927189988624,
+      "loss": 6.2748,
+      "step": 6139
+    },
+    {
+      "epoch": 2.0955631399317407,
+      "grad_norm": 3.0708155632019043,
+      "learning_rate": 0.0003014789533560865,
+      "loss": 6.2729,
+      "step": 6140
+    },
+    {
+      "epoch": 2.095904436860068,
+      "grad_norm": 2.921039342880249,
+      "learning_rate": 0.00030136518771331055,
+      "loss": 5.9834,
+      "step": 6141
+    },
+    {
+      "epoch": 2.096245733788396,
+      "grad_norm": 2.7934975624084473,
+      "learning_rate": 0.0003012514220705347,
+      "loss": 5.9395,
+      "step": 6142
+    },
+    {
+      "epoch": 2.0965870307167234,
+      "grad_norm": 2.899808645248413,
+      "learning_rate": 0.0003011376564277588,
+      "loss": 6.1141,
+      "step": 6143
+    },
+    {
+      "epoch": 2.096928327645051,
+      "grad_norm": 2.8703773021698,
+      "learning_rate": 0.0003010238907849829,
+      "loss": 6.3106,
+      "step": 6144
+    },
+    {
+      "epoch": 2.0972696245733786,
+      "grad_norm": 2.8744289875030518,
+      "learning_rate": 0.0003009101251422071,
+      "loss": 6.4198,
+      "step": 6145
+    },
+    {
+      "epoch": 2.0976109215017065,
+      "grad_norm": 2.801576852798462,
+      "learning_rate": 0.0003007963594994312,
+      "loss": 6.0057,
+      "step": 6146
+    },
+    {
+      "epoch": 2.0979522184300343,
+      "grad_norm": 2.772566080093384,
+      "learning_rate": 0.0003006825938566553,
+      "loss": 5.799,
+      "step": 6147
+    },
+    {
+      "epoch": 2.0982935153583617,
+      "grad_norm": 2.8693549633026123,
+      "learning_rate": 0.00030056882821387945,
+      "loss": 6.134,
+      "step": 6148
+    },
+    {
+      "epoch": 2.0986348122866896,
+      "grad_norm": 1.9791274070739746,
+      "learning_rate": 0.00030045506257110355,
+      "loss": 3.122,
+      "step": 6149
+    },
+    {
+      "epoch": 2.098976109215017,
+      "grad_norm": 2.8072032928466797,
+      "learning_rate": 0.00030034129692832766,
+      "loss": 5.8741,
+      "step": 6150
+    },
+    {
+      "epoch": 2.099317406143345,
+      "grad_norm": 2.8439536094665527,
+      "learning_rate": 0.00030022753128555176,
+      "loss": 6.4318,
+      "step": 6151
+    },
+    {
+      "epoch": 2.0996587030716722,
+      "grad_norm": 2.8671748638153076,
+      "learning_rate": 0.00030011376564277587,
+      "loss": 6.1086,
+      "step": 6152
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.7963056564331055,
+      "learning_rate": 0.0003,
+      "loss": 6.0888,
+      "step": 6153
+    },
+    {
+      "epoch": 2.1003412969283275,
+      "grad_norm": 2.9660708904266357,
+      "learning_rate": 0.00029988623435722413,
+      "loss": 6.0207,
+      "step": 6154
+    },
+    {
+      "epoch": 2.1006825938566553,
+      "grad_norm": 2.9538464546203613,
+      "learning_rate": 0.00029977246871444824,
+      "loss": 6.5961,
+      "step": 6155
+    },
+    {
+      "epoch": 2.1010238907849828,
+      "grad_norm": 2.916219472885132,
+      "learning_rate": 0.00029965870307167234,
+      "loss": 6.4961,
+      "step": 6156
+    },
+    {
+      "epoch": 2.1013651877133106,
+      "grad_norm": 3.0762720108032227,
+      "learning_rate": 0.0002995449374288965,
+      "loss": 5.8819,
+      "step": 6157
+    },
+    {
+      "epoch": 2.101706484641638,
+      "grad_norm": 2.9753077030181885,
+      "learning_rate": 0.0002994311717861206,
+      "loss": 6.0732,
+      "step": 6158
+    },
+    {
+      "epoch": 2.102047781569966,
+      "grad_norm": 3.163635492324829,
+      "learning_rate": 0.00029931740614334477,
+      "loss": 5.752,
+      "step": 6159
+    },
+    {
+      "epoch": 2.1023890784982937,
+      "grad_norm": 2.8499581813812256,
+      "learning_rate": 0.0002992036405005688,
+      "loss": 6.1047,
+      "step": 6160
+    },
+    {
+      "epoch": 2.102730375426621,
+      "grad_norm": 2.870997190475464,
+      "learning_rate": 0.0002990898748577929,
+      "loss": 6.2264,
+      "step": 6161
+    },
+    {
+      "epoch": 2.103071672354949,
+      "grad_norm": 2.834073305130005,
+      "learning_rate": 0.0002989761092150171,
+      "loss": 6.223,
+      "step": 6162
+    },
+    {
+      "epoch": 2.1034129692832764,
+      "grad_norm": 2.8211171627044678,
+      "learning_rate": 0.0002988623435722412,
+      "loss": 6.8218,
+      "step": 6163
+    },
+    {
+      "epoch": 2.103754266211604,
+      "grad_norm": 2.9839437007904053,
+      "learning_rate": 0.0002987485779294653,
+      "loss": 6.177,
+      "step": 6164
+    },
+    {
+      "epoch": 2.1040955631399316,
+      "grad_norm": 2.881976366043091,
+      "learning_rate": 0.00029863481228668945,
+      "loss": 6.3719,
+      "step": 6165
+    },
+    {
+      "epoch": 2.1044368600682595,
+      "grad_norm": 2.873582601547241,
+      "learning_rate": 0.00029852104664391355,
+      "loss": 6.6528,
+      "step": 6166
+    },
+    {
+      "epoch": 2.104778156996587,
+      "grad_norm": 2.877270460128784,
+      "learning_rate": 0.00029840728100113766,
+      "loss": 6.2307,
+      "step": 6167
+    },
+    {
+      "epoch": 2.1051194539249147,
+      "grad_norm": 2.885765314102173,
+      "learning_rate": 0.0002982935153583618,
+      "loss": 6.1185,
+      "step": 6168
+    },
+    {
+      "epoch": 2.105460750853242,
+      "grad_norm": 2.835249662399292,
+      "learning_rate": 0.0002981797497155859,
+      "loss": 6.2276,
+      "step": 6169
+    },
+    {
+      "epoch": 2.10580204778157,
+      "grad_norm": 2.9260549545288086,
+      "learning_rate": 0.00029806598407281,
+      "loss": 6.3071,
+      "step": 6170
+    },
+    {
+      "epoch": 2.1061433447098974,
+      "grad_norm": 2.8237810134887695,
+      "learning_rate": 0.00029795221843003413,
+      "loss": 6.1465,
+      "step": 6171
+    },
+    {
+      "epoch": 2.1064846416382252,
+      "grad_norm": 2.822054862976074,
+      "learning_rate": 0.00029783845278725824,
+      "loss": 5.8652,
+      "step": 6172
+    },
+    {
+      "epoch": 2.106825938566553,
+      "grad_norm": 2.801043748855591,
+      "learning_rate": 0.00029772468714448234,
+      "loss": 6.0659,
+      "step": 6173
+    },
+    {
+      "epoch": 2.1071672354948805,
+      "grad_norm": 2.7532236576080322,
+      "learning_rate": 0.0002976109215017065,
+      "loss": 6.1017,
+      "step": 6174
+    },
+    {
+      "epoch": 2.1075085324232083,
+      "grad_norm": 2.8876118659973145,
+      "learning_rate": 0.0002974971558589306,
+      "loss": 5.7315,
+      "step": 6175
+    },
+    {
+      "epoch": 2.1078498293515358,
+      "grad_norm": 2.9598159790039062,
+      "learning_rate": 0.0002973833902161547,
+      "loss": 6.4794,
+      "step": 6176
+    },
+    {
+      "epoch": 2.1081911262798636,
+      "grad_norm": 2.756242513656616,
+      "learning_rate": 0.00029726962457337887,
+      "loss": 6.6465,
+      "step": 6177
+    },
+    {
+      "epoch": 2.108532423208191,
+      "grad_norm": 3.125593662261963,
+      "learning_rate": 0.000297155858930603,
+      "loss": 6.2238,
+      "step": 6178
+    },
+    {
+      "epoch": 2.108873720136519,
+      "grad_norm": 2.907468557357788,
+      "learning_rate": 0.0002970420932878271,
+      "loss": 6.4868,
+      "step": 6179
+    },
+    {
+      "epoch": 2.1092150170648463,
+      "grad_norm": 4.4180755615234375,
+      "learning_rate": 0.0002969283276450512,
+      "loss": 5.2227,
+      "step": 6180
+    },
+    {
+      "epoch": 2.109556313993174,
+      "grad_norm": 2.873162269592285,
+      "learning_rate": 0.0002968145620022753,
+      "loss": 6.7799,
+      "step": 6181
+    },
+    {
+      "epoch": 2.1098976109215015,
+      "grad_norm": 2.8836138248443604,
+      "learning_rate": 0.00029670079635949945,
+      "loss": 5.8275,
+      "step": 6182
+    },
+    {
+      "epoch": 2.1102389078498294,
+      "grad_norm": 2.93564510345459,
+      "learning_rate": 0.00029658703071672356,
+      "loss": 5.3932,
+      "step": 6183
+    },
+    {
+      "epoch": 2.1105802047781568,
+      "grad_norm": 2.858703374862671,
+      "learning_rate": 0.00029647326507394766,
+      "loss": 6.2049,
+      "step": 6184
+    },
+    {
+      "epoch": 2.1109215017064846,
+      "grad_norm": 2.9091365337371826,
+      "learning_rate": 0.0002963594994311718,
+      "loss": 6.3892,
+      "step": 6185
+    },
+    {
+      "epoch": 2.1112627986348125,
+      "grad_norm": 2.96952486038208,
+      "learning_rate": 0.0002962457337883959,
+      "loss": 5.8229,
+      "step": 6186
+    },
+    {
+      "epoch": 2.11160409556314,
+      "grad_norm": 2.865198850631714,
+      "learning_rate": 0.00029613196814562003,
+      "loss": 6.4365,
+      "step": 6187
+    },
+    {
+      "epoch": 2.1119453924914677,
+      "grad_norm": 2.7699646949768066,
+      "learning_rate": 0.0002960182025028442,
+      "loss": 6.4912,
+      "step": 6188
+    },
+    {
+      "epoch": 2.112286689419795,
+      "grad_norm": 2.8436672687530518,
+      "learning_rate": 0.00029590443686006824,
+      "loss": 5.2912,
+      "step": 6189
+    },
+    {
+      "epoch": 2.112627986348123,
+      "grad_norm": 2.8417892456054688,
+      "learning_rate": 0.00029579067121729234,
+      "loss": 6.3945,
+      "step": 6190
+    },
+    {
+      "epoch": 2.1129692832764504,
+      "grad_norm": 6.457067966461182,
+      "learning_rate": 0.0002956769055745165,
+      "loss": 3.8569,
+      "step": 6191
+    },
+    {
+      "epoch": 2.1133105802047782,
+      "grad_norm": 4.360135555267334,
+      "learning_rate": 0.0002955631399317406,
+      "loss": 5.5622,
+      "step": 6192
+    },
+    {
+      "epoch": 2.1136518771331056,
+      "grad_norm": 3.0096099376678467,
+      "learning_rate": 0.0002954493742889647,
+      "loss": 5.8473,
+      "step": 6193
+    },
+    {
+      "epoch": 2.1139931740614335,
+      "grad_norm": 3.391237497329712,
+      "learning_rate": 0.00029533560864618887,
+      "loss": 5.4066,
+      "step": 6194
+    },
+    {
+      "epoch": 2.114334470989761,
+      "grad_norm": 2.9052274227142334,
+      "learning_rate": 0.000295221843003413,
+      "loss": 5.9522,
+      "step": 6195
+    },
+    {
+      "epoch": 2.1146757679180888,
+      "grad_norm": 2.8499536514282227,
+      "learning_rate": 0.0002951080773606371,
+      "loss": 5.3388,
+      "step": 6196
+    },
+    {
+      "epoch": 2.115017064846416,
+      "grad_norm": 2.8930346965789795,
+      "learning_rate": 0.00029499431171786124,
+      "loss": 6.1808,
+      "step": 6197
+    },
+    {
+      "epoch": 2.115358361774744,
+      "grad_norm": 8.893372535705566,
+      "learning_rate": 0.00029488054607508535,
+      "loss": 5.8041,
+      "step": 6198
+    },
+    {
+      "epoch": 2.115699658703072,
+      "grad_norm": 2.8673787117004395,
+      "learning_rate": 0.0002947667804323094,
+      "loss": 6.1321,
+      "step": 6199
+    },
+    {
+      "epoch": 2.1160409556313993,
+      "grad_norm": 2.806966543197632,
+      "learning_rate": 0.00029465301478953356,
+      "loss": 6.6198,
+      "step": 6200
+    },
+    {
+      "epoch": 2.116382252559727,
+      "grad_norm": 2.9241514205932617,
+      "learning_rate": 0.00029453924914675766,
+      "loss": 6.4543,
+      "step": 6201
+    },
+    {
+      "epoch": 2.1167235494880545,
+      "grad_norm": 2.939549684524536,
+      "learning_rate": 0.0002944254835039818,
+      "loss": 5.8664,
+      "step": 6202
+    },
+    {
+      "epoch": 2.1170648464163824,
+      "grad_norm": 2.817613124847412,
+      "learning_rate": 0.0002943117178612059,
+      "loss": 6.4852,
+      "step": 6203
+    },
+    {
+      "epoch": 2.11740614334471,
+      "grad_norm": 2.9055352210998535,
+      "learning_rate": 0.00029419795221843003,
+      "loss": 6.1691,
+      "step": 6204
+    },
+    {
+      "epoch": 2.1177474402730376,
+      "grad_norm": 2.756702184677124,
+      "learning_rate": 0.0002940841865756542,
+      "loss": 6.3137,
+      "step": 6205
+    },
+    {
+      "epoch": 2.118088737201365,
+      "grad_norm": 2.957275152206421,
+      "learning_rate": 0.0002939704209328783,
+      "loss": 6.5439,
+      "step": 6206
+    },
+    {
+      "epoch": 2.118430034129693,
+      "grad_norm": 2.8155031204223633,
+      "learning_rate": 0.0002938566552901024,
+      "loss": 6.0326,
+      "step": 6207
+    },
+    {
+      "epoch": 2.1187713310580203,
+      "grad_norm": 12.575434684753418,
+      "learning_rate": 0.00029374288964732656,
+      "loss": 4.9286,
+      "step": 6208
+    },
+    {
+      "epoch": 2.119112627986348,
+      "grad_norm": 5.587793350219727,
+      "learning_rate": 0.0002936291240045506,
+      "loss": 4.9998,
+      "step": 6209
+    },
+    {
+      "epoch": 2.1194539249146755,
+      "grad_norm": 3.0160374641418457,
+      "learning_rate": 0.0002935153583617747,
+      "loss": 5.9745,
+      "step": 6210
+    },
+    {
+      "epoch": 2.1197952218430034,
+      "grad_norm": 2.93703293800354,
+      "learning_rate": 0.0002934015927189989,
+      "loss": 6.4411,
+      "step": 6211
+    },
+    {
+      "epoch": 2.1201365187713312,
+      "grad_norm": 2.9282970428466797,
+      "learning_rate": 0.000293287827076223,
+      "loss": 6.1064,
+      "step": 6212
+    },
+    {
+      "epoch": 2.1204778156996587,
+      "grad_norm": 2.965182304382324,
+      "learning_rate": 0.0002931740614334471,
+      "loss": 6.5645,
+      "step": 6213
+    },
+    {
+      "epoch": 2.1208191126279865,
+      "grad_norm": 2.8125925064086914,
+      "learning_rate": 0.00029306029579067124,
+      "loss": 6.1774,
+      "step": 6214
+    },
+    {
+      "epoch": 2.121160409556314,
+      "grad_norm": 2.8314590454101562,
+      "learning_rate": 0.00029294653014789535,
+      "loss": 5.887,
+      "step": 6215
+    },
+    {
+      "epoch": 2.1215017064846418,
+      "grad_norm": 2.745131492614746,
+      "learning_rate": 0.00029283276450511945,
+      "loss": 6.5378,
+      "step": 6216
+    },
+    {
+      "epoch": 2.121843003412969,
+      "grad_norm": 2.8095803260803223,
+      "learning_rate": 0.0002927189988623436,
+      "loss": 6.3276,
+      "step": 6217
+    },
+    {
+      "epoch": 2.122184300341297,
+      "grad_norm": 2.9379096031188965,
+      "learning_rate": 0.0002926052332195677,
+      "loss": 6.4891,
+      "step": 6218
+    },
+    {
+      "epoch": 2.1225255972696244,
+      "grad_norm": 2.851707696914673,
+      "learning_rate": 0.00029249146757679177,
+      "loss": 6.2868,
+      "step": 6219
+    },
+    {
+      "epoch": 2.1228668941979523,
+      "grad_norm": 3.480109214782715,
+      "learning_rate": 0.0002923777019340159,
+      "loss": 5.4423,
+      "step": 6220
+    },
+    {
+      "epoch": 2.1232081911262797,
+      "grad_norm": 3.0105674266815186,
+      "learning_rate": 0.00029226393629124003,
+      "loss": 5.9214,
+      "step": 6221
+    },
+    {
+      "epoch": 2.1235494880546075,
+      "grad_norm": 2.8100273609161377,
+      "learning_rate": 0.0002921501706484642,
+      "loss": 6.3837,
+      "step": 6222
+    },
+    {
+      "epoch": 2.123890784982935,
+      "grad_norm": 2.886622190475464,
+      "learning_rate": 0.0002920364050056883,
+      "loss": 5.8483,
+      "step": 6223
+    },
+    {
+      "epoch": 2.124232081911263,
+      "grad_norm": 2.9184138774871826,
+      "learning_rate": 0.0002919226393629124,
+      "loss": 6.2556,
+      "step": 6224
+    },
+    {
+      "epoch": 2.1245733788395906,
+      "grad_norm": 3.3458776473999023,
+      "learning_rate": 0.00029180887372013656,
+      "loss": 4.7381,
+      "step": 6225
+    },
+    {
+      "epoch": 2.124914675767918,
+      "grad_norm": 2.833589553833008,
+      "learning_rate": 0.00029169510807736066,
+      "loss": 6.5108,
+      "step": 6226
+    },
+    {
+      "epoch": 2.125255972696246,
+      "grad_norm": 2.8403658866882324,
+      "learning_rate": 0.00029158134243458477,
+      "loss": 6.0265,
+      "step": 6227
+    },
+    {
+      "epoch": 2.1255972696245733,
+      "grad_norm": 2.9905974864959717,
+      "learning_rate": 0.0002914675767918089,
+      "loss": 5.9997,
+      "step": 6228
+    },
+    {
+      "epoch": 2.125938566552901,
+      "grad_norm": 3.724816083908081,
+      "learning_rate": 0.000291353811149033,
+      "loss": 5.0568,
+      "step": 6229
+    },
+    {
+      "epoch": 2.1262798634812285,
+      "grad_norm": 2.8078196048736572,
+      "learning_rate": 0.0002912400455062571,
+      "loss": 6.1329,
+      "step": 6230
+    },
+    {
+      "epoch": 2.1266211604095564,
+      "grad_norm": 2.8691749572753906,
+      "learning_rate": 0.00029112627986348124,
+      "loss": 6.2066,
+      "step": 6231
+    },
+    {
+      "epoch": 2.126962457337884,
+      "grad_norm": 2.830688953399658,
+      "learning_rate": 0.00029101251422070535,
+      "loss": 6.0778,
+      "step": 6232
+    },
+    {
+      "epoch": 2.1273037542662117,
+      "grad_norm": 2.865365505218506,
+      "learning_rate": 0.00029089874857792945,
+      "loss": 6.1605,
+      "step": 6233
+    },
+    {
+      "epoch": 2.127645051194539,
+      "grad_norm": 2.6853408813476562,
+      "learning_rate": 0.0002907849829351536,
+      "loss": 5.9814,
+      "step": 6234
+    },
+    {
+      "epoch": 2.127986348122867,
+      "grad_norm": 2.7538394927978516,
+      "learning_rate": 0.0002906712172923777,
+      "loss": 5.2563,
+      "step": 6235
+    },
+    {
+      "epoch": 2.1283276450511943,
+      "grad_norm": 3.2029075622558594,
+      "learning_rate": 0.0002905574516496018,
+      "loss": 5.537,
+      "step": 6236
+    },
+    {
+      "epoch": 2.128668941979522,
+      "grad_norm": 2.8655052185058594,
+      "learning_rate": 0.000290443686006826,
+      "loss": 6.3069,
+      "step": 6237
+    },
+    {
+      "epoch": 2.12901023890785,
+      "grad_norm": 2.836099624633789,
+      "learning_rate": 0.00029032992036405003,
+      "loss": 6.1336,
+      "step": 6238
+    },
+    {
+      "epoch": 2.1293515358361774,
+      "grad_norm": 2.8297688961029053,
+      "learning_rate": 0.00029021615472127414,
+      "loss": 5.9636,
+      "step": 6239
+    },
+    {
+      "epoch": 2.1296928327645053,
+      "grad_norm": 2.8413031101226807,
+      "learning_rate": 0.0002901023890784983,
+      "loss": 6.4108,
+      "step": 6240
+    },
+    {
+      "epoch": 2.1300341296928327,
+      "grad_norm": 2.8722383975982666,
+      "learning_rate": 0.0002899886234357224,
+      "loss": 6.5657,
+      "step": 6241
+    },
+    {
+      "epoch": 2.1303754266211605,
+      "grad_norm": 2.8211677074432373,
+      "learning_rate": 0.00028987485779294656,
+      "loss": 6.3421,
+      "step": 6242
+    },
+    {
+      "epoch": 2.130716723549488,
+      "grad_norm": 3.3281304836273193,
+      "learning_rate": 0.00028976109215017067,
+      "loss": 5.4648,
+      "step": 6243
+    },
+    {
+      "epoch": 2.131058020477816,
+      "grad_norm": 2.8957037925720215,
+      "learning_rate": 0.00028964732650739477,
+      "loss": 6.5093,
+      "step": 6244
+    },
+    {
+      "epoch": 2.131399317406143,
+      "grad_norm": 2.93902587890625,
+      "learning_rate": 0.00028953356086461893,
+      "loss": 6.1025,
+      "step": 6245
+    },
+    {
+      "epoch": 2.131740614334471,
+      "grad_norm": 2.877804756164551,
+      "learning_rate": 0.00028941979522184303,
+      "loss": 5.8682,
+      "step": 6246
+    },
+    {
+      "epoch": 2.1320819112627984,
+      "grad_norm": 2.8391692638397217,
+      "learning_rate": 0.00028930602957906714,
+      "loss": 6.1784,
+      "step": 6247
+    },
+    {
+      "epoch": 2.1324232081911263,
+      "grad_norm": 2.83154559135437,
+      "learning_rate": 0.00028919226393629124,
+      "loss": 6.0821,
+      "step": 6248
+    },
+    {
+      "epoch": 2.1327645051194537,
+      "grad_norm": 2.8575069904327393,
+      "learning_rate": 0.00028907849829351535,
+      "loss": 6.1532,
+      "step": 6249
+    },
+    {
+      "epoch": 2.1331058020477816,
+      "grad_norm": 2.8776323795318604,
+      "learning_rate": 0.00028896473265073945,
+      "loss": 6.1705,
+      "step": 6250
+    },
+    {
+      "epoch": 2.1334470989761094,
+      "grad_norm": 2.809744119644165,
+      "learning_rate": 0.0002888509670079636,
+      "loss": 5.6912,
+      "step": 6251
+    },
+    {
+      "epoch": 2.133788395904437,
+      "grad_norm": 2.8350303173065186,
+      "learning_rate": 0.0002887372013651877,
+      "loss": 6.6106,
+      "step": 6252
+    },
+    {
+      "epoch": 2.1341296928327647,
+      "grad_norm": 2.8744351863861084,
+      "learning_rate": 0.0002886234357224118,
+      "loss": 5.9893,
+      "step": 6253
+    },
+    {
+      "epoch": 2.134470989761092,
+      "grad_norm": 2.87178897857666,
+      "learning_rate": 0.000288509670079636,
+      "loss": 5.9063,
+      "step": 6254
+    },
+    {
+      "epoch": 2.13481228668942,
+      "grad_norm": 2.9190359115600586,
+      "learning_rate": 0.0002883959044368601,
+      "loss": 5.8424,
+      "step": 6255
+    },
+    {
+      "epoch": 2.1351535836177473,
+      "grad_norm": 2.8117921352386475,
+      "learning_rate": 0.0002882821387940842,
+      "loss": 6.3904,
+      "step": 6256
+    },
+    {
+      "epoch": 2.135494880546075,
+      "grad_norm": 2.7705960273742676,
+      "learning_rate": 0.00028816837315130835,
+      "loss": 6.1901,
+      "step": 6257
+    },
+    {
+      "epoch": 2.1358361774744026,
+      "grad_norm": 2.7972941398620605,
+      "learning_rate": 0.0002880546075085324,
+      "loss": 6.4878,
+      "step": 6258
+    },
+    {
+      "epoch": 2.1361774744027304,
+      "grad_norm": 2.915985345840454,
+      "learning_rate": 0.0002879408418657565,
+      "loss": 6.319,
+      "step": 6259
+    },
+    {
+      "epoch": 2.136518771331058,
+      "grad_norm": 3.0406363010406494,
+      "learning_rate": 0.00028782707622298067,
+      "loss": 5.3564,
+      "step": 6260
+    },
+    {
+      "epoch": 2.1368600682593857,
+      "grad_norm": 2.915806531906128,
+      "learning_rate": 0.00028771331058020477,
+      "loss": 6.8073,
+      "step": 6261
+    },
+    {
+      "epoch": 2.137201365187713,
+      "grad_norm": 2.840094804763794,
+      "learning_rate": 0.0002875995449374289,
+      "loss": 6.0616,
+      "step": 6262
+    },
+    {
+      "epoch": 2.137542662116041,
+      "grad_norm": 3.044546365737915,
+      "learning_rate": 0.00028748577929465304,
+      "loss": 5.9823,
+      "step": 6263
+    },
+    {
+      "epoch": 2.137883959044369,
+      "grad_norm": 2.8069345951080322,
+      "learning_rate": 0.00028737201365187714,
+      "loss": 5.8899,
+      "step": 6264
+    },
+    {
+      "epoch": 2.138225255972696,
+      "grad_norm": 2.8434646129608154,
+      "learning_rate": 0.0002872582480091013,
+      "loss": 6.471,
+      "step": 6265
+    },
+    {
+      "epoch": 2.138566552901024,
+      "grad_norm": 2.9248239994049072,
+      "learning_rate": 0.0002871444823663254,
+      "loss": 5.8759,
+      "step": 6266
+    },
+    {
+      "epoch": 2.1389078498293514,
+      "grad_norm": 2.838510036468506,
+      "learning_rate": 0.00028703071672354946,
+      "loss": 6.3492,
+      "step": 6267
+    },
+    {
+      "epoch": 2.1392491467576793,
+      "grad_norm": 2.9336483478546143,
+      "learning_rate": 0.0002869169510807736,
+      "loss": 6.0126,
+      "step": 6268
+    },
+    {
+      "epoch": 2.1395904436860067,
+      "grad_norm": 2.8124420642852783,
+      "learning_rate": 0.0002868031854379977,
+      "loss": 5.8763,
+      "step": 6269
+    },
+    {
+      "epoch": 2.1399317406143346,
+      "grad_norm": 2.821249485015869,
+      "learning_rate": 0.0002866894197952218,
+      "loss": 6.0869,
+      "step": 6270
+    },
+    {
+      "epoch": 2.140273037542662,
+      "grad_norm": 2.825366258621216,
+      "learning_rate": 0.000286575654152446,
+      "loss": 6.1345,
+      "step": 6271
+    },
+    {
+      "epoch": 2.14061433447099,
+      "grad_norm": 2.799079418182373,
+      "learning_rate": 0.0002864618885096701,
+      "loss": 5.919,
+      "step": 6272
+    },
+    {
+      "epoch": 2.140955631399317,
+      "grad_norm": 2.9337072372436523,
+      "learning_rate": 0.0002863481228668942,
+      "loss": 5.927,
+      "step": 6273
+    },
+    {
+      "epoch": 2.141296928327645,
+      "grad_norm": 2.8965418338775635,
+      "learning_rate": 0.00028623435722411835,
+      "loss": 6.1504,
+      "step": 6274
+    },
+    {
+      "epoch": 2.1416382252559725,
+      "grad_norm": 3.0523440837860107,
+      "learning_rate": 0.00028612059158134246,
+      "loss": 6.0062,
+      "step": 6275
+    },
+    {
+      "epoch": 2.1419795221843003,
+      "grad_norm": 2.986406087875366,
+      "learning_rate": 0.00028600682593856656,
+      "loss": 5.8823,
+      "step": 6276
+    },
+    {
+      "epoch": 2.142320819112628,
+      "grad_norm": 2.9808123111724854,
+      "learning_rate": 0.00028589306029579067,
+      "loss": 6.4025,
+      "step": 6277
+    },
+    {
+      "epoch": 2.1426621160409556,
+      "grad_norm": 2.6804819107055664,
+      "learning_rate": 0.00028577929465301477,
+      "loss": 5.7797,
+      "step": 6278
+    },
+    {
+      "epoch": 2.1430034129692834,
+      "grad_norm": 3.432440996170044,
+      "learning_rate": 0.0002856655290102389,
+      "loss": 4.9929,
+      "step": 6279
+    },
+    {
+      "epoch": 2.143344709897611,
+      "grad_norm": 3.000901222229004,
+      "learning_rate": 0.00028555176336746304,
+      "loss": 6.2861,
+      "step": 6280
+    },
+    {
+      "epoch": 2.1436860068259387,
+      "grad_norm": 2.8772642612457275,
+      "learning_rate": 0.00028543799772468714,
+      "loss": 6.4075,
+      "step": 6281
+    },
+    {
+      "epoch": 2.144027303754266,
+      "grad_norm": 2.7788705825805664,
+      "learning_rate": 0.00028532423208191125,
+      "loss": 4.92,
+      "step": 6282
+    },
+    {
+      "epoch": 2.144368600682594,
+      "grad_norm": 3.084559917449951,
+      "learning_rate": 0.0002852104664391354,
+      "loss": 5.9546,
+      "step": 6283
+    },
+    {
+      "epoch": 2.1447098976109213,
+      "grad_norm": 3.004912853240967,
+      "learning_rate": 0.0002850967007963595,
+      "loss": 6.2623,
+      "step": 6284
+    },
+    {
+      "epoch": 2.145051194539249,
+      "grad_norm": 4.325676441192627,
+      "learning_rate": 0.00028498293515358367,
+      "loss": 4.1026,
+      "step": 6285
+    },
+    {
+      "epoch": 2.1453924914675766,
+      "grad_norm": 2.822035551071167,
+      "learning_rate": 0.0002848691695108078,
+      "loss": 6.1718,
+      "step": 6286
+    },
+    {
+      "epoch": 2.1457337883959045,
+      "grad_norm": 3.587118148803711,
+      "learning_rate": 0.0002847554038680318,
+      "loss": 4.5978,
+      "step": 6287
+    },
+    {
+      "epoch": 2.146075085324232,
+      "grad_norm": 3.637235403060913,
+      "learning_rate": 0.000284641638225256,
+      "loss": 5.1437,
+      "step": 6288
+    },
+    {
+      "epoch": 2.1464163822525597,
+      "grad_norm": 3.1141090393066406,
+      "learning_rate": 0.0002845278725824801,
+      "loss": 4.2409,
+      "step": 6289
+    },
+    {
+      "epoch": 2.1467576791808876,
+      "grad_norm": 2.9125001430511475,
+      "learning_rate": 0.0002844141069397042,
+      "loss": 6.4584,
+      "step": 6290
+    },
+    {
+      "epoch": 2.147098976109215,
+      "grad_norm": 2.8799238204956055,
+      "learning_rate": 0.00028430034129692835,
+      "loss": 6.1453,
+      "step": 6291
+    },
+    {
+      "epoch": 2.147440273037543,
+      "grad_norm": 2.945450782775879,
+      "learning_rate": 0.00028418657565415246,
+      "loss": 5.9845,
+      "step": 6292
+    },
+    {
+      "epoch": 2.14778156996587,
+      "grad_norm": 2.812429428100586,
+      "learning_rate": 0.00028407281001137656,
+      "loss": 5.7118,
+      "step": 6293
+    },
+    {
+      "epoch": 2.148122866894198,
+      "grad_norm": 2.8083250522613525,
+      "learning_rate": 0.0002839590443686007,
+      "loss": 6.1992,
+      "step": 6294
+    },
+    {
+      "epoch": 2.1484641638225255,
+      "grad_norm": 2.8323144912719727,
+      "learning_rate": 0.00028384527872582483,
+      "loss": 6.9573,
+      "step": 6295
+    },
+    {
+      "epoch": 2.1488054607508533,
+      "grad_norm": 2.777940273284912,
+      "learning_rate": 0.0002837315130830489,
+      "loss": 5.7647,
+      "step": 6296
+    },
+    {
+      "epoch": 2.1491467576791807,
+      "grad_norm": 2.9180915355682373,
+      "learning_rate": 0.00028361774744027304,
+      "loss": 6.115,
+      "step": 6297
+    },
+    {
+      "epoch": 2.1494880546075086,
+      "grad_norm": 2.861795663833618,
+      "learning_rate": 0.00028350398179749714,
+      "loss": 5.9888,
+      "step": 6298
+    },
+    {
+      "epoch": 2.149829351535836,
+      "grad_norm": 2.804222822189331,
+      "learning_rate": 0.00028339021615472125,
+      "loss": 6.1902,
+      "step": 6299
+    },
+    {
+      "epoch": 2.150170648464164,
+      "grad_norm": 2.753507375717163,
+      "learning_rate": 0.0002832764505119454,
+      "loss": 6.6358,
+      "step": 6300
+    },
+    {
+      "epoch": 2.1505119453924912,
+      "grad_norm": 2.8499131202697754,
+      "learning_rate": 0.0002831626848691695,
+      "loss": 6.6611,
+      "step": 6301
+    },
+    {
+      "epoch": 2.150853242320819,
+      "grad_norm": 2.9446632862091064,
+      "learning_rate": 0.0002830489192263936,
+      "loss": 6.9951,
+      "step": 6302
+    },
+    {
+      "epoch": 2.151194539249147,
+      "grad_norm": 2.8921797275543213,
+      "learning_rate": 0.0002829351535836178,
+      "loss": 6.4307,
+      "step": 6303
+    },
+    {
+      "epoch": 2.1515358361774743,
+      "grad_norm": 2.776242971420288,
+      "learning_rate": 0.0002828213879408419,
+      "loss": 6.5389,
+      "step": 6304
+    },
+    {
+      "epoch": 2.151877133105802,
+      "grad_norm": 3.0470118522644043,
+      "learning_rate": 0.00028270762229806604,
+      "loss": 5.9021,
+      "step": 6305
+    },
+    {
+      "epoch": 2.1522184300341296,
+      "grad_norm": 2.870765209197998,
+      "learning_rate": 0.0002825938566552901,
+      "loss": 5.767,
+      "step": 6306
+    },
+    {
+      "epoch": 2.1525597269624575,
+      "grad_norm": 2.9112579822540283,
+      "learning_rate": 0.0002824800910125142,
+      "loss": 5.9856,
+      "step": 6307
+    },
+    {
+      "epoch": 2.152901023890785,
+      "grad_norm": 2.8619847297668457,
+      "learning_rate": 0.00028236632536973835,
+      "loss": 6.447,
+      "step": 6308
+    },
+    {
+      "epoch": 2.1532423208191127,
+      "grad_norm": 2.7564167976379395,
+      "learning_rate": 0.00028225255972696246,
+      "loss": 6.2936,
+      "step": 6309
+    },
+    {
+      "epoch": 2.15358361774744,
+      "grad_norm": 2.8190064430236816,
+      "learning_rate": 0.00028213879408418656,
+      "loss": 6.035,
+      "step": 6310
+    },
+    {
+      "epoch": 2.153924914675768,
+      "grad_norm": 2.4524943828582764,
+      "learning_rate": 0.0002820250284414107,
+      "loss": 3.2047,
+      "step": 6311
+    },
+    {
+      "epoch": 2.1542662116040954,
+      "grad_norm": 2.8662970066070557,
+      "learning_rate": 0.00028191126279863483,
+      "loss": 4.9013,
+      "step": 6312
+    },
+    {
+      "epoch": 2.154607508532423,
+      "grad_norm": 2.9255385398864746,
+      "learning_rate": 0.00028179749715585893,
+      "loss": 6.6828,
+      "step": 6313
+    },
+    {
+      "epoch": 2.1549488054607506,
+      "grad_norm": 2.917696475982666,
+      "learning_rate": 0.0002816837315130831,
+      "loss": 4.9657,
+      "step": 6314
+    },
+    {
+      "epoch": 2.1552901023890785,
+      "grad_norm": 2.835268259048462,
+      "learning_rate": 0.0002815699658703072,
+      "loss": 6.6023,
+      "step": 6315
+    },
+    {
+      "epoch": 2.1556313993174063,
+      "grad_norm": 2.958221435546875,
+      "learning_rate": 0.00028145620022753125,
+      "loss": 5.9026,
+      "step": 6316
+    },
+    {
+      "epoch": 2.1559726962457337,
+      "grad_norm": 2.8288214206695557,
+      "learning_rate": 0.0002813424345847554,
+      "loss": 5.8854,
+      "step": 6317
+    },
+    {
+      "epoch": 2.1563139931740616,
+      "grad_norm": 3.5853185653686523,
+      "learning_rate": 0.0002812286689419795,
+      "loss": 5.0054,
+      "step": 6318
+    },
+    {
+      "epoch": 2.156655290102389,
+      "grad_norm": 2.9272825717926025,
+      "learning_rate": 0.0002811149032992036,
+      "loss": 5.7341,
+      "step": 6319
+    },
+    {
+      "epoch": 2.156996587030717,
+      "grad_norm": 2.8925740718841553,
+      "learning_rate": 0.0002810011376564278,
+      "loss": 6.3701,
+      "step": 6320
+    },
+    {
+      "epoch": 2.1573378839590442,
+      "grad_norm": 2.798673152923584,
+      "learning_rate": 0.0002808873720136519,
+      "loss": 5.85,
+      "step": 6321
+    },
+    {
+      "epoch": 2.157679180887372,
+      "grad_norm": 2.7680532932281494,
+      "learning_rate": 0.000280773606370876,
+      "loss": 6.048,
+      "step": 6322
+    },
+    {
+      "epoch": 2.1580204778156995,
+      "grad_norm": 2.8404247760772705,
+      "learning_rate": 0.00028065984072810015,
+      "loss": 5.9817,
+      "step": 6323
+    },
+    {
+      "epoch": 2.1583617747440274,
+      "grad_norm": 3.323213815689087,
+      "learning_rate": 0.00028054607508532425,
+      "loss": 4.1193,
+      "step": 6324
+    },
+    {
+      "epoch": 2.1587030716723548,
+      "grad_norm": 3.4093360900878906,
+      "learning_rate": 0.0002804323094425484,
+      "loss": 4.3087,
+      "step": 6325
+    },
+    {
+      "epoch": 2.1590443686006826,
+      "grad_norm": 2.8730340003967285,
+      "learning_rate": 0.00028031854379977246,
+      "loss": 6.0826,
+      "step": 6326
+    },
+    {
+      "epoch": 2.15938566552901,
+      "grad_norm": 2.8777995109558105,
+      "learning_rate": 0.00028020477815699656,
+      "loss": 6.4141,
+      "step": 6327
+    },
+    {
+      "epoch": 2.159726962457338,
+      "grad_norm": 2.9867589473724365,
+      "learning_rate": 0.0002800910125142207,
+      "loss": 6.3958,
+      "step": 6328
+    },
+    {
+      "epoch": 2.1600682593856657,
+      "grad_norm": 2.889984369277954,
+      "learning_rate": 0.00027997724687144483,
+      "loss": 5.8593,
+      "step": 6329
+    },
+    {
+      "epoch": 2.160409556313993,
+      "grad_norm": 2.9526164531707764,
+      "learning_rate": 0.00027986348122866893,
+      "loss": 5.9088,
+      "step": 6330
+    },
+    {
+      "epoch": 2.160750853242321,
+      "grad_norm": 4.757111072540283,
+      "learning_rate": 0.0002797497155858931,
+      "loss": 5.8323,
+      "step": 6331
+    },
+    {
+      "epoch": 2.1610921501706484,
+      "grad_norm": 2.8701725006103516,
+      "learning_rate": 0.0002796359499431172,
+      "loss": 6.536,
+      "step": 6332
+    },
+    {
+      "epoch": 2.1614334470989762,
+      "grad_norm": 2.827101230621338,
+      "learning_rate": 0.0002795221843003413,
+      "loss": 5.6708,
+      "step": 6333
+    },
+    {
+      "epoch": 2.1617747440273036,
+      "grad_norm": 3.0027883052825928,
+      "learning_rate": 0.00027940841865756546,
+      "loss": 5.5514,
+      "step": 6334
+    },
+    {
+      "epoch": 2.1621160409556315,
+      "grad_norm": 3.244901418685913,
+      "learning_rate": 0.0002792946530147895,
+      "loss": 5.0491,
+      "step": 6335
+    },
+    {
+      "epoch": 2.162457337883959,
+      "grad_norm": 2.8494021892547607,
+      "learning_rate": 0.0002791808873720136,
+      "loss": 6.3514,
+      "step": 6336
+    },
+    {
+      "epoch": 2.1627986348122867,
+      "grad_norm": 2.9061434268951416,
+      "learning_rate": 0.0002790671217292378,
+      "loss": 6.184,
+      "step": 6337
+    },
+    {
+      "epoch": 2.163139931740614,
+      "grad_norm": 2.848724126815796,
+      "learning_rate": 0.0002789533560864619,
+      "loss": 6.5265,
+      "step": 6338
+    },
+    {
+      "epoch": 2.163481228668942,
+      "grad_norm": 3.203709602355957,
+      "learning_rate": 0.000278839590443686,
+      "loss": 6.0744,
+      "step": 6339
+    },
+    {
+      "epoch": 2.1638225255972694,
+      "grad_norm": 2.862330198287964,
+      "learning_rate": 0.00027872582480091015,
+      "loss": 6.3895,
+      "step": 6340
+    },
+    {
+      "epoch": 2.1641638225255972,
+      "grad_norm": 4.710589408874512,
+      "learning_rate": 0.00027861205915813425,
+      "loss": 4.9895,
+      "step": 6341
+    },
+    {
+      "epoch": 2.164505119453925,
+      "grad_norm": 2.8820667266845703,
+      "learning_rate": 0.00027849829351535836,
+      "loss": 5.7645,
+      "step": 6342
+    },
+    {
+      "epoch": 2.1648464163822525,
+      "grad_norm": 3.1783809661865234,
+      "learning_rate": 0.0002783845278725825,
+      "loss": 5.7725,
+      "step": 6343
+    },
+    {
+      "epoch": 2.1651877133105804,
+      "grad_norm": 2.8277227878570557,
+      "learning_rate": 0.0002782707622298066,
+      "loss": 6.6817,
+      "step": 6344
+    },
+    {
+      "epoch": 2.1655290102389078,
+      "grad_norm": 2.84521222114563,
+      "learning_rate": 0.00027815699658703067,
+      "loss": 5.903,
+      "step": 6345
+    },
+    {
+      "epoch": 2.1658703071672356,
+      "grad_norm": 2.908289909362793,
+      "learning_rate": 0.00027804323094425483,
+      "loss": 5.6247,
+      "step": 6346
+    },
+    {
+      "epoch": 2.166211604095563,
+      "grad_norm": 3.011632204055786,
+      "learning_rate": 0.00027792946530147893,
+      "loss": 6.108,
+      "step": 6347
+    },
+    {
+      "epoch": 2.166552901023891,
+      "grad_norm": 2.8673269748687744,
+      "learning_rate": 0.0002778156996587031,
+      "loss": 6.2876,
+      "step": 6348
+    },
+    {
+      "epoch": 2.1668941979522183,
+      "grad_norm": 2.8036434650421143,
+      "learning_rate": 0.0002777019340159272,
+      "loss": 6.3993,
+      "step": 6349
+    },
+    {
+      "epoch": 2.167235494880546,
+      "grad_norm": 2.8587241172790527,
+      "learning_rate": 0.0002775881683731513,
+      "loss": 6.0677,
+      "step": 6350
+    },
+    {
+      "epoch": 2.1675767918088735,
+      "grad_norm": 3.036465883255005,
+      "learning_rate": 0.00027747440273037546,
+      "loss": 5.7799,
+      "step": 6351
+    },
+    {
+      "epoch": 2.1679180887372014,
+      "grad_norm": 2.8174219131469727,
+      "learning_rate": 0.00027736063708759957,
+      "loss": 5.7702,
+      "step": 6352
+    },
+    {
+      "epoch": 2.168259385665529,
+      "grad_norm": 2.7940962314605713,
+      "learning_rate": 0.00027724687144482367,
+      "loss": 6.4102,
+      "step": 6353
+    },
+    {
+      "epoch": 2.1686006825938566,
+      "grad_norm": 4.701676368713379,
+      "learning_rate": 0.00027713310580204783,
+      "loss": 4.1398,
+      "step": 6354
+    },
+    {
+      "epoch": 2.1689419795221845,
+      "grad_norm": 2.993300676345825,
+      "learning_rate": 0.0002770193401592719,
+      "loss": 6.4199,
+      "step": 6355
+    },
+    {
+      "epoch": 2.169283276450512,
+      "grad_norm": 2.90317440032959,
+      "learning_rate": 0.000276905574516496,
+      "loss": 6.436,
+      "step": 6356
+    },
+    {
+      "epoch": 2.1696245733788397,
+      "grad_norm": 2.9535815715789795,
+      "learning_rate": 0.00027679180887372015,
+      "loss": 6.5237,
+      "step": 6357
+    },
+    {
+      "epoch": 2.169965870307167,
+      "grad_norm": 2.869474172592163,
+      "learning_rate": 0.00027667804323094425,
+      "loss": 6.0359,
+      "step": 6358
+    },
+    {
+      "epoch": 2.170307167235495,
+      "grad_norm": 2.888397693634033,
+      "learning_rate": 0.00027656427758816836,
+      "loss": 6.4275,
+      "step": 6359
+    },
+    {
+      "epoch": 2.1706484641638224,
+      "grad_norm": 2.8712522983551025,
+      "learning_rate": 0.0002764505119453925,
+      "loss": 6.1997,
+      "step": 6360
+    },
+    {
+      "epoch": 2.1709897610921502,
+      "grad_norm": 2.934269905090332,
+      "learning_rate": 0.0002763367463026166,
+      "loss": 6.4548,
+      "step": 6361
+    },
+    {
+      "epoch": 2.1713310580204777,
+      "grad_norm": 2.752089262008667,
+      "learning_rate": 0.0002762229806598407,
+      "loss": 6.0606,
+      "step": 6362
+    },
+    {
+      "epoch": 2.1716723549488055,
+      "grad_norm": 3.460238456726074,
+      "learning_rate": 0.0002761092150170649,
+      "loss": 4.7184,
+      "step": 6363
+    },
+    {
+      "epoch": 2.172013651877133,
+      "grad_norm": 2.8735015392303467,
+      "learning_rate": 0.00027599544937428894,
+      "loss": 5.8198,
+      "step": 6364
+    },
+    {
+      "epoch": 2.1723549488054608,
+      "grad_norm": 2.812506675720215,
+      "learning_rate": 0.00027588168373151304,
+      "loss": 5.6997,
+      "step": 6365
+    },
+    {
+      "epoch": 2.172696245733788,
+      "grad_norm": 2.769024133682251,
+      "learning_rate": 0.0002757679180887372,
+      "loss": 6.1244,
+      "step": 6366
+    },
+    {
+      "epoch": 2.173037542662116,
+      "grad_norm": 2.8082642555236816,
+      "learning_rate": 0.0002756541524459613,
+      "loss": 6.0761,
+      "step": 6367
+    },
+    {
+      "epoch": 2.173378839590444,
+      "grad_norm": 2.7583067417144775,
+      "learning_rate": 0.00027554038680318546,
+      "loss": 5.7482,
+      "step": 6368
+    },
+    {
+      "epoch": 2.1737201365187713,
+      "grad_norm": 2.982508659362793,
+      "learning_rate": 0.00027542662116040957,
+      "loss": 5.0801,
+      "step": 6369
+    },
+    {
+      "epoch": 2.174061433447099,
+      "grad_norm": 2.7821574211120605,
+      "learning_rate": 0.0002753128555176337,
+      "loss": 6.4252,
+      "step": 6370
+    },
+    {
+      "epoch": 2.1744027303754265,
+      "grad_norm": 2.783104419708252,
+      "learning_rate": 0.00027519908987485783,
+      "loss": 5.8358,
+      "step": 6371
+    },
+    {
+      "epoch": 2.1747440273037544,
+      "grad_norm": 2.8212077617645264,
+      "learning_rate": 0.00027508532423208194,
+      "loss": 6.0909,
+      "step": 6372
+    },
+    {
+      "epoch": 2.175085324232082,
+      "grad_norm": 2.8396553993225098,
+      "learning_rate": 0.00027497155858930604,
+      "loss": 6.2717,
+      "step": 6373
+    },
+    {
+      "epoch": 2.1754266211604096,
+      "grad_norm": 4.277924060821533,
+      "learning_rate": 0.00027485779294653015,
+      "loss": 5.4965,
+      "step": 6374
+    },
+    {
+      "epoch": 2.175767918088737,
+      "grad_norm": 2.8166744709014893,
+      "learning_rate": 0.00027474402730375425,
+      "loss": 5.7125,
+      "step": 6375
+    },
+    {
+      "epoch": 2.176109215017065,
+      "grad_norm": 2.8981945514678955,
+      "learning_rate": 0.00027463026166097836,
+      "loss": 5.2261,
+      "step": 6376
+    },
+    {
+      "epoch": 2.1764505119453923,
+      "grad_norm": 2.9114558696746826,
+      "learning_rate": 0.0002745164960182025,
+      "loss": 6.0895,
+      "step": 6377
+    },
+    {
+      "epoch": 2.17679180887372,
+      "grad_norm": 3.33648681640625,
+      "learning_rate": 0.0002744027303754266,
+      "loss": 5.1431,
+      "step": 6378
+    },
+    {
+      "epoch": 2.1771331058020476,
+      "grad_norm": 2.8457000255584717,
+      "learning_rate": 0.0002742889647326507,
+      "loss": 6.706,
+      "step": 6379
+    },
+    {
+      "epoch": 2.1774744027303754,
+      "grad_norm": 2.829103946685791,
+      "learning_rate": 0.0002741751990898749,
+      "loss": 6.4637,
+      "step": 6380
+    },
+    {
+      "epoch": 2.1778156996587033,
+      "grad_norm": 2.912069797515869,
+      "learning_rate": 0.000274061433447099,
+      "loss": 6.7109,
+      "step": 6381
+    },
+    {
+      "epoch": 2.1781569965870307,
+      "grad_norm": 2.960214853286743,
+      "learning_rate": 0.0002739476678043231,
+      "loss": 5.9655,
+      "step": 6382
+    },
+    {
+      "epoch": 2.1784982935153585,
+      "grad_norm": 3.0913033485412598,
+      "learning_rate": 0.00027383390216154725,
+      "loss": 5.843,
+      "step": 6383
+    },
+    {
+      "epoch": 2.178839590443686,
+      "grad_norm": 2.850743293762207,
+      "learning_rate": 0.0002737201365187713,
+      "loss": 4.9693,
+      "step": 6384
+    },
+    {
+      "epoch": 2.1791808873720138,
+      "grad_norm": 2.808077335357666,
+      "learning_rate": 0.0002736063708759954,
+      "loss": 6.0217,
+      "step": 6385
+    },
+    {
+      "epoch": 2.179522184300341,
+      "grad_norm": 2.9761393070220947,
+      "learning_rate": 0.00027349260523321957,
+      "loss": 4.5135,
+      "step": 6386
+    },
+    {
+      "epoch": 2.179863481228669,
+      "grad_norm": 2.9906089305877686,
+      "learning_rate": 0.0002733788395904437,
+      "loss": 6.1202,
+      "step": 6387
+    },
+    {
+      "epoch": 2.1802047781569964,
+      "grad_norm": 3.4929113388061523,
+      "learning_rate": 0.00027326507394766783,
+      "loss": 5.8859,
+      "step": 6388
+    },
+    {
+      "epoch": 2.1805460750853243,
+      "grad_norm": 2.846465826034546,
+      "learning_rate": 0.00027315130830489194,
+      "loss": 6.0745,
+      "step": 6389
+    },
+    {
+      "epoch": 2.1808873720136517,
+      "grad_norm": 2.907808780670166,
+      "learning_rate": 0.00027303754266211604,
+      "loss": 5.983,
+      "step": 6390
+    },
+    {
+      "epoch": 2.1812286689419795,
+      "grad_norm": 2.858454942703247,
+      "learning_rate": 0.0002729237770193402,
+      "loss": 6.0968,
+      "step": 6391
+    },
+    {
+      "epoch": 2.181569965870307,
+      "grad_norm": 2.9495015144348145,
+      "learning_rate": 0.0002728100113765643,
+      "loss": 6.0371,
+      "step": 6392
+    },
+    {
+      "epoch": 2.181911262798635,
+      "grad_norm": 3.2573885917663574,
+      "learning_rate": 0.0002726962457337884,
+      "loss": 5.8426,
+      "step": 6393
+    },
+    {
+      "epoch": 2.1822525597269626,
+      "grad_norm": 4.468471527099609,
+      "learning_rate": 0.0002725824800910125,
+      "loss": 5.1957,
+      "step": 6394
+    },
+    {
+      "epoch": 2.18259385665529,
+      "grad_norm": 2.860956907272339,
+      "learning_rate": 0.0002724687144482366,
+      "loss": 5.7918,
+      "step": 6395
+    },
+    {
+      "epoch": 2.182935153583618,
+      "grad_norm": 2.8932464122772217,
+      "learning_rate": 0.00027235494880546073,
+      "loss": 6.7351,
+      "step": 6396
+    },
+    {
+      "epoch": 2.1832764505119453,
+      "grad_norm": 3.92891001701355,
+      "learning_rate": 0.0002722411831626849,
+      "loss": 4.5889,
+      "step": 6397
+    },
+    {
+      "epoch": 2.183617747440273,
+      "grad_norm": 1.9834781885147095,
+      "learning_rate": 0.000272127417519909,
+      "loss": 3.1828,
+      "step": 6398
+    },
+    {
+      "epoch": 2.1839590443686006,
+      "grad_norm": 2.90181827545166,
+      "learning_rate": 0.0002720136518771331,
+      "loss": 6.1316,
+      "step": 6399
+    },
+    {
+      "epoch": 2.1843003412969284,
+      "grad_norm": 3.912259817123413,
+      "learning_rate": 0.00027189988623435726,
+      "loss": 5.1916,
+      "step": 6400
+    },
+    {
+      "epoch": 2.184641638225256,
+      "grad_norm": 2.8160667419433594,
+      "learning_rate": 0.00027178612059158136,
+      "loss": 5.7501,
+      "step": 6401
+    },
+    {
+      "epoch": 2.1849829351535837,
+      "grad_norm": 2.872741222381592,
+      "learning_rate": 0.00027167235494880547,
+      "loss": 6.3647,
+      "step": 6402
+    },
+    {
+      "epoch": 2.185324232081911,
+      "grad_norm": 3.7568233013153076,
+      "learning_rate": 0.00027155858930602957,
+      "loss": 5.5898,
+      "step": 6403
+    },
+    {
+      "epoch": 2.185665529010239,
+      "grad_norm": 2.84272837638855,
+      "learning_rate": 0.0002714448236632537,
+      "loss": 5.3566,
+      "step": 6404
+    },
+    {
+      "epoch": 2.1860068259385663,
+      "grad_norm": 2.7461118698120117,
+      "learning_rate": 0.0002713310580204778,
+      "loss": 5.7219,
+      "step": 6405
+    },
+    {
+      "epoch": 2.186348122866894,
+      "grad_norm": 2.8305881023406982,
+      "learning_rate": 0.00027121729237770194,
+      "loss": 6.4276,
+      "step": 6406
+    },
+    {
+      "epoch": 2.186689419795222,
+      "grad_norm": 2.7954049110412598,
+      "learning_rate": 0.00027110352673492604,
+      "loss": 6.167,
+      "step": 6407
+    },
+    {
+      "epoch": 2.1870307167235494,
+      "grad_norm": 2.9570295810699463,
+      "learning_rate": 0.0002709897610921502,
+      "loss": 5.7226,
+      "step": 6408
+    },
+    {
+      "epoch": 2.1873720136518773,
+      "grad_norm": 2.7967288494110107,
+      "learning_rate": 0.0002708759954493743,
+      "loss": 6.2319,
+      "step": 6409
+    },
+    {
+      "epoch": 2.1877133105802047,
+      "grad_norm": 6.396007061004639,
+      "learning_rate": 0.0002707622298065984,
+      "loss": 5.3139,
+      "step": 6410
+    },
+    {
+      "epoch": 2.1880546075085325,
+      "grad_norm": 2.839141607284546,
+      "learning_rate": 0.00027064846416382257,
+      "loss": 6.323,
+      "step": 6411
+    },
+    {
+      "epoch": 2.18839590443686,
+      "grad_norm": 2.803288459777832,
+      "learning_rate": 0.0002705346985210467,
+      "loss": 6.2097,
+      "step": 6412
+    },
+    {
+      "epoch": 2.188737201365188,
+      "grad_norm": 2.8960049152374268,
+      "learning_rate": 0.00027042093287827073,
+      "loss": 5.969,
+      "step": 6413
+    },
+    {
+      "epoch": 2.189078498293515,
+      "grad_norm": 2.7831318378448486,
+      "learning_rate": 0.0002703071672354949,
+      "loss": 6.3809,
+      "step": 6414
+    },
+    {
+      "epoch": 2.189419795221843,
+      "grad_norm": 2.8870432376861572,
+      "learning_rate": 0.000270193401592719,
+      "loss": 6.6029,
+      "step": 6415
+    },
+    {
+      "epoch": 2.1897610921501705,
+      "grad_norm": 2.961520195007324,
+      "learning_rate": 0.0002700796359499431,
+      "loss": 6.0527,
+      "step": 6416
+    },
+    {
+      "epoch": 2.1901023890784983,
+      "grad_norm": 3.088958978652954,
+      "learning_rate": 0.00026996587030716726,
+      "loss": 5.7707,
+      "step": 6417
+    },
+    {
+      "epoch": 2.1904436860068257,
+      "grad_norm": 2.74513840675354,
+      "learning_rate": 0.00026985210466439136,
+      "loss": 5.923,
+      "step": 6418
+    },
+    {
+      "epoch": 2.1907849829351536,
+      "grad_norm": 3.138380765914917,
+      "learning_rate": 0.00026973833902161547,
+      "loss": 5.148,
+      "step": 6419
+    },
+    {
+      "epoch": 2.1911262798634814,
+      "grad_norm": 3.0224714279174805,
+      "learning_rate": 0.0002696245733788396,
+      "loss": 6.7062,
+      "step": 6420
+    },
+    {
+      "epoch": 2.191467576791809,
+      "grad_norm": 2.8007893562316895,
+      "learning_rate": 0.00026951080773606373,
+      "loss": 6.122,
+      "step": 6421
+    },
+    {
+      "epoch": 2.1918088737201367,
+      "grad_norm": 2.802536725997925,
+      "learning_rate": 0.00026939704209328784,
+      "loss": 6.1242,
+      "step": 6422
+    },
+    {
+      "epoch": 2.192150170648464,
+      "grad_norm": 2.7604641914367676,
+      "learning_rate": 0.00026928327645051194,
+      "loss": 5.7296,
+      "step": 6423
+    },
+    {
+      "epoch": 2.192491467576792,
+      "grad_norm": 3.111950159072876,
+      "learning_rate": 0.00026916951080773605,
+      "loss": 5.6642,
+      "step": 6424
+    },
+    {
+      "epoch": 2.1928327645051193,
+      "grad_norm": 2.8447318077087402,
+      "learning_rate": 0.00026905574516496015,
+      "loss": 5.7298,
+      "step": 6425
+    },
+    {
+      "epoch": 2.193174061433447,
+      "grad_norm": 2.7198376655578613,
+      "learning_rate": 0.0002689419795221843,
+      "loss": 5.9149,
+      "step": 6426
+    },
+    {
+      "epoch": 2.1935153583617746,
+      "grad_norm": 2.7918670177459717,
+      "learning_rate": 0.0002688282138794084,
+      "loss": 6.2658,
+      "step": 6427
+    },
+    {
+      "epoch": 2.1938566552901024,
+      "grad_norm": 3.1611740589141846,
+      "learning_rate": 0.0002687144482366326,
+      "loss": 4.734,
+      "step": 6428
+    },
+    {
+      "epoch": 2.19419795221843,
+      "grad_norm": 2.8715784549713135,
+      "learning_rate": 0.0002686006825938567,
+      "loss": 6.3364,
+      "step": 6429
+    },
+    {
+      "epoch": 2.1945392491467577,
+      "grad_norm": 2.958388328552246,
+      "learning_rate": 0.0002684869169510808,
+      "loss": 6.582,
+      "step": 6430
+    },
+    {
+      "epoch": 2.194880546075085,
+      "grad_norm": 3.085158109664917,
+      "learning_rate": 0.00026837315130830494,
+      "loss": 5.759,
+      "step": 6431
+    },
+    {
+      "epoch": 2.195221843003413,
+      "grad_norm": 2.9389808177948,
+      "learning_rate": 0.00026825938566552905,
+      "loss": 5.6247,
+      "step": 6432
+    },
+    {
+      "epoch": 2.195563139931741,
+      "grad_norm": 4.009513854980469,
+      "learning_rate": 0.0002681456200227531,
+      "loss": 4.9773,
+      "step": 6433
+    },
+    {
+      "epoch": 2.195904436860068,
+      "grad_norm": 2.732940196990967,
+      "learning_rate": 0.00026803185437997726,
+      "loss": 5.8776,
+      "step": 6434
+    },
+    {
+      "epoch": 2.196245733788396,
+      "grad_norm": 2.8661234378814697,
+      "learning_rate": 0.00026791808873720136,
+      "loss": 6.5482,
+      "step": 6435
+    },
+    {
+      "epoch": 2.1965870307167235,
+      "grad_norm": 2.915010690689087,
+      "learning_rate": 0.00026780432309442547,
+      "loss": 6.4809,
+      "step": 6436
+    },
+    {
+      "epoch": 2.1969283276450513,
+      "grad_norm": 2.878875255584717,
+      "learning_rate": 0.0002676905574516496,
+      "loss": 6.1371,
+      "step": 6437
+    },
+    {
+      "epoch": 2.1972696245733787,
+      "grad_norm": 3.3660991191864014,
+      "learning_rate": 0.00026757679180887373,
+      "loss": 6.0642,
+      "step": 6438
+    },
+    {
+      "epoch": 2.1976109215017066,
+      "grad_norm": 2.9850966930389404,
+      "learning_rate": 0.00026746302616609784,
+      "loss": 5.561,
+      "step": 6439
+    },
+    {
+      "epoch": 2.197952218430034,
+      "grad_norm": 2.8198862075805664,
+      "learning_rate": 0.000267349260523322,
+      "loss": 6.2445,
+      "step": 6440
+    },
+    {
+      "epoch": 2.198293515358362,
+      "grad_norm": 3.0752429962158203,
+      "learning_rate": 0.0002672354948805461,
+      "loss": 5.486,
+      "step": 6441
+    },
+    {
+      "epoch": 2.198634812286689,
+      "grad_norm": 2.9619510173797607,
+      "learning_rate": 0.00026712172923777015,
+      "loss": 5.3807,
+      "step": 6442
+    },
+    {
+      "epoch": 2.198976109215017,
+      "grad_norm": 2.8163340091705322,
+      "learning_rate": 0.0002670079635949943,
+      "loss": 6.5359,
+      "step": 6443
+    },
+    {
+      "epoch": 2.1993174061433445,
+      "grad_norm": 2.8125815391540527,
+      "learning_rate": 0.0002668941979522184,
+      "loss": 6.039,
+      "step": 6444
+    },
+    {
+      "epoch": 2.1996587030716723,
+      "grad_norm": 2.7999801635742188,
+      "learning_rate": 0.0002667804323094425,
+      "loss": 6.1154,
+      "step": 6445
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.8857717514038086,
+      "learning_rate": 0.0002666666666666667,
+      "loss": 5.9834,
+      "step": 6446
+    },
+    {
+      "epoch": 2.2003412969283276,
+      "grad_norm": 2.783299446105957,
+      "learning_rate": 0.0002665529010238908,
+      "loss": 6.3258,
+      "step": 6447
+    },
+    {
+      "epoch": 2.2006825938566554,
+      "grad_norm": 2.7714779376983643,
+      "learning_rate": 0.00026643913538111494,
+      "loss": 6.0787,
+      "step": 6448
+    },
+    {
+      "epoch": 2.201023890784983,
+      "grad_norm": 2.6970479488372803,
+      "learning_rate": 0.00026632536973833905,
+      "loss": 6.2198,
+      "step": 6449
+    },
+    {
+      "epoch": 2.2013651877133107,
+      "grad_norm": 2.753514051437378,
+      "learning_rate": 0.00026621160409556315,
+      "loss": 5.8685,
+      "step": 6450
+    },
+    {
+      "epoch": 2.201706484641638,
+      "grad_norm": 2.866901159286499,
+      "learning_rate": 0.0002660978384527873,
+      "loss": 6.4346,
+      "step": 6451
+    },
+    {
+      "epoch": 2.202047781569966,
+      "grad_norm": 2.863551139831543,
+      "learning_rate": 0.00026598407281001136,
+      "loss": 5.7247,
+      "step": 6452
+    },
+    {
+      "epoch": 2.2023890784982934,
+      "grad_norm": 2.88744854927063,
+      "learning_rate": 0.00026587030716723547,
+      "loss": 5.9772,
+      "step": 6453
+    },
+    {
+      "epoch": 2.202730375426621,
+      "grad_norm": 3.1774702072143555,
+      "learning_rate": 0.00026575654152445963,
+      "loss": 5.8586,
+      "step": 6454
+    },
+    {
+      "epoch": 2.2030716723549486,
+      "grad_norm": 2.86907958984375,
+      "learning_rate": 0.00026564277588168373,
+      "loss": 6.8104,
+      "step": 6455
+    },
+    {
+      "epoch": 2.2034129692832765,
+      "grad_norm": 3.6283483505249023,
+      "learning_rate": 0.00026552901023890784,
+      "loss": 4.1912,
+      "step": 6456
+    },
+    {
+      "epoch": 2.203754266211604,
+      "grad_norm": 2.87131404876709,
+      "learning_rate": 0.000265415244596132,
+      "loss": 6.1613,
+      "step": 6457
+    },
+    {
+      "epoch": 2.2040955631399317,
+      "grad_norm": 2.80851149559021,
+      "learning_rate": 0.0002653014789533561,
+      "loss": 5.8281,
+      "step": 6458
+    },
+    {
+      "epoch": 2.2044368600682596,
+      "grad_norm": 2.9386231899261475,
+      "learning_rate": 0.0002651877133105802,
+      "loss": 6.7991,
+      "step": 6459
+    },
+    {
+      "epoch": 2.204778156996587,
+      "grad_norm": 2.9392945766448975,
+      "learning_rate": 0.00026507394766780437,
+      "loss": 6.1992,
+      "step": 6460
+    },
+    {
+      "epoch": 2.205119453924915,
+      "grad_norm": 2.8204219341278076,
+      "learning_rate": 0.00026496018202502847,
+      "loss": 6.0136,
+      "step": 6461
+    },
+    {
+      "epoch": 2.2054607508532422,
+      "grad_norm": 2.826657295227051,
+      "learning_rate": 0.0002648464163822525,
+      "loss": 5.8623,
+      "step": 6462
+    },
+    {
+      "epoch": 2.20580204778157,
+      "grad_norm": 2.8905653953552246,
+      "learning_rate": 0.0002647326507394767,
+      "loss": 6.2452,
+      "step": 6463
+    },
+    {
+      "epoch": 2.2061433447098975,
+      "grad_norm": 2.7385172843933105,
+      "learning_rate": 0.0002646188850967008,
+      "loss": 5.9435,
+      "step": 6464
+    },
+    {
+      "epoch": 2.2064846416382253,
+      "grad_norm": 2.80544114112854,
+      "learning_rate": 0.0002645051194539249,
+      "loss": 5.9384,
+      "step": 6465
+    },
+    {
+      "epoch": 2.2068259385665527,
+      "grad_norm": 2.877896785736084,
+      "learning_rate": 0.00026439135381114905,
+      "loss": 5.8879,
+      "step": 6466
+    },
+    {
+      "epoch": 2.2071672354948806,
+      "grad_norm": 2.7671401500701904,
+      "learning_rate": 0.00026427758816837315,
+      "loss": 6.2006,
+      "step": 6467
+    },
+    {
+      "epoch": 2.207508532423208,
+      "grad_norm": 2.8723201751708984,
+      "learning_rate": 0.00026416382252559726,
+      "loss": 6.475,
+      "step": 6468
+    },
+    {
+      "epoch": 2.207849829351536,
+      "grad_norm": 2.728503704071045,
+      "learning_rate": 0.0002640500568828214,
+      "loss": 5.9604,
+      "step": 6469
+    },
+    {
+      "epoch": 2.2081911262798632,
+      "grad_norm": 2.96976900100708,
+      "learning_rate": 0.0002639362912400455,
+      "loss": 5.7367,
+      "step": 6470
+    },
+    {
+      "epoch": 2.208532423208191,
+      "grad_norm": 5.522859573364258,
+      "learning_rate": 0.00026382252559726963,
+      "loss": 4.7263,
+      "step": 6471
+    },
+    {
+      "epoch": 2.208873720136519,
+      "grad_norm": 3.003938674926758,
+      "learning_rate": 0.00026370875995449373,
+      "loss": 5.9086,
+      "step": 6472
+    },
+    {
+      "epoch": 2.2092150170648464,
+      "grad_norm": 2.7639968395233154,
+      "learning_rate": 0.00026359499431171784,
+      "loss": 5.8062,
+      "step": 6473
+    },
+    {
+      "epoch": 2.209556313993174,
+      "grad_norm": 2.8196215629577637,
+      "learning_rate": 0.000263481228668942,
+      "loss": 6.236,
+      "step": 6474
+    },
+    {
+      "epoch": 2.2098976109215016,
+      "grad_norm": 6.847387790679932,
+      "learning_rate": 0.0002633674630261661,
+      "loss": 5.3268,
+      "step": 6475
+    },
+    {
+      "epoch": 2.2102389078498295,
+      "grad_norm": 2.8416600227355957,
+      "learning_rate": 0.0002632536973833902,
+      "loss": 5.8468,
+      "step": 6476
+    },
+    {
+      "epoch": 2.210580204778157,
+      "grad_norm": 7.3218278884887695,
+      "learning_rate": 0.00026313993174061437,
+      "loss": 4.5238,
+      "step": 6477
+    },
+    {
+      "epoch": 2.2109215017064847,
+      "grad_norm": 2.832921266555786,
+      "learning_rate": 0.00026302616609783847,
+      "loss": 5.7101,
+      "step": 6478
+    },
+    {
+      "epoch": 2.211262798634812,
+      "grad_norm": 2.810177803039551,
+      "learning_rate": 0.0002629124004550626,
+      "loss": 5.9935,
+      "step": 6479
+    },
+    {
+      "epoch": 2.21160409556314,
+      "grad_norm": 2.747629404067993,
+      "learning_rate": 0.00026279863481228674,
+      "loss": 6.4255,
+      "step": 6480
+    },
+    {
+      "epoch": 2.2119453924914674,
+      "grad_norm": 2.820760726928711,
+      "learning_rate": 0.0002626848691695108,
+      "loss": 5.8886,
+      "step": 6481
+    },
+    {
+      "epoch": 2.2122866894197952,
+      "grad_norm": 3.057654857635498,
+      "learning_rate": 0.0002625711035267349,
+      "loss": 5.4938,
+      "step": 6482
+    },
+    {
+      "epoch": 2.2126279863481226,
+      "grad_norm": 3.943042755126953,
+      "learning_rate": 0.00026245733788395905,
+      "loss": 5.5298,
+      "step": 6483
+    },
+    {
+      "epoch": 2.2129692832764505,
+      "grad_norm": 2.72080135345459,
+      "learning_rate": 0.00026234357224118316,
+      "loss": 6.1378,
+      "step": 6484
+    },
+    {
+      "epoch": 2.2133105802047783,
+      "grad_norm": 2.7657735347747803,
+      "learning_rate": 0.00026222980659840726,
+      "loss": 6.1694,
+      "step": 6485
+    },
+    {
+      "epoch": 2.2136518771331057,
+      "grad_norm": 2.795632839202881,
+      "learning_rate": 0.0002621160409556314,
+      "loss": 6.1866,
+      "step": 6486
+    },
+    {
+      "epoch": 2.2139931740614336,
+      "grad_norm": 2.779759407043457,
+      "learning_rate": 0.0002620022753128555,
+      "loss": 5.7196,
+      "step": 6487
+    },
+    {
+      "epoch": 2.214334470989761,
+      "grad_norm": 2.8835349082946777,
+      "learning_rate": 0.00026188850967007963,
+      "loss": 6.2627,
+      "step": 6488
+    },
+    {
+      "epoch": 2.214675767918089,
+      "grad_norm": 5.428569316864014,
+      "learning_rate": 0.0002617747440273038,
+      "loss": 5.1553,
+      "step": 6489
+    },
+    {
+      "epoch": 2.2150170648464163,
+      "grad_norm": 2.7858505249023438,
+      "learning_rate": 0.0002616609783845279,
+      "loss": 6.2648,
+      "step": 6490
+    },
+    {
+      "epoch": 2.215358361774744,
+      "grad_norm": 2.827878475189209,
+      "learning_rate": 0.000261547212741752,
+      "loss": 6.1736,
+      "step": 6491
+    },
+    {
+      "epoch": 2.2156996587030715,
+      "grad_norm": 2.7949113845825195,
+      "learning_rate": 0.0002614334470989761,
+      "loss": 6.4478,
+      "step": 6492
+    },
+    {
+      "epoch": 2.2160409556313994,
+      "grad_norm": 3.1486802101135254,
+      "learning_rate": 0.0002613196814562002,
+      "loss": 5.8937,
+      "step": 6493
+    },
+    {
+      "epoch": 2.2163822525597268,
+      "grad_norm": 2.8336243629455566,
+      "learning_rate": 0.00026120591581342437,
+      "loss": 6.8206,
+      "step": 6494
+    },
+    {
+      "epoch": 2.2167235494880546,
+      "grad_norm": 2.8043153285980225,
+      "learning_rate": 0.00026109215017064847,
+      "loss": 6.1785,
+      "step": 6495
+    },
+    {
+      "epoch": 2.217064846416382,
+      "grad_norm": 2.822309732437134,
+      "learning_rate": 0.0002609783845278726,
+      "loss": 5.5858,
+      "step": 6496
+    },
+    {
+      "epoch": 2.21740614334471,
+      "grad_norm": 3.1378068923950195,
+      "learning_rate": 0.00026086461888509674,
+      "loss": 5.7572,
+      "step": 6497
+    },
+    {
+      "epoch": 2.2177474402730377,
+      "grad_norm": 2.852703332901001,
+      "learning_rate": 0.00026075085324232084,
+      "loss": 6.8021,
+      "step": 6498
+    },
+    {
+      "epoch": 2.218088737201365,
+      "grad_norm": 2.8655879497528076,
+      "learning_rate": 0.00026063708759954495,
+      "loss": 5.816,
+      "step": 6499
+    },
+    {
+      "epoch": 2.218430034129693,
+      "grad_norm": 2.730862855911255,
+      "learning_rate": 0.0002605233219567691,
+      "loss": 6.2995,
+      "step": 6500
+    },
+    {
+      "epoch": 2.2187713310580204,
+      "grad_norm": 2.8955562114715576,
+      "learning_rate": 0.00026040955631399316,
+      "loss": 6.3926,
+      "step": 6501
+    },
+    {
+      "epoch": 2.2191126279863482,
+      "grad_norm": 2.799837350845337,
+      "learning_rate": 0.00026029579067121726,
+      "loss": 6.1799,
+      "step": 6502
+    },
+    {
+      "epoch": 2.2194539249146756,
+      "grad_norm": 2.7385647296905518,
+      "learning_rate": 0.0002601820250284414,
+      "loss": 6.1926,
+      "step": 6503
+    },
+    {
+      "epoch": 2.2197952218430035,
+      "grad_norm": 2.808077096939087,
+      "learning_rate": 0.0002600682593856655,
+      "loss": 6.1108,
+      "step": 6504
+    },
+    {
+      "epoch": 2.220136518771331,
+      "grad_norm": 2.8860621452331543,
+      "learning_rate": 0.00025995449374288963,
+      "loss": 6.0537,
+      "step": 6505
+    },
+    {
+      "epoch": 2.2204778156996587,
+      "grad_norm": 2.850010871887207,
+      "learning_rate": 0.0002598407281001138,
+      "loss": 6.5343,
+      "step": 6506
+    },
+    {
+      "epoch": 2.220819112627986,
+      "grad_norm": 2.805081367492676,
+      "learning_rate": 0.0002597269624573379,
+      "loss": 6.2704,
+      "step": 6507
+    },
+    {
+      "epoch": 2.221160409556314,
+      "grad_norm": 3.074279308319092,
+      "learning_rate": 0.000259613196814562,
+      "loss": 5.9484,
+      "step": 6508
+    },
+    {
+      "epoch": 2.2215017064846414,
+      "grad_norm": 2.8078973293304443,
+      "learning_rate": 0.00025949943117178616,
+      "loss": 5.8503,
+      "step": 6509
+    },
+    {
+      "epoch": 2.2218430034129693,
+      "grad_norm": 2.807574987411499,
+      "learning_rate": 0.0002593856655290102,
+      "loss": 5.7379,
+      "step": 6510
+    },
+    {
+      "epoch": 2.222184300341297,
+      "grad_norm": 2.841683864593506,
+      "learning_rate": 0.00025927189988623437,
+      "loss": 6.2235,
+      "step": 6511
+    },
+    {
+      "epoch": 2.2225255972696245,
+      "grad_norm": 2.873955726623535,
+      "learning_rate": 0.0002591581342434585,
+      "loss": 5.6057,
+      "step": 6512
+    },
+    {
+      "epoch": 2.2228668941979524,
+      "grad_norm": 2.908351421356201,
+      "learning_rate": 0.0002590443686006826,
+      "loss": 5.904,
+      "step": 6513
+    },
+    {
+      "epoch": 2.2232081911262798,
+      "grad_norm": 7.525314807891846,
+      "learning_rate": 0.00025893060295790674,
+      "loss": 5.3918,
+      "step": 6514
+    },
+    {
+      "epoch": 2.2235494880546076,
+      "grad_norm": 2.931612253189087,
+      "learning_rate": 0.00025881683731513084,
+      "loss": 5.8895,
+      "step": 6515
+    },
+    {
+      "epoch": 2.223890784982935,
+      "grad_norm": 2.975277900695801,
+      "learning_rate": 0.00025870307167235495,
+      "loss": 6.1299,
+      "step": 6516
+    },
+    {
+      "epoch": 2.224232081911263,
+      "grad_norm": 2.8416590690612793,
+      "learning_rate": 0.0002585893060295791,
+      "loss": 6.2545,
+      "step": 6517
+    },
+    {
+      "epoch": 2.2245733788395903,
+      "grad_norm": 2.888201951980591,
+      "learning_rate": 0.0002584755403868032,
+      "loss": 6.193,
+      "step": 6518
+    },
+    {
+      "epoch": 2.224914675767918,
+      "grad_norm": 2.7630908489227295,
+      "learning_rate": 0.0002583617747440273,
+      "loss": 5.9981,
+      "step": 6519
+    },
+    {
+      "epoch": 2.2252559726962455,
+      "grad_norm": 2.9234695434570312,
+      "learning_rate": 0.0002582480091012514,
+      "loss": 6.2954,
+      "step": 6520
+    },
+    {
+      "epoch": 2.2255972696245734,
+      "grad_norm": 3.6878268718719482,
+      "learning_rate": 0.0002581342434584755,
+      "loss": 5.1989,
+      "step": 6521
+    },
+    {
+      "epoch": 2.225938566552901,
+      "grad_norm": 2.8937203884124756,
+      "learning_rate": 0.00025802047781569963,
+      "loss": 6.15,
+      "step": 6522
+    },
+    {
+      "epoch": 2.2262798634812286,
+      "grad_norm": 2.8810667991638184,
+      "learning_rate": 0.0002579067121729238,
+      "loss": 6.2176,
+      "step": 6523
+    },
+    {
+      "epoch": 2.2266211604095565,
+      "grad_norm": 3.0311129093170166,
+      "learning_rate": 0.0002577929465301479,
+      "loss": 5.6466,
+      "step": 6524
+    },
+    {
+      "epoch": 2.226962457337884,
+      "grad_norm": 3.263695001602173,
+      "learning_rate": 0.000257679180887372,
+      "loss": 4.1231,
+      "step": 6525
+    },
+    {
+      "epoch": 2.2273037542662117,
+      "grad_norm": 2.8339765071868896,
+      "learning_rate": 0.00025756541524459616,
+      "loss": 6.2549,
+      "step": 6526
+    },
+    {
+      "epoch": 2.227645051194539,
+      "grad_norm": 2.7960495948791504,
+      "learning_rate": 0.00025745164960182026,
+      "loss": 6.3462,
+      "step": 6527
+    },
+    {
+      "epoch": 2.227986348122867,
+      "grad_norm": 2.9964165687561035,
+      "learning_rate": 0.00025733788395904437,
+      "loss": 5.2995,
+      "step": 6528
+    },
+    {
+      "epoch": 2.2283276450511944,
+      "grad_norm": 2.8184449672698975,
+      "learning_rate": 0.00025722411831626853,
+      "loss": 6.5219,
+      "step": 6529
+    },
+    {
+      "epoch": 2.2286689419795223,
+      "grad_norm": 2.7724874019622803,
+      "learning_rate": 0.0002571103526734926,
+      "loss": 6.0435,
+      "step": 6530
+    },
+    {
+      "epoch": 2.2290102389078497,
+      "grad_norm": 3.961784601211548,
+      "learning_rate": 0.0002569965870307167,
+      "loss": 5.7937,
+      "step": 6531
+    },
+    {
+      "epoch": 2.2293515358361775,
+      "grad_norm": 2.875619888305664,
+      "learning_rate": 0.00025688282138794084,
+      "loss": 6.6192,
+      "step": 6532
+    },
+    {
+      "epoch": 2.229692832764505,
+      "grad_norm": 2.870389699935913,
+      "learning_rate": 0.00025676905574516495,
+      "loss": 6.4559,
+      "step": 6533
+    },
+    {
+      "epoch": 2.2300341296928328,
+      "grad_norm": 3.843813896179199,
+      "learning_rate": 0.0002566552901023891,
+      "loss": 4.74,
+      "step": 6534
+    },
+    {
+      "epoch": 2.2303754266211606,
+      "grad_norm": 3.107849597930908,
+      "learning_rate": 0.0002565415244596132,
+      "loss": 5.6221,
+      "step": 6535
+    },
+    {
+      "epoch": 2.230716723549488,
+      "grad_norm": 2.7651352882385254,
+      "learning_rate": 0.0002564277588168373,
+      "loss": 6.2125,
+      "step": 6536
+    },
+    {
+      "epoch": 2.231058020477816,
+      "grad_norm": 2.9481201171875,
+      "learning_rate": 0.0002563139931740615,
+      "loss": 5.7552,
+      "step": 6537
+    },
+    {
+      "epoch": 2.2313993174061433,
+      "grad_norm": 2.9435575008392334,
+      "learning_rate": 0.0002562002275312856,
+      "loss": 6.1869,
+      "step": 6538
+    },
+    {
+      "epoch": 2.231740614334471,
+      "grad_norm": 2.826810359954834,
+      "learning_rate": 0.0002560864618885097,
+      "loss": 6.1379,
+      "step": 6539
+    },
+    {
+      "epoch": 2.2320819112627985,
+      "grad_norm": 2.7598679065704346,
+      "learning_rate": 0.0002559726962457338,
+      "loss": 6.0253,
+      "step": 6540
+    },
+    {
+      "epoch": 2.2324232081911264,
+      "grad_norm": 2.83957576751709,
+      "learning_rate": 0.0002558589306029579,
+      "loss": 5.9333,
+      "step": 6541
+    },
+    {
+      "epoch": 2.232764505119454,
+      "grad_norm": 3.761160135269165,
+      "learning_rate": 0.000255745164960182,
+      "loss": 5.4008,
+      "step": 6542
+    },
+    {
+      "epoch": 2.2331058020477816,
+      "grad_norm": 2.7840142250061035,
+      "learning_rate": 0.00025563139931740616,
+      "loss": 6.4198,
+      "step": 6543
+    },
+    {
+      "epoch": 2.233447098976109,
+      "grad_norm": 2.7931652069091797,
+      "learning_rate": 0.00025551763367463026,
+      "loss": 6.5151,
+      "step": 6544
+    },
+    {
+      "epoch": 2.233788395904437,
+      "grad_norm": 2.733041286468506,
+      "learning_rate": 0.00025540386803185437,
+      "loss": 6.168,
+      "step": 6545
+    },
+    {
+      "epoch": 2.2341296928327643,
+      "grad_norm": 3.0885894298553467,
+      "learning_rate": 0.00025529010238907853,
+      "loss": 5.4425,
+      "step": 6546
+    },
+    {
+      "epoch": 2.234470989761092,
+      "grad_norm": 2.77445912361145,
+      "learning_rate": 0.00025517633674630263,
+      "loss": 6.0673,
+      "step": 6547
+    },
+    {
+      "epoch": 2.23481228668942,
+      "grad_norm": 2.836225986480713,
+      "learning_rate": 0.00025506257110352674,
+      "loss": 6.3828,
+      "step": 6548
+    },
+    {
+      "epoch": 2.2351535836177474,
+      "grad_norm": 2.814570903778076,
+      "learning_rate": 0.00025494880546075084,
+      "loss": 6.2877,
+      "step": 6549
+    },
+    {
+      "epoch": 2.2354948805460753,
+      "grad_norm": 2.966297149658203,
+      "learning_rate": 0.00025483503981797495,
+      "loss": 6.2828,
+      "step": 6550
+    },
+    {
+      "epoch": 2.2358361774744027,
+      "grad_norm": 2.8446590900421143,
+      "learning_rate": 0.00025472127417519905,
+      "loss": 5.9261,
+      "step": 6551
+    },
+    {
+      "epoch": 2.2361774744027305,
+      "grad_norm": 2.9468929767608643,
+      "learning_rate": 0.0002546075085324232,
+      "loss": 6.1299,
+      "step": 6552
+    },
+    {
+      "epoch": 2.236518771331058,
+      "grad_norm": 2.959243059158325,
+      "learning_rate": 0.0002544937428896473,
+      "loss": 5.9142,
+      "step": 6553
+    },
+    {
+      "epoch": 2.2368600682593858,
+      "grad_norm": 2.908616781234741,
+      "learning_rate": 0.0002543799772468715,
+      "loss": 5.9464,
+      "step": 6554
+    },
+    {
+      "epoch": 2.237201365187713,
+      "grad_norm": 2.857999563217163,
+      "learning_rate": 0.0002542662116040956,
+      "loss": 6.5652,
+      "step": 6555
+    },
+    {
+      "epoch": 2.237542662116041,
+      "grad_norm": 2.829493522644043,
+      "learning_rate": 0.0002541524459613197,
+      "loss": 5.7772,
+      "step": 6556
+    },
+    {
+      "epoch": 2.2378839590443684,
+      "grad_norm": 2.8293933868408203,
+      "learning_rate": 0.00025403868031854385,
+      "loss": 6.0328,
+      "step": 6557
+    },
+    {
+      "epoch": 2.2382252559726963,
+      "grad_norm": 2.949575662612915,
+      "learning_rate": 0.00025392491467576795,
+      "loss": 6.5024,
+      "step": 6558
+    },
+    {
+      "epoch": 2.2385665529010237,
+      "grad_norm": 3.105191946029663,
+      "learning_rate": 0.000253811149032992,
+      "loss": 6.3816,
+      "step": 6559
+    },
+    {
+      "epoch": 2.2389078498293515,
+      "grad_norm": 2.836264133453369,
+      "learning_rate": 0.00025369738339021616,
+      "loss": 6.4266,
+      "step": 6560
+    },
+    {
+      "epoch": 2.2392491467576794,
+      "grad_norm": 2.8078081607818604,
+      "learning_rate": 0.00025358361774744027,
+      "loss": 6.5147,
+      "step": 6561
+    },
+    {
+      "epoch": 2.239590443686007,
+      "grad_norm": 2.8312695026397705,
+      "learning_rate": 0.00025346985210466437,
+      "loss": 6.6833,
+      "step": 6562
+    },
+    {
+      "epoch": 2.2399317406143346,
+      "grad_norm": 3.007437229156494,
+      "learning_rate": 0.00025335608646188853,
+      "loss": 5.3381,
+      "step": 6563
+    },
+    {
+      "epoch": 2.240273037542662,
+      "grad_norm": 3.071810722351074,
+      "learning_rate": 0.00025324232081911263,
+      "loss": 5.3215,
+      "step": 6564
+    },
+    {
+      "epoch": 2.24061433447099,
+      "grad_norm": 2.7744386196136475,
+      "learning_rate": 0.00025312855517633674,
+      "loss": 6.1504,
+      "step": 6565
+    },
+    {
+      "epoch": 2.2409556313993173,
+      "grad_norm": 2.710651159286499,
+      "learning_rate": 0.0002530147895335609,
+      "loss": 6.2226,
+      "step": 6566
+    },
+    {
+      "epoch": 2.241296928327645,
+      "grad_norm": 2.8886449337005615,
+      "learning_rate": 0.000252901023890785,
+      "loss": 6.6402,
+      "step": 6567
+    },
+    {
+      "epoch": 2.2416382252559726,
+      "grad_norm": 2.7804300785064697,
+      "learning_rate": 0.0002527872582480091,
+      "loss": 5.756,
+      "step": 6568
+    },
+    {
+      "epoch": 2.2419795221843004,
+      "grad_norm": 2.8473620414733887,
+      "learning_rate": 0.0002526734926052332,
+      "loss": 6.4581,
+      "step": 6569
+    },
+    {
+      "epoch": 2.242320819112628,
+      "grad_norm": 2.7306015491485596,
+      "learning_rate": 0.0002525597269624573,
+      "loss": 5.8941,
+      "step": 6570
+    },
+    {
+      "epoch": 2.2426621160409557,
+      "grad_norm": 4.8221516609191895,
+      "learning_rate": 0.0002524459613196814,
+      "loss": 4.8643,
+      "step": 6571
+    },
+    {
+      "epoch": 2.243003412969283,
+      "grad_norm": 2.7513792514801025,
+      "learning_rate": 0.0002523321956769056,
+      "loss": 5.9781,
+      "step": 6572
+    },
+    {
+      "epoch": 2.243344709897611,
+      "grad_norm": 2.6935529708862305,
+      "learning_rate": 0.0002522184300341297,
+      "loss": 6.2471,
+      "step": 6573
+    },
+    {
+      "epoch": 2.2436860068259388,
+      "grad_norm": 2.7685117721557617,
+      "learning_rate": 0.00025210466439135385,
+      "loss": 6.2081,
+      "step": 6574
+    },
+    {
+      "epoch": 2.244027303754266,
+      "grad_norm": 2.770594596862793,
+      "learning_rate": 0.00025199089874857795,
+      "loss": 5.9571,
+      "step": 6575
+    },
+    {
+      "epoch": 2.244368600682594,
+      "grad_norm": 2.826627731323242,
+      "learning_rate": 0.00025187713310580206,
+      "loss": 5.9076,
+      "step": 6576
+    },
+    {
+      "epoch": 2.2447098976109214,
+      "grad_norm": 2.8154921531677246,
+      "learning_rate": 0.0002517633674630262,
+      "loss": 5.9773,
+      "step": 6577
+    },
+    {
+      "epoch": 2.2450511945392493,
+      "grad_norm": 2.681960105895996,
+      "learning_rate": 0.00025164960182025027,
+      "loss": 5.4657,
+      "step": 6578
+    },
+    {
+      "epoch": 2.2453924914675767,
+      "grad_norm": 2.780457019805908,
+      "learning_rate": 0.00025153583617747437,
+      "loss": 6.2352,
+      "step": 6579
+    },
+    {
+      "epoch": 2.2457337883959045,
+      "grad_norm": 2.8633296489715576,
+      "learning_rate": 0.00025142207053469853,
+      "loss": 6.102,
+      "step": 6580
+    },
+    {
+      "epoch": 2.246075085324232,
+      "grad_norm": 3.040717124938965,
+      "learning_rate": 0.00025130830489192264,
+      "loss": 5.2462,
+      "step": 6581
+    },
+    {
+      "epoch": 2.24641638225256,
+      "grad_norm": 2.7666218280792236,
+      "learning_rate": 0.00025119453924914674,
+      "loss": 5.4919,
+      "step": 6582
+    },
+    {
+      "epoch": 2.246757679180887,
+      "grad_norm": 2.838775157928467,
+      "learning_rate": 0.0002510807736063709,
+      "loss": 6.2391,
+      "step": 6583
+    },
+    {
+      "epoch": 2.247098976109215,
+      "grad_norm": 2.8280463218688965,
+      "learning_rate": 0.000250967007963595,
+      "loss": 6.3613,
+      "step": 6584
+    },
+    {
+      "epoch": 2.2474402730375425,
+      "grad_norm": 2.8792102336883545,
+      "learning_rate": 0.0002508532423208191,
+      "loss": 5.8355,
+      "step": 6585
+    },
+    {
+      "epoch": 2.2477815699658703,
+      "grad_norm": 2.7618908882141113,
+      "learning_rate": 0.00025073947667804327,
+      "loss": 6.0076,
+      "step": 6586
+    },
+    {
+      "epoch": 2.248122866894198,
+      "grad_norm": 2.7827494144439697,
+      "learning_rate": 0.0002506257110352674,
+      "loss": 4.3666,
+      "step": 6587
+    },
+    {
+      "epoch": 2.2484641638225256,
+      "grad_norm": 2.926374673843384,
+      "learning_rate": 0.0002505119453924914,
+      "loss": 6.048,
+      "step": 6588
+    },
+    {
+      "epoch": 2.2488054607508534,
+      "grad_norm": 2.982754945755005,
+      "learning_rate": 0.0002503981797497156,
+      "loss": 5.8397,
+      "step": 6589
+    },
+    {
+      "epoch": 2.249146757679181,
+      "grad_norm": 2.973355770111084,
+      "learning_rate": 0.0002502844141069397,
+      "loss": 5.4961,
+      "step": 6590
+    },
+    {
+      "epoch": 2.2494880546075087,
+      "grad_norm": 2.7500498294830322,
+      "learning_rate": 0.0002501706484641638,
+      "loss": 5.631,
+      "step": 6591
+    },
+    {
+      "epoch": 2.249829351535836,
+      "grad_norm": 2.882366180419922,
+      "learning_rate": 0.00025005688282138795,
+      "loss": 5.8569,
+      "step": 6592
+    },
+    {
+      "epoch": 2.250170648464164,
+      "grad_norm": 2.73225736618042,
+      "learning_rate": 0.00024994311717861206,
+      "loss": 6.4002,
+      "step": 6593
+    },
+    {
+      "epoch": 2.2505119453924913,
+      "grad_norm": 2.800004720687866,
+      "learning_rate": 0.0002498293515358362,
+      "loss": 6.4107,
+      "step": 6594
+    },
+    {
+      "epoch": 2.250853242320819,
+      "grad_norm": 2.6945483684539795,
+      "learning_rate": 0.00024971558589306027,
+      "loss": 6.414,
+      "step": 6595
+    },
+    {
+      "epoch": 2.2511945392491466,
+      "grad_norm": 2.850060224533081,
+      "learning_rate": 0.0002496018202502844,
+      "loss": 5.8978,
+      "step": 6596
+    },
+    {
+      "epoch": 2.2515358361774744,
+      "grad_norm": 2.810472249984741,
+      "learning_rate": 0.00024948805460750853,
+      "loss": 5.7288,
+      "step": 6597
+    },
+    {
+      "epoch": 2.2518771331058023,
+      "grad_norm": 4.052561283111572,
+      "learning_rate": 0.00024937428896473264,
+      "loss": 5.1013,
+      "step": 6598
+    },
+    {
+      "epoch": 2.2522184300341297,
+      "grad_norm": 3.0097126960754395,
+      "learning_rate": 0.0002492605233219568,
+      "loss": 5.8071,
+      "step": 6599
+    },
+    {
+      "epoch": 2.252559726962457,
+      "grad_norm": 2.8566362857818604,
+      "learning_rate": 0.0002491467576791809,
+      "loss": 6.1921,
+      "step": 6600
+    },
+    {
+      "epoch": 2.252901023890785,
+      "grad_norm": 3.514784336090088,
+      "learning_rate": 0.000249032992036405,
+      "loss": 4.8535,
+      "step": 6601
+    },
+    {
+      "epoch": 2.253242320819113,
+      "grad_norm": 3.010627031326294,
+      "learning_rate": 0.0002489192263936291,
+      "loss": 6.0636,
+      "step": 6602
+    },
+    {
+      "epoch": 2.25358361774744,
+      "grad_norm": 2.843026876449585,
+      "learning_rate": 0.00024880546075085327,
+      "loss": 5.8824,
+      "step": 6603
+    },
+    {
+      "epoch": 2.253924914675768,
+      "grad_norm": 2.940535306930542,
+      "learning_rate": 0.0002486916951080774,
+      "loss": 5.8778,
+      "step": 6604
+    },
+    {
+      "epoch": 2.2542662116040955,
+      "grad_norm": 2.917555570602417,
+      "learning_rate": 0.0002485779294653015,
+      "loss": 6.1297,
+      "step": 6605
+    },
+    {
+      "epoch": 2.2546075085324233,
+      "grad_norm": 3.041822910308838,
+      "learning_rate": 0.0002484641638225256,
+      "loss": 5.5592,
+      "step": 6606
+    },
+    {
+      "epoch": 2.2549488054607507,
+      "grad_norm": 2.8139913082122803,
+      "learning_rate": 0.00024835039817974974,
+      "loss": 6.2674,
+      "step": 6607
+    },
+    {
+      "epoch": 2.2552901023890786,
+      "grad_norm": 2.7989606857299805,
+      "learning_rate": 0.00024823663253697385,
+      "loss": 5.9394,
+      "step": 6608
+    },
+    {
+      "epoch": 2.255631399317406,
+      "grad_norm": 2.64013671875,
+      "learning_rate": 0.00024812286689419795,
+      "loss": 6.1189,
+      "step": 6609
+    },
+    {
+      "epoch": 2.255972696245734,
+      "grad_norm": 2.7237131595611572,
+      "learning_rate": 0.00024800910125142206,
+      "loss": 5.7884,
+      "step": 6610
+    },
+    {
+      "epoch": 2.2563139931740617,
+      "grad_norm": 4.8718719482421875,
+      "learning_rate": 0.00024789533560864616,
+      "loss": 6.2122,
+      "step": 6611
+    },
+    {
+      "epoch": 2.256655290102389,
+      "grad_norm": 3.137288808822632,
+      "learning_rate": 0.0002477815699658703,
+      "loss": 6.0289,
+      "step": 6612
+    },
+    {
+      "epoch": 2.2569965870307165,
+      "grad_norm": 2.8663458824157715,
+      "learning_rate": 0.00024766780432309443,
+      "loss": 5.903,
+      "step": 6613
+    },
+    {
+      "epoch": 2.2573378839590443,
+      "grad_norm": 2.8547210693359375,
+      "learning_rate": 0.0002475540386803186,
+      "loss": 6.4314,
+      "step": 6614
+    },
+    {
+      "epoch": 2.257679180887372,
+      "grad_norm": 2.7361135482788086,
+      "learning_rate": 0.00024744027303754264,
+      "loss": 5.6617,
+      "step": 6615
+    },
+    {
+      "epoch": 2.2580204778156996,
+      "grad_norm": 3.2815380096435547,
+      "learning_rate": 0.0002473265073947668,
+      "loss": 5.6537,
+      "step": 6616
+    },
+    {
+      "epoch": 2.2583617747440274,
+      "grad_norm": 2.7974159717559814,
+      "learning_rate": 0.0002472127417519909,
+      "loss": 6.072,
+      "step": 6617
+    },
+    {
+      "epoch": 2.258703071672355,
+      "grad_norm": 2.8251793384552,
+      "learning_rate": 0.000247098976109215,
+      "loss": 6.4367,
+      "step": 6618
+    },
+    {
+      "epoch": 2.2590443686006827,
+      "grad_norm": 2.9200313091278076,
+      "learning_rate": 0.00024698521046643917,
+      "loss": 5.9348,
+      "step": 6619
+    },
+    {
+      "epoch": 2.25938566552901,
+      "grad_norm": 3.057755708694458,
+      "learning_rate": 0.00024687144482366327,
+      "loss": 5.5646,
+      "step": 6620
+    },
+    {
+      "epoch": 2.259726962457338,
+      "grad_norm": 2.739042043685913,
+      "learning_rate": 0.0002467576791808874,
+      "loss": 6.5243,
+      "step": 6621
+    },
+    {
+      "epoch": 2.2600682593856654,
+      "grad_norm": 2.8239619731903076,
+      "learning_rate": 0.0002466439135381115,
+      "loss": 5.7505,
+      "step": 6622
+    },
+    {
+      "epoch": 2.260409556313993,
+      "grad_norm": 2.8290717601776123,
+      "learning_rate": 0.00024653014789533564,
+      "loss": 5.9059,
+      "step": 6623
+    },
+    {
+      "epoch": 2.260750853242321,
+      "grad_norm": 3.1619184017181396,
+      "learning_rate": 0.00024641638225255974,
+      "loss": 5.6338,
+      "step": 6624
+    },
+    {
+      "epoch": 2.2610921501706485,
+      "grad_norm": 2.940603256225586,
+      "learning_rate": 0.00024630261660978385,
+      "loss": 5.4515,
+      "step": 6625
+    },
+    {
+      "epoch": 2.261433447098976,
+      "grad_norm": 3.02312970161438,
+      "learning_rate": 0.00024618885096700795,
+      "loss": 5.0054,
+      "step": 6626
+    },
+    {
+      "epoch": 2.2617747440273037,
+      "grad_norm": 2.7911484241485596,
+      "learning_rate": 0.0002460750853242321,
+      "loss": 6.7703,
+      "step": 6627
+    },
+    {
+      "epoch": 2.2621160409556316,
+      "grad_norm": 2.8118064403533936,
+      "learning_rate": 0.0002459613196814562,
+      "loss": 5.8083,
+      "step": 6628
+    },
+    {
+      "epoch": 2.262457337883959,
+      "grad_norm": 2.7824995517730713,
+      "learning_rate": 0.0002458475540386803,
+      "loss": 5.9023,
+      "step": 6629
+    },
+    {
+      "epoch": 2.262798634812287,
+      "grad_norm": 2.8783822059631348,
+      "learning_rate": 0.00024573378839590443,
+      "loss": 6.2359,
+      "step": 6630
+    },
+    {
+      "epoch": 2.2631399317406142,
+      "grad_norm": 2.9949822425842285,
+      "learning_rate": 0.00024562002275312853,
+      "loss": 5.0531,
+      "step": 6631
+    },
+    {
+      "epoch": 2.263481228668942,
+      "grad_norm": 2.8121888637542725,
+      "learning_rate": 0.0002455062571103527,
+      "loss": 5.86,
+      "step": 6632
+    },
+    {
+      "epoch": 2.2638225255972695,
+      "grad_norm": 2.776623249053955,
+      "learning_rate": 0.0002453924914675768,
+      "loss": 6.3007,
+      "step": 6633
+    },
+    {
+      "epoch": 2.2641638225255973,
+      "grad_norm": 2.7938060760498047,
+      "learning_rate": 0.0002452787258248009,
+      "loss": 6.3209,
+      "step": 6634
+    },
+    {
+      "epoch": 2.2645051194539247,
+      "grad_norm": 2.750595808029175,
+      "learning_rate": 0.000245164960182025,
+      "loss": 6.1808,
+      "step": 6635
+    },
+    {
+      "epoch": 2.2648464163822526,
+      "grad_norm": 2.7990636825561523,
+      "learning_rate": 0.00024505119453924917,
+      "loss": 6.4628,
+      "step": 6636
+    },
+    {
+      "epoch": 2.2651877133105804,
+      "grad_norm": 2.674833059310913,
+      "learning_rate": 0.00024493742889647327,
+      "loss": 6.1304,
+      "step": 6637
+    },
+    {
+      "epoch": 2.265529010238908,
+      "grad_norm": 2.6972098350524902,
+      "learning_rate": 0.0002448236632536974,
+      "loss": 6.4638,
+      "step": 6638
+    },
+    {
+      "epoch": 2.2658703071672353,
+      "grad_norm": 2.8341851234436035,
+      "learning_rate": 0.0002447098976109215,
+      "loss": 6.2938,
+      "step": 6639
+    },
+    {
+      "epoch": 2.266211604095563,
+      "grad_norm": 2.8352243900299072,
+      "learning_rate": 0.00024459613196814564,
+      "loss": 5.7276,
+      "step": 6640
+    },
+    {
+      "epoch": 2.266552901023891,
+      "grad_norm": 2.8336145877838135,
+      "learning_rate": 0.00024448236632536975,
+      "loss": 5.9019,
+      "step": 6641
+    },
+    {
+      "epoch": 2.2668941979522184,
+      "grad_norm": 2.872852087020874,
+      "learning_rate": 0.00024436860068259385,
+      "loss": 6.4402,
+      "step": 6642
+    },
+    {
+      "epoch": 2.267235494880546,
+      "grad_norm": 2.7706029415130615,
+      "learning_rate": 0.000244254835039818,
+      "loss": 5.992,
+      "step": 6643
+    },
+    {
+      "epoch": 2.2675767918088736,
+      "grad_norm": 2.805307149887085,
+      "learning_rate": 0.00024414106939704206,
+      "loss": 5.5684,
+      "step": 6644
+    },
+    {
+      "epoch": 2.2679180887372015,
+      "grad_norm": 2.671431303024292,
+      "learning_rate": 0.00024402730375426622,
+      "loss": 5.8068,
+      "step": 6645
+    },
+    {
+      "epoch": 2.268259385665529,
+      "grad_norm": 2.8474628925323486,
+      "learning_rate": 0.00024391353811149032,
+      "loss": 6.6582,
+      "step": 6646
+    },
+    {
+      "epoch": 2.2686006825938567,
+      "grad_norm": 2.825212240219116,
+      "learning_rate": 0.00024379977246871446,
+      "loss": 6.1744,
+      "step": 6647
+    },
+    {
+      "epoch": 2.268941979522184,
+      "grad_norm": 2.828723669052124,
+      "learning_rate": 0.0002436860068259386,
+      "loss": 5.794,
+      "step": 6648
+    },
+    {
+      "epoch": 2.269283276450512,
+      "grad_norm": 2.701645612716675,
+      "learning_rate": 0.00024357224118316267,
+      "loss": 6.222,
+      "step": 6649
+    },
+    {
+      "epoch": 2.26962457337884,
+      "grad_norm": 2.8495328426361084,
+      "learning_rate": 0.0002434584755403868,
+      "loss": 6.0397,
+      "step": 6650
+    },
+    {
+      "epoch": 2.2699658703071672,
+      "grad_norm": 2.8856194019317627,
+      "learning_rate": 0.00024334470989761093,
+      "loss": 5.6817,
+      "step": 6651
+    },
+    {
+      "epoch": 2.2703071672354946,
+      "grad_norm": 2.8644917011260986,
+      "learning_rate": 0.00024323094425483506,
+      "loss": 5.5872,
+      "step": 6652
+    },
+    {
+      "epoch": 2.2706484641638225,
+      "grad_norm": 2.808652877807617,
+      "learning_rate": 0.00024311717861205917,
+      "loss": 5.4768,
+      "step": 6653
+    },
+    {
+      "epoch": 2.2709897610921503,
+      "grad_norm": 3.034461259841919,
+      "learning_rate": 0.00024300341296928327,
+      "loss": 6.1287,
+      "step": 6654
+    },
+    {
+      "epoch": 2.2713310580204777,
+      "grad_norm": 2.723137140274048,
+      "learning_rate": 0.0002428896473265074,
+      "loss": 6.1159,
+      "step": 6655
+    },
+    {
+      "epoch": 2.2716723549488056,
+      "grad_norm": 2.77011775970459,
+      "learning_rate": 0.0002427758816837315,
+      "loss": 6.3846,
+      "step": 6656
+    },
+    {
+      "epoch": 2.272013651877133,
+      "grad_norm": 2.8131182193756104,
+      "learning_rate": 0.00024266211604095564,
+      "loss": 6.6823,
+      "step": 6657
+    },
+    {
+      "epoch": 2.272354948805461,
+      "grad_norm": 2.815347194671631,
+      "learning_rate": 0.00024254835039817977,
+      "loss": 6.3332,
+      "step": 6658
+    },
+    {
+      "epoch": 2.2726962457337883,
+      "grad_norm": 2.8586032390594482,
+      "learning_rate": 0.00024243458475540385,
+      "loss": 6.2077,
+      "step": 6659
+    },
+    {
+      "epoch": 2.273037542662116,
+      "grad_norm": 2.8561007976531982,
+      "learning_rate": 0.00024232081911262798,
+      "loss": 5.5016,
+      "step": 6660
+    },
+    {
+      "epoch": 2.2733788395904435,
+      "grad_norm": 2.8218114376068115,
+      "learning_rate": 0.00024220705346985212,
+      "loss": 6.3183,
+      "step": 6661
+    },
+    {
+      "epoch": 2.2737201365187714,
+      "grad_norm": 2.9033164978027344,
+      "learning_rate": 0.00024209328782707625,
+      "loss": 5.2998,
+      "step": 6662
+    },
+    {
+      "epoch": 2.274061433447099,
+      "grad_norm": 2.784006357192993,
+      "learning_rate": 0.00024197952218430033,
+      "loss": 6.1072,
+      "step": 6663
+    },
+    {
+      "epoch": 2.2744027303754266,
+      "grad_norm": 2.8394157886505127,
+      "learning_rate": 0.00024186575654152446,
+      "loss": 5.9447,
+      "step": 6664
+    },
+    {
+      "epoch": 2.274744027303754,
+      "grad_norm": 2.8039710521698,
+      "learning_rate": 0.0002417519908987486,
+      "loss": 6.2788,
+      "step": 6665
+    },
+    {
+      "epoch": 2.275085324232082,
+      "grad_norm": 2.781151533126831,
+      "learning_rate": 0.0002416382252559727,
+      "loss": 6.0061,
+      "step": 6666
+    },
+    {
+      "epoch": 2.2754266211604097,
+      "grad_norm": 2.732513427734375,
+      "learning_rate": 0.00024152445961319683,
+      "loss": 6.1009,
+      "step": 6667
+    },
+    {
+      "epoch": 2.275767918088737,
+      "grad_norm": 2.8295493125915527,
+      "learning_rate": 0.00024141069397042093,
+      "loss": 5.4378,
+      "step": 6668
+    },
+    {
+      "epoch": 2.276109215017065,
+      "grad_norm": 2.9463069438934326,
+      "learning_rate": 0.00024129692832764504,
+      "loss": 5.8396,
+      "step": 6669
+    },
+    {
+      "epoch": 2.2764505119453924,
+      "grad_norm": 2.698521375656128,
+      "learning_rate": 0.00024118316268486917,
+      "loss": 5.6336,
+      "step": 6670
+    },
+    {
+      "epoch": 2.2767918088737202,
+      "grad_norm": 4.356855869293213,
+      "learning_rate": 0.0002410693970420933,
+      "loss": 4.6489,
+      "step": 6671
+    },
+    {
+      "epoch": 2.2771331058020476,
+      "grad_norm": 2.87615704536438,
+      "learning_rate": 0.00024095563139931743,
+      "loss": 6.4171,
+      "step": 6672
+    },
+    {
+      "epoch": 2.2774744027303755,
+      "grad_norm": 2.763654947280884,
+      "learning_rate": 0.0002408418657565415,
+      "loss": 6.1001,
+      "step": 6673
+    },
+    {
+      "epoch": 2.277815699658703,
+      "grad_norm": 2.8707218170166016,
+      "learning_rate": 0.00024072810011376564,
+      "loss": 5.8888,
+      "step": 6674
+    },
+    {
+      "epoch": 2.2781569965870307,
+      "grad_norm": 2.8277971744537354,
+      "learning_rate": 0.00024061433447098977,
+      "loss": 5.6618,
+      "step": 6675
+    },
+    {
+      "epoch": 2.2784982935153586,
+      "grad_norm": 2.787092447280884,
+      "learning_rate": 0.00024050056882821388,
+      "loss": 6.5541,
+      "step": 6676
+    },
+    {
+      "epoch": 2.278839590443686,
+      "grad_norm": 3.1452341079711914,
+      "learning_rate": 0.000240386803185438,
+      "loss": 4.8793,
+      "step": 6677
+    },
+    {
+      "epoch": 2.2791808873720134,
+      "grad_norm": 2.8025355339050293,
+      "learning_rate": 0.00024027303754266212,
+      "loss": 5.8608,
+      "step": 6678
+    },
+    {
+      "epoch": 2.2795221843003413,
+      "grad_norm": 2.7934341430664062,
+      "learning_rate": 0.00024015927189988622,
+      "loss": 6.2955,
+      "step": 6679
+    },
+    {
+      "epoch": 2.279863481228669,
+      "grad_norm": 2.744135618209839,
+      "learning_rate": 0.00024004550625711035,
+      "loss": 6.4455,
+      "step": 6680
+    },
+    {
+      "epoch": 2.2802047781569965,
+      "grad_norm": 2.76705002784729,
+      "learning_rate": 0.00023993174061433449,
+      "loss": 5.3531,
+      "step": 6681
+    },
+    {
+      "epoch": 2.2805460750853244,
+      "grad_norm": 2.9204037189483643,
+      "learning_rate": 0.00023981797497155862,
+      "loss": 6.5813,
+      "step": 6682
+    },
+    {
+      "epoch": 2.2808873720136518,
+      "grad_norm": 2.7960617542266846,
+      "learning_rate": 0.0002397042093287827,
+      "loss": 6.678,
+      "step": 6683
+    },
+    {
+      "epoch": 2.2812286689419796,
+      "grad_norm": 2.736600399017334,
+      "learning_rate": 0.00023959044368600683,
+      "loss": 6.0339,
+      "step": 6684
+    },
+    {
+      "epoch": 2.281569965870307,
+      "grad_norm": 3.3651063442230225,
+      "learning_rate": 0.00023947667804323096,
+      "loss": 5.7733,
+      "step": 6685
+    },
+    {
+      "epoch": 2.281911262798635,
+      "grad_norm": 2.7875654697418213,
+      "learning_rate": 0.00023936291240045506,
+      "loss": 5.9666,
+      "step": 6686
+    },
+    {
+      "epoch": 2.2822525597269623,
+      "grad_norm": 2.8029191493988037,
+      "learning_rate": 0.0002392491467576792,
+      "loss": 6.4682,
+      "step": 6687
+    },
+    {
+      "epoch": 2.28259385665529,
+      "grad_norm": 2.731361150741577,
+      "learning_rate": 0.0002391353811149033,
+      "loss": 6.54,
+      "step": 6688
+    },
+    {
+      "epoch": 2.282935153583618,
+      "grad_norm": 2.759530782699585,
+      "learning_rate": 0.0002390216154721274,
+      "loss": 6.4203,
+      "step": 6689
+    },
+    {
+      "epoch": 2.2832764505119454,
+      "grad_norm": 2.8398633003234863,
+      "learning_rate": 0.00023890784982935154,
+      "loss": 6.7545,
+      "step": 6690
+    },
+    {
+      "epoch": 2.283617747440273,
+      "grad_norm": 4.068755626678467,
+      "learning_rate": 0.00023879408418657567,
+      "loss": 5.7944,
+      "step": 6691
+    },
+    {
+      "epoch": 2.2839590443686006,
+      "grad_norm": 2.7552263736724854,
+      "learning_rate": 0.0002386803185437998,
+      "loss": 6.4369,
+      "step": 6692
+    },
+    {
+      "epoch": 2.2843003412969285,
+      "grad_norm": 2.8534114360809326,
+      "learning_rate": 0.00023856655290102388,
+      "loss": 5.6889,
+      "step": 6693
+    },
+    {
+      "epoch": 2.284641638225256,
+      "grad_norm": 2.8475186824798584,
+      "learning_rate": 0.000238452787258248,
+      "loss": 5.7206,
+      "step": 6694
+    },
+    {
+      "epoch": 2.2849829351535837,
+      "grad_norm": 2.8369054794311523,
+      "learning_rate": 0.00023833902161547214,
+      "loss": 5.6994,
+      "step": 6695
+    },
+    {
+      "epoch": 2.285324232081911,
+      "grad_norm": 2.86474609375,
+      "learning_rate": 0.00023822525597269625,
+      "loss": 6.6088,
+      "step": 6696
+    },
+    {
+      "epoch": 2.285665529010239,
+      "grad_norm": 2.8578531742095947,
+      "learning_rate": 0.00023811149032992038,
+      "loss": 6.2553,
+      "step": 6697
+    },
+    {
+      "epoch": 2.2860068259385664,
+      "grad_norm": 2.844667673110962,
+      "learning_rate": 0.00023799772468714449,
+      "loss": 5.9709,
+      "step": 6698
+    },
+    {
+      "epoch": 2.2863481228668943,
+      "grad_norm": 2.900826930999756,
+      "learning_rate": 0.0002378839590443686,
+      "loss": 6.7451,
+      "step": 6699
+    },
+    {
+      "epoch": 2.2866894197952217,
+      "grad_norm": 2.819692373275757,
+      "learning_rate": 0.00023777019340159272,
+      "loss": 5.666,
+      "step": 6700
+    },
+    {
+      "epoch": 2.2870307167235495,
+      "grad_norm": 2.7792670726776123,
+      "learning_rate": 0.00023765642775881686,
+      "loss": 6.7826,
+      "step": 6701
+    },
+    {
+      "epoch": 2.2873720136518774,
+      "grad_norm": 2.857227087020874,
+      "learning_rate": 0.00023754266211604096,
+      "loss": 5.8068,
+      "step": 6702
+    },
+    {
+      "epoch": 2.2877133105802048,
+      "grad_norm": 2.8718831539154053,
+      "learning_rate": 0.00023742889647326507,
+      "loss": 6.0465,
+      "step": 6703
+    },
+    {
+      "epoch": 2.288054607508532,
+      "grad_norm": 2.8006720542907715,
+      "learning_rate": 0.0002373151308304892,
+      "loss": 6.0748,
+      "step": 6704
+    },
+    {
+      "epoch": 2.28839590443686,
+      "grad_norm": 2.717714548110962,
+      "learning_rate": 0.00023720136518771333,
+      "loss": 6.2189,
+      "step": 6705
+    },
+    {
+      "epoch": 2.288737201365188,
+      "grad_norm": 2.80198073387146,
+      "learning_rate": 0.00023708759954493743,
+      "loss": 6.7073,
+      "step": 6706
+    },
+    {
+      "epoch": 2.2890784982935153,
+      "grad_norm": 2.960460901260376,
+      "learning_rate": 0.00023697383390216154,
+      "loss": 5.2394,
+      "step": 6707
+    },
+    {
+      "epoch": 2.289419795221843,
+      "grad_norm": 2.737926483154297,
+      "learning_rate": 0.00023686006825938567,
+      "loss": 6.5359,
+      "step": 6708
+    },
+    {
+      "epoch": 2.2897610921501705,
+      "grad_norm": 2.870438814163208,
+      "learning_rate": 0.00023674630261660978,
+      "loss": 6.5294,
+      "step": 6709
+    },
+    {
+      "epoch": 2.2901023890784984,
+      "grad_norm": 2.9853978157043457,
+      "learning_rate": 0.0002366325369738339,
+      "loss": 6.5247,
+      "step": 6710
+    },
+    {
+      "epoch": 2.290443686006826,
+      "grad_norm": 2.814044713973999,
+      "learning_rate": 0.00023651877133105804,
+      "loss": 6.4487,
+      "step": 6711
+    },
+    {
+      "epoch": 2.2907849829351536,
+      "grad_norm": 2.9334306716918945,
+      "learning_rate": 0.00023640500568828215,
+      "loss": 5.1492,
+      "step": 6712
+    },
+    {
+      "epoch": 2.291126279863481,
+      "grad_norm": 2.8538413047790527,
+      "learning_rate": 0.00023629124004550625,
+      "loss": 6.6194,
+      "step": 6713
+    },
+    {
+      "epoch": 2.291467576791809,
+      "grad_norm": 2.9160053730010986,
+      "learning_rate": 0.00023617747440273038,
+      "loss": 5.5613,
+      "step": 6714
+    },
+    {
+      "epoch": 2.2918088737201368,
+      "grad_norm": 2.795891046524048,
+      "learning_rate": 0.00023606370875995451,
+      "loss": 6.1532,
+      "step": 6715
+    },
+    {
+      "epoch": 2.292150170648464,
+      "grad_norm": 2.7880423069000244,
+      "learning_rate": 0.00023594994311717862,
+      "loss": 5.6862,
+      "step": 6716
+    },
+    {
+      "epoch": 2.2924914675767916,
+      "grad_norm": 2.863717794418335,
+      "learning_rate": 0.00023583617747440272,
+      "loss": 5.6751,
+      "step": 6717
+    },
+    {
+      "epoch": 2.2928327645051194,
+      "grad_norm": 2.9119949340820312,
+      "learning_rate": 0.00023572241183162686,
+      "loss": 5.5249,
+      "step": 6718
+    },
+    {
+      "epoch": 2.2931740614334473,
+      "grad_norm": 2.8489530086517334,
+      "learning_rate": 0.00023560864618885096,
+      "loss": 6.3403,
+      "step": 6719
+    },
+    {
+      "epoch": 2.2935153583617747,
+      "grad_norm": 3.275228261947632,
+      "learning_rate": 0.0002354948805460751,
+      "loss": 5.1688,
+      "step": 6720
+    },
+    {
+      "epoch": 2.2938566552901025,
+      "grad_norm": 2.758864164352417,
+      "learning_rate": 0.00023538111490329923,
+      "loss": 6.241,
+      "step": 6721
+    },
+    {
+      "epoch": 2.29419795221843,
+      "grad_norm": 2.7460415363311768,
+      "learning_rate": 0.00023526734926052333,
+      "loss": 5.9875,
+      "step": 6722
+    },
+    {
+      "epoch": 2.2945392491467578,
+      "grad_norm": 2.8659422397613525,
+      "learning_rate": 0.00023515358361774744,
+      "loss": 6.1308,
+      "step": 6723
+    },
+    {
+      "epoch": 2.294880546075085,
+      "grad_norm": 8.827744483947754,
+      "learning_rate": 0.00023503981797497157,
+      "loss": 6.2413,
+      "step": 6724
+    },
+    {
+      "epoch": 2.295221843003413,
+      "grad_norm": 2.7797117233276367,
+      "learning_rate": 0.0002349260523321957,
+      "loss": 6.1305,
+      "step": 6725
+    },
+    {
+      "epoch": 2.2955631399317404,
+      "grad_norm": 2.6930477619171143,
+      "learning_rate": 0.0002348122866894198,
+      "loss": 6.202,
+      "step": 6726
+    },
+    {
+      "epoch": 2.2959044368600683,
+      "grad_norm": 2.801419258117676,
+      "learning_rate": 0.0002346985210466439,
+      "loss": 6.2353,
+      "step": 6727
+    },
+    {
+      "epoch": 2.296245733788396,
+      "grad_norm": 2.8180465698242188,
+      "learning_rate": 0.00023458475540386804,
+      "loss": 6.1544,
+      "step": 6728
+    },
+    {
+      "epoch": 2.2965870307167235,
+      "grad_norm": 2.789066791534424,
+      "learning_rate": 0.00023447098976109215,
+      "loss": 5.9909,
+      "step": 6729
+    },
+    {
+      "epoch": 2.296928327645051,
+      "grad_norm": 2.7527191638946533,
+      "learning_rate": 0.00023435722411831628,
+      "loss": 6.1873,
+      "step": 6730
+    },
+    {
+      "epoch": 2.297269624573379,
+      "grad_norm": 2.798922061920166,
+      "learning_rate": 0.0002342434584755404,
+      "loss": 6.2598,
+      "step": 6731
+    },
+    {
+      "epoch": 2.2976109215017066,
+      "grad_norm": 2.9935081005096436,
+      "learning_rate": 0.0002341296928327645,
+      "loss": 5.3781,
+      "step": 6732
+    },
+    {
+      "epoch": 2.297952218430034,
+      "grad_norm": 2.8289272785186768,
+      "learning_rate": 0.00023401592718998862,
+      "loss": 6.3926,
+      "step": 6733
+    },
+    {
+      "epoch": 2.298293515358362,
+      "grad_norm": 2.873661756515503,
+      "learning_rate": 0.00023390216154721275,
+      "loss": 5.2289,
+      "step": 6734
+    },
+    {
+      "epoch": 2.2986348122866893,
+      "grad_norm": 2.7585761547088623,
+      "learning_rate": 0.00023378839590443688,
+      "loss": 6.4229,
+      "step": 6735
+    },
+    {
+      "epoch": 2.298976109215017,
+      "grad_norm": 2.7429072856903076,
+      "learning_rate": 0.00023367463026166096,
+      "loss": 6.1628,
+      "step": 6736
+    },
+    {
+      "epoch": 2.2993174061433446,
+      "grad_norm": 2.8091378211975098,
+      "learning_rate": 0.0002335608646188851,
+      "loss": 6.2183,
+      "step": 6737
+    },
+    {
+      "epoch": 2.2996587030716724,
+      "grad_norm": 2.830545425415039,
+      "learning_rate": 0.00023344709897610923,
+      "loss": 6.1877,
+      "step": 6738
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.6861021518707275,
+      "learning_rate": 0.00023333333333333333,
+      "loss": 4.8636,
+      "step": 6739
+    },
+    {
+      "epoch": 2.3003412969283277,
+      "grad_norm": 2.8198957443237305,
+      "learning_rate": 0.00023321956769055746,
+      "loss": 6.044,
+      "step": 6740
+    },
+    {
+      "epoch": 2.3006825938566555,
+      "grad_norm": 2.8199527263641357,
+      "learning_rate": 0.00023310580204778157,
+      "loss": 6.2105,
+      "step": 6741
+    },
+    {
+      "epoch": 2.301023890784983,
+      "grad_norm": 2.9623944759368896,
+      "learning_rate": 0.00023299203640500567,
+      "loss": 5.464,
+      "step": 6742
+    },
+    {
+      "epoch": 2.3013651877133103,
+      "grad_norm": 2.8883421421051025,
+      "learning_rate": 0.0002328782707622298,
+      "loss": 5.7796,
+      "step": 6743
+    },
+    {
+      "epoch": 2.301706484641638,
+      "grad_norm": 2.6974728107452393,
+      "learning_rate": 0.00023276450511945394,
+      "loss": 6.3523,
+      "step": 6744
+    },
+    {
+      "epoch": 2.302047781569966,
+      "grad_norm": 2.7741858959198,
+      "learning_rate": 0.00023265073947667807,
+      "loss": 6.0041,
+      "step": 6745
+    },
+    {
+      "epoch": 2.3023890784982934,
+      "grad_norm": 2.771564245223999,
+      "learning_rate": 0.00023253697383390215,
+      "loss": 6.3403,
+      "step": 6746
+    },
+    {
+      "epoch": 2.3027303754266213,
+      "grad_norm": 2.801663875579834,
+      "learning_rate": 0.00023242320819112628,
+      "loss": 5.9708,
+      "step": 6747
+    },
+    {
+      "epoch": 2.3030716723549487,
+      "grad_norm": 2.906155586242676,
+      "learning_rate": 0.0002323094425483504,
+      "loss": 5.9501,
+      "step": 6748
+    },
+    {
+      "epoch": 2.3034129692832765,
+      "grad_norm": 2.7590839862823486,
+      "learning_rate": 0.00023219567690557452,
+      "loss": 6.2378,
+      "step": 6749
+    },
+    {
+      "epoch": 2.303754266211604,
+      "grad_norm": 2.7589144706726074,
+      "learning_rate": 0.00023208191126279865,
+      "loss": 6.4537,
+      "step": 6750
+    },
+    {
+      "epoch": 2.304095563139932,
+      "grad_norm": 2.7965896129608154,
+      "learning_rate": 0.00023196814562002275,
+      "loss": 5.9146,
+      "step": 6751
+    },
+    {
+      "epoch": 2.304436860068259,
+      "grad_norm": 2.827988862991333,
+      "learning_rate": 0.00023185437997724686,
+      "loss": 6.109,
+      "step": 6752
+    },
+    {
+      "epoch": 2.304778156996587,
+      "grad_norm": 2.7847177982330322,
+      "learning_rate": 0.000231740614334471,
+      "loss": 6.2527,
+      "step": 6753
+    },
+    {
+      "epoch": 2.305119453924915,
+      "grad_norm": 2.7947700023651123,
+      "learning_rate": 0.00023162684869169512,
+      "loss": 5.7674,
+      "step": 6754
+    },
+    {
+      "epoch": 2.3054607508532423,
+      "grad_norm": 2.828664779663086,
+      "learning_rate": 0.00023151308304891925,
+      "loss": 6.086,
+      "step": 6755
+    },
+    {
+      "epoch": 2.3058020477815697,
+      "grad_norm": 2.7907068729400635,
+      "learning_rate": 0.00023139931740614333,
+      "loss": 6.1609,
+      "step": 6756
+    },
+    {
+      "epoch": 2.3061433447098976,
+      "grad_norm": 2.880263566970825,
+      "learning_rate": 0.00023128555176336746,
+      "loss": 5.7513,
+      "step": 6757
+    },
+    {
+      "epoch": 2.3064846416382254,
+      "grad_norm": 2.764763832092285,
+      "learning_rate": 0.0002311717861205916,
+      "loss": 6.418,
+      "step": 6758
+    },
+    {
+      "epoch": 2.306825938566553,
+      "grad_norm": 5.498814105987549,
+      "learning_rate": 0.0002310580204778157,
+      "loss": 4.4087,
+      "step": 6759
+    },
+    {
+      "epoch": 2.3071672354948807,
+      "grad_norm": 2.8534750938415527,
+      "learning_rate": 0.00023094425483503983,
+      "loss": 6.3137,
+      "step": 6760
+    },
+    {
+      "epoch": 2.307508532423208,
+      "grad_norm": 2.6376616954803467,
+      "learning_rate": 0.00023083048919226394,
+      "loss": 6.0889,
+      "step": 6761
+    },
+    {
+      "epoch": 2.307849829351536,
+      "grad_norm": 3.473785877227783,
+      "learning_rate": 0.00023071672354948804,
+      "loss": 5.482,
+      "step": 6762
+    },
+    {
+      "epoch": 2.3081911262798633,
+      "grad_norm": 3.221046209335327,
+      "learning_rate": 0.00023060295790671217,
+      "loss": 5.6875,
+      "step": 6763
+    },
+    {
+      "epoch": 2.308532423208191,
+      "grad_norm": 2.871093273162842,
+      "learning_rate": 0.0002304891922639363,
+      "loss": 6.1357,
+      "step": 6764
+    },
+    {
+      "epoch": 2.3088737201365186,
+      "grad_norm": 4.9006667137146,
+      "learning_rate": 0.00023037542662116044,
+      "loss": 5.354,
+      "step": 6765
+    },
+    {
+      "epoch": 2.3092150170648464,
+      "grad_norm": 5.541250228881836,
+      "learning_rate": 0.00023026166097838452,
+      "loss": 5.1683,
+      "step": 6766
+    },
+    {
+      "epoch": 2.3095563139931743,
+      "grad_norm": 2.8012752532958984,
+      "learning_rate": 0.00023014789533560865,
+      "loss": 6.1139,
+      "step": 6767
+    },
+    {
+      "epoch": 2.3098976109215017,
+      "grad_norm": 2.7383131980895996,
+      "learning_rate": 0.00023003412969283278,
+      "loss": 5.1907,
+      "step": 6768
+    },
+    {
+      "epoch": 2.310238907849829,
+      "grad_norm": 2.7988059520721436,
+      "learning_rate": 0.00022992036405005689,
+      "loss": 5.9688,
+      "step": 6769
+    },
+    {
+      "epoch": 2.310580204778157,
+      "grad_norm": 2.8070616722106934,
+      "learning_rate": 0.000229806598407281,
+      "loss": 6.0292,
+      "step": 6770
+    },
+    {
+      "epoch": 2.310921501706485,
+      "grad_norm": 2.855241537094116,
+      "learning_rate": 0.00022969283276450512,
+      "loss": 5.8921,
+      "step": 6771
+    },
+    {
+      "epoch": 2.311262798634812,
+      "grad_norm": 6.3953142166137695,
+      "learning_rate": 0.00022957906712172923,
+      "loss": 5.0532,
+      "step": 6772
+    },
+    {
+      "epoch": 2.31160409556314,
+      "grad_norm": 6.097601890563965,
+      "learning_rate": 0.00022946530147895336,
+      "loss": 4.7717,
+      "step": 6773
+    },
+    {
+      "epoch": 2.3119453924914675,
+      "grad_norm": 4.208084583282471,
+      "learning_rate": 0.0002293515358361775,
+      "loss": 3.4272,
+      "step": 6774
+    },
+    {
+      "epoch": 2.3122866894197953,
+      "grad_norm": 2.8885231018066406,
+      "learning_rate": 0.0002292377701934016,
+      "loss": 6.0002,
+      "step": 6775
+    },
+    {
+      "epoch": 2.3126279863481227,
+      "grad_norm": 2.9425899982452393,
+      "learning_rate": 0.0002291240045506257,
+      "loss": 5.8151,
+      "step": 6776
+    },
+    {
+      "epoch": 2.3129692832764506,
+      "grad_norm": 2.7507827281951904,
+      "learning_rate": 0.00022901023890784983,
+      "loss": 6.2093,
+      "step": 6777
+    },
+    {
+      "epoch": 2.313310580204778,
+      "grad_norm": 2.732609272003174,
+      "learning_rate": 0.00022889647326507397,
+      "loss": 5.5137,
+      "step": 6778
+    },
+    {
+      "epoch": 2.313651877133106,
+      "grad_norm": 2.8268463611602783,
+      "learning_rate": 0.00022878270762229807,
+      "loss": 5.6401,
+      "step": 6779
+    },
+    {
+      "epoch": 2.3139931740614337,
+      "grad_norm": 2.7519233226776123,
+      "learning_rate": 0.00022866894197952218,
+      "loss": 6.2487,
+      "step": 6780
+    },
+    {
+      "epoch": 2.314334470989761,
+      "grad_norm": 2.7072274684906006,
+      "learning_rate": 0.0002285551763367463,
+      "loss": 5.4718,
+      "step": 6781
+    },
+    {
+      "epoch": 2.3146757679180885,
+      "grad_norm": 2.7394325733184814,
+      "learning_rate": 0.0002284414106939704,
+      "loss": 6.1958,
+      "step": 6782
+    },
+    {
+      "epoch": 2.3150170648464163,
+      "grad_norm": 2.812884569168091,
+      "learning_rate": 0.00022832764505119454,
+      "loss": 6.1972,
+      "step": 6783
+    },
+    {
+      "epoch": 2.315358361774744,
+      "grad_norm": 2.7412703037261963,
+      "learning_rate": 0.00022821387940841868,
+      "loss": 6.1518,
+      "step": 6784
+    },
+    {
+      "epoch": 2.3156996587030716,
+      "grad_norm": 2.686539888381958,
+      "learning_rate": 0.00022810011376564278,
+      "loss": 6.5791,
+      "step": 6785
+    },
+    {
+      "epoch": 2.3160409556313994,
+      "grad_norm": 2.780839443206787,
+      "learning_rate": 0.0002279863481228669,
+      "loss": 5.8385,
+      "step": 6786
+    },
+    {
+      "epoch": 2.316382252559727,
+      "grad_norm": 2.8812310695648193,
+      "learning_rate": 0.00022787258248009102,
+      "loss": 4.8785,
+      "step": 6787
+    },
+    {
+      "epoch": 2.3167235494880547,
+      "grad_norm": 2.8832690715789795,
+      "learning_rate": 0.00022775881683731515,
+      "loss": 6.5055,
+      "step": 6788
+    },
+    {
+      "epoch": 2.317064846416382,
+      "grad_norm": 2.7269794940948486,
+      "learning_rate": 0.00022764505119453926,
+      "loss": 6.2457,
+      "step": 6789
+    },
+    {
+      "epoch": 2.31740614334471,
+      "grad_norm": 2.879974126815796,
+      "learning_rate": 0.00022753128555176336,
+      "loss": 5.7098,
+      "step": 6790
+    },
+    {
+      "epoch": 2.3177474402730374,
+      "grad_norm": 2.8237853050231934,
+      "learning_rate": 0.0002274175199089875,
+      "loss": 5.9351,
+      "step": 6791
+    },
+    {
+      "epoch": 2.318088737201365,
+      "grad_norm": 2.740983247756958,
+      "learning_rate": 0.0002273037542662116,
+      "loss": 6.1166,
+      "step": 6792
+    },
+    {
+      "epoch": 2.318430034129693,
+      "grad_norm": 2.9319350719451904,
+      "learning_rate": 0.00022718998862343573,
+      "loss": 6.4873,
+      "step": 6793
+    },
+    {
+      "epoch": 2.3187713310580205,
+      "grad_norm": 2.7880241870880127,
+      "learning_rate": 0.00022707622298065986,
+      "loss": 5.8139,
+      "step": 6794
+    },
+    {
+      "epoch": 2.319112627986348,
+      "grad_norm": 2.693197727203369,
+      "learning_rate": 0.00022696245733788397,
+      "loss": 6.0517,
+      "step": 6795
+    },
+    {
+      "epoch": 2.3194539249146757,
+      "grad_norm": 2.8118865489959717,
+      "learning_rate": 0.00022684869169510807,
+      "loss": 6.0578,
+      "step": 6796
+    },
+    {
+      "epoch": 2.3197952218430036,
+      "grad_norm": 3.474181890487671,
+      "learning_rate": 0.0002267349260523322,
+      "loss": 4.9987,
+      "step": 6797
+    },
+    {
+      "epoch": 2.320136518771331,
+      "grad_norm": 2.758629083633423,
+      "learning_rate": 0.00022662116040955634,
+      "loss": 6.036,
+      "step": 6798
+    },
+    {
+      "epoch": 2.320477815699659,
+      "grad_norm": 2.87188720703125,
+      "learning_rate": 0.00022650739476678044,
+      "loss": 6.5413,
+      "step": 6799
+    },
+    {
+      "epoch": 2.3208191126279862,
+      "grad_norm": 2.650592803955078,
+      "learning_rate": 0.00022639362912400455,
+      "loss": 6.0238,
+      "step": 6800
+    },
+    {
+      "epoch": 2.321160409556314,
+      "grad_norm": 2.7691047191619873,
+      "learning_rate": 0.00022627986348122868,
+      "loss": 6.3138,
+      "step": 6801
+    },
+    {
+      "epoch": 2.3215017064846415,
+      "grad_norm": 2.800083637237549,
+      "learning_rate": 0.00022616609783845278,
+      "loss": 6.0257,
+      "step": 6802
+    },
+    {
+      "epoch": 2.3218430034129693,
+      "grad_norm": 2.9052236080169678,
+      "learning_rate": 0.00022605233219567691,
+      "loss": 5.8308,
+      "step": 6803
+    },
+    {
+      "epoch": 2.3221843003412967,
+      "grad_norm": 3.039931058883667,
+      "learning_rate": 0.00022593856655290102,
+      "loss": 5.0397,
+      "step": 6804
+    },
+    {
+      "epoch": 2.3225255972696246,
+      "grad_norm": 2.8269150257110596,
+      "learning_rate": 0.00022582480091012515,
+      "loss": 5.7589,
+      "step": 6805
+    },
+    {
+      "epoch": 2.3228668941979524,
+      "grad_norm": 2.7746031284332275,
+      "learning_rate": 0.00022571103526734926,
+      "loss": 5.6901,
+      "step": 6806
+    },
+    {
+      "epoch": 2.32320819112628,
+      "grad_norm": 1.9601713418960571,
+      "learning_rate": 0.0002255972696245734,
+      "loss": 3.2623,
+      "step": 6807
+    },
+    {
+      "epoch": 2.3235494880546073,
+      "grad_norm": 2.8860151767730713,
+      "learning_rate": 0.00022548350398179752,
+      "loss": 5.5029,
+      "step": 6808
+    },
+    {
+      "epoch": 2.323890784982935,
+      "grad_norm": 2.7598114013671875,
+      "learning_rate": 0.0002253697383390216,
+      "loss": 5.9686,
+      "step": 6809
+    },
+    {
+      "epoch": 2.324232081911263,
+      "grad_norm": 2.8497135639190674,
+      "learning_rate": 0.00022525597269624573,
+      "loss": 6.0748,
+      "step": 6810
+    },
+    {
+      "epoch": 2.3245733788395904,
+      "grad_norm": 2.8299918174743652,
+      "learning_rate": 0.00022514220705346986,
+      "loss": 6.3305,
+      "step": 6811
+    },
+    {
+      "epoch": 2.324914675767918,
+      "grad_norm": 2.794346332550049,
+      "learning_rate": 0.00022502844141069397,
+      "loss": 5.8338,
+      "step": 6812
+    },
+    {
+      "epoch": 2.3252559726962456,
+      "grad_norm": 2.809251546859741,
+      "learning_rate": 0.0002249146757679181,
+      "loss": 6.1905,
+      "step": 6813
+    },
+    {
+      "epoch": 2.3255972696245735,
+      "grad_norm": 3.0918211936950684,
+      "learning_rate": 0.0002248009101251422,
+      "loss": 6.2205,
+      "step": 6814
+    },
+    {
+      "epoch": 2.325938566552901,
+      "grad_norm": 6.421435832977295,
+      "learning_rate": 0.00022468714448236634,
+      "loss": 4.1736,
+      "step": 6815
+    },
+    {
+      "epoch": 2.3262798634812287,
+      "grad_norm": 3.02647066116333,
+      "learning_rate": 0.00022457337883959044,
+      "loss": 5.8284,
+      "step": 6816
+    },
+    {
+      "epoch": 2.326621160409556,
+      "grad_norm": 2.943453550338745,
+      "learning_rate": 0.00022445961319681457,
+      "loss": 5.329,
+      "step": 6817
+    },
+    {
+      "epoch": 2.326962457337884,
+      "grad_norm": 2.718707323074341,
+      "learning_rate": 0.0002243458475540387,
+      "loss": 6.4537,
+      "step": 6818
+    },
+    {
+      "epoch": 2.327303754266212,
+      "grad_norm": 2.7228763103485107,
+      "learning_rate": 0.00022423208191126278,
+      "loss": 5.5298,
+      "step": 6819
+    },
+    {
+      "epoch": 2.3276450511945392,
+      "grad_norm": 2.7870371341705322,
+      "learning_rate": 0.00022411831626848692,
+      "loss": 5.9528,
+      "step": 6820
+    },
+    {
+      "epoch": 2.3279863481228666,
+      "grad_norm": 2.7378652095794678,
+      "learning_rate": 0.00022400455062571105,
+      "loss": 5.9546,
+      "step": 6821
+    },
+    {
+      "epoch": 2.3283276450511945,
+      "grad_norm": 2.816159725189209,
+      "learning_rate": 0.00022389078498293515,
+      "loss": 6.0227,
+      "step": 6822
+    },
+    {
+      "epoch": 2.3286689419795223,
+      "grad_norm": 3.0372061729431152,
+      "learning_rate": 0.00022377701934015928,
+      "loss": 5.1104,
+      "step": 6823
+    },
+    {
+      "epoch": 2.3290102389078498,
+      "grad_norm": 3.7163870334625244,
+      "learning_rate": 0.0002236632536973834,
+      "loss": 5.6614,
+      "step": 6824
+    },
+    {
+      "epoch": 2.3293515358361776,
+      "grad_norm": 2.7196543216705322,
+      "learning_rate": 0.00022354948805460752,
+      "loss": 6.4318,
+      "step": 6825
+    },
+    {
+      "epoch": 2.329692832764505,
+      "grad_norm": 2.8511674404144287,
+      "learning_rate": 0.00022343572241183163,
+      "loss": 5.7164,
+      "step": 6826
+    },
+    {
+      "epoch": 2.330034129692833,
+      "grad_norm": 2.838578462600708,
+      "learning_rate": 0.00022332195676905576,
+      "loss": 5.8845,
+      "step": 6827
+    },
+    {
+      "epoch": 2.3303754266211603,
+      "grad_norm": 3.3302481174468994,
+      "learning_rate": 0.0002232081911262799,
+      "loss": 5.2908,
+      "step": 6828
+    },
+    {
+      "epoch": 2.330716723549488,
+      "grad_norm": 2.730325222015381,
+      "learning_rate": 0.00022309442548350397,
+      "loss": 6.1103,
+      "step": 6829
+    },
+    {
+      "epoch": 2.3310580204778155,
+      "grad_norm": 2.734823226928711,
+      "learning_rate": 0.0002229806598407281,
+      "loss": 6.5294,
+      "step": 6830
+    },
+    {
+      "epoch": 2.3313993174061434,
+      "grad_norm": 2.777035713195801,
+      "learning_rate": 0.00022286689419795223,
+      "loss": 6.0642,
+      "step": 6831
+    },
+    {
+      "epoch": 2.331740614334471,
+      "grad_norm": 2.896202802658081,
+      "learning_rate": 0.00022275312855517634,
+      "loss": 5.5833,
+      "step": 6832
+    },
+    {
+      "epoch": 2.3320819112627986,
+      "grad_norm": 2.8327341079711914,
+      "learning_rate": 0.00022263936291240047,
+      "loss": 5.7914,
+      "step": 6833
+    },
+    {
+      "epoch": 2.3324232081911265,
+      "grad_norm": 7.732255935668945,
+      "learning_rate": 0.00022252559726962457,
+      "loss": 3.887,
+      "step": 6834
+    },
+    {
+      "epoch": 2.332764505119454,
+      "grad_norm": 2.896789312362671,
+      "learning_rate": 0.00022241183162684868,
+      "loss": 6.0555,
+      "step": 6835
+    },
+    {
+      "epoch": 2.3331058020477817,
+      "grad_norm": 2.7806904315948486,
+      "learning_rate": 0.0002222980659840728,
+      "loss": 6.4753,
+      "step": 6836
+    },
+    {
+      "epoch": 2.333447098976109,
+      "grad_norm": 2.836212635040283,
+      "learning_rate": 0.00022218430034129694,
+      "loss": 6.3409,
+      "step": 6837
+    },
+    {
+      "epoch": 2.333788395904437,
+      "grad_norm": 2.7770421504974365,
+      "learning_rate": 0.00022207053469852108,
+      "loss": 6.5263,
+      "step": 6838
+    },
+    {
+      "epoch": 2.3341296928327644,
+      "grad_norm": 2.7537899017333984,
+      "learning_rate": 0.00022195676905574515,
+      "loss": 6.3918,
+      "step": 6839
+    },
+    {
+      "epoch": 2.3344709897610922,
+      "grad_norm": 2.9366977214813232,
+      "learning_rate": 0.00022184300341296929,
+      "loss": 5.7525,
+      "step": 6840
+    },
+    {
+      "epoch": 2.3348122866894196,
+      "grad_norm": 3.4186301231384277,
+      "learning_rate": 0.00022172923777019342,
+      "loss": 5.2059,
+      "step": 6841
+    },
+    {
+      "epoch": 2.3351535836177475,
+      "grad_norm": 2.832503080368042,
+      "learning_rate": 0.00022161547212741752,
+      "loss": 5.9087,
+      "step": 6842
+    },
+    {
+      "epoch": 2.335494880546075,
+      "grad_norm": 2.741806983947754,
+      "learning_rate": 0.00022150170648464163,
+      "loss": 6.3415,
+      "step": 6843
+    },
+    {
+      "epoch": 2.3358361774744028,
+      "grad_norm": 2.8064048290252686,
+      "learning_rate": 0.00022138794084186576,
+      "loss": 6.0018,
+      "step": 6844
+    },
+    {
+      "epoch": 2.3361774744027306,
+      "grad_norm": 2.7633867263793945,
+      "learning_rate": 0.00022127417519908986,
+      "loss": 6.0155,
+      "step": 6845
+    },
+    {
+      "epoch": 2.336518771331058,
+      "grad_norm": 2.8978452682495117,
+      "learning_rate": 0.000221160409556314,
+      "loss": 5.8894,
+      "step": 6846
+    },
+    {
+      "epoch": 2.336860068259386,
+      "grad_norm": 2.9551198482513428,
+      "learning_rate": 0.00022104664391353813,
+      "loss": 4.3913,
+      "step": 6847
+    },
+    {
+      "epoch": 2.3372013651877133,
+      "grad_norm": 2.845735549926758,
+      "learning_rate": 0.00022093287827076223,
+      "loss": 5.5712,
+      "step": 6848
+    },
+    {
+      "epoch": 2.337542662116041,
+      "grad_norm": 2.8292698860168457,
+      "learning_rate": 0.00022081911262798634,
+      "loss": 6.0293,
+      "step": 6849
+    },
+    {
+      "epoch": 2.3378839590443685,
+      "grad_norm": 2.7965333461761475,
+      "learning_rate": 0.00022070534698521047,
+      "loss": 5.9881,
+      "step": 6850
+    },
+    {
+      "epoch": 2.3382252559726964,
+      "grad_norm": 2.763213872909546,
+      "learning_rate": 0.0002205915813424346,
+      "loss": 6.1471,
+      "step": 6851
+    },
+    {
+      "epoch": 2.3385665529010238,
+      "grad_norm": 3.1655666828155518,
+      "learning_rate": 0.0002204778156996587,
+      "loss": 5.4231,
+      "step": 6852
+    },
+    {
+      "epoch": 2.3389078498293516,
+      "grad_norm": 2.8350963592529297,
+      "learning_rate": 0.0002203640500568828,
+      "loss": 5.9476,
+      "step": 6853
+    },
+    {
+      "epoch": 2.339249146757679,
+      "grad_norm": 2.731945514678955,
+      "learning_rate": 0.00022025028441410694,
+      "loss": 5.8654,
+      "step": 6854
+    },
+    {
+      "epoch": 2.339590443686007,
+      "grad_norm": 2.7220661640167236,
+      "learning_rate": 0.00022013651877133105,
+      "loss": 6.4109,
+      "step": 6855
+    },
+    {
+      "epoch": 2.3399317406143343,
+      "grad_norm": 2.7597591876983643,
+      "learning_rate": 0.00022002275312855518,
+      "loss": 6.377,
+      "step": 6856
+    },
+    {
+      "epoch": 2.340273037542662,
+      "grad_norm": 2.7260942459106445,
+      "learning_rate": 0.0002199089874857793,
+      "loss": 5.9268,
+      "step": 6857
+    },
+    {
+      "epoch": 2.34061433447099,
+      "grad_norm": 2.8388872146606445,
+      "learning_rate": 0.00021979522184300342,
+      "loss": 5.8152,
+      "step": 6858
+    },
+    {
+      "epoch": 2.3409556313993174,
+      "grad_norm": 2.754322052001953,
+      "learning_rate": 0.00021968145620022752,
+      "loss": 6.2267,
+      "step": 6859
+    },
+    {
+      "epoch": 2.3412969283276452,
+      "grad_norm": 2.73994779586792,
+      "learning_rate": 0.00021956769055745166,
+      "loss": 6.0755,
+      "step": 6860
+    },
+    {
+      "epoch": 2.3416382252559726,
+      "grad_norm": 2.6968204975128174,
+      "learning_rate": 0.0002194539249146758,
+      "loss": 6.0136,
+      "step": 6861
+    },
+    {
+      "epoch": 2.3419795221843005,
+      "grad_norm": 2.7095813751220703,
+      "learning_rate": 0.0002193401592718999,
+      "loss": 5.9502,
+      "step": 6862
+    },
+    {
+      "epoch": 2.342320819112628,
+      "grad_norm": 4.785019397735596,
+      "learning_rate": 0.000219226393629124,
+      "loss": 4.9616,
+      "step": 6863
+    },
+    {
+      "epoch": 2.3426621160409558,
+      "grad_norm": 2.7408225536346436,
+      "learning_rate": 0.00021911262798634813,
+      "loss": 6.1393,
+      "step": 6864
+    },
+    {
+      "epoch": 2.343003412969283,
+      "grad_norm": 2.8245584964752197,
+      "learning_rate": 0.00021899886234357223,
+      "loss": 6.7514,
+      "step": 6865
+    },
+    {
+      "epoch": 2.343344709897611,
+      "grad_norm": 2.7768466472625732,
+      "learning_rate": 0.00021888509670079637,
+      "loss": 6.0515,
+      "step": 6866
+    },
+    {
+      "epoch": 2.3436860068259384,
+      "grad_norm": 2.7474124431610107,
+      "learning_rate": 0.0002187713310580205,
+      "loss": 6.5632,
+      "step": 6867
+    },
+    {
+      "epoch": 2.3440273037542663,
+      "grad_norm": 2.730879545211792,
+      "learning_rate": 0.0002186575654152446,
+      "loss": 6.0169,
+      "step": 6868
+    },
+    {
+      "epoch": 2.3443686006825937,
+      "grad_norm": 2.721271514892578,
+      "learning_rate": 0.0002185437997724687,
+      "loss": 6.3651,
+      "step": 6869
+    },
+    {
+      "epoch": 2.3447098976109215,
+      "grad_norm": 2.727299690246582,
+      "learning_rate": 0.00021843003412969284,
+      "loss": 6.0875,
+      "step": 6870
+    },
+    {
+      "epoch": 2.3450511945392494,
+      "grad_norm": 2.7979040145874023,
+      "learning_rate": 0.00021831626848691697,
+      "loss": 5.3848,
+      "step": 6871
+    },
+    {
+      "epoch": 2.345392491467577,
+      "grad_norm": 2.7405858039855957,
+      "learning_rate": 0.00021820250284414108,
+      "loss": 5.9585,
+      "step": 6872
+    },
+    {
+      "epoch": 2.3457337883959046,
+      "grad_norm": 2.7488925457000732,
+      "learning_rate": 0.00021808873720136518,
+      "loss": 6.4375,
+      "step": 6873
+    },
+    {
+      "epoch": 2.346075085324232,
+      "grad_norm": 2.80232310295105,
+      "learning_rate": 0.00021797497155858931,
+      "loss": 6.0547,
+      "step": 6874
+    },
+    {
+      "epoch": 2.34641638225256,
+      "grad_norm": 2.803708791732788,
+      "learning_rate": 0.00021786120591581342,
+      "loss": 5.5016,
+      "step": 6875
+    },
+    {
+      "epoch": 2.3467576791808873,
+      "grad_norm": 2.724703550338745,
+      "learning_rate": 0.00021774744027303755,
+      "loss": 6.1233,
+      "step": 6876
+    },
+    {
+      "epoch": 2.347098976109215,
+      "grad_norm": 2.8196821212768555,
+      "learning_rate": 0.00021763367463026166,
+      "loss": 5.8877,
+      "step": 6877
+    },
+    {
+      "epoch": 2.3474402730375425,
+      "grad_norm": 3.931154251098633,
+      "learning_rate": 0.0002175199089874858,
+      "loss": 5.5875,
+      "step": 6878
+    },
+    {
+      "epoch": 2.3477815699658704,
+      "grad_norm": 3.5176618099212646,
+      "learning_rate": 0.0002174061433447099,
+      "loss": 4.2908,
+      "step": 6879
+    },
+    {
+      "epoch": 2.348122866894198,
+      "grad_norm": 2.752065896987915,
+      "learning_rate": 0.00021729237770193403,
+      "loss": 5.6804,
+      "step": 6880
+    },
+    {
+      "epoch": 2.3484641638225257,
+      "grad_norm": 2.8360445499420166,
+      "learning_rate": 0.00021717861205915816,
+      "loss": 6.1461,
+      "step": 6881
+    },
+    {
+      "epoch": 2.348805460750853,
+      "grad_norm": 2.739427328109741,
+      "learning_rate": 0.00021706484641638224,
+      "loss": 6.4064,
+      "step": 6882
+    },
+    {
+      "epoch": 2.349146757679181,
+      "grad_norm": 3.3373124599456787,
+      "learning_rate": 0.00021695108077360637,
+      "loss": 6.0023,
+      "step": 6883
+    },
+    {
+      "epoch": 2.3494880546075088,
+      "grad_norm": 2.7712957859039307,
+      "learning_rate": 0.0002168373151308305,
+      "loss": 6.3348,
+      "step": 6884
+    },
+    {
+      "epoch": 2.349829351535836,
+      "grad_norm": 3.1796998977661133,
+      "learning_rate": 0.0002167235494880546,
+      "loss": 5.1784,
+      "step": 6885
+    },
+    {
+      "epoch": 2.350170648464164,
+      "grad_norm": 2.7970149517059326,
+      "learning_rate": 0.00021660978384527874,
+      "loss": 6.1688,
+      "step": 6886
+    },
+    {
+      "epoch": 2.3505119453924914,
+      "grad_norm": 2.6617581844329834,
+      "learning_rate": 0.00021649601820250284,
+      "loss": 6.3794,
+      "step": 6887
+    },
+    {
+      "epoch": 2.3508532423208193,
+      "grad_norm": 2.752316474914551,
+      "learning_rate": 0.00021638225255972697,
+      "loss": 5.9957,
+      "step": 6888
+    },
+    {
+      "epoch": 2.3511945392491467,
+      "grad_norm": 2.7273192405700684,
+      "learning_rate": 0.00021626848691695108,
+      "loss": 6.0486,
+      "step": 6889
+    },
+    {
+      "epoch": 2.3515358361774745,
+      "grad_norm": 2.755157947540283,
+      "learning_rate": 0.0002161547212741752,
+      "loss": 6.2515,
+      "step": 6890
+    },
+    {
+      "epoch": 2.351877133105802,
+      "grad_norm": 2.757171154022217,
+      "learning_rate": 0.00021604095563139934,
+      "loss": 6.3392,
+      "step": 6891
+    },
+    {
+      "epoch": 2.35221843003413,
+      "grad_norm": 2.830286979675293,
+      "learning_rate": 0.00021592718998862342,
+      "loss": 6.4345,
+      "step": 6892
+    },
+    {
+      "epoch": 2.352559726962457,
+      "grad_norm": 2.98397159576416,
+      "learning_rate": 0.00021581342434584755,
+      "loss": 5.7659,
+      "step": 6893
+    },
+    {
+      "epoch": 2.352901023890785,
+      "grad_norm": 2.784120798110962,
+      "learning_rate": 0.00021569965870307168,
+      "loss": 5.3936,
+      "step": 6894
+    },
+    {
+      "epoch": 2.3532423208191124,
+      "grad_norm": 2.79896879196167,
+      "learning_rate": 0.0002155858930602958,
+      "loss": 6.6646,
+      "step": 6895
+    },
+    {
+      "epoch": 2.3535836177474403,
+      "grad_norm": 2.7373616695404053,
+      "learning_rate": 0.00021547212741751992,
+      "loss": 6.4598,
+      "step": 6896
+    },
+    {
+      "epoch": 2.353924914675768,
+      "grad_norm": 2.7969889640808105,
+      "learning_rate": 0.00021535836177474403,
+      "loss": 5.9267,
+      "step": 6897
+    },
+    {
+      "epoch": 2.3542662116040955,
+      "grad_norm": 2.9162516593933105,
+      "learning_rate": 0.00021524459613196816,
+      "loss": 5.6167,
+      "step": 6898
+    },
+    {
+      "epoch": 2.3546075085324234,
+      "grad_norm": 2.745880365371704,
+      "learning_rate": 0.00021513083048919226,
+      "loss": 5.9538,
+      "step": 6899
+    },
+    {
+      "epoch": 2.354948805460751,
+      "grad_norm": 2.799854278564453,
+      "learning_rate": 0.0002150170648464164,
+      "loss": 4.8416,
+      "step": 6900
+    },
+    {
+      "epoch": 2.3552901023890787,
+      "grad_norm": 2.84338116645813,
+      "learning_rate": 0.00021490329920364053,
+      "loss": 6.2332,
+      "step": 6901
+    },
+    {
+      "epoch": 2.355631399317406,
+      "grad_norm": 2.713643789291382,
+      "learning_rate": 0.0002147895335608646,
+      "loss": 6.3292,
+      "step": 6902
+    },
+    {
+      "epoch": 2.355972696245734,
+      "grad_norm": 2.7593953609466553,
+      "learning_rate": 0.00021467576791808874,
+      "loss": 6.1654,
+      "step": 6903
+    },
+    {
+      "epoch": 2.3563139931740613,
+      "grad_norm": 2.7301793098449707,
+      "learning_rate": 0.00021456200227531287,
+      "loss": 5.8449,
+      "step": 6904
+    },
+    {
+      "epoch": 2.356655290102389,
+      "grad_norm": 2.736713409423828,
+      "learning_rate": 0.00021444823663253697,
+      "loss": 6.0168,
+      "step": 6905
+    },
+    {
+      "epoch": 2.3569965870307166,
+      "grad_norm": 2.8782119750976562,
+      "learning_rate": 0.0002143344709897611,
+      "loss": 6.1835,
+      "step": 6906
+    },
+    {
+      "epoch": 2.3573378839590444,
+      "grad_norm": 2.7333459854125977,
+      "learning_rate": 0.0002142207053469852,
+      "loss": 6.9639,
+      "step": 6907
+    },
+    {
+      "epoch": 2.357679180887372,
+      "grad_norm": 2.8486573696136475,
+      "learning_rate": 0.00021410693970420934,
+      "loss": 5.9174,
+      "step": 6908
+    },
+    {
+      "epoch": 2.3580204778156997,
+      "grad_norm": 2.6959550380706787,
+      "learning_rate": 0.00021399317406143345,
+      "loss": 6.3947,
+      "step": 6909
+    },
+    {
+      "epoch": 2.3583617747440275,
+      "grad_norm": 2.771045207977295,
+      "learning_rate": 0.00021387940841865758,
+      "loss": 5.9439,
+      "step": 6910
+    },
+    {
+      "epoch": 2.358703071672355,
+      "grad_norm": 2.7466630935668945,
+      "learning_rate": 0.00021376564277588169,
+      "loss": 5.7553,
+      "step": 6911
+    },
+    {
+      "epoch": 2.359044368600683,
+      "grad_norm": 2.725827217102051,
+      "learning_rate": 0.0002136518771331058,
+      "loss": 5.7062,
+      "step": 6912
+    },
+    {
+      "epoch": 2.35938566552901,
+      "grad_norm": 2.763960123062134,
+      "learning_rate": 0.00021353811149032992,
+      "loss": 6.1409,
+      "step": 6913
+    },
+    {
+      "epoch": 2.359726962457338,
+      "grad_norm": 2.7459399700164795,
+      "learning_rate": 0.00021342434584755405,
+      "loss": 6.2805,
+      "step": 6914
+    },
+    {
+      "epoch": 2.3600682593856654,
+      "grad_norm": 2.6126558780670166,
+      "learning_rate": 0.00021331058020477816,
+      "loss": 5.6644,
+      "step": 6915
+    },
+    {
+      "epoch": 2.3604095563139933,
+      "grad_norm": 2.730069160461426,
+      "learning_rate": 0.00021319681456200226,
+      "loss": 5.9614,
+      "step": 6916
+    },
+    {
+      "epoch": 2.3607508532423207,
+      "grad_norm": 2.6941754817962646,
+      "learning_rate": 0.0002130830489192264,
+      "loss": 6.082,
+      "step": 6917
+    },
+    {
+      "epoch": 2.3610921501706486,
+      "grad_norm": 2.685556173324585,
+      "learning_rate": 0.00021296928327645053,
+      "loss": 5.7493,
+      "step": 6918
+    },
+    {
+      "epoch": 2.361433447098976,
+      "grad_norm": 2.748690366744995,
+      "learning_rate": 0.00021285551763367463,
+      "loss": 6.2632,
+      "step": 6919
+    },
+    {
+      "epoch": 2.361774744027304,
+      "grad_norm": 2.6726412773132324,
+      "learning_rate": 0.00021274175199089877,
+      "loss": 5.5741,
+      "step": 6920
+    },
+    {
+      "epoch": 2.362116040955631,
+      "grad_norm": 2.778284788131714,
+      "learning_rate": 0.00021262798634812287,
+      "loss": 5.3961,
+      "step": 6921
+    },
+    {
+      "epoch": 2.362457337883959,
+      "grad_norm": 2.7935075759887695,
+      "learning_rate": 0.00021251422070534698,
+      "loss": 6.1253,
+      "step": 6922
+    },
+    {
+      "epoch": 2.362798634812287,
+      "grad_norm": 2.765655279159546,
+      "learning_rate": 0.0002124004550625711,
+      "loss": 6.6871,
+      "step": 6923
+    },
+    {
+      "epoch": 2.3631399317406143,
+      "grad_norm": 2.659959554672241,
+      "learning_rate": 0.00021228668941979524,
+      "loss": 6.0147,
+      "step": 6924
+    },
+    {
+      "epoch": 2.363481228668942,
+      "grad_norm": 2.7546181678771973,
+      "learning_rate": 0.00021217292377701934,
+      "loss": 6.5698,
+      "step": 6925
+    },
+    {
+      "epoch": 2.3638225255972696,
+      "grad_norm": 2.8323190212249756,
+      "learning_rate": 0.00021205915813424345,
+      "loss": 5.7636,
+      "step": 6926
+    },
+    {
+      "epoch": 2.3641638225255974,
+      "grad_norm": 2.7773520946502686,
+      "learning_rate": 0.00021194539249146758,
+      "loss": 6.3198,
+      "step": 6927
+    },
+    {
+      "epoch": 2.364505119453925,
+      "grad_norm": 2.6490156650543213,
+      "learning_rate": 0.0002118316268486917,
+      "loss": 5.7899,
+      "step": 6928
+    },
+    {
+      "epoch": 2.3648464163822527,
+      "grad_norm": 2.878728151321411,
+      "learning_rate": 0.00021171786120591582,
+      "loss": 5.5224,
+      "step": 6929
+    },
+    {
+      "epoch": 2.36518771331058,
+      "grad_norm": 2.7923343181610107,
+      "learning_rate": 0.00021160409556313995,
+      "loss": 6.871,
+      "step": 6930
+    },
+    {
+      "epoch": 2.365529010238908,
+      "grad_norm": 4.340980052947998,
+      "learning_rate": 0.00021149032992036406,
+      "loss": 5.0883,
+      "step": 6931
+    },
+    {
+      "epoch": 2.3658703071672353,
+      "grad_norm": 2.780867338180542,
+      "learning_rate": 0.00021137656427758816,
+      "loss": 5.614,
+      "step": 6932
+    },
+    {
+      "epoch": 2.366211604095563,
+      "grad_norm": 2.7651314735412598,
+      "learning_rate": 0.0002112627986348123,
+      "loss": 6.4194,
+      "step": 6933
+    },
+    {
+      "epoch": 2.3665529010238906,
+      "grad_norm": 2.7465429306030273,
+      "learning_rate": 0.00021114903299203642,
+      "loss": 5.8274,
+      "step": 6934
+    },
+    {
+      "epoch": 2.3668941979522184,
+      "grad_norm": 2.8030600547790527,
+      "learning_rate": 0.00021103526734926053,
+      "loss": 5.9884,
+      "step": 6935
+    },
+    {
+      "epoch": 2.3672354948805463,
+      "grad_norm": 2.78486704826355,
+      "learning_rate": 0.00021092150170648463,
+      "loss": 6.4359,
+      "step": 6936
+    },
+    {
+      "epoch": 2.3675767918088737,
+      "grad_norm": 2.7584781646728516,
+      "learning_rate": 0.00021080773606370877,
+      "loss": 6.4127,
+      "step": 6937
+    },
+    {
+      "epoch": 2.3679180887372016,
+      "grad_norm": 2.7212891578674316,
+      "learning_rate": 0.00021069397042093287,
+      "loss": 6.8108,
+      "step": 6938
+    },
+    {
+      "epoch": 2.368259385665529,
+      "grad_norm": 2.8116931915283203,
+      "learning_rate": 0.000210580204778157,
+      "loss": 6.0632,
+      "step": 6939
+    },
+    {
+      "epoch": 2.368600682593857,
+      "grad_norm": 2.847207546234131,
+      "learning_rate": 0.00021046643913538114,
+      "loss": 5.8893,
+      "step": 6940
+    },
+    {
+      "epoch": 2.368941979522184,
+      "grad_norm": 2.7255427837371826,
+      "learning_rate": 0.00021035267349260524,
+      "loss": 6.3627,
+      "step": 6941
+    },
+    {
+      "epoch": 2.369283276450512,
+      "grad_norm": 3.3348500728607178,
+      "learning_rate": 0.00021023890784982934,
+      "loss": 6.4244,
+      "step": 6942
+    },
+    {
+      "epoch": 2.3696245733788395,
+      "grad_norm": 2.6929471492767334,
+      "learning_rate": 0.00021012514220705348,
+      "loss": 6.4524,
+      "step": 6943
+    },
+    {
+      "epoch": 2.3699658703071673,
+      "grad_norm": 2.598615884780884,
+      "learning_rate": 0.0002100113765642776,
+      "loss": 5.8728,
+      "step": 6944
+    },
+    {
+      "epoch": 2.3703071672354947,
+      "grad_norm": 2.8167927265167236,
+      "learning_rate": 0.0002098976109215017,
+      "loss": 5.5652,
+      "step": 6945
+    },
+    {
+      "epoch": 2.3706484641638226,
+      "grad_norm": 3.9340548515319824,
+      "learning_rate": 0.00020978384527872582,
+      "loss": 5.4056,
+      "step": 6946
+    },
+    {
+      "epoch": 2.37098976109215,
+      "grad_norm": 2.9168150424957275,
+      "learning_rate": 0.00020967007963594995,
+      "loss": 5.4847,
+      "step": 6947
+    },
+    {
+      "epoch": 2.371331058020478,
+      "grad_norm": 2.80441951751709,
+      "learning_rate": 0.00020955631399317406,
+      "loss": 6.661,
+      "step": 6948
+    },
+    {
+      "epoch": 2.3716723549488057,
+      "grad_norm": 2.7382166385650635,
+      "learning_rate": 0.0002094425483503982,
+      "loss": 5.718,
+      "step": 6949
+    },
+    {
+      "epoch": 2.372013651877133,
+      "grad_norm": 2.878997802734375,
+      "learning_rate": 0.0002093287827076223,
+      "loss": 6.9364,
+      "step": 6950
+    },
+    {
+      "epoch": 2.372354948805461,
+      "grad_norm": 2.7554514408111572,
+      "learning_rate": 0.00020921501706484642,
+      "loss": 6.7696,
+      "step": 6951
+    },
+    {
+      "epoch": 2.3726962457337883,
+      "grad_norm": 2.6601312160491943,
+      "learning_rate": 0.00020910125142207053,
+      "loss": 6.2882,
+      "step": 6952
+    },
+    {
+      "epoch": 2.373037542662116,
+      "grad_norm": 2.9576683044433594,
+      "learning_rate": 0.00020898748577929466,
+      "loss": 5.6812,
+      "step": 6953
+    },
+    {
+      "epoch": 2.3733788395904436,
+      "grad_norm": 2.700512170791626,
+      "learning_rate": 0.0002088737201365188,
+      "loss": 6.0231,
+      "step": 6954
+    },
+    {
+      "epoch": 2.3737201365187715,
+      "grad_norm": 2.6124114990234375,
+      "learning_rate": 0.00020875995449374287,
+      "loss": 5.9432,
+      "step": 6955
+    },
+    {
+      "epoch": 2.374061433447099,
+      "grad_norm": 4.633599281311035,
+      "learning_rate": 0.000208646188850967,
+      "loss": 5.6296,
+      "step": 6956
+    },
+    {
+      "epoch": 2.3744027303754267,
+      "grad_norm": 2.7569868564605713,
+      "learning_rate": 0.00020853242320819114,
+      "loss": 6.6165,
+      "step": 6957
+    },
+    {
+      "epoch": 2.374744027303754,
+      "grad_norm": 2.804227590560913,
+      "learning_rate": 0.00020841865756541524,
+      "loss": 5.4564,
+      "step": 6958
+    },
+    {
+      "epoch": 2.375085324232082,
+      "grad_norm": 2.7688915729522705,
+      "learning_rate": 0.00020830489192263937,
+      "loss": 6.1656,
+      "step": 6959
+    },
+    {
+      "epoch": 2.3754266211604094,
+      "grad_norm": 2.7455203533172607,
+      "learning_rate": 0.00020819112627986348,
+      "loss": 6.0186,
+      "step": 6960
+    },
+    {
+      "epoch": 2.375767918088737,
+      "grad_norm": 2.7759294509887695,
+      "learning_rate": 0.0002080773606370876,
+      "loss": 5.8901,
+      "step": 6961
+    },
+    {
+      "epoch": 2.376109215017065,
+      "grad_norm": 2.660414218902588,
+      "learning_rate": 0.00020796359499431171,
+      "loss": 6.0713,
+      "step": 6962
+    },
+    {
+      "epoch": 2.3764505119453925,
+      "grad_norm": 2.7604238986968994,
+      "learning_rate": 0.00020784982935153585,
+      "loss": 6.2344,
+      "step": 6963
+    },
+    {
+      "epoch": 2.3767918088737203,
+      "grad_norm": 2.65087890625,
+      "learning_rate": 0.00020773606370875998,
+      "loss": 6.0383,
+      "step": 6964
+    },
+    {
+      "epoch": 2.3771331058020477,
+      "grad_norm": 2.73121976852417,
+      "learning_rate": 0.00020762229806598406,
+      "loss": 5.8064,
+      "step": 6965
+    },
+    {
+      "epoch": 2.3774744027303756,
+      "grad_norm": 11.710968017578125,
+      "learning_rate": 0.0002075085324232082,
+      "loss": 5.0296,
+      "step": 6966
+    },
+    {
+      "epoch": 2.377815699658703,
+      "grad_norm": 2.5780153274536133,
+      "learning_rate": 0.00020739476678043232,
+      "loss": 6.0447,
+      "step": 6967
+    },
+    {
+      "epoch": 2.378156996587031,
+      "grad_norm": 2.7825236320495605,
+      "learning_rate": 0.00020728100113765643,
+      "loss": 6.6443,
+      "step": 6968
+    },
+    {
+      "epoch": 2.3784982935153582,
+      "grad_norm": 2.7635421752929688,
+      "learning_rate": 0.00020716723549488056,
+      "loss": 5.4036,
+      "step": 6969
+    },
+    {
+      "epoch": 2.378839590443686,
+      "grad_norm": 2.749699592590332,
+      "learning_rate": 0.00020705346985210466,
+      "loss": 5.763,
+      "step": 6970
+    },
+    {
+      "epoch": 2.3791808873720135,
+      "grad_norm": 2.686196804046631,
+      "learning_rate": 0.0002069397042093288,
+      "loss": 5.9363,
+      "step": 6971
+    },
+    {
+      "epoch": 2.3795221843003413,
+      "grad_norm": 2.801814317703247,
+      "learning_rate": 0.0002068259385665529,
+      "loss": 6.4167,
+      "step": 6972
+    },
+    {
+      "epoch": 2.3798634812286688,
+      "grad_norm": 2.927759885787964,
+      "learning_rate": 0.00020671217292377703,
+      "loss": 5.5384,
+      "step": 6973
+    },
+    {
+      "epoch": 2.3802047781569966,
+      "grad_norm": 5.355523109436035,
+      "learning_rate": 0.00020659840728100116,
+      "loss": 4.3588,
+      "step": 6974
+    },
+    {
+      "epoch": 2.3805460750853245,
+      "grad_norm": 6.205738544464111,
+      "learning_rate": 0.00020648464163822524,
+      "loss": 5.5929,
+      "step": 6975
+    },
+    {
+      "epoch": 2.380887372013652,
+      "grad_norm": 2.8085031509399414,
+      "learning_rate": 0.00020637087599544937,
+      "loss": 6.4081,
+      "step": 6976
+    },
+    {
+      "epoch": 2.3812286689419797,
+      "grad_norm": 2.7193679809570312,
+      "learning_rate": 0.0002062571103526735,
+      "loss": 6.0632,
+      "step": 6977
+    },
+    {
+      "epoch": 2.381569965870307,
+      "grad_norm": 2.8042826652526855,
+      "learning_rate": 0.0002061433447098976,
+      "loss": 5.7394,
+      "step": 6978
+    },
+    {
+      "epoch": 2.381911262798635,
+      "grad_norm": 2.7161612510681152,
+      "learning_rate": 0.00020602957906712174,
+      "loss": 6.55,
+      "step": 6979
+    },
+    {
+      "epoch": 2.3822525597269624,
+      "grad_norm": 2.6074378490448,
+      "learning_rate": 0.00020591581342434585,
+      "loss": 5.7456,
+      "step": 6980
+    },
+    {
+      "epoch": 2.38259385665529,
+      "grad_norm": 2.804840087890625,
+      "learning_rate": 0.00020580204778156998,
+      "loss": 4.9974,
+      "step": 6981
+    },
+    {
+      "epoch": 2.3829351535836176,
+      "grad_norm": 2.746267795562744,
+      "learning_rate": 0.00020568828213879408,
+      "loss": 6.2011,
+      "step": 6982
+    },
+    {
+      "epoch": 2.3832764505119455,
+      "grad_norm": 2.72701096534729,
+      "learning_rate": 0.00020557451649601822,
+      "loss": 6.2791,
+      "step": 6983
+    },
+    {
+      "epoch": 2.383617747440273,
+      "grad_norm": 2.6555674076080322,
+      "learning_rate": 0.00020546075085324232,
+      "loss": 5.5011,
+      "step": 6984
+    },
+    {
+      "epoch": 2.3839590443686007,
+      "grad_norm": 2.7322537899017334,
+      "learning_rate": 0.00020534698521046643,
+      "loss": 5.9383,
+      "step": 6985
+    },
+    {
+      "epoch": 2.384300341296928,
+      "grad_norm": 2.779167652130127,
+      "learning_rate": 0.00020523321956769056,
+      "loss": 6.5133,
+      "step": 6986
+    },
+    {
+      "epoch": 2.384641638225256,
+      "grad_norm": 2.8352394104003906,
+      "learning_rate": 0.0002051194539249147,
+      "loss": 6.4203,
+      "step": 6987
+    },
+    {
+      "epoch": 2.384982935153584,
+      "grad_norm": 2.740598440170288,
+      "learning_rate": 0.0002050056882821388,
+      "loss": 5.7477,
+      "step": 6988
+    },
+    {
+      "epoch": 2.3853242320819112,
+      "grad_norm": 2.779649496078491,
+      "learning_rate": 0.0002048919226393629,
+      "loss": 6.0869,
+      "step": 6989
+    },
+    {
+      "epoch": 2.385665529010239,
+      "grad_norm": 2.779036521911621,
+      "learning_rate": 0.00020477815699658703,
+      "loss": 5.5761,
+      "step": 6990
+    },
+    {
+      "epoch": 2.3860068259385665,
+      "grad_norm": 2.7772841453552246,
+      "learning_rate": 0.00020466439135381116,
+      "loss": 6.7393,
+      "step": 6991
+    },
+    {
+      "epoch": 2.3863481228668944,
+      "grad_norm": 2.7716453075408936,
+      "learning_rate": 0.00020455062571103527,
+      "loss": 6.1316,
+      "step": 6992
+    },
+    {
+      "epoch": 2.3866894197952218,
+      "grad_norm": 3.0707526206970215,
+      "learning_rate": 0.0002044368600682594,
+      "loss": 5.781,
+      "step": 6993
+    },
+    {
+      "epoch": 2.3870307167235496,
+      "grad_norm": 2.663928270339966,
+      "learning_rate": 0.0002043230944254835,
+      "loss": 6.2102,
+      "step": 6994
+    },
+    {
+      "epoch": 2.387372013651877,
+      "grad_norm": 2.7111122608184814,
+      "learning_rate": 0.0002042093287827076,
+      "loss": 6.2127,
+      "step": 6995
+    },
+    {
+      "epoch": 2.387713310580205,
+      "grad_norm": 2.7795569896698,
+      "learning_rate": 0.00020409556313993174,
+      "loss": 5.7084,
+      "step": 6996
+    },
+    {
+      "epoch": 2.3880546075085323,
+      "grad_norm": 2.881326198577881,
+      "learning_rate": 0.00020398179749715588,
+      "loss": 5.929,
+      "step": 6997
+    },
+    {
+      "epoch": 2.38839590443686,
+      "grad_norm": 2.725304365158081,
+      "learning_rate": 0.00020386803185437998,
+      "loss": 6.2795,
+      "step": 6998
+    },
+    {
+      "epoch": 2.3887372013651875,
+      "grad_norm": 2.7434685230255127,
+      "learning_rate": 0.00020375426621160409,
+      "loss": 6.0435,
+      "step": 6999
+    },
+    {
+      "epoch": 2.3890784982935154,
+      "grad_norm": 2.676453113555908,
+      "learning_rate": 0.00020364050056882822,
+      "loss": 6.1874,
+      "step": 7000
+    },
+    {
+      "epoch": 2.3894197952218432,
+      "grad_norm": 2.7269630432128906,
+      "learning_rate": 0.00020352673492605235,
+      "loss": 5.7898,
+      "step": 7001
+    },
+    {
+      "epoch": 2.3897610921501706,
+      "grad_norm": 2.72902774810791,
+      "learning_rate": 0.00020341296928327645,
+      "loss": 5.5644,
+      "step": 7002
+    },
+    {
+      "epoch": 2.3901023890784985,
+      "grad_norm": 2.750995397567749,
+      "learning_rate": 0.0002032992036405006,
+      "loss": 6.2748,
+      "step": 7003
+    },
+    {
+      "epoch": 2.390443686006826,
+      "grad_norm": 2.7395596504211426,
+      "learning_rate": 0.0002031854379977247,
+      "loss": 6.0858,
+      "step": 7004
+    },
+    {
+      "epoch": 2.3907849829351537,
+      "grad_norm": 2.73018217086792,
+      "learning_rate": 0.0002030716723549488,
+      "loss": 5.9797,
+      "step": 7005
+    },
+    {
+      "epoch": 2.391126279863481,
+      "grad_norm": 2.739351511001587,
+      "learning_rate": 0.00020295790671217293,
+      "loss": 6.2691,
+      "step": 7006
+    },
+    {
+      "epoch": 2.391467576791809,
+      "grad_norm": 2.804325580596924,
+      "learning_rate": 0.00020284414106939706,
+      "loss": 6.5061,
+      "step": 7007
+    },
+    {
+      "epoch": 2.3918088737201364,
+      "grad_norm": 2.734001636505127,
+      "learning_rate": 0.00020273037542662117,
+      "loss": 6.3185,
+      "step": 7008
+    },
+    {
+      "epoch": 2.3921501706484642,
+      "grad_norm": 2.743891954421997,
+      "learning_rate": 0.00020261660978384527,
+      "loss": 6.3436,
+      "step": 7009
+    },
+    {
+      "epoch": 2.3924914675767917,
+      "grad_norm": 3.208683729171753,
+      "learning_rate": 0.0002025028441410694,
+      "loss": 4.3648,
+      "step": 7010
+    },
+    {
+      "epoch": 2.3928327645051195,
+      "grad_norm": 2.7099032402038574,
+      "learning_rate": 0.00020238907849829353,
+      "loss": 5.1361,
+      "step": 7011
+    },
+    {
+      "epoch": 2.393174061433447,
+      "grad_norm": 2.8060643672943115,
+      "learning_rate": 0.00020227531285551764,
+      "loss": 6.1124,
+      "step": 7012
+    },
+    {
+      "epoch": 2.3935153583617748,
+      "grad_norm": 2.7645676136016846,
+      "learning_rate": 0.00020216154721274177,
+      "loss": 5.3761,
+      "step": 7013
+    },
+    {
+      "epoch": 2.3938566552901026,
+      "grad_norm": 2.6478114128112793,
+      "learning_rate": 0.00020204778156996588,
+      "loss": 5.6484,
+      "step": 7014
+    },
+    {
+      "epoch": 2.39419795221843,
+      "grad_norm": 2.8686115741729736,
+      "learning_rate": 0.00020193401592718998,
+      "loss": 5.5769,
+      "step": 7015
+    },
+    {
+      "epoch": 2.394539249146758,
+      "grad_norm": 2.700092315673828,
+      "learning_rate": 0.00020182025028441411,
+      "loss": 6.5028,
+      "step": 7016
+    },
+    {
+      "epoch": 2.3948805460750853,
+      "grad_norm": 2.810314655303955,
+      "learning_rate": 0.00020170648464163825,
+      "loss": 6.574,
+      "step": 7017
+    },
+    {
+      "epoch": 2.395221843003413,
+      "grad_norm": 2.6746256351470947,
+      "learning_rate": 0.00020159271899886232,
+      "loss": 5.7884,
+      "step": 7018
+    },
+    {
+      "epoch": 2.3955631399317405,
+      "grad_norm": 2.735712766647339,
+      "learning_rate": 0.00020147895335608646,
+      "loss": 6.5509,
+      "step": 7019
+    },
+    {
+      "epoch": 2.3959044368600684,
+      "grad_norm": 2.7703819274902344,
+      "learning_rate": 0.0002013651877133106,
+      "loss": 6.1054,
+      "step": 7020
+    },
+    {
+      "epoch": 2.396245733788396,
+      "grad_norm": 2.840043306350708,
+      "learning_rate": 0.00020125142207053472,
+      "loss": 6.2688,
+      "step": 7021
+    },
+    {
+      "epoch": 2.3965870307167236,
+      "grad_norm": 2.789001226425171,
+      "learning_rate": 0.00020113765642775882,
+      "loss": 6.1622,
+      "step": 7022
+    },
+    {
+      "epoch": 2.396928327645051,
+      "grad_norm": 2.6123547554016113,
+      "learning_rate": 0.00020102389078498293,
+      "loss": 5.9342,
+      "step": 7023
+    },
+    {
+      "epoch": 2.397269624573379,
+      "grad_norm": 2.793732166290283,
+      "learning_rate": 0.00020091012514220706,
+      "loss": 5.487,
+      "step": 7024
+    },
+    {
+      "epoch": 2.3976109215017063,
+      "grad_norm": 2.722079277038574,
+      "learning_rate": 0.00020079635949943117,
+      "loss": 6.1717,
+      "step": 7025
+    },
+    {
+      "epoch": 2.397952218430034,
+      "grad_norm": 2.798908233642578,
+      "learning_rate": 0.0002006825938566553,
+      "loss": 6.3019,
+      "step": 7026
+    },
+    {
+      "epoch": 2.398293515358362,
+      "grad_norm": 2.8467278480529785,
+      "learning_rate": 0.00020056882821387943,
+      "loss": 5.7511,
+      "step": 7027
+    },
+    {
+      "epoch": 2.3986348122866894,
+      "grad_norm": 2.7143468856811523,
+      "learning_rate": 0.0002004550625711035,
+      "loss": 5.4081,
+      "step": 7028
+    },
+    {
+      "epoch": 2.3989761092150172,
+      "grad_norm": 2.916228771209717,
+      "learning_rate": 0.00020034129692832764,
+      "loss": 5.913,
+      "step": 7029
+    },
+    {
+      "epoch": 2.3993174061433447,
+      "grad_norm": 2.752556562423706,
+      "learning_rate": 0.00020022753128555177,
+      "loss": 6.2246,
+      "step": 7030
+    },
+    {
+      "epoch": 2.3996587030716725,
+      "grad_norm": 2.748680591583252,
+      "learning_rate": 0.0002001137656427759,
+      "loss": 6.3106,
+      "step": 7031
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.7810096740722656,
+      "learning_rate": 0.0002,
+      "loss": 6.5981,
+      "step": 7032
+    },
+    {
+      "epoch": 2.4003412969283278,
+      "grad_norm": 2.9772250652313232,
+      "learning_rate": 0.00019988623435722411,
+      "loss": 4.0665,
+      "step": 7033
+    },
+    {
+      "epoch": 2.400682593856655,
+      "grad_norm": 2.8036811351776123,
+      "learning_rate": 0.00019977246871444825,
+      "loss": 6.0835,
+      "step": 7034
+    },
+    {
+      "epoch": 2.401023890784983,
+      "grad_norm": 2.714700698852539,
+      "learning_rate": 0.00019965870307167235,
+      "loss": 6.4391,
+      "step": 7035
+    },
+    {
+      "epoch": 2.4013651877133104,
+      "grad_norm": 2.8477370738983154,
+      "learning_rate": 0.00019954493742889648,
+      "loss": 5.4793,
+      "step": 7036
+    },
+    {
+      "epoch": 2.4017064846416383,
+      "grad_norm": 2.8073480129241943,
+      "learning_rate": 0.00019943117178612062,
+      "loss": 5.7709,
+      "step": 7037
+    },
+    {
+      "epoch": 2.4020477815699657,
+      "grad_norm": 2.8629024028778076,
+      "learning_rate": 0.0001993174061433447,
+      "loss": 5.7823,
+      "step": 7038
+    },
+    {
+      "epoch": 2.4023890784982935,
+      "grad_norm": 2.7898664474487305,
+      "learning_rate": 0.00019920364050056883,
+      "loss": 6.2151,
+      "step": 7039
+    },
+    {
+      "epoch": 2.4027303754266214,
+      "grad_norm": 3.052391290664673,
+      "learning_rate": 0.00019908987485779296,
+      "loss": 4.0887,
+      "step": 7040
+    },
+    {
+      "epoch": 2.403071672354949,
+      "grad_norm": 2.876765489578247,
+      "learning_rate": 0.00019897610921501706,
+      "loss": 5.6112,
+      "step": 7041
+    },
+    {
+      "epoch": 2.4034129692832766,
+      "grad_norm": 2.745173931121826,
+      "learning_rate": 0.0001988623435722412,
+      "loss": 6.1771,
+      "step": 7042
+    },
+    {
+      "epoch": 2.403754266211604,
+      "grad_norm": 2.908230781555176,
+      "learning_rate": 0.0001987485779294653,
+      "loss": 5.1346,
+      "step": 7043
+    },
+    {
+      "epoch": 2.404095563139932,
+      "grad_norm": 2.827298402786255,
+      "learning_rate": 0.00019863481228668943,
+      "loss": 6.1908,
+      "step": 7044
+    },
+    {
+      "epoch": 2.4044368600682593,
+      "grad_norm": 2.6762778759002686,
+      "learning_rate": 0.00019852104664391354,
+      "loss": 6.1944,
+      "step": 7045
+    },
+    {
+      "epoch": 2.404778156996587,
+      "grad_norm": 2.9926464557647705,
+      "learning_rate": 0.00019840728100113767,
+      "loss": 4.689,
+      "step": 7046
+    },
+    {
+      "epoch": 2.4051194539249146,
+      "grad_norm": 2.697849750518799,
+      "learning_rate": 0.0001982935153583618,
+      "loss": 6.0584,
+      "step": 7047
+    },
+    {
+      "epoch": 2.4054607508532424,
+      "grad_norm": 3.2848122119903564,
+      "learning_rate": 0.00019817974971558588,
+      "loss": 5.2301,
+      "step": 7048
+    },
+    {
+      "epoch": 2.40580204778157,
+      "grad_norm": 3.0285191535949707,
+      "learning_rate": 0.00019806598407281,
+      "loss": 4.5672,
+      "step": 7049
+    },
+    {
+      "epoch": 2.4061433447098977,
+      "grad_norm": 2.768855094909668,
+      "learning_rate": 0.00019795221843003414,
+      "loss": 6.4452,
+      "step": 7050
+    },
+    {
+      "epoch": 2.406484641638225,
+      "grad_norm": 2.6719319820404053,
+      "learning_rate": 0.00019783845278725825,
+      "loss": 6.2495,
+      "step": 7051
+    },
+    {
+      "epoch": 2.406825938566553,
+      "grad_norm": 2.812721014022827,
+      "learning_rate": 0.00019772468714448235,
+      "loss": 5.9707,
+      "step": 7052
+    },
+    {
+      "epoch": 2.4071672354948808,
+      "grad_norm": 3.698657751083374,
+      "learning_rate": 0.00019761092150170648,
+      "loss": 5.3361,
+      "step": 7053
+    },
+    {
+      "epoch": 2.407508532423208,
+      "grad_norm": 2.74291729927063,
+      "learning_rate": 0.00019749715585893062,
+      "loss": 6.0255,
+      "step": 7054
+    },
+    {
+      "epoch": 2.407849829351536,
+      "grad_norm": 2.770951986312866,
+      "learning_rate": 0.00019738339021615472,
+      "loss": 6.6123,
+      "step": 7055
+    },
+    {
+      "epoch": 2.4081911262798634,
+      "grad_norm": 2.8375635147094727,
+      "learning_rate": 0.00019726962457337885,
+      "loss": 6.4971,
+      "step": 7056
+    },
+    {
+      "epoch": 2.4085324232081913,
+      "grad_norm": 2.986063003540039,
+      "learning_rate": 0.00019715585893060296,
+      "loss": 5.8929,
+      "step": 7057
+    },
+    {
+      "epoch": 2.4088737201365187,
+      "grad_norm": 4.404201030731201,
+      "learning_rate": 0.00019704209328782706,
+      "loss": 4.4985,
+      "step": 7058
+    },
+    {
+      "epoch": 2.4092150170648465,
+      "grad_norm": 2.6988513469696045,
+      "learning_rate": 0.0001969283276450512,
+      "loss": 5.791,
+      "step": 7059
+    },
+    {
+      "epoch": 2.409556313993174,
+      "grad_norm": 2.7154700756073,
+      "learning_rate": 0.00019681456200227533,
+      "loss": 6.4737,
+      "step": 7060
+    },
+    {
+      "epoch": 2.409897610921502,
+      "grad_norm": 2.815117597579956,
+      "learning_rate": 0.00019670079635949943,
+      "loss": 6.0718,
+      "step": 7061
+    },
+    {
+      "epoch": 2.410238907849829,
+      "grad_norm": 2.831260919570923,
+      "learning_rate": 0.00019658703071672354,
+      "loss": 5.3611,
+      "step": 7062
+    },
+    {
+      "epoch": 2.410580204778157,
+      "grad_norm": 1.9146329164505005,
+      "learning_rate": 0.00019647326507394767,
+      "loss": 2.7539,
+      "step": 7063
+    },
+    {
+      "epoch": 2.4109215017064844,
+      "grad_norm": 2.497642755508423,
+      "learning_rate": 0.0001963594994311718,
+      "loss": 5.5512,
+      "step": 7064
+    },
+    {
+      "epoch": 2.4112627986348123,
+      "grad_norm": 2.6266098022460938,
+      "learning_rate": 0.0001962457337883959,
+      "loss": 5.7939,
+      "step": 7065
+    },
+    {
+      "epoch": 2.41160409556314,
+      "grad_norm": 2.7678215503692627,
+      "learning_rate": 0.00019613196814562004,
+      "loss": 6.1091,
+      "step": 7066
+    },
+    {
+      "epoch": 2.4119453924914676,
+      "grad_norm": 3.060725212097168,
+      "learning_rate": 0.00019601820250284414,
+      "loss": 6.1383,
+      "step": 7067
+    },
+    {
+      "epoch": 2.4122866894197954,
+      "grad_norm": 2.825890302658081,
+      "learning_rate": 0.00019590443686006825,
+      "loss": 6.1728,
+      "step": 7068
+    },
+    {
+      "epoch": 2.412627986348123,
+      "grad_norm": 2.731437921524048,
+      "learning_rate": 0.00019579067121729238,
+      "loss": 6.2054,
+      "step": 7069
+    },
+    {
+      "epoch": 2.4129692832764507,
+      "grad_norm": 2.756383180618286,
+      "learning_rate": 0.0001956769055745165,
+      "loss": 5.4519,
+      "step": 7070
+    },
+    {
+      "epoch": 2.413310580204778,
+      "grad_norm": 2.7772936820983887,
+      "learning_rate": 0.00019556313993174062,
+      "loss": 5.739,
+      "step": 7071
+    },
+    {
+      "epoch": 2.413651877133106,
+      "grad_norm": 2.736973285675049,
+      "learning_rate": 0.00019544937428896472,
+      "loss": 5.8474,
+      "step": 7072
+    },
+    {
+      "epoch": 2.4139931740614333,
+      "grad_norm": 2.7424445152282715,
+      "learning_rate": 0.00019533560864618885,
+      "loss": 6.2036,
+      "step": 7073
+    },
+    {
+      "epoch": 2.414334470989761,
+      "grad_norm": 2.801053285598755,
+      "learning_rate": 0.00019522184300341299,
+      "loss": 6.0991,
+      "step": 7074
+    },
+    {
+      "epoch": 2.4146757679180886,
+      "grad_norm": 3.4473876953125,
+      "learning_rate": 0.0001951080773606371,
+      "loss": 5.6659,
+      "step": 7075
+    },
+    {
+      "epoch": 2.4150170648464164,
+      "grad_norm": 2.782585620880127,
+      "learning_rate": 0.00019499431171786122,
+      "loss": 5.653,
+      "step": 7076
+    },
+    {
+      "epoch": 2.415358361774744,
+      "grad_norm": 2.6526317596435547,
+      "learning_rate": 0.00019488054607508533,
+      "loss": 5.88,
+      "step": 7077
+    },
+    {
+      "epoch": 2.4156996587030717,
+      "grad_norm": 2.770357608795166,
+      "learning_rate": 0.00019476678043230943,
+      "loss": 5.7505,
+      "step": 7078
+    },
+    {
+      "epoch": 2.4160409556313995,
+      "grad_norm": 4.893315315246582,
+      "learning_rate": 0.00019465301478953357,
+      "loss": 4.9911,
+      "step": 7079
+    },
+    {
+      "epoch": 2.416382252559727,
+      "grad_norm": 2.6807408332824707,
+      "learning_rate": 0.0001945392491467577,
+      "loss": 6.1131,
+      "step": 7080
+    },
+    {
+      "epoch": 2.416723549488055,
+      "grad_norm": 2.761153221130371,
+      "learning_rate": 0.0001944254835039818,
+      "loss": 5.9863,
+      "step": 7081
+    },
+    {
+      "epoch": 2.417064846416382,
+      "grad_norm": 2.669250965118408,
+      "learning_rate": 0.0001943117178612059,
+      "loss": 5.7527,
+      "step": 7082
+    },
+    {
+      "epoch": 2.41740614334471,
+      "grad_norm": 2.7944650650024414,
+      "learning_rate": 0.00019419795221843004,
+      "loss": 6.0305,
+      "step": 7083
+    },
+    {
+      "epoch": 2.4177474402730375,
+      "grad_norm": 2.851085662841797,
+      "learning_rate": 0.00019408418657565417,
+      "loss": 5.8564,
+      "step": 7084
+    },
+    {
+      "epoch": 2.4180887372013653,
+      "grad_norm": 2.7912724018096924,
+      "learning_rate": 0.00019397042093287828,
+      "loss": 6.6317,
+      "step": 7085
+    },
+    {
+      "epoch": 2.4184300341296927,
+      "grad_norm": 2.802370309829712,
+      "learning_rate": 0.00019385665529010238,
+      "loss": 6.2561,
+      "step": 7086
+    },
+    {
+      "epoch": 2.4187713310580206,
+      "grad_norm": 2.675826072692871,
+      "learning_rate": 0.0001937428896473265,
+      "loss": 6.4923,
+      "step": 7087
+    },
+    {
+      "epoch": 2.419112627986348,
+      "grad_norm": 2.732436418533325,
+      "learning_rate": 0.00019362912400455062,
+      "loss": 5.8842,
+      "step": 7088
+    },
+    {
+      "epoch": 2.419453924914676,
+      "grad_norm": 2.6516783237457275,
+      "learning_rate": 0.00019351535836177475,
+      "loss": 6.1319,
+      "step": 7089
+    },
+    {
+      "epoch": 2.419795221843003,
+      "grad_norm": 2.7802767753601074,
+      "learning_rate": 0.00019340159271899888,
+      "loss": 6.2859,
+      "step": 7090
+    },
+    {
+      "epoch": 2.420136518771331,
+      "grad_norm": 2.645561695098877,
+      "learning_rate": 0.00019328782707622296,
+      "loss": 6.2186,
+      "step": 7091
+    },
+    {
+      "epoch": 2.420477815699659,
+      "grad_norm": 2.762645721435547,
+      "learning_rate": 0.0001931740614334471,
+      "loss": 5.679,
+      "step": 7092
+    },
+    {
+      "epoch": 2.4208191126279863,
+      "grad_norm": 2.7711853981018066,
+      "learning_rate": 0.00019306029579067122,
+      "loss": 6.269,
+      "step": 7093
+    },
+    {
+      "epoch": 2.421160409556314,
+      "grad_norm": 2.817148208618164,
+      "learning_rate": 0.00019294653014789536,
+      "loss": 5.9867,
+      "step": 7094
+    },
+    {
+      "epoch": 2.4215017064846416,
+      "grad_norm": 3.189115285873413,
+      "learning_rate": 0.00019283276450511946,
+      "loss": 5.4579,
+      "step": 7095
+    },
+    {
+      "epoch": 2.4218430034129694,
+      "grad_norm": 2.7254624366760254,
+      "learning_rate": 0.00019271899886234357,
+      "loss": 6.1964,
+      "step": 7096
+    },
+    {
+      "epoch": 2.422184300341297,
+      "grad_norm": 2.8384273052215576,
+      "learning_rate": 0.0001926052332195677,
+      "loss": 5.0824,
+      "step": 7097
+    },
+    {
+      "epoch": 2.4225255972696247,
+      "grad_norm": 2.7569327354431152,
+      "learning_rate": 0.0001924914675767918,
+      "loss": 5.8933,
+      "step": 7098
+    },
+    {
+      "epoch": 2.422866894197952,
+      "grad_norm": 2.6978225708007812,
+      "learning_rate": 0.00019237770193401594,
+      "loss": 4.7832,
+      "step": 7099
+    },
+    {
+      "epoch": 2.42320819112628,
+      "grad_norm": 2.7177062034606934,
+      "learning_rate": 0.00019226393629124007,
+      "loss": 5.4131,
+      "step": 7100
+    },
+    {
+      "epoch": 2.4235494880546073,
+      "grad_norm": 2.7541306018829346,
+      "learning_rate": 0.00019215017064846415,
+      "loss": 6.3098,
+      "step": 7101
+    },
+    {
+      "epoch": 2.423890784982935,
+      "grad_norm": 2.7490100860595703,
+      "learning_rate": 0.00019203640500568828,
+      "loss": 6.5531,
+      "step": 7102
+    },
+    {
+      "epoch": 2.4242320819112626,
+      "grad_norm": 2.74153208732605,
+      "learning_rate": 0.0001919226393629124,
+      "loss": 6.1171,
+      "step": 7103
+    },
+    {
+      "epoch": 2.4245733788395905,
+      "grad_norm": 2.7653117179870605,
+      "learning_rate": 0.00019180887372013654,
+      "loss": 5.8081,
+      "step": 7104
+    },
+    {
+      "epoch": 2.4249146757679183,
+      "grad_norm": 2.7591192722320557,
+      "learning_rate": 0.00019169510807736065,
+      "loss": 5.5008,
+      "step": 7105
+    },
+    {
+      "epoch": 2.4252559726962457,
+      "grad_norm": 2.5488815307617188,
+      "learning_rate": 0.00019158134243458475,
+      "loss": 6.4281,
+      "step": 7106
+    },
+    {
+      "epoch": 2.4255972696245736,
+      "grad_norm": 2.22991943359375,
+      "learning_rate": 0.00019146757679180888,
+      "loss": 2.1561,
+      "step": 7107
+    },
+    {
+      "epoch": 2.425938566552901,
+      "grad_norm": 3.231452465057373,
+      "learning_rate": 0.000191353811149033,
+      "loss": 5.4485,
+      "step": 7108
+    },
+    {
+      "epoch": 2.426279863481229,
+      "grad_norm": 3.1210076808929443,
+      "learning_rate": 0.00019124004550625712,
+      "loss": 4.5784,
+      "step": 7109
+    },
+    {
+      "epoch": 2.426621160409556,
+      "grad_norm": 2.622694253921509,
+      "learning_rate": 0.00019112627986348125,
+      "loss": 5.6532,
+      "step": 7110
+    },
+    {
+      "epoch": 2.426962457337884,
+      "grad_norm": 2.8294200897216797,
+      "learning_rate": 0.00019101251422070533,
+      "loss": 6.0965,
+      "step": 7111
+    },
+    {
+      "epoch": 2.4273037542662115,
+      "grad_norm": 2.737205982208252,
+      "learning_rate": 0.00019089874857792946,
+      "loss": 6.3284,
+      "step": 7112
+    },
+    {
+      "epoch": 2.4276450511945393,
+      "grad_norm": 2.6656839847564697,
+      "learning_rate": 0.0001907849829351536,
+      "loss": 6.3298,
+      "step": 7113
+    },
+    {
+      "epoch": 2.4279863481228667,
+      "grad_norm": 2.7031219005584717,
+      "learning_rate": 0.00019067121729237773,
+      "loss": 6.1422,
+      "step": 7114
+    },
+    {
+      "epoch": 2.4283276450511946,
+      "grad_norm": 2.792492389678955,
+      "learning_rate": 0.00019055745164960183,
+      "loss": 6.1246,
+      "step": 7115
+    },
+    {
+      "epoch": 2.428668941979522,
+      "grad_norm": 2.8730626106262207,
+      "learning_rate": 0.00019044368600682594,
+      "loss": 5.2022,
+      "step": 7116
+    },
+    {
+      "epoch": 2.42901023890785,
+      "grad_norm": 2.832547903060913,
+      "learning_rate": 0.00019032992036405007,
+      "loss": 5.8241,
+      "step": 7117
+    },
+    {
+      "epoch": 2.4293515358361777,
+      "grad_norm": 2.7533719539642334,
+      "learning_rate": 0.00019021615472127417,
+      "loss": 6.1243,
+      "step": 7118
+    },
+    {
+      "epoch": 2.429692832764505,
+      "grad_norm": 2.7654356956481934,
+      "learning_rate": 0.0001901023890784983,
+      "loss": 6.5303,
+      "step": 7119
+    },
+    {
+      "epoch": 2.430034129692833,
+      "grad_norm": 2.7612063884735107,
+      "learning_rate": 0.0001899886234357224,
+      "loss": 6.8023,
+      "step": 7120
+    },
+    {
+      "epoch": 2.4303754266211604,
+      "grad_norm": 2.8074557781219482,
+      "learning_rate": 0.00018987485779294652,
+      "loss": 5.4001,
+      "step": 7121
+    },
+    {
+      "epoch": 2.430716723549488,
+      "grad_norm": 2.6893937587738037,
+      "learning_rate": 0.00018976109215017065,
+      "loss": 5.6546,
+      "step": 7122
+    },
+    {
+      "epoch": 2.4310580204778156,
+      "grad_norm": 2.6178319454193115,
+      "learning_rate": 0.00018964732650739478,
+      "loss": 6.3508,
+      "step": 7123
+    },
+    {
+      "epoch": 2.4313993174061435,
+      "grad_norm": 2.972472906112671,
+      "learning_rate": 0.0001895335608646189,
+      "loss": 5.156,
+      "step": 7124
+    },
+    {
+      "epoch": 2.431740614334471,
+      "grad_norm": 2.7686290740966797,
+      "learning_rate": 0.000189419795221843,
+      "loss": 6.5945,
+      "step": 7125
+    },
+    {
+      "epoch": 2.4320819112627987,
+      "grad_norm": 2.8814847469329834,
+      "learning_rate": 0.00018930602957906712,
+      "loss": 5.6923,
+      "step": 7126
+    },
+    {
+      "epoch": 2.432423208191126,
+      "grad_norm": 1.9587899446487427,
+      "learning_rate": 0.00018919226393629125,
+      "loss": 3.0517,
+      "step": 7127
+    },
+    {
+      "epoch": 2.432764505119454,
+      "grad_norm": 2.7158172130584717,
+      "learning_rate": 0.00018907849829351536,
+      "loss": 5.734,
+      "step": 7128
+    },
+    {
+      "epoch": 2.4331058020477814,
+      "grad_norm": 2.719918727874756,
+      "learning_rate": 0.0001889647326507395,
+      "loss": 6.0725,
+      "step": 7129
+    },
+    {
+      "epoch": 2.4334470989761092,
+      "grad_norm": 3.0201733112335205,
+      "learning_rate": 0.0001888509670079636,
+      "loss": 5.2013,
+      "step": 7130
+    },
+    {
+      "epoch": 2.433788395904437,
+      "grad_norm": 2.701390266418457,
+      "learning_rate": 0.0001887372013651877,
+      "loss": 5.9561,
+      "step": 7131
+    },
+    {
+      "epoch": 2.4341296928327645,
+      "grad_norm": 2.7162129878997803,
+      "learning_rate": 0.00018862343572241183,
+      "loss": 5.5845,
+      "step": 7132
+    },
+    {
+      "epoch": 2.4344709897610923,
+      "grad_norm": 2.755629777908325,
+      "learning_rate": 0.00018850967007963596,
+      "loss": 5.9108,
+      "step": 7133
+    },
+    {
+      "epoch": 2.4348122866894197,
+      "grad_norm": 2.744990110397339,
+      "learning_rate": 0.0001883959044368601,
+      "loss": 6.0868,
+      "step": 7134
+    },
+    {
+      "epoch": 2.4351535836177476,
+      "grad_norm": 2.6877083778381348,
+      "learning_rate": 0.00018828213879408417,
+      "loss": 6.2003,
+      "step": 7135
+    },
+    {
+      "epoch": 2.435494880546075,
+      "grad_norm": 2.8432061672210693,
+      "learning_rate": 0.0001881683731513083,
+      "loss": 4.8051,
+      "step": 7136
+    },
+    {
+      "epoch": 2.435836177474403,
+      "grad_norm": 2.6555399894714355,
+      "learning_rate": 0.00018805460750853244,
+      "loss": 6.0686,
+      "step": 7137
+    },
+    {
+      "epoch": 2.4361774744027302,
+      "grad_norm": 2.7505736351013184,
+      "learning_rate": 0.00018794084186575654,
+      "loss": 6.1821,
+      "step": 7138
+    },
+    {
+      "epoch": 2.436518771331058,
+      "grad_norm": 2.7092573642730713,
+      "learning_rate": 0.00018782707622298068,
+      "loss": 6.1421,
+      "step": 7139
+    },
+    {
+      "epoch": 2.4368600682593855,
+      "grad_norm": 2.7411575317382812,
+      "learning_rate": 0.00018771331058020478,
+      "loss": 5.4851,
+      "step": 7140
+    },
+    {
+      "epoch": 2.4372013651877134,
+      "grad_norm": 2.853494882583618,
+      "learning_rate": 0.00018759954493742888,
+      "loss": 5.2069,
+      "step": 7141
+    },
+    {
+      "epoch": 2.4375426621160408,
+      "grad_norm": 2.8510758876800537,
+      "learning_rate": 0.00018748577929465302,
+      "loss": 5.1078,
+      "step": 7142
+    },
+    {
+      "epoch": 2.4378839590443686,
+      "grad_norm": 2.6476118564605713,
+      "learning_rate": 0.00018737201365187715,
+      "loss": 5.4156,
+      "step": 7143
+    },
+    {
+      "epoch": 2.4382252559726965,
+      "grad_norm": 2.692403793334961,
+      "learning_rate": 0.00018725824800910125,
+      "loss": 5.7651,
+      "step": 7144
+    },
+    {
+      "epoch": 2.438566552901024,
+      "grad_norm": 3.758758783340454,
+      "learning_rate": 0.00018714448236632536,
+      "loss": 5.061,
+      "step": 7145
+    },
+    {
+      "epoch": 2.4389078498293517,
+      "grad_norm": 2.8809521198272705,
+      "learning_rate": 0.0001870307167235495,
+      "loss": 5.8695,
+      "step": 7146
+    },
+    {
+      "epoch": 2.439249146757679,
+      "grad_norm": 3.381040573120117,
+      "learning_rate": 0.00018691695108077362,
+      "loss": 3.9914,
+      "step": 7147
+    },
+    {
+      "epoch": 2.439590443686007,
+      "grad_norm": 2.7137439250946045,
+      "learning_rate": 0.00018680318543799773,
+      "loss": 6.4293,
+      "step": 7148
+    },
+    {
+      "epoch": 2.4399317406143344,
+      "grad_norm": 2.7312674522399902,
+      "learning_rate": 0.00018668941979522186,
+      "loss": 5.4181,
+      "step": 7149
+    },
+    {
+      "epoch": 2.4402730375426622,
+      "grad_norm": 2.8323845863342285,
+      "learning_rate": 0.00018657565415244596,
+      "loss": 6.078,
+      "step": 7150
+    },
+    {
+      "epoch": 2.4406143344709896,
+      "grad_norm": 2.6376965045928955,
+      "learning_rate": 0.00018646188850967007,
+      "loss": 5.5007,
+      "step": 7151
+    },
+    {
+      "epoch": 2.4409556313993175,
+      "grad_norm": 2.6789960861206055,
+      "learning_rate": 0.0001863481228668942,
+      "loss": 5.7487,
+      "step": 7152
+    },
+    {
+      "epoch": 2.441296928327645,
+      "grad_norm": 2.624680995941162,
+      "learning_rate": 0.00018623435722411833,
+      "loss": 6.5413,
+      "step": 7153
+    },
+    {
+      "epoch": 2.4416382252559727,
+      "grad_norm": 2.9074723720550537,
+      "learning_rate": 0.00018612059158134244,
+      "loss": 5.3683,
+      "step": 7154
+    },
+    {
+      "epoch": 2.4419795221843,
+      "grad_norm": 2.71537446975708,
+      "learning_rate": 0.00018600682593856654,
+      "loss": 5.7032,
+      "step": 7155
+    },
+    {
+      "epoch": 2.442320819112628,
+      "grad_norm": 3.114678144454956,
+      "learning_rate": 0.00018589306029579068,
+      "loss": 5.9269,
+      "step": 7156
+    },
+    {
+      "epoch": 2.442662116040956,
+      "grad_norm": 2.7557051181793213,
+      "learning_rate": 0.0001857792946530148,
+      "loss": 6.3845,
+      "step": 7157
+    },
+    {
+      "epoch": 2.4430034129692833,
+      "grad_norm": 2.7388687133789062,
+      "learning_rate": 0.0001856655290102389,
+      "loss": 6.0723,
+      "step": 7158
+    },
+    {
+      "epoch": 2.443344709897611,
+      "grad_norm": 2.7414541244506836,
+      "learning_rate": 0.00018555176336746302,
+      "loss": 6.4165,
+      "step": 7159
+    },
+    {
+      "epoch": 2.4436860068259385,
+      "grad_norm": 2.7356631755828857,
+      "learning_rate": 0.00018543799772468715,
+      "loss": 6.1855,
+      "step": 7160
+    },
+    {
+      "epoch": 2.4440273037542664,
+      "grad_norm": 2.845184564590454,
+      "learning_rate": 0.00018532423208191125,
+      "loss": 5.8645,
+      "step": 7161
+    },
+    {
+      "epoch": 2.4443686006825938,
+      "grad_norm": 2.6517839431762695,
+      "learning_rate": 0.0001852104664391354,
+      "loss": 6.2252,
+      "step": 7162
+    },
+    {
+      "epoch": 2.4447098976109216,
+      "grad_norm": 2.7228269577026367,
+      "learning_rate": 0.00018509670079635952,
+      "loss": 6.3114,
+      "step": 7163
+    },
+    {
+      "epoch": 2.445051194539249,
+      "grad_norm": 2.680921792984009,
+      "learning_rate": 0.0001849829351535836,
+      "loss": 6.3173,
+      "step": 7164
+    },
+    {
+      "epoch": 2.445392491467577,
+      "grad_norm": 2.817765235900879,
+      "learning_rate": 0.00018486916951080773,
+      "loss": 6.0299,
+      "step": 7165
+    },
+    {
+      "epoch": 2.4457337883959043,
+      "grad_norm": 4.926273822784424,
+      "learning_rate": 0.00018475540386803186,
+      "loss": 5.256,
+      "step": 7166
+    },
+    {
+      "epoch": 2.446075085324232,
+      "grad_norm": 2.7189366817474365,
+      "learning_rate": 0.000184641638225256,
+      "loss": 5.6313,
+      "step": 7167
+    },
+    {
+      "epoch": 2.4464163822525595,
+      "grad_norm": 6.03790807723999,
+      "learning_rate": 0.0001845278725824801,
+      "loss": 4.8653,
+      "step": 7168
+    },
+    {
+      "epoch": 2.4467576791808874,
+      "grad_norm": 2.8209898471832275,
+      "learning_rate": 0.0001844141069397042,
+      "loss": 6.1202,
+      "step": 7169
+    },
+    {
+      "epoch": 2.4470989761092152,
+      "grad_norm": 2.895895004272461,
+      "learning_rate": 0.00018430034129692833,
+      "loss": 5.8068,
+      "step": 7170
+    },
+    {
+      "epoch": 2.4474402730375426,
+      "grad_norm": 2.669280529022217,
+      "learning_rate": 0.00018418657565415244,
+      "loss": 5.6466,
+      "step": 7171
+    },
+    {
+      "epoch": 2.4477815699658705,
+      "grad_norm": 2.70436429977417,
+      "learning_rate": 0.00018407281001137657,
+      "loss": 5.8402,
+      "step": 7172
+    },
+    {
+      "epoch": 2.448122866894198,
+      "grad_norm": 2.793987989425659,
+      "learning_rate": 0.0001839590443686007,
+      "loss": 6.507,
+      "step": 7173
+    },
+    {
+      "epoch": 2.4484641638225257,
+      "grad_norm": 2.6550915241241455,
+      "learning_rate": 0.00018384527872582478,
+      "loss": 5.8959,
+      "step": 7174
+    },
+    {
+      "epoch": 2.448805460750853,
+      "grad_norm": 2.6808414459228516,
+      "learning_rate": 0.00018373151308304891,
+      "loss": 6.6288,
+      "step": 7175
+    },
+    {
+      "epoch": 2.449146757679181,
+      "grad_norm": 2.6597981452941895,
+      "learning_rate": 0.00018361774744027305,
+      "loss": 5.4804,
+      "step": 7176
+    },
+    {
+      "epoch": 2.4494880546075084,
+      "grad_norm": 2.7056760787963867,
+      "learning_rate": 0.00018350398179749718,
+      "loss": 6.5242,
+      "step": 7177
+    },
+    {
+      "epoch": 2.4498293515358363,
+      "grad_norm": 2.821537733078003,
+      "learning_rate": 0.00018339021615472128,
+      "loss": 5.9898,
+      "step": 7178
+    },
+    {
+      "epoch": 2.4501706484641637,
+      "grad_norm": 4.618656635284424,
+      "learning_rate": 0.0001832764505119454,
+      "loss": 5.2683,
+      "step": 7179
+    },
+    {
+      "epoch": 2.4505119453924915,
+      "grad_norm": 2.7883718013763428,
+      "learning_rate": 0.00018316268486916952,
+      "loss": 6.4302,
+      "step": 7180
+    },
+    {
+      "epoch": 2.450853242320819,
+      "grad_norm": 2.782613515853882,
+      "learning_rate": 0.00018304891922639362,
+      "loss": 6.0204,
+      "step": 7181
+    },
+    {
+      "epoch": 2.4511945392491468,
+      "grad_norm": 3.4366047382354736,
+      "learning_rate": 0.00018293515358361776,
+      "loss": 4.9788,
+      "step": 7182
+    },
+    {
+      "epoch": 2.4515358361774746,
+      "grad_norm": 2.758683681488037,
+      "learning_rate": 0.0001828213879408419,
+      "loss": 6.1169,
+      "step": 7183
+    },
+    {
+      "epoch": 2.451877133105802,
+      "grad_norm": 2.7406606674194336,
+      "learning_rate": 0.00018270762229806597,
+      "loss": 6.1129,
+      "step": 7184
+    },
+    {
+      "epoch": 2.45221843003413,
+      "grad_norm": 2.7950448989868164,
+      "learning_rate": 0.0001825938566552901,
+      "loss": 6.2123,
+      "step": 7185
+    },
+    {
+      "epoch": 2.4525597269624573,
+      "grad_norm": 2.6972243785858154,
+      "learning_rate": 0.00018248009101251423,
+      "loss": 5.9016,
+      "step": 7186
+    },
+    {
+      "epoch": 2.452901023890785,
+      "grad_norm": 2.7879507541656494,
+      "learning_rate": 0.00018236632536973836,
+      "loss": 6.3663,
+      "step": 7187
+    },
+    {
+      "epoch": 2.4532423208191125,
+      "grad_norm": 2.7534303665161133,
+      "learning_rate": 0.00018225255972696247,
+      "loss": 6.1968,
+      "step": 7188
+    },
+    {
+      "epoch": 2.4535836177474404,
+      "grad_norm": 2.755431652069092,
+      "learning_rate": 0.00018213879408418657,
+      "loss": 5.782,
+      "step": 7189
+    },
+    {
+      "epoch": 2.453924914675768,
+      "grad_norm": 2.6141881942749023,
+      "learning_rate": 0.0001820250284414107,
+      "loss": 5.532,
+      "step": 7190
+    },
+    {
+      "epoch": 2.4542662116040956,
+      "grad_norm": 3.2307145595550537,
+      "learning_rate": 0.0001819112627986348,
+      "loss": 5.0953,
+      "step": 7191
+    },
+    {
+      "epoch": 2.454607508532423,
+      "grad_norm": 2.6708171367645264,
+      "learning_rate": 0.00018179749715585894,
+      "loss": 6.0323,
+      "step": 7192
+    },
+    {
+      "epoch": 2.454948805460751,
+      "grad_norm": 2.6066534519195557,
+      "learning_rate": 0.00018168373151308305,
+      "loss": 6.0203,
+      "step": 7193
+    },
+    {
+      "epoch": 2.4552901023890783,
+      "grad_norm": 2.7354557514190674,
+      "learning_rate": 0.00018156996587030715,
+      "loss": 6.1075,
+      "step": 7194
+    },
+    {
+      "epoch": 2.455631399317406,
+      "grad_norm": 2.2138519287109375,
+      "learning_rate": 0.00018145620022753128,
+      "loss": 4.2661,
+      "step": 7195
+    },
+    {
+      "epoch": 2.455972696245734,
+      "grad_norm": 2.87853741645813,
+      "learning_rate": 0.00018134243458475542,
+      "loss": 6.0552,
+      "step": 7196
+    },
+    {
+      "epoch": 2.4563139931740614,
+      "grad_norm": 2.7033181190490723,
+      "learning_rate": 0.00018122866894197955,
+      "loss": 6.5583,
+      "step": 7197
+    },
+    {
+      "epoch": 2.4566552901023893,
+      "grad_norm": 2.6118319034576416,
+      "learning_rate": 0.00018111490329920363,
+      "loss": 6.1334,
+      "step": 7198
+    },
+    {
+      "epoch": 2.4569965870307167,
+      "grad_norm": 2.641845941543579,
+      "learning_rate": 0.00018100113765642776,
+      "loss": 6.1401,
+      "step": 7199
+    },
+    {
+      "epoch": 2.4573378839590445,
+      "grad_norm": 2.8019094467163086,
+      "learning_rate": 0.0001808873720136519,
+      "loss": 5.518,
+      "step": 7200
+    },
+    {
+      "epoch": 2.457679180887372,
+      "grad_norm": 4.355568885803223,
+      "learning_rate": 0.000180773606370876,
+      "loss": 4.7298,
+      "step": 7201
+    },
+    {
+      "epoch": 2.4580204778156998,
+      "grad_norm": 3.0501973628997803,
+      "learning_rate": 0.00018065984072810013,
+      "loss": 5.1193,
+      "step": 7202
+    },
+    {
+      "epoch": 2.458361774744027,
+      "grad_norm": 2.716735363006592,
+      "learning_rate": 0.00018054607508532423,
+      "loss": 5.868,
+      "step": 7203
+    },
+    {
+      "epoch": 2.458703071672355,
+      "grad_norm": 2.653223991394043,
+      "learning_rate": 0.00018043230944254834,
+      "loss": 5.8835,
+      "step": 7204
+    },
+    {
+      "epoch": 2.4590443686006824,
+      "grad_norm": 2.7722156047821045,
+      "learning_rate": 0.00018031854379977247,
+      "loss": 6.1643,
+      "step": 7205
+    },
+    {
+      "epoch": 2.4593856655290103,
+      "grad_norm": 4.1703619956970215,
+      "learning_rate": 0.0001802047781569966,
+      "loss": 5.0742,
+      "step": 7206
+    },
+    {
+      "epoch": 2.4597269624573377,
+      "grad_norm": 2.799798011779785,
+      "learning_rate": 0.00018009101251422073,
+      "loss": 5.3693,
+      "step": 7207
+    },
+    {
+      "epoch": 2.4600682593856655,
+      "grad_norm": 2.693559169769287,
+      "learning_rate": 0.0001799772468714448,
+      "loss": 6.1607,
+      "step": 7208
+    },
+    {
+      "epoch": 2.4604095563139934,
+      "grad_norm": 2.786590576171875,
+      "learning_rate": 0.00017986348122866894,
+      "loss": 6.3078,
+      "step": 7209
+    },
+    {
+      "epoch": 2.460750853242321,
+      "grad_norm": 6.34663200378418,
+      "learning_rate": 0.00017974971558589307,
+      "loss": 3.7846,
+      "step": 7210
+    },
+    {
+      "epoch": 2.4610921501706486,
+      "grad_norm": 2.779407024383545,
+      "learning_rate": 0.00017963594994311718,
+      "loss": 6.0981,
+      "step": 7211
+    },
+    {
+      "epoch": 2.461433447098976,
+      "grad_norm": 2.8530948162078857,
+      "learning_rate": 0.0001795221843003413,
+      "loss": 6.5747,
+      "step": 7212
+    },
+    {
+      "epoch": 2.461774744027304,
+      "grad_norm": 2.747483491897583,
+      "learning_rate": 0.00017940841865756542,
+      "loss": 6.037,
+      "step": 7213
+    },
+    {
+      "epoch": 2.4621160409556313,
+      "grad_norm": 2.689389228820801,
+      "learning_rate": 0.00017929465301478952,
+      "loss": 5.8297,
+      "step": 7214
+    },
+    {
+      "epoch": 2.462457337883959,
+      "grad_norm": 2.734438419342041,
+      "learning_rate": 0.00017918088737201365,
+      "loss": 6.2479,
+      "step": 7215
+    },
+    {
+      "epoch": 2.4627986348122866,
+      "grad_norm": 5.3711466789245605,
+      "learning_rate": 0.00017906712172923779,
+      "loss": 4.8514,
+      "step": 7216
+    },
+    {
+      "epoch": 2.4631399317406144,
+      "grad_norm": 2.606839418411255,
+      "learning_rate": 0.00017895335608646192,
+      "loss": 6.2493,
+      "step": 7217
+    },
+    {
+      "epoch": 2.463481228668942,
+      "grad_norm": 2.6620965003967285,
+      "learning_rate": 0.000178839590443686,
+      "loss": 6.1907,
+      "step": 7218
+    },
+    {
+      "epoch": 2.4638225255972697,
+      "grad_norm": 2.8028299808502197,
+      "learning_rate": 0.00017872582480091013,
+      "loss": 6.3521,
+      "step": 7219
+    },
+    {
+      "epoch": 2.464163822525597,
+      "grad_norm": 2.6073973178863525,
+      "learning_rate": 0.00017861205915813426,
+      "loss": 5.6279,
+      "step": 7220
+    },
+    {
+      "epoch": 2.464505119453925,
+      "grad_norm": 3.001814603805542,
+      "learning_rate": 0.00017849829351535836,
+      "loss": 4.9883,
+      "step": 7221
+    },
+    {
+      "epoch": 2.4648464163822528,
+      "grad_norm": 2.9250714778900146,
+      "learning_rate": 0.0001783845278725825,
+      "loss": 5.5733,
+      "step": 7222
+    },
+    {
+      "epoch": 2.46518771331058,
+      "grad_norm": 2.743499994277954,
+      "learning_rate": 0.0001782707622298066,
+      "loss": 6.2804,
+      "step": 7223
+    },
+    {
+      "epoch": 2.465529010238908,
+      "grad_norm": 2.7742860317230225,
+      "learning_rate": 0.0001781569965870307,
+      "loss": 6.0203,
+      "step": 7224
+    },
+    {
+      "epoch": 2.4658703071672354,
+      "grad_norm": 2.818325996398926,
+      "learning_rate": 0.00017804323094425484,
+      "loss": 5.6781,
+      "step": 7225
+    },
+    {
+      "epoch": 2.4662116040955633,
+      "grad_norm": 2.795591115951538,
+      "learning_rate": 0.00017792946530147897,
+      "loss": 5.602,
+      "step": 7226
+    },
+    {
+      "epoch": 2.4665529010238907,
+      "grad_norm": 2.807055950164795,
+      "learning_rate": 0.00017781569965870305,
+      "loss": 5.1593,
+      "step": 7227
+    },
+    {
+      "epoch": 2.4668941979522185,
+      "grad_norm": 2.7123217582702637,
+      "learning_rate": 0.00017770193401592718,
+      "loss": 6.4722,
+      "step": 7228
+    },
+    {
+      "epoch": 2.467235494880546,
+      "grad_norm": 2.759861469268799,
+      "learning_rate": 0.0001775881683731513,
+      "loss": 5.4875,
+      "step": 7229
+    },
+    {
+      "epoch": 2.467576791808874,
+      "grad_norm": 2.78224515914917,
+      "learning_rate": 0.00017747440273037544,
+      "loss": 5.79,
+      "step": 7230
+    },
+    {
+      "epoch": 2.467918088737201,
+      "grad_norm": 2.7341177463531494,
+      "learning_rate": 0.00017736063708759955,
+      "loss": 5.9504,
+      "step": 7231
+    },
+    {
+      "epoch": 2.468259385665529,
+      "grad_norm": 2.6890923976898193,
+      "learning_rate": 0.00017724687144482365,
+      "loss": 6.5827,
+      "step": 7232
+    },
+    {
+      "epoch": 2.4686006825938565,
+      "grad_norm": 2.8176991939544678,
+      "learning_rate": 0.0001771331058020478,
+      "loss": 5.8202,
+      "step": 7233
+    },
+    {
+      "epoch": 2.4689419795221843,
+      "grad_norm": 2.746105670928955,
+      "learning_rate": 0.0001770193401592719,
+      "loss": 6.5549,
+      "step": 7234
+    },
+    {
+      "epoch": 2.469283276450512,
+      "grad_norm": 2.7601068019866943,
+      "learning_rate": 0.00017690557451649602,
+      "loss": 5.8594,
+      "step": 7235
+    },
+    {
+      "epoch": 2.4696245733788396,
+      "grad_norm": 2.550632953643799,
+      "learning_rate": 0.00017679180887372016,
+      "loss": 4.9317,
+      "step": 7236
+    },
+    {
+      "epoch": 2.4699658703071674,
+      "grad_norm": 2.6904959678649902,
+      "learning_rate": 0.00017667804323094423,
+      "loss": 6.5549,
+      "step": 7237
+    },
+    {
+      "epoch": 2.470307167235495,
+      "grad_norm": 2.70029616355896,
+      "learning_rate": 0.00017656427758816837,
+      "loss": 5.7531,
+      "step": 7238
+    },
+    {
+      "epoch": 2.4706484641638227,
+      "grad_norm": 2.799412250518799,
+      "learning_rate": 0.0001764505119453925,
+      "loss": 5.6873,
+      "step": 7239
+    },
+    {
+      "epoch": 2.47098976109215,
+      "grad_norm": 2.8735549449920654,
+      "learning_rate": 0.00017633674630261663,
+      "loss": 4.2249,
+      "step": 7240
+    },
+    {
+      "epoch": 2.471331058020478,
+      "grad_norm": 2.010611057281494,
+      "learning_rate": 0.00017622298065984073,
+      "loss": 2.7766,
+      "step": 7241
+    },
+    {
+      "epoch": 2.4716723549488053,
+      "grad_norm": 2.699479579925537,
+      "learning_rate": 0.00017610921501706484,
+      "loss": 5.8529,
+      "step": 7242
+    },
+    {
+      "epoch": 2.472013651877133,
+      "grad_norm": 2.7273783683776855,
+      "learning_rate": 0.00017599544937428897,
+      "loss": 6.4994,
+      "step": 7243
+    },
+    {
+      "epoch": 2.4723549488054606,
+      "grad_norm": 2.768795967102051,
+      "learning_rate": 0.00017588168373151308,
+      "loss": 6.4167,
+      "step": 7244
+    },
+    {
+      "epoch": 2.4726962457337884,
+      "grad_norm": 2.7882094383239746,
+      "learning_rate": 0.0001757679180887372,
+      "loss": 5.8934,
+      "step": 7245
+    },
+    {
+      "epoch": 2.473037542662116,
+      "grad_norm": 2.8088643550872803,
+      "learning_rate": 0.00017565415244596134,
+      "loss": 6.0832,
+      "step": 7246
+    },
+    {
+      "epoch": 2.4733788395904437,
+      "grad_norm": 2.7583882808685303,
+      "learning_rate": 0.00017554038680318542,
+      "loss": 6.4278,
+      "step": 7247
+    },
+    {
+      "epoch": 2.4737201365187715,
+      "grad_norm": 2.710397720336914,
+      "learning_rate": 0.00017542662116040955,
+      "loss": 5.9064,
+      "step": 7248
+    },
+    {
+      "epoch": 2.474061433447099,
+      "grad_norm": 3.8351919651031494,
+      "learning_rate": 0.00017531285551763368,
+      "loss": 5.1969,
+      "step": 7249
+    },
+    {
+      "epoch": 2.474402730375427,
+      "grad_norm": 2.7258529663085938,
+      "learning_rate": 0.00017519908987485781,
+      "loss": 5.6392,
+      "step": 7250
+    },
+    {
+      "epoch": 2.474744027303754,
+      "grad_norm": 2.758955717086792,
+      "learning_rate": 0.00017508532423208192,
+      "loss": 6.3959,
+      "step": 7251
+    },
+    {
+      "epoch": 2.475085324232082,
+      "grad_norm": 2.727679967880249,
+      "learning_rate": 0.00017497155858930602,
+      "loss": 5.689,
+      "step": 7252
+    },
+    {
+      "epoch": 2.4754266211604095,
+      "grad_norm": 2.7123825550079346,
+      "learning_rate": 0.00017485779294653016,
+      "loss": 6.4214,
+      "step": 7253
+    },
+    {
+      "epoch": 2.4757679180887373,
+      "grad_norm": 2.7021186351776123,
+      "learning_rate": 0.00017474402730375426,
+      "loss": 5.8807,
+      "step": 7254
+    },
+    {
+      "epoch": 2.4761092150170647,
+      "grad_norm": 2.6904120445251465,
+      "learning_rate": 0.0001746302616609784,
+      "loss": 6.3117,
+      "step": 7255
+    },
+    {
+      "epoch": 2.4764505119453926,
+      "grad_norm": 2.68574595451355,
+      "learning_rate": 0.00017451649601820253,
+      "loss": 6.2527,
+      "step": 7256
+    },
+    {
+      "epoch": 2.47679180887372,
+      "grad_norm": 2.733830213546753,
+      "learning_rate": 0.0001744027303754266,
+      "loss": 6.2249,
+      "step": 7257
+    },
+    {
+      "epoch": 2.477133105802048,
+      "grad_norm": 2.7638471126556396,
+      "learning_rate": 0.00017428896473265074,
+      "loss": 6.5467,
+      "step": 7258
+    },
+    {
+      "epoch": 2.4774744027303752,
+      "grad_norm": 2.627199172973633,
+      "learning_rate": 0.00017417519908987487,
+      "loss": 6.2666,
+      "step": 7259
+    },
+    {
+      "epoch": 2.477815699658703,
+      "grad_norm": 2.6954307556152344,
+      "learning_rate": 0.000174061433447099,
+      "loss": 5.407,
+      "step": 7260
+    },
+    {
+      "epoch": 2.478156996587031,
+      "grad_norm": 2.6798486709594727,
+      "learning_rate": 0.00017394766780432308,
+      "loss": 6.3605,
+      "step": 7261
+    },
+    {
+      "epoch": 2.4784982935153583,
+      "grad_norm": 2.6780683994293213,
+      "learning_rate": 0.0001738339021615472,
+      "loss": 5.8869,
+      "step": 7262
+    },
+    {
+      "epoch": 2.478839590443686,
+      "grad_norm": 2.620546340942383,
+      "learning_rate": 0.00017372013651877134,
+      "loss": 6.0429,
+      "step": 7263
+    },
+    {
+      "epoch": 2.4791808873720136,
+      "grad_norm": 3.052701711654663,
+      "learning_rate": 0.00017360637087599545,
+      "loss": 5.4183,
+      "step": 7264
+    },
+    {
+      "epoch": 2.4795221843003414,
+      "grad_norm": 2.6496641635894775,
+      "learning_rate": 0.00017349260523321958,
+      "loss": 5.8464,
+      "step": 7265
+    },
+    {
+      "epoch": 2.479863481228669,
+      "grad_norm": 2.715848684310913,
+      "learning_rate": 0.00017337883959044368,
+      "loss": 6.6531,
+      "step": 7266
+    },
+    {
+      "epoch": 2.4802047781569967,
+      "grad_norm": 2.6739721298217773,
+      "learning_rate": 0.0001732650739476678,
+      "loss": 6.264,
+      "step": 7267
+    },
+    {
+      "epoch": 2.480546075085324,
+      "grad_norm": 2.7070438861846924,
+      "learning_rate": 0.00017315130830489192,
+      "loss": 6.047,
+      "step": 7268
+    },
+    {
+      "epoch": 2.480887372013652,
+      "grad_norm": 2.752788782119751,
+      "learning_rate": 0.00017303754266211605,
+      "loss": 6.0782,
+      "step": 7269
+    },
+    {
+      "epoch": 2.4812286689419794,
+      "grad_norm": 2.7496471405029297,
+      "learning_rate": 0.00017292377701934018,
+      "loss": 6.4748,
+      "step": 7270
+    },
+    {
+      "epoch": 2.481569965870307,
+      "grad_norm": 2.7307982444763184,
+      "learning_rate": 0.00017281001137656426,
+      "loss": 6.6512,
+      "step": 7271
+    },
+    {
+      "epoch": 2.4819112627986346,
+      "grad_norm": 2.725822925567627,
+      "learning_rate": 0.0001726962457337884,
+      "loss": 6.5072,
+      "step": 7272
+    },
+    {
+      "epoch": 2.4822525597269625,
+      "grad_norm": 2.7123241424560547,
+      "learning_rate": 0.00017258248009101253,
+      "loss": 5.8875,
+      "step": 7273
+    },
+    {
+      "epoch": 2.4825938566552903,
+      "grad_norm": 2.633855104446411,
+      "learning_rate": 0.00017246871444823663,
+      "loss": 6.0088,
+      "step": 7274
+    },
+    {
+      "epoch": 2.4829351535836177,
+      "grad_norm": 4.553799629211426,
+      "learning_rate": 0.00017235494880546076,
+      "loss": 4.3472,
+      "step": 7275
+    },
+    {
+      "epoch": 2.4832764505119456,
+      "grad_norm": 2.7347793579101562,
+      "learning_rate": 0.00017224118316268487,
+      "loss": 5.2237,
+      "step": 7276
+    },
+    {
+      "epoch": 2.483617747440273,
+      "grad_norm": 2.723522424697876,
+      "learning_rate": 0.00017212741751990897,
+      "loss": 6.5071,
+      "step": 7277
+    },
+    {
+      "epoch": 2.483959044368601,
+      "grad_norm": 2.6944844722747803,
+      "learning_rate": 0.0001720136518771331,
+      "loss": 6.1664,
+      "step": 7278
+    },
+    {
+      "epoch": 2.4843003412969282,
+      "grad_norm": 2.7033700942993164,
+      "learning_rate": 0.00017189988623435724,
+      "loss": 6.2606,
+      "step": 7279
+    },
+    {
+      "epoch": 2.484641638225256,
+      "grad_norm": 3.5506386756896973,
+      "learning_rate": 0.00017178612059158137,
+      "loss": 4.5449,
+      "step": 7280
+    },
+    {
+      "epoch": 2.4849829351535835,
+      "grad_norm": 2.705688714981079,
+      "learning_rate": 0.00017167235494880545,
+      "loss": 6.3225,
+      "step": 7281
+    },
+    {
+      "epoch": 2.4853242320819113,
+      "grad_norm": 2.7204155921936035,
+      "learning_rate": 0.00017155858930602958,
+      "loss": 6.047,
+      "step": 7282
+    },
+    {
+      "epoch": 2.4856655290102387,
+      "grad_norm": 2.656323194503784,
+      "learning_rate": 0.0001714448236632537,
+      "loss": 6.1217,
+      "step": 7283
+    },
+    {
+      "epoch": 2.4860068259385666,
+      "grad_norm": 3.497453451156616,
+      "learning_rate": 0.00017133105802047782,
+      "loss": 5.0882,
+      "step": 7284
+    },
+    {
+      "epoch": 2.486348122866894,
+      "grad_norm": 2.69559383392334,
+      "learning_rate": 0.00017121729237770195,
+      "loss": 5.7036,
+      "step": 7285
+    },
+    {
+      "epoch": 2.486689419795222,
+      "grad_norm": 2.734285354614258,
+      "learning_rate": 0.00017110352673492605,
+      "loss": 5.8757,
+      "step": 7286
+    },
+    {
+      "epoch": 2.4870307167235497,
+      "grad_norm": 2.6888885498046875,
+      "learning_rate": 0.00017098976109215016,
+      "loss": 5.7562,
+      "step": 7287
+    },
+    {
+      "epoch": 2.487372013651877,
+      "grad_norm": 2.7632334232330322,
+      "learning_rate": 0.0001708759954493743,
+      "loss": 6.3917,
+      "step": 7288
+    },
+    {
+      "epoch": 2.487713310580205,
+      "grad_norm": 2.7298285961151123,
+      "learning_rate": 0.00017076222980659842,
+      "loss": 5.9822,
+      "step": 7289
+    },
+    {
+      "epoch": 2.4880546075085324,
+      "grad_norm": 2.7448837757110596,
+      "learning_rate": 0.00017064846416382255,
+      "loss": 6.1619,
+      "step": 7290
+    },
+    {
+      "epoch": 2.48839590443686,
+      "grad_norm": 2.6427669525146484,
+      "learning_rate": 0.00017053469852104663,
+      "loss": 5.762,
+      "step": 7291
+    },
+    {
+      "epoch": 2.4887372013651876,
+      "grad_norm": 2.733144998550415,
+      "learning_rate": 0.00017042093287827076,
+      "loss": 6.4147,
+      "step": 7292
+    },
+    {
+      "epoch": 2.4890784982935155,
+      "grad_norm": 2.6906228065490723,
+      "learning_rate": 0.0001703071672354949,
+      "loss": 6.2987,
+      "step": 7293
+    },
+    {
+      "epoch": 2.489419795221843,
+      "grad_norm": 2.6950948238372803,
+      "learning_rate": 0.000170193401592719,
+      "loss": 5.6162,
+      "step": 7294
+    },
+    {
+      "epoch": 2.4897610921501707,
+      "grad_norm": 2.7234177589416504,
+      "learning_rate": 0.00017007963594994313,
+      "loss": 5.8311,
+      "step": 7295
+    },
+    {
+      "epoch": 2.490102389078498,
+      "grad_norm": 2.7544310092926025,
+      "learning_rate": 0.00016996587030716724,
+      "loss": 6.3541,
+      "step": 7296
+    },
+    {
+      "epoch": 2.490443686006826,
+      "grad_norm": 2.595151424407959,
+      "learning_rate": 0.00016985210466439134,
+      "loss": 6.0401,
+      "step": 7297
+    },
+    {
+      "epoch": 2.4907849829351534,
+      "grad_norm": 3.826329469680786,
+      "learning_rate": 0.00016973833902161548,
+      "loss": 4.7459,
+      "step": 7298
+    },
+    {
+      "epoch": 2.4911262798634812,
+      "grad_norm": 2.8267500400543213,
+      "learning_rate": 0.0001696245733788396,
+      "loss": 5.6925,
+      "step": 7299
+    },
+    {
+      "epoch": 2.491467576791809,
+      "grad_norm": 2.7589223384857178,
+      "learning_rate": 0.0001695108077360637,
+      "loss": 5.8866,
+      "step": 7300
+    },
+    {
+      "epoch": 2.4918088737201365,
+      "grad_norm": 2.703211545944214,
+      "learning_rate": 0.00016939704209328782,
+      "loss": 6.3634,
+      "step": 7301
+    },
+    {
+      "epoch": 2.4921501706484643,
+      "grad_norm": 2.917485237121582,
+      "learning_rate": 0.00016928327645051195,
+      "loss": 5.7224,
+      "step": 7302
+    },
+    {
+      "epoch": 2.4924914675767917,
+      "grad_norm": 2.6582095623016357,
+      "learning_rate": 0.00016916951080773608,
+      "loss": 5.8915,
+      "step": 7303
+    },
+    {
+      "epoch": 2.4928327645051196,
+      "grad_norm": 2.8652775287628174,
+      "learning_rate": 0.00016905574516496019,
+      "loss": 3.9218,
+      "step": 7304
+    },
+    {
+      "epoch": 2.493174061433447,
+      "grad_norm": 2.739563226699829,
+      "learning_rate": 0.0001689419795221843,
+      "loss": 6.6156,
+      "step": 7305
+    },
+    {
+      "epoch": 2.493515358361775,
+      "grad_norm": 2.928471803665161,
+      "learning_rate": 0.00016882821387940842,
+      "loss": 5.1992,
+      "step": 7306
+    },
+    {
+      "epoch": 2.4938566552901023,
+      "grad_norm": 2.6852288246154785,
+      "learning_rate": 0.00016871444823663253,
+      "loss": 5.9716,
+      "step": 7307
+    },
+    {
+      "epoch": 2.49419795221843,
+      "grad_norm": 2.7359983921051025,
+      "learning_rate": 0.00016860068259385666,
+      "loss": 5.5148,
+      "step": 7308
+    },
+    {
+      "epoch": 2.4945392491467575,
+      "grad_norm": 2.796595335006714,
+      "learning_rate": 0.0001684869169510808,
+      "loss": 6.9963,
+      "step": 7309
+    },
+    {
+      "epoch": 2.4948805460750854,
+      "grad_norm": 2.682931900024414,
+      "learning_rate": 0.0001683731513083049,
+      "loss": 5.4007,
+      "step": 7310
+    },
+    {
+      "epoch": 2.4952218430034128,
+      "grad_norm": 1.8263013362884521,
+      "learning_rate": 0.000168259385665529,
+      "loss": 3.1275,
+      "step": 7311
+    },
+    {
+      "epoch": 2.4955631399317406,
+      "grad_norm": 2.828277587890625,
+      "learning_rate": 0.00016814562002275313,
+      "loss": 6.0928,
+      "step": 7312
+    },
+    {
+      "epoch": 2.4959044368600685,
+      "grad_norm": 2.7260549068450928,
+      "learning_rate": 0.00016803185437997727,
+      "loss": 5.7029,
+      "step": 7313
+    },
+    {
+      "epoch": 2.496245733788396,
+      "grad_norm": 2.6961669921875,
+      "learning_rate": 0.00016791808873720137,
+      "loss": 5.9228,
+      "step": 7314
+    },
+    {
+      "epoch": 2.4965870307167237,
+      "grad_norm": 2.813055992126465,
+      "learning_rate": 0.00016780432309442548,
+      "loss": 6.1379,
+      "step": 7315
+    },
+    {
+      "epoch": 2.496928327645051,
+      "grad_norm": 2.6785166263580322,
+      "learning_rate": 0.0001676905574516496,
+      "loss": 5.7098,
+      "step": 7316
+    },
+    {
+      "epoch": 2.497269624573379,
+      "grad_norm": 2.698068857192993,
+      "learning_rate": 0.0001675767918088737,
+      "loss": 6.0382,
+      "step": 7317
+    },
+    {
+      "epoch": 2.4976109215017064,
+      "grad_norm": 2.680922508239746,
+      "learning_rate": 0.00016746302616609785,
+      "loss": 6.1616,
+      "step": 7318
+    },
+    {
+      "epoch": 2.4979522184300342,
+      "grad_norm": 2.7005722522735596,
+      "learning_rate": 0.00016734926052332198,
+      "loss": 5.649,
+      "step": 7319
+    },
+    {
+      "epoch": 2.4982935153583616,
+      "grad_norm": 2.744039535522461,
+      "learning_rate": 0.00016723549488054606,
+      "loss": 6.2951,
+      "step": 7320
+    },
+    {
+      "epoch": 2.4986348122866895,
+      "grad_norm": 2.7696053981781006,
+      "learning_rate": 0.0001671217292377702,
+      "loss": 6.2824,
+      "step": 7321
+    },
+    {
+      "epoch": 2.498976109215017,
+      "grad_norm": 2.9358487129211426,
+      "learning_rate": 0.00016700796359499432,
+      "loss": 5.1869,
+      "step": 7322
+    },
+    {
+      "epoch": 2.4993174061433447,
+      "grad_norm": 2.740104913711548,
+      "learning_rate": 0.00016689419795221845,
+      "loss": 6.0786,
+      "step": 7323
+    },
+    {
+      "epoch": 2.499658703071672,
+      "grad_norm": 2.702230215072632,
+      "learning_rate": 0.00016678043230944256,
+      "loss": 6.4051,
+      "step": 7324
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.933879375457764,
+      "learning_rate": 0.00016666666666666666,
+      "loss": 4.6992,
+      "step": 7325
+    },
+    {
+      "epoch": 2.500341296928328,
+      "grad_norm": 2.6900596618652344,
+      "learning_rate": 0.0001665529010238908,
+      "loss": 5.8764,
+      "step": 7326
+    },
+    {
+      "epoch": 2.5006825938566553,
+      "grad_norm": 2.7726457118988037,
+      "learning_rate": 0.0001664391353811149,
+      "loss": 5.6187,
+      "step": 7327
+    },
+    {
+      "epoch": 2.5010238907849827,
+      "grad_norm": 1.9055275917053223,
+      "learning_rate": 0.00016632536973833903,
+      "loss": 3.4067,
+      "step": 7328
+    },
+    {
+      "epoch": 2.5013651877133105,
+      "grad_norm": 2.7264151573181152,
+      "learning_rate": 0.00016621160409556316,
+      "loss": 6.0756,
+      "step": 7329
+    },
+    {
+      "epoch": 2.5017064846416384,
+      "grad_norm": 2.759608030319214,
+      "learning_rate": 0.00016609783845278724,
+      "loss": 6.1258,
+      "step": 7330
+    },
+    {
+      "epoch": 2.5020477815699658,
+      "grad_norm": 2.6936237812042236,
+      "learning_rate": 0.00016598407281001137,
+      "loss": 6.2095,
+      "step": 7331
+    },
+    {
+      "epoch": 2.5023890784982936,
+      "grad_norm": 2.6739792823791504,
+      "learning_rate": 0.0001658703071672355,
+      "loss": 6.5549,
+      "step": 7332
+    },
+    {
+      "epoch": 2.502730375426621,
+      "grad_norm": 2.725817918777466,
+      "learning_rate": 0.00016575654152445964,
+      "loss": 6.1217,
+      "step": 7333
+    },
+    {
+      "epoch": 2.503071672354949,
+      "grad_norm": 2.785630702972412,
+      "learning_rate": 0.00016564277588168371,
+      "loss": 6.0126,
+      "step": 7334
+    },
+    {
+      "epoch": 2.5034129692832767,
+      "grad_norm": 2.7225522994995117,
+      "learning_rate": 0.00016552901023890785,
+      "loss": 5.9088,
+      "step": 7335
+    },
+    {
+      "epoch": 2.503754266211604,
+      "grad_norm": 2.7138113975524902,
+      "learning_rate": 0.00016541524459613198,
+      "loss": 6.5287,
+      "step": 7336
+    },
+    {
+      "epoch": 2.5040955631399315,
+      "grad_norm": 2.659165143966675,
+      "learning_rate": 0.00016530147895335608,
+      "loss": 6.1559,
+      "step": 7337
+    },
+    {
+      "epoch": 2.5044368600682594,
+      "grad_norm": 2.6596744060516357,
+      "learning_rate": 0.00016518771331058022,
+      "loss": 5.9957,
+      "step": 7338
+    },
+    {
+      "epoch": 2.5047781569965872,
+      "grad_norm": 2.6613821983337402,
+      "learning_rate": 0.00016507394766780432,
+      "loss": 5.6435,
+      "step": 7339
+    },
+    {
+      "epoch": 2.5051194539249146,
+      "grad_norm": 2.7147886753082275,
+      "learning_rate": 0.00016496018202502842,
+      "loss": 6.0681,
+      "step": 7340
+    },
+    {
+      "epoch": 2.505460750853242,
+      "grad_norm": 2.6107845306396484,
+      "learning_rate": 0.00016484641638225256,
+      "loss": 5.9106,
+      "step": 7341
+    },
+    {
+      "epoch": 2.50580204778157,
+      "grad_norm": 2.7579963207244873,
+      "learning_rate": 0.0001647326507394767,
+      "loss": 6.6005,
+      "step": 7342
+    },
+    {
+      "epoch": 2.5061433447098977,
+      "grad_norm": 2.604501962661743,
+      "learning_rate": 0.00016461888509670082,
+      "loss": 5.8982,
+      "step": 7343
+    },
+    {
+      "epoch": 2.506484641638225,
+      "grad_norm": 2.7372636795043945,
+      "learning_rate": 0.0001645051194539249,
+      "loss": 5.3334,
+      "step": 7344
+    },
+    {
+      "epoch": 2.506825938566553,
+      "grad_norm": 2.69775390625,
+      "learning_rate": 0.00016439135381114903,
+      "loss": 6.0477,
+      "step": 7345
+    },
+    {
+      "epoch": 2.5071672354948804,
+      "grad_norm": 2.698512077331543,
+      "learning_rate": 0.00016427758816837316,
+      "loss": 5.5654,
+      "step": 7346
+    },
+    {
+      "epoch": 2.5075085324232083,
+      "grad_norm": 2.6294450759887695,
+      "learning_rate": 0.00016416382252559727,
+      "loss": 6.1314,
+      "step": 7347
+    },
+    {
+      "epoch": 2.507849829351536,
+      "grad_norm": 2.6949124336242676,
+      "learning_rate": 0.0001640500568828214,
+      "loss": 6.6314,
+      "step": 7348
+    },
+    {
+      "epoch": 2.5081911262798635,
+      "grad_norm": 2.7231664657592773,
+      "learning_rate": 0.0001639362912400455,
+      "loss": 6.1183,
+      "step": 7349
+    },
+    {
+      "epoch": 2.508532423208191,
+      "grad_norm": 2.9584968090057373,
+      "learning_rate": 0.0001638225255972696,
+      "loss": 5.1231,
+      "step": 7350
+    },
+    {
+      "epoch": 2.5088737201365188,
+      "grad_norm": 2.7133231163024902,
+      "learning_rate": 0.00016370875995449374,
+      "loss": 5.5295,
+      "step": 7351
+    },
+    {
+      "epoch": 2.5092150170648466,
+      "grad_norm": 2.7822914123535156,
+      "learning_rate": 0.00016359499431171787,
+      "loss": 5.4579,
+      "step": 7352
+    },
+    {
+      "epoch": 2.509556313993174,
+      "grad_norm": 2.6988959312438965,
+      "learning_rate": 0.000163481228668942,
+      "loss": 6.0726,
+      "step": 7353
+    },
+    {
+      "epoch": 2.5098976109215014,
+      "grad_norm": 2.623605489730835,
+      "learning_rate": 0.00016336746302616608,
+      "loss": 6.1312,
+      "step": 7354
+    },
+    {
+      "epoch": 2.5102389078498293,
+      "grad_norm": 2.7255163192749023,
+      "learning_rate": 0.00016325369738339022,
+      "loss": 5.9188,
+      "step": 7355
+    },
+    {
+      "epoch": 2.510580204778157,
+      "grad_norm": 2.6522679328918457,
+      "learning_rate": 0.00016313993174061435,
+      "loss": 6.2107,
+      "step": 7356
+    },
+    {
+      "epoch": 2.5109215017064845,
+      "grad_norm": 2.733840227127075,
+      "learning_rate": 0.00016302616609783845,
+      "loss": 6.0067,
+      "step": 7357
+    },
+    {
+      "epoch": 2.5112627986348124,
+      "grad_norm": 2.6680688858032227,
+      "learning_rate": 0.00016291240045506259,
+      "loss": 6.2216,
+      "step": 7358
+    },
+    {
+      "epoch": 2.51160409556314,
+      "grad_norm": 3.2532639503479004,
+      "learning_rate": 0.0001627986348122867,
+      "loss": 5.0936,
+      "step": 7359
+    },
+    {
+      "epoch": 2.5119453924914676,
+      "grad_norm": 2.6566052436828613,
+      "learning_rate": 0.0001626848691695108,
+      "loss": 6.2296,
+      "step": 7360
+    },
+    {
+      "epoch": 2.5122866894197955,
+      "grad_norm": 3.307908296585083,
+      "learning_rate": 0.00016257110352673493,
+      "loss": 4.3073,
+      "step": 7361
+    },
+    {
+      "epoch": 2.512627986348123,
+      "grad_norm": 2.7235147953033447,
+      "learning_rate": 0.00016245733788395906,
+      "loss": 5.967,
+      "step": 7362
+    },
+    {
+      "epoch": 2.5129692832764503,
+      "grad_norm": 2.7577085494995117,
+      "learning_rate": 0.0001623435722411832,
+      "loss": 4.6936,
+      "step": 7363
+    },
+    {
+      "epoch": 2.513310580204778,
+      "grad_norm": 2.6942477226257324,
+      "learning_rate": 0.00016222980659840727,
+      "loss": 6.1181,
+      "step": 7364
+    },
+    {
+      "epoch": 2.513651877133106,
+      "grad_norm": 2.6965997219085693,
+      "learning_rate": 0.0001621160409556314,
+      "loss": 4.4271,
+      "step": 7365
+    },
+    {
+      "epoch": 2.5139931740614334,
+      "grad_norm": 2.5623295307159424,
+      "learning_rate": 0.00016200227531285553,
+      "loss": 5.8977,
+      "step": 7366
+    },
+    {
+      "epoch": 2.514334470989761,
+      "grad_norm": 2.7341978549957275,
+      "learning_rate": 0.00016188850967007964,
+      "loss": 5.7355,
+      "step": 7367
+    },
+    {
+      "epoch": 2.5146757679180887,
+      "grad_norm": 2.686908721923828,
+      "learning_rate": 0.00016177474402730374,
+      "loss": 5.5919,
+      "step": 7368
+    },
+    {
+      "epoch": 2.5150170648464165,
+      "grad_norm": 2.768826961517334,
+      "learning_rate": 0.00016166097838452787,
+      "loss": 6.317,
+      "step": 7369
+    },
+    {
+      "epoch": 2.515358361774744,
+      "grad_norm": 2.7448062896728516,
+      "learning_rate": 0.00016154721274175198,
+      "loss": 5.9812,
+      "step": 7370
+    },
+    {
+      "epoch": 2.5156996587030718,
+      "grad_norm": 2.6533572673797607,
+      "learning_rate": 0.0001614334470989761,
+      "loss": 5.7661,
+      "step": 7371
+    },
+    {
+      "epoch": 2.516040955631399,
+      "grad_norm": 2.7210354804992676,
+      "learning_rate": 0.00016131968145620024,
+      "loss": 5.9004,
+      "step": 7372
+    },
+    {
+      "epoch": 2.516382252559727,
+      "grad_norm": 2.7211415767669678,
+      "learning_rate": 0.00016120591581342435,
+      "loss": 6.3239,
+      "step": 7373
+    },
+    {
+      "epoch": 2.516723549488055,
+      "grad_norm": 2.718151092529297,
+      "learning_rate": 0.00016109215017064845,
+      "loss": 6.2132,
+      "step": 7374
+    },
+    {
+      "epoch": 2.5170648464163823,
+      "grad_norm": 2.657224416732788,
+      "learning_rate": 0.00016097838452787259,
+      "loss": 5.9553,
+      "step": 7375
+    },
+    {
+      "epoch": 2.5174061433447097,
+      "grad_norm": 2.738079071044922,
+      "learning_rate": 0.00016086461888509672,
+      "loss": 5.824,
+      "step": 7376
+    },
+    {
+      "epoch": 2.5177474402730375,
+      "grad_norm": 2.7597906589508057,
+      "learning_rate": 0.00016075085324232082,
+      "loss": 6.2877,
+      "step": 7377
+    },
+    {
+      "epoch": 2.5180887372013654,
+      "grad_norm": 2.7214534282684326,
+      "learning_rate": 0.00016063708759954493,
+      "loss": 6.262,
+      "step": 7378
+    },
+    {
+      "epoch": 2.518430034129693,
+      "grad_norm": 2.705698013305664,
+      "learning_rate": 0.00016052332195676906,
+      "loss": 6.2512,
+      "step": 7379
+    },
+    {
+      "epoch": 2.51877133105802,
+      "grad_norm": 2.636011838912964,
+      "learning_rate": 0.00016040955631399316,
+      "loss": 4.9735,
+      "step": 7380
+    },
+    {
+      "epoch": 2.519112627986348,
+      "grad_norm": 3.66536021232605,
+      "learning_rate": 0.0001602957906712173,
+      "loss": 3.9323,
+      "step": 7381
+    },
+    {
+      "epoch": 2.519453924914676,
+      "grad_norm": 2.7328991889953613,
+      "learning_rate": 0.00016018202502844143,
+      "loss": 4.5109,
+      "step": 7382
+    },
+    {
+      "epoch": 2.5197952218430033,
+      "grad_norm": 2.689422369003296,
+      "learning_rate": 0.00016006825938566553,
+      "loss": 5.9128,
+      "step": 7383
+    },
+    {
+      "epoch": 2.520136518771331,
+      "grad_norm": 3.5838780403137207,
+      "learning_rate": 0.00015995449374288964,
+      "loss": 4.2017,
+      "step": 7384
+    },
+    {
+      "epoch": 2.5204778156996586,
+      "grad_norm": 2.720184087753296,
+      "learning_rate": 0.00015984072810011377,
+      "loss": 5.6297,
+      "step": 7385
+    },
+    {
+      "epoch": 2.5208191126279864,
+      "grad_norm": 2.6871819496154785,
+      "learning_rate": 0.0001597269624573379,
+      "loss": 5.9124,
+      "step": 7386
+    },
+    {
+      "epoch": 2.5211604095563143,
+      "grad_norm": 2.7035536766052246,
+      "learning_rate": 0.000159613196814562,
+      "loss": 6.2627,
+      "step": 7387
+    },
+    {
+      "epoch": 2.5215017064846417,
+      "grad_norm": 2.7028918266296387,
+      "learning_rate": 0.0001594994311717861,
+      "loss": 5.8171,
+      "step": 7388
+    },
+    {
+      "epoch": 2.521843003412969,
+      "grad_norm": 2.6863887310028076,
+      "learning_rate": 0.00015938566552901024,
+      "loss": 6.7491,
+      "step": 7389
+    },
+    {
+      "epoch": 2.522184300341297,
+      "grad_norm": 2.64327073097229,
+      "learning_rate": 0.00015927189988623435,
+      "loss": 6.236,
+      "step": 7390
+    },
+    {
+      "epoch": 2.5225255972696248,
+      "grad_norm": 2.628070592880249,
+      "learning_rate": 0.00015915813424345848,
+      "loss": 5.51,
+      "step": 7391
+    },
+    {
+      "epoch": 2.522866894197952,
+      "grad_norm": 2.6996893882751465,
+      "learning_rate": 0.00015904436860068261,
+      "loss": 6.1538,
+      "step": 7392
+    },
+    {
+      "epoch": 2.5232081911262796,
+      "grad_norm": 4.886688709259033,
+      "learning_rate": 0.00015893060295790672,
+      "loss": 4.6864,
+      "step": 7393
+    },
+    {
+      "epoch": 2.5235494880546074,
+      "grad_norm": 2.6805944442749023,
+      "learning_rate": 0.00015881683731513082,
+      "loss": 6.105,
+      "step": 7394
+    },
+    {
+      "epoch": 2.5238907849829353,
+      "grad_norm": 2.6910324096679688,
+      "learning_rate": 0.00015870307167235496,
+      "loss": 6.266,
+      "step": 7395
+    },
+    {
+      "epoch": 2.5242320819112627,
+      "grad_norm": 2.6817996501922607,
+      "learning_rate": 0.0001585893060295791,
+      "loss": 6.4578,
+      "step": 7396
+    },
+    {
+      "epoch": 2.5245733788395905,
+      "grad_norm": 3.1869544982910156,
+      "learning_rate": 0.0001584755403868032,
+      "loss": 5.4998,
+      "step": 7397
+    },
+    {
+      "epoch": 2.524914675767918,
+      "grad_norm": 2.6462249755859375,
+      "learning_rate": 0.0001583617747440273,
+      "loss": 6.3286,
+      "step": 7398
+    },
+    {
+      "epoch": 2.525255972696246,
+      "grad_norm": 2.711543321609497,
+      "learning_rate": 0.00015824800910125143,
+      "loss": 6.6815,
+      "step": 7399
+    },
+    {
+      "epoch": 2.5255972696245736,
+      "grad_norm": 2.6534812450408936,
+      "learning_rate": 0.00015813424345847553,
+      "loss": 6.3155,
+      "step": 7400
+    },
+    {
+      "epoch": 2.525938566552901,
+      "grad_norm": 2.68021559715271,
+      "learning_rate": 0.00015802047781569967,
+      "loss": 6.126,
+      "step": 7401
+    },
+    {
+      "epoch": 2.5262798634812285,
+      "grad_norm": 2.7050492763519287,
+      "learning_rate": 0.00015790671217292377,
+      "loss": 5.7176,
+      "step": 7402
+    },
+    {
+      "epoch": 2.5266211604095563,
+      "grad_norm": 2.6450881958007812,
+      "learning_rate": 0.0001577929465301479,
+      "loss": 6.5606,
+      "step": 7403
+    },
+    {
+      "epoch": 2.526962457337884,
+      "grad_norm": 2.6857900619506836,
+      "learning_rate": 0.000157679180887372,
+      "loss": 6.2598,
+      "step": 7404
+    },
+    {
+      "epoch": 2.5273037542662116,
+      "grad_norm": 2.6581039428710938,
+      "learning_rate": 0.00015756541524459614,
+      "loss": 6.5876,
+      "step": 7405
+    },
+    {
+      "epoch": 2.527645051194539,
+      "grad_norm": 2.6261470317840576,
+      "learning_rate": 0.00015745164960182027,
+      "loss": 5.4513,
+      "step": 7406
+    },
+    {
+      "epoch": 2.527986348122867,
+      "grad_norm": 2.684361696243286,
+      "learning_rate": 0.00015733788395904435,
+      "loss": 5.8607,
+      "step": 7407
+    },
+    {
+      "epoch": 2.5283276450511947,
+      "grad_norm": 2.6788904666900635,
+      "learning_rate": 0.00015722411831626848,
+      "loss": 6.32,
+      "step": 7408
+    },
+    {
+      "epoch": 2.528668941979522,
+      "grad_norm": 2.700685977935791,
+      "learning_rate": 0.00015711035267349261,
+      "loss": 5.7209,
+      "step": 7409
+    },
+    {
+      "epoch": 2.52901023890785,
+      "grad_norm": 2.7809455394744873,
+      "learning_rate": 0.00015699658703071672,
+      "loss": 6.0776,
+      "step": 7410
+    },
+    {
+      "epoch": 2.5293515358361773,
+      "grad_norm": 2.6642520427703857,
+      "learning_rate": 0.00015688282138794085,
+      "loss": 5.6266,
+      "step": 7411
+    },
+    {
+      "epoch": 2.529692832764505,
+      "grad_norm": 2.7473788261413574,
+      "learning_rate": 0.00015676905574516496,
+      "loss": 5.0825,
+      "step": 7412
+    },
+    {
+      "epoch": 2.530034129692833,
+      "grad_norm": 2.7476108074188232,
+      "learning_rate": 0.0001566552901023891,
+      "loss": 6.2024,
+      "step": 7413
+    },
+    {
+      "epoch": 2.5303754266211604,
+      "grad_norm": 2.6110386848449707,
+      "learning_rate": 0.0001565415244596132,
+      "loss": 6.1209,
+      "step": 7414
+    },
+    {
+      "epoch": 2.530716723549488,
+      "grad_norm": 2.9755313396453857,
+      "learning_rate": 0.00015642775881683733,
+      "loss": 5.5796,
+      "step": 7415
+    },
+    {
+      "epoch": 2.5310580204778157,
+      "grad_norm": 2.710559844970703,
+      "learning_rate": 0.00015631399317406146,
+      "loss": 5.7508,
+      "step": 7416
+    },
+    {
+      "epoch": 2.5313993174061435,
+      "grad_norm": 2.636697292327881,
+      "learning_rate": 0.00015620022753128554,
+      "loss": 6.1145,
+      "step": 7417
+    },
+    {
+      "epoch": 2.531740614334471,
+      "grad_norm": 2.8945558071136475,
+      "learning_rate": 0.00015608646188850967,
+      "loss": 5.4822,
+      "step": 7418
+    },
+    {
+      "epoch": 2.532081911262799,
+      "grad_norm": 2.7583422660827637,
+      "learning_rate": 0.0001559726962457338,
+      "loss": 6.3568,
+      "step": 7419
+    },
+    {
+      "epoch": 2.532423208191126,
+      "grad_norm": 2.6713101863861084,
+      "learning_rate": 0.0001558589306029579,
+      "loss": 5.554,
+      "step": 7420
+    },
+    {
+      "epoch": 2.532764505119454,
+      "grad_norm": 2.626268148422241,
+      "learning_rate": 0.00015574516496018204,
+      "loss": 5.8425,
+      "step": 7421
+    },
+    {
+      "epoch": 2.5331058020477815,
+      "grad_norm": 2.9670071601867676,
+      "learning_rate": 0.00015563139931740614,
+      "loss": 5.2788,
+      "step": 7422
+    },
+    {
+      "epoch": 2.5334470989761093,
+      "grad_norm": 2.694166660308838,
+      "learning_rate": 0.00015551763367463025,
+      "loss": 5.982,
+      "step": 7423
+    },
+    {
+      "epoch": 2.5337883959044367,
+      "grad_norm": 2.6969735622406006,
+      "learning_rate": 0.00015540386803185438,
+      "loss": 6.4294,
+      "step": 7424
+    },
+    {
+      "epoch": 2.5341296928327646,
+      "grad_norm": 2.7275049686431885,
+      "learning_rate": 0.0001552901023890785,
+      "loss": 6.0112,
+      "step": 7425
+    },
+    {
+      "epoch": 2.5344709897610924,
+      "grad_norm": 2.7712514400482178,
+      "learning_rate": 0.00015517633674630264,
+      "loss": 6.2063,
+      "step": 7426
+    },
+    {
+      "epoch": 2.53481228668942,
+      "grad_norm": 3.0031471252441406,
+      "learning_rate": 0.00015506257110352672,
+      "loss": 4.5911,
+      "step": 7427
+    },
+    {
+      "epoch": 2.5351535836177472,
+      "grad_norm": 2.857637882232666,
+      "learning_rate": 0.00015494880546075085,
+      "loss": 4.5048,
+      "step": 7428
+    },
+    {
+      "epoch": 2.535494880546075,
+      "grad_norm": 2.757772207260132,
+      "learning_rate": 0.00015483503981797498,
+      "loss": 5.9304,
+      "step": 7429
+    },
+    {
+      "epoch": 2.535836177474403,
+      "grad_norm": 2.77624773979187,
+      "learning_rate": 0.0001547212741751991,
+      "loss": 5.2982,
+      "step": 7430
+    },
+    {
+      "epoch": 2.5361774744027303,
+      "grad_norm": 2.9393277168273926,
+      "learning_rate": 0.00015460750853242322,
+      "loss": 5.1273,
+      "step": 7431
+    },
+    {
+      "epoch": 2.536518771331058,
+      "grad_norm": 2.715731143951416,
+      "learning_rate": 0.00015449374288964733,
+      "loss": 6.2645,
+      "step": 7432
+    },
+    {
+      "epoch": 2.5368600682593856,
+      "grad_norm": 2.7153773307800293,
+      "learning_rate": 0.00015437997724687143,
+      "loss": 5.8669,
+      "step": 7433
+    },
+    {
+      "epoch": 2.5372013651877134,
+      "grad_norm": 2.890367031097412,
+      "learning_rate": 0.00015426621160409556,
+      "loss": 5.8867,
+      "step": 7434
+    },
+    {
+      "epoch": 2.537542662116041,
+      "grad_norm": 2.7217259407043457,
+      "learning_rate": 0.0001541524459613197,
+      "loss": 6.2305,
+      "step": 7435
+    },
+    {
+      "epoch": 2.5378839590443687,
+      "grad_norm": 5.296313285827637,
+      "learning_rate": 0.00015403868031854383,
+      "loss": 3.7479,
+      "step": 7436
+    },
+    {
+      "epoch": 2.538225255972696,
+      "grad_norm": 2.703120231628418,
+      "learning_rate": 0.0001539249146757679,
+      "loss": 6.5971,
+      "step": 7437
+    },
+    {
+      "epoch": 2.538566552901024,
+      "grad_norm": 3.4888710975646973,
+      "learning_rate": 0.00015381114903299204,
+      "loss": 4.5474,
+      "step": 7438
+    },
+    {
+      "epoch": 2.538907849829352,
+      "grad_norm": 2.7433698177337646,
+      "learning_rate": 0.00015369738339021617,
+      "loss": 5.4895,
+      "step": 7439
+    },
+    {
+      "epoch": 2.539249146757679,
+      "grad_norm": 2.715934991836548,
+      "learning_rate": 0.00015358361774744027,
+      "loss": 4.5857,
+      "step": 7440
+    },
+    {
+      "epoch": 2.5395904436860066,
+      "grad_norm": 2.6325573921203613,
+      "learning_rate": 0.00015346985210466438,
+      "loss": 6.0054,
+      "step": 7441
+    },
+    {
+      "epoch": 2.5399317406143345,
+      "grad_norm": 2.666834831237793,
+      "learning_rate": 0.0001533560864618885,
+      "loss": 6.4198,
+      "step": 7442
+    },
+    {
+      "epoch": 2.5402730375426623,
+      "grad_norm": 2.6133065223693848,
+      "learning_rate": 0.00015324232081911262,
+      "loss": 6.1836,
+      "step": 7443
+    },
+    {
+      "epoch": 2.5406143344709897,
+      "grad_norm": 2.7008965015411377,
+      "learning_rate": 0.00015312855517633675,
+      "loss": 6.5665,
+      "step": 7444
+    },
+    {
+      "epoch": 2.5409556313993176,
+      "grad_norm": 2.5905418395996094,
+      "learning_rate": 0.00015301478953356088,
+      "loss": 5.7239,
+      "step": 7445
+    },
+    {
+      "epoch": 2.541296928327645,
+      "grad_norm": 2.6478431224823,
+      "learning_rate": 0.00015290102389078499,
+      "loss": 6.0945,
+      "step": 7446
+    },
+    {
+      "epoch": 2.541638225255973,
+      "grad_norm": 2.7686331272125244,
+      "learning_rate": 0.0001527872582480091,
+      "loss": 5.9458,
+      "step": 7447
+    },
+    {
+      "epoch": 2.5419795221843002,
+      "grad_norm": 2.7820048332214355,
+      "learning_rate": 0.00015267349260523322,
+      "loss": 5.3187,
+      "step": 7448
+    },
+    {
+      "epoch": 2.542320819112628,
+      "grad_norm": 2.7252564430236816,
+      "learning_rate": 0.00015255972696245735,
+      "loss": 6.1386,
+      "step": 7449
+    },
+    {
+      "epoch": 2.5426621160409555,
+      "grad_norm": 3.9234542846679688,
+      "learning_rate": 0.00015244596131968146,
+      "loss": 4.9931,
+      "step": 7450
+    },
+    {
+      "epoch": 2.5430034129692833,
+      "grad_norm": 2.7013142108917236,
+      "learning_rate": 0.00015233219567690556,
+      "loss": 6.5903,
+      "step": 7451
+    },
+    {
+      "epoch": 2.543344709897611,
+      "grad_norm": 2.7234249114990234,
+      "learning_rate": 0.0001522184300341297,
+      "loss": 6.5083,
+      "step": 7452
+    },
+    {
+      "epoch": 2.5436860068259386,
+      "grad_norm": 2.646691083908081,
+      "learning_rate": 0.0001521046643913538,
+      "loss": 5.8912,
+      "step": 7453
+    },
+    {
+      "epoch": 2.544027303754266,
+      "grad_norm": 2.6651551723480225,
+      "learning_rate": 0.00015199089874857793,
+      "loss": 6.4359,
+      "step": 7454
+    },
+    {
+      "epoch": 2.544368600682594,
+      "grad_norm": 2.7203242778778076,
+      "learning_rate": 0.00015187713310580207,
+      "loss": 6.1125,
+      "step": 7455
+    },
+    {
+      "epoch": 2.5447098976109217,
+      "grad_norm": 2.7022976875305176,
+      "learning_rate": 0.00015176336746302617,
+      "loss": 6.3177,
+      "step": 7456
+    },
+    {
+      "epoch": 2.545051194539249,
+      "grad_norm": 3.036482334136963,
+      "learning_rate": 0.00015164960182025028,
+      "loss": 5.8322,
+      "step": 7457
+    },
+    {
+      "epoch": 2.545392491467577,
+      "grad_norm": 2.6736273765563965,
+      "learning_rate": 0.0001515358361774744,
+      "loss": 5.683,
+      "step": 7458
+    },
+    {
+      "epoch": 2.5457337883959044,
+      "grad_norm": 2.6975038051605225,
+      "learning_rate": 0.00015142207053469854,
+      "loss": 5.9692,
+      "step": 7459
+    },
+    {
+      "epoch": 2.546075085324232,
+      "grad_norm": 2.6708292961120605,
+      "learning_rate": 0.00015130830489192264,
+      "loss": 6.2471,
+      "step": 7460
+    },
+    {
+      "epoch": 2.5464163822525596,
+      "grad_norm": 4.273525714874268,
+      "learning_rate": 0.00015119453924914675,
+      "loss": 5.0722,
+      "step": 7461
+    },
+    {
+      "epoch": 2.5467576791808875,
+      "grad_norm": 2.717935085296631,
+      "learning_rate": 0.00015108077360637088,
+      "loss": 6.2979,
+      "step": 7462
+    },
+    {
+      "epoch": 2.547098976109215,
+      "grad_norm": 2.620076894760132,
+      "learning_rate": 0.00015096700796359499,
+      "loss": 6.0223,
+      "step": 7463
+    },
+    {
+      "epoch": 2.5474402730375427,
+      "grad_norm": 2.6915009021759033,
+      "learning_rate": 0.00015085324232081912,
+      "loss": 6.0069,
+      "step": 7464
+    },
+    {
+      "epoch": 2.5477815699658706,
+      "grad_norm": 2.7233173847198486,
+      "learning_rate": 0.00015073947667804325,
+      "loss": 5.3867,
+      "step": 7465
+    },
+    {
+      "epoch": 2.548122866894198,
+      "grad_norm": 2.6472926139831543,
+      "learning_rate": 0.00015062571103526736,
+      "loss": 6.8144,
+      "step": 7466
+    },
+    {
+      "epoch": 2.5484641638225254,
+      "grad_norm": 2.7000956535339355,
+      "learning_rate": 0.00015051194539249146,
+      "loss": 5.9423,
+      "step": 7467
+    },
+    {
+      "epoch": 2.5488054607508532,
+      "grad_norm": 2.6225948333740234,
+      "learning_rate": 0.0001503981797497156,
+      "loss": 6.0564,
+      "step": 7468
+    },
+    {
+      "epoch": 2.549146757679181,
+      "grad_norm": 3.0721187591552734,
+      "learning_rate": 0.00015028441410693972,
+      "loss": 4.9398,
+      "step": 7469
+    },
+    {
+      "epoch": 2.5494880546075085,
+      "grad_norm": 2.621379852294922,
+      "learning_rate": 0.00015017064846416383,
+      "loss": 6.065,
+      "step": 7470
+    },
+    {
+      "epoch": 2.5498293515358363,
+      "grad_norm": 2.6615641117095947,
+      "learning_rate": 0.00015005688282138793,
+      "loss": 6.5425,
+      "step": 7471
+    },
+    {
+      "epoch": 2.5501706484641637,
+      "grad_norm": 2.7236409187316895,
+      "learning_rate": 0.00014994311717861207,
+      "loss": 6.1923,
+      "step": 7472
+    },
+    {
+      "epoch": 2.5505119453924916,
+      "grad_norm": 2.776329517364502,
+      "learning_rate": 0.00014982935153583617,
+      "loss": 6.3152,
+      "step": 7473
+    },
+    {
+      "epoch": 2.550853242320819,
+      "grad_norm": 2.7233684062957764,
+      "learning_rate": 0.0001497155858930603,
+      "loss": 5.9222,
+      "step": 7474
+    },
+    {
+      "epoch": 2.551194539249147,
+      "grad_norm": 2.578994035720825,
+      "learning_rate": 0.0001496018202502844,
+      "loss": 5.9317,
+      "step": 7475
+    },
+    {
+      "epoch": 2.5515358361774743,
+      "grad_norm": 2.6251556873321533,
+      "learning_rate": 0.00014948805460750854,
+      "loss": 6.1575,
+      "step": 7476
+    },
+    {
+      "epoch": 2.551877133105802,
+      "grad_norm": 2.599755048751831,
+      "learning_rate": 0.00014937428896473265,
+      "loss": 6.3228,
+      "step": 7477
+    },
+    {
+      "epoch": 2.55221843003413,
+      "grad_norm": 2.6746912002563477,
+      "learning_rate": 0.00014926052332195678,
+      "loss": 6.0529,
+      "step": 7478
+    },
+    {
+      "epoch": 2.5525597269624574,
+      "grad_norm": 2.6559457778930664,
+      "learning_rate": 0.0001491467576791809,
+      "loss": 5.4679,
+      "step": 7479
+    },
+    {
+      "epoch": 2.5529010238907848,
+      "grad_norm": 2.8027467727661133,
+      "learning_rate": 0.000149032992036405,
+      "loss": 5.8367,
+      "step": 7480
+    },
+    {
+      "epoch": 2.5532423208191126,
+      "grad_norm": 2.701709747314453,
+      "learning_rate": 0.00014891922639362912,
+      "loss": 6.8529,
+      "step": 7481
+    },
+    {
+      "epoch": 2.5535836177474405,
+      "grad_norm": 2.6528892517089844,
+      "learning_rate": 0.00014880546075085325,
+      "loss": 5.5392,
+      "step": 7482
+    },
+    {
+      "epoch": 2.553924914675768,
+      "grad_norm": 2.7285146713256836,
+      "learning_rate": 0.00014869169510807736,
+      "loss": 6.1253,
+      "step": 7483
+    },
+    {
+      "epoch": 2.5542662116040957,
+      "grad_norm": 2.646544933319092,
+      "learning_rate": 0.0001485779294653015,
+      "loss": 6.2362,
+      "step": 7484
+    },
+    {
+      "epoch": 2.554607508532423,
+      "grad_norm": 2.632233142852783,
+      "learning_rate": 0.0001484641638225256,
+      "loss": 5.9352,
+      "step": 7485
+    },
+    {
+      "epoch": 2.554948805460751,
+      "grad_norm": 2.641140937805176,
+      "learning_rate": 0.00014835039817974973,
+      "loss": 6.2673,
+      "step": 7486
+    },
+    {
+      "epoch": 2.5552901023890784,
+      "grad_norm": 2.7124555110931396,
+      "learning_rate": 0.00014823663253697383,
+      "loss": 5.4315,
+      "step": 7487
+    },
+    {
+      "epoch": 2.5556313993174062,
+      "grad_norm": 2.6860406398773193,
+      "learning_rate": 0.00014812286689419796,
+      "loss": 6.3044,
+      "step": 7488
+    },
+    {
+      "epoch": 2.5559726962457336,
+      "grad_norm": 2.6111605167388916,
+      "learning_rate": 0.0001480091012514221,
+      "loss": 6.2049,
+      "step": 7489
+    },
+    {
+      "epoch": 2.5563139931740615,
+      "grad_norm": 2.7284209728240967,
+      "learning_rate": 0.00014789533560864617,
+      "loss": 6.7773,
+      "step": 7490
+    },
+    {
+      "epoch": 2.5566552901023893,
+      "grad_norm": 2.732853651046753,
+      "learning_rate": 0.0001477815699658703,
+      "loss": 5.6464,
+      "step": 7491
+    },
+    {
+      "epoch": 2.5569965870307167,
+      "grad_norm": 2.888129949569702,
+      "learning_rate": 0.00014766780432309444,
+      "loss": 5.4964,
+      "step": 7492
+    },
+    {
+      "epoch": 2.557337883959044,
+      "grad_norm": 2.723909616470337,
+      "learning_rate": 0.00014755403868031854,
+      "loss": 6.3293,
+      "step": 7493
+    },
+    {
+      "epoch": 2.557679180887372,
+      "grad_norm": 2.6410248279571533,
+      "learning_rate": 0.00014744027303754267,
+      "loss": 5.7546,
+      "step": 7494
+    },
+    {
+      "epoch": 2.5580204778157,
+      "grad_norm": 2.708897829055786,
+      "learning_rate": 0.00014732650739476678,
+      "loss": 5.9368,
+      "step": 7495
+    },
+    {
+      "epoch": 2.5583617747440273,
+      "grad_norm": 2.6864473819732666,
+      "learning_rate": 0.0001472127417519909,
+      "loss": 6.3272,
+      "step": 7496
+    },
+    {
+      "epoch": 2.558703071672355,
+      "grad_norm": 2.6929290294647217,
+      "learning_rate": 0.00014709897610921502,
+      "loss": 6.0851,
+      "step": 7497
+    },
+    {
+      "epoch": 2.5590443686006825,
+      "grad_norm": 2.725696325302124,
+      "learning_rate": 0.00014698521046643915,
+      "loss": 6.092,
+      "step": 7498
+    },
+    {
+      "epoch": 2.5593856655290104,
+      "grad_norm": 2.703749418258667,
+      "learning_rate": 0.00014687144482366328,
+      "loss": 5.9719,
+      "step": 7499
+    },
+    {
+      "epoch": 2.5597269624573378,
+      "grad_norm": 3.734013319015503,
+      "learning_rate": 0.00014675767918088736,
+      "loss": 4.0252,
+      "step": 7500
+    },
+    {
+      "epoch": 2.5600682593856656,
+      "grad_norm": 2.6439578533172607,
+      "learning_rate": 0.0001466439135381115,
+      "loss": 5.7126,
+      "step": 7501
+    },
+    {
+      "epoch": 2.560409556313993,
+      "grad_norm": 2.6712331771850586,
+      "learning_rate": 0.00014653014789533562,
+      "loss": 6.0862,
+      "step": 7502
+    },
+    {
+      "epoch": 2.560750853242321,
+      "grad_norm": 2.6167094707489014,
+      "learning_rate": 0.00014641638225255973,
+      "loss": 6.5083,
+      "step": 7503
+    },
+    {
+      "epoch": 2.5610921501706487,
+      "grad_norm": 2.631967544555664,
+      "learning_rate": 0.00014630261660978386,
+      "loss": 5.9288,
+      "step": 7504
+    },
+    {
+      "epoch": 2.561433447098976,
+      "grad_norm": 2.8037562370300293,
+      "learning_rate": 0.00014618885096700796,
+      "loss": 5.0377,
+      "step": 7505
+    },
+    {
+      "epoch": 2.5617747440273035,
+      "grad_norm": 2.6779751777648926,
+      "learning_rate": 0.0001460750853242321,
+      "loss": 6.2956,
+      "step": 7506
+    },
+    {
+      "epoch": 2.5621160409556314,
+      "grad_norm": 2.6464788913726807,
+      "learning_rate": 0.0001459613196814562,
+      "loss": 5.9083,
+      "step": 7507
+    },
+    {
+      "epoch": 2.5624573378839592,
+      "grad_norm": 2.7481160163879395,
+      "learning_rate": 0.00014584755403868033,
+      "loss": 6.2735,
+      "step": 7508
+    },
+    {
+      "epoch": 2.5627986348122866,
+      "grad_norm": 2.6728515625,
+      "learning_rate": 0.00014573378839590444,
+      "loss": 6.0976,
+      "step": 7509
+    },
+    {
+      "epoch": 2.5631399317406145,
+      "grad_norm": 3.7015953063964844,
+      "learning_rate": 0.00014562002275312854,
+      "loss": 3.9438,
+      "step": 7510
+    },
+    {
+      "epoch": 2.563481228668942,
+      "grad_norm": 4.3790693283081055,
+      "learning_rate": 0.00014550625711035267,
+      "loss": 4.7137,
+      "step": 7511
+    },
+    {
+      "epoch": 2.5638225255972698,
+      "grad_norm": 2.6874852180480957,
+      "learning_rate": 0.0001453924914675768,
+      "loss": 6.2475,
+      "step": 7512
+    },
+    {
+      "epoch": 2.564163822525597,
+      "grad_norm": 2.693321943283081,
+      "learning_rate": 0.0001452787258248009,
+      "loss": 6.2983,
+      "step": 7513
+    },
+    {
+      "epoch": 2.564505119453925,
+      "grad_norm": 2.621973752975464,
+      "learning_rate": 0.00014516496018202502,
+      "loss": 6.1115,
+      "step": 7514
+    },
+    {
+      "epoch": 2.5648464163822524,
+      "grad_norm": 2.658691167831421,
+      "learning_rate": 0.00014505119453924915,
+      "loss": 5.8714,
+      "step": 7515
+    },
+    {
+      "epoch": 2.5651877133105803,
+      "grad_norm": 2.660804271697998,
+      "learning_rate": 0.00014493742889647328,
+      "loss": 6.6816,
+      "step": 7516
+    },
+    {
+      "epoch": 2.565529010238908,
+      "grad_norm": 2.6108481884002686,
+      "learning_rate": 0.00014482366325369739,
+      "loss": 5.8989,
+      "step": 7517
+    },
+    {
+      "epoch": 2.5658703071672355,
+      "grad_norm": 2.703063488006592,
+      "learning_rate": 0.00014470989761092152,
+      "loss": 5.9069,
+      "step": 7518
+    },
+    {
+      "epoch": 2.566211604095563,
+      "grad_norm": 2.910573959350586,
+      "learning_rate": 0.00014459613196814562,
+      "loss": 5.1331,
+      "step": 7519
+    },
+    {
+      "epoch": 2.5665529010238908,
+      "grad_norm": 2.6214263439178467,
+      "learning_rate": 0.00014448236632536973,
+      "loss": 6.2886,
+      "step": 7520
+    },
+    {
+      "epoch": 2.5668941979522186,
+      "grad_norm": 2.604752779006958,
+      "learning_rate": 0.00014436860068259386,
+      "loss": 6.0879,
+      "step": 7521
+    },
+    {
+      "epoch": 2.567235494880546,
+      "grad_norm": 2.679976224899292,
+      "learning_rate": 0.000144254835039818,
+      "loss": 6.2263,
+      "step": 7522
+    },
+    {
+      "epoch": 2.567576791808874,
+      "grad_norm": 2.592036485671997,
+      "learning_rate": 0.0001441410693970421,
+      "loss": 6.2702,
+      "step": 7523
+    },
+    {
+      "epoch": 2.5679180887372013,
+      "grad_norm": 2.691864490509033,
+      "learning_rate": 0.0001440273037542662,
+      "loss": 6.3147,
+      "step": 7524
+    },
+    {
+      "epoch": 2.568259385665529,
+      "grad_norm": 2.749069929122925,
+      "learning_rate": 0.00014391353811149033,
+      "loss": 5.9712,
+      "step": 7525
+    },
+    {
+      "epoch": 2.5686006825938565,
+      "grad_norm": 2.7472119331359863,
+      "learning_rate": 0.00014379977246871444,
+      "loss": 6.6555,
+      "step": 7526
+    },
+    {
+      "epoch": 2.5689419795221844,
+      "grad_norm": 2.7437517642974854,
+      "learning_rate": 0.00014368600682593857,
+      "loss": 5.8816,
+      "step": 7527
+    },
+    {
+      "epoch": 2.569283276450512,
+      "grad_norm": 2.7553818225860596,
+      "learning_rate": 0.0001435722411831627,
+      "loss": 5.6382,
+      "step": 7528
+    },
+    {
+      "epoch": 2.5696245733788396,
+      "grad_norm": 2.6735825538635254,
+      "learning_rate": 0.0001434584755403868,
+      "loss": 5.8185,
+      "step": 7529
+    },
+    {
+      "epoch": 2.5699658703071675,
+      "grad_norm": 2.8182194232940674,
+      "learning_rate": 0.0001433447098976109,
+      "loss": 5.0823,
+      "step": 7530
+    },
+    {
+      "epoch": 2.570307167235495,
+      "grad_norm": 2.7155697345733643,
+      "learning_rate": 0.00014323094425483504,
+      "loss": 5.3627,
+      "step": 7531
+    },
+    {
+      "epoch": 2.5706484641638223,
+      "grad_norm": 2.6774308681488037,
+      "learning_rate": 0.00014311717861205918,
+      "loss": 5.8861,
+      "step": 7532
+    },
+    {
+      "epoch": 2.57098976109215,
+      "grad_norm": 2.782548189163208,
+      "learning_rate": 0.00014300341296928328,
+      "loss": 5.6668,
+      "step": 7533
+    },
+    {
+      "epoch": 2.571331058020478,
+      "grad_norm": 2.7212352752685547,
+      "learning_rate": 0.00014288964732650739,
+      "loss": 6.0719,
+      "step": 7534
+    },
+    {
+      "epoch": 2.5716723549488054,
+      "grad_norm": 2.695491313934326,
+      "learning_rate": 0.00014277588168373152,
+      "loss": 6.0384,
+      "step": 7535
+    },
+    {
+      "epoch": 2.5720136518771333,
+      "grad_norm": 2.6470556259155273,
+      "learning_rate": 0.00014266211604095562,
+      "loss": 5.7107,
+      "step": 7536
+    },
+    {
+      "epoch": 2.5723549488054607,
+      "grad_norm": 2.6767444610595703,
+      "learning_rate": 0.00014254835039817976,
+      "loss": 5.8553,
+      "step": 7537
+    },
+    {
+      "epoch": 2.5726962457337885,
+      "grad_norm": 2.643503189086914,
+      "learning_rate": 0.0001424345847554039,
+      "loss": 6.1327,
+      "step": 7538
+    },
+    {
+      "epoch": 2.573037542662116,
+      "grad_norm": 2.6985726356506348,
+      "learning_rate": 0.000142320819112628,
+      "loss": 5.551,
+      "step": 7539
+    },
+    {
+      "epoch": 2.573378839590444,
+      "grad_norm": 2.657045364379883,
+      "learning_rate": 0.0001422070534698521,
+      "loss": 5.9295,
+      "step": 7540
+    },
+    {
+      "epoch": 2.573720136518771,
+      "grad_norm": 2.6590969562530518,
+      "learning_rate": 0.00014209328782707623,
+      "loss": 5.922,
+      "step": 7541
+    },
+    {
+      "epoch": 2.574061433447099,
+      "grad_norm": 2.6547012329101562,
+      "learning_rate": 0.00014197952218430036,
+      "loss": 6.1562,
+      "step": 7542
+    },
+    {
+      "epoch": 2.574402730375427,
+      "grad_norm": 2.6703603267669678,
+      "learning_rate": 0.00014186575654152444,
+      "loss": 5.9321,
+      "step": 7543
+    },
+    {
+      "epoch": 2.5747440273037543,
+      "grad_norm": 2.9982120990753174,
+      "learning_rate": 0.00014175199089874857,
+      "loss": 5.0249,
+      "step": 7544
+    },
+    {
+      "epoch": 2.5750853242320817,
+      "grad_norm": 2.565819263458252,
+      "learning_rate": 0.0001416382252559727,
+      "loss": 5.3644,
+      "step": 7545
+    },
+    {
+      "epoch": 2.5754266211604095,
+      "grad_norm": 2.6522724628448486,
+      "learning_rate": 0.0001415244596131968,
+      "loss": 6.1545,
+      "step": 7546
+    },
+    {
+      "epoch": 2.5757679180887374,
+      "grad_norm": 2.639831781387329,
+      "learning_rate": 0.00014141069397042094,
+      "loss": 5.8925,
+      "step": 7547
+    },
+    {
+      "epoch": 2.576109215017065,
+      "grad_norm": 2.7260892391204834,
+      "learning_rate": 0.00014129692832764505,
+      "loss": 6.6641,
+      "step": 7548
+    },
+    {
+      "epoch": 2.5764505119453927,
+      "grad_norm": 2.7644131183624268,
+      "learning_rate": 0.00014118316268486918,
+      "loss": 5.8881,
+      "step": 7549
+    },
+    {
+      "epoch": 2.57679180887372,
+      "grad_norm": 2.671483039855957,
+      "learning_rate": 0.00014106939704209328,
+      "loss": 6.047,
+      "step": 7550
+    },
+    {
+      "epoch": 2.577133105802048,
+      "grad_norm": 2.5989768505096436,
+      "learning_rate": 0.00014095563139931741,
+      "loss": 6.5126,
+      "step": 7551
+    },
+    {
+      "epoch": 2.5774744027303753,
+      "grad_norm": 2.6520800590515137,
+      "learning_rate": 0.00014084186575654155,
+      "loss": 5.6839,
+      "step": 7552
+    },
+    {
+      "epoch": 2.577815699658703,
+      "grad_norm": 2.7389848232269287,
+      "learning_rate": 0.00014072810011376562,
+      "loss": 6.3082,
+      "step": 7553
+    },
+    {
+      "epoch": 2.5781569965870306,
+      "grad_norm": 2.6571834087371826,
+      "learning_rate": 0.00014061433447098976,
+      "loss": 2.0284,
+      "step": 7554
+    },
+    {
+      "epoch": 2.5784982935153584,
+      "grad_norm": 2.616753339767456,
+      "learning_rate": 0.0001405005688282139,
+      "loss": 5.1408,
+      "step": 7555
+    },
+    {
+      "epoch": 2.5788395904436863,
+      "grad_norm": 2.6577699184417725,
+      "learning_rate": 0.000140386803185438,
+      "loss": 6.3211,
+      "step": 7556
+    },
+    {
+      "epoch": 2.5791808873720137,
+      "grad_norm": 2.816687822341919,
+      "learning_rate": 0.00014027303754266213,
+      "loss": 6.3943,
+      "step": 7557
+    },
+    {
+      "epoch": 2.579522184300341,
+      "grad_norm": 2.7218778133392334,
+      "learning_rate": 0.00014015927189988623,
+      "loss": 6.1001,
+      "step": 7558
+    },
+    {
+      "epoch": 2.579863481228669,
+      "grad_norm": 2.7211270332336426,
+      "learning_rate": 0.00014004550625711036,
+      "loss": 5.817,
+      "step": 7559
+    },
+    {
+      "epoch": 2.580204778156997,
+      "grad_norm": 2.6363632678985596,
+      "learning_rate": 0.00013993174061433447,
+      "loss": 6.3371,
+      "step": 7560
+    },
+    {
+      "epoch": 2.580546075085324,
+      "grad_norm": 2.642151355743408,
+      "learning_rate": 0.0001398179749715586,
+      "loss": 6.1385,
+      "step": 7561
+    },
+    {
+      "epoch": 2.580887372013652,
+      "grad_norm": 3.0516741275787354,
+      "learning_rate": 0.00013970420932878273,
+      "loss": 5.0531,
+      "step": 7562
+    },
+    {
+      "epoch": 2.5812286689419794,
+      "grad_norm": 2.72440505027771,
+      "learning_rate": 0.0001395904436860068,
+      "loss": 6.3252,
+      "step": 7563
+    },
+    {
+      "epoch": 2.5815699658703073,
+      "grad_norm": 2.5998916625976562,
+      "learning_rate": 0.00013947667804323094,
+      "loss": 5.7731,
+      "step": 7564
+    },
+    {
+      "epoch": 2.5819112627986347,
+      "grad_norm": 2.963616371154785,
+      "learning_rate": 0.00013936291240045507,
+      "loss": 4.7545,
+      "step": 7565
+    },
+    {
+      "epoch": 2.5822525597269625,
+      "grad_norm": 2.8518779277801514,
+      "learning_rate": 0.00013924914675767918,
+      "loss": 5.6993,
+      "step": 7566
+    },
+    {
+      "epoch": 2.58259385665529,
+      "grad_norm": 2.6663856506347656,
+      "learning_rate": 0.0001391353811149033,
+      "loss": 4.3027,
+      "step": 7567
+    },
+    {
+      "epoch": 2.582935153583618,
+      "grad_norm": 2.5740647315979004,
+      "learning_rate": 0.00013902161547212741,
+      "loss": 6.0794,
+      "step": 7568
+    },
+    {
+      "epoch": 2.5832764505119457,
+      "grad_norm": 2.7939860820770264,
+      "learning_rate": 0.00013890784982935155,
+      "loss": 5.3234,
+      "step": 7569
+    },
+    {
+      "epoch": 2.583617747440273,
+      "grad_norm": 2.7503066062927246,
+      "learning_rate": 0.00013879408418657565,
+      "loss": 6.3546,
+      "step": 7570
+    },
+    {
+      "epoch": 2.5839590443686005,
+      "grad_norm": 2.6611952781677246,
+      "learning_rate": 0.00013868031854379978,
+      "loss": 5.9687,
+      "step": 7571
+    },
+    {
+      "epoch": 2.5843003412969283,
+      "grad_norm": 2.741959810256958,
+      "learning_rate": 0.00013856655290102392,
+      "loss": 5.5272,
+      "step": 7572
+    },
+    {
+      "epoch": 2.584641638225256,
+      "grad_norm": 2.7685506343841553,
+      "learning_rate": 0.000138452787258248,
+      "loss": 5.9441,
+      "step": 7573
+    },
+    {
+      "epoch": 2.5849829351535836,
+      "grad_norm": 2.6498281955718994,
+      "learning_rate": 0.00013833902161547213,
+      "loss": 5.6183,
+      "step": 7574
+    },
+    {
+      "epoch": 2.5853242320819114,
+      "grad_norm": 2.6913111209869385,
+      "learning_rate": 0.00013822525597269626,
+      "loss": 6.2333,
+      "step": 7575
+    },
+    {
+      "epoch": 2.585665529010239,
+      "grad_norm": 2.6091015338897705,
+      "learning_rate": 0.00013811149032992036,
+      "loss": 6.4434,
+      "step": 7576
+    },
+    {
+      "epoch": 2.5860068259385667,
+      "grad_norm": 2.665632486343384,
+      "learning_rate": 0.00013799772468714447,
+      "loss": 6.0747,
+      "step": 7577
+    },
+    {
+      "epoch": 2.586348122866894,
+      "grad_norm": 2.7666921615600586,
+      "learning_rate": 0.0001378839590443686,
+      "loss": 6.2239,
+      "step": 7578
+    },
+    {
+      "epoch": 2.586689419795222,
+      "grad_norm": 2.6611504554748535,
+      "learning_rate": 0.00013777019340159273,
+      "loss": 6.6124,
+      "step": 7579
+    },
+    {
+      "epoch": 2.5870307167235493,
+      "grad_norm": 2.631537437438965,
+      "learning_rate": 0.00013765642775881684,
+      "loss": 6.2377,
+      "step": 7580
+    },
+    {
+      "epoch": 2.587372013651877,
+      "grad_norm": 2.7600057125091553,
+      "learning_rate": 0.00013754266211604097,
+      "loss": 5.9184,
+      "step": 7581
+    },
+    {
+      "epoch": 2.587713310580205,
+      "grad_norm": 2.6380791664123535,
+      "learning_rate": 0.00013742889647326507,
+      "loss": 6.3217,
+      "step": 7582
+    },
+    {
+      "epoch": 2.5880546075085324,
+      "grad_norm": 2.6468701362609863,
+      "learning_rate": 0.00013731513083048918,
+      "loss": 5.9776,
+      "step": 7583
+    },
+    {
+      "epoch": 2.58839590443686,
+      "grad_norm": 2.6581053733825684,
+      "learning_rate": 0.0001372013651877133,
+      "loss": 6.3691,
+      "step": 7584
+    },
+    {
+      "epoch": 2.5887372013651877,
+      "grad_norm": 2.5985143184661865,
+      "learning_rate": 0.00013708759954493744,
+      "loss": 6.0739,
+      "step": 7585
+    },
+    {
+      "epoch": 2.5890784982935156,
+      "grad_norm": 2.678396701812744,
+      "learning_rate": 0.00013697383390216155,
+      "loss": 6.2776,
+      "step": 7586
+    },
+    {
+      "epoch": 2.589419795221843,
+      "grad_norm": 2.7191009521484375,
+      "learning_rate": 0.00013686006825938565,
+      "loss": 6.1219,
+      "step": 7587
+    },
+    {
+      "epoch": 2.589761092150171,
+      "grad_norm": 2.738563299179077,
+      "learning_rate": 0.00013674630261660978,
+      "loss": 6.3903,
+      "step": 7588
+    },
+    {
+      "epoch": 2.590102389078498,
+      "grad_norm": 2.5974106788635254,
+      "learning_rate": 0.00013663253697383392,
+      "loss": 6.0863,
+      "step": 7589
+    },
+    {
+      "epoch": 2.590443686006826,
+      "grad_norm": 2.727854013442993,
+      "learning_rate": 0.00013651877133105802,
+      "loss": 5.466,
+      "step": 7590
+    },
+    {
+      "epoch": 2.5907849829351535,
+      "grad_norm": 2.6561155319213867,
+      "learning_rate": 0.00013640500568828215,
+      "loss": 5.8661,
+      "step": 7591
+    },
+    {
+      "epoch": 2.5911262798634813,
+      "grad_norm": 2.671494483947754,
+      "learning_rate": 0.00013629124004550626,
+      "loss": 5.6121,
+      "step": 7592
+    },
+    {
+      "epoch": 2.5914675767918087,
+      "grad_norm": 2.713571786880493,
+      "learning_rate": 0.00013617747440273036,
+      "loss": 6.0892,
+      "step": 7593
+    },
+    {
+      "epoch": 2.5918088737201366,
+      "grad_norm": 2.6231658458709717,
+      "learning_rate": 0.0001360637087599545,
+      "loss": 5.5849,
+      "step": 7594
+    },
+    {
+      "epoch": 2.5921501706484644,
+      "grad_norm": 4.061588287353516,
+      "learning_rate": 0.00013594994311717863,
+      "loss": 4.2174,
+      "step": 7595
+    },
+    {
+      "epoch": 2.592491467576792,
+      "grad_norm": 2.6725544929504395,
+      "learning_rate": 0.00013583617747440273,
+      "loss": 5.5262,
+      "step": 7596
+    },
+    {
+      "epoch": 2.5928327645051192,
+      "grad_norm": 3.150416612625122,
+      "learning_rate": 0.00013572241183162684,
+      "loss": 4.9608,
+      "step": 7597
+    },
+    {
+      "epoch": 2.593174061433447,
+      "grad_norm": 2.6914148330688477,
+      "learning_rate": 0.00013560864618885097,
+      "loss": 6.0067,
+      "step": 7598
+    },
+    {
+      "epoch": 2.593515358361775,
+      "grad_norm": 2.676713705062866,
+      "learning_rate": 0.0001354948805460751,
+      "loss": 6.4768,
+      "step": 7599
+    },
+    {
+      "epoch": 2.5938566552901023,
+      "grad_norm": 2.7440037727355957,
+      "learning_rate": 0.0001353811149032992,
+      "loss": 5.4661,
+      "step": 7600
+    },
+    {
+      "epoch": 2.59419795221843,
+      "grad_norm": 2.6399929523468018,
+      "learning_rate": 0.00013526734926052334,
+      "loss": 6.3171,
+      "step": 7601
+    },
+    {
+      "epoch": 2.5945392491467576,
+      "grad_norm": 2.746882438659668,
+      "learning_rate": 0.00013515358361774744,
+      "loss": 6.0754,
+      "step": 7602
+    },
+    {
+      "epoch": 2.5948805460750854,
+      "grad_norm": 2.638324022293091,
+      "learning_rate": 0.00013503981797497155,
+      "loss": 5.8914,
+      "step": 7603
+    },
+    {
+      "epoch": 2.595221843003413,
+      "grad_norm": 2.618677854537964,
+      "learning_rate": 0.00013492605233219568,
+      "loss": 6.0271,
+      "step": 7604
+    },
+    {
+      "epoch": 2.5955631399317407,
+      "grad_norm": 2.606271743774414,
+      "learning_rate": 0.0001348122866894198,
+      "loss": 4.9815,
+      "step": 7605
+    },
+    {
+      "epoch": 2.595904436860068,
+      "grad_norm": 3.786407470703125,
+      "learning_rate": 0.00013469852104664392,
+      "loss": 4.8562,
+      "step": 7606
+    },
+    {
+      "epoch": 2.596245733788396,
+      "grad_norm": 3.0423500537872314,
+      "learning_rate": 0.00013458475540386802,
+      "loss": 4.8359,
+      "step": 7607
+    },
+    {
+      "epoch": 2.596587030716724,
+      "grad_norm": 3.009005308151245,
+      "learning_rate": 0.00013447098976109215,
+      "loss": 4.6837,
+      "step": 7608
+    },
+    {
+      "epoch": 2.596928327645051,
+      "grad_norm": 2.638024091720581,
+      "learning_rate": 0.0001343572241183163,
+      "loss": 5.9415,
+      "step": 7609
+    },
+    {
+      "epoch": 2.5972696245733786,
+      "grad_norm": 2.710456371307373,
+      "learning_rate": 0.0001342434584755404,
+      "loss": 5.729,
+      "step": 7610
+    },
+    {
+      "epoch": 2.5976109215017065,
+      "grad_norm": 2.8217124938964844,
+      "learning_rate": 0.00013412969283276452,
+      "loss": 5.2043,
+      "step": 7611
+    },
+    {
+      "epoch": 2.5979522184300343,
+      "grad_norm": 2.6994223594665527,
+      "learning_rate": 0.00013401592718998863,
+      "loss": 5.7662,
+      "step": 7612
+    },
+    {
+      "epoch": 2.5982935153583617,
+      "grad_norm": 2.637618064880371,
+      "learning_rate": 0.00013390216154721273,
+      "loss": 5.9881,
+      "step": 7613
+    },
+    {
+      "epoch": 2.5986348122866896,
+      "grad_norm": 2.62662935256958,
+      "learning_rate": 0.00013378839590443687,
+      "loss": 5.9036,
+      "step": 7614
+    },
+    {
+      "epoch": 2.598976109215017,
+      "grad_norm": 2.6811110973358154,
+      "learning_rate": 0.000133674630261661,
+      "loss": 5.7672,
+      "step": 7615
+    },
+    {
+      "epoch": 2.599317406143345,
+      "grad_norm": 2.62439227104187,
+      "learning_rate": 0.00013356086461888508,
+      "loss": 6.594,
+      "step": 7616
+    },
+    {
+      "epoch": 2.5996587030716722,
+      "grad_norm": 2.640967607498169,
+      "learning_rate": 0.0001334470989761092,
+      "loss": 6.0876,
+      "step": 7617
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 2.6701509952545166,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 6.2532,
+      "step": 7618
+    },
+    {
+      "epoch": 2.6003412969283275,
+      "grad_norm": 2.628978967666626,
+      "learning_rate": 0.00013321956769055747,
+      "loss": 6.0009,
+      "step": 7619
+    },
+    {
+      "epoch": 2.6006825938566553,
+      "grad_norm": 3.894216775894165,
+      "learning_rate": 0.00013310580204778158,
+      "loss": 5.3864,
+      "step": 7620
+    },
+    {
+      "epoch": 2.601023890784983,
+      "grad_norm": 2.637104034423828,
+      "learning_rate": 0.00013299203640500568,
+      "loss": 6.1776,
+      "step": 7621
+    },
+    {
+      "epoch": 2.6013651877133106,
+      "grad_norm": 2.7615833282470703,
+      "learning_rate": 0.00013287827076222981,
+      "loss": 6.4661,
+      "step": 7622
+    },
+    {
+      "epoch": 2.601706484641638,
+      "grad_norm": 3.8650755882263184,
+      "learning_rate": 0.00013276450511945392,
+      "loss": 5.7443,
+      "step": 7623
+    },
+    {
+      "epoch": 2.602047781569966,
+      "grad_norm": 2.71655011177063,
+      "learning_rate": 0.00013265073947667805,
+      "loss": 6.3148,
+      "step": 7624
+    },
+    {
+      "epoch": 2.6023890784982937,
+      "grad_norm": 2.681018114089966,
+      "learning_rate": 0.00013253697383390218,
+      "loss": 6.3714,
+      "step": 7625
+    },
+    {
+      "epoch": 2.602730375426621,
+      "grad_norm": 2.730727195739746,
+      "learning_rate": 0.00013242320819112626,
+      "loss": 6.1858,
+      "step": 7626
+    },
+    {
+      "epoch": 2.603071672354949,
+      "grad_norm": 2.642637014389038,
+      "learning_rate": 0.0001323094425483504,
+      "loss": 5.5325,
+      "step": 7627
+    },
+    {
+      "epoch": 2.6034129692832764,
+      "grad_norm": 2.673652172088623,
+      "learning_rate": 0.00013219567690557452,
+      "loss": 6.3912,
+      "step": 7628
+    },
+    {
+      "epoch": 2.603754266211604,
+      "grad_norm": 2.6778078079223633,
+      "learning_rate": 0.00013208191126279863,
+      "loss": 6.2991,
+      "step": 7629
+    },
+    {
+      "epoch": 2.6040955631399316,
+      "grad_norm": 2.6532318592071533,
+      "learning_rate": 0.00013196814562002276,
+      "loss": 5.6819,
+      "step": 7630
+    },
+    {
+      "epoch": 2.6044368600682595,
+      "grad_norm": 3.36240291595459,
+      "learning_rate": 0.00013185437997724687,
+      "loss": 4.7713,
+      "step": 7631
+    },
+    {
+      "epoch": 2.604778156996587,
+      "grad_norm": 2.656712532043457,
+      "learning_rate": 0.000131740614334471,
+      "loss": 5.4265,
+      "step": 7632
+    },
+    {
+      "epoch": 2.6051194539249147,
+      "grad_norm": 2.6337647438049316,
+      "learning_rate": 0.0001316268486916951,
+      "loss": 6.0607,
+      "step": 7633
+    },
+    {
+      "epoch": 2.6054607508532426,
+      "grad_norm": 2.639120578765869,
+      "learning_rate": 0.00013151308304891924,
+      "loss": 5.8628,
+      "step": 7634
+    },
+    {
+      "epoch": 2.60580204778157,
+      "grad_norm": 2.6694211959838867,
+      "learning_rate": 0.00013139931740614337,
+      "loss": 6.3282,
+      "step": 7635
+    },
+    {
+      "epoch": 2.6061433447098974,
+      "grad_norm": 2.556114435195923,
+      "learning_rate": 0.00013128555176336745,
+      "loss": 5.6204,
+      "step": 7636
+    },
+    {
+      "epoch": 2.6064846416382252,
+      "grad_norm": 2.6191563606262207,
+      "learning_rate": 0.00013117178612059158,
+      "loss": 4.6587,
+      "step": 7637
+    },
+    {
+      "epoch": 2.606825938566553,
+      "grad_norm": 2.7384796142578125,
+      "learning_rate": 0.0001310580204778157,
+      "loss": 5.1479,
+      "step": 7638
+    },
+    {
+      "epoch": 2.6071672354948805,
+      "grad_norm": 2.6779263019561768,
+      "learning_rate": 0.00013094425483503981,
+      "loss": 5.5593,
+      "step": 7639
+    },
+    {
+      "epoch": 2.6075085324232083,
+      "grad_norm": 2.643181324005127,
+      "learning_rate": 0.00013083048919226395,
+      "loss": 6.0678,
+      "step": 7640
+    },
+    {
+      "epoch": 2.6078498293515358,
+      "grad_norm": 2.6609034538269043,
+      "learning_rate": 0.00013071672354948805,
+      "loss": 5.7646,
+      "step": 7641
+    },
+    {
+      "epoch": 2.6081911262798636,
+      "grad_norm": 2.6710946559906006,
+      "learning_rate": 0.00013060295790671218,
+      "loss": 5.7591,
+      "step": 7642
+    },
+    {
+      "epoch": 2.608532423208191,
+      "grad_norm": 2.7074642181396484,
+      "learning_rate": 0.0001304891922639363,
+      "loss": 5.8659,
+      "step": 7643
+    },
+    {
+      "epoch": 2.608873720136519,
+      "grad_norm": 2.806108236312866,
+      "learning_rate": 0.00013037542662116042,
+      "loss": 5.3068,
+      "step": 7644
+    },
+    {
+      "epoch": 2.6092150170648463,
+      "grad_norm": 2.8876841068267822,
+      "learning_rate": 0.00013026166097838455,
+      "loss": 4.9614,
+      "step": 7645
+    },
+    {
+      "epoch": 2.609556313993174,
+      "grad_norm": 2.6452441215515137,
+      "learning_rate": 0.00013014789533560863,
+      "loss": 4.6235,
+      "step": 7646
+    },
+    {
+      "epoch": 2.609897610921502,
+      "grad_norm": 2.64217209815979,
+      "learning_rate": 0.00013003412969283276,
+      "loss": 5.6776,
+      "step": 7647
+    },
+    {
+      "epoch": 2.6102389078498294,
+      "grad_norm": 2.6999876499176025,
+      "learning_rate": 0.0001299203640500569,
+      "loss": 5.8503,
+      "step": 7648
+    },
+    {
+      "epoch": 2.6105802047781568,
+      "grad_norm": 2.6543946266174316,
+      "learning_rate": 0.000129806598407281,
+      "loss": 6.0889,
+      "step": 7649
+    },
+    {
+      "epoch": 2.6109215017064846,
+      "grad_norm": 2.6520276069641113,
+      "learning_rate": 0.0001296928327645051,
+      "loss": 5.8006,
+      "step": 7650
+    },
+    {
+      "epoch": 2.6112627986348125,
+      "grad_norm": 2.520744562149048,
+      "learning_rate": 0.00012957906712172924,
+      "loss": 6.255,
+      "step": 7651
+    },
+    {
+      "epoch": 2.61160409556314,
+      "grad_norm": 2.6043193340301514,
+      "learning_rate": 0.00012946530147895337,
+      "loss": 5.8413,
+      "step": 7652
+    },
+    {
+      "epoch": 2.6119453924914677,
+      "grad_norm": 2.659938335418701,
+      "learning_rate": 0.00012935153583617747,
+      "loss": 5.6511,
+      "step": 7653
+    },
+    {
+      "epoch": 2.612286689419795,
+      "grad_norm": 2.6185102462768555,
+      "learning_rate": 0.0001292377701934016,
+      "loss": 6.55,
+      "step": 7654
+    },
+    {
+      "epoch": 2.612627986348123,
+      "grad_norm": 2.686222553253174,
+      "learning_rate": 0.0001291240045506257,
+      "loss": 5.7808,
+      "step": 7655
+    },
+    {
+      "epoch": 2.6129692832764504,
+      "grad_norm": 2.712735414505005,
+      "learning_rate": 0.00012901023890784982,
+      "loss": 5.6224,
+      "step": 7656
+    },
+    {
+      "epoch": 2.6133105802047782,
+      "grad_norm": 2.8246679306030273,
+      "learning_rate": 0.00012889647326507395,
+      "loss": 4.6806,
+      "step": 7657
+    },
+    {
+      "epoch": 2.6136518771331056,
+      "grad_norm": 2.6839938163757324,
+      "learning_rate": 0.00012878270762229808,
+      "loss": 5.9373,
+      "step": 7658
+    },
+    {
+      "epoch": 2.6139931740614335,
+      "grad_norm": 2.625178575515747,
+      "learning_rate": 0.00012866894197952218,
+      "loss": 6.0065,
+      "step": 7659
+    },
+    {
+      "epoch": 2.6143344709897613,
+      "grad_norm": 2.6909379959106445,
+      "learning_rate": 0.0001285551763367463,
+      "loss": 6.3687,
+      "step": 7660
+    },
+    {
+      "epoch": 2.6146757679180888,
+      "grad_norm": 2.5886874198913574,
+      "learning_rate": 0.00012844141069397042,
+      "loss": 5.5438,
+      "step": 7661
+    },
+    {
+      "epoch": 2.615017064846416,
+      "grad_norm": 2.6835293769836426,
+      "learning_rate": 0.00012832764505119455,
+      "loss": 6.2562,
+      "step": 7662
+    },
+    {
+      "epoch": 2.615358361774744,
+      "grad_norm": 2.7177085876464844,
+      "learning_rate": 0.00012821387940841866,
+      "loss": 6.1722,
+      "step": 7663
+    },
+    {
+      "epoch": 2.615699658703072,
+      "grad_norm": 2.651545763015747,
+      "learning_rate": 0.0001281001137656428,
+      "loss": 6.2457,
+      "step": 7664
+    },
+    {
+      "epoch": 2.6160409556313993,
+      "grad_norm": 2.6499643325805664,
+      "learning_rate": 0.0001279863481228669,
+      "loss": 5.9292,
+      "step": 7665
+    },
+    {
+      "epoch": 2.616382252559727,
+      "grad_norm": 2.6566162109375,
+      "learning_rate": 0.000127872582480091,
+      "loss": 5.9867,
+      "step": 7666
+    },
+    {
+      "epoch": 2.6167235494880545,
+      "grad_norm": 2.6786088943481445,
+      "learning_rate": 0.00012775881683731513,
+      "loss": 6.447,
+      "step": 7667
+    },
+    {
+      "epoch": 2.6170648464163824,
+      "grad_norm": 2.659654140472412,
+      "learning_rate": 0.00012764505119453926,
+      "loss": 5.8441,
+      "step": 7668
+    },
+    {
+      "epoch": 2.61740614334471,
+      "grad_norm": 2.6184911727905273,
+      "learning_rate": 0.00012753128555176337,
+      "loss": 5.8193,
+      "step": 7669
+    },
+    {
+      "epoch": 2.6177474402730376,
+      "grad_norm": 2.677710771560669,
+      "learning_rate": 0.00012741751990898747,
+      "loss": 6.0518,
+      "step": 7670
+    },
+    {
+      "epoch": 2.618088737201365,
+      "grad_norm": 2.705430030822754,
+      "learning_rate": 0.0001273037542662116,
+      "loss": 6.224,
+      "step": 7671
+    },
+    {
+      "epoch": 2.618430034129693,
+      "grad_norm": 2.7365448474884033,
+      "learning_rate": 0.00012718998862343574,
+      "loss": 6.3756,
+      "step": 7672
+    },
+    {
+      "epoch": 2.6187713310580207,
+      "grad_norm": 2.6876437664031982,
+      "learning_rate": 0.00012707622298065984,
+      "loss": 6.07,
+      "step": 7673
+    },
+    {
+      "epoch": 2.619112627986348,
+      "grad_norm": 2.6189560890197754,
+      "learning_rate": 0.00012696245733788398,
+      "loss": 6.1245,
+      "step": 7674
+    },
+    {
+      "epoch": 2.6194539249146755,
+      "grad_norm": 3.7583062648773193,
+      "learning_rate": 0.00012684869169510808,
+      "loss": 5.6039,
+      "step": 7675
+    },
+    {
+      "epoch": 2.6197952218430034,
+      "grad_norm": 2.7220373153686523,
+      "learning_rate": 0.00012673492605233219,
+      "loss": 6.0232,
+      "step": 7676
+    },
+    {
+      "epoch": 2.6201365187713312,
+      "grad_norm": 2.691070556640625,
+      "learning_rate": 0.00012662116040955632,
+      "loss": 6.2148,
+      "step": 7677
+    },
+    {
+      "epoch": 2.6204778156996587,
+      "grad_norm": 2.6679129600524902,
+      "learning_rate": 0.00012650739476678045,
+      "loss": 6.1872,
+      "step": 7678
+    },
+    {
+      "epoch": 2.6208191126279865,
+      "grad_norm": 2.7614474296569824,
+      "learning_rate": 0.00012639362912400455,
+      "loss": 5.669,
+      "step": 7679
+    },
+    {
+      "epoch": 2.621160409556314,
+      "grad_norm": 2.7429089546203613,
+      "learning_rate": 0.00012627986348122866,
+      "loss": 5.5727,
+      "step": 7680
+    },
+    {
+      "epoch": 2.6215017064846418,
+      "grad_norm": 2.684476613998413,
+      "learning_rate": 0.0001261660978384528,
+      "loss": 5.7475,
+      "step": 7681
+    },
+    {
+      "epoch": 2.621843003412969,
+      "grad_norm": 2.627525806427002,
+      "learning_rate": 0.00012605233219567692,
+      "loss": 6.2481,
+      "step": 7682
+    },
+    {
+      "epoch": 2.622184300341297,
+      "grad_norm": 2.691311836242676,
+      "learning_rate": 0.00012593856655290103,
+      "loss": 6.0962,
+      "step": 7683
+    },
+    {
+      "epoch": 2.6225255972696244,
+      "grad_norm": 2.6626622676849365,
+      "learning_rate": 0.00012582480091012513,
+      "loss": 6.2863,
+      "step": 7684
+    },
+    {
+      "epoch": 2.6228668941979523,
+      "grad_norm": 2.6201605796813965,
+      "learning_rate": 0.00012571103526734927,
+      "loss": 5.6584,
+      "step": 7685
+    },
+    {
+      "epoch": 2.62320819112628,
+      "grad_norm": 2.704043388366699,
+      "learning_rate": 0.00012559726962457337,
+      "loss": 6.1213,
+      "step": 7686
+    },
+    {
+      "epoch": 2.6235494880546075,
+      "grad_norm": 2.5595293045043945,
+      "learning_rate": 0.0001254835039817975,
+      "loss": 6.3741,
+      "step": 7687
+    },
+    {
+      "epoch": 2.623890784982935,
+      "grad_norm": 5.192468643188477,
+      "learning_rate": 0.00012536973833902163,
+      "loss": 4.9172,
+      "step": 7688
+    },
+    {
+      "epoch": 2.624232081911263,
+      "grad_norm": 2.794823169708252,
+      "learning_rate": 0.0001252559726962457,
+      "loss": 5.9293,
+      "step": 7689
+    },
+    {
+      "epoch": 2.6245733788395906,
+      "grad_norm": 2.6778604984283447,
+      "learning_rate": 0.00012514220705346984,
+      "loss": 6.2208,
+      "step": 7690
+    },
+    {
+      "epoch": 2.624914675767918,
+      "grad_norm": 2.6402392387390137,
+      "learning_rate": 0.00012502844141069398,
+      "loss": 5.5919,
+      "step": 7691
+    },
+    {
+      "epoch": 2.625255972696246,
+      "grad_norm": 2.6135995388031006,
+      "learning_rate": 0.0001249146757679181,
+      "loss": 6.367,
+      "step": 7692
+    },
+    {
+      "epoch": 2.6255972696245733,
+      "grad_norm": 2.745082378387451,
+      "learning_rate": 0.0001248009101251422,
+      "loss": 5.8043,
+      "step": 7693
+    },
+    {
+      "epoch": 2.625938566552901,
+      "grad_norm": 2.596097946166992,
+      "learning_rate": 0.00012468714448236632,
+      "loss": 5.4991,
+      "step": 7694
+    },
+    {
+      "epoch": 2.6262798634812285,
+      "grad_norm": 2.6917331218719482,
+      "learning_rate": 0.00012457337883959045,
+      "loss": 6.1673,
+      "step": 7695
+    },
+    {
+      "epoch": 2.6266211604095564,
+      "grad_norm": 2.624852180480957,
+      "learning_rate": 0.00012445961319681456,
+      "loss": 6.381,
+      "step": 7696
+    },
+    {
+      "epoch": 2.626962457337884,
+      "grad_norm": 2.734668493270874,
+      "learning_rate": 0.0001243458475540387,
+      "loss": 5.8872,
+      "step": 7697
+    },
+    {
+      "epoch": 2.6273037542662117,
+      "grad_norm": 2.7033309936523438,
+      "learning_rate": 0.0001242320819112628,
+      "loss": 5.7082,
+      "step": 7698
+    },
+    {
+      "epoch": 2.6276450511945395,
+      "grad_norm": 2.640166759490967,
+      "learning_rate": 0.00012411831626848692,
+      "loss": 5.962,
+      "step": 7699
+    },
+    {
+      "epoch": 2.627986348122867,
+      "grad_norm": 2.584916591644287,
+      "learning_rate": 0.00012400455062571103,
+      "loss": 6.1359,
+      "step": 7700
+    },
+    {
+      "epoch": 2.6283276450511943,
+      "grad_norm": 2.798306465148926,
+      "learning_rate": 0.00012389078498293516,
+      "loss": 6.2166,
+      "step": 7701
+    },
+    {
+      "epoch": 2.628668941979522,
+      "grad_norm": 2.539496660232544,
+      "learning_rate": 0.0001237770193401593,
+      "loss": 5.7527,
+      "step": 7702
+    },
+    {
+      "epoch": 2.62901023890785,
+      "grad_norm": 2.606154680252075,
+      "learning_rate": 0.0001236632536973834,
+      "loss": 6.1712,
+      "step": 7703
+    },
+    {
+      "epoch": 2.6293515358361774,
+      "grad_norm": 4.360844612121582,
+      "learning_rate": 0.0001235494880546075,
+      "loss": 5.0891,
+      "step": 7704
+    },
+    {
+      "epoch": 2.6296928327645053,
+      "grad_norm": 2.640214443206787,
+      "learning_rate": 0.00012343572241183164,
+      "loss": 5.8549,
+      "step": 7705
+    },
+    {
+      "epoch": 2.6300341296928327,
+      "grad_norm": 2.8788836002349854,
+      "learning_rate": 0.00012332195676905574,
+      "loss": 5.118,
+      "step": 7706
+    },
+    {
+      "epoch": 2.6303754266211605,
+      "grad_norm": 2.637471914291382,
+      "learning_rate": 0.00012320819112627987,
+      "loss": 6.1626,
+      "step": 7707
+    },
+    {
+      "epoch": 2.630716723549488,
+      "grad_norm": 2.6709108352661133,
+      "learning_rate": 0.00012309442548350398,
+      "loss": 5.7018,
+      "step": 7708
+    },
+    {
+      "epoch": 2.631058020477816,
+      "grad_norm": 2.575793504714966,
+      "learning_rate": 0.0001229806598407281,
+      "loss": 5.8412,
+      "step": 7709
+    },
+    {
+      "epoch": 2.631399317406143,
+      "grad_norm": 2.7106573581695557,
+      "learning_rate": 0.00012286689419795221,
+      "loss": 5.151,
+      "step": 7710
+    },
+    {
+      "epoch": 2.631740614334471,
+      "grad_norm": 2.6999032497406006,
+      "learning_rate": 0.00012275312855517635,
+      "loss": 6.1057,
+      "step": 7711
+    },
+    {
+      "epoch": 2.632081911262799,
+      "grad_norm": 2.700033664703369,
+      "learning_rate": 0.00012263936291240045,
+      "loss": 6.1486,
+      "step": 7712
+    },
+    {
+      "epoch": 2.6324232081911263,
+      "grad_norm": 2.705191135406494,
+      "learning_rate": 0.00012252559726962458,
+      "loss": 5.427,
+      "step": 7713
+    },
+    {
+      "epoch": 2.6327645051194537,
+      "grad_norm": 2.5965819358825684,
+      "learning_rate": 0.0001224118316268487,
+      "loss": 6.0963,
+      "step": 7714
+    },
+    {
+      "epoch": 2.6331058020477816,
+      "grad_norm": 2.7236809730529785,
+      "learning_rate": 0.00012229806598407282,
+      "loss": 5.8642,
+      "step": 7715
+    },
+    {
+      "epoch": 2.6334470989761094,
+      "grad_norm": 2.6513259410858154,
+      "learning_rate": 0.00012218430034129693,
+      "loss": 6.547,
+      "step": 7716
+    },
+    {
+      "epoch": 2.633788395904437,
+      "grad_norm": 2.6121363639831543,
+      "learning_rate": 0.00012207053469852103,
+      "loss": 5.9717,
+      "step": 7717
+    },
+    {
+      "epoch": 2.6341296928327647,
+      "grad_norm": 2.756356716156006,
+      "learning_rate": 0.00012195676905574516,
+      "loss": 6.0155,
+      "step": 7718
+    },
+    {
+      "epoch": 2.634470989761092,
+      "grad_norm": 2.6240437030792236,
+      "learning_rate": 0.0001218430034129693,
+      "loss": 5.6416,
+      "step": 7719
+    },
+    {
+      "epoch": 2.63481228668942,
+      "grad_norm": 2.6901957988739014,
+      "learning_rate": 0.0001217292377701934,
+      "loss": 6.0908,
+      "step": 7720
+    },
+    {
+      "epoch": 2.6351535836177473,
+      "grad_norm": 2.5926854610443115,
+      "learning_rate": 0.00012161547212741753,
+      "loss": 5.8315,
+      "step": 7721
+    },
+    {
+      "epoch": 2.635494880546075,
+      "grad_norm": 2.6455578804016113,
+      "learning_rate": 0.00012150170648464164,
+      "loss": 6.1936,
+      "step": 7722
+    },
+    {
+      "epoch": 2.6358361774744026,
+      "grad_norm": 2.6685874462127686,
+      "learning_rate": 0.00012138794084186575,
+      "loss": 6.2457,
+      "step": 7723
+    },
+    {
+      "epoch": 2.6361774744027304,
+      "grad_norm": 2.5576179027557373,
+      "learning_rate": 0.00012127417519908989,
+      "loss": 6.1212,
+      "step": 7724
+    },
+    {
+      "epoch": 2.6365187713310583,
+      "grad_norm": 2.6035420894622803,
+      "learning_rate": 0.00012116040955631399,
+      "loss": 6.173,
+      "step": 7725
+    },
+    {
+      "epoch": 2.6368600682593857,
+      "grad_norm": 2.7565255165100098,
+      "learning_rate": 0.00012104664391353812,
+      "loss": 5.909,
+      "step": 7726
+    },
+    {
+      "epoch": 2.637201365187713,
+      "grad_norm": 2.3989455699920654,
+      "learning_rate": 0.00012093287827076223,
+      "loss": 5.1077,
+      "step": 7727
+    },
+    {
+      "epoch": 2.637542662116041,
+      "grad_norm": 2.7297885417938232,
+      "learning_rate": 0.00012081911262798635,
+      "loss": 5.1734,
+      "step": 7728
+    },
+    {
+      "epoch": 2.637883959044369,
+      "grad_norm": 2.634373664855957,
+      "learning_rate": 0.00012070534698521047,
+      "loss": 5.418,
+      "step": 7729
+    },
+    {
+      "epoch": 2.638225255972696,
+      "grad_norm": 2.557570457458496,
+      "learning_rate": 0.00012059158134243458,
+      "loss": 6.1528,
+      "step": 7730
+    },
+    {
+      "epoch": 2.638566552901024,
+      "grad_norm": 2.647575855255127,
+      "learning_rate": 0.00012047781569965872,
+      "loss": 6.0692,
+      "step": 7731
+    },
+    {
+      "epoch": 2.6389078498293514,
+      "grad_norm": 2.5773942470550537,
+      "learning_rate": 0.00012036405005688282,
+      "loss": 5.6637,
+      "step": 7732
+    },
+    {
+      "epoch": 2.6392491467576793,
+      "grad_norm": 2.692434549331665,
+      "learning_rate": 0.00012025028441410694,
+      "loss": 6.029,
+      "step": 7733
+    },
+    {
+      "epoch": 2.6395904436860067,
+      "grad_norm": 2.6368980407714844,
+      "learning_rate": 0.00012013651877133106,
+      "loss": 6.3336,
+      "step": 7734
+    },
+    {
+      "epoch": 2.6399317406143346,
+      "grad_norm": 2.6980204582214355,
+      "learning_rate": 0.00012002275312855518,
+      "loss": 6.1817,
+      "step": 7735
+    },
+    {
+      "epoch": 2.640273037542662,
+      "grad_norm": 2.6697869300842285,
+      "learning_rate": 0.00011990898748577931,
+      "loss": 6.2257,
+      "step": 7736
+    },
+    {
+      "epoch": 2.64061433447099,
+      "grad_norm": 2.5750112533569336,
+      "learning_rate": 0.00011979522184300341,
+      "loss": 5.8854,
+      "step": 7737
+    },
+    {
+      "epoch": 2.6409556313993177,
+      "grad_norm": 2.6087377071380615,
+      "learning_rate": 0.00011968145620022753,
+      "loss": 5.9283,
+      "step": 7738
+    },
+    {
+      "epoch": 2.641296928327645,
+      "grad_norm": 2.607609272003174,
+      "learning_rate": 0.00011956769055745165,
+      "loss": 6.1513,
+      "step": 7739
+    },
+    {
+      "epoch": 2.6416382252559725,
+      "grad_norm": 2.6486034393310547,
+      "learning_rate": 0.00011945392491467577,
+      "loss": 5.6541,
+      "step": 7740
+    },
+    {
+      "epoch": 2.6419795221843003,
+      "grad_norm": 2.6967036724090576,
+      "learning_rate": 0.0001193401592718999,
+      "loss": 6.3013,
+      "step": 7741
+    },
+    {
+      "epoch": 2.642320819112628,
+      "grad_norm": 2.7555899620056152,
+      "learning_rate": 0.000119226393629124,
+      "loss": 5.9503,
+      "step": 7742
+    },
+    {
+      "epoch": 2.6426621160409556,
+      "grad_norm": 2.6998205184936523,
+      "learning_rate": 0.00011911262798634812,
+      "loss": 5.8985,
+      "step": 7743
+    },
+    {
+      "epoch": 2.6430034129692834,
+      "grad_norm": 2.662689208984375,
+      "learning_rate": 0.00011899886234357224,
+      "loss": 6.3067,
+      "step": 7744
+    },
+    {
+      "epoch": 2.643344709897611,
+      "grad_norm": 2.587956190109253,
+      "learning_rate": 0.00011888509670079636,
+      "loss": 6.137,
+      "step": 7745
+    },
+    {
+      "epoch": 2.6436860068259387,
+      "grad_norm": 2.9233665466308594,
+      "learning_rate": 0.00011877133105802048,
+      "loss": 5.8408,
+      "step": 7746
+    },
+    {
+      "epoch": 2.644027303754266,
+      "grad_norm": 2.765820264816284,
+      "learning_rate": 0.0001186575654152446,
+      "loss": 6.4021,
+      "step": 7747
+    },
+    {
+      "epoch": 2.644368600682594,
+      "grad_norm": 2.7021546363830566,
+      "learning_rate": 0.00011854379977246872,
+      "loss": 6.3432,
+      "step": 7748
+    },
+    {
+      "epoch": 2.6447098976109213,
+      "grad_norm": 2.628324031829834,
+      "learning_rate": 0.00011843003412969284,
+      "loss": 5.5596,
+      "step": 7749
+    },
+    {
+      "epoch": 2.645051194539249,
+      "grad_norm": 2.6931304931640625,
+      "learning_rate": 0.00011831626848691695,
+      "loss": 6.4549,
+      "step": 7750
+    },
+    {
+      "epoch": 2.645392491467577,
+      "grad_norm": 2.3119070529937744,
+      "learning_rate": 0.00011820250284414107,
+      "loss": 4.2596,
+      "step": 7751
+    },
+    {
+      "epoch": 2.6457337883959045,
+      "grad_norm": 2.604924440383911,
+      "learning_rate": 0.00011808873720136519,
+      "loss": 5.8979,
+      "step": 7752
+    },
+    {
+      "epoch": 2.646075085324232,
+      "grad_norm": 2.697232723236084,
+      "learning_rate": 0.00011797497155858931,
+      "loss": 5.1108,
+      "step": 7753
+    },
+    {
+      "epoch": 2.6464163822525597,
+      "grad_norm": 2.6660239696502686,
+      "learning_rate": 0.00011786120591581343,
+      "loss": 6.2313,
+      "step": 7754
+    },
+    {
+      "epoch": 2.6467576791808876,
+      "grad_norm": 2.5658226013183594,
+      "learning_rate": 0.00011774744027303755,
+      "loss": 5.9833,
+      "step": 7755
+    },
+    {
+      "epoch": 2.647098976109215,
+      "grad_norm": 2.658269166946411,
+      "learning_rate": 0.00011763367463026167,
+      "loss": 5.9185,
+      "step": 7756
+    },
+    {
+      "epoch": 2.647440273037543,
+      "grad_norm": 2.5440711975097656,
+      "learning_rate": 0.00011751990898748578,
+      "loss": 5.47,
+      "step": 7757
+    },
+    {
+      "epoch": 2.64778156996587,
+      "grad_norm": 2.6822307109832764,
+      "learning_rate": 0.0001174061433447099,
+      "loss": 6.2749,
+      "step": 7758
+    },
+    {
+      "epoch": 2.648122866894198,
+      "grad_norm": 2.758451461791992,
+      "learning_rate": 0.00011729237770193402,
+      "loss": 6.2035,
+      "step": 7759
+    },
+    {
+      "epoch": 2.6484641638225255,
+      "grad_norm": 2.591820240020752,
+      "learning_rate": 0.00011717861205915814,
+      "loss": 6.4704,
+      "step": 7760
+    },
+    {
+      "epoch": 2.6488054607508533,
+      "grad_norm": 2.8156046867370605,
+      "learning_rate": 0.00011706484641638224,
+      "loss": 6.4334,
+      "step": 7761
+    },
+    {
+      "epoch": 2.6491467576791807,
+      "grad_norm": 2.6117329597473145,
+      "learning_rate": 0.00011695108077360638,
+      "loss": 6.3525,
+      "step": 7762
+    },
+    {
+      "epoch": 2.6494880546075086,
+      "grad_norm": 2.6518020629882812,
+      "learning_rate": 0.00011683731513083048,
+      "loss": 5.9983,
+      "step": 7763
+    },
+    {
+      "epoch": 2.6498293515358364,
+      "grad_norm": 2.606523275375366,
+      "learning_rate": 0.00011672354948805461,
+      "loss": 5.6909,
+      "step": 7764
+    },
+    {
+      "epoch": 2.650170648464164,
+      "grad_norm": 2.7865447998046875,
+      "learning_rate": 0.00011660978384527873,
+      "loss": 3.9492,
+      "step": 7765
+    },
+    {
+      "epoch": 2.6505119453924912,
+      "grad_norm": 2.6306912899017334,
+      "learning_rate": 0.00011649601820250284,
+      "loss": 5.3691,
+      "step": 7766
+    },
+    {
+      "epoch": 2.650853242320819,
+      "grad_norm": 2.5862772464752197,
+      "learning_rate": 0.00011638225255972697,
+      "loss": 6.0451,
+      "step": 7767
+    },
+    {
+      "epoch": 2.651194539249147,
+      "grad_norm": 2.6710424423217773,
+      "learning_rate": 0.00011626848691695107,
+      "loss": 6.339,
+      "step": 7768
+    },
+    {
+      "epoch": 2.6515358361774743,
+      "grad_norm": 2.731431484222412,
+      "learning_rate": 0.0001161547212741752,
+      "loss": 6.5248,
+      "step": 7769
+    },
+    {
+      "epoch": 2.651877133105802,
+      "grad_norm": 2.6861982345581055,
+      "learning_rate": 0.00011604095563139932,
+      "loss": 5.8373,
+      "step": 7770
+    },
+    {
+      "epoch": 2.6522184300341296,
+      "grad_norm": 2.622677803039551,
+      "learning_rate": 0.00011592718998862343,
+      "loss": 6.0151,
+      "step": 7771
+    },
+    {
+      "epoch": 2.6525597269624575,
+      "grad_norm": 2.653522491455078,
+      "learning_rate": 0.00011581342434584756,
+      "loss": 6.2623,
+      "step": 7772
+    },
+    {
+      "epoch": 2.652901023890785,
+      "grad_norm": 2.724036693572998,
+      "learning_rate": 0.00011569965870307167,
+      "loss": 5.8716,
+      "step": 7773
+    },
+    {
+      "epoch": 2.6532423208191127,
+      "grad_norm": 2.6914572715759277,
+      "learning_rate": 0.0001155858930602958,
+      "loss": 6.0782,
+      "step": 7774
+    },
+    {
+      "epoch": 2.65358361774744,
+      "grad_norm": 2.528007984161377,
+      "learning_rate": 0.00011547212741751992,
+      "loss": 5.6457,
+      "step": 7775
+    },
+    {
+      "epoch": 2.653924914675768,
+      "grad_norm": 2.635348081588745,
+      "learning_rate": 0.00011535836177474402,
+      "loss": 6.2753,
+      "step": 7776
+    },
+    {
+      "epoch": 2.654266211604096,
+      "grad_norm": 2.6857311725616455,
+      "learning_rate": 0.00011524459613196815,
+      "loss": 5.6202,
+      "step": 7777
+    },
+    {
+      "epoch": 2.654607508532423,
+      "grad_norm": 2.612197160720825,
+      "learning_rate": 0.00011513083048919226,
+      "loss": 5.9211,
+      "step": 7778
+    },
+    {
+      "epoch": 2.6549488054607506,
+      "grad_norm": 2.6178486347198486,
+      "learning_rate": 0.00011501706484641639,
+      "loss": 4.9022,
+      "step": 7779
+    },
+    {
+      "epoch": 2.6552901023890785,
+      "grad_norm": 2.6042556762695312,
+      "learning_rate": 0.0001149032992036405,
+      "loss": 6.2243,
+      "step": 7780
+    },
+    {
+      "epoch": 2.6556313993174063,
+      "grad_norm": 2.5869898796081543,
+      "learning_rate": 0.00011478953356086461,
+      "loss": 6.0258,
+      "step": 7781
+    },
+    {
+      "epoch": 2.6559726962457337,
+      "grad_norm": 2.7476601600646973,
+      "learning_rate": 0.00011467576791808875,
+      "loss": 5.2693,
+      "step": 7782
+    },
+    {
+      "epoch": 2.6563139931740616,
+      "grad_norm": 2.492094039916992,
+      "learning_rate": 0.00011456200227531285,
+      "loss": 6.0099,
+      "step": 7783
+    },
+    {
+      "epoch": 2.656655290102389,
+      "grad_norm": 2.6337618827819824,
+      "learning_rate": 0.00011444823663253698,
+      "loss": 6.8608,
+      "step": 7784
+    },
+    {
+      "epoch": 2.656996587030717,
+      "grad_norm": 2.559250593185425,
+      "learning_rate": 0.00011433447098976109,
+      "loss": 5.9022,
+      "step": 7785
+    },
+    {
+      "epoch": 2.6573378839590442,
+      "grad_norm": 2.6352384090423584,
+      "learning_rate": 0.0001142207053469852,
+      "loss": 6.3508,
+      "step": 7786
+    },
+    {
+      "epoch": 2.657679180887372,
+      "grad_norm": 2.701533317565918,
+      "learning_rate": 0.00011410693970420934,
+      "loss": 6.1916,
+      "step": 7787
+    },
+    {
+      "epoch": 2.6580204778156995,
+      "grad_norm": 2.6848697662353516,
+      "learning_rate": 0.00011399317406143344,
+      "loss": 6.2393,
+      "step": 7788
+    },
+    {
+      "epoch": 2.6583617747440274,
+      "grad_norm": 3.9241063594818115,
+      "learning_rate": 0.00011387940841865758,
+      "loss": 5.0438,
+      "step": 7789
+    },
+    {
+      "epoch": 2.658703071672355,
+      "grad_norm": 2.6796340942382812,
+      "learning_rate": 0.00011376564277588168,
+      "loss": 5.3564,
+      "step": 7790
+    },
+    {
+      "epoch": 2.6590443686006826,
+      "grad_norm": 2.6478474140167236,
+      "learning_rate": 0.0001136518771331058,
+      "loss": 6.3117,
+      "step": 7791
+    },
+    {
+      "epoch": 2.65938566552901,
+      "grad_norm": 2.6212310791015625,
+      "learning_rate": 0.00011353811149032993,
+      "loss": 6.6302,
+      "step": 7792
+    },
+    {
+      "epoch": 2.659726962457338,
+      "grad_norm": 2.6432223320007324,
+      "learning_rate": 0.00011342434584755404,
+      "loss": 6.1271,
+      "step": 7793
+    },
+    {
+      "epoch": 2.6600682593856657,
+      "grad_norm": 2.6077170372009277,
+      "learning_rate": 0.00011331058020477817,
+      "loss": 6.1778,
+      "step": 7794
+    },
+    {
+      "epoch": 2.660409556313993,
+      "grad_norm": 2.6975767612457275,
+      "learning_rate": 0.00011319681456200227,
+      "loss": 6.2841,
+      "step": 7795
+    },
+    {
+      "epoch": 2.660750853242321,
+      "grad_norm": 2.6964032649993896,
+      "learning_rate": 0.00011308304891922639,
+      "loss": 5.8834,
+      "step": 7796
+    },
+    {
+      "epoch": 2.6610921501706484,
+      "grad_norm": 2.594709873199463,
+      "learning_rate": 0.00011296928327645051,
+      "loss": 5.8787,
+      "step": 7797
+    },
+    {
+      "epoch": 2.6614334470989762,
+      "grad_norm": 2.6809041500091553,
+      "learning_rate": 0.00011285551763367463,
+      "loss": 6.3782,
+      "step": 7798
+    },
+    {
+      "epoch": 2.6617747440273036,
+      "grad_norm": 2.5971288681030273,
+      "learning_rate": 0.00011274175199089876,
+      "loss": 6.1397,
+      "step": 7799
+    },
+    {
+      "epoch": 2.6621160409556315,
+      "grad_norm": 2.648439407348633,
+      "learning_rate": 0.00011262798634812287,
+      "loss": 5.7752,
+      "step": 7800
+    },
+    {
+      "epoch": 2.662457337883959,
+      "grad_norm": 2.61542010307312,
+      "learning_rate": 0.00011251422070534698,
+      "loss": 5.6754,
+      "step": 7801
+    },
+    {
+      "epoch": 2.6627986348122867,
+      "grad_norm": 2.5414392948150635,
+      "learning_rate": 0.0001124004550625711,
+      "loss": 5.9982,
+      "step": 7802
+    },
+    {
+      "epoch": 2.6631399317406146,
+      "grad_norm": 2.6229312419891357,
+      "learning_rate": 0.00011228668941979522,
+      "loss": 6.584,
+      "step": 7803
+    },
+    {
+      "epoch": 2.663481228668942,
+      "grad_norm": 2.5685508251190186,
+      "learning_rate": 0.00011217292377701935,
+      "loss": 6.1668,
+      "step": 7804
+    },
+    {
+      "epoch": 2.6638225255972694,
+      "grad_norm": 2.811805486679077,
+      "learning_rate": 0.00011205915813424346,
+      "loss": 5.735,
+      "step": 7805
+    },
+    {
+      "epoch": 2.6641638225255972,
+      "grad_norm": 2.5972208976745605,
+      "learning_rate": 0.00011194539249146758,
+      "loss": 6.849,
+      "step": 7806
+    },
+    {
+      "epoch": 2.664505119453925,
+      "grad_norm": 2.659641742706299,
+      "learning_rate": 0.0001118316268486917,
+      "loss": 6.3136,
+      "step": 7807
+    },
+    {
+      "epoch": 2.6648464163822525,
+      "grad_norm": 2.7042298316955566,
+      "learning_rate": 0.00011171786120591581,
+      "loss": 6.0122,
+      "step": 7808
+    },
+    {
+      "epoch": 2.6651877133105804,
+      "grad_norm": 2.617111921310425,
+      "learning_rate": 0.00011160409556313995,
+      "loss": 6.1822,
+      "step": 7809
+    },
+    {
+      "epoch": 2.6655290102389078,
+      "grad_norm": 2.545041084289551,
+      "learning_rate": 0.00011149032992036405,
+      "loss": 6.1464,
+      "step": 7810
+    },
+    {
+      "epoch": 2.6658703071672356,
+      "grad_norm": 3.4677798748016357,
+      "learning_rate": 0.00011137656427758817,
+      "loss": 4.4972,
+      "step": 7811
+    },
+    {
+      "epoch": 2.666211604095563,
+      "grad_norm": 2.6456897258758545,
+      "learning_rate": 0.00011126279863481229,
+      "loss": 5.4503,
+      "step": 7812
+    },
+    {
+      "epoch": 2.666552901023891,
+      "grad_norm": 2.6556036472320557,
+      "learning_rate": 0.0001111490329920364,
+      "loss": 6.0158,
+      "step": 7813
+    },
+    {
+      "epoch": 2.6668941979522183,
+      "grad_norm": 2.6038191318511963,
+      "learning_rate": 0.00011103526734926054,
+      "loss": 6.0356,
+      "step": 7814
+    },
+    {
+      "epoch": 2.667235494880546,
+      "grad_norm": 2.673539400100708,
+      "learning_rate": 0.00011092150170648464,
+      "loss": 5.9474,
+      "step": 7815
+    },
+    {
+      "epoch": 2.667576791808874,
+      "grad_norm": 2.661170721054077,
+      "learning_rate": 0.00011080773606370876,
+      "loss": 6.2451,
+      "step": 7816
+    },
+    {
+      "epoch": 2.6679180887372014,
+      "grad_norm": 2.639185905456543,
+      "learning_rate": 0.00011069397042093288,
+      "loss": 5.3801,
+      "step": 7817
+    },
+    {
+      "epoch": 2.668259385665529,
+      "grad_norm": 2.672795057296753,
+      "learning_rate": 0.000110580204778157,
+      "loss": 6.2753,
+      "step": 7818
+    },
+    {
+      "epoch": 2.6686006825938566,
+      "grad_norm": 1.828800916671753,
+      "learning_rate": 0.00011046643913538112,
+      "loss": 2.4337,
+      "step": 7819
+    },
+    {
+      "epoch": 2.6689419795221845,
+      "grad_norm": 2.646155834197998,
+      "learning_rate": 0.00011035267349260524,
+      "loss": 5.8717,
+      "step": 7820
+    },
+    {
+      "epoch": 2.669283276450512,
+      "grad_norm": 2.6443183422088623,
+      "learning_rate": 0.00011023890784982935,
+      "loss": 6.0079,
+      "step": 7821
+    },
+    {
+      "epoch": 2.6696245733788397,
+      "grad_norm": 2.595067262649536,
+      "learning_rate": 0.00011012514220705347,
+      "loss": 5.6408,
+      "step": 7822
+    },
+    {
+      "epoch": 2.669965870307167,
+      "grad_norm": 2.655027151107788,
+      "learning_rate": 0.00011001137656427759,
+      "loss": 5.8955,
+      "step": 7823
+    },
+    {
+      "epoch": 2.670307167235495,
+      "grad_norm": 2.6794850826263428,
+      "learning_rate": 0.00010989761092150171,
+      "loss": 5.8458,
+      "step": 7824
+    },
+    {
+      "epoch": 2.6706484641638224,
+      "grad_norm": 2.556561231613159,
+      "learning_rate": 0.00010978384527872583,
+      "loss": 5.9302,
+      "step": 7825
+    },
+    {
+      "epoch": 2.6709897610921502,
+      "grad_norm": 2.651296854019165,
+      "learning_rate": 0.00010967007963594995,
+      "loss": 6.2365,
+      "step": 7826
+    },
+    {
+      "epoch": 2.6713310580204777,
+      "grad_norm": 2.6557466983795166,
+      "learning_rate": 0.00010955631399317406,
+      "loss": 5.845,
+      "step": 7827
+    },
+    {
+      "epoch": 2.6716723549488055,
+      "grad_norm": 2.8386025428771973,
+      "learning_rate": 0.00010944254835039818,
+      "loss": 5.3838,
+      "step": 7828
+    },
+    {
+      "epoch": 2.6720136518771334,
+      "grad_norm": 2.6636369228363037,
+      "learning_rate": 0.0001093287827076223,
+      "loss": 6.2458,
+      "step": 7829
+    },
+    {
+      "epoch": 2.6723549488054608,
+      "grad_norm": 2.6208786964416504,
+      "learning_rate": 0.00010921501706484642,
+      "loss": 6.4429,
+      "step": 7830
+    },
+    {
+      "epoch": 2.672696245733788,
+      "grad_norm": 2.61503005027771,
+      "learning_rate": 0.00010910125142207054,
+      "loss": 6.0356,
+      "step": 7831
+    },
+    {
+      "epoch": 2.673037542662116,
+      "grad_norm": 2.5759787559509277,
+      "learning_rate": 0.00010898748577929466,
+      "loss": 5.025,
+      "step": 7832
+    },
+    {
+      "epoch": 2.673378839590444,
+      "grad_norm": 2.680743932723999,
+      "learning_rate": 0.00010887372013651878,
+      "loss": 5.651,
+      "step": 7833
+    },
+    {
+      "epoch": 2.6737201365187713,
+      "grad_norm": 2.4130074977874756,
+      "learning_rate": 0.0001087599544937429,
+      "loss": 4.161,
+      "step": 7834
+    },
+    {
+      "epoch": 2.674061433447099,
+      "grad_norm": 2.479400634765625,
+      "learning_rate": 0.00010864618885096701,
+      "loss": 4.6418,
+      "step": 7835
+    },
+    {
+      "epoch": 2.6744027303754265,
+      "grad_norm": 2.6607160568237305,
+      "learning_rate": 0.00010853242320819112,
+      "loss": 5.682,
+      "step": 7836
+    },
+    {
+      "epoch": 2.6747440273037544,
+      "grad_norm": 2.6387665271759033,
+      "learning_rate": 0.00010841865756541525,
+      "loss": 6.2866,
+      "step": 7837
+    },
+    {
+      "epoch": 2.675085324232082,
+      "grad_norm": 2.6623637676239014,
+      "learning_rate": 0.00010830489192263937,
+      "loss": 5.5277,
+      "step": 7838
+    },
+    {
+      "epoch": 2.6754266211604096,
+      "grad_norm": 2.7224626541137695,
+      "learning_rate": 0.00010819112627986349,
+      "loss": 6.5055,
+      "step": 7839
+    },
+    {
+      "epoch": 2.675767918088737,
+      "grad_norm": 1.7368601560592651,
+      "learning_rate": 0.0001080773606370876,
+      "loss": 2.5353,
+      "step": 7840
+    },
+    {
+      "epoch": 2.676109215017065,
+      "grad_norm": 2.5209169387817383,
+      "learning_rate": 0.00010796359499431171,
+      "loss": 5.4228,
+      "step": 7841
+    },
+    {
+      "epoch": 2.6764505119453927,
+      "grad_norm": 2.6286282539367676,
+      "learning_rate": 0.00010784982935153584,
+      "loss": 6.5636,
+      "step": 7842
+    },
+    {
+      "epoch": 2.67679180887372,
+      "grad_norm": 2.581543207168579,
+      "learning_rate": 0.00010773606370875996,
+      "loss": 4.6108,
+      "step": 7843
+    },
+    {
+      "epoch": 2.6771331058020476,
+      "grad_norm": 2.5444607734680176,
+      "learning_rate": 0.00010762229806598408,
+      "loss": 6.3867,
+      "step": 7844
+    },
+    {
+      "epoch": 2.6774744027303754,
+      "grad_norm": 2.6205897331237793,
+      "learning_rate": 0.0001075085324232082,
+      "loss": 6.3425,
+      "step": 7845
+    },
+    {
+      "epoch": 2.6778156996587033,
+      "grad_norm": 2.6083576679229736,
+      "learning_rate": 0.0001073947667804323,
+      "loss": 5.435,
+      "step": 7846
+    },
+    {
+      "epoch": 2.6781569965870307,
+      "grad_norm": 2.6391236782073975,
+      "learning_rate": 0.00010728100113765643,
+      "loss": 5.6658,
+      "step": 7847
+    },
+    {
+      "epoch": 2.6784982935153585,
+      "grad_norm": 2.71447491645813,
+      "learning_rate": 0.00010716723549488055,
+      "loss": 5.6652,
+      "step": 7848
+    },
+    {
+      "epoch": 2.678839590443686,
+      "grad_norm": 2.63681697845459,
+      "learning_rate": 0.00010705346985210467,
+      "loss": 5.891,
+      "step": 7849
+    },
+    {
+      "epoch": 2.6791808873720138,
+      "grad_norm": 2.910919427871704,
+      "learning_rate": 0.00010693970420932879,
+      "loss": 5.8579,
+      "step": 7850
+    },
+    {
+      "epoch": 2.679522184300341,
+      "grad_norm": 2.6132237911224365,
+      "learning_rate": 0.0001068259385665529,
+      "loss": 5.8708,
+      "step": 7851
+    },
+    {
+      "epoch": 2.679863481228669,
+      "grad_norm": 2.6288161277770996,
+      "learning_rate": 0.00010671217292377703,
+      "loss": 5.7728,
+      "step": 7852
+    },
+    {
+      "epoch": 2.6802047781569964,
+      "grad_norm": 2.5175042152404785,
+      "learning_rate": 0.00010659840728100113,
+      "loss": 5.9085,
+      "step": 7853
+    },
+    {
+      "epoch": 2.6805460750853243,
+      "grad_norm": 2.57769775390625,
+      "learning_rate": 0.00010648464163822526,
+      "loss": 6.2867,
+      "step": 7854
+    },
+    {
+      "epoch": 2.680887372013652,
+      "grad_norm": 2.5458014011383057,
+      "learning_rate": 0.00010637087599544938,
+      "loss": 6.0821,
+      "step": 7855
+    },
+    {
+      "epoch": 2.6812286689419795,
+      "grad_norm": 2.641937732696533,
+      "learning_rate": 0.00010625711035267349,
+      "loss": 6.295,
+      "step": 7856
+    },
+    {
+      "epoch": 2.681569965870307,
+      "grad_norm": 2.6056764125823975,
+      "learning_rate": 0.00010614334470989762,
+      "loss": 5.8905,
+      "step": 7857
+    },
+    {
+      "epoch": 2.681911262798635,
+      "grad_norm": 2.6439177989959717,
+      "learning_rate": 0.00010602957906712172,
+      "loss": 6.2158,
+      "step": 7858
+    },
+    {
+      "epoch": 2.6822525597269626,
+      "grad_norm": 2.680771589279175,
+      "learning_rate": 0.00010591581342434586,
+      "loss": 5.609,
+      "step": 7859
+    },
+    {
+      "epoch": 2.68259385665529,
+      "grad_norm": 2.6440012454986572,
+      "learning_rate": 0.00010580204778156998,
+      "loss": 5.7455,
+      "step": 7860
+    },
+    {
+      "epoch": 2.682935153583618,
+      "grad_norm": 2.6077284812927246,
+      "learning_rate": 0.00010568828213879408,
+      "loss": 6.4567,
+      "step": 7861
+    },
+    {
+      "epoch": 2.6832764505119453,
+      "grad_norm": 2.5259697437286377,
+      "learning_rate": 0.00010557451649601821,
+      "loss": 5.9364,
+      "step": 7862
+    },
+    {
+      "epoch": 2.683617747440273,
+      "grad_norm": 2.647272825241089,
+      "learning_rate": 0.00010546075085324232,
+      "loss": 6.4563,
+      "step": 7863
+    },
+    {
+      "epoch": 2.6839590443686006,
+      "grad_norm": 2.600893020629883,
+      "learning_rate": 0.00010534698521046644,
+      "loss": 5.9343,
+      "step": 7864
+    },
+    {
+      "epoch": 2.6843003412969284,
+      "grad_norm": 2.8401756286621094,
+      "learning_rate": 0.00010523321956769057,
+      "loss": 4.4045,
+      "step": 7865
+    },
+    {
+      "epoch": 2.684641638225256,
+      "grad_norm": 2.6061348915100098,
+      "learning_rate": 0.00010511945392491467,
+      "loss": 4.6237,
+      "step": 7866
+    },
+    {
+      "epoch": 2.6849829351535837,
+      "grad_norm": 2.599360466003418,
+      "learning_rate": 0.0001050056882821388,
+      "loss": 6.0888,
+      "step": 7867
+    },
+    {
+      "epoch": 2.6853242320819115,
+      "grad_norm": 2.5830085277557373,
+      "learning_rate": 0.00010489192263936291,
+      "loss": 5.5227,
+      "step": 7868
+    },
+    {
+      "epoch": 2.685665529010239,
+      "grad_norm": 2.7643754482269287,
+      "learning_rate": 0.00010477815699658703,
+      "loss": 5.0325,
+      "step": 7869
+    },
+    {
+      "epoch": 2.6860068259385663,
+      "grad_norm": 2.723921537399292,
+      "learning_rate": 0.00010466439135381115,
+      "loss": 5.5498,
+      "step": 7870
+    },
+    {
+      "epoch": 2.686348122866894,
+      "grad_norm": 2.556098461151123,
+      "learning_rate": 0.00010455062571103526,
+      "loss": 5.6835,
+      "step": 7871
+    },
+    {
+      "epoch": 2.686689419795222,
+      "grad_norm": 2.629362106323242,
+      "learning_rate": 0.0001044368600682594,
+      "loss": 6.4194,
+      "step": 7872
+    },
+    {
+      "epoch": 2.6870307167235494,
+      "grad_norm": 2.6125411987304688,
+      "learning_rate": 0.0001043230944254835,
+      "loss": 6.0512,
+      "step": 7873
+    },
+    {
+      "epoch": 2.6873720136518773,
+      "grad_norm": 2.6440958976745605,
+      "learning_rate": 0.00010420932878270762,
+      "loss": 6.0304,
+      "step": 7874
+    },
+    {
+      "epoch": 2.6877133105802047,
+      "grad_norm": 2.598612070083618,
+      "learning_rate": 0.00010409556313993174,
+      "loss": 5.6832,
+      "step": 7875
+    },
+    {
+      "epoch": 2.6880546075085325,
+      "grad_norm": 2.6545569896698,
+      "learning_rate": 0.00010398179749715586,
+      "loss": 6.4576,
+      "step": 7876
+    },
+    {
+      "epoch": 2.68839590443686,
+      "grad_norm": 2.641326427459717,
+      "learning_rate": 0.00010386803185437999,
+      "loss": 6.3398,
+      "step": 7877
+    },
+    {
+      "epoch": 2.688737201365188,
+      "grad_norm": 2.6488940715789795,
+      "learning_rate": 0.0001037542662116041,
+      "loss": 5.7814,
+      "step": 7878
+    },
+    {
+      "epoch": 2.689078498293515,
+      "grad_norm": 2.5873377323150635,
+      "learning_rate": 0.00010364050056882821,
+      "loss": 6.4943,
+      "step": 7879
+    },
+    {
+      "epoch": 2.689419795221843,
+      "grad_norm": 2.5985260009765625,
+      "learning_rate": 0.00010352673492605233,
+      "loss": 6.0007,
+      "step": 7880
+    },
+    {
+      "epoch": 2.689761092150171,
+      "grad_norm": 2.6462574005126953,
+      "learning_rate": 0.00010341296928327645,
+      "loss": 6.4716,
+      "step": 7881
+    },
+    {
+      "epoch": 2.6901023890784983,
+      "grad_norm": 2.649360418319702,
+      "learning_rate": 0.00010329920364050058,
+      "loss": 5.7194,
+      "step": 7882
+    },
+    {
+      "epoch": 2.6904436860068257,
+      "grad_norm": 2.5704243183135986,
+      "learning_rate": 0.00010318543799772469,
+      "loss": 5.9094,
+      "step": 7883
+    },
+    {
+      "epoch": 2.6907849829351536,
+      "grad_norm": 2.7193174362182617,
+      "learning_rate": 0.0001030716723549488,
+      "loss": 6.4267,
+      "step": 7884
+    },
+    {
+      "epoch": 2.6911262798634814,
+      "grad_norm": 2.567355155944824,
+      "learning_rate": 0.00010295790671217292,
+      "loss": 5.5288,
+      "step": 7885
+    },
+    {
+      "epoch": 2.691467576791809,
+      "grad_norm": 2.8648149967193604,
+      "learning_rate": 0.00010284414106939704,
+      "loss": 5.4581,
+      "step": 7886
+    },
+    {
+      "epoch": 2.6918088737201367,
+      "grad_norm": 2.6088626384735107,
+      "learning_rate": 0.00010273037542662116,
+      "loss": 6.3214,
+      "step": 7887
+    },
+    {
+      "epoch": 2.692150170648464,
+      "grad_norm": 2.7508292198181152,
+      "learning_rate": 0.00010261660978384528,
+      "loss": 6.3367,
+      "step": 7888
+    },
+    {
+      "epoch": 2.692491467576792,
+      "grad_norm": 2.5980231761932373,
+      "learning_rate": 0.0001025028441410694,
+      "loss": 5.7762,
+      "step": 7889
+    },
+    {
+      "epoch": 2.6928327645051193,
+      "grad_norm": 2.668694496154785,
+      "learning_rate": 0.00010238907849829352,
+      "loss": 5.5341,
+      "step": 7890
+    },
+    {
+      "epoch": 2.693174061433447,
+      "grad_norm": 2.5616753101348877,
+      "learning_rate": 0.00010227531285551763,
+      "loss": 6.0325,
+      "step": 7891
+    },
+    {
+      "epoch": 2.6935153583617746,
+      "grad_norm": 5.180400371551514,
+      "learning_rate": 0.00010216154721274175,
+      "loss": 4.5402,
+      "step": 7892
+    },
+    {
+      "epoch": 2.6938566552901024,
+      "grad_norm": 2.6964242458343506,
+      "learning_rate": 0.00010204778156996587,
+      "loss": 6.8569,
+      "step": 7893
+    },
+    {
+      "epoch": 2.6941979522184303,
+      "grad_norm": 2.5725672245025635,
+      "learning_rate": 0.00010193401592718999,
+      "loss": 5.7403,
+      "step": 7894
+    },
+    {
+      "epoch": 2.6945392491467577,
+      "grad_norm": 2.649010181427002,
+      "learning_rate": 0.00010182025028441411,
+      "loss": 6.1924,
+      "step": 7895
+    },
+    {
+      "epoch": 2.694880546075085,
+      "grad_norm": 2.6579694747924805,
+      "learning_rate": 0.00010170648464163823,
+      "loss": 5.7039,
+      "step": 7896
+    },
+    {
+      "epoch": 2.695221843003413,
+      "grad_norm": 2.6075141429901123,
+      "learning_rate": 0.00010159271899886235,
+      "loss": 6.356,
+      "step": 7897
+    },
+    {
+      "epoch": 2.695563139931741,
+      "grad_norm": 2.6528830528259277,
+      "learning_rate": 0.00010147895335608646,
+      "loss": 5.4291,
+      "step": 7898
+    },
+    {
+      "epoch": 2.695904436860068,
+      "grad_norm": 2.625629425048828,
+      "learning_rate": 0.00010136518771331058,
+      "loss": 6.2645,
+      "step": 7899
+    },
+    {
+      "epoch": 2.696245733788396,
+      "grad_norm": 2.710127830505371,
+      "learning_rate": 0.0001012514220705347,
+      "loss": 4.9046,
+      "step": 7900
+    },
+    {
+      "epoch": 2.6965870307167235,
+      "grad_norm": 2.4944448471069336,
+      "learning_rate": 0.00010113765642775882,
+      "loss": 5.518,
+      "step": 7901
+    },
+    {
+      "epoch": 2.6969283276450513,
+      "grad_norm": 2.6292831897735596,
+      "learning_rate": 0.00010102389078498294,
+      "loss": 6.1254,
+      "step": 7902
+    },
+    {
+      "epoch": 2.6972696245733787,
+      "grad_norm": 2.5295937061309814,
+      "learning_rate": 0.00010091012514220706,
+      "loss": 6.5135,
+      "step": 7903
+    },
+    {
+      "epoch": 2.6976109215017066,
+      "grad_norm": 2.602799415588379,
+      "learning_rate": 0.00010079635949943116,
+      "loss": 5.8567,
+      "step": 7904
+    },
+    {
+      "epoch": 2.697952218430034,
+      "grad_norm": 2.6777186393737793,
+      "learning_rate": 0.0001006825938566553,
+      "loss": 6.0356,
+      "step": 7905
+    },
+    {
+      "epoch": 2.698293515358362,
+      "grad_norm": 2.5789544582366943,
+      "learning_rate": 0.00010056882821387941,
+      "loss": 5.3017,
+      "step": 7906
+    },
+    {
+      "epoch": 2.6986348122866897,
+      "grad_norm": 2.565323829650879,
+      "learning_rate": 0.00010045506257110353,
+      "loss": 6.0801,
+      "step": 7907
+    },
+    {
+      "epoch": 2.698976109215017,
+      "grad_norm": 2.089174270629883,
+      "learning_rate": 0.00010034129692832765,
+      "loss": 3.2135,
+      "step": 7908
+    },
+    {
+      "epoch": 2.6993174061433445,
+      "grad_norm": 2.6627211570739746,
+      "learning_rate": 0.00010022753128555175,
+      "loss": 5.925,
+      "step": 7909
+    },
+    {
+      "epoch": 2.6996587030716723,
+      "grad_norm": 2.5624616146087646,
+      "learning_rate": 0.00010011376564277589,
+      "loss": 5.6055,
+      "step": 7910
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.5882010459899902,
+      "learning_rate": 0.0001,
+      "loss": 4.6856,
+      "step": 7911
+    },
+    {
+      "epoch": 2.7003412969283276,
+      "grad_norm": 2.569596290588379,
+      "learning_rate": 9.988623435722412e-05,
+      "loss": 5.86,
+      "step": 7912
+    },
+    {
+      "epoch": 2.7006825938566554,
+      "grad_norm": 2.565866708755493,
+      "learning_rate": 9.977246871444824e-05,
+      "loss": 6.1467,
+      "step": 7913
+    },
+    {
+      "epoch": 2.701023890784983,
+      "grad_norm": 2.524012327194214,
+      "learning_rate": 9.965870307167235e-05,
+      "loss": 5.9074,
+      "step": 7914
+    },
+    {
+      "epoch": 2.7013651877133107,
+      "grad_norm": 1.8367999792099,
+      "learning_rate": 9.954493742889648e-05,
+      "loss": 3.4121,
+      "step": 7915
+    },
+    {
+      "epoch": 2.701706484641638,
+      "grad_norm": 2.5629944801330566,
+      "learning_rate": 9.94311717861206e-05,
+      "loss": 6.5814,
+      "step": 7916
+    },
+    {
+      "epoch": 2.702047781569966,
+      "grad_norm": 2.7166037559509277,
+      "learning_rate": 9.931740614334472e-05,
+      "loss": 6.285,
+      "step": 7917
+    },
+    {
+      "epoch": 2.7023890784982934,
+      "grad_norm": 2.672518730163574,
+      "learning_rate": 9.920364050056883e-05,
+      "loss": 3.7134,
+      "step": 7918
+    },
+    {
+      "epoch": 2.702730375426621,
+      "grad_norm": 2.4576282501220703,
+      "learning_rate": 9.908987485779294e-05,
+      "loss": 4.0857,
+      "step": 7919
+    },
+    {
+      "epoch": 2.703071672354949,
+      "grad_norm": 2.5885422229766846,
+      "learning_rate": 9.897610921501707e-05,
+      "loss": 5.9866,
+      "step": 7920
+    },
+    {
+      "epoch": 2.7034129692832765,
+      "grad_norm": 2.686434030532837,
+      "learning_rate": 9.886234357224118e-05,
+      "loss": 5.7955,
+      "step": 7921
+    },
+    {
+      "epoch": 2.703754266211604,
+      "grad_norm": 2.6147704124450684,
+      "learning_rate": 9.874857792946531e-05,
+      "loss": 6.1341,
+      "step": 7922
+    },
+    {
+      "epoch": 2.7040955631399317,
+      "grad_norm": 3.224640369415283,
+      "learning_rate": 9.863481228668943e-05,
+      "loss": 4.9066,
+      "step": 7923
+    },
+    {
+      "epoch": 2.7044368600682596,
+      "grad_norm": 2.6708364486694336,
+      "learning_rate": 9.852104664391353e-05,
+      "loss": 5.7106,
+      "step": 7924
+    },
+    {
+      "epoch": 2.704778156996587,
+      "grad_norm": 2.5874176025390625,
+      "learning_rate": 9.840728100113766e-05,
+      "loss": 4.865,
+      "step": 7925
+    },
+    {
+      "epoch": 2.705119453924915,
+      "grad_norm": 2.6783151626586914,
+      "learning_rate": 9.829351535836177e-05,
+      "loss": 4.9307,
+      "step": 7926
+    },
+    {
+      "epoch": 2.7054607508532422,
+      "grad_norm": 2.5469775199890137,
+      "learning_rate": 9.81797497155859e-05,
+      "loss": 5.7146,
+      "step": 7927
+    },
+    {
+      "epoch": 2.70580204778157,
+      "grad_norm": 2.5792226791381836,
+      "learning_rate": 9.806598407281002e-05,
+      "loss": 5.7424,
+      "step": 7928
+    },
+    {
+      "epoch": 2.7061433447098975,
+      "grad_norm": 2.562011241912842,
+      "learning_rate": 9.795221843003412e-05,
+      "loss": 5.7029,
+      "step": 7929
+    },
+    {
+      "epoch": 2.7064846416382253,
+      "grad_norm": 3.7608301639556885,
+      "learning_rate": 9.783845278725826e-05,
+      "loss": 4.2989,
+      "step": 7930
+    },
+    {
+      "epoch": 2.7068259385665527,
+      "grad_norm": 2.5663902759552,
+      "learning_rate": 9.772468714448236e-05,
+      "loss": 6.4621,
+      "step": 7931
+    },
+    {
+      "epoch": 2.7071672354948806,
+      "grad_norm": 2.6280720233917236,
+      "learning_rate": 9.761092150170649e-05,
+      "loss": 6.3645,
+      "step": 7932
+    },
+    {
+      "epoch": 2.7075085324232084,
+      "grad_norm": 2.5684332847595215,
+      "learning_rate": 9.749715585893061e-05,
+      "loss": 5.8003,
+      "step": 7933
+    },
+    {
+      "epoch": 2.707849829351536,
+      "grad_norm": 2.643038034439087,
+      "learning_rate": 9.738339021615472e-05,
+      "loss": 6.4916,
+      "step": 7934
+    },
+    {
+      "epoch": 2.7081911262798632,
+      "grad_norm": 2.559818983078003,
+      "learning_rate": 9.726962457337885e-05,
+      "loss": 6.41,
+      "step": 7935
+    },
+    {
+      "epoch": 2.708532423208191,
+      "grad_norm": 2.636819839477539,
+      "learning_rate": 9.715585893060295e-05,
+      "loss": 5.2788,
+      "step": 7936
+    },
+    {
+      "epoch": 2.708873720136519,
+      "grad_norm": 2.651435613632202,
+      "learning_rate": 9.704209328782709e-05,
+      "loss": 5.8778,
+      "step": 7937
+    },
+    {
+      "epoch": 2.7092150170648464,
+      "grad_norm": 2.6740775108337402,
+      "learning_rate": 9.692832764505119e-05,
+      "loss": 4.9705,
+      "step": 7938
+    },
+    {
+      "epoch": 2.709556313993174,
+      "grad_norm": 2.6301076412200928,
+      "learning_rate": 9.681456200227531e-05,
+      "loss": 6.1048,
+      "step": 7939
+    },
+    {
+      "epoch": 2.7098976109215016,
+      "grad_norm": 2.613173007965088,
+      "learning_rate": 9.670079635949944e-05,
+      "loss": 5.3901,
+      "step": 7940
+    },
+    {
+      "epoch": 2.7102389078498295,
+      "grad_norm": 2.5331554412841797,
+      "learning_rate": 9.658703071672355e-05,
+      "loss": 5.4303,
+      "step": 7941
+    },
+    {
+      "epoch": 2.710580204778157,
+      "grad_norm": 2.602365016937256,
+      "learning_rate": 9.647326507394768e-05,
+      "loss": 6.2944,
+      "step": 7942
+    },
+    {
+      "epoch": 2.7109215017064847,
+      "grad_norm": 2.6677892208099365,
+      "learning_rate": 9.635949943117178e-05,
+      "loss": 5.7442,
+      "step": 7943
+    },
+    {
+      "epoch": 2.711262798634812,
+      "grad_norm": 2.62030291557312,
+      "learning_rate": 9.62457337883959e-05,
+      "loss": 6.2834,
+      "step": 7944
+    },
+    {
+      "epoch": 2.71160409556314,
+      "grad_norm": 2.5873615741729736,
+      "learning_rate": 9.613196814562003e-05,
+      "loss": 5.7214,
+      "step": 7945
+    },
+    {
+      "epoch": 2.711945392491468,
+      "grad_norm": 2.6523163318634033,
+      "learning_rate": 9.601820250284414e-05,
+      "loss": 5.5457,
+      "step": 7946
+    },
+    {
+      "epoch": 2.7122866894197952,
+      "grad_norm": 2.6249499320983887,
+      "learning_rate": 9.590443686006827e-05,
+      "loss": 5.5907,
+      "step": 7947
+    },
+    {
+      "epoch": 2.7126279863481226,
+      "grad_norm": 2.6475677490234375,
+      "learning_rate": 9.579067121729238e-05,
+      "loss": 6.2695,
+      "step": 7948
+    },
+    {
+      "epoch": 2.7129692832764505,
+      "grad_norm": 2.638388156890869,
+      "learning_rate": 9.56769055745165e-05,
+      "loss": 6.0452,
+      "step": 7949
+    },
+    {
+      "epoch": 2.7133105802047783,
+      "grad_norm": 2.5917885303497314,
+      "learning_rate": 9.556313993174063e-05,
+      "loss": 5.4567,
+      "step": 7950
+    },
+    {
+      "epoch": 2.7136518771331057,
+      "grad_norm": 2.65228533744812,
+      "learning_rate": 9.544937428896473e-05,
+      "loss": 6.3895,
+      "step": 7951
+    },
+    {
+      "epoch": 2.7139931740614336,
+      "grad_norm": 2.614609956741333,
+      "learning_rate": 9.533560864618886e-05,
+      "loss": 6.4066,
+      "step": 7952
+    },
+    {
+      "epoch": 2.714334470989761,
+      "grad_norm": 2.717390298843384,
+      "learning_rate": 9.522184300341297e-05,
+      "loss": 4.8762,
+      "step": 7953
+    },
+    {
+      "epoch": 2.714675767918089,
+      "grad_norm": 2.660956859588623,
+      "learning_rate": 9.510807736063709e-05,
+      "loss": 5.9635,
+      "step": 7954
+    },
+    {
+      "epoch": 2.7150170648464163,
+      "grad_norm": 2.5839521884918213,
+      "learning_rate": 9.49943117178612e-05,
+      "loss": 5.8931,
+      "step": 7955
+    },
+    {
+      "epoch": 2.715358361774744,
+      "grad_norm": 2.445621967315674,
+      "learning_rate": 9.488054607508532e-05,
+      "loss": 3.551,
+      "step": 7956
+    },
+    {
+      "epoch": 2.7156996587030715,
+      "grad_norm": 2.659536838531494,
+      "learning_rate": 9.476678043230946e-05,
+      "loss": 6.4926,
+      "step": 7957
+    },
+    {
+      "epoch": 2.7160409556313994,
+      "grad_norm": 2.6747255325317383,
+      "learning_rate": 9.465301478953356e-05,
+      "loss": 6.5505,
+      "step": 7958
+    },
+    {
+      "epoch": 2.716382252559727,
+      "grad_norm": 2.663691520690918,
+      "learning_rate": 9.453924914675768e-05,
+      "loss": 6.1892,
+      "step": 7959
+    },
+    {
+      "epoch": 2.7167235494880546,
+      "grad_norm": 2.6698076725006104,
+      "learning_rate": 9.44254835039818e-05,
+      "loss": 5.9412,
+      "step": 7960
+    },
+    {
+      "epoch": 2.717064846416382,
+      "grad_norm": 2.622624635696411,
+      "learning_rate": 9.431171786120592e-05,
+      "loss": 5.906,
+      "step": 7961
+    },
+    {
+      "epoch": 2.71740614334471,
+      "grad_norm": 2.6831278800964355,
+      "learning_rate": 9.419795221843005e-05,
+      "loss": 5.8821,
+      "step": 7962
+    },
+    {
+      "epoch": 2.7177474402730377,
+      "grad_norm": 2.650803565979004,
+      "learning_rate": 9.408418657565415e-05,
+      "loss": 5.7527,
+      "step": 7963
+    },
+    {
+      "epoch": 2.718088737201365,
+      "grad_norm": 2.567662477493286,
+      "learning_rate": 9.397042093287827e-05,
+      "loss": 5.8555,
+      "step": 7964
+    },
+    {
+      "epoch": 2.718430034129693,
+      "grad_norm": 2.6113736629486084,
+      "learning_rate": 9.385665529010239e-05,
+      "loss": 5.5252,
+      "step": 7965
+    },
+    {
+      "epoch": 2.7187713310580204,
+      "grad_norm": 2.6406290531158447,
+      "learning_rate": 9.374288964732651e-05,
+      "loss": 6.5528,
+      "step": 7966
+    },
+    {
+      "epoch": 2.7191126279863482,
+      "grad_norm": 2.571793556213379,
+      "learning_rate": 9.362912400455063e-05,
+      "loss": 5.9012,
+      "step": 7967
+    },
+    {
+      "epoch": 2.7194539249146756,
+      "grad_norm": 2.543337345123291,
+      "learning_rate": 9.351535836177475e-05,
+      "loss": 5.6792,
+      "step": 7968
+    },
+    {
+      "epoch": 2.7197952218430035,
+      "grad_norm": 2.667567014694214,
+      "learning_rate": 9.340159271899886e-05,
+      "loss": 6.1126,
+      "step": 7969
+    },
+    {
+      "epoch": 2.720136518771331,
+      "grad_norm": 2.7619335651397705,
+      "learning_rate": 9.328782707622298e-05,
+      "loss": 4.8842,
+      "step": 7970
+    },
+    {
+      "epoch": 2.7204778156996587,
+      "grad_norm": 2.702486753463745,
+      "learning_rate": 9.31740614334471e-05,
+      "loss": 5.1622,
+      "step": 7971
+    },
+    {
+      "epoch": 2.7208191126279866,
+      "grad_norm": 2.543321132659912,
+      "learning_rate": 9.306029579067122e-05,
+      "loss": 6.5731,
+      "step": 7972
+    },
+    {
+      "epoch": 2.721160409556314,
+      "grad_norm": 2.6276233196258545,
+      "learning_rate": 9.294653014789534e-05,
+      "loss": 6.0075,
+      "step": 7973
+    },
+    {
+      "epoch": 2.7215017064846414,
+      "grad_norm": 2.569683074951172,
+      "learning_rate": 9.283276450511946e-05,
+      "loss": 5.8531,
+      "step": 7974
+    },
+    {
+      "epoch": 2.7218430034129693,
+      "grad_norm": 2.636974334716797,
+      "learning_rate": 9.271899886234357e-05,
+      "loss": 5.6186,
+      "step": 7975
+    },
+    {
+      "epoch": 2.722184300341297,
+      "grad_norm": 2.7969841957092285,
+      "learning_rate": 9.26052332195677e-05,
+      "loss": 5.0165,
+      "step": 7976
+    },
+    {
+      "epoch": 2.7225255972696245,
+      "grad_norm": 2.6524500846862793,
+      "learning_rate": 9.24914675767918e-05,
+      "loss": 5.6564,
+      "step": 7977
+    },
+    {
+      "epoch": 2.7228668941979524,
+      "grad_norm": 2.553088903427124,
+      "learning_rate": 9.237770193401593e-05,
+      "loss": 6.0369,
+      "step": 7978
+    },
+    {
+      "epoch": 2.7232081911262798,
+      "grad_norm": 2.4714152812957764,
+      "learning_rate": 9.226393629124005e-05,
+      "loss": 5.8547,
+      "step": 7979
+    },
+    {
+      "epoch": 2.7235494880546076,
+      "grad_norm": 2.634051561355591,
+      "learning_rate": 9.215017064846417e-05,
+      "loss": 6.1899,
+      "step": 7980
+    },
+    {
+      "epoch": 2.723890784982935,
+      "grad_norm": 2.6114561557769775,
+      "learning_rate": 9.203640500568829e-05,
+      "loss": 6.6533,
+      "step": 7981
+    },
+    {
+      "epoch": 2.724232081911263,
+      "grad_norm": 2.653291702270508,
+      "learning_rate": 9.192263936291239e-05,
+      "loss": 6.1926,
+      "step": 7982
+    },
+    {
+      "epoch": 2.7245733788395903,
+      "grad_norm": 2.614102363586426,
+      "learning_rate": 9.180887372013652e-05,
+      "loss": 6.1669,
+      "step": 7983
+    },
+    {
+      "epoch": 2.724914675767918,
+      "grad_norm": 2.6618130207061768,
+      "learning_rate": 9.169510807736064e-05,
+      "loss": 4.0645,
+      "step": 7984
+    },
+    {
+      "epoch": 2.725255972696246,
+      "grad_norm": 2.57953143119812,
+      "learning_rate": 9.158134243458476e-05,
+      "loss": 6.201,
+      "step": 7985
+    },
+    {
+      "epoch": 2.7255972696245734,
+      "grad_norm": 2.664710760116577,
+      "learning_rate": 9.146757679180888e-05,
+      "loss": 6.319,
+      "step": 7986
+    },
+    {
+      "epoch": 2.725938566552901,
+      "grad_norm": 2.681464672088623,
+      "learning_rate": 9.135381114903298e-05,
+      "loss": 5.8685,
+      "step": 7987
+    },
+    {
+      "epoch": 2.7262798634812286,
+      "grad_norm": 2.5687036514282227,
+      "learning_rate": 9.124004550625712e-05,
+      "loss": 6.3869,
+      "step": 7988
+    },
+    {
+      "epoch": 2.7266211604095565,
+      "grad_norm": 2.604609489440918,
+      "learning_rate": 9.112627986348123e-05,
+      "loss": 5.8885,
+      "step": 7989
+    },
+    {
+      "epoch": 2.726962457337884,
+      "grad_norm": 2.5861759185791016,
+      "learning_rate": 9.101251422070535e-05,
+      "loss": 5.1008,
+      "step": 7990
+    },
+    {
+      "epoch": 2.7273037542662117,
+      "grad_norm": 2.619473934173584,
+      "learning_rate": 9.089874857792947e-05,
+      "loss": 6.0467,
+      "step": 7991
+    },
+    {
+      "epoch": 2.727645051194539,
+      "grad_norm": 2.6422886848449707,
+      "learning_rate": 9.078498293515358e-05,
+      "loss": 6.595,
+      "step": 7992
+    },
+    {
+      "epoch": 2.727986348122867,
+      "grad_norm": 2.61552357673645,
+      "learning_rate": 9.067121729237771e-05,
+      "loss": 5.6713,
+      "step": 7993
+    },
+    {
+      "epoch": 2.7283276450511944,
+      "grad_norm": 2.6517817974090576,
+      "learning_rate": 9.055745164960181e-05,
+      "loss": 6.3159,
+      "step": 7994
+    },
+    {
+      "epoch": 2.7286689419795223,
+      "grad_norm": 2.8126118183135986,
+      "learning_rate": 9.044368600682594e-05,
+      "loss": 4.8574,
+      "step": 7995
+    },
+    {
+      "epoch": 2.7290102389078497,
+      "grad_norm": 2.5647075176239014,
+      "learning_rate": 9.032992036405006e-05,
+      "loss": 5.692,
+      "step": 7996
+    },
+    {
+      "epoch": 2.7293515358361775,
+      "grad_norm": 2.641845941543579,
+      "learning_rate": 9.021615472127417e-05,
+      "loss": 6.3246,
+      "step": 7997
+    },
+    {
+      "epoch": 2.7296928327645054,
+      "grad_norm": 2.609525203704834,
+      "learning_rate": 9.01023890784983e-05,
+      "loss": 6.4175,
+      "step": 7998
+    },
+    {
+      "epoch": 2.7300341296928328,
+      "grad_norm": 2.601841688156128,
+      "learning_rate": 8.99886234357224e-05,
+      "loss": 6.5798,
+      "step": 7999
+    },
+    {
+      "epoch": 2.73037542662116,
+      "grad_norm": 2.6355714797973633,
+      "learning_rate": 8.987485779294654e-05,
+      "loss": 6.3919,
+      "step": 8000
+    },
+    {
+      "epoch": 2.730716723549488,
+      "grad_norm": 2.583918333053589,
+      "learning_rate": 8.976109215017066e-05,
+      "loss": 6.3685,
+      "step": 8001
+    },
+    {
+      "epoch": 2.731058020477816,
+      "grad_norm": 2.569997549057007,
+      "learning_rate": 8.964732650739476e-05,
+      "loss": 6.1898,
+      "step": 8002
+    },
+    {
+      "epoch": 2.7313993174061433,
+      "grad_norm": 2.597790002822876,
+      "learning_rate": 8.953356086461889e-05,
+      "loss": 6.2867,
+      "step": 8003
+    },
+    {
+      "epoch": 2.731740614334471,
+      "grad_norm": 2.638692855834961,
+      "learning_rate": 8.9419795221843e-05,
+      "loss": 5.6171,
+      "step": 8004
+    },
+    {
+      "epoch": 2.7320819112627985,
+      "grad_norm": 2.7340500354766846,
+      "learning_rate": 8.930602957906713e-05,
+      "loss": 4.4383,
+      "step": 8005
+    },
+    {
+      "epoch": 2.7324232081911264,
+      "grad_norm": 2.652092933654785,
+      "learning_rate": 8.919226393629125e-05,
+      "loss": 6.168,
+      "step": 8006
+    },
+    {
+      "epoch": 2.732764505119454,
+      "grad_norm": 2.593048572540283,
+      "learning_rate": 8.907849829351535e-05,
+      "loss": 6.4116,
+      "step": 8007
+    },
+    {
+      "epoch": 2.7331058020477816,
+      "grad_norm": 2.6806387901306152,
+      "learning_rate": 8.896473265073949e-05,
+      "loss": 5.6523,
+      "step": 8008
+    },
+    {
+      "epoch": 2.733447098976109,
+      "grad_norm": 2.5880985260009766,
+      "learning_rate": 8.885096700796359e-05,
+      "loss": 6.137,
+      "step": 8009
+    },
+    {
+      "epoch": 2.733788395904437,
+      "grad_norm": 2.6142632961273193,
+      "learning_rate": 8.873720136518772e-05,
+      "loss": 6.0497,
+      "step": 8010
+    },
+    {
+      "epoch": 2.7341296928327647,
+      "grad_norm": 2.7263357639312744,
+      "learning_rate": 8.862343572241183e-05,
+      "loss": 6.2217,
+      "step": 8011
+    },
+    {
+      "epoch": 2.734470989761092,
+      "grad_norm": 2.6077632904052734,
+      "learning_rate": 8.850967007963595e-05,
+      "loss": 5.723,
+      "step": 8012
+    },
+    {
+      "epoch": 2.7348122866894196,
+      "grad_norm": 2.4901204109191895,
+      "learning_rate": 8.839590443686008e-05,
+      "loss": 5.1386,
+      "step": 8013
+    },
+    {
+      "epoch": 2.7351535836177474,
+      "grad_norm": 2.620048999786377,
+      "learning_rate": 8.828213879408418e-05,
+      "loss": 6.5681,
+      "step": 8014
+    },
+    {
+      "epoch": 2.7354948805460753,
+      "grad_norm": 2.5127553939819336,
+      "learning_rate": 8.816837315130831e-05,
+      "loss": 3.7144,
+      "step": 8015
+    },
+    {
+      "epoch": 2.7358361774744027,
+      "grad_norm": 2.677717447280884,
+      "learning_rate": 8.805460750853242e-05,
+      "loss": 5.7862,
+      "step": 8016
+    },
+    {
+      "epoch": 2.7361774744027305,
+      "grad_norm": 2.5813772678375244,
+      "learning_rate": 8.794084186575654e-05,
+      "loss": 6.38,
+      "step": 8017
+    },
+    {
+      "epoch": 2.736518771331058,
+      "grad_norm": 2.593780279159546,
+      "learning_rate": 8.782707622298067e-05,
+      "loss": 6.2665,
+      "step": 8018
+    },
+    {
+      "epoch": 2.7368600682593858,
+      "grad_norm": 2.6182265281677246,
+      "learning_rate": 8.771331058020478e-05,
+      "loss": 5.6717,
+      "step": 8019
+    },
+    {
+      "epoch": 2.737201365187713,
+      "grad_norm": 2.615816593170166,
+      "learning_rate": 8.759954493742891e-05,
+      "loss": 6.4123,
+      "step": 8020
+    },
+    {
+      "epoch": 2.737542662116041,
+      "grad_norm": 2.6214540004730225,
+      "learning_rate": 8.748577929465301e-05,
+      "loss": 6.0376,
+      "step": 8021
+    },
+    {
+      "epoch": 2.7378839590443684,
+      "grad_norm": 2.631145715713501,
+      "learning_rate": 8.737201365187713e-05,
+      "loss": 5.6407,
+      "step": 8022
+    },
+    {
+      "epoch": 2.7382252559726963,
+      "grad_norm": 2.52872371673584,
+      "learning_rate": 8.725824800910126e-05,
+      "loss": 6.0498,
+      "step": 8023
+    },
+    {
+      "epoch": 2.738566552901024,
+      "grad_norm": 2.6615374088287354,
+      "learning_rate": 8.714448236632537e-05,
+      "loss": 5.4418,
+      "step": 8024
+    },
+    {
+      "epoch": 2.7389078498293515,
+      "grad_norm": 4.1157917976379395,
+      "learning_rate": 8.70307167235495e-05,
+      "loss": 4.1903,
+      "step": 8025
+    },
+    {
+      "epoch": 2.739249146757679,
+      "grad_norm": 2.548863410949707,
+      "learning_rate": 8.69169510807736e-05,
+      "loss": 6.1133,
+      "step": 8026
+    },
+    {
+      "epoch": 2.739590443686007,
+      "grad_norm": 2.564734935760498,
+      "learning_rate": 8.680318543799772e-05,
+      "loss": 6.0158,
+      "step": 8027
+    },
+    {
+      "epoch": 2.7399317406143346,
+      "grad_norm": 2.6325366497039795,
+      "learning_rate": 8.668941979522184e-05,
+      "loss": 6.132,
+      "step": 8028
+    },
+    {
+      "epoch": 2.740273037542662,
+      "grad_norm": 2.528510808944702,
+      "learning_rate": 8.657565415244596e-05,
+      "loss": 6.0174,
+      "step": 8029
+    },
+    {
+      "epoch": 2.74061433447099,
+      "grad_norm": 2.521850347518921,
+      "learning_rate": 8.646188850967009e-05,
+      "loss": 5.7658,
+      "step": 8030
+    },
+    {
+      "epoch": 2.7409556313993173,
+      "grad_norm": 2.5946974754333496,
+      "learning_rate": 8.63481228668942e-05,
+      "loss": 6.4626,
+      "step": 8031
+    },
+    {
+      "epoch": 2.741296928327645,
+      "grad_norm": 2.647458076477051,
+      "learning_rate": 8.623435722411832e-05,
+      "loss": 6.2879,
+      "step": 8032
+    },
+    {
+      "epoch": 2.7416382252559726,
+      "grad_norm": 2.7117865085601807,
+      "learning_rate": 8.612059158134243e-05,
+      "loss": 5.5557,
+      "step": 8033
+    },
+    {
+      "epoch": 2.7419795221843004,
+      "grad_norm": 2.601431131362915,
+      "learning_rate": 8.600682593856655e-05,
+      "loss": 5.8308,
+      "step": 8034
+    },
+    {
+      "epoch": 2.742320819112628,
+      "grad_norm": 2.799042224884033,
+      "learning_rate": 8.589306029579068e-05,
+      "loss": 5.5987,
+      "step": 8035
+    },
+    {
+      "epoch": 2.7426621160409557,
+      "grad_norm": 2.671846389770508,
+      "learning_rate": 8.577929465301479e-05,
+      "loss": 6.4048,
+      "step": 8036
+    },
+    {
+      "epoch": 2.7430034129692835,
+      "grad_norm": 2.5593509674072266,
+      "learning_rate": 8.566552901023891e-05,
+      "loss": 5.5486,
+      "step": 8037
+    },
+    {
+      "epoch": 2.743344709897611,
+      "grad_norm": 2.5567145347595215,
+      "learning_rate": 8.555176336746303e-05,
+      "loss": 6.0707,
+      "step": 8038
+    },
+    {
+      "epoch": 2.7436860068259383,
+      "grad_norm": 2.59529185295105,
+      "learning_rate": 8.543799772468715e-05,
+      "loss": 5.8477,
+      "step": 8039
+    },
+    {
+      "epoch": 2.744027303754266,
+      "grad_norm": 2.6873974800109863,
+      "learning_rate": 8.532423208191128e-05,
+      "loss": 6.008,
+      "step": 8040
+    },
+    {
+      "epoch": 2.744368600682594,
+      "grad_norm": 2.6299967765808105,
+      "learning_rate": 8.521046643913538e-05,
+      "loss": 5.65,
+      "step": 8041
+    },
+    {
+      "epoch": 2.7447098976109214,
+      "grad_norm": 2.6703577041625977,
+      "learning_rate": 8.50967007963595e-05,
+      "loss": 6.3595,
+      "step": 8042
+    },
+    {
+      "epoch": 2.7450511945392493,
+      "grad_norm": 2.5916976928710938,
+      "learning_rate": 8.498293515358362e-05,
+      "loss": 6.1119,
+      "step": 8043
+    },
+    {
+      "epoch": 2.7453924914675767,
+      "grad_norm": 2.577881097793579,
+      "learning_rate": 8.486916951080774e-05,
+      "loss": 5.5663,
+      "step": 8044
+    },
+    {
+      "epoch": 2.7457337883959045,
+      "grad_norm": 2.5749382972717285,
+      "learning_rate": 8.475540386803186e-05,
+      "loss": 5.7458,
+      "step": 8045
+    },
+    {
+      "epoch": 2.746075085324232,
+      "grad_norm": 2.5341036319732666,
+      "learning_rate": 8.464163822525597e-05,
+      "loss": 5.6241,
+      "step": 8046
+    },
+    {
+      "epoch": 2.74641638225256,
+      "grad_norm": 2.6867024898529053,
+      "learning_rate": 8.452787258248009e-05,
+      "loss": 6.3056,
+      "step": 8047
+    },
+    {
+      "epoch": 2.746757679180887,
+      "grad_norm": 2.5458154678344727,
+      "learning_rate": 8.441410693970421e-05,
+      "loss": 6.0838,
+      "step": 8048
+    },
+    {
+      "epoch": 2.747098976109215,
+      "grad_norm": 2.6471140384674072,
+      "learning_rate": 8.430034129692833e-05,
+      "loss": 5.9484,
+      "step": 8049
+    },
+    {
+      "epoch": 2.747440273037543,
+      "grad_norm": 2.6072444915771484,
+      "learning_rate": 8.418657565415245e-05,
+      "loss": 6.4437,
+      "step": 8050
+    },
+    {
+      "epoch": 2.7477815699658703,
+      "grad_norm": 2.5881097316741943,
+      "learning_rate": 8.407281001137657e-05,
+      "loss": 5.7884,
+      "step": 8051
+    },
+    {
+      "epoch": 2.7481228668941977,
+      "grad_norm": 2.7298855781555176,
+      "learning_rate": 8.395904436860069e-05,
+      "loss": 4.845,
+      "step": 8052
+    },
+    {
+      "epoch": 2.7484641638225256,
+      "grad_norm": 2.646090507507324,
+      "learning_rate": 8.38452787258248e-05,
+      "loss": 6.2684,
+      "step": 8053
+    },
+    {
+      "epoch": 2.7488054607508534,
+      "grad_norm": 2.5868003368377686,
+      "learning_rate": 8.373151308304892e-05,
+      "loss": 5.8335,
+      "step": 8054
+    },
+    {
+      "epoch": 2.749146757679181,
+      "grad_norm": 2.60349440574646,
+      "learning_rate": 8.361774744027303e-05,
+      "loss": 5.6704,
+      "step": 8055
+    },
+    {
+      "epoch": 2.7494880546075087,
+      "grad_norm": 2.5340113639831543,
+      "learning_rate": 8.350398179749716e-05,
+      "loss": 5.5741,
+      "step": 8056
+    },
+    {
+      "epoch": 2.749829351535836,
+      "grad_norm": 2.6440858840942383,
+      "learning_rate": 8.339021615472128e-05,
+      "loss": 5.4331,
+      "step": 8057
+    },
+    {
+      "epoch": 2.750170648464164,
+      "grad_norm": 2.588313579559326,
+      "learning_rate": 8.32764505119454e-05,
+      "loss": 6.2252,
+      "step": 8058
+    },
+    {
+      "epoch": 2.7505119453924913,
+      "grad_norm": 2.574767827987671,
+      "learning_rate": 8.316268486916952e-05,
+      "loss": 6.3772,
+      "step": 8059
+    },
+    {
+      "epoch": 2.750853242320819,
+      "grad_norm": 2.5740933418273926,
+      "learning_rate": 8.304891922639362e-05,
+      "loss": 6.231,
+      "step": 8060
+    },
+    {
+      "epoch": 2.7511945392491466,
+      "grad_norm": 2.542821168899536,
+      "learning_rate": 8.293515358361775e-05,
+      "loss": 5.886,
+      "step": 8061
+    },
+    {
+      "epoch": 2.7515358361774744,
+      "grad_norm": 2.542856454849243,
+      "learning_rate": 8.282138794084186e-05,
+      "loss": 5.9924,
+      "step": 8062
+    },
+    {
+      "epoch": 2.7518771331058023,
+      "grad_norm": 2.618867874145508,
+      "learning_rate": 8.270762229806599e-05,
+      "loss": 5.3548,
+      "step": 8063
+    },
+    {
+      "epoch": 2.7522184300341297,
+      "grad_norm": 2.605292797088623,
+      "learning_rate": 8.259385665529011e-05,
+      "loss": 5.7645,
+      "step": 8064
+    },
+    {
+      "epoch": 2.752559726962457,
+      "grad_norm": 2.6475188732147217,
+      "learning_rate": 8.248009101251421e-05,
+      "loss": 6.4186,
+      "step": 8065
+    },
+    {
+      "epoch": 2.752901023890785,
+      "grad_norm": 2.5117578506469727,
+      "learning_rate": 8.236632536973834e-05,
+      "loss": 5.997,
+      "step": 8066
+    },
+    {
+      "epoch": 2.753242320819113,
+      "grad_norm": 2.5080535411834717,
+      "learning_rate": 8.225255972696245e-05,
+      "loss": 6.1888,
+      "step": 8067
+    },
+    {
+      "epoch": 2.75358361774744,
+      "grad_norm": 2.5350992679595947,
+      "learning_rate": 8.213879408418658e-05,
+      "loss": 5.9308,
+      "step": 8068
+    },
+    {
+      "epoch": 2.753924914675768,
+      "grad_norm": 2.5473434925079346,
+      "learning_rate": 8.20250284414107e-05,
+      "loss": 6.1776,
+      "step": 8069
+    },
+    {
+      "epoch": 2.7542662116040955,
+      "grad_norm": 2.4866843223571777,
+      "learning_rate": 8.19112627986348e-05,
+      "loss": 4.3249,
+      "step": 8070
+    },
+    {
+      "epoch": 2.7546075085324233,
+      "grad_norm": 2.6929423809051514,
+      "learning_rate": 8.179749715585894e-05,
+      "loss": 4.4646,
+      "step": 8071
+    },
+    {
+      "epoch": 2.7549488054607507,
+      "grad_norm": 2.553462505340576,
+      "learning_rate": 8.168373151308304e-05,
+      "loss": 6.4205,
+      "step": 8072
+    },
+    {
+      "epoch": 2.7552901023890786,
+      "grad_norm": 2.5952062606811523,
+      "learning_rate": 8.156996587030717e-05,
+      "loss": 6.2428,
+      "step": 8073
+    },
+    {
+      "epoch": 2.755631399317406,
+      "grad_norm": 2.697514057159424,
+      "learning_rate": 8.145620022753129e-05,
+      "loss": 6.3499,
+      "step": 8074
+    },
+    {
+      "epoch": 2.755972696245734,
+      "grad_norm": 2.558379888534546,
+      "learning_rate": 8.13424345847554e-05,
+      "loss": 5.889,
+      "step": 8075
+    },
+    {
+      "epoch": 2.7563139931740617,
+      "grad_norm": 2.633807420730591,
+      "learning_rate": 8.122866894197953e-05,
+      "loss": 5.544,
+      "step": 8076
+    },
+    {
+      "epoch": 2.756655290102389,
+      "grad_norm": 2.470966100692749,
+      "learning_rate": 8.111490329920363e-05,
+      "loss": 4.4105,
+      "step": 8077
+    },
+    {
+      "epoch": 2.7569965870307165,
+      "grad_norm": 2.6610970497131348,
+      "learning_rate": 8.100113765642777e-05,
+      "loss": 6.4596,
+      "step": 8078
+    },
+    {
+      "epoch": 2.7573378839590443,
+      "grad_norm": 2.5555026531219482,
+      "learning_rate": 8.088737201365187e-05,
+      "loss": 6.3518,
+      "step": 8079
+    },
+    {
+      "epoch": 2.757679180887372,
+      "grad_norm": 2.5696256160736084,
+      "learning_rate": 8.077360637087599e-05,
+      "loss": 6.0327,
+      "step": 8080
+    },
+    {
+      "epoch": 2.7580204778156996,
+      "grad_norm": 2.542940855026245,
+      "learning_rate": 8.065984072810012e-05,
+      "loss": 5.54,
+      "step": 8081
+    },
+    {
+      "epoch": 2.7583617747440274,
+      "grad_norm": 2.5934388637542725,
+      "learning_rate": 8.054607508532423e-05,
+      "loss": 6.237,
+      "step": 8082
+    },
+    {
+      "epoch": 2.758703071672355,
+      "grad_norm": 2.5500986576080322,
+      "learning_rate": 8.043230944254836e-05,
+      "loss": 6.0471,
+      "step": 8083
+    },
+    {
+      "epoch": 2.7590443686006827,
+      "grad_norm": 2.556598663330078,
+      "learning_rate": 8.031854379977246e-05,
+      "loss": 5.7077,
+      "step": 8084
+    },
+    {
+      "epoch": 2.75938566552901,
+      "grad_norm": 2.6231422424316406,
+      "learning_rate": 8.020477815699658e-05,
+      "loss": 5.1179,
+      "step": 8085
+    },
+    {
+      "epoch": 2.759726962457338,
+      "grad_norm": 2.5738203525543213,
+      "learning_rate": 8.009101251422071e-05,
+      "loss": 6.3482,
+      "step": 8086
+    },
+    {
+      "epoch": 2.7600682593856654,
+      "grad_norm": 2.713271379470825,
+      "learning_rate": 7.997724687144482e-05,
+      "loss": 6.5092,
+      "step": 8087
+    },
+    {
+      "epoch": 2.760409556313993,
+      "grad_norm": 2.689073085784912,
+      "learning_rate": 7.986348122866895e-05,
+      "loss": 5.0591,
+      "step": 8088
+    },
+    {
+      "epoch": 2.760750853242321,
+      "grad_norm": 2.620436429977417,
+      "learning_rate": 7.974971558589306e-05,
+      "loss": 6.0436,
+      "step": 8089
+    },
+    {
+      "epoch": 2.7610921501706485,
+      "grad_norm": 2.5438833236694336,
+      "learning_rate": 7.963594994311717e-05,
+      "loss": 5.7818,
+      "step": 8090
+    },
+    {
+      "epoch": 2.761433447098976,
+      "grad_norm": 2.6242706775665283,
+      "learning_rate": 7.952218430034131e-05,
+      "loss": 5.8769,
+      "step": 8091
+    },
+    {
+      "epoch": 2.7617747440273037,
+      "grad_norm": 2.4857864379882812,
+      "learning_rate": 7.940841865756541e-05,
+      "loss": 5.7621,
+      "step": 8092
+    },
+    {
+      "epoch": 2.7621160409556316,
+      "grad_norm": 2.627063035964966,
+      "learning_rate": 7.929465301478954e-05,
+      "loss": 6.45,
+      "step": 8093
+    },
+    {
+      "epoch": 2.762457337883959,
+      "grad_norm": 2.4991109371185303,
+      "learning_rate": 7.918088737201365e-05,
+      "loss": 5.7273,
+      "step": 8094
+    },
+    {
+      "epoch": 2.762798634812287,
+      "grad_norm": 2.532963991165161,
+      "learning_rate": 7.906712172923777e-05,
+      "loss": 6.2703,
+      "step": 8095
+    },
+    {
+      "epoch": 2.7631399317406142,
+      "grad_norm": 2.5222792625427246,
+      "learning_rate": 7.895335608646189e-05,
+      "loss": 5.554,
+      "step": 8096
+    },
+    {
+      "epoch": 2.763481228668942,
+      "grad_norm": 2.6223628520965576,
+      "learning_rate": 7.8839590443686e-05,
+      "loss": 6.0185,
+      "step": 8097
+    },
+    {
+      "epoch": 2.7638225255972695,
+      "grad_norm": 2.598527669906616,
+      "learning_rate": 7.872582480091014e-05,
+      "loss": 5.8662,
+      "step": 8098
+    },
+    {
+      "epoch": 2.7641638225255973,
+      "grad_norm": 2.4972290992736816,
+      "learning_rate": 7.861205915813424e-05,
+      "loss": 6.2221,
+      "step": 8099
+    },
+    {
+      "epoch": 2.7645051194539247,
+      "grad_norm": 2.5884251594543457,
+      "learning_rate": 7.849829351535836e-05,
+      "loss": 6.2226,
+      "step": 8100
+    },
+    {
+      "epoch": 2.7648464163822526,
+      "grad_norm": 2.6210410594940186,
+      "learning_rate": 7.838452787258248e-05,
+      "loss": 5.378,
+      "step": 8101
+    },
+    {
+      "epoch": 2.7651877133105804,
+      "grad_norm": 2.380128860473633,
+      "learning_rate": 7.82707622298066e-05,
+      "loss": 4.1676,
+      "step": 8102
+    },
+    {
+      "epoch": 2.765529010238908,
+      "grad_norm": 2.6382617950439453,
+      "learning_rate": 7.815699658703073e-05,
+      "loss": 6.0014,
+      "step": 8103
+    },
+    {
+      "epoch": 2.7658703071672353,
+      "grad_norm": 2.6179182529449463,
+      "learning_rate": 7.804323094425483e-05,
+      "loss": 5.9414,
+      "step": 8104
+    },
+    {
+      "epoch": 2.766211604095563,
+      "grad_norm": 2.5637147426605225,
+      "learning_rate": 7.792946530147895e-05,
+      "loss": 5.5923,
+      "step": 8105
+    },
+    {
+      "epoch": 2.766552901023891,
+      "grad_norm": 2.6299054622650146,
+      "learning_rate": 7.781569965870307e-05,
+      "loss": 5.6322,
+      "step": 8106
+    },
+    {
+      "epoch": 2.7668941979522184,
+      "grad_norm": 2.7012157440185547,
+      "learning_rate": 7.770193401592719e-05,
+      "loss": 5.6976,
+      "step": 8107
+    },
+    {
+      "epoch": 2.767235494880546,
+      "grad_norm": 2.60090970993042,
+      "learning_rate": 7.758816837315132e-05,
+      "loss": 6.3271,
+      "step": 8108
+    },
+    {
+      "epoch": 2.7675767918088736,
+      "grad_norm": 2.5655224323272705,
+      "learning_rate": 7.747440273037543e-05,
+      "loss": 5.833,
+      "step": 8109
+    },
+    {
+      "epoch": 2.7679180887372015,
+      "grad_norm": 2.630000114440918,
+      "learning_rate": 7.736063708759954e-05,
+      "loss": 5.7833,
+      "step": 8110
+    },
+    {
+      "epoch": 2.768259385665529,
+      "grad_norm": 2.582270860671997,
+      "learning_rate": 7.724687144482366e-05,
+      "loss": 5.4806,
+      "step": 8111
+    },
+    {
+      "epoch": 2.7686006825938567,
+      "grad_norm": 2.481131076812744,
+      "learning_rate": 7.713310580204778e-05,
+      "loss": 4.7699,
+      "step": 8112
+    },
+    {
+      "epoch": 2.768941979522184,
+      "grad_norm": 1.8021996021270752,
+      "learning_rate": 7.701934015927191e-05,
+      "loss": 3.0427,
+      "step": 8113
+    },
+    {
+      "epoch": 2.769283276450512,
+      "grad_norm": 2.502584457397461,
+      "learning_rate": 7.690557451649602e-05,
+      "loss": 6.344,
+      "step": 8114
+    },
+    {
+      "epoch": 2.76962457337884,
+      "grad_norm": 2.3312814235687256,
+      "learning_rate": 7.679180887372014e-05,
+      "loss": 4.1687,
+      "step": 8115
+    },
+    {
+      "epoch": 2.7699658703071672,
+      "grad_norm": 2.548377752304077,
+      "learning_rate": 7.667804323094426e-05,
+      "loss": 5.0895,
+      "step": 8116
+    },
+    {
+      "epoch": 2.7703071672354946,
+      "grad_norm": 2.6071834564208984,
+      "learning_rate": 7.656427758816837e-05,
+      "loss": 5.0985,
+      "step": 8117
+    },
+    {
+      "epoch": 2.7706484641638225,
+      "grad_norm": 2.3951501846313477,
+      "learning_rate": 7.645051194539249e-05,
+      "loss": 4.862,
+      "step": 8118
+    },
+    {
+      "epoch": 2.7709897610921503,
+      "grad_norm": 2.610914945602417,
+      "learning_rate": 7.633674630261661e-05,
+      "loss": 6.137,
+      "step": 8119
+    },
+    {
+      "epoch": 2.7713310580204777,
+      "grad_norm": 2.4345929622650146,
+      "learning_rate": 7.622298065984073e-05,
+      "loss": 4.7642,
+      "step": 8120
+    },
+    {
+      "epoch": 2.7716723549488056,
+      "grad_norm": 2.550601005554199,
+      "learning_rate": 7.610921501706485e-05,
+      "loss": 5.3778,
+      "step": 8121
+    },
+    {
+      "epoch": 2.772013651877133,
+      "grad_norm": 2.6055097579956055,
+      "learning_rate": 7.599544937428897e-05,
+      "loss": 5.7184,
+      "step": 8122
+    },
+    {
+      "epoch": 2.772354948805461,
+      "grad_norm": 2.512861490249634,
+      "learning_rate": 7.588168373151309e-05,
+      "loss": 5.2337,
+      "step": 8123
+    },
+    {
+      "epoch": 2.7726962457337883,
+      "grad_norm": 2.71311354637146,
+      "learning_rate": 7.57679180887372e-05,
+      "loss": 6.3602,
+      "step": 8124
+    },
+    {
+      "epoch": 2.773037542662116,
+      "grad_norm": 2.6916327476501465,
+      "learning_rate": 7.565415244596132e-05,
+      "loss": 6.0187,
+      "step": 8125
+    },
+    {
+      "epoch": 2.7733788395904435,
+      "grad_norm": 2.6451501846313477,
+      "learning_rate": 7.554038680318544e-05,
+      "loss": 5.8547,
+      "step": 8126
+    },
+    {
+      "epoch": 2.7737201365187714,
+      "grad_norm": 2.736882209777832,
+      "learning_rate": 7.542662116040956e-05,
+      "loss": 6.2056,
+      "step": 8127
+    },
+    {
+      "epoch": 2.774061433447099,
+      "grad_norm": 2.5844454765319824,
+      "learning_rate": 7.531285551763368e-05,
+      "loss": 5.795,
+      "step": 8128
+    },
+    {
+      "epoch": 2.7744027303754266,
+      "grad_norm": 2.4675440788269043,
+      "learning_rate": 7.51990898748578e-05,
+      "loss": 5.4714,
+      "step": 8129
+    },
+    {
+      "epoch": 2.774744027303754,
+      "grad_norm": 2.4359679222106934,
+      "learning_rate": 7.508532423208191e-05,
+      "loss": 5.8461,
+      "step": 8130
+    },
+    {
+      "epoch": 2.775085324232082,
+      "grad_norm": 2.678617477416992,
+      "learning_rate": 7.497155858930603e-05,
+      "loss": 5.3954,
+      "step": 8131
+    },
+    {
+      "epoch": 2.7754266211604097,
+      "grad_norm": 2.575972557067871,
+      "learning_rate": 7.485779294653015e-05,
+      "loss": 6.3859,
+      "step": 8132
+    },
+    {
+      "epoch": 2.775767918088737,
+      "grad_norm": 2.5873968601226807,
+      "learning_rate": 7.474402730375427e-05,
+      "loss": 6.2744,
+      "step": 8133
+    },
+    {
+      "epoch": 2.776109215017065,
+      "grad_norm": 2.5679585933685303,
+      "learning_rate": 7.463026166097839e-05,
+      "loss": 6.3518,
+      "step": 8134
+    },
+    {
+      "epoch": 2.7764505119453924,
+      "grad_norm": 2.5673532485961914,
+      "learning_rate": 7.45164960182025e-05,
+      "loss": 5.3267,
+      "step": 8135
+    },
+    {
+      "epoch": 2.7767918088737202,
+      "grad_norm": 2.5870144367218018,
+      "learning_rate": 7.440273037542663e-05,
+      "loss": 5.5847,
+      "step": 8136
+    },
+    {
+      "epoch": 2.7771331058020476,
+      "grad_norm": 2.542771100997925,
+      "learning_rate": 7.428896473265074e-05,
+      "loss": 4.9482,
+      "step": 8137
+    },
+    {
+      "epoch": 2.7774744027303755,
+      "grad_norm": 2.619039535522461,
+      "learning_rate": 7.417519908987486e-05,
+      "loss": 6.4865,
+      "step": 8138
+    },
+    {
+      "epoch": 2.777815699658703,
+      "grad_norm": 2.628445863723755,
+      "learning_rate": 7.406143344709898e-05,
+      "loss": 5.8237,
+      "step": 8139
+    },
+    {
+      "epoch": 2.7781569965870307,
+      "grad_norm": 2.5685930252075195,
+      "learning_rate": 7.394766780432309e-05,
+      "loss": 6.1367,
+      "step": 8140
+    },
+    {
+      "epoch": 2.7784982935153586,
+      "grad_norm": 2.6697919368743896,
+      "learning_rate": 7.383390216154722e-05,
+      "loss": 5.8577,
+      "step": 8141
+    },
+    {
+      "epoch": 2.778839590443686,
+      "grad_norm": 2.6898036003112793,
+      "learning_rate": 7.372013651877134e-05,
+      "loss": 6.3594,
+      "step": 8142
+    },
+    {
+      "epoch": 2.7791808873720134,
+      "grad_norm": 2.5270895957946777,
+      "learning_rate": 7.360637087599546e-05,
+      "loss": 6.1053,
+      "step": 8143
+    },
+    {
+      "epoch": 2.7795221843003413,
+      "grad_norm": 2.6355865001678467,
+      "learning_rate": 7.349260523321957e-05,
+      "loss": 4.7732,
+      "step": 8144
+    },
+    {
+      "epoch": 2.779863481228669,
+      "grad_norm": 2.624152898788452,
+      "learning_rate": 7.337883959044368e-05,
+      "loss": 6.3116,
+      "step": 8145
+    },
+    {
+      "epoch": 2.7802047781569965,
+      "grad_norm": 2.588355302810669,
+      "learning_rate": 7.326507394766781e-05,
+      "loss": 5.3328,
+      "step": 8146
+    },
+    {
+      "epoch": 2.7805460750853244,
+      "grad_norm": 2.633571147918701,
+      "learning_rate": 7.315130830489193e-05,
+      "loss": 6.1316,
+      "step": 8147
+    },
+    {
+      "epoch": 2.7808873720136518,
+      "grad_norm": 2.573197364807129,
+      "learning_rate": 7.303754266211605e-05,
+      "loss": 6.7343,
+      "step": 8148
+    },
+    {
+      "epoch": 2.7812286689419796,
+      "grad_norm": 2.558769464492798,
+      "learning_rate": 7.292377701934017e-05,
+      "loss": 5.357,
+      "step": 8149
+    },
+    {
+      "epoch": 2.781569965870307,
+      "grad_norm": 2.6176862716674805,
+      "learning_rate": 7.281001137656427e-05,
+      "loss": 5.6288,
+      "step": 8150
+    },
+    {
+      "epoch": 2.781911262798635,
+      "grad_norm": 2.510423183441162,
+      "learning_rate": 7.26962457337884e-05,
+      "loss": 6.4119,
+      "step": 8151
+    },
+    {
+      "epoch": 2.7822525597269623,
+      "grad_norm": 2.7808709144592285,
+      "learning_rate": 7.258248009101251e-05,
+      "loss": 4.3593,
+      "step": 8152
+    },
+    {
+      "epoch": 2.78259385665529,
+      "grad_norm": 2.575263261795044,
+      "learning_rate": 7.246871444823664e-05,
+      "loss": 6.4656,
+      "step": 8153
+    },
+    {
+      "epoch": 2.782935153583618,
+      "grad_norm": 2.588742256164551,
+      "learning_rate": 7.235494880546076e-05,
+      "loss": 5.9384,
+      "step": 8154
+    },
+    {
+      "epoch": 2.7832764505119454,
+      "grad_norm": 2.6123979091644287,
+      "learning_rate": 7.224118316268486e-05,
+      "loss": 6.0583,
+      "step": 8155
+    },
+    {
+      "epoch": 2.783617747440273,
+      "grad_norm": 2.4757981300354004,
+      "learning_rate": 7.2127417519909e-05,
+      "loss": 5.5663,
+      "step": 8156
+    },
+    {
+      "epoch": 2.7839590443686006,
+      "grad_norm": 2.5429530143737793,
+      "learning_rate": 7.20136518771331e-05,
+      "loss": 5.9294,
+      "step": 8157
+    },
+    {
+      "epoch": 2.7843003412969285,
+      "grad_norm": 2.5413453578948975,
+      "learning_rate": 7.189988623435722e-05,
+      "loss": 5.9371,
+      "step": 8158
+    },
+    {
+      "epoch": 2.784641638225256,
+      "grad_norm": 2.5826592445373535,
+      "learning_rate": 7.178612059158135e-05,
+      "loss": 5.9754,
+      "step": 8159
+    },
+    {
+      "epoch": 2.7849829351535837,
+      "grad_norm": 2.4613709449768066,
+      "learning_rate": 7.167235494880546e-05,
+      "loss": 5.7554,
+      "step": 8160
+    },
+    {
+      "epoch": 2.785324232081911,
+      "grad_norm": 2.59067964553833,
+      "learning_rate": 7.155858930602959e-05,
+      "loss": 6.3009,
+      "step": 8161
+    },
+    {
+      "epoch": 2.785665529010239,
+      "grad_norm": 2.543574810028076,
+      "learning_rate": 7.144482366325369e-05,
+      "loss": 5.7074,
+      "step": 8162
+    },
+    {
+      "epoch": 2.7860068259385664,
+      "grad_norm": 2.595146656036377,
+      "learning_rate": 7.133105802047781e-05,
+      "loss": 6.4757,
+      "step": 8163
+    },
+    {
+      "epoch": 2.7863481228668943,
+      "grad_norm": 5.569501876831055,
+      "learning_rate": 7.121729237770194e-05,
+      "loss": 4.9309,
+      "step": 8164
+    },
+    {
+      "epoch": 2.7866894197952217,
+      "grad_norm": 2.49450945854187,
+      "learning_rate": 7.110352673492605e-05,
+      "loss": 5.1546,
+      "step": 8165
+    },
+    {
+      "epoch": 2.7870307167235495,
+      "grad_norm": 2.642744541168213,
+      "learning_rate": 7.098976109215018e-05,
+      "loss": 6.0856,
+      "step": 8166
+    },
+    {
+      "epoch": 2.7873720136518774,
+      "grad_norm": 2.5893845558166504,
+      "learning_rate": 7.087599544937429e-05,
+      "loss": 6.0408,
+      "step": 8167
+    },
+    {
+      "epoch": 2.7877133105802048,
+      "grad_norm": 2.603260040283203,
+      "learning_rate": 7.07622298065984e-05,
+      "loss": 5.7834,
+      "step": 8168
+    },
+    {
+      "epoch": 2.788054607508532,
+      "grad_norm": 2.669589042663574,
+      "learning_rate": 7.064846416382252e-05,
+      "loss": 5.6021,
+      "step": 8169
+    },
+    {
+      "epoch": 2.78839590443686,
+      "grad_norm": 2.57565975189209,
+      "learning_rate": 7.053469852104664e-05,
+      "loss": 5.8934,
+      "step": 8170
+    },
+    {
+      "epoch": 2.788737201365188,
+      "grad_norm": 2.455730438232422,
+      "learning_rate": 7.042093287827077e-05,
+      "loss": 5.5977,
+      "step": 8171
+    },
+    {
+      "epoch": 2.7890784982935153,
+      "grad_norm": 2.560124635696411,
+      "learning_rate": 7.030716723549488e-05,
+      "loss": 5.4134,
+      "step": 8172
+    },
+    {
+      "epoch": 2.789419795221843,
+      "grad_norm": 2.573716163635254,
+      "learning_rate": 7.0193401592719e-05,
+      "loss": 5.9382,
+      "step": 8173
+    },
+    {
+      "epoch": 2.7897610921501705,
+      "grad_norm": 2.4947168827056885,
+      "learning_rate": 7.007963594994311e-05,
+      "loss": 5.3002,
+      "step": 8174
+    },
+    {
+      "epoch": 2.7901023890784984,
+      "grad_norm": 2.6110596656799316,
+      "learning_rate": 6.996587030716723e-05,
+      "loss": 6.0533,
+      "step": 8175
+    },
+    {
+      "epoch": 2.790443686006826,
+      "grad_norm": 2.5457863807678223,
+      "learning_rate": 6.985210466439137e-05,
+      "loss": 6.1391,
+      "step": 8176
+    },
+    {
+      "epoch": 2.7907849829351536,
+      "grad_norm": 2.6228179931640625,
+      "learning_rate": 6.973833902161547e-05,
+      "loss": 6.0603,
+      "step": 8177
+    },
+    {
+      "epoch": 2.791126279863481,
+      "grad_norm": 2.5657594203948975,
+      "learning_rate": 6.962457337883959e-05,
+      "loss": 5.8292,
+      "step": 8178
+    },
+    {
+      "epoch": 2.791467576791809,
+      "grad_norm": 2.512878894805908,
+      "learning_rate": 6.951080773606371e-05,
+      "loss": 5.5296,
+      "step": 8179
+    },
+    {
+      "epoch": 2.7918088737201368,
+      "grad_norm": 2.526179075241089,
+      "learning_rate": 6.939704209328783e-05,
+      "loss": 5.4596,
+      "step": 8180
+    },
+    {
+      "epoch": 2.792150170648464,
+      "grad_norm": 2.5494656562805176,
+      "learning_rate": 6.928327645051196e-05,
+      "loss": 5.7413,
+      "step": 8181
+    },
+    {
+      "epoch": 2.7924914675767916,
+      "grad_norm": 2.623551368713379,
+      "learning_rate": 6.916951080773606e-05,
+      "loss": 5.9198,
+      "step": 8182
+    },
+    {
+      "epoch": 2.7928327645051194,
+      "grad_norm": 2.609656572341919,
+      "learning_rate": 6.905574516496018e-05,
+      "loss": 6.1618,
+      "step": 8183
+    },
+    {
+      "epoch": 2.7931740614334473,
+      "grad_norm": 3.3349082469940186,
+      "learning_rate": 6.89419795221843e-05,
+      "loss": 4.9097,
+      "step": 8184
+    },
+    {
+      "epoch": 2.7935153583617747,
+      "grad_norm": 2.6144847869873047,
+      "learning_rate": 6.882821387940842e-05,
+      "loss": 6.3871,
+      "step": 8185
+    },
+    {
+      "epoch": 2.7938566552901025,
+      "grad_norm": 2.5866599082946777,
+      "learning_rate": 6.871444823663254e-05,
+      "loss": 5.3323,
+      "step": 8186
+    },
+    {
+      "epoch": 2.79419795221843,
+      "grad_norm": 2.489753246307373,
+      "learning_rate": 6.860068259385666e-05,
+      "loss": 5.2399,
+      "step": 8187
+    },
+    {
+      "epoch": 2.7945392491467578,
+      "grad_norm": 2.5292809009552,
+      "learning_rate": 6.848691695108077e-05,
+      "loss": 4.0588,
+      "step": 8188
+    },
+    {
+      "epoch": 2.794880546075085,
+      "grad_norm": 2.519035577774048,
+      "learning_rate": 6.837315130830489e-05,
+      "loss": 5.8272,
+      "step": 8189
+    },
+    {
+      "epoch": 2.795221843003413,
+      "grad_norm": 2.5505032539367676,
+      "learning_rate": 6.825938566552901e-05,
+      "loss": 6.0068,
+      "step": 8190
+    },
+    {
+      "epoch": 2.7955631399317404,
+      "grad_norm": 2.5838968753814697,
+      "learning_rate": 6.814562002275313e-05,
+      "loss": 6.1091,
+      "step": 8191
+    },
+    {
+      "epoch": 2.7959044368600683,
+      "grad_norm": 4.409801006317139,
+      "learning_rate": 6.803185437997725e-05,
+      "loss": 4.6617,
+      "step": 8192
+    },
+    {
+      "epoch": 2.796245733788396,
+      "grad_norm": 2.558980703353882,
+      "learning_rate": 6.791808873720137e-05,
+      "loss": 5.8792,
+      "step": 8193
+    },
+    {
+      "epoch": 2.7965870307167235,
+      "grad_norm": 2.547933340072632,
+      "learning_rate": 6.780432309442548e-05,
+      "loss": 6.102,
+      "step": 8194
+    },
+    {
+      "epoch": 2.796928327645051,
+      "grad_norm": 2.719900608062744,
+      "learning_rate": 6.76905574516496e-05,
+      "loss": 5.5898,
+      "step": 8195
+    },
+    {
+      "epoch": 2.797269624573379,
+      "grad_norm": 2.68369722366333,
+      "learning_rate": 6.757679180887372e-05,
+      "loss": 5.1553,
+      "step": 8196
+    },
+    {
+      "epoch": 2.7976109215017066,
+      "grad_norm": 2.6060352325439453,
+      "learning_rate": 6.746302616609784e-05,
+      "loss": 6.2941,
+      "step": 8197
+    },
+    {
+      "epoch": 2.797952218430034,
+      "grad_norm": 2.6711232662200928,
+      "learning_rate": 6.734926052332196e-05,
+      "loss": 6.4384,
+      "step": 8198
+    },
+    {
+      "epoch": 2.798293515358362,
+      "grad_norm": 2.690762996673584,
+      "learning_rate": 6.723549488054608e-05,
+      "loss": 6.0928,
+      "step": 8199
+    },
+    {
+      "epoch": 2.7986348122866893,
+      "grad_norm": 2.6682162284851074,
+      "learning_rate": 6.71217292377702e-05,
+      "loss": 6.2376,
+      "step": 8200
+    },
+    {
+      "epoch": 2.798976109215017,
+      "grad_norm": 2.683518171310425,
+      "learning_rate": 6.700796359499431e-05,
+      "loss": 6.319,
+      "step": 8201
+    },
+    {
+      "epoch": 2.7993174061433446,
+      "grad_norm": 2.5631539821624756,
+      "learning_rate": 6.689419795221843e-05,
+      "loss": 6.1925,
+      "step": 8202
+    },
+    {
+      "epoch": 2.7996587030716724,
+      "grad_norm": 2.5291433334350586,
+      "learning_rate": 6.678043230944254e-05,
+      "loss": 6.2034,
+      "step": 8203
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 2.4929697513580322,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 5.6572,
+      "step": 8204
+    },
+    {
+      "epoch": 2.8003412969283277,
+      "grad_norm": 2.6479387283325195,
+      "learning_rate": 6.655290102389079e-05,
+      "loss": 5.844,
+      "step": 8205
+    },
+    {
+      "epoch": 2.8006825938566555,
+      "grad_norm": 3.730424404144287,
+      "learning_rate": 6.643913538111491e-05,
+      "loss": 3.1324,
+      "step": 8206
+    },
+    {
+      "epoch": 2.801023890784983,
+      "grad_norm": 2.496727228164673,
+      "learning_rate": 6.632536973833903e-05,
+      "loss": 5.9789,
+      "step": 8207
+    },
+    {
+      "epoch": 2.8013651877133103,
+      "grad_norm": 2.9955997467041016,
+      "learning_rate": 6.621160409556313e-05,
+      "loss": 4.9055,
+      "step": 8208
+    },
+    {
+      "epoch": 2.801706484641638,
+      "grad_norm": 2.6087541580200195,
+      "learning_rate": 6.609783845278726e-05,
+      "loss": 6.3396,
+      "step": 8209
+    },
+    {
+      "epoch": 2.802047781569966,
+      "grad_norm": 2.530484676361084,
+      "learning_rate": 6.598407281001138e-05,
+      "loss": 5.6127,
+      "step": 8210
+    },
+    {
+      "epoch": 2.8023890784982934,
+      "grad_norm": 2.2731897830963135,
+      "learning_rate": 6.58703071672355e-05,
+      "loss": 3.3329,
+      "step": 8211
+    },
+    {
+      "epoch": 2.8027303754266213,
+      "grad_norm": 2.5415871143341064,
+      "learning_rate": 6.575654152445962e-05,
+      "loss": 5.8304,
+      "step": 8212
+    },
+    {
+      "epoch": 2.8030716723549487,
+      "grad_norm": 2.5074963569641113,
+      "learning_rate": 6.564277588168372e-05,
+      "loss": 5.6721,
+      "step": 8213
+    },
+    {
+      "epoch": 2.8034129692832765,
+      "grad_norm": 2.5421886444091797,
+      "learning_rate": 6.552901023890785e-05,
+      "loss": 6.2318,
+      "step": 8214
+    },
+    {
+      "epoch": 2.803754266211604,
+      "grad_norm": 2.5332021713256836,
+      "learning_rate": 6.541524459613197e-05,
+      "loss": 5.7743,
+      "step": 8215
+    },
+    {
+      "epoch": 2.804095563139932,
+      "grad_norm": 2.539001226425171,
+      "learning_rate": 6.530147895335609e-05,
+      "loss": 5.5197,
+      "step": 8216
+    },
+    {
+      "epoch": 2.804436860068259,
+      "grad_norm": 2.585606336593628,
+      "learning_rate": 6.518771331058021e-05,
+      "loss": 5.9402,
+      "step": 8217
+    },
+    {
+      "epoch": 2.804778156996587,
+      "grad_norm": 2.5669174194335938,
+      "learning_rate": 6.507394766780432e-05,
+      "loss": 5.8436,
+      "step": 8218
+    },
+    {
+      "epoch": 2.805119453924915,
+      "grad_norm": 2.6501104831695557,
+      "learning_rate": 6.496018202502845e-05,
+      "loss": 6.0262,
+      "step": 8219
+    },
+    {
+      "epoch": 2.8054607508532423,
+      "grad_norm": 2.6469192504882812,
+      "learning_rate": 6.484641638225255e-05,
+      "loss": 6.1936,
+      "step": 8220
+    },
+    {
+      "epoch": 2.8058020477815697,
+      "grad_norm": 2.4249207973480225,
+      "learning_rate": 6.473265073947668e-05,
+      "loss": 5.074,
+      "step": 8221
+    },
+    {
+      "epoch": 2.8061433447098976,
+      "grad_norm": 2.5924134254455566,
+      "learning_rate": 6.46188850967008e-05,
+      "loss": 6.188,
+      "step": 8222
+    },
+    {
+      "epoch": 2.8064846416382254,
+      "grad_norm": 2.559102773666382,
+      "learning_rate": 6.450511945392491e-05,
+      "loss": 5.493,
+      "step": 8223
+    },
+    {
+      "epoch": 2.806825938566553,
+      "grad_norm": 2.6225290298461914,
+      "learning_rate": 6.439135381114904e-05,
+      "loss": 5.5104,
+      "step": 8224
+    },
+    {
+      "epoch": 2.8071672354948807,
+      "grad_norm": 2.5742197036743164,
+      "learning_rate": 6.427758816837314e-05,
+      "loss": 6.0183,
+      "step": 8225
+    },
+    {
+      "epoch": 2.807508532423208,
+      "grad_norm": 2.534911870956421,
+      "learning_rate": 6.416382252559728e-05,
+      "loss": 5.6932,
+      "step": 8226
+    },
+    {
+      "epoch": 2.807849829351536,
+      "grad_norm": 2.5558876991271973,
+      "learning_rate": 6.40500568828214e-05,
+      "loss": 5.7026,
+      "step": 8227
+    },
+    {
+      "epoch": 2.8081911262798633,
+      "grad_norm": 2.528634786605835,
+      "learning_rate": 6.39362912400455e-05,
+      "loss": 5.7845,
+      "step": 8228
+    },
+    {
+      "epoch": 2.808532423208191,
+      "grad_norm": 2.934544324874878,
+      "learning_rate": 6.382252559726963e-05,
+      "loss": 4.4868,
+      "step": 8229
+    },
+    {
+      "epoch": 2.8088737201365186,
+      "grad_norm": 2.5007529258728027,
+      "learning_rate": 6.370875995449374e-05,
+      "loss": 4.6475,
+      "step": 8230
+    },
+    {
+      "epoch": 2.8092150170648464,
+      "grad_norm": 2.566521644592285,
+      "learning_rate": 6.359499431171787e-05,
+      "loss": 5.7933,
+      "step": 8231
+    },
+    {
+      "epoch": 2.8095563139931743,
+      "grad_norm": 2.5641417503356934,
+      "learning_rate": 6.348122866894199e-05,
+      "loss": 5.4627,
+      "step": 8232
+    },
+    {
+      "epoch": 2.8098976109215017,
+      "grad_norm": 2.489718437194824,
+      "learning_rate": 6.336746302616609e-05,
+      "loss": 5.6577,
+      "step": 8233
+    },
+    {
+      "epoch": 2.810238907849829,
+      "grad_norm": 2.5271124839782715,
+      "learning_rate": 6.325369738339022e-05,
+      "loss": 5.8988,
+      "step": 8234
+    },
+    {
+      "epoch": 2.810580204778157,
+      "grad_norm": 2.544497013092041,
+      "learning_rate": 6.313993174061433e-05,
+      "loss": 6.0608,
+      "step": 8235
+    },
+    {
+      "epoch": 2.810921501706485,
+      "grad_norm": 3.1858813762664795,
+      "learning_rate": 6.302616609783846e-05,
+      "loss": 4.988,
+      "step": 8236
+    },
+    {
+      "epoch": 2.811262798634812,
+      "grad_norm": 2.5615243911743164,
+      "learning_rate": 6.291240045506257e-05,
+      "loss": 5.9591,
+      "step": 8237
+    },
+    {
+      "epoch": 2.81160409556314,
+      "grad_norm": 2.4544830322265625,
+      "learning_rate": 6.279863481228669e-05,
+      "loss": 6.0645,
+      "step": 8238
+    },
+    {
+      "epoch": 2.8119453924914675,
+      "grad_norm": 2.5661332607269287,
+      "learning_rate": 6.268486916951082e-05,
+      "loss": 4.455,
+      "step": 8239
+    },
+    {
+      "epoch": 2.8122866894197953,
+      "grad_norm": 2.561377763748169,
+      "learning_rate": 6.257110352673492e-05,
+      "loss": 6.0717,
+      "step": 8240
+    },
+    {
+      "epoch": 2.8126279863481227,
+      "grad_norm": 2.537170648574829,
+      "learning_rate": 6.245733788395905e-05,
+      "loss": 4.6822,
+      "step": 8241
+    },
+    {
+      "epoch": 2.8129692832764506,
+      "grad_norm": 2.4561476707458496,
+      "learning_rate": 6.234357224118316e-05,
+      "loss": 5.5428,
+      "step": 8242
+    },
+    {
+      "epoch": 2.813310580204778,
+      "grad_norm": 2.530233383178711,
+      "learning_rate": 6.222980659840728e-05,
+      "loss": 6.0164,
+      "step": 8243
+    },
+    {
+      "epoch": 2.813651877133106,
+      "grad_norm": 2.61234974861145,
+      "learning_rate": 6.21160409556314e-05,
+      "loss": 5.5028,
+      "step": 8244
+    },
+    {
+      "epoch": 2.8139931740614337,
+      "grad_norm": 2.5273263454437256,
+      "learning_rate": 6.200227531285551e-05,
+      "loss": 5.8513,
+      "step": 8245
+    },
+    {
+      "epoch": 2.814334470989761,
+      "grad_norm": 2.5896804332733154,
+      "learning_rate": 6.188850967007965e-05,
+      "loss": 5.3815,
+      "step": 8246
+    },
+    {
+      "epoch": 2.8146757679180885,
+      "grad_norm": 2.605863332748413,
+      "learning_rate": 6.177474402730375e-05,
+      "loss": 5.9044,
+      "step": 8247
+    },
+    {
+      "epoch": 2.8150170648464163,
+      "grad_norm": 2.4373059272766113,
+      "learning_rate": 6.166097838452787e-05,
+      "loss": 5.8908,
+      "step": 8248
+    },
+    {
+      "epoch": 2.815358361774744,
+      "grad_norm": 2.578674793243408,
+      "learning_rate": 6.154721274175199e-05,
+      "loss": 6.1842,
+      "step": 8249
+    },
+    {
+      "epoch": 2.8156996587030716,
+      "grad_norm": 2.639101028442383,
+      "learning_rate": 6.143344709897611e-05,
+      "loss": 5.4184,
+      "step": 8250
+    },
+    {
+      "epoch": 2.8160409556313994,
+      "grad_norm": 2.5843536853790283,
+      "learning_rate": 6.131968145620023e-05,
+      "loss": 6.1219,
+      "step": 8251
+    },
+    {
+      "epoch": 2.816382252559727,
+      "grad_norm": 2.562448263168335,
+      "learning_rate": 6.120591581342434e-05,
+      "loss": 5.6921,
+      "step": 8252
+    },
+    {
+      "epoch": 2.8167235494880547,
+      "grad_norm": 2.592428207397461,
+      "learning_rate": 6.109215017064846e-05,
+      "loss": 6.1349,
+      "step": 8253
+    },
+    {
+      "epoch": 2.817064846416382,
+      "grad_norm": 2.594329357147217,
+      "learning_rate": 6.097838452787258e-05,
+      "loss": 6.4242,
+      "step": 8254
+    },
+    {
+      "epoch": 2.81740614334471,
+      "grad_norm": 2.5750083923339844,
+      "learning_rate": 6.08646188850967e-05,
+      "loss": 5.7389,
+      "step": 8255
+    },
+    {
+      "epoch": 2.8177474402730374,
+      "grad_norm": 2.5281083583831787,
+      "learning_rate": 6.075085324232082e-05,
+      "loss": 5.9597,
+      "step": 8256
+    },
+    {
+      "epoch": 2.818088737201365,
+      "grad_norm": 2.5450093746185303,
+      "learning_rate": 6.063708759954494e-05,
+      "loss": 5.7704,
+      "step": 8257
+    },
+    {
+      "epoch": 2.818430034129693,
+      "grad_norm": 2.50821852684021,
+      "learning_rate": 6.052332195676906e-05,
+      "loss": 5.5946,
+      "step": 8258
+    },
+    {
+      "epoch": 2.8187713310580205,
+      "grad_norm": 2.6139910221099854,
+      "learning_rate": 6.0409556313993174e-05,
+      "loss": 6.1745,
+      "step": 8259
+    },
+    {
+      "epoch": 2.819112627986348,
+      "grad_norm": 2.6088783740997314,
+      "learning_rate": 6.029579067121729e-05,
+      "loss": 5.8251,
+      "step": 8260
+    },
+    {
+      "epoch": 2.8194539249146757,
+      "grad_norm": 2.6340761184692383,
+      "learning_rate": 6.018202502844141e-05,
+      "loss": 4.9621,
+      "step": 8261
+    },
+    {
+      "epoch": 2.8197952218430036,
+      "grad_norm": 2.5626347064971924,
+      "learning_rate": 6.006825938566553e-05,
+      "loss": 5.829,
+      "step": 8262
+    },
+    {
+      "epoch": 2.820136518771331,
+      "grad_norm": 2.573450803756714,
+      "learning_rate": 5.9954493742889654e-05,
+      "loss": 5.9689,
+      "step": 8263
+    },
+    {
+      "epoch": 2.820477815699659,
+      "grad_norm": 2.446423053741455,
+      "learning_rate": 5.9840728100113766e-05,
+      "loss": 5.088,
+      "step": 8264
+    },
+    {
+      "epoch": 2.8208191126279862,
+      "grad_norm": 2.590911865234375,
+      "learning_rate": 5.9726962457337885e-05,
+      "loss": 5.8315,
+      "step": 8265
+    },
+    {
+      "epoch": 2.821160409556314,
+      "grad_norm": 2.581800699234009,
+      "learning_rate": 5.9613196814562e-05,
+      "loss": 6.0672,
+      "step": 8266
+    },
+    {
+      "epoch": 2.8215017064846415,
+      "grad_norm": 2.534433126449585,
+      "learning_rate": 5.949943117178612e-05,
+      "loss": 5.7578,
+      "step": 8267
+    },
+    {
+      "epoch": 2.8218430034129693,
+      "grad_norm": 2.6216630935668945,
+      "learning_rate": 5.938566552901024e-05,
+      "loss": 5.6905,
+      "step": 8268
+    },
+    {
+      "epoch": 2.8221843003412967,
+      "grad_norm": 2.5895118713378906,
+      "learning_rate": 5.927189988623436e-05,
+      "loss": 6.1605,
+      "step": 8269
+    },
+    {
+      "epoch": 2.8225255972696246,
+      "grad_norm": 2.5925073623657227,
+      "learning_rate": 5.915813424345848e-05,
+      "loss": 6.0352,
+      "step": 8270
+    },
+    {
+      "epoch": 2.8228668941979524,
+      "grad_norm": 2.5645461082458496,
+      "learning_rate": 5.9044368600682596e-05,
+      "loss": 5.7567,
+      "step": 8271
+    },
+    {
+      "epoch": 2.82320819112628,
+      "grad_norm": 2.479318857192993,
+      "learning_rate": 5.8930602957906714e-05,
+      "loss": 5.6575,
+      "step": 8272
+    },
+    {
+      "epoch": 2.8235494880546073,
+      "grad_norm": 2.6021623611450195,
+      "learning_rate": 5.881683731513083e-05,
+      "loss": 4.8975,
+      "step": 8273
+    },
+    {
+      "epoch": 2.823890784982935,
+      "grad_norm": 2.5270769596099854,
+      "learning_rate": 5.870307167235495e-05,
+      "loss": 6.0327,
+      "step": 8274
+    },
+    {
+      "epoch": 2.824232081911263,
+      "grad_norm": 2.6603550910949707,
+      "learning_rate": 5.858930602957907e-05,
+      "loss": 6.38,
+      "step": 8275
+    },
+    {
+      "epoch": 2.8245733788395904,
+      "grad_norm": 2.523515462875366,
+      "learning_rate": 5.847554038680319e-05,
+      "loss": 5.7957,
+      "step": 8276
+    },
+    {
+      "epoch": 2.824914675767918,
+      "grad_norm": 2.602630615234375,
+      "learning_rate": 5.8361774744027307e-05,
+      "loss": 6.3797,
+      "step": 8277
+    },
+    {
+      "epoch": 2.8252559726962456,
+      "grad_norm": 3.1169798374176025,
+      "learning_rate": 5.824800910125142e-05,
+      "loss": 4.2113,
+      "step": 8278
+    },
+    {
+      "epoch": 2.8255972696245735,
+      "grad_norm": 2.580352306365967,
+      "learning_rate": 5.813424345847554e-05,
+      "loss": 6.75,
+      "step": 8279
+    },
+    {
+      "epoch": 2.825938566552901,
+      "grad_norm": 2.5643091201782227,
+      "learning_rate": 5.802047781569966e-05,
+      "loss": 5.7626,
+      "step": 8280
+    },
+    {
+      "epoch": 2.8262798634812287,
+      "grad_norm": 2.5562422275543213,
+      "learning_rate": 5.790671217292378e-05,
+      "loss": 6.2436,
+      "step": 8281
+    },
+    {
+      "epoch": 2.826621160409556,
+      "grad_norm": 2.547008514404297,
+      "learning_rate": 5.77929465301479e-05,
+      "loss": 6.1454,
+      "step": 8282
+    },
+    {
+      "epoch": 2.826962457337884,
+      "grad_norm": 2.544355869293213,
+      "learning_rate": 5.767918088737201e-05,
+      "loss": 6.3044,
+      "step": 8283
+    },
+    {
+      "epoch": 2.827303754266212,
+      "grad_norm": 2.6169373989105225,
+      "learning_rate": 5.756541524459613e-05,
+      "loss": 5.7332,
+      "step": 8284
+    },
+    {
+      "epoch": 2.8276450511945392,
+      "grad_norm": 2.581423759460449,
+      "learning_rate": 5.745164960182025e-05,
+      "loss": 5.234,
+      "step": 8285
+    },
+    {
+      "epoch": 2.8279863481228666,
+      "grad_norm": 2.5888426303863525,
+      "learning_rate": 5.733788395904437e-05,
+      "loss": 6.8817,
+      "step": 8286
+    },
+    {
+      "epoch": 2.8283276450511945,
+      "grad_norm": 2.491856098175049,
+      "learning_rate": 5.722411831626849e-05,
+      "loss": 5.7124,
+      "step": 8287
+    },
+    {
+      "epoch": 2.8286689419795223,
+      "grad_norm": 2.546119213104248,
+      "learning_rate": 5.71103526734926e-05,
+      "loss": 5.9523,
+      "step": 8288
+    },
+    {
+      "epoch": 2.8290102389078498,
+      "grad_norm": 2.61152720451355,
+      "learning_rate": 5.699658703071672e-05,
+      "loss": 5.9196,
+      "step": 8289
+    },
+    {
+      "epoch": 2.8293515358361776,
+      "grad_norm": 2.7438294887542725,
+      "learning_rate": 5.688282138794084e-05,
+      "loss": 5.4704,
+      "step": 8290
+    },
+    {
+      "epoch": 2.829692832764505,
+      "grad_norm": 2.6315395832061768,
+      "learning_rate": 5.6769055745164965e-05,
+      "loss": 5.323,
+      "step": 8291
+    },
+    {
+      "epoch": 2.830034129692833,
+      "grad_norm": 2.675381660461426,
+      "learning_rate": 5.6655290102389084e-05,
+      "loss": 5.9233,
+      "step": 8292
+    },
+    {
+      "epoch": 2.8303754266211603,
+      "grad_norm": 2.479240894317627,
+      "learning_rate": 5.6541524459613196e-05,
+      "loss": 5.9549,
+      "step": 8293
+    },
+    {
+      "epoch": 2.830716723549488,
+      "grad_norm": 2.5712265968322754,
+      "learning_rate": 5.6427758816837314e-05,
+      "loss": 6.1167,
+      "step": 8294
+    },
+    {
+      "epoch": 2.8310580204778155,
+      "grad_norm": 2.5504567623138428,
+      "learning_rate": 5.631399317406143e-05,
+      "loss": 6.1586,
+      "step": 8295
+    },
+    {
+      "epoch": 2.8313993174061434,
+      "grad_norm": 2.505173444747925,
+      "learning_rate": 5.620022753128555e-05,
+      "loss": 6.0159,
+      "step": 8296
+    },
+    {
+      "epoch": 2.831740614334471,
+      "grad_norm": 2.5844502449035645,
+      "learning_rate": 5.6086461888509676e-05,
+      "loss": 6.1771,
+      "step": 8297
+    },
+    {
+      "epoch": 2.8320819112627986,
+      "grad_norm": 2.5144166946411133,
+      "learning_rate": 5.597269624573379e-05,
+      "loss": 6.1626,
+      "step": 8298
+    },
+    {
+      "epoch": 2.832423208191126,
+      "grad_norm": 2.5902957916259766,
+      "learning_rate": 5.585893060295791e-05,
+      "loss": 5.6985,
+      "step": 8299
+    },
+    {
+      "epoch": 2.832764505119454,
+      "grad_norm": 2.809479236602783,
+      "learning_rate": 5.5745164960182025e-05,
+      "loss": 4.2213,
+      "step": 8300
+    },
+    {
+      "epoch": 2.8331058020477817,
+      "grad_norm": 2.563436269760132,
+      "learning_rate": 5.5631399317406144e-05,
+      "loss": 5.617,
+      "step": 8301
+    },
+    {
+      "epoch": 2.833447098976109,
+      "grad_norm": 2.593966484069824,
+      "learning_rate": 5.551763367463027e-05,
+      "loss": 5.7472,
+      "step": 8302
+    },
+    {
+      "epoch": 2.833788395904437,
+      "grad_norm": 2.590921640396118,
+      "learning_rate": 5.540386803185438e-05,
+      "loss": 5.1288,
+      "step": 8303
+    },
+    {
+      "epoch": 2.8341296928327644,
+      "grad_norm": 2.558959722518921,
+      "learning_rate": 5.52901023890785e-05,
+      "loss": 5.8822,
+      "step": 8304
+    },
+    {
+      "epoch": 2.8344709897610922,
+      "grad_norm": 2.5313167572021484,
+      "learning_rate": 5.517633674630262e-05,
+      "loss": 6.0652,
+      "step": 8305
+    },
+    {
+      "epoch": 2.8348122866894196,
+      "grad_norm": 2.59818172454834,
+      "learning_rate": 5.5062571103526736e-05,
+      "loss": 5.8414,
+      "step": 8306
+    },
+    {
+      "epoch": 2.8351535836177475,
+      "grad_norm": 2.5654821395874023,
+      "learning_rate": 5.4948805460750855e-05,
+      "loss": 5.9746,
+      "step": 8307
+    },
+    {
+      "epoch": 2.835494880546075,
+      "grad_norm": 2.573014974594116,
+      "learning_rate": 5.483503981797497e-05,
+      "loss": 6.3835,
+      "step": 8308
+    },
+    {
+      "epoch": 2.8358361774744028,
+      "grad_norm": 2.426826238632202,
+      "learning_rate": 5.472127417519909e-05,
+      "loss": 5.2187,
+      "step": 8309
+    },
+    {
+      "epoch": 2.8361774744027306,
+      "grad_norm": 2.524761199951172,
+      "learning_rate": 5.460750853242321e-05,
+      "loss": 6.1975,
+      "step": 8310
+    },
+    {
+      "epoch": 2.836518771331058,
+      "grad_norm": 2.5690839290618896,
+      "learning_rate": 5.449374288964733e-05,
+      "loss": 5.8855,
+      "step": 8311
+    },
+    {
+      "epoch": 2.8368600682593854,
+      "grad_norm": 2.8350605964660645,
+      "learning_rate": 5.437997724687145e-05,
+      "loss": 4.5688,
+      "step": 8312
+    },
+    {
+      "epoch": 2.8372013651877133,
+      "grad_norm": 2.7381675243377686,
+      "learning_rate": 5.426621160409556e-05,
+      "loss": 5.0834,
+      "step": 8313
+    },
+    {
+      "epoch": 2.837542662116041,
+      "grad_norm": 2.534423828125,
+      "learning_rate": 5.4152445961319684e-05,
+      "loss": 6.1567,
+      "step": 8314
+    },
+    {
+      "epoch": 2.8378839590443685,
+      "grad_norm": 2.4711785316467285,
+      "learning_rate": 5.40386803185438e-05,
+      "loss": 5.8467,
+      "step": 8315
+    },
+    {
+      "epoch": 2.8382252559726964,
+      "grad_norm": 2.48486328125,
+      "learning_rate": 5.392491467576792e-05,
+      "loss": 5.5474,
+      "step": 8316
+    },
+    {
+      "epoch": 2.8385665529010238,
+      "grad_norm": 2.6153039932250977,
+      "learning_rate": 5.381114903299204e-05,
+      "loss": 6.1734,
+      "step": 8317
+    },
+    {
+      "epoch": 2.8389078498293516,
+      "grad_norm": 2.5097267627716064,
+      "learning_rate": 5.369738339021615e-05,
+      "loss": 5.9648,
+      "step": 8318
+    },
+    {
+      "epoch": 2.839249146757679,
+      "grad_norm": 2.603243827819824,
+      "learning_rate": 5.3583617747440277e-05,
+      "loss": 5.7459,
+      "step": 8319
+    },
+    {
+      "epoch": 2.839590443686007,
+      "grad_norm": 2.543027877807617,
+      "learning_rate": 5.3469852104664395e-05,
+      "loss": 6.1973,
+      "step": 8320
+    },
+    {
+      "epoch": 2.8399317406143343,
+      "grad_norm": 2.5744729042053223,
+      "learning_rate": 5.3356086461888514e-05,
+      "loss": 6.2723,
+      "step": 8321
+    },
+    {
+      "epoch": 2.840273037542662,
+      "grad_norm": 2.5325405597686768,
+      "learning_rate": 5.324232081911263e-05,
+      "loss": 6.2439,
+      "step": 8322
+    },
+    {
+      "epoch": 2.84061433447099,
+      "grad_norm": 2.521803617477417,
+      "learning_rate": 5.3128555176336744e-05,
+      "loss": 4.9205,
+      "step": 8323
+    },
+    {
+      "epoch": 2.8409556313993174,
+      "grad_norm": 2.6305558681488037,
+      "learning_rate": 5.301478953356086e-05,
+      "loss": 6.7376,
+      "step": 8324
+    },
+    {
+      "epoch": 2.841296928327645,
+      "grad_norm": 2.589456558227539,
+      "learning_rate": 5.290102389078499e-05,
+      "loss": 5.3846,
+      "step": 8325
+    },
+    {
+      "epoch": 2.8416382252559726,
+      "grad_norm": 2.5877840518951416,
+      "learning_rate": 5.2787258248009106e-05,
+      "loss": 5.3142,
+      "step": 8326
+    },
+    {
+      "epoch": 2.8419795221843005,
+      "grad_norm": 2.562431573867798,
+      "learning_rate": 5.267349260523322e-05,
+      "loss": 5.6504,
+      "step": 8327
+    },
+    {
+      "epoch": 2.842320819112628,
+      "grad_norm": 2.539492607116699,
+      "learning_rate": 5.2559726962457336e-05,
+      "loss": 4.5076,
+      "step": 8328
+    },
+    {
+      "epoch": 2.8426621160409558,
+      "grad_norm": 2.5125441551208496,
+      "learning_rate": 5.2445961319681455e-05,
+      "loss": 6.1126,
+      "step": 8329
+    },
+    {
+      "epoch": 2.843003412969283,
+      "grad_norm": 2.546273708343506,
+      "learning_rate": 5.233219567690557e-05,
+      "loss": 5.7161,
+      "step": 8330
+    },
+    {
+      "epoch": 2.843344709897611,
+      "grad_norm": 2.596364736557007,
+      "learning_rate": 5.22184300341297e-05,
+      "loss": 6.1584,
+      "step": 8331
+    },
+    {
+      "epoch": 2.8436860068259384,
+      "grad_norm": 2.556121349334717,
+      "learning_rate": 5.210466439135381e-05,
+      "loss": 5.4851,
+      "step": 8332
+    },
+    {
+      "epoch": 2.8440273037542663,
+      "grad_norm": 2.4971086978912354,
+      "learning_rate": 5.199089874857793e-05,
+      "loss": 6.1228,
+      "step": 8333
+    },
+    {
+      "epoch": 2.8443686006825937,
+      "grad_norm": 2.5893213748931885,
+      "learning_rate": 5.187713310580205e-05,
+      "loss": 4.826,
+      "step": 8334
+    },
+    {
+      "epoch": 2.8447098976109215,
+      "grad_norm": 2.562465190887451,
+      "learning_rate": 5.1763367463026166e-05,
+      "loss": 4.6726,
+      "step": 8335
+    },
+    {
+      "epoch": 2.8450511945392494,
+      "grad_norm": 2.575310707092285,
+      "learning_rate": 5.164960182025029e-05,
+      "loss": 6.3681,
+      "step": 8336
+    },
+    {
+      "epoch": 2.845392491467577,
+      "grad_norm": 3.0285212993621826,
+      "learning_rate": 5.15358361774744e-05,
+      "loss": 5.0006,
+      "step": 8337
+    },
+    {
+      "epoch": 2.845733788395904,
+      "grad_norm": 2.7478995323181152,
+      "learning_rate": 5.142207053469852e-05,
+      "loss": 5.4962,
+      "step": 8338
+    },
+    {
+      "epoch": 2.846075085324232,
+      "grad_norm": 2.4442386627197266,
+      "learning_rate": 5.130830489192264e-05,
+      "loss": 5.7498,
+      "step": 8339
+    },
+    {
+      "epoch": 2.84641638225256,
+      "grad_norm": 2.5991361141204834,
+      "learning_rate": 5.119453924914676e-05,
+      "loss": 5.419,
+      "step": 8340
+    },
+    {
+      "epoch": 2.8467576791808873,
+      "grad_norm": 2.547792434692383,
+      "learning_rate": 5.108077360637088e-05,
+      "loss": 5.9812,
+      "step": 8341
+    },
+    {
+      "epoch": 2.847098976109215,
+      "grad_norm": 2.6672022342681885,
+      "learning_rate": 5.0967007963594995e-05,
+      "loss": 4.3484,
+      "step": 8342
+    },
+    {
+      "epoch": 2.8474402730375425,
+      "grad_norm": 2.547494649887085,
+      "learning_rate": 5.0853242320819114e-05,
+      "loss": 6.2943,
+      "step": 8343
+    },
+    {
+      "epoch": 2.8477815699658704,
+      "grad_norm": 2.4189071655273438,
+      "learning_rate": 5.073947667804323e-05,
+      "loss": 5.0494,
+      "step": 8344
+    },
+    {
+      "epoch": 2.848122866894198,
+      "grad_norm": 2.471524477005005,
+      "learning_rate": 5.062571103526735e-05,
+      "loss": 5.5994,
+      "step": 8345
+    },
+    {
+      "epoch": 2.8484641638225257,
+      "grad_norm": 2.490009069442749,
+      "learning_rate": 5.051194539249147e-05,
+      "loss": 5.8388,
+      "step": 8346
+    },
+    {
+      "epoch": 2.848805460750853,
+      "grad_norm": 2.542736053466797,
+      "learning_rate": 5.039817974971558e-05,
+      "loss": 5.7709,
+      "step": 8347
+    },
+    {
+      "epoch": 2.849146757679181,
+      "grad_norm": 2.5808589458465576,
+      "learning_rate": 5.0284414106939706e-05,
+      "loss": 6.7117,
+      "step": 8348
+    },
+    {
+      "epoch": 2.8494880546075088,
+      "grad_norm": 2.597972869873047,
+      "learning_rate": 5.0170648464163825e-05,
+      "loss": 6.4791,
+      "step": 8349
+    },
+    {
+      "epoch": 2.849829351535836,
+      "grad_norm": 2.577030658721924,
+      "learning_rate": 5.005688282138794e-05,
+      "loss": 6.3102,
+      "step": 8350
+    },
+    {
+      "epoch": 2.8501706484641636,
+      "grad_norm": 2.47444486618042,
+      "learning_rate": 4.994311717861206e-05,
+      "loss": 5.5967,
+      "step": 8351
+    },
+    {
+      "epoch": 2.8505119453924914,
+      "grad_norm": 2.5281221866607666,
+      "learning_rate": 4.982935153583617e-05,
+      "loss": 5.4821,
+      "step": 8352
+    },
+    {
+      "epoch": 2.8508532423208193,
+      "grad_norm": 2.5494225025177,
+      "learning_rate": 4.97155858930603e-05,
+      "loss": 6.4092,
+      "step": 8353
+    },
+    {
+      "epoch": 2.8511945392491467,
+      "grad_norm": 2.561234712600708,
+      "learning_rate": 4.960182025028442e-05,
+      "loss": 5.7412,
+      "step": 8354
+    },
+    {
+      "epoch": 2.8515358361774745,
+      "grad_norm": 3.5011022090911865,
+      "learning_rate": 4.9488054607508536e-05,
+      "loss": 5.4677,
+      "step": 8355
+    },
+    {
+      "epoch": 2.851877133105802,
+      "grad_norm": 2.60447359085083,
+      "learning_rate": 4.9374288964732654e-05,
+      "loss": 6.2699,
+      "step": 8356
+    },
+    {
+      "epoch": 2.85221843003413,
+      "grad_norm": 2.556480884552002,
+      "learning_rate": 4.9260523321956766e-05,
+      "loss": 5.9936,
+      "step": 8357
+    },
+    {
+      "epoch": 2.852559726962457,
+      "grad_norm": 2.508237600326538,
+      "learning_rate": 4.9146757679180884e-05,
+      "loss": 5.9199,
+      "step": 8358
+    },
+    {
+      "epoch": 2.852901023890785,
+      "grad_norm": 2.5938949584960938,
+      "learning_rate": 4.903299203640501e-05,
+      "loss": 6.2548,
+      "step": 8359
+    },
+    {
+      "epoch": 2.8532423208191124,
+      "grad_norm": 2.570672035217285,
+      "learning_rate": 4.891922639362913e-05,
+      "loss": 6.0058,
+      "step": 8360
+    },
+    {
+      "epoch": 2.8535836177474403,
+      "grad_norm": 2.583148717880249,
+      "learning_rate": 4.8805460750853247e-05,
+      "loss": 6.0509,
+      "step": 8361
+    },
+    {
+      "epoch": 2.853924914675768,
+      "grad_norm": 2.4389407634735107,
+      "learning_rate": 4.869169510807736e-05,
+      "loss": 6.2846,
+      "step": 8362
+    },
+    {
+      "epoch": 2.8542662116040955,
+      "grad_norm": 2.574002981185913,
+      "learning_rate": 4.857792946530148e-05,
+      "loss": 5.7958,
+      "step": 8363
+    },
+    {
+      "epoch": 2.854607508532423,
+      "grad_norm": 2.5079104900360107,
+      "learning_rate": 4.8464163822525595e-05,
+      "loss": 6.4898,
+      "step": 8364
+    },
+    {
+      "epoch": 2.854948805460751,
+      "grad_norm": 2.6473233699798584,
+      "learning_rate": 4.835039817974972e-05,
+      "loss": 5.9305,
+      "step": 8365
+    },
+    {
+      "epoch": 2.8552901023890787,
+      "grad_norm": 2.551525831222534,
+      "learning_rate": 4.823663253697384e-05,
+      "loss": 5.7632,
+      "step": 8366
+    },
+    {
+      "epoch": 2.855631399317406,
+      "grad_norm": 2.5253288745880127,
+      "learning_rate": 4.812286689419795e-05,
+      "loss": 6.1019,
+      "step": 8367
+    },
+    {
+      "epoch": 2.855972696245734,
+      "grad_norm": 2.598867654800415,
+      "learning_rate": 4.800910125142207e-05,
+      "loss": 6.0308,
+      "step": 8368
+    },
+    {
+      "epoch": 2.8563139931740613,
+      "grad_norm": 2.4677653312683105,
+      "learning_rate": 4.789533560864619e-05,
+      "loss": 5.5467,
+      "step": 8369
+    },
+    {
+      "epoch": 2.856655290102389,
+      "grad_norm": 2.6016745567321777,
+      "learning_rate": 4.778156996587031e-05,
+      "loss": 5.6461,
+      "step": 8370
+    },
+    {
+      "epoch": 2.8569965870307166,
+      "grad_norm": 2.613708734512329,
+      "learning_rate": 4.766780432309443e-05,
+      "loss": 6.6418,
+      "step": 8371
+    },
+    {
+      "epoch": 2.8573378839590444,
+      "grad_norm": 2.5926055908203125,
+      "learning_rate": 4.755403868031854e-05,
+      "loss": 5.8853,
+      "step": 8372
+    },
+    {
+      "epoch": 2.857679180887372,
+      "grad_norm": 2.541905403137207,
+      "learning_rate": 4.744027303754266e-05,
+      "loss": 6.4039,
+      "step": 8373
+    },
+    {
+      "epoch": 2.8580204778156997,
+      "grad_norm": 2.5121161937713623,
+      "learning_rate": 4.732650739476678e-05,
+      "loss": 5.7596,
+      "step": 8374
+    },
+    {
+      "epoch": 2.8583617747440275,
+      "grad_norm": 2.536179780960083,
+      "learning_rate": 4.72127417519909e-05,
+      "loss": 6.1308,
+      "step": 8375
+    },
+    {
+      "epoch": 2.858703071672355,
+      "grad_norm": 7.43336296081543,
+      "learning_rate": 4.7098976109215024e-05,
+      "loss": 5.0863,
+      "step": 8376
+    },
+    {
+      "epoch": 2.8590443686006823,
+      "grad_norm": 2.5915966033935547,
+      "learning_rate": 4.6985210466439136e-05,
+      "loss": 6.191,
+      "step": 8377
+    },
+    {
+      "epoch": 2.85938566552901,
+      "grad_norm": 2.5173799991607666,
+      "learning_rate": 4.6871444823663254e-05,
+      "loss": 6.0567,
+      "step": 8378
+    },
+    {
+      "epoch": 2.859726962457338,
+      "grad_norm": 2.5869557857513428,
+      "learning_rate": 4.675767918088737e-05,
+      "loss": 5.7846,
+      "step": 8379
+    },
+    {
+      "epoch": 2.8600682593856654,
+      "grad_norm": 2.5171236991882324,
+      "learning_rate": 4.664391353811149e-05,
+      "loss": 6.236,
+      "step": 8380
+    },
+    {
+      "epoch": 2.8604095563139933,
+      "grad_norm": 2.6231682300567627,
+      "learning_rate": 4.653014789533561e-05,
+      "loss": 5.6616,
+      "step": 8381
+    },
+    {
+      "epoch": 2.8607508532423207,
+      "grad_norm": 2.555541753768921,
+      "learning_rate": 4.641638225255973e-05,
+      "loss": 6.3776,
+      "step": 8382
+    },
+    {
+      "epoch": 2.8610921501706486,
+      "grad_norm": 2.5504350662231445,
+      "learning_rate": 4.630261660978385e-05,
+      "loss": 6.1276,
+      "step": 8383
+    },
+    {
+      "epoch": 2.861433447098976,
+      "grad_norm": 2.523817300796509,
+      "learning_rate": 4.6188850967007965e-05,
+      "loss": 5.6059,
+      "step": 8384
+    },
+    {
+      "epoch": 2.861774744027304,
+      "grad_norm": 2.5135388374328613,
+      "learning_rate": 4.6075085324232084e-05,
+      "loss": 5.924,
+      "step": 8385
+    },
+    {
+      "epoch": 2.862116040955631,
+      "grad_norm": 2.5756356716156006,
+      "learning_rate": 4.5961319681456195e-05,
+      "loss": 5.9326,
+      "step": 8386
+    },
+    {
+      "epoch": 2.862457337883959,
+      "grad_norm": 2.5897839069366455,
+      "learning_rate": 4.584755403868032e-05,
+      "loss": 6.2464,
+      "step": 8387
+    },
+    {
+      "epoch": 2.862798634812287,
+      "grad_norm": 2.522111177444458,
+      "learning_rate": 4.573378839590444e-05,
+      "loss": 5.7213,
+      "step": 8388
+    },
+    {
+      "epoch": 2.8631399317406143,
+      "grad_norm": 2.5438270568847656,
+      "learning_rate": 4.562002275312856e-05,
+      "loss": 5.5916,
+      "step": 8389
+    },
+    {
+      "epoch": 2.8634812286689417,
+      "grad_norm": 2.5333592891693115,
+      "learning_rate": 4.5506257110352676e-05,
+      "loss": 6.1719,
+      "step": 8390
+    },
+    {
+      "epoch": 2.8638225255972696,
+      "grad_norm": 2.5089774131774902,
+      "learning_rate": 4.539249146757679e-05,
+      "loss": 6.1262,
+      "step": 8391
+    },
+    {
+      "epoch": 2.8641638225255974,
+      "grad_norm": 2.4316928386688232,
+      "learning_rate": 4.5278725824800906e-05,
+      "loss": 4.518,
+      "step": 8392
+    },
+    {
+      "epoch": 2.864505119453925,
+      "grad_norm": 2.5709829330444336,
+      "learning_rate": 4.516496018202503e-05,
+      "loss": 6.1676,
+      "step": 8393
+    },
+    {
+      "epoch": 2.8648464163822527,
+      "grad_norm": 2.5119333267211914,
+      "learning_rate": 4.505119453924915e-05,
+      "loss": 4.8229,
+      "step": 8394
+    },
+    {
+      "epoch": 2.86518771331058,
+      "grad_norm": 2.5550503730773926,
+      "learning_rate": 4.493742889647327e-05,
+      "loss": 5.6363,
+      "step": 8395
+    },
+    {
+      "epoch": 2.865529010238908,
+      "grad_norm": 2.5302844047546387,
+      "learning_rate": 4.482366325369738e-05,
+      "loss": 5.3337,
+      "step": 8396
+    },
+    {
+      "epoch": 2.8658703071672353,
+      "grad_norm": 2.5047590732574463,
+      "learning_rate": 4.47098976109215e-05,
+      "loss": 6.5914,
+      "step": 8397
+    },
+    {
+      "epoch": 2.866211604095563,
+      "grad_norm": 2.5279815196990967,
+      "learning_rate": 4.4596131968145624e-05,
+      "loss": 6.4273,
+      "step": 8398
+    },
+    {
+      "epoch": 2.8665529010238906,
+      "grad_norm": 2.603135347366333,
+      "learning_rate": 4.448236632536974e-05,
+      "loss": 6.5059,
+      "step": 8399
+    },
+    {
+      "epoch": 2.8668941979522184,
+      "grad_norm": 2.582097053527832,
+      "learning_rate": 4.436860068259386e-05,
+      "loss": 5.8008,
+      "step": 8400
+    },
+    {
+      "epoch": 2.8672354948805463,
+      "grad_norm": 2.5195491313934326,
+      "learning_rate": 4.425483503981797e-05,
+      "loss": 6.622,
+      "step": 8401
+    },
+    {
+      "epoch": 2.8675767918088737,
+      "grad_norm": 2.5339550971984863,
+      "learning_rate": 4.414106939704209e-05,
+      "loss": 6.4584,
+      "step": 8402
+    },
+    {
+      "epoch": 2.867918088737201,
+      "grad_norm": 2.524705648422241,
+      "learning_rate": 4.402730375426621e-05,
+      "loss": 5.2988,
+      "step": 8403
+    },
+    {
+      "epoch": 2.868259385665529,
+      "grad_norm": 2.5974693298339844,
+      "learning_rate": 4.3913538111490335e-05,
+      "loss": 6.255,
+      "step": 8404
+    },
+    {
+      "epoch": 2.868600682593857,
+      "grad_norm": 2.66146183013916,
+      "learning_rate": 4.3799772468714454e-05,
+      "loss": 4.7422,
+      "step": 8405
+    },
+    {
+      "epoch": 2.868941979522184,
+      "grad_norm": 2.778348922729492,
+      "learning_rate": 4.3686006825938565e-05,
+      "loss": 5.1854,
+      "step": 8406
+    },
+    {
+      "epoch": 2.869283276450512,
+      "grad_norm": 2.8919944763183594,
+      "learning_rate": 4.3572241183162684e-05,
+      "loss": 3.8649,
+      "step": 8407
+    },
+    {
+      "epoch": 2.8696245733788395,
+      "grad_norm": 2.5984740257263184,
+      "learning_rate": 4.34584755403868e-05,
+      "loss": 5.6027,
+      "step": 8408
+    },
+    {
+      "epoch": 2.8699658703071673,
+      "grad_norm": 2.449209690093994,
+      "learning_rate": 4.334470989761092e-05,
+      "loss": 5.0734,
+      "step": 8409
+    },
+    {
+      "epoch": 2.8703071672354947,
+      "grad_norm": 2.4918715953826904,
+      "learning_rate": 4.3230944254835046e-05,
+      "loss": 5.4387,
+      "step": 8410
+    },
+    {
+      "epoch": 2.8706484641638226,
+      "grad_norm": 2.528923749923706,
+      "learning_rate": 4.311717861205916e-05,
+      "loss": 5.9894,
+      "step": 8411
+    },
+    {
+      "epoch": 2.87098976109215,
+      "grad_norm": 2.5853374004364014,
+      "learning_rate": 4.3003412969283276e-05,
+      "loss": 6.5989,
+      "step": 8412
+    },
+    {
+      "epoch": 2.871331058020478,
+      "grad_norm": 2.531834363937378,
+      "learning_rate": 4.2889647326507395e-05,
+      "loss": 6.2346,
+      "step": 8413
+    },
+    {
+      "epoch": 2.8716723549488057,
+      "grad_norm": 2.5188536643981934,
+      "learning_rate": 4.277588168373151e-05,
+      "loss": 5.9237,
+      "step": 8414
+    },
+    {
+      "epoch": 2.872013651877133,
+      "grad_norm": 2.5645134449005127,
+      "learning_rate": 4.266211604095564e-05,
+      "loss": 5.9186,
+      "step": 8415
+    },
+    {
+      "epoch": 2.8723549488054605,
+      "grad_norm": 2.4560606479644775,
+      "learning_rate": 4.254835039817975e-05,
+      "loss": 5.6281,
+      "step": 8416
+    },
+    {
+      "epoch": 2.8726962457337883,
+      "grad_norm": 2.49910306930542,
+      "learning_rate": 4.243458475540387e-05,
+      "loss": 6.1462,
+      "step": 8417
+    },
+    {
+      "epoch": 2.873037542662116,
+      "grad_norm": 2.547170877456665,
+      "learning_rate": 4.232081911262799e-05,
+      "loss": 5.9647,
+      "step": 8418
+    },
+    {
+      "epoch": 2.8733788395904436,
+      "grad_norm": 2.5908734798431396,
+      "learning_rate": 4.2207053469852106e-05,
+      "loss": 5.909,
+      "step": 8419
+    },
+    {
+      "epoch": 2.8737201365187715,
+      "grad_norm": 2.5554044246673584,
+      "learning_rate": 4.2093287827076224e-05,
+      "loss": 5.9551,
+      "step": 8420
+    },
+    {
+      "epoch": 2.874061433447099,
+      "grad_norm": 2.5563385486602783,
+      "learning_rate": 4.197952218430034e-05,
+      "loss": 6.0185,
+      "step": 8421
+    },
+    {
+      "epoch": 2.8744027303754267,
+      "grad_norm": 2.525693893432617,
+      "learning_rate": 4.186575654152446e-05,
+      "loss": 5.7964,
+      "step": 8422
+    },
+    {
+      "epoch": 2.874744027303754,
+      "grad_norm": 2.5117642879486084,
+      "learning_rate": 4.175199089874858e-05,
+      "loss": 5.5265,
+      "step": 8423
+    },
+    {
+      "epoch": 2.875085324232082,
+      "grad_norm": 2.5649948120117188,
+      "learning_rate": 4.16382252559727e-05,
+      "loss": 5.9956,
+      "step": 8424
+    },
+    {
+      "epoch": 2.8754266211604094,
+      "grad_norm": 2.43780255317688,
+      "learning_rate": 4.152445961319681e-05,
+      "loss": 6.1702,
+      "step": 8425
+    },
+    {
+      "epoch": 2.875767918088737,
+      "grad_norm": 2.5544321537017822,
+      "learning_rate": 4.141069397042093e-05,
+      "loss": 5.5664,
+      "step": 8426
+    },
+    {
+      "epoch": 2.876109215017065,
+      "grad_norm": 2.5855114459991455,
+      "learning_rate": 4.1296928327645054e-05,
+      "loss": 6.168,
+      "step": 8427
+    },
+    {
+      "epoch": 2.8764505119453925,
+      "grad_norm": 2.5726702213287354,
+      "learning_rate": 4.118316268486917e-05,
+      "loss": 5.963,
+      "step": 8428
+    },
+    {
+      "epoch": 2.87679180887372,
+      "grad_norm": 2.6072628498077393,
+      "learning_rate": 4.106939704209329e-05,
+      "loss": 6.4451,
+      "step": 8429
+    },
+    {
+      "epoch": 2.8771331058020477,
+      "grad_norm": 2.586865186691284,
+      "learning_rate": 4.09556313993174e-05,
+      "loss": 6.5633,
+      "step": 8430
+    },
+    {
+      "epoch": 2.8774744027303756,
+      "grad_norm": 2.524959087371826,
+      "learning_rate": 4.084186575654152e-05,
+      "loss": 6.5356,
+      "step": 8431
+    },
+    {
+      "epoch": 2.877815699658703,
+      "grad_norm": 2.5504794120788574,
+      "learning_rate": 4.0728100113765646e-05,
+      "loss": 5.9479,
+      "step": 8432
+    },
+    {
+      "epoch": 2.878156996587031,
+      "grad_norm": 2.483062505722046,
+      "learning_rate": 4.0614334470989765e-05,
+      "loss": 6.0917,
+      "step": 8433
+    },
+    {
+      "epoch": 2.8784982935153582,
+      "grad_norm": 2.506287097930908,
+      "learning_rate": 4.050056882821388e-05,
+      "loss": 6.0076,
+      "step": 8434
+    },
+    {
+      "epoch": 2.878839590443686,
+      "grad_norm": 2.6461734771728516,
+      "learning_rate": 4.0386803185437995e-05,
+      "loss": 5.6557,
+      "step": 8435
+    },
+    {
+      "epoch": 2.8791808873720135,
+      "grad_norm": 2.628688335418701,
+      "learning_rate": 4.0273037542662113e-05,
+      "loss": 6.4374,
+      "step": 8436
+    },
+    {
+      "epoch": 2.8795221843003413,
+      "grad_norm": 2.541585683822632,
+      "learning_rate": 4.015927189988623e-05,
+      "loss": 6.1847,
+      "step": 8437
+    },
+    {
+      "epoch": 2.8798634812286688,
+      "grad_norm": 2.503748655319214,
+      "learning_rate": 4.004550625711036e-05,
+      "loss": 5.7091,
+      "step": 8438
+    },
+    {
+      "epoch": 2.8802047781569966,
+      "grad_norm": 2.540815591812134,
+      "learning_rate": 3.9931740614334476e-05,
+      "loss": 5.6816,
+      "step": 8439
+    },
+    {
+      "epoch": 2.8805460750853245,
+      "grad_norm": 2.56005597114563,
+      "learning_rate": 3.981797497155859e-05,
+      "loss": 5.871,
+      "step": 8440
+    },
+    {
+      "epoch": 2.880887372013652,
+      "grad_norm": 2.4481024742126465,
+      "learning_rate": 3.9704209328782706e-05,
+      "loss": 5.8886,
+      "step": 8441
+    },
+    {
+      "epoch": 2.8812286689419793,
+      "grad_norm": 2.5519607067108154,
+      "learning_rate": 3.9590443686006824e-05,
+      "loss": 5.7186,
+      "step": 8442
+    },
+    {
+      "epoch": 2.881569965870307,
+      "grad_norm": 2.5279147624969482,
+      "learning_rate": 3.947667804323094e-05,
+      "loss": 5.7121,
+      "step": 8443
+    },
+    {
+      "epoch": 2.881911262798635,
+      "grad_norm": 2.5419199466705322,
+      "learning_rate": 3.936291240045507e-05,
+      "loss": 6.1733,
+      "step": 8444
+    },
+    {
+      "epoch": 2.8822525597269624,
+      "grad_norm": 2.4785492420196533,
+      "learning_rate": 3.924914675767918e-05,
+      "loss": 5.7262,
+      "step": 8445
+    },
+    {
+      "epoch": 2.88259385665529,
+      "grad_norm": 2.5293445587158203,
+      "learning_rate": 3.91353811149033e-05,
+      "loss": 6.0911,
+      "step": 8446
+    },
+    {
+      "epoch": 2.8829351535836176,
+      "grad_norm": 2.6208319664001465,
+      "learning_rate": 3.902161547212742e-05,
+      "loss": 6.4624,
+      "step": 8447
+    },
+    {
+      "epoch": 2.8832764505119455,
+      "grad_norm": 2.5119760036468506,
+      "learning_rate": 3.8907849829351535e-05,
+      "loss": 5.5541,
+      "step": 8448
+    },
+    {
+      "epoch": 2.883617747440273,
+      "grad_norm": 2.55000638961792,
+      "learning_rate": 3.879408418657566e-05,
+      "loss": 6.5788,
+      "step": 8449
+    },
+    {
+      "epoch": 2.8839590443686007,
+      "grad_norm": 2.545145273208618,
+      "learning_rate": 3.868031854379977e-05,
+      "loss": 6.2904,
+      "step": 8450
+    },
+    {
+      "epoch": 2.884300341296928,
+      "grad_norm": 2.5100696086883545,
+      "learning_rate": 3.856655290102389e-05,
+      "loss": 6.0244,
+      "step": 8451
+    },
+    {
+      "epoch": 2.884641638225256,
+      "grad_norm": 2.528092384338379,
+      "learning_rate": 3.845278725824801e-05,
+      "loss": 4.7265,
+      "step": 8452
+    },
+    {
+      "epoch": 2.884982935153584,
+      "grad_norm": 2.526163339614868,
+      "learning_rate": 3.833902161547213e-05,
+      "loss": 5.983,
+      "step": 8453
+    },
+    {
+      "epoch": 2.8853242320819112,
+      "grad_norm": 2.3863651752471924,
+      "learning_rate": 3.8225255972696246e-05,
+      "loss": 4.602,
+      "step": 8454
+    },
+    {
+      "epoch": 2.8856655290102387,
+      "grad_norm": 2.4391610622406006,
+      "learning_rate": 3.8111490329920365e-05,
+      "loss": 5.4737,
+      "step": 8455
+    },
+    {
+      "epoch": 2.8860068259385665,
+      "grad_norm": 2.584704875946045,
+      "learning_rate": 3.799772468714448e-05,
+      "loss": 5.8695,
+      "step": 8456
+    },
+    {
+      "epoch": 2.8863481228668944,
+      "grad_norm": 2.5321240425109863,
+      "learning_rate": 3.78839590443686e-05,
+      "loss": 5.7091,
+      "step": 8457
+    },
+    {
+      "epoch": 2.8866894197952218,
+      "grad_norm": 1.8349437713623047,
+      "learning_rate": 3.777019340159272e-05,
+      "loss": 3.3003,
+      "step": 8458
+    },
+    {
+      "epoch": 2.8870307167235496,
+      "grad_norm": 2.5653040409088135,
+      "learning_rate": 3.765642775881684e-05,
+      "loss": 6.2374,
+      "step": 8459
+    },
+    {
+      "epoch": 2.887372013651877,
+      "grad_norm": 2.696136236190796,
+      "learning_rate": 3.754266211604096e-05,
+      "loss": 5.7614,
+      "step": 8460
+    },
+    {
+      "epoch": 2.887713310580205,
+      "grad_norm": 2.485445499420166,
+      "learning_rate": 3.7428896473265076e-05,
+      "loss": 4.7444,
+      "step": 8461
+    },
+    {
+      "epoch": 2.8880546075085323,
+      "grad_norm": 2.547309398651123,
+      "learning_rate": 3.7315130830489194e-05,
+      "loss": 6.003,
+      "step": 8462
+    },
+    {
+      "epoch": 2.88839590443686,
+      "grad_norm": 2.4875457286834717,
+      "learning_rate": 3.720136518771331e-05,
+      "loss": 6.3583,
+      "step": 8463
+    },
+    {
+      "epoch": 2.8887372013651875,
+      "grad_norm": 2.5078489780426025,
+      "learning_rate": 3.708759954493743e-05,
+      "loss": 5.2605,
+      "step": 8464
+    },
+    {
+      "epoch": 2.8890784982935154,
+      "grad_norm": 2.540022373199463,
+      "learning_rate": 3.697383390216154e-05,
+      "loss": 6.3982,
+      "step": 8465
+    },
+    {
+      "epoch": 2.8894197952218432,
+      "grad_norm": 2.497128963470459,
+      "learning_rate": 3.686006825938567e-05,
+      "loss": 5.7944,
+      "step": 8466
+    },
+    {
+      "epoch": 2.8897610921501706,
+      "grad_norm": 2.550382614135742,
+      "learning_rate": 3.674630261660979e-05,
+      "loss": 6.0483,
+      "step": 8467
+    },
+    {
+      "epoch": 2.890102389078498,
+      "grad_norm": 2.4876744747161865,
+      "learning_rate": 3.6632536973833905e-05,
+      "loss": 5.8666,
+      "step": 8468
+    },
+    {
+      "epoch": 2.890443686006826,
+      "grad_norm": 2.4927735328674316,
+      "learning_rate": 3.6518771331058024e-05,
+      "loss": 5.7912,
+      "step": 8469
+    },
+    {
+      "epoch": 2.8907849829351537,
+      "grad_norm": 2.4995381832122803,
+      "learning_rate": 3.6405005688282136e-05,
+      "loss": 5.6446,
+      "step": 8470
+    },
+    {
+      "epoch": 2.891126279863481,
+      "grad_norm": 2.485089063644409,
+      "learning_rate": 3.6291240045506254e-05,
+      "loss": 6.2976,
+      "step": 8471
+    },
+    {
+      "epoch": 2.891467576791809,
+      "grad_norm": 2.451723337173462,
+      "learning_rate": 3.617747440273038e-05,
+      "loss": 5.6149,
+      "step": 8472
+    },
+    {
+      "epoch": 2.8918088737201364,
+      "grad_norm": 2.487278938293457,
+      "learning_rate": 3.60637087599545e-05,
+      "loss": 5.683,
+      "step": 8473
+    },
+    {
+      "epoch": 2.8921501706484642,
+      "grad_norm": 2.5790092945098877,
+      "learning_rate": 3.594994311717861e-05,
+      "loss": 5.1775,
+      "step": 8474
+    },
+    {
+      "epoch": 2.8924914675767917,
+      "grad_norm": 2.5560343265533447,
+      "learning_rate": 3.583617747440273e-05,
+      "loss": 5.4929,
+      "step": 8475
+    },
+    {
+      "epoch": 2.8928327645051195,
+      "grad_norm": 2.5981833934783936,
+      "learning_rate": 3.5722411831626847e-05,
+      "loss": 5.6103,
+      "step": 8476
+    },
+    {
+      "epoch": 2.893174061433447,
+      "grad_norm": 2.4823999404907227,
+      "learning_rate": 3.560864618885097e-05,
+      "loss": 5.6562,
+      "step": 8477
+    },
+    {
+      "epoch": 2.8935153583617748,
+      "grad_norm": 2.4643754959106445,
+      "learning_rate": 3.549488054607509e-05,
+      "loss": 6.0838,
+      "step": 8478
+    },
+    {
+      "epoch": 2.8938566552901026,
+      "grad_norm": 2.5785889625549316,
+      "learning_rate": 3.53811149032992e-05,
+      "loss": 5.6819,
+      "step": 8479
+    },
+    {
+      "epoch": 2.89419795221843,
+      "grad_norm": 2.451211452484131,
+      "learning_rate": 3.526734926052332e-05,
+      "loss": 5.2065,
+      "step": 8480
+    },
+    {
+      "epoch": 2.8945392491467574,
+      "grad_norm": 2.5396275520324707,
+      "learning_rate": 3.515358361774744e-05,
+      "loss": 6.2585,
+      "step": 8481
+    },
+    {
+      "epoch": 2.8948805460750853,
+      "grad_norm": 2.5079259872436523,
+      "learning_rate": 3.503981797497156e-05,
+      "loss": 5.9257,
+      "step": 8482
+    },
+    {
+      "epoch": 2.895221843003413,
+      "grad_norm": 2.489826202392578,
+      "learning_rate": 3.492605233219568e-05,
+      "loss": 5.6171,
+      "step": 8483
+    },
+    {
+      "epoch": 2.8955631399317405,
+      "grad_norm": 2.556549310684204,
+      "learning_rate": 3.4812286689419794e-05,
+      "loss": 5.8783,
+      "step": 8484
+    },
+    {
+      "epoch": 2.8959044368600684,
+      "grad_norm": 2.518484354019165,
+      "learning_rate": 3.469852104664391e-05,
+      "loss": 5.9915,
+      "step": 8485
+    },
+    {
+      "epoch": 2.896245733788396,
+      "grad_norm": 2.582871675491333,
+      "learning_rate": 3.458475540386803e-05,
+      "loss": 6.5008,
+      "step": 8486
+    },
+    {
+      "epoch": 2.8965870307167236,
+      "grad_norm": 2.5308432579040527,
+      "learning_rate": 3.447098976109215e-05,
+      "loss": 6.2562,
+      "step": 8487
+    },
+    {
+      "epoch": 2.896928327645051,
+      "grad_norm": 2.4115796089172363,
+      "learning_rate": 3.435722411831627e-05,
+      "loss": 4.786,
+      "step": 8488
+    },
+    {
+      "epoch": 2.897269624573379,
+      "grad_norm": 2.572611093521118,
+      "learning_rate": 3.424345847554039e-05,
+      "loss": 6.1011,
+      "step": 8489
+    },
+    {
+      "epoch": 2.8976109215017063,
+      "grad_norm": 2.5578956604003906,
+      "learning_rate": 3.4129692832764505e-05,
+      "loss": 6.097,
+      "step": 8490
+    },
+    {
+      "epoch": 2.897952218430034,
+      "grad_norm": 2.5361123085021973,
+      "learning_rate": 3.4015927189988624e-05,
+      "loss": 5.612,
+      "step": 8491
+    },
+    {
+      "epoch": 2.898293515358362,
+      "grad_norm": 2.522542715072632,
+      "learning_rate": 3.390216154721274e-05,
+      "loss": 6.287,
+      "step": 8492
+    },
+    {
+      "epoch": 2.8986348122866894,
+      "grad_norm": 2.521287202835083,
+      "learning_rate": 3.378839590443686e-05,
+      "loss": 6.448,
+      "step": 8493
+    },
+    {
+      "epoch": 2.898976109215017,
+      "grad_norm": 2.5632736682891846,
+      "learning_rate": 3.367463026166098e-05,
+      "loss": 6.7079,
+      "step": 8494
+    },
+    {
+      "epoch": 2.8993174061433447,
+      "grad_norm": 2.497307300567627,
+      "learning_rate": 3.35608646188851e-05,
+      "loss": 5.8383,
+      "step": 8495
+    },
+    {
+      "epoch": 2.8996587030716725,
+      "grad_norm": 2.5428879261016846,
+      "learning_rate": 3.3447098976109216e-05,
+      "loss": 6.4592,
+      "step": 8496
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 2.5488977432250977,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 5.6038,
+      "step": 8497
+    },
+    {
+      "epoch": 2.9003412969283278,
+      "grad_norm": 2.5850279331207275,
+      "learning_rate": 3.3219567690557453e-05,
+      "loss": 6.1336,
+      "step": 8498
+    },
+    {
+      "epoch": 2.900682593856655,
+      "grad_norm": 2.4947924613952637,
+      "learning_rate": 3.3105802047781565e-05,
+      "loss": 5.3009,
+      "step": 8499
+    },
+    {
+      "epoch": 2.901023890784983,
+      "grad_norm": 2.5952110290527344,
+      "learning_rate": 3.299203640500569e-05,
+      "loss": 5.9782,
+      "step": 8500
+    },
+    {
+      "epoch": 2.9013651877133104,
+      "grad_norm": 2.5056798458099365,
+      "learning_rate": 3.287827076222981e-05,
+      "loss": 5.4483,
+      "step": 8501
+    },
+    {
+      "epoch": 2.9017064846416383,
+      "grad_norm": 2.5299181938171387,
+      "learning_rate": 3.276450511945393e-05,
+      "loss": 6.3067,
+      "step": 8502
+    },
+    {
+      "epoch": 2.9020477815699657,
+      "grad_norm": 2.47225284576416,
+      "learning_rate": 3.2650739476678046e-05,
+      "loss": 5.9558,
+      "step": 8503
+    },
+    {
+      "epoch": 2.9023890784982935,
+      "grad_norm": 2.519516944885254,
+      "learning_rate": 3.253697383390216e-05,
+      "loss": 5.8317,
+      "step": 8504
+    },
+    {
+      "epoch": 2.9027303754266214,
+      "grad_norm": 2.538496255874634,
+      "learning_rate": 3.2423208191126276e-05,
+      "loss": 6.2876,
+      "step": 8505
+    },
+    {
+      "epoch": 2.903071672354949,
+      "grad_norm": 1.7910349369049072,
+      "learning_rate": 3.23094425483504e-05,
+      "loss": 3.0475,
+      "step": 8506
+    },
+    {
+      "epoch": 2.903412969283276,
+      "grad_norm": 2.440401077270508,
+      "learning_rate": 3.219567690557452e-05,
+      "loss": 6.2855,
+      "step": 8507
+    },
+    {
+      "epoch": 2.903754266211604,
+      "grad_norm": 2.4906156063079834,
+      "learning_rate": 3.208191126279864e-05,
+      "loss": 5.975,
+      "step": 8508
+    },
+    {
+      "epoch": 2.904095563139932,
+      "grad_norm": 2.4778618812561035,
+      "learning_rate": 3.196814562002275e-05,
+      "loss": 5.3293,
+      "step": 8509
+    },
+    {
+      "epoch": 2.9044368600682593,
+      "grad_norm": 2.5149154663085938,
+      "learning_rate": 3.185437997724687e-05,
+      "loss": 5.6077,
+      "step": 8510
+    },
+    {
+      "epoch": 2.904778156996587,
+      "grad_norm": 2.450698137283325,
+      "learning_rate": 3.1740614334470994e-05,
+      "loss": 5.5491,
+      "step": 8511
+    },
+    {
+      "epoch": 2.9051194539249146,
+      "grad_norm": 2.482588768005371,
+      "learning_rate": 3.162684869169511e-05,
+      "loss": 4.8993,
+      "step": 8512
+    },
+    {
+      "epoch": 2.9054607508532424,
+      "grad_norm": 2.5095643997192383,
+      "learning_rate": 3.151308304891923e-05,
+      "loss": 5.3198,
+      "step": 8513
+    },
+    {
+      "epoch": 2.90580204778157,
+      "grad_norm": 2.603271484375,
+      "learning_rate": 3.139931740614334e-05,
+      "loss": 6.0782,
+      "step": 8514
+    },
+    {
+      "epoch": 2.9061433447098977,
+      "grad_norm": 2.445197105407715,
+      "learning_rate": 3.128555176336746e-05,
+      "loss": 6.073,
+      "step": 8515
+    },
+    {
+      "epoch": 2.906484641638225,
+      "grad_norm": 2.500511407852173,
+      "learning_rate": 3.117178612059158e-05,
+      "loss": 5.7921,
+      "step": 8516
+    },
+    {
+      "epoch": 2.906825938566553,
+      "grad_norm": 2.6357011795043945,
+      "learning_rate": 3.10580204778157e-05,
+      "loss": 5.9283,
+      "step": 8517
+    },
+    {
+      "epoch": 2.9071672354948808,
+      "grad_norm": 2.5531325340270996,
+      "learning_rate": 3.094425483503982e-05,
+      "loss": 5.7393,
+      "step": 8518
+    },
+    {
+      "epoch": 2.907508532423208,
+      "grad_norm": 2.4458229541778564,
+      "learning_rate": 3.0830489192263935e-05,
+      "loss": 5.7482,
+      "step": 8519
+    },
+    {
+      "epoch": 2.9078498293515356,
+      "grad_norm": 2.4953360557556152,
+      "learning_rate": 3.0716723549488054e-05,
+      "loss": 5.7151,
+      "step": 8520
+    },
+    {
+      "epoch": 2.9081911262798634,
+      "grad_norm": 2.5121448040008545,
+      "learning_rate": 3.060295790671217e-05,
+      "loss": 5.8653,
+      "step": 8521
+    },
+    {
+      "epoch": 2.9085324232081913,
+      "grad_norm": 2.5156524181365967,
+      "learning_rate": 3.048919226393629e-05,
+      "loss": 6.4765,
+      "step": 8522
+    },
+    {
+      "epoch": 2.9088737201365187,
+      "grad_norm": 2.5591423511505127,
+      "learning_rate": 3.037542662116041e-05,
+      "loss": 6.2486,
+      "step": 8523
+    },
+    {
+      "epoch": 2.9092150170648465,
+      "grad_norm": 2.460395336151123,
+      "learning_rate": 3.026166097838453e-05,
+      "loss": 5.5661,
+      "step": 8524
+    },
+    {
+      "epoch": 2.909556313993174,
+      "grad_norm": 2.557854413986206,
+      "learning_rate": 3.0147895335608646e-05,
+      "loss": 6.0055,
+      "step": 8525
+    },
+    {
+      "epoch": 2.909897610921502,
+      "grad_norm": 2.463956594467163,
+      "learning_rate": 3.0034129692832765e-05,
+      "loss": 5.6952,
+      "step": 8526
+    },
+    {
+      "epoch": 2.910238907849829,
+      "grad_norm": 2.577049493789673,
+      "learning_rate": 2.9920364050056883e-05,
+      "loss": 6.4598,
+      "step": 8527
+    },
+    {
+      "epoch": 2.910580204778157,
+      "grad_norm": 2.5568654537200928,
+      "learning_rate": 2.9806598407281e-05,
+      "loss": 5.7478,
+      "step": 8528
+    },
+    {
+      "epoch": 2.9109215017064844,
+      "grad_norm": 2.576349973678589,
+      "learning_rate": 2.969283276450512e-05,
+      "loss": 6.455,
+      "step": 8529
+    },
+    {
+      "epoch": 2.9112627986348123,
+      "grad_norm": 2.497002601623535,
+      "learning_rate": 2.957906712172924e-05,
+      "loss": 5.4758,
+      "step": 8530
+    },
+    {
+      "epoch": 2.91160409556314,
+      "grad_norm": 2.4823288917541504,
+      "learning_rate": 2.9465301478953357e-05,
+      "loss": 6.516,
+      "step": 8531
+    },
+    {
+      "epoch": 2.9119453924914676,
+      "grad_norm": 2.561558246612549,
+      "learning_rate": 2.9351535836177476e-05,
+      "loss": 6.5232,
+      "step": 8532
+    },
+    {
+      "epoch": 2.912286689419795,
+      "grad_norm": 2.498506546020508,
+      "learning_rate": 2.9237770193401594e-05,
+      "loss": 5.9132,
+      "step": 8533
+    },
+    {
+      "epoch": 2.912627986348123,
+      "grad_norm": 2.6166279315948486,
+      "learning_rate": 2.912400455062571e-05,
+      "loss": 6.2046,
+      "step": 8534
+    },
+    {
+      "epoch": 2.9129692832764507,
+      "grad_norm": 2.5233633518218994,
+      "learning_rate": 2.901023890784983e-05,
+      "loss": 5.4725,
+      "step": 8535
+    },
+    {
+      "epoch": 2.913310580204778,
+      "grad_norm": 2.531095266342163,
+      "learning_rate": 2.889647326507395e-05,
+      "loss": 5.3685,
+      "step": 8536
+    },
+    {
+      "epoch": 2.913651877133106,
+      "grad_norm": 2.4622771739959717,
+      "learning_rate": 2.8782707622298065e-05,
+      "loss": 6.094,
+      "step": 8537
+    },
+    {
+      "epoch": 2.9139931740614333,
+      "grad_norm": 2.830827474594116,
+      "learning_rate": 2.8668941979522186e-05,
+      "loss": 5.0274,
+      "step": 8538
+    },
+    {
+      "epoch": 2.914334470989761,
+      "grad_norm": 2.535754442214966,
+      "learning_rate": 2.85551763367463e-05,
+      "loss": 5.6579,
+      "step": 8539
+    },
+    {
+      "epoch": 2.9146757679180886,
+      "grad_norm": 2.4490978717803955,
+      "learning_rate": 2.844141069397042e-05,
+      "loss": 5.3953,
+      "step": 8540
+    },
+    {
+      "epoch": 2.9150170648464164,
+      "grad_norm": 2.5437428951263428,
+      "learning_rate": 2.8327645051194542e-05,
+      "loss": 5.3498,
+      "step": 8541
+    },
+    {
+      "epoch": 2.915358361774744,
+      "grad_norm": 2.509765863418579,
+      "learning_rate": 2.8213879408418657e-05,
+      "loss": 5.8658,
+      "step": 8542
+    },
+    {
+      "epoch": 2.9156996587030717,
+      "grad_norm": 2.643817186355591,
+      "learning_rate": 2.8100113765642776e-05,
+      "loss": 4.4102,
+      "step": 8543
+    },
+    {
+      "epoch": 2.9160409556313995,
+      "grad_norm": 2.594874858856201,
+      "learning_rate": 2.7986348122866894e-05,
+      "loss": 5.9731,
+      "step": 8544
+    },
+    {
+      "epoch": 2.916382252559727,
+      "grad_norm": 2.5506644248962402,
+      "learning_rate": 2.7872582480091013e-05,
+      "loss": 5.838,
+      "step": 8545
+    },
+    {
+      "epoch": 2.9167235494880543,
+      "grad_norm": 2.4979164600372314,
+      "learning_rate": 2.7758816837315134e-05,
+      "loss": 6.2731,
+      "step": 8546
+    },
+    {
+      "epoch": 2.917064846416382,
+      "grad_norm": 2.5016138553619385,
+      "learning_rate": 2.764505119453925e-05,
+      "loss": 6.0537,
+      "step": 8547
+    },
+    {
+      "epoch": 2.91740614334471,
+      "grad_norm": 2.495339870452881,
+      "learning_rate": 2.7531285551763368e-05,
+      "loss": 6.1774,
+      "step": 8548
+    },
+    {
+      "epoch": 2.9177474402730375,
+      "grad_norm": 2.524669647216797,
+      "learning_rate": 2.7417519908987487e-05,
+      "loss": 5.435,
+      "step": 8549
+    },
+    {
+      "epoch": 2.9180887372013653,
+      "grad_norm": 2.4896841049194336,
+      "learning_rate": 2.7303754266211605e-05,
+      "loss": 5.6336,
+      "step": 8550
+    },
+    {
+      "epoch": 2.9184300341296927,
+      "grad_norm": 2.673529624938965,
+      "learning_rate": 2.7189988623435724e-05,
+      "loss": 5.4314,
+      "step": 8551
+    },
+    {
+      "epoch": 2.9187713310580206,
+      "grad_norm": 2.626699447631836,
+      "learning_rate": 2.7076222980659842e-05,
+      "loss": 5.5602,
+      "step": 8552
+    },
+    {
+      "epoch": 2.919112627986348,
+      "grad_norm": 2.520334243774414,
+      "learning_rate": 2.696245733788396e-05,
+      "loss": 5.9856,
+      "step": 8553
+    },
+    {
+      "epoch": 2.919453924914676,
+      "grad_norm": 2.5466573238372803,
+      "learning_rate": 2.6848691695108076e-05,
+      "loss": 5.8577,
+      "step": 8554
+    },
+    {
+      "epoch": 2.919795221843003,
+      "grad_norm": 2.380843162536621,
+      "learning_rate": 2.6734926052332198e-05,
+      "loss": 6.0584,
+      "step": 8555
+    },
+    {
+      "epoch": 2.920136518771331,
+      "grad_norm": 2.4532737731933594,
+      "learning_rate": 2.6621160409556316e-05,
+      "loss": 5.8358,
+      "step": 8556
+    },
+    {
+      "epoch": 2.920477815699659,
+      "grad_norm": 2.493149518966675,
+      "learning_rate": 2.650739476678043e-05,
+      "loss": 6.2587,
+      "step": 8557
+    },
+    {
+      "epoch": 2.9208191126279863,
+      "grad_norm": 2.247823476791382,
+      "learning_rate": 2.6393629124004553e-05,
+      "loss": 4.0832,
+      "step": 8558
+    },
+    {
+      "epoch": 2.9211604095563137,
+      "grad_norm": 2.4827280044555664,
+      "learning_rate": 2.6279863481228668e-05,
+      "loss": 5.8317,
+      "step": 8559
+    },
+    {
+      "epoch": 2.9215017064846416,
+      "grad_norm": 2.504302978515625,
+      "learning_rate": 2.6166097838452787e-05,
+      "loss": 5.7784,
+      "step": 8560
+    },
+    {
+      "epoch": 2.9218430034129694,
+      "grad_norm": 2.515334367752075,
+      "learning_rate": 2.6052332195676905e-05,
+      "loss": 5.533,
+      "step": 8561
+    },
+    {
+      "epoch": 2.922184300341297,
+      "grad_norm": 2.5521583557128906,
+      "learning_rate": 2.5938566552901024e-05,
+      "loss": 5.6752,
+      "step": 8562
+    },
+    {
+      "epoch": 2.9225255972696247,
+      "grad_norm": 2.538944959640503,
+      "learning_rate": 2.5824800910125145e-05,
+      "loss": 6.2666,
+      "step": 8563
+    },
+    {
+      "epoch": 2.922866894197952,
+      "grad_norm": 5.985457897186279,
+      "learning_rate": 2.571103526734926e-05,
+      "loss": 4.8984,
+      "step": 8564
+    },
+    {
+      "epoch": 2.92320819112628,
+      "grad_norm": 2.5538158416748047,
+      "learning_rate": 2.559726962457338e-05,
+      "loss": 5.5896,
+      "step": 8565
+    },
+    {
+      "epoch": 2.9235494880546073,
+      "grad_norm": 2.5504369735717773,
+      "learning_rate": 2.5483503981797498e-05,
+      "loss": 5.352,
+      "step": 8566
+    },
+    {
+      "epoch": 2.923890784982935,
+      "grad_norm": 2.536086082458496,
+      "learning_rate": 2.5369738339021616e-05,
+      "loss": 5.7882,
+      "step": 8567
+    },
+    {
+      "epoch": 2.9242320819112626,
+      "grad_norm": 2.487410068511963,
+      "learning_rate": 2.5255972696245735e-05,
+      "loss": 6.1777,
+      "step": 8568
+    },
+    {
+      "epoch": 2.9245733788395905,
+      "grad_norm": 2.55271577835083,
+      "learning_rate": 2.5142207053469853e-05,
+      "loss": 6.6132,
+      "step": 8569
+    },
+    {
+      "epoch": 2.9249146757679183,
+      "grad_norm": 2.5541188716888428,
+      "learning_rate": 2.502844141069397e-05,
+      "loss": 4.3692,
+      "step": 8570
+    },
+    {
+      "epoch": 2.9252559726962457,
+      "grad_norm": 2.487154006958008,
+      "learning_rate": 2.4914675767918087e-05,
+      "loss": 6.1641,
+      "step": 8571
+    },
+    {
+      "epoch": 2.925597269624573,
+      "grad_norm": 2.4862101078033447,
+      "learning_rate": 2.480091012514221e-05,
+      "loss": 6.3252,
+      "step": 8572
+    },
+    {
+      "epoch": 2.925938566552901,
+      "grad_norm": 2.580897569656372,
+      "learning_rate": 2.4687144482366327e-05,
+      "loss": 5.0885,
+      "step": 8573
+    },
+    {
+      "epoch": 2.926279863481229,
+      "grad_norm": 2.546022653579712,
+      "learning_rate": 2.4573378839590442e-05,
+      "loss": 5.7194,
+      "step": 8574
+    },
+    {
+      "epoch": 2.926621160409556,
+      "grad_norm": 2.55588436126709,
+      "learning_rate": 2.4459613196814564e-05,
+      "loss": 6.2847,
+      "step": 8575
+    },
+    {
+      "epoch": 2.926962457337884,
+      "grad_norm": 2.4706380367279053,
+      "learning_rate": 2.434584755403868e-05,
+      "loss": 5.9013,
+      "step": 8576
+    },
+    {
+      "epoch": 2.9273037542662115,
+      "grad_norm": 2.4694294929504395,
+      "learning_rate": 2.4232081911262798e-05,
+      "loss": 5.854,
+      "step": 8577
+    },
+    {
+      "epoch": 2.9276450511945393,
+      "grad_norm": 2.516496181488037,
+      "learning_rate": 2.411831626848692e-05,
+      "loss": 5.9697,
+      "step": 8578
+    },
+    {
+      "epoch": 2.9279863481228667,
+      "grad_norm": 2.545330762863159,
+      "learning_rate": 2.4004550625711035e-05,
+      "loss": 6.2448,
+      "step": 8579
+    },
+    {
+      "epoch": 2.9283276450511946,
+      "grad_norm": 2.4520492553710938,
+      "learning_rate": 2.3890784982935157e-05,
+      "loss": 5.3888,
+      "step": 8580
+    },
+    {
+      "epoch": 2.928668941979522,
+      "grad_norm": 2.5857083797454834,
+      "learning_rate": 2.377701934015927e-05,
+      "loss": 6.2331,
+      "step": 8581
+    },
+    {
+      "epoch": 2.92901023890785,
+      "grad_norm": 2.9561941623687744,
+      "learning_rate": 2.366325369738339e-05,
+      "loss": 4.9944,
+      "step": 8582
+    },
+    {
+      "epoch": 2.9293515358361777,
+      "grad_norm": 2.5174217224121094,
+      "learning_rate": 2.3549488054607512e-05,
+      "loss": 5.897,
+      "step": 8583
+    },
+    {
+      "epoch": 2.929692832764505,
+      "grad_norm": 2.585947275161743,
+      "learning_rate": 2.3435722411831627e-05,
+      "loss": 6.0279,
+      "step": 8584
+    },
+    {
+      "epoch": 2.9300341296928325,
+      "grad_norm": 2.439552068710327,
+      "learning_rate": 2.3321956769055746e-05,
+      "loss": 5.9078,
+      "step": 8585
+    },
+    {
+      "epoch": 2.9303754266211604,
+      "grad_norm": 2.4478914737701416,
+      "learning_rate": 2.3208191126279864e-05,
+      "loss": 5.6497,
+      "step": 8586
+    },
+    {
+      "epoch": 2.930716723549488,
+      "grad_norm": 2.537717342376709,
+      "learning_rate": 2.3094425483503983e-05,
+      "loss": 6.0121,
+      "step": 8587
+    },
+    {
+      "epoch": 2.9310580204778156,
+      "grad_norm": 2.4616639614105225,
+      "learning_rate": 2.2980659840728098e-05,
+      "loss": 5.9547,
+      "step": 8588
+    },
+    {
+      "epoch": 2.9313993174061435,
+      "grad_norm": 2.514845371246338,
+      "learning_rate": 2.286689419795222e-05,
+      "loss": 6.3048,
+      "step": 8589
+    },
+    {
+      "epoch": 2.931740614334471,
+      "grad_norm": 2.4787206649780273,
+      "learning_rate": 2.2753128555176338e-05,
+      "loss": 6.2414,
+      "step": 8590
+    },
+    {
+      "epoch": 2.9320819112627987,
+      "grad_norm": 2.4855761528015137,
+      "learning_rate": 2.2639362912400453e-05,
+      "loss": 6.3916,
+      "step": 8591
+    },
+    {
+      "epoch": 2.932423208191126,
+      "grad_norm": 2.6569809913635254,
+      "learning_rate": 2.2525597269624575e-05,
+      "loss": 6.2394,
+      "step": 8592
+    },
+    {
+      "epoch": 2.932764505119454,
+      "grad_norm": 2.4555177688598633,
+      "learning_rate": 2.241183162684869e-05,
+      "loss": 5.8789,
+      "step": 8593
+    },
+    {
+      "epoch": 2.9331058020477814,
+      "grad_norm": 2.477505683898926,
+      "learning_rate": 2.2298065984072812e-05,
+      "loss": 5.7146,
+      "step": 8594
+    },
+    {
+      "epoch": 2.9334470989761092,
+      "grad_norm": 2.5200958251953125,
+      "learning_rate": 2.218430034129693e-05,
+      "loss": 6.0025,
+      "step": 8595
+    },
+    {
+      "epoch": 2.933788395904437,
+      "grad_norm": 2.5263843536376953,
+      "learning_rate": 2.2070534698521046e-05,
+      "loss": 6.2893,
+      "step": 8596
+    },
+    {
+      "epoch": 2.9341296928327645,
+      "grad_norm": 2.552253007888794,
+      "learning_rate": 2.1956769055745168e-05,
+      "loss": 6.0527,
+      "step": 8597
+    },
+    {
+      "epoch": 2.934470989761092,
+      "grad_norm": 2.494657039642334,
+      "learning_rate": 2.1843003412969283e-05,
+      "loss": 5.55,
+      "step": 8598
+    },
+    {
+      "epoch": 2.9348122866894197,
+      "grad_norm": 2.446608543395996,
+      "learning_rate": 2.17292377701934e-05,
+      "loss": 5.6856,
+      "step": 8599
+    },
+    {
+      "epoch": 2.9351535836177476,
+      "grad_norm": 2.4764108657836914,
+      "learning_rate": 2.1615472127417523e-05,
+      "loss": 6.2355,
+      "step": 8600
+    },
+    {
+      "epoch": 2.935494880546075,
+      "grad_norm": 2.570817232131958,
+      "learning_rate": 2.1501706484641638e-05,
+      "loss": 6.3309,
+      "step": 8601
+    },
+    {
+      "epoch": 2.935836177474403,
+      "grad_norm": 2.5633604526519775,
+      "learning_rate": 2.1387940841865757e-05,
+      "loss": 5.7691,
+      "step": 8602
+    },
+    {
+      "epoch": 2.9361774744027302,
+      "grad_norm": 2.5183494091033936,
+      "learning_rate": 2.1274175199089875e-05,
+      "loss": 4.9273,
+      "step": 8603
+    },
+    {
+      "epoch": 2.936518771331058,
+      "grad_norm": 2.4383068084716797,
+      "learning_rate": 2.1160409556313994e-05,
+      "loss": 5.0008,
+      "step": 8604
+    },
+    {
+      "epoch": 2.9368600682593855,
+      "grad_norm": 2.5003600120544434,
+      "learning_rate": 2.1046643913538112e-05,
+      "loss": 5.857,
+      "step": 8605
+    },
+    {
+      "epoch": 2.9372013651877134,
+      "grad_norm": 2.480581283569336,
+      "learning_rate": 2.093287827076223e-05,
+      "loss": 5.1716,
+      "step": 8606
+    },
+    {
+      "epoch": 2.9375426621160408,
+      "grad_norm": 2.540728807449341,
+      "learning_rate": 2.081911262798635e-05,
+      "loss": 6.2848,
+      "step": 8607
+    },
+    {
+      "epoch": 2.9378839590443686,
+      "grad_norm": 2.5028791427612305,
+      "learning_rate": 2.0705346985210464e-05,
+      "loss": 5.7391,
+      "step": 8608
+    },
+    {
+      "epoch": 2.9382252559726965,
+      "grad_norm": 2.4789538383483887,
+      "learning_rate": 2.0591581342434586e-05,
+      "loss": 5.223,
+      "step": 8609
+    },
+    {
+      "epoch": 2.938566552901024,
+      "grad_norm": 2.4967267513275146,
+      "learning_rate": 2.04778156996587e-05,
+      "loss": 5.668,
+      "step": 8610
+    },
+    {
+      "epoch": 2.9389078498293513,
+      "grad_norm": 2.519948959350586,
+      "learning_rate": 2.0364050056882823e-05,
+      "loss": 5.9026,
+      "step": 8611
+    },
+    {
+      "epoch": 2.939249146757679,
+      "grad_norm": 2.4573259353637695,
+      "learning_rate": 2.025028441410694e-05,
+      "loss": 5.8025,
+      "step": 8612
+    },
+    {
+      "epoch": 2.939590443686007,
+      "grad_norm": 2.5251667499542236,
+      "learning_rate": 2.0136518771331057e-05,
+      "loss": 5.5212,
+      "step": 8613
+    },
+    {
+      "epoch": 2.9399317406143344,
+      "grad_norm": 2.6210274696350098,
+      "learning_rate": 2.002275312855518e-05,
+      "loss": 5.3951,
+      "step": 8614
+    },
+    {
+      "epoch": 2.9402730375426622,
+      "grad_norm": 2.532759189605713,
+      "learning_rate": 1.9908987485779294e-05,
+      "loss": 5.797,
+      "step": 8615
+    },
+    {
+      "epoch": 2.9406143344709896,
+      "grad_norm": 2.4565927982330322,
+      "learning_rate": 1.9795221843003412e-05,
+      "loss": 5.4728,
+      "step": 8616
+    },
+    {
+      "epoch": 2.9409556313993175,
+      "grad_norm": 2.4736921787261963,
+      "learning_rate": 1.9681456200227534e-05,
+      "loss": 5.9357,
+      "step": 8617
+    },
+    {
+      "epoch": 2.941296928327645,
+      "grad_norm": 2.5852224826812744,
+      "learning_rate": 1.956769055745165e-05,
+      "loss": 4.7119,
+      "step": 8618
+    },
+    {
+      "epoch": 2.9416382252559727,
+      "grad_norm": 2.5384979248046875,
+      "learning_rate": 1.9453924914675768e-05,
+      "loss": 5.4497,
+      "step": 8619
+    },
+    {
+      "epoch": 2.9419795221843,
+      "grad_norm": 2.4566867351531982,
+      "learning_rate": 1.9340159271899886e-05,
+      "loss": 5.7871,
+      "step": 8620
+    },
+    {
+      "epoch": 2.942320819112628,
+      "grad_norm": 2.4984500408172607,
+      "learning_rate": 1.9226393629124005e-05,
+      "loss": 5.9535,
+      "step": 8621
+    },
+    {
+      "epoch": 2.942662116040956,
+      "grad_norm": 2.4898228645324707,
+      "learning_rate": 1.9112627986348123e-05,
+      "loss": 5.7868,
+      "step": 8622
+    },
+    {
+      "epoch": 2.9430034129692833,
+      "grad_norm": 2.410717248916626,
+      "learning_rate": 1.899886234357224e-05,
+      "loss": 4.692,
+      "step": 8623
+    },
+    {
+      "epoch": 2.9433447098976107,
+      "grad_norm": 2.5031991004943848,
+      "learning_rate": 1.888509670079636e-05,
+      "loss": 6.0732,
+      "step": 8624
+    },
+    {
+      "epoch": 2.9436860068259385,
+      "grad_norm": 2.5316109657287598,
+      "learning_rate": 1.877133105802048e-05,
+      "loss": 5.5196,
+      "step": 8625
+    },
+    {
+      "epoch": 2.9440273037542664,
+      "grad_norm": 2.44270658493042,
+      "learning_rate": 1.8657565415244597e-05,
+      "loss": 5.973,
+      "step": 8626
+    },
+    {
+      "epoch": 2.9443686006825938,
+      "grad_norm": 2.500929355621338,
+      "learning_rate": 1.8543799772468716e-05,
+      "loss": 6.2362,
+      "step": 8627
+    },
+    {
+      "epoch": 2.9447098976109216,
+      "grad_norm": 2.4967660903930664,
+      "learning_rate": 1.8430034129692834e-05,
+      "loss": 6.0688,
+      "step": 8628
+    },
+    {
+      "epoch": 2.945051194539249,
+      "grad_norm": 2.551168441772461,
+      "learning_rate": 1.8316268486916953e-05,
+      "loss": 6.4435,
+      "step": 8629
+    },
+    {
+      "epoch": 2.945392491467577,
+      "grad_norm": 2.529707431793213,
+      "learning_rate": 1.8202502844141068e-05,
+      "loss": 6.0493,
+      "step": 8630
+    },
+    {
+      "epoch": 2.9457337883959043,
+      "grad_norm": 2.394730567932129,
+      "learning_rate": 1.808873720136519e-05,
+      "loss": 5.1711,
+      "step": 8631
+    },
+    {
+      "epoch": 2.946075085324232,
+      "grad_norm": 2.5961592197418213,
+      "learning_rate": 1.7974971558589305e-05,
+      "loss": 6.0267,
+      "step": 8632
+    },
+    {
+      "epoch": 2.9464163822525595,
+      "grad_norm": 2.4679746627807617,
+      "learning_rate": 1.7861205915813423e-05,
+      "loss": 5.6721,
+      "step": 8633
+    },
+    {
+      "epoch": 2.9467576791808874,
+      "grad_norm": 2.5363996028900146,
+      "learning_rate": 1.7747440273037545e-05,
+      "loss": 5.6486,
+      "step": 8634
+    },
+    {
+      "epoch": 2.9470989761092152,
+      "grad_norm": 2.4979100227355957,
+      "learning_rate": 1.763367463026166e-05,
+      "loss": 6.2355,
+      "step": 8635
+    },
+    {
+      "epoch": 2.9474402730375426,
+      "grad_norm": 2.5005037784576416,
+      "learning_rate": 1.751990898748578e-05,
+      "loss": 6.0746,
+      "step": 8636
+    },
+    {
+      "epoch": 2.94778156996587,
+      "grad_norm": 2.6224281787872314,
+      "learning_rate": 1.7406143344709897e-05,
+      "loss": 6.1915,
+      "step": 8637
+    },
+    {
+      "epoch": 2.948122866894198,
+      "grad_norm": 2.4762070178985596,
+      "learning_rate": 1.7292377701934016e-05,
+      "loss": 5.9801,
+      "step": 8638
+    },
+    {
+      "epoch": 2.9484641638225257,
+      "grad_norm": 2.5673601627349854,
+      "learning_rate": 1.7178612059158134e-05,
+      "loss": 5.5908,
+      "step": 8639
+    },
+    {
+      "epoch": 2.948805460750853,
+      "grad_norm": 2.461416482925415,
+      "learning_rate": 1.7064846416382253e-05,
+      "loss": 5.2207,
+      "step": 8640
+    },
+    {
+      "epoch": 2.949146757679181,
+      "grad_norm": 2.4866373538970947,
+      "learning_rate": 1.695108077360637e-05,
+      "loss": 6.422,
+      "step": 8641
+    },
+    {
+      "epoch": 2.9494880546075084,
+      "grad_norm": 2.6868081092834473,
+      "learning_rate": 1.683731513083049e-05,
+      "loss": 4.5656,
+      "step": 8642
+    },
+    {
+      "epoch": 2.9498293515358363,
+      "grad_norm": 2.542024612426758,
+      "learning_rate": 1.6723549488054608e-05,
+      "loss": 5.2073,
+      "step": 8643
+    },
+    {
+      "epoch": 2.9501706484641637,
+      "grad_norm": 2.4783406257629395,
+      "learning_rate": 1.6609783845278727e-05,
+      "loss": 5.8559,
+      "step": 8644
+    },
+    {
+      "epoch": 2.9505119453924915,
+      "grad_norm": 2.5044283866882324,
+      "learning_rate": 1.6496018202502845e-05,
+      "loss": 6.2763,
+      "step": 8645
+    },
+    {
+      "epoch": 2.950853242320819,
+      "grad_norm": 2.472458600997925,
+      "learning_rate": 1.6382252559726964e-05,
+      "loss": 6.0363,
+      "step": 8646
+    },
+    {
+      "epoch": 2.9511945392491468,
+      "grad_norm": 2.4711015224456787,
+      "learning_rate": 1.626848691695108e-05,
+      "loss": 5.4181,
+      "step": 8647
+    },
+    {
+      "epoch": 2.9515358361774746,
+      "grad_norm": 2.5318753719329834,
+      "learning_rate": 1.61547212741752e-05,
+      "loss": 6.4278,
+      "step": 8648
+    },
+    {
+      "epoch": 2.951877133105802,
+      "grad_norm": 2.4760727882385254,
+      "learning_rate": 1.604095563139932e-05,
+      "loss": 6.1243,
+      "step": 8649
+    },
+    {
+      "epoch": 2.9522184300341294,
+      "grad_norm": 2.4930100440979004,
+      "learning_rate": 1.5927189988623434e-05,
+      "loss": 5.0605,
+      "step": 8650
+    },
+    {
+      "epoch": 2.9525597269624573,
+      "grad_norm": 2.477997064590454,
+      "learning_rate": 1.5813424345847556e-05,
+      "loss": 6.0705,
+      "step": 8651
+    },
+    {
+      "epoch": 2.952901023890785,
+      "grad_norm": 2.1888606548309326,
+      "learning_rate": 1.569965870307167e-05,
+      "loss": 4.6192,
+      "step": 8652
+    },
+    {
+      "epoch": 2.9532423208191125,
+      "grad_norm": 2.3284385204315186,
+      "learning_rate": 1.558589306029579e-05,
+      "loss": 4.0104,
+      "step": 8653
+    },
+    {
+      "epoch": 2.9535836177474404,
+      "grad_norm": 2.4472768306732178,
+      "learning_rate": 1.547212741751991e-05,
+      "loss": 5.7057,
+      "step": 8654
+    },
+    {
+      "epoch": 2.953924914675768,
+      "grad_norm": 2.4223828315734863,
+      "learning_rate": 1.5358361774744027e-05,
+      "loss": 5.7879,
+      "step": 8655
+    },
+    {
+      "epoch": 2.9542662116040956,
+      "grad_norm": 2.5395631790161133,
+      "learning_rate": 1.5244596131968145e-05,
+      "loss": 5.7686,
+      "step": 8656
+    },
+    {
+      "epoch": 2.954607508532423,
+      "grad_norm": 2.530900239944458,
+      "learning_rate": 1.5130830489192265e-05,
+      "loss": 4.8828,
+      "step": 8657
+    },
+    {
+      "epoch": 2.954948805460751,
+      "grad_norm": 2.5554354190826416,
+      "learning_rate": 1.5017064846416382e-05,
+      "loss": 6.6185,
+      "step": 8658
+    },
+    {
+      "epoch": 2.9552901023890783,
+      "grad_norm": 2.492499828338623,
+      "learning_rate": 1.49032992036405e-05,
+      "loss": 5.1967,
+      "step": 8659
+    },
+    {
+      "epoch": 2.955631399317406,
+      "grad_norm": 2.4863228797912598,
+      "learning_rate": 1.478953356086462e-05,
+      "loss": 5.9918,
+      "step": 8660
+    },
+    {
+      "epoch": 2.955972696245734,
+      "grad_norm": 2.4008748531341553,
+      "learning_rate": 1.4675767918088738e-05,
+      "loss": 5.112,
+      "step": 8661
+    },
+    {
+      "epoch": 2.9563139931740614,
+      "grad_norm": 2.5058634281158447,
+      "learning_rate": 1.4562002275312855e-05,
+      "loss": 5.9349,
+      "step": 8662
+    },
+    {
+      "epoch": 2.956655290102389,
+      "grad_norm": 2.5977463722229004,
+      "learning_rate": 1.4448236632536975e-05,
+      "loss": 5.3876,
+      "step": 8663
+    },
+    {
+      "epoch": 2.9569965870307167,
+      "grad_norm": 1.7508560419082642,
+      "learning_rate": 1.4334470989761093e-05,
+      "loss": 2.9253,
+      "step": 8664
+    },
+    {
+      "epoch": 2.9573378839590445,
+      "grad_norm": 2.433396816253662,
+      "learning_rate": 1.422070534698521e-05,
+      "loss": 6.2839,
+      "step": 8665
+    },
+    {
+      "epoch": 2.957679180887372,
+      "grad_norm": 2.559898614883423,
+      "learning_rate": 1.4106939704209329e-05,
+      "loss": 5.909,
+      "step": 8666
+    },
+    {
+      "epoch": 2.9580204778156998,
+      "grad_norm": 2.5637035369873047,
+      "learning_rate": 1.3993174061433447e-05,
+      "loss": 5.9518,
+      "step": 8667
+    },
+    {
+      "epoch": 2.958361774744027,
+      "grad_norm": 2.566162586212158,
+      "learning_rate": 1.3879408418657567e-05,
+      "loss": 6.0647,
+      "step": 8668
+    },
+    {
+      "epoch": 2.958703071672355,
+      "grad_norm": 2.502931594848633,
+      "learning_rate": 1.3765642775881684e-05,
+      "loss": 4.3806,
+      "step": 8669
+    },
+    {
+      "epoch": 2.9590443686006824,
+      "grad_norm": 2.453782320022583,
+      "learning_rate": 1.3651877133105803e-05,
+      "loss": 5.6523,
+      "step": 8670
+    },
+    {
+      "epoch": 2.9593856655290103,
+      "grad_norm": 2.585204601287842,
+      "learning_rate": 1.3538111490329921e-05,
+      "loss": 5.9991,
+      "step": 8671
+    },
+    {
+      "epoch": 2.9597269624573377,
+      "grad_norm": 2.713543176651001,
+      "learning_rate": 1.3424345847554038e-05,
+      "loss": 4.8614,
+      "step": 8672
+    },
+    {
+      "epoch": 2.9600682593856655,
+      "grad_norm": 2.511810541152954,
+      "learning_rate": 1.3310580204778158e-05,
+      "loss": 6.274,
+      "step": 8673
+    },
+    {
+      "epoch": 2.9604095563139934,
+      "grad_norm": 2.4611916542053223,
+      "learning_rate": 1.3196814562002277e-05,
+      "loss": 5.9661,
+      "step": 8674
+    },
+    {
+      "epoch": 2.960750853242321,
+      "grad_norm": 2.5248091220855713,
+      "learning_rate": 1.3083048919226393e-05,
+      "loss": 6.2989,
+      "step": 8675
+    },
+    {
+      "epoch": 2.961092150170648,
+      "grad_norm": 2.4829485416412354,
+      "learning_rate": 1.2969283276450512e-05,
+      "loss": 5.8782,
+      "step": 8676
+    },
+    {
+      "epoch": 2.961433447098976,
+      "grad_norm": 2.432086229324341,
+      "learning_rate": 1.285551763367463e-05,
+      "loss": 5.5988,
+      "step": 8677
+    },
+    {
+      "epoch": 2.961774744027304,
+      "grad_norm": 2.426537036895752,
+      "learning_rate": 1.2741751990898749e-05,
+      "loss": 4.3375,
+      "step": 8678
+    },
+    {
+      "epoch": 2.9621160409556313,
+      "grad_norm": 2.4175045490264893,
+      "learning_rate": 1.2627986348122867e-05,
+      "loss": 5.8644,
+      "step": 8679
+    },
+    {
+      "epoch": 2.962457337883959,
+      "grad_norm": 2.5056276321411133,
+      "learning_rate": 1.2514220705346986e-05,
+      "loss": 5.4849,
+      "step": 8680
+    },
+    {
+      "epoch": 2.9627986348122866,
+      "grad_norm": 2.4787650108337402,
+      "learning_rate": 1.2400455062571104e-05,
+      "loss": 5.944,
+      "step": 8681
+    },
+    {
+      "epoch": 2.9631399317406144,
+      "grad_norm": 2.4695487022399902,
+      "learning_rate": 1.2286689419795221e-05,
+      "loss": 5.8351,
+      "step": 8682
+    },
+    {
+      "epoch": 2.963481228668942,
+      "grad_norm": 2.573592185974121,
+      "learning_rate": 1.217292377701934e-05,
+      "loss": 5.9925,
+      "step": 8683
+    },
+    {
+      "epoch": 2.9638225255972697,
+      "grad_norm": 2.236001968383789,
+      "learning_rate": 1.205915813424346e-05,
+      "loss": 4.7678,
+      "step": 8684
+    },
+    {
+      "epoch": 2.964163822525597,
+      "grad_norm": 2.475437641143799,
+      "learning_rate": 1.1945392491467578e-05,
+      "loss": 6.0264,
+      "step": 8685
+    },
+    {
+      "epoch": 2.964505119453925,
+      "grad_norm": 2.467595100402832,
+      "learning_rate": 1.1831626848691695e-05,
+      "loss": 5.2569,
+      "step": 8686
+    },
+    {
+      "epoch": 2.9648464163822528,
+      "grad_norm": 2.4960720539093018,
+      "learning_rate": 1.1717861205915814e-05,
+      "loss": 5.4893,
+      "step": 8687
+    },
+    {
+      "epoch": 2.96518771331058,
+      "grad_norm": 2.579531192779541,
+      "learning_rate": 1.1604095563139932e-05,
+      "loss": 6.2595,
+      "step": 8688
+    },
+    {
+      "epoch": 2.9655290102389076,
+      "grad_norm": 2.4547293186187744,
+      "learning_rate": 1.1490329920364049e-05,
+      "loss": 6.4029,
+      "step": 8689
+    },
+    {
+      "epoch": 2.9658703071672354,
+      "grad_norm": 2.6123619079589844,
+      "learning_rate": 1.1376564277588169e-05,
+      "loss": 5.7366,
+      "step": 8690
+    },
+    {
+      "epoch": 2.9662116040955633,
+      "grad_norm": 2.4635937213897705,
+      "learning_rate": 1.1262798634812288e-05,
+      "loss": 5.4398,
+      "step": 8691
+    },
+    {
+      "epoch": 2.9665529010238907,
+      "grad_norm": 2.44295597076416,
+      "learning_rate": 1.1149032992036406e-05,
+      "loss": 5.6055,
+      "step": 8692
+    },
+    {
+      "epoch": 2.9668941979522185,
+      "grad_norm": 2.5429694652557373,
+      "learning_rate": 1.1035267349260523e-05,
+      "loss": 5.9528,
+      "step": 8693
+    },
+    {
+      "epoch": 2.967235494880546,
+      "grad_norm": 2.481144428253174,
+      "learning_rate": 1.0921501706484641e-05,
+      "loss": 6.0664,
+      "step": 8694
+    },
+    {
+      "epoch": 2.967576791808874,
+      "grad_norm": 2.5129120349884033,
+      "learning_rate": 1.0807736063708762e-05,
+      "loss": 5.1419,
+      "step": 8695
+    },
+    {
+      "epoch": 2.967918088737201,
+      "grad_norm": 2.522629737854004,
+      "learning_rate": 1.0693970420932878e-05,
+      "loss": 6.0011,
+      "step": 8696
+    },
+    {
+      "epoch": 2.968259385665529,
+      "grad_norm": 2.5712890625,
+      "learning_rate": 1.0580204778156997e-05,
+      "loss": 5.9975,
+      "step": 8697
+    },
+    {
+      "epoch": 2.9686006825938565,
+      "grad_norm": 2.445847511291504,
+      "learning_rate": 1.0466439135381115e-05,
+      "loss": 5.2381,
+      "step": 8698
+    },
+    {
+      "epoch": 2.9689419795221843,
+      "grad_norm": 2.4890472888946533,
+      "learning_rate": 1.0352673492605232e-05,
+      "loss": 6.3443,
+      "step": 8699
+    },
+    {
+      "epoch": 2.969283276450512,
+      "grad_norm": 2.519037961959839,
+      "learning_rate": 1.023890784982935e-05,
+      "loss": 4.9883,
+      "step": 8700
+    },
+    {
+      "epoch": 2.9696245733788396,
+      "grad_norm": 2.496644973754883,
+      "learning_rate": 1.012514220705347e-05,
+      "loss": 5.4765,
+      "step": 8701
+    },
+    {
+      "epoch": 2.969965870307167,
+      "grad_norm": 2.4772489070892334,
+      "learning_rate": 1.001137656427759e-05,
+      "loss": 6.0936,
+      "step": 8702
+    },
+    {
+      "epoch": 2.970307167235495,
+      "grad_norm": 2.4817209243774414,
+      "learning_rate": 9.897610921501706e-06,
+      "loss": 5.833,
+      "step": 8703
+    },
+    {
+      "epoch": 2.9706484641638227,
+      "grad_norm": 2.488058567047119,
+      "learning_rate": 9.783845278725825e-06,
+      "loss": 6.3268,
+      "step": 8704
+    },
+    {
+      "epoch": 2.97098976109215,
+      "grad_norm": 2.6241326332092285,
+      "learning_rate": 9.670079635949943e-06,
+      "loss": 6.1281,
+      "step": 8705
+    },
+    {
+      "epoch": 2.971331058020478,
+      "grad_norm": 2.4947798252105713,
+      "learning_rate": 9.556313993174062e-06,
+      "loss": 6.1815,
+      "step": 8706
+    },
+    {
+      "epoch": 2.9716723549488053,
+      "grad_norm": 2.588179349899292,
+      "learning_rate": 9.44254835039818e-06,
+      "loss": 6.1517,
+      "step": 8707
+    },
+    {
+      "epoch": 2.972013651877133,
+      "grad_norm": 2.5421693325042725,
+      "learning_rate": 9.328782707622299e-06,
+      "loss": 5.874,
+      "step": 8708
+    },
+    {
+      "epoch": 2.972354948805461,
+      "grad_norm": 2.4261326789855957,
+      "learning_rate": 9.215017064846417e-06,
+      "loss": 5.1937,
+      "step": 8709
+    },
+    {
+      "epoch": 2.9726962457337884,
+      "grad_norm": 4.1984076499938965,
+      "learning_rate": 9.101251422070534e-06,
+      "loss": 4.4323,
+      "step": 8710
+    },
+    {
+      "epoch": 2.973037542662116,
+      "grad_norm": 2.5183022022247314,
+      "learning_rate": 8.987485779294652e-06,
+      "loss": 6.3584,
+      "step": 8711
+    },
+    {
+      "epoch": 2.9733788395904437,
+      "grad_norm": 2.593282699584961,
+      "learning_rate": 8.873720136518773e-06,
+      "loss": 6.1431,
+      "step": 8712
+    },
+    {
+      "epoch": 2.9737201365187715,
+      "grad_norm": 2.50671124458313,
+      "learning_rate": 8.75995449374289e-06,
+      "loss": 5.9708,
+      "step": 8713
+    },
+    {
+      "epoch": 2.974061433447099,
+      "grad_norm": 2.5852482318878174,
+      "learning_rate": 8.646188850967008e-06,
+      "loss": 6.1701,
+      "step": 8714
+    },
+    {
+      "epoch": 2.9744027303754264,
+      "grad_norm": 2.4211649894714355,
+      "learning_rate": 8.532423208191126e-06,
+      "loss": 5.8361,
+      "step": 8715
+    },
+    {
+      "epoch": 2.974744027303754,
+      "grad_norm": 2.5099081993103027,
+      "learning_rate": 8.418657565415245e-06,
+      "loss": 5.9594,
+      "step": 8716
+    },
+    {
+      "epoch": 2.975085324232082,
+      "grad_norm": 2.5803756713867188,
+      "learning_rate": 8.304891922639363e-06,
+      "loss": 6.0941,
+      "step": 8717
+    },
+    {
+      "epoch": 2.9754266211604095,
+      "grad_norm": 2.493959665298462,
+      "learning_rate": 8.191126279863482e-06,
+      "loss": 6.2058,
+      "step": 8718
+    },
+    {
+      "epoch": 2.9757679180887373,
+      "grad_norm": 2.3790855407714844,
+      "learning_rate": 8.0773606370876e-06,
+      "loss": 4.5639,
+      "step": 8719
+    },
+    {
+      "epoch": 2.9761092150170647,
+      "grad_norm": 2.463736057281494,
+      "learning_rate": 7.963594994311717e-06,
+      "loss": 5.2897,
+      "step": 8720
+    },
+    {
+      "epoch": 2.9764505119453926,
+      "grad_norm": 2.368633270263672,
+      "learning_rate": 7.849829351535836e-06,
+      "loss": 4.8404,
+      "step": 8721
+    },
+    {
+      "epoch": 2.9767918088737204,
+      "grad_norm": 2.4608263969421387,
+      "learning_rate": 7.736063708759956e-06,
+      "loss": 6.1187,
+      "step": 8722
+    },
+    {
+      "epoch": 2.977133105802048,
+      "grad_norm": 2.5266082286834717,
+      "learning_rate": 7.622298065984073e-06,
+      "loss": 4.6866,
+      "step": 8723
+    },
+    {
+      "epoch": 2.9774744027303752,
+      "grad_norm": 2.4819984436035156,
+      "learning_rate": 7.508532423208191e-06,
+      "loss": 5.8985,
+      "step": 8724
+    },
+    {
+      "epoch": 2.977815699658703,
+      "grad_norm": 2.534745931625366,
+      "learning_rate": 7.39476678043231e-06,
+      "loss": 6.0181,
+      "step": 8725
+    },
+    {
+      "epoch": 2.978156996587031,
+      "grad_norm": 2.522770643234253,
+      "learning_rate": 7.281001137656427e-06,
+      "loss": 4.879,
+      "step": 8726
+    },
+    {
+      "epoch": 2.9784982935153583,
+      "grad_norm": 2.298532009124756,
+      "learning_rate": 7.167235494880547e-06,
+      "loss": 4.4423,
+      "step": 8727
+    },
+    {
+      "epoch": 2.9788395904436857,
+      "grad_norm": 2.4776766300201416,
+      "learning_rate": 7.053469852104664e-06,
+      "loss": 5.5068,
+      "step": 8728
+    },
+    {
+      "epoch": 2.9791808873720136,
+      "grad_norm": 2.411564588546753,
+      "learning_rate": 6.939704209328784e-06,
+      "loss": 5.7158,
+      "step": 8729
+    },
+    {
+      "epoch": 2.9795221843003414,
+      "grad_norm": 2.4252982139587402,
+      "learning_rate": 6.825938566552901e-06,
+      "loss": 4.9397,
+      "step": 8730
+    },
+    {
+      "epoch": 2.979863481228669,
+      "grad_norm": 2.472968339920044,
+      "learning_rate": 6.712172923777019e-06,
+      "loss": 4.6723,
+      "step": 8731
+    },
+    {
+      "epoch": 2.9802047781569967,
+      "grad_norm": 2.463764190673828,
+      "learning_rate": 6.598407281001138e-06,
+      "loss": 5.8935,
+      "step": 8732
+    },
+    {
+      "epoch": 2.980546075085324,
+      "grad_norm": 2.4861762523651123,
+      "learning_rate": 6.484641638225256e-06,
+      "loss": 5.6101,
+      "step": 8733
+    },
+    {
+      "epoch": 2.980887372013652,
+      "grad_norm": 2.5116894245147705,
+      "learning_rate": 6.370875995449374e-06,
+      "loss": 6.019,
+      "step": 8734
+    },
+    {
+      "epoch": 2.98122866894198,
+      "grad_norm": 2.589373826980591,
+      "learning_rate": 6.257110352673493e-06,
+      "loss": 5.3002,
+      "step": 8735
+    },
+    {
+      "epoch": 2.981569965870307,
+      "grad_norm": 2.5636868476867676,
+      "learning_rate": 6.1433447098976105e-06,
+      "loss": 6.0956,
+      "step": 8736
+    },
+    {
+      "epoch": 2.9819112627986346,
+      "grad_norm": 2.4157183170318604,
+      "learning_rate": 6.02957906712173e-06,
+      "loss": 6.1327,
+      "step": 8737
+    },
+    {
+      "epoch": 2.9822525597269625,
+      "grad_norm": 2.5133705139160156,
+      "learning_rate": 5.9158134243458475e-06,
+      "loss": 6.5907,
+      "step": 8738
+    },
+    {
+      "epoch": 2.9825938566552903,
+      "grad_norm": 2.6293768882751465,
+      "learning_rate": 5.802047781569966e-06,
+      "loss": 6.3252,
+      "step": 8739
+    },
+    {
+      "epoch": 2.9829351535836177,
+      "grad_norm": 2.530759334564209,
+      "learning_rate": 5.6882821387940845e-06,
+      "loss": 5.8731,
+      "step": 8740
+    },
+    {
+      "epoch": 2.983276450511945,
+      "grad_norm": 2.5379390716552734,
+      "learning_rate": 5.574516496018203e-06,
+      "loss": 5.9245,
+      "step": 8741
+    },
+    {
+      "epoch": 2.983617747440273,
+      "grad_norm": 2.5322394371032715,
+      "learning_rate": 5.460750853242321e-06,
+      "loss": 4.8229,
+      "step": 8742
+    },
+    {
+      "epoch": 2.983959044368601,
+      "grad_norm": 2.562993049621582,
+      "learning_rate": 5.346985210466439e-06,
+      "loss": 6.2476,
+      "step": 8743
+    },
+    {
+      "epoch": 2.9843003412969282,
+      "grad_norm": 2.5595932006835938,
+      "learning_rate": 5.233219567690558e-06,
+      "loss": 6.0822,
+      "step": 8744
+    },
+    {
+      "epoch": 2.984641638225256,
+      "grad_norm": 2.542302131652832,
+      "learning_rate": 5.119453924914675e-06,
+      "loss": 5.6779,
+      "step": 8745
+    },
+    {
+      "epoch": 2.9849829351535835,
+      "grad_norm": 2.3574278354644775,
+      "learning_rate": 5.005688282138795e-06,
+      "loss": 5.5255,
+      "step": 8746
+    },
+    {
+      "epoch": 2.9853242320819113,
+      "grad_norm": 2.4841222763061523,
+      "learning_rate": 4.891922639362912e-06,
+      "loss": 5.7245,
+      "step": 8747
+    },
+    {
+      "epoch": 2.985665529010239,
+      "grad_norm": 2.4711790084838867,
+      "learning_rate": 4.778156996587031e-06,
+      "loss": 6.1008,
+      "step": 8748
+    },
+    {
+      "epoch": 2.9860068259385666,
+      "grad_norm": 2.472411632537842,
+      "learning_rate": 4.664391353811149e-06,
+      "loss": 5.9053,
+      "step": 8749
+    },
+    {
+      "epoch": 2.986348122866894,
+      "grad_norm": 2.540469169616699,
+      "learning_rate": 4.550625711035267e-06,
+      "loss": 5.4633,
+      "step": 8750
+    },
+    {
+      "epoch": 2.986689419795222,
+      "grad_norm": 2.4350697994232178,
+      "learning_rate": 4.436860068259386e-06,
+      "loss": 5.7402,
+      "step": 8751
+    },
+    {
+      "epoch": 2.9870307167235497,
+      "grad_norm": 2.4693779945373535,
+      "learning_rate": 4.323094425483504e-06,
+      "loss": 5.52,
+      "step": 8752
+    },
+    {
+      "epoch": 2.987372013651877,
+      "grad_norm": 2.453641414642334,
+      "learning_rate": 4.2093287827076224e-06,
+      "loss": 6.0574,
+      "step": 8753
+    },
+    {
+      "epoch": 2.9877133105802045,
+      "grad_norm": 2.4429779052734375,
+      "learning_rate": 4.095563139931741e-06,
+      "loss": 5.5031,
+      "step": 8754
+    },
+    {
+      "epoch": 2.9880546075085324,
+      "grad_norm": 2.479944944381714,
+      "learning_rate": 3.9817974971558586e-06,
+      "loss": 5.6135,
+      "step": 8755
+    },
+    {
+      "epoch": 2.98839590443686,
+      "grad_norm": 2.4810147285461426,
+      "learning_rate": 3.868031854379978e-06,
+      "loss": 6.4582,
+      "step": 8756
+    },
+    {
+      "epoch": 2.9887372013651876,
+      "grad_norm": 2.5106160640716553,
+      "learning_rate": 3.7542662116040956e-06,
+      "loss": 6.1153,
+      "step": 8757
+    },
+    {
+      "epoch": 2.9890784982935155,
+      "grad_norm": 2.4401440620422363,
+      "learning_rate": 3.6405005688282136e-06,
+      "loss": 5.8295,
+      "step": 8758
+    },
+    {
+      "epoch": 2.989419795221843,
+      "grad_norm": 2.520796298980713,
+      "learning_rate": 3.526734926052332e-06,
+      "loss": 6.0053,
+      "step": 8759
+    },
+    {
+      "epoch": 2.9897610921501707,
+      "grad_norm": 2.531611204147339,
+      "learning_rate": 3.4129692832764506e-06,
+      "loss": 6.1409,
+      "step": 8760
+    },
+    {
+      "epoch": 2.9901023890784986,
+      "grad_norm": 2.4728920459747314,
+      "learning_rate": 3.299203640500569e-06,
+      "loss": 5.6671,
+      "step": 8761
+    },
+    {
+      "epoch": 2.990443686006826,
+      "grad_norm": 2.493472099304199,
+      "learning_rate": 3.185437997724687e-06,
+      "loss": 6.0225,
+      "step": 8762
+    },
+    {
+      "epoch": 2.9907849829351534,
+      "grad_norm": 2.4957869052886963,
+      "learning_rate": 3.0716723549488053e-06,
+      "loss": 6.2367,
+      "step": 8763
+    },
+    {
+      "epoch": 2.9911262798634812,
+      "grad_norm": 2.530949115753174,
+      "learning_rate": 2.9579067121729238e-06,
+      "loss": 5.9412,
+      "step": 8764
+    },
+    {
+      "epoch": 2.991467576791809,
+      "grad_norm": 2.5535237789154053,
+      "learning_rate": 2.8441410693970423e-06,
+      "loss": 5.8223,
+      "step": 8765
+    },
+    {
+      "epoch": 2.9918088737201365,
+      "grad_norm": 2.5407912731170654,
+      "learning_rate": 2.7303754266211603e-06,
+      "loss": 5.8673,
+      "step": 8766
+    },
+    {
+      "epoch": 2.992150170648464,
+      "grad_norm": 2.4849934577941895,
+      "learning_rate": 2.616609783845279e-06,
+      "loss": 6.2063,
+      "step": 8767
+    },
+    {
+      "epoch": 2.9924914675767917,
+      "grad_norm": 2.459456443786621,
+      "learning_rate": 2.5028441410693973e-06,
+      "loss": 5.7936,
+      "step": 8768
+    },
+    {
+      "epoch": 2.9928327645051196,
+      "grad_norm": 2.41933012008667,
+      "learning_rate": 2.3890784982935154e-06,
+      "loss": 6.1125,
+      "step": 8769
+    },
+    {
+      "epoch": 2.993174061433447,
+      "grad_norm": 2.5088438987731934,
+      "learning_rate": 2.2753128555176335e-06,
+      "loss": 5.7709,
+      "step": 8770
+    },
+    {
+      "epoch": 2.993515358361775,
+      "grad_norm": 2.481630563735962,
+      "learning_rate": 2.161547212741752e-06,
+      "loss": 5.1395,
+      "step": 8771
+    },
+    {
+      "epoch": 2.9938566552901023,
+      "grad_norm": 2.4694602489471436,
+      "learning_rate": 2.0477815699658705e-06,
+      "loss": 6.408,
+      "step": 8772
+    },
+    {
+      "epoch": 2.99419795221843,
+      "grad_norm": 2.609255313873291,
+      "learning_rate": 1.934015927189989e-06,
+      "loss": 5.2451,
+      "step": 8773
+    },
+    {
+      "epoch": 2.994539249146758,
+      "grad_norm": 2.562502145767212,
+      "learning_rate": 1.8202502844141068e-06,
+      "loss": 6.0563,
+      "step": 8774
+    },
+    {
+      "epoch": 2.9948805460750854,
+      "grad_norm": 2.526449203491211,
+      "learning_rate": 1.7064846416382253e-06,
+      "loss": 5.7829,
+      "step": 8775
+    },
+    {
+      "epoch": 2.9952218430034128,
+      "grad_norm": 2.5452816486358643,
+      "learning_rate": 1.5927189988623436e-06,
+      "loss": 6.2753,
+      "step": 8776
+    },
+    {
+      "epoch": 2.9955631399317406,
+      "grad_norm": 2.4952499866485596,
+      "learning_rate": 1.4789533560864619e-06,
+      "loss": 6.2114,
+      "step": 8777
+    },
+    {
+      "epoch": 2.9959044368600685,
+      "grad_norm": 2.254758358001709,
+      "learning_rate": 1.3651877133105802e-06,
+      "loss": 4.1049,
+      "step": 8778
+    },
+    {
+      "epoch": 2.996245733788396,
+      "grad_norm": 2.518864870071411,
+      "learning_rate": 1.2514220705346987e-06,
+      "loss": 6.5108,
+      "step": 8779
+    },
+    {
+      "epoch": 2.9965870307167233,
+      "grad_norm": 2.4607439041137695,
+      "learning_rate": 1.1376564277588167e-06,
+      "loss": 5.4417,
+      "step": 8780
+    },
+    {
+      "epoch": 2.996928327645051,
+      "grad_norm": 2.581373453140259,
+      "learning_rate": 1.0238907849829352e-06,
+      "loss": 5.4841,
+      "step": 8781
+    },
+    {
+      "epoch": 2.997269624573379,
+      "grad_norm": 2.3811047077178955,
+      "learning_rate": 9.101251422070534e-07,
+      "loss": 5.2381,
+      "step": 8782
+    },
+    {
+      "epoch": 2.9976109215017064,
+      "grad_norm": 2.4670047760009766,
+      "learning_rate": 7.963594994311718e-07,
+      "loss": 6.3373,
+      "step": 8783
+    },
+    {
+      "epoch": 2.9979522184300342,
+      "grad_norm": 2.462789297103882,
+      "learning_rate": 6.825938566552901e-07,
+      "loss": 5.4766,
+      "step": 8784
+    },
+    {
+      "epoch": 2.9982935153583616,
+      "grad_norm": 2.467465877532959,
+      "learning_rate": 5.688282138794084e-07,
+      "loss": 6.3852,
+      "step": 8785
+    },
+    {
+      "epoch": 2.9986348122866895,
+      "grad_norm": 2.4952261447906494,
+      "learning_rate": 4.550625711035267e-07,
+      "loss": 5.9034,
+      "step": 8786
+    },
+    {
+      "epoch": 2.9989761092150173,
+      "grad_norm": 2.497133731842041,
+      "learning_rate": 3.4129692832764504e-07,
+      "loss": 5.4161,
+      "step": 8787
+    },
+    {
+      "epoch": 2.9993174061433447,
+      "grad_norm": 2.377657890319824,
+      "learning_rate": 2.2753128555176335e-07,
+      "loss": 5.5288,
+      "step": 8788
+    },
+    {
+      "epoch": 2.999658703071672,
+      "grad_norm": 2.2658233642578125,
+      "learning_rate": 1.1376564277588168e-07,
+      "loss": 3.8865,
+      "step": 8789
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 2.476238965988159,
+      "learning_rate": 0.0,
+      "loss": 5.7151,
+      "step": 8790
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 8790,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 120000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 7.53227897518424e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}