{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.913043478260869,
  "eval_steps": 500,
  "global_step": 366,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.016304347826086956,
      "grad_norm": 45.78702163696289,
      "learning_rate": 5.0000000000000004e-08,
      "loss": 4.8845,
      "step": 1
    },
    {
      "epoch": 0.03260869565217391,
      "grad_norm": 45.26464080810547,
      "learning_rate": 1.0000000000000001e-07,
      "loss": 4.8307,
      "step": 2
    },
    {
      "epoch": 0.04891304347826087,
      "grad_norm": 45.96015167236328,
      "learning_rate": 1.5000000000000002e-07,
      "loss": 4.8322,
      "step": 3
    },
    {
      "epoch": 0.06521739130434782,
      "grad_norm": 45.86792755126953,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 4.8841,
      "step": 4
    },
    {
      "epoch": 0.08152173913043478,
      "grad_norm": 44.19926071166992,
      "learning_rate": 2.5000000000000004e-07,
      "loss": 4.8013,
      "step": 5
    },
    {
      "epoch": 0.09782608695652174,
      "grad_norm": 45.76038360595703,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 4.8112,
      "step": 6
    },
    {
      "epoch": 0.11413043478260869,
      "grad_norm": 45.73483657836914,
      "learning_rate": 3.5000000000000004e-07,
      "loss": 4.8235,
      "step": 7
    },
    {
      "epoch": 0.13043478260869565,
      "grad_norm": 46.2863655090332,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 4.895,
      "step": 8
    },
    {
      "epoch": 0.14673913043478262,
      "grad_norm": 44.06720733642578,
      "learning_rate": 4.5000000000000003e-07,
      "loss": 4.7696,
      "step": 9
    },
    {
      "epoch": 0.16304347826086957,
      "grad_norm": 43.8859977722168,
      "learning_rate": 5.000000000000001e-07,
      "loss": 4.7098,
      "step": 10
    },
    {
      "epoch": 0.1793478260869565,
      "grad_norm": 44.059043884277344,
      "learning_rate": 5.5e-07,
      "loss": 4.7519,
      "step": 11
    },
    {
      "epoch": 0.1956521739130435,
      "grad_norm": 43.980201721191406,
      "learning_rate": 6.000000000000001e-07,
      "loss": 4.7465,
      "step": 12
    },
    {
      "epoch": 0.21195652173913043,
      "grad_norm": 40.88735580444336,
      "learning_rate": 6.5e-07,
      "loss": 4.5416,
      "step": 13
    },
    {
      "epoch": 0.22826086956521738,
      "grad_norm": 39.40138244628906,
      "learning_rate": 7.000000000000001e-07,
      "loss": 4.5273,
      "step": 14
    },
    {
      "epoch": 0.24456521739130435,
      "grad_norm": 37.21616744995117,
      "learning_rate": 7.5e-07,
      "loss": 4.4206,
      "step": 15
    },
    {
      "epoch": 0.2608695652173913,
      "grad_norm": 35.060447692871094,
      "learning_rate": 8.000000000000001e-07,
      "loss": 4.3318,
      "step": 16
    },
    {
      "epoch": 0.27717391304347827,
      "grad_norm": 30.492183685302734,
      "learning_rate": 8.500000000000001e-07,
      "loss": 4.0841,
      "step": 17
    },
    {
      "epoch": 0.29347826086956524,
      "grad_norm": 28.49239730834961,
      "learning_rate": 9.000000000000001e-07,
      "loss": 4.0979,
      "step": 18
    },
    {
      "epoch": 0.30978260869565216,
      "grad_norm": 24.898632049560547,
      "learning_rate": 9.500000000000001e-07,
      "loss": 3.8752,
      "step": 19
    },
    {
      "epoch": 0.32608695652173914,
      "grad_norm": 22.521434783935547,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 3.77,
      "step": 20
    },
    {
      "epoch": 0.3423913043478261,
      "grad_norm": 20.311031341552734,
      "learning_rate": 1.0500000000000001e-06,
      "loss": 3.5973,
      "step": 21
    },
    {
      "epoch": 0.358695652173913,
      "grad_norm": 19.707839965820312,
      "learning_rate": 1.1e-06,
      "loss": 3.6102,
      "step": 22
    },
    {
      "epoch": 0.375,
      "grad_norm": 18.431489944458008,
      "learning_rate": 1.1500000000000002e-06,
      "loss": 3.4147,
      "step": 23
    },
    {
      "epoch": 0.391304347826087,
      "grad_norm": 17.78904151916504,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 3.2665,
      "step": 24
    },
    {
      "epoch": 0.4076086956521739,
      "grad_norm": 18.014617919921875,
      "learning_rate": 1.25e-06,
      "loss": 3.1941,
      "step": 25
    },
    {
      "epoch": 0.42391304347826086,
      "grad_norm": 17.8990478515625,
      "learning_rate": 1.3e-06,
      "loss": 3.0377,
      "step": 26
    },
    {
      "epoch": 0.44021739130434784,
      "grad_norm": 18.60847282409668,
      "learning_rate": 1.3500000000000002e-06,
      "loss": 2.8489,
      "step": 27
    },
    {
      "epoch": 0.45652173913043476,
      "grad_norm": 18.562536239624023,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 2.6229,
      "step": 28
    },
    {
      "epoch": 0.47282608695652173,
      "grad_norm": 19.00884246826172,
      "learning_rate": 1.45e-06,
      "loss": 2.4521,
      "step": 29
    },
    {
      "epoch": 0.4891304347826087,
      "grad_norm": 20.419940948486328,
      "learning_rate": 1.5e-06,
      "loss": 2.3907,
      "step": 30
    },
    {
      "epoch": 0.5054347826086957,
      "grad_norm": 21.181713104248047,
      "learning_rate": 1.5500000000000002e-06,
      "loss": 2.2032,
      "step": 31
    },
    {
      "epoch": 0.5217391304347826,
      "grad_norm": 20.592416763305664,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 2.004,
      "step": 32
    },
    {
      "epoch": 0.5380434782608695,
      "grad_norm": 20.40924835205078,
      "learning_rate": 1.6500000000000003e-06,
      "loss": 1.8373,
      "step": 33
    },
    {
      "epoch": 0.5543478260869565,
      "grad_norm": 18.38524627685547,
      "learning_rate": 1.7000000000000002e-06,
      "loss": 1.6218,
      "step": 34
    },
    {
      "epoch": 0.5706521739130435,
      "grad_norm": 18.922391891479492,
      "learning_rate": 1.75e-06,
      "loss": 1.4708,
      "step": 35
    },
    {
      "epoch": 0.5869565217391305,
      "grad_norm": 20.200698852539062,
      "learning_rate": 1.8000000000000001e-06,
      "loss": 1.3044,
      "step": 36
    },
    {
      "epoch": 0.6032608695652174,
      "grad_norm": 20.70162582397461,
      "learning_rate": 1.85e-06,
      "loss": 1.1445,
      "step": 37
    },
    {
      "epoch": 0.6195652173913043,
      "grad_norm": 18.18869400024414,
      "learning_rate": 1.9000000000000002e-06,
      "loss": 0.9504,
      "step": 38
    },
    {
      "epoch": 0.6358695652173914,
      "grad_norm": 15.211593627929688,
      "learning_rate": 1.9500000000000004e-06,
      "loss": 0.8083,
      "step": 39
    },
    {
      "epoch": 0.6521739130434783,
      "grad_norm": 12.71890640258789,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.6981,
      "step": 40
    },
    {
      "epoch": 0.6684782608695652,
      "grad_norm": 10.869053840637207,
      "learning_rate": 2.05e-06,
      "loss": 0.6018,
      "step": 41
    },
    {
      "epoch": 0.6847826086956522,
      "grad_norm": 9.5787353515625,
      "learning_rate": 2.1000000000000002e-06,
      "loss": 0.4899,
      "step": 42
    },
    {
      "epoch": 0.7010869565217391,
      "grad_norm": 9.319293975830078,
      "learning_rate": 2.15e-06,
      "loss": 0.3986,
      "step": 43
    },
    {
      "epoch": 0.717391304347826,
      "grad_norm": 9.404044151306152,
      "learning_rate": 2.2e-06,
      "loss": 0.3084,
      "step": 44
    },
    {
      "epoch": 0.7336956521739131,
      "grad_norm": 7.995025634765625,
      "learning_rate": 2.25e-06,
      "loss": 0.2373,
      "step": 45
    },
    {
      "epoch": 0.75,
      "grad_norm": 4.505397319793701,
      "learning_rate": 2.3000000000000004e-06,
      "loss": 0.1527,
      "step": 46
    },
    {
      "epoch": 0.7663043478260869,
      "grad_norm": 2.5063579082489014,
      "learning_rate": 2.35e-06,
      "loss": 0.1097,
      "step": 47
    },
    {
      "epoch": 0.782608695652174,
      "grad_norm": 1.5846028327941895,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.102,
      "step": 48
    },
    {
      "epoch": 0.7989130434782609,
      "grad_norm": 1.1286852359771729,
      "learning_rate": 2.4500000000000003e-06,
      "loss": 0.093,
      "step": 49
    },
    {
      "epoch": 0.8152173913043478,
      "grad_norm": 1.0311343669891357,
      "learning_rate": 2.5e-06,
      "loss": 0.0778,
      "step": 50
    },
    {
      "epoch": 0.8315217391304348,
      "grad_norm": 0.6458576917648315,
      "learning_rate": 2.55e-06,
      "loss": 0.0883,
      "step": 51
    },
    {
      "epoch": 0.8478260869565217,
      "grad_norm": 0.727554202079773,
      "learning_rate": 2.6e-06,
      "loss": 0.0852,
      "step": 52
    },
    {
      "epoch": 0.8641304347826086,
      "grad_norm": 0.619137167930603,
      "learning_rate": 2.6500000000000005e-06,
      "loss": 0.0697,
      "step": 53
    },
    {
      "epoch": 0.8804347826086957,
      "grad_norm": 0.38241881132125854,
      "learning_rate": 2.7000000000000004e-06,
      "loss": 0.0665,
      "step": 54
    },
    {
      "epoch": 0.8967391304347826,
      "grad_norm": 0.541621744632721,
      "learning_rate": 2.7500000000000004e-06,
      "loss": 0.0658,
      "step": 55
    },
    {
      "epoch": 0.9130434782608695,
      "grad_norm": 0.3869657516479492,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 0.0728,
      "step": 56
    },
    {
      "epoch": 0.9293478260869565,
      "grad_norm": 0.3768727481365204,
      "learning_rate": 2.85e-06,
      "loss": 0.0741,
      "step": 57
    },
    {
      "epoch": 0.9456521739130435,
      "grad_norm": 0.31400591135025024,
      "learning_rate": 2.9e-06,
      "loss": 0.0682,
      "step": 58
    },
    {
      "epoch": 0.9619565217391305,
      "grad_norm": 0.3604981303215027,
      "learning_rate": 2.95e-06,
      "loss": 0.0652,
      "step": 59
    },
    {
      "epoch": 0.9782608695652174,
      "grad_norm": 0.4383264482021332,
      "learning_rate": 3e-06,
      "loss": 0.0703,
      "step": 60
    },
    {
      "epoch": 0.9945652173913043,
      "grad_norm": 0.310332328081131,
      "learning_rate": 3.05e-06,
      "loss": 0.0664,
      "step": 61
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.310332328081131,
      "learning_rate": 3.1000000000000004e-06,
      "loss": 0.0509,
      "step": 62
    },
    {
      "epoch": 1.016304347826087,
      "grad_norm": 0.665212869644165,
      "learning_rate": 3.1500000000000003e-06,
      "loss": 0.0633,
      "step": 63
    },
    {
      "epoch": 1.0326086956521738,
      "grad_norm": 0.3108278512954712,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.0648,
      "step": 64
    },
    {
      "epoch": 1.048913043478261,
      "grad_norm": 0.5986258387565613,
      "learning_rate": 3.2500000000000002e-06,
      "loss": 0.0769,
      "step": 65
    },
    {
      "epoch": 1.065217391304348,
      "grad_norm": 0.6987417936325073,
      "learning_rate": 3.3000000000000006e-06,
      "loss": 0.0728,
      "step": 66
    },
    {
      "epoch": 1.0815217391304348,
      "grad_norm": 0.36069774627685547,
      "learning_rate": 3.3500000000000005e-06,
      "loss": 0.0647,
      "step": 67
    },
    {
      "epoch": 1.0978260869565217,
      "grad_norm": 0.37604954838752747,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 0.0571,
      "step": 68
    },
    {
      "epoch": 1.1141304347826086,
      "grad_norm": 0.2985791862010956,
      "learning_rate": 3.45e-06,
      "loss": 0.0629,
      "step": 69
    },
    {
      "epoch": 1.1304347826086956,
      "grad_norm": 0.3454388380050659,
      "learning_rate": 3.5e-06,
      "loss": 0.0644,
      "step": 70
    },
    {
      "epoch": 1.1467391304347827,
      "grad_norm": 0.3371462821960449,
      "learning_rate": 3.5500000000000003e-06,
      "loss": 0.0645,
      "step": 71
    },
    {
      "epoch": 1.1630434782608696,
      "grad_norm": 0.27834194898605347,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 0.0678,
      "step": 72
    },
    {
      "epoch": 1.1793478260869565,
      "grad_norm": 0.26285555958747864,
      "learning_rate": 3.65e-06,
      "loss": 0.0586,
      "step": 73
    },
    {
      "epoch": 1.1956521739130435,
      "grad_norm": 0.34152188897132874,
      "learning_rate": 3.7e-06,
      "loss": 0.061,
      "step": 74
    },
    {
      "epoch": 1.2119565217391304,
      "grad_norm": 0.2939279079437256,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.0643,
      "step": 75
    },
    {
      "epoch": 1.2282608695652173,
      "grad_norm": 0.395220547914505,
      "learning_rate": 3.8000000000000005e-06,
      "loss": 0.0644,
      "step": 76
    },
    {
      "epoch": 1.2445652173913044,
      "grad_norm": 0.29400259256362915,
      "learning_rate": 3.85e-06,
      "loss": 0.0562,
      "step": 77
    },
    {
      "epoch": 1.2608695652173914,
      "grad_norm": 0.25938117504119873,
      "learning_rate": 3.900000000000001e-06,
      "loss": 0.0658,
      "step": 78
    },
    {
      "epoch": 1.2771739130434783,
      "grad_norm": 0.36941587924957275,
      "learning_rate": 3.95e-06,
      "loss": 0.0678,
      "step": 79
    },
    {
      "epoch": 1.2934782608695652,
      "grad_norm": 0.26572781801223755,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.0588,
      "step": 80
    },
    {
      "epoch": 1.309782608695652,
      "grad_norm": 0.22964634001255035,
      "learning_rate": 4.05e-06,
      "loss": 0.0612,
      "step": 81
    },
    {
      "epoch": 1.3260869565217392,
      "grad_norm": 0.24455289542675018,
      "learning_rate": 4.1e-06,
      "loss": 0.059,
      "step": 82
    },
    {
      "epoch": 1.3423913043478262,
      "grad_norm": 0.3925253748893738,
      "learning_rate": 4.15e-06,
      "loss": 0.0648,
      "step": 83
    },
    {
      "epoch": 1.358695652173913,
      "grad_norm": 0.24822917580604553,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 0.0618,
      "step": 84
    },
    {
      "epoch": 1.375,
      "grad_norm": 0.2522635757923126,
      "learning_rate": 4.25e-06,
      "loss": 0.0568,
      "step": 85
    },
    {
      "epoch": 1.391304347826087,
      "grad_norm": 0.2465311586856842,
      "learning_rate": 4.3e-06,
      "loss": 0.0613,
      "step": 86
    },
    {
      "epoch": 1.4076086956521738,
      "grad_norm": 0.2514893412590027,
      "learning_rate": 4.350000000000001e-06,
      "loss": 0.0574,
      "step": 87
    },
    {
      "epoch": 1.4239130434782608,
      "grad_norm": 0.2320777177810669,
      "learning_rate": 4.4e-06,
      "loss": 0.0502,
      "step": 88
    },
    {
      "epoch": 1.440217391304348,
      "grad_norm": 0.2494516372680664,
      "learning_rate": 4.450000000000001e-06,
      "loss": 0.0685,
      "step": 89
    },
    {
      "epoch": 1.4565217391304348,
      "grad_norm": 0.244571715593338,
      "learning_rate": 4.5e-06,
      "loss": 0.0588,
      "step": 90
    },
    {
      "epoch": 1.4728260869565217,
      "grad_norm": 0.22765810787677765,
      "learning_rate": 4.5500000000000005e-06,
      "loss": 0.0591,
      "step": 91
    },
    {
      "epoch": 1.4891304347826086,
      "grad_norm": 0.2349582016468048,
      "learning_rate": 4.600000000000001e-06,
      "loss": 0.0572,
      "step": 92
    },
    {
      "epoch": 1.5054347826086958,
      "grad_norm": 0.228654682636261,
      "learning_rate": 4.65e-06,
      "loss": 0.0593,
      "step": 93
    },
    {
      "epoch": 1.5217391304347827,
      "grad_norm": 0.3090372681617737,
      "learning_rate": 4.7e-06,
      "loss": 0.0592,
      "step": 94
    },
    {
      "epoch": 1.5380434782608696,
      "grad_norm": 0.2116968333721161,
      "learning_rate": 4.75e-06,
      "loss": 0.0575,
      "step": 95
    },
    {
      "epoch": 1.5543478260869565,
      "grad_norm": 0.32294484972953796,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.062,
      "step": 96
    },
    {
      "epoch": 1.5706521739130435,
      "grad_norm": 0.35240596532821655,
      "learning_rate": 4.85e-06,
      "loss": 0.0561,
      "step": 97
    },
    {
      "epoch": 1.5869565217391304,
      "grad_norm": 0.24272935092449188,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 0.0591,
      "step": 98
    },
    {
      "epoch": 1.6032608695652173,
      "grad_norm": 0.26952850818634033,
      "learning_rate": 4.95e-06,
      "loss": 0.0557,
      "step": 99
    },
    {
      "epoch": 1.6195652173913042,
      "grad_norm": 0.2775789201259613,
      "learning_rate": 5e-06,
      "loss": 0.0624,
      "step": 100
    },
    {
      "epoch": 1.6358695652173914,
      "grad_norm": 0.2162039428949356,
      "learning_rate": 4.999825642177387e-06,
      "loss": 0.0582,
      "step": 101
    },
    {
      "epoch": 1.6521739130434783,
      "grad_norm": 0.38436195254325867,
      "learning_rate": 4.999302593030069e-06,
      "loss": 0.0618,
      "step": 102
    },
    {
      "epoch": 1.6684782608695652,
      "grad_norm": 0.2860764265060425,
      "learning_rate": 4.998430925516213e-06,
      "loss": 0.0637,
      "step": 103
    },
    {
      "epoch": 1.6847826086956523,
      "grad_norm": 0.2180222123861313,
      "learning_rate": 4.99721076122146e-06,
      "loss": 0.0565,
      "step": 104
    },
    {
      "epoch": 1.7010869565217392,
      "grad_norm": 0.30026087164878845,
      "learning_rate": 4.995642270341961e-06,
      "loss": 0.0615,
      "step": 105
    },
    {
      "epoch": 1.7173913043478262,
      "grad_norm": 0.27657821774482727,
      "learning_rate": 4.99372567166064e-06,
      "loss": 0.0661,
      "step": 106
    },
    {
      "epoch": 1.733695652173913,
      "grad_norm": 0.23324432969093323,
      "learning_rate": 4.991461232516675e-06,
      "loss": 0.0563,
      "step": 107
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.2670609951019287,
      "learning_rate": 4.98884926876821e-06,
      "loss": 0.0671,
      "step": 108
    },
    {
      "epoch": 1.766304347826087,
      "grad_norm": 0.2405894696712494,
      "learning_rate": 4.9858901447482924e-06,
      "loss": 0.0571,
      "step": 109
    },
    {
      "epoch": 1.7826086956521738,
      "grad_norm": 0.23271754384040833,
      "learning_rate": 4.982584273214061e-06,
      "loss": 0.0598,
      "step": 110
    },
    {
      "epoch": 1.7989130434782608,
      "grad_norm": 0.21662941575050354,
      "learning_rate": 4.978932115289165e-06,
      "loss": 0.0562,
      "step": 111
    },
    {
      "epoch": 1.8152173913043477,
      "grad_norm": 0.27656564116477966,
      "learning_rate": 4.974934180399447e-06,
      "loss": 0.0531,
      "step": 112
    },
    {
      "epoch": 1.8315217391304348,
      "grad_norm": 0.3287908434867859,
      "learning_rate": 4.970591026201884e-06,
      "loss": 0.0522,
      "step": 113
    },
    {
      "epoch": 1.8478260869565217,
      "grad_norm": 0.1879669576883316,
      "learning_rate": 4.965903258506806e-06,
      "loss": 0.0625,
      "step": 114
    },
    {
      "epoch": 1.8641304347826086,
      "grad_norm": 0.2798942029476166,
      "learning_rate": 4.9608715311933865e-06,
      "loss": 0.0546,
      "step": 115
    },
    {
      "epoch": 1.8804347826086958,
      "grad_norm": 0.31130167841911316,
      "learning_rate": 4.955496546118439e-06,
      "loss": 0.0576,
      "step": 116
    },
    {
      "epoch": 1.8967391304347827,
      "grad_norm": 0.23934200406074524,
      "learning_rate": 4.949779053018519e-06,
      "loss": 0.0522,
      "step": 117
    },
    {
      "epoch": 1.9130434782608696,
      "grad_norm": 0.2684226632118225,
      "learning_rate": 4.943719849405347e-06,
      "loss": 0.0596,
      "step": 118
    },
    {
      "epoch": 1.9293478260869565,
      "grad_norm": 0.31276896595954895,
      "learning_rate": 4.937319780454559e-06,
      "loss": 0.0609,
      "step": 119
    },
    {
      "epoch": 1.9456521739130435,
      "grad_norm": 0.2928497791290283,
      "learning_rate": 4.930579738887827e-06,
      "loss": 0.0515,
      "step": 120
    },
    {
      "epoch": 1.9619565217391304,
      "grad_norm": 0.25487199425697327,
      "learning_rate": 4.923500664848327e-06,
      "loss": 0.0565,
      "step": 121
    },
    {
      "epoch": 1.9782608695652173,
      "grad_norm": 0.19076956808567047,
      "learning_rate": 4.9160835457696075e-06,
      "loss": 0.0541,
      "step": 122
    },
    {
      "epoch": 1.9945652173913042,
      "grad_norm": 0.36202409863471985,
      "learning_rate": 4.9083294162378545e-06,
      "loss": 0.0572,
      "step": 123
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.36202409863471985,
      "learning_rate": 4.900239357847582e-06,
      "loss": 0.0578,
      "step": 124
    },
    {
      "epoch": 2.016304347826087,
      "grad_norm": 0.5418670177459717,
      "learning_rate": 4.891814499050762e-06,
      "loss": 0.0504,
      "step": 125
    },
    {
      "epoch": 2.032608695652174,
      "grad_norm": 0.24317026138305664,
      "learning_rate": 4.883056014999423e-06,
      "loss": 0.0557,
      "step": 126
    },
    {
      "epoch": 2.0489130434782608,
      "grad_norm": 0.23591496050357819,
      "learning_rate": 4.873965127381734e-06,
      "loss": 0.0567,
      "step": 127
    },
    {
      "epoch": 2.0652173913043477,
      "grad_norm": 0.26839157938957214,
      "learning_rate": 4.864543104251587e-06,
      "loss": 0.0489,
      "step": 128
    },
    {
      "epoch": 2.0815217391304346,
      "grad_norm": 0.23912706971168518,
      "learning_rate": 4.854791259851735e-06,
      "loss": 0.0519,
      "step": 129
    },
    {
      "epoch": 2.097826086956522,
      "grad_norm": 0.22615790367126465,
      "learning_rate": 4.844710954430464e-06,
      "loss": 0.0533,
      "step": 130
    },
    {
      "epoch": 2.114130434782609,
      "grad_norm": 0.18881212174892426,
      "learning_rate": 4.834303594051854e-06,
      "loss": 0.0509,
      "step": 131
    },
    {
      "epoch": 2.130434782608696,
      "grad_norm": 0.20288890600204468,
      "learning_rate": 4.823570630399665e-06,
      "loss": 0.0515,
      "step": 132
    },
    {
      "epoch": 2.1467391304347827,
      "grad_norm": 0.2495969533920288,
      "learning_rate": 4.812513560574832e-06,
      "loss": 0.0593,
      "step": 133
    },
    {
      "epoch": 2.1630434782608696,
      "grad_norm": 0.25076112151145935,
      "learning_rate": 4.8011339268866505e-06,
      "loss": 0.0546,
      "step": 134
    },
    {
      "epoch": 2.1793478260869565,
      "grad_norm": 0.26761433482170105,
      "learning_rate": 4.789433316637644e-06,
      "loss": 0.0494,
      "step": 135
    },
    {
      "epoch": 2.1956521739130435,
      "grad_norm": 0.22908158600330353,
      "learning_rate": 4.777413361902152e-06,
      "loss": 0.0556,
      "step": 136
    },
    {
      "epoch": 2.2119565217391304,
      "grad_norm": 0.24989894032478333,
      "learning_rate": 4.765075739298683e-06,
      "loss": 0.0457,
      "step": 137
    },
    {
      "epoch": 2.2282608695652173,
      "grad_norm": 0.19970546662807465,
      "learning_rate": 4.752422169756048e-06,
      "loss": 0.0483,
      "step": 138
    },
    {
      "epoch": 2.244565217391304,
      "grad_norm": 0.2547377347946167,
      "learning_rate": 4.739454418273314e-06,
      "loss": 0.0562,
      "step": 139
    },
    {
      "epoch": 2.260869565217391,
      "grad_norm": 0.30050888657569885,
      "learning_rate": 4.726174293673612e-06,
      "loss": 0.0474,
      "step": 140
    },
    {
      "epoch": 2.2771739130434785,
      "grad_norm": 0.20995593070983887,
      "learning_rate": 4.712583648351827e-06,
      "loss": 0.0486,
      "step": 141
    },
    {
      "epoch": 2.2934782608695654,
      "grad_norm": 0.24649159610271454,
      "learning_rate": 4.698684378016223e-06,
      "loss": 0.0521,
      "step": 142
    },
    {
      "epoch": 2.3097826086956523,
      "grad_norm": 0.23422546684741974,
      "learning_rate": 4.684478421424007e-06,
      "loss": 0.0521,
      "step": 143
    },
    {
      "epoch": 2.3260869565217392,
      "grad_norm": 0.21018289029598236,
      "learning_rate": 4.669967760110908e-06,
      "loss": 0.0503,
      "step": 144
    },
    {
      "epoch": 2.342391304347826,
      "grad_norm": 0.20515775680541992,
      "learning_rate": 4.655154418114774e-06,
      "loss": 0.0545,
      "step": 145
    },
    {
      "epoch": 2.358695652173913,
      "grad_norm": 0.2237028032541275,
      "learning_rate": 4.6400404616932505e-06,
      "loss": 0.052,
      "step": 146
    },
    {
      "epoch": 2.375,
      "grad_norm": 0.24619129300117493,
      "learning_rate": 4.624627999035564e-06,
      "loss": 0.059,
      "step": 147
    },
    {
      "epoch": 2.391304347826087,
      "grad_norm": 0.20323814451694489,
      "learning_rate": 4.608919179968457e-06,
      "loss": 0.0534,
      "step": 148
    },
    {
      "epoch": 2.407608695652174,
      "grad_norm": 0.21665000915527344,
      "learning_rate": 4.592916195656322e-06,
      "loss": 0.0505,
      "step": 149
    },
    {
      "epoch": 2.4239130434782608,
      "grad_norm": 0.21209590137004852,
      "learning_rate": 4.576621278295558e-06,
      "loss": 0.055,
      "step": 150
    },
    {
      "epoch": 2.4402173913043477,
      "grad_norm": 0.19232606887817383,
      "learning_rate": 4.5600367008032135e-06,
      "loss": 0.0533,
      "step": 151
    },
    {
      "epoch": 2.4565217391304346,
      "grad_norm": 0.22477513551712036,
      "learning_rate": 4.543164776499945e-06,
      "loss": 0.0544,
      "step": 152
    },
    {
      "epoch": 2.4728260869565215,
      "grad_norm": 0.21310777962207794,
      "learning_rate": 4.5260078587873416e-06,
      "loss": 0.0534,
      "step": 153
    },
    {
      "epoch": 2.489130434782609,
      "grad_norm": 0.2207050621509552,
      "learning_rate": 4.508568340819654e-06,
      "loss": 0.0516,
      "step": 154
    },
    {
      "epoch": 2.505434782608696,
      "grad_norm": 0.22430899739265442,
      "learning_rate": 4.490848655169986e-06,
      "loss": 0.0552,
      "step": 155
    },
    {
      "epoch": 2.5217391304347827,
      "grad_norm": 0.2665068805217743,
      "learning_rate": 4.472851273490985e-06,
      "loss": 0.0517,
      "step": 156
    },
    {
      "epoch": 2.5380434782608696,
      "grad_norm": 0.20904089510440826,
      "learning_rate": 4.454578706170075e-06,
      "loss": 0.0477,
      "step": 157
    },
    {
      "epoch": 2.5543478260869565,
      "grad_norm": 0.21974575519561768,
      "learning_rate": 4.436033501979299e-06,
      "loss": 0.0465,
      "step": 158
    },
    {
      "epoch": 2.5706521739130435,
      "grad_norm": 0.2737596333026886,
      "learning_rate": 4.417218247719794e-06,
      "loss": 0.0443,
      "step": 159
    },
    {
      "epoch": 2.5869565217391304,
      "grad_norm": 0.19713006913661957,
      "learning_rate": 4.398135567860972e-06,
      "loss": 0.0565,
      "step": 160
    },
    {
      "epoch": 2.6032608695652173,
      "grad_norm": 0.24970802664756775,
      "learning_rate": 4.378788124174441e-06,
      "loss": 0.0473,
      "step": 161
    },
    {
      "epoch": 2.619565217391304,
      "grad_norm": 0.20846787095069885,
      "learning_rate": 4.359178615362725e-06,
      "loss": 0.0511,
      "step": 162
    },
    {
      "epoch": 2.6358695652173916,
      "grad_norm": 0.33354219794273376,
      "learning_rate": 4.33930977668283e-06,
      "loss": 0.0541,
      "step": 163
    },
    {
      "epoch": 2.6521739130434785,
      "grad_norm": 0.23585955798625946,
      "learning_rate": 4.319184379564716e-06,
      "loss": 0.0534,
      "step": 164
    },
    {
      "epoch": 2.6684782608695654,
      "grad_norm": 0.26923519372940063,
      "learning_rate": 4.298805231224721e-06,
      "loss": 0.052,
      "step": 165
    },
    {
      "epoch": 2.6847826086956523,
      "grad_norm": 0.20466569066047668,
      "learning_rate": 4.278175174273989e-06,
      "loss": 0.0532,
      "step": 166
    },
    {
      "epoch": 2.7010869565217392,
      "grad_norm": 0.27908194065093994,
      "learning_rate": 4.257297086321967e-06,
      "loss": 0.0497,
      "step": 167
    },
    {
      "epoch": 2.717391304347826,
      "grad_norm": 0.29970937967300415,
      "learning_rate": 4.236173879575022e-06,
      "loss": 0.0502,
      "step": 168
    },
    {
      "epoch": 2.733695652173913,
      "grad_norm": 0.2322796732187271,
      "learning_rate": 4.2148085004302205e-06,
      "loss": 0.054,
      "step": 169
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.1984764039516449,
      "learning_rate": 4.1932039290643534e-06,
      "loss": 0.0491,
      "step": 170
    },
    {
      "epoch": 2.766304347826087,
      "grad_norm": 0.20822829008102417,
      "learning_rate": 4.1713631790182366e-06,
      "loss": 0.0555,
      "step": 171
    },
    {
      "epoch": 2.782608695652174,
      "grad_norm": 0.2537862956523895,
      "learning_rate": 4.149289296776369e-06,
      "loss": 0.0479,
      "step": 172
    },
    {
      "epoch": 2.7989130434782608,
      "grad_norm": 0.20386171340942383,
      "learning_rate": 4.126985361341984e-06,
      "loss": 0.049,
      "step": 173
    },
    {
      "epoch": 2.8152173913043477,
      "grad_norm": 0.28424644470214844,
      "learning_rate": 4.104454483807579e-06,
      "loss": 0.0481,
      "step": 174
    },
    {
      "epoch": 2.8315217391304346,
      "grad_norm": 0.2668056786060333,
      "learning_rate": 4.0816998069209516e-06,
      "loss": 0.0528,
      "step": 175
    },
    {
      "epoch": 2.8478260869565215,
      "grad_norm": 0.2771640419960022,
      "learning_rate": 4.058724504646834e-06,
      "loss": 0.0673,
      "step": 176
    },
    {
      "epoch": 2.8641304347826084,
      "grad_norm": 0.26568353176116943,
      "learning_rate": 4.0355317817241705e-06,
      "loss": 0.0438,
      "step": 177
    },
    {
      "epoch": 2.880434782608696,
      "grad_norm": 0.21058987081050873,
      "learning_rate": 4.012124873219094e-06,
      "loss": 0.0499,
      "step": 178
    },
    {
      "epoch": 2.8967391304347827,
      "grad_norm": 0.27742889523506165,
      "learning_rate": 3.988507044073687e-06,
      "loss": 0.047,
      "step": 179
    },
    {
      "epoch": 2.9130434782608696,
      "grad_norm": 0.21633900701999664,
      "learning_rate": 3.964681588650562e-06,
      "loss": 0.0485,
      "step": 180
    },
    {
      "epoch": 2.9293478260869565,
      "grad_norm": 0.21723268926143646,
      "learning_rate": 3.940651830273342e-06,
      "loss": 0.0535,
      "step": 181
    },
    {
      "epoch": 2.9456521739130435,
      "grad_norm": 0.21648381650447845,
      "learning_rate": 3.916421120763106e-06,
      "loss": 0.0469,
      "step": 182
    },
    {
      "epoch": 2.9619565217391304,
      "grad_norm": 0.2256888449192047,
      "learning_rate": 3.891992839970855e-06,
      "loss": 0.0469,
      "step": 183
    },
    {
      "epoch": 2.9782608695652173,
      "grad_norm": 0.3136005699634552,
      "learning_rate": 3.8673703953060685e-06,
      "loss": 0.0517,
      "step": 184
    },
    {
      "epoch": 2.994565217391304,
      "grad_norm": 0.2358432561159134,
      "learning_rate": 3.8425572212614155e-06,
      "loss": 0.0511,
      "step": 185
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.4989979565143585,
      "learning_rate": 3.817556778933697e-06,
      "loss": 0.0502,
      "step": 186
    },
    {
      "epoch": 3.016304347826087,
      "grad_norm": 0.24074898660182953,
      "learning_rate": 3.792372555541064e-06,
      "loss": 0.0446,
      "step": 187
    },
    {
      "epoch": 3.032608695652174,
      "grad_norm": 0.23359227180480957,
      "learning_rate": 3.7670080639366e-06,
      "loss": 0.0444,
      "step": 188
    },
    {
      "epoch": 3.0489130434782608,
      "grad_norm": 0.21096405386924744,
      "learning_rate": 3.741466842118327e-06,
      "loss": 0.0492,
      "step": 189
    },
    {
      "epoch": 3.0652173913043477,
      "grad_norm": 0.18678264319896698,
      "learning_rate": 3.7157524527357036e-06,
      "loss": 0.0458,
      "step": 190
    },
    {
      "epoch": 3.0815217391304346,
      "grad_norm": 0.21836838126182556,
      "learning_rate": 3.6898684825926845e-06,
      "loss": 0.0471,
      "step": 191
    },
    {
      "epoch": 3.097826086956522,
      "grad_norm": 0.21061812341213226,
      "learning_rate": 3.663818542147409e-06,
      "loss": 0.0469,
      "step": 192
    },
    {
      "epoch": 3.114130434782609,
      "grad_norm": 0.2022838145494461,
      "learning_rate": 3.6376062650085918e-06,
      "loss": 0.0457,
      "step": 193
    },
    {
      "epoch": 3.130434782608696,
      "grad_norm": 0.21767382323741913,
      "learning_rate": 3.61123530742869e-06,
      "loss": 0.0468,
      "step": 194
    },
    {
      "epoch": 3.1467391304347827,
      "grad_norm": 0.2330484390258789,
      "learning_rate": 3.5847093477938955e-06,
      "loss": 0.0437,
      "step": 195
    },
    {
      "epoch": 3.1630434782608696,
      "grad_norm": 0.21127860248088837,
      "learning_rate": 3.5580320861110627e-06,
      "loss": 0.0459,
      "step": 196
    },
    {
      "epoch": 3.1793478260869565,
      "grad_norm": 0.20847401022911072,
      "learning_rate": 3.5312072434915983e-06,
      "loss": 0.0451,
      "step": 197
    },
    {
      "epoch": 3.1956521739130435,
      "grad_norm": 0.20767691731452942,
      "learning_rate": 3.5042385616324243e-06,
      "loss": 0.0434,
      "step": 198
    },
    {
      "epoch": 3.2119565217391304,
      "grad_norm": 0.21209655702114105,
      "learning_rate": 3.477129802294057e-06,
      "loss": 0.0418,
      "step": 199
    },
    {
      "epoch": 3.2282608695652173,
      "grad_norm": 0.21463370323181152,
      "learning_rate": 3.4498847467759e-06,
      "loss": 0.0408,
      "step": 200
    },
    {
      "epoch": 3.244565217391304,
      "grad_norm": 0.19667920470237732,
      "learning_rate": 3.4225071953887977e-06,
      "loss": 0.0476,
      "step": 201
    },
    {
      "epoch": 3.260869565217391,
      "grad_norm": 0.2126225084066391,
      "learning_rate": 3.3950009669249502e-06,
      "loss": 0.0438,
      "step": 202
    },
    {
      "epoch": 3.2771739130434785,
      "grad_norm": 0.21141058206558228,
      "learning_rate": 3.3673698981252385e-06,
      "loss": 0.0443,
      "step": 203
    },
    {
      "epoch": 3.2934782608695654,
      "grad_norm": 0.3029347360134125,
      "learning_rate": 3.3396178431440572e-06,
      "loss": 0.0549,
      "step": 204
    },
    {
      "epoch": 3.3097826086956523,
      "grad_norm": 0.20408247411251068,
      "learning_rate": 3.3117486730117092e-06,
      "loss": 0.0441,
      "step": 205
    },
    {
      "epoch": 3.3260869565217392,
      "grad_norm": 0.19485828280448914,
      "learning_rate": 3.283766275094454e-06,
      "loss": 0.0484,
      "step": 206
    },
    {
      "epoch": 3.342391304347826,
      "grad_norm": 0.23771652579307556,
      "learning_rate": 3.255674552552267e-06,
      "loss": 0.0478,
      "step": 207
    },
    {
      "epoch": 3.358695652173913,
      "grad_norm": 0.21850208938121796,
      "learning_rate": 3.227477423794412e-06,
      "loss": 0.0487,
      "step": 208
    },
    {
      "epoch": 3.375,
      "grad_norm": 0.23703360557556152,
      "learning_rate": 3.1991788219328657e-06,
      "loss": 0.0505,
      "step": 209
    },
    {
      "epoch": 3.391304347826087,
      "grad_norm": 0.21323414146900177,
      "learning_rate": 3.1707826942337124e-06,
      "loss": 0.0398,
      "step": 210
    },
    {
      "epoch": 3.407608695652174,
      "grad_norm": 0.28237420320510864,
      "learning_rate": 3.142293001566548e-06,
      "loss": 0.0429,
      "step": 211
    },
    {
      "epoch": 3.4239130434782608,
      "grad_norm": 0.24118061363697052,
      "learning_rate": 3.1137137178519983e-06,
      "loss": 0.0441,
      "step": 212
    },
    {
      "epoch": 3.4402173913043477,
      "grad_norm": 0.20298974215984344,
      "learning_rate": 3.085048829507406e-06,
      "loss": 0.0429,
      "step": 213
    },
    {
      "epoch": 3.4565217391304346,
      "grad_norm": 0.2397722750902176,
      "learning_rate": 3.056302334890786e-06,
      "loss": 0.0487,
      "step": 214
    },
    {
      "epoch": 3.4728260869565215,
      "grad_norm": 0.24425837397575378,
      "learning_rate": 3.027478243743106e-06,
      "loss": 0.0465,
      "step": 215
    },
    {
      "epoch": 3.489130434782609,
      "grad_norm": 0.2465757578611374,
      "learning_rate": 2.9985805766289815e-06,
      "loss": 0.0415,
      "step": 216
    },
    {
      "epoch": 3.505434782608696,
      "grad_norm": 0.2629953920841217,
      "learning_rate": 2.9696133643758663e-06,
      "loss": 0.0456,
      "step": 217
    },
    {
      "epoch": 3.5217391304347827,
      "grad_norm": 0.23994584381580353,
      "learning_rate": 2.940580647511805e-06,
      "loss": 0.0467,
      "step": 218
    },
    {
      "epoch": 3.5380434782608696,
      "grad_norm": 0.29615387320518494,
      "learning_rate": 2.911486475701835e-06,
      "loss": 0.0468,
      "step": 219
    },
    {
      "epoch": 3.5543478260869565,
      "grad_norm": 0.28419968485832214,
      "learning_rate": 2.8823349071831154e-06,
      "loss": 0.0479,
      "step": 220
    },
    {
      "epoch": 3.5706521739130435,
      "grad_norm": 0.2224627584218979,
      "learning_rate": 2.853130008198855e-06,
      "loss": 0.0436,
      "step": 221
    },
    {
      "epoch": 3.5869565217391304,
      "grad_norm": 0.2682191729545593,
      "learning_rate": 2.8238758524311316e-06,
      "loss": 0.0439,
      "step": 222
    },
    {
      "epoch": 3.6032608695652173,
      "grad_norm": 0.2094065397977829,
      "learning_rate": 2.7945765204326664e-06,
      "loss": 0.0472,
      "step": 223
    },
    {
      "epoch": 3.619565217391304,
      "grad_norm": 0.22561469674110413,
      "learning_rate": 2.7652360990576457e-06,
      "loss": 0.0426,
      "step": 224
    },
    {
      "epoch": 3.6358695652173916,
      "grad_norm": 0.1920589804649353,
      "learning_rate": 2.735858680891656e-06,
      "loss": 0.0465,
      "step": 225
    },
    {
      "epoch": 3.6521739130434785,
      "grad_norm": 0.2274886816740036,
      "learning_rate": 2.7064483636808314e-06,
      "loss": 0.0471,
      "step": 226
    },
    {
      "epoch": 3.6684782608695654,
      "grad_norm": 0.27154678106307983,
      "learning_rate": 2.677009249760268e-06,
      "loss": 0.0537,
      "step": 227
    },
    {
      "epoch": 3.6847826086956523,
      "grad_norm": 0.19993992149829865,
      "learning_rate": 2.6475454454818072e-06,
      "loss": 0.0423,
      "step": 228
    },
    {
      "epoch": 3.7010869565217392,
      "grad_norm": 0.26778337359428406,
      "learning_rate": 2.6180610606412587e-06,
      "loss": 0.0502,
      "step": 229
    },
    {
      "epoch": 3.717391304347826,
      "grad_norm": 0.22767065465450287,
      "learning_rate": 2.5885602079051354e-06,
      "loss": 0.0447,
      "step": 230
    },
    {
      "epoch": 3.733695652173913,
      "grad_norm": 0.2405000478029251,
      "learning_rate": 2.559047002236995e-06,
      "loss": 0.0463,
      "step": 231
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.23877452313899994,
      "learning_rate": 2.529525560323462e-06,
      "loss": 0.0472,
      "step": 232
    },
    {
      "epoch": 3.766304347826087,
      "grad_norm": 0.2590119540691376,
      "learning_rate": 2.5e-06,
      "loss": 0.0474,
      "step": 233
    },
    {
      "epoch": 3.782608695652174,
      "grad_norm": 0.2477138489484787,
      "learning_rate": 2.470474439676539e-06,
      "loss": 0.0461,
      "step": 234
    },
    {
      "epoch": 3.7989130434782608,
      "grad_norm": 0.23399893939495087,
      "learning_rate": 2.4409529977630052e-06,
      "loss": 0.0446,
      "step": 235
    },
    {
      "epoch": 3.8152173913043477,
      "grad_norm": 0.24198178946971893,
      "learning_rate": 2.411439792094866e-06,
      "loss": 0.0482,
      "step": 236
    },
    {
      "epoch": 3.8315217391304346,
      "grad_norm": 0.22425131499767303,
      "learning_rate": 2.381938939358742e-06,
      "loss": 0.0477,
      "step": 237
    },
    {
      "epoch": 3.8478260869565215,
      "grad_norm": 0.21231728792190552,
      "learning_rate": 2.3524545545181936e-06,
      "loss": 0.0437,
      "step": 238
    },
    {
      "epoch": 3.8641304347826084,
      "grad_norm": 0.24554376304149628,
      "learning_rate": 2.322990750239733e-06,
      "loss": 0.0408,
      "step": 239
    },
    {
      "epoch": 3.880434782608696,
      "grad_norm": 0.20181581377983093,
      "learning_rate": 2.2935516363191695e-06,
      "loss": 0.0465,
      "step": 240
    },
    {
      "epoch": 3.8967391304347827,
      "grad_norm": 0.23197294771671295,
      "learning_rate": 2.2641413191083445e-06,
      "loss": 0.0437,
      "step": 241
    },
    {
      "epoch": 3.9130434782608696,
      "grad_norm": 0.18796613812446594,
      "learning_rate": 2.234763900942355e-06,
      "loss": 0.0427,
      "step": 242
    },
    {
      "epoch": 3.9293478260869565,
      "grad_norm": 0.2553354799747467,
      "learning_rate": 2.2054234795673336e-06,
      "loss": 0.0526,
      "step": 243
    },
    {
      "epoch": 3.9456521739130435,
      "grad_norm": 0.20340357720851898,
      "learning_rate": 2.1761241475688697e-06,
      "loss": 0.0405,
      "step": 244
    },
    {
      "epoch": 3.9619565217391304,
      "grad_norm": 0.2950204908847809,
      "learning_rate": 2.146869991801146e-06,
      "loss": 0.0469,
      "step": 245
    },
    {
      "epoch": 3.9782608695652173,
      "grad_norm": 0.226835697889328,
      "learning_rate": 2.117665092816885e-06,
      "loss": 0.0489,
      "step": 246
    },
    {
      "epoch": 3.994565217391304,
      "grad_norm": 0.2127283215522766,
      "learning_rate": 2.088513524298165e-06,
      "loss": 0.0444,
      "step": 247
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.2127283215522766,
      "learning_rate": 2.059419352488196e-06,
      "loss": 0.0514,
      "step": 248
    },
    {
      "epoch": 4.016304347826087,
      "grad_norm": 0.5015725493431091,
      "learning_rate": 2.030386635624135e-06,
      "loss": 0.0402,
      "step": 249
    },
    {
      "epoch": 4.032608695652174,
      "grad_norm": 0.21895018219947815,
      "learning_rate": 2.0014194233710193e-06,
      "loss": 0.0427,
      "step": 250
    },
    {
      "epoch": 4.048913043478261,
      "grad_norm": 0.2607857584953308,
      "learning_rate": 1.972521756256895e-06,
      "loss": 0.0352,
      "step": 251
    },
    {
      "epoch": 4.065217391304348,
      "grad_norm": 0.2097238153219223,
      "learning_rate": 1.9436976651092143e-06,
      "loss": 0.0467,
      "step": 252
    },
    {
      "epoch": 4.081521739130435,
      "grad_norm": 0.21395832300186157,
      "learning_rate": 1.9149511704925945e-06,
      "loss": 0.0436,
      "step": 253
    },
    {
      "epoch": 4.0978260869565215,
      "grad_norm": 0.24933360517024994,
      "learning_rate": 1.8862862821480023e-06,
      "loss": 0.0409,
      "step": 254
    },
    {
      "epoch": 4.114130434782608,
      "grad_norm": 0.24037982523441315,
      "learning_rate": 1.8577069984334522e-06,
      "loss": 0.0445,
      "step": 255
    },
    {
      "epoch": 4.130434782608695,
      "grad_norm": 0.22139646112918854,
      "learning_rate": 1.829217305766289e-06,
      "loss": 0.041,
      "step": 256
    },
    {
      "epoch": 4.146739130434782,
      "grad_norm": 0.23465107381343842,
      "learning_rate": 1.8008211780671353e-06,
      "loss": 0.0456,
      "step": 257
    },
    {
      "epoch": 4.163043478260869,
      "grad_norm": 0.23133054375648499,
      "learning_rate": 1.772522576205589e-06,
      "loss": 0.0428,
      "step": 258
    },
    {
      "epoch": 4.179347826086956,
      "grad_norm": 0.25171801447868347,
      "learning_rate": 1.7443254474477328e-06,
      "loss": 0.0387,
      "step": 259
    },
    {
      "epoch": 4.195652173913044,
      "grad_norm": 0.22239144146442413,
      "learning_rate": 1.7162337249055478e-06,
      "loss": 0.0417,
      "step": 260
    },
    {
      "epoch": 4.211956521739131,
      "grad_norm": 0.24063155055046082,
      "learning_rate": 1.6882513269882916e-06,
      "loss": 0.0451,
      "step": 261
    },
    {
      "epoch": 4.228260869565218,
      "grad_norm": 0.24732705950737,
      "learning_rate": 1.6603821568559436e-06,
      "loss": 0.0428,
      "step": 262
    },
    {
      "epoch": 4.244565217391305,
      "grad_norm": 0.2546471953392029,
      "learning_rate": 1.6326301018747623e-06,
      "loss": 0.0378,
      "step": 263
    },
    {
      "epoch": 4.260869565217392,
      "grad_norm": 0.2471722811460495,
      "learning_rate": 1.6049990330750508e-06,
      "loss": 0.0401,
      "step": 264
    },
    {
      "epoch": 4.2771739130434785,
      "grad_norm": 0.20250661671161652,
      "learning_rate": 1.5774928046112025e-06,
      "loss": 0.0404,
      "step": 265
    },
    {
      "epoch": 4.293478260869565,
      "grad_norm": 0.21679657697677612,
      "learning_rate": 1.5501152532241003e-06,
      "loss": 0.0432,
      "step": 266
    },
    {
      "epoch": 4.309782608695652,
      "grad_norm": 0.25477948784828186,
      "learning_rate": 1.522870197705943e-06,
      "loss": 0.0484,
      "step": 267
    },
    {
      "epoch": 4.326086956521739,
      "grad_norm": 0.23823678493499756,
      "learning_rate": 1.495761438367577e-06,
      "loss": 0.038,
      "step": 268
    },
    {
      "epoch": 4.342391304347826,
      "grad_norm": 0.19264428317546844,
      "learning_rate": 1.4687927565084023e-06,
      "loss": 0.0463,
      "step": 269
    },
    {
      "epoch": 4.358695652173913,
      "grad_norm": 0.2232806533575058,
      "learning_rate": 1.4419679138889379e-06,
      "loss": 0.0367,
      "step": 270
    },
    {
      "epoch": 4.375,
      "grad_norm": 0.227358877658844,
      "learning_rate": 1.415290652206105e-06,
      "loss": 0.0393,
      "step": 271
    },
    {
      "epoch": 4.391304347826087,
      "grad_norm": 0.21266399323940277,
      "learning_rate": 1.3887646925713116e-06,
      "loss": 0.0427,
      "step": 272
    },
    {
      "epoch": 4.407608695652174,
      "grad_norm": 0.219986230134964,
      "learning_rate": 1.3623937349914093e-06,
      "loss": 0.0391,
      "step": 273
    },
    {
      "epoch": 4.423913043478261,
      "grad_norm": 0.20911090075969696,
      "learning_rate": 1.3361814578525922e-06,
      "loss": 0.0414,
      "step": 274
    },
    {
      "epoch": 4.440217391304348,
      "grad_norm": 0.20766796171665192,
      "learning_rate": 1.3101315174073162e-06,
      "loss": 0.0427,
      "step": 275
    },
    {
      "epoch": 4.456521739130435,
      "grad_norm": 0.2213449478149414,
      "learning_rate": 1.2842475472642969e-06,
      "loss": 0.0412,
      "step": 276
    },
    {
      "epoch": 4.4728260869565215,
      "grad_norm": 0.21896588802337646,
      "learning_rate": 1.258533157881674e-06,
      "loss": 0.0419,
      "step": 277
    },
    {
      "epoch": 4.489130434782608,
      "grad_norm": 0.23026618361473083,
      "learning_rate": 1.2329919360634003e-06,
      "loss": 0.0391,
      "step": 278
    },
    {
      "epoch": 4.505434782608695,
      "grad_norm": 0.2366199791431427,
      "learning_rate": 1.2076274444589361e-06,
      "loss": 0.0451,
      "step": 279
    },
    {
      "epoch": 4.521739130434782,
      "grad_norm": 0.21407200396060944,
      "learning_rate": 1.182443221066303e-06,
      "loss": 0.0445,
      "step": 280
    },
    {
      "epoch": 4.538043478260869,
      "grad_norm": 0.22052663564682007,
      "learning_rate": 1.1574427787385853e-06,
      "loss": 0.0454,
      "step": 281
    },
    {
      "epoch": 4.554347826086957,
      "grad_norm": 0.23922421038150787,
      "learning_rate": 1.1326296046939334e-06,
      "loss": 0.0386,
      "step": 282
    },
    {
      "epoch": 4.570652173913043,
      "grad_norm": 0.21938639879226685,
      "learning_rate": 1.1080071600291453e-06,
      "loss": 0.0424,
      "step": 283
    },
    {
      "epoch": 4.586956521739131,
      "grad_norm": 0.23158493638038635,
      "learning_rate": 1.083578879236895e-06,
      "loss": 0.0425,
      "step": 284
    },
    {
      "epoch": 4.603260869565218,
      "grad_norm": 0.22377173602581024,
      "learning_rate": 1.0593481697266582e-06,
      "loss": 0.0431,
      "step": 285
    },
    {
      "epoch": 4.619565217391305,
      "grad_norm": 0.21283438801765442,
      "learning_rate": 1.0353184113494386e-06,
      "loss": 0.0383,
      "step": 286
    },
    {
      "epoch": 4.635869565217392,
      "grad_norm": 0.24626271426677704,
      "learning_rate": 1.0114929559263122e-06,
      "loss": 0.04,
      "step": 287
    },
    {
      "epoch": 4.6521739130434785,
      "grad_norm": 0.23055487871170044,
      "learning_rate": 9.878751267809069e-07,
      "loss": 0.0438,
      "step": 288
    },
    {
      "epoch": 4.668478260869565,
      "grad_norm": 0.21289804577827454,
      "learning_rate": 9.644682182758305e-07,
      "loss": 0.044,
      "step": 289
    },
    {
      "epoch": 4.684782608695652,
      "grad_norm": 0.25633224844932556,
      "learning_rate": 9.412754953531664e-07,
      "loss": 0.0387,
      "step": 290
    },
    {
      "epoch": 4.701086956521739,
      "grad_norm": 0.21574127674102783,
      "learning_rate": 9.183001930790483e-07,
      "loss": 0.0408,
      "step": 291
    },
    {
      "epoch": 4.717391304347826,
      "grad_norm": 0.2526251971721649,
      "learning_rate": 8.955455161924217e-07,
      "loss": 0.0431,
      "step": 292
    },
    {
      "epoch": 4.733695652173913,
      "grad_norm": 0.2219126969575882,
      "learning_rate": 8.730146386580157e-07,
      "loss": 0.0382,
      "step": 293
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.22477059066295624,
      "learning_rate": 8.507107032236323e-07,
      "loss": 0.0387,
      "step": 294
    },
    {
      "epoch": 4.766304347826087,
      "grad_norm": 0.23549960553646088,
      "learning_rate": 8.286368209817644e-07,
      "loss": 0.042,
      "step": 295
    },
    {
      "epoch": 4.782608695652174,
      "grad_norm": 0.24367327988147736,
      "learning_rate": 8.067960709356479e-07,
      "loss": 0.0405,
      "step": 296
    },
    {
      "epoch": 4.798913043478261,
      "grad_norm": 0.25587689876556396,
      "learning_rate": 7.851914995697801e-07,
      "loss": 0.0416,
      "step": 297
    },
    {
      "epoch": 4.815217391304348,
      "grad_norm": 0.270684152841568,
      "learning_rate": 7.638261204249783e-07,
      "loss": 0.0394,
      "step": 298
    },
    {
      "epoch": 4.831521739130435,
      "grad_norm": 0.22876906394958496,
      "learning_rate": 7.427029136780333e-07,
      "loss": 0.0369,
      "step": 299
    },
    {
      "epoch": 4.8478260869565215,
      "grad_norm": 0.29206568002700806,
      "learning_rate": 7.218248257260127e-07,
      "loss": 0.0503,
      "step": 300
    },
    {
      "epoch": 4.864130434782608,
      "grad_norm": 0.22232329845428467,
      "learning_rate": 7.011947687752804e-07,
      "loss": 0.0407,
      "step": 301
    },
    {
      "epoch": 4.880434782608695,
      "grad_norm": 0.2493494153022766,
      "learning_rate": 6.808156204352845e-07,
      "loss": 0.0394,
      "step": 302
    },
    {
      "epoch": 4.896739130434782,
      "grad_norm": 0.20831190049648285,
      "learning_rate": 6.60690223317171e-07,
      "loss": 0.0406,
      "step": 303
    },
    {
      "epoch": 4.913043478260869,
      "grad_norm": 0.22676558792591095,
      "learning_rate": 6.40821384637276e-07,
      "loss": 0.0365,
      "step": 304
    },
    {
      "epoch": 4.929347826086957,
      "grad_norm": 0.2518041133880615,
      "learning_rate": 6.212118758255595e-07,
      "loss": 0.0475,
      "step": 305
    },
    {
      "epoch": 4.945652173913043,
      "grad_norm": 0.23944813013076782,
      "learning_rate": 6.018644321390288e-07,
      "loss": 0.042,
      "step": 306
    },
    {
      "epoch": 4.961956521739131,
      "grad_norm": 0.2410481721162796,
      "learning_rate": 5.827817522802065e-07,
      "loss": 0.0389,
      "step": 307
    },
    {
      "epoch": 4.978260869565218,
      "grad_norm": 0.2525772750377655,
      "learning_rate": 5.639664980207024e-07,
      "loss": 0.0406,
      "step": 308
    },
    {
      "epoch": 4.994565217391305,
      "grad_norm": 0.276832640171051,
      "learning_rate": 5.454212938299256e-07,
      "loss": 0.0439,
      "step": 309
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.276832640171051,
      "learning_rate": 5.271487265090163e-07,
      "loss": 0.0354,
      "step": 310
    },
    {
      "epoch": 5.016304347826087,
      "grad_norm": 0.5199080109596252,
      "learning_rate": 5.091513448300142e-07,
      "loss": 0.0407,
      "step": 311
    },
    {
      "epoch": 5.032608695652174,
      "grad_norm": 0.270175963640213,
      "learning_rate": 4.914316591803475e-07,
      "loss": 0.0402,
      "step": 312
    },
    {
      "epoch": 5.048913043478261,
      "grad_norm": 0.22836650907993317,
      "learning_rate": 4.739921412126591e-07,
      "loss": 0.0382,
      "step": 313
    },
    {
      "epoch": 5.065217391304348,
      "grad_norm": 0.22557897865772247,
      "learning_rate": 4.5683522350005505e-07,
      "loss": 0.0388,
      "step": 314
    },
    {
      "epoch": 5.081521739130435,
      "grad_norm": 0.23105387389659882,
      "learning_rate": 4.399632991967867e-07,
      "loss": 0.0362,
      "step": 315
    },
    {
      "epoch": 5.0978260869565215,
      "grad_norm": 0.26361796259880066,
      "learning_rate": 4.23378721704443e-07,
      "loss": 0.0466,
      "step": 316
    },
    {
      "epoch": 5.114130434782608,
      "grad_norm": 0.2395378202199936,
      "learning_rate": 4.070838043436787e-07,
      "loss": 0.041,
      "step": 317
    },
    {
      "epoch": 5.130434782608695,
      "grad_norm": 0.20131447911262512,
      "learning_rate": 3.910808200315433e-07,
      "loss": 0.0388,
      "step": 318
    },
    {
      "epoch": 5.146739130434782,
      "grad_norm": 0.2295176386833191,
      "learning_rate": 3.753720009644371e-07,
      "loss": 0.037,
      "step": 319
    },
    {
      "epoch": 5.163043478260869,
      "grad_norm": 0.24505485594272614,
      "learning_rate": 3.5995953830675004e-07,
      "loss": 0.043,
      "step": 320
    },
    {
      "epoch": 5.179347826086956,
      "grad_norm": 0.2312752902507782,
      "learning_rate": 3.448455818852267e-07,
      "loss": 0.0369,
      "step": 321
    },
    {
      "epoch": 5.195652173913044,
      "grad_norm": 0.21873357892036438,
      "learning_rate": 3.3003223988909234e-07,
      "loss": 0.0398,
      "step": 322
    },
    {
      "epoch": 5.211956521739131,
      "grad_norm": 0.22271527349948883,
      "learning_rate": 3.1552157857599327e-07,
      "loss": 0.0377,
      "step": 323
    },
    {
      "epoch": 5.228260869565218,
      "grad_norm": 0.22470292448997498,
      "learning_rate": 3.0131562198377763e-07,
      "loss": 0.0411,
      "step": 324
    },
    {
      "epoch": 5.244565217391305,
      "grad_norm": 0.22748963534832,
      "learning_rate": 2.874163516481732e-07,
      "loss": 0.0346,
      "step": 325
    },
    {
      "epoch": 5.260869565217392,
      "grad_norm": 0.27392780780792236,
      "learning_rate": 2.7382570632638853e-07,
      "loss": 0.0426,
      "step": 326
    },
    {
      "epoch": 5.2771739130434785,
      "grad_norm": 0.2509923577308655,
      "learning_rate": 2.605455817266861e-07,
      "loss": 0.0375,
      "step": 327
    },
    {
      "epoch": 5.293478260869565,
      "grad_norm": 0.20978674292564392,
      "learning_rate": 2.4757783024395244e-07,
      "loss": 0.0333,
      "step": 328
    },
    {
      "epoch": 5.309782608695652,
      "grad_norm": 0.23130132257938385,
      "learning_rate": 2.3492426070131746e-07,
      "loss": 0.0429,
      "step": 329
    },
    {
      "epoch": 5.326086956521739,
      "grad_norm": 0.2261745035648346,
      "learning_rate": 2.2258663809784892e-07,
      "loss": 0.0376,
      "step": 330
    },
    {
      "epoch": 5.342391304347826,
      "grad_norm": 0.21033048629760742,
      "learning_rate": 2.1056668336235624e-07,
      "loss": 0.039,
      "step": 331
    },
    {
      "epoch": 5.358695652173913,
      "grad_norm": 0.23950928449630737,
      "learning_rate": 1.9886607311334987e-07,
      "loss": 0.0375,
      "step": 332
    },
    {
      "epoch": 5.375,
      "grad_norm": 0.2400600165128708,
      "learning_rate": 1.8748643942516882e-07,
      "loss": 0.0447,
      "step": 333
    },
    {
      "epoch": 5.391304347826087,
      "grad_norm": 0.24937079846858978,
      "learning_rate": 1.764293696003358e-07,
      "loss": 0.0439,
      "step": 334
    },
    {
      "epoch": 5.407608695652174,
      "grad_norm": 0.2523324489593506,
      "learning_rate": 1.656964059481453e-07,
      "loss": 0.0368,
      "step": 335
    },
    {
      "epoch": 5.423913043478261,
      "grad_norm": 0.24901482462882996,
      "learning_rate": 1.552890455695369e-07,
      "loss": 0.0382,
      "step": 336
    },
    {
      "epoch": 5.440217391304348,
      "grad_norm": 0.21542105078697205,
      "learning_rate": 1.4520874014826464e-07,
      "loss": 0.04,
      "step": 337
    },
    {
      "epoch": 5.456521739130435,
      "grad_norm": 0.24464815855026245,
      "learning_rate": 1.3545689574841341e-07,
      "loss": 0.0358,
      "step": 338
    },
    {
      "epoch": 5.4728260869565215,
      "grad_norm": 0.24528922140598297,
      "learning_rate": 1.2603487261826726e-07,
      "loss": 0.038,
      "step": 339
    },
    {
      "epoch": 5.489130434782608,
      "grad_norm": 0.22918996214866638,
      "learning_rate": 1.1694398500057714e-07,
      "loss": 0.0421,
      "step": 340
    },
    {
      "epoch": 5.505434782608695,
      "grad_norm": 0.2410709410905838,
      "learning_rate": 1.081855009492383e-07,
      "loss": 0.0395,
      "step": 341
    },
    {
      "epoch": 5.521739130434782,
      "grad_norm": 0.22620242834091187,
      "learning_rate": 9.976064215241859e-08,
      "loss": 0.0392,
      "step": 342
    },
    {
      "epoch": 5.538043478260869,
      "grad_norm": 0.24372537434101105,
      "learning_rate": 9.167058376214621e-08,
      "loss": 0.0379,
      "step": 343
    },
    {
      "epoch": 5.554347826086957,
      "grad_norm": 0.21486566960811615,
      "learning_rate": 8.391645423039357e-08,
      "loss": 0.0377,
      "step": 344
    },
    {
      "epoch": 5.570652173913043,
      "grad_norm": 0.22310517728328705,
      "learning_rate": 7.649933515167407e-08,
      "loss": 0.0386,
      "step": 345
    },
    {
      "epoch": 5.586956521739131,
      "grad_norm": 0.23468376696109772,
      "learning_rate": 6.94202611121736e-08,
      "loss": 0.0398,
      "step": 346
    },
    {
      "epoch": 5.603260869565218,
      "grad_norm": 0.24592317640781403,
      "learning_rate": 6.268021954544095e-08,
      "loss": 0.0346,
      "step": 347
    },
    {
      "epoch": 5.619565217391305,
      "grad_norm": 0.2240305244922638,
      "learning_rate": 5.628015059465364e-08,
      "loss": 0.0483,
      "step": 348
    },
    {
      "epoch": 5.635869565217392,
      "grad_norm": 0.2843106687068939,
      "learning_rate": 5.022094698148072e-08,
      "loss": 0.0427,
      "step": 349
    },
    {
      "epoch": 5.6521739130434785,
      "grad_norm": 0.2484791874885559,
      "learning_rate": 4.450345388156141e-08,
      "loss": 0.0429,
      "step": 350
    },
    {
      "epoch": 5.668478260869565,
      "grad_norm": 0.23303182423114777,
      "learning_rate": 3.9128468806614304e-08,
      "loss": 0.0355,
      "step": 351
    },
    {
      "epoch": 5.684782608695652,
      "grad_norm": 0.22100433707237244,
      "learning_rate": 3.4096741493194196e-08,
      "loss": 0.0386,
      "step": 352
    },
    {
      "epoch": 5.701086956521739,
      "grad_norm": 0.2393917292356491,
      "learning_rate": 2.940897379811597e-08,
      "loss": 0.0495,
      "step": 353
    },
    {
      "epoch": 5.717391304347826,
      "grad_norm": 0.27596771717071533,
      "learning_rate": 2.506581960055432e-08,
      "loss": 0.0388,
      "step": 354
    },
    {
      "epoch": 5.733695652173913,
      "grad_norm": 0.23731505870819092,
      "learning_rate": 2.106788471083615e-08,
      "loss": 0.039,
      "step": 355
    },
    {
      "epoch": 5.75,
      "grad_norm": 0.23420853912830353,
      "learning_rate": 1.7415726785939836e-08,
      "loss": 0.0466,
      "step": 356
    },
    {
      "epoch": 5.766304347826087,
      "grad_norm": 0.2571556866168976,
      "learning_rate": 1.4109855251708272e-08,
      "loss": 0.0398,
      "step": 357
    },
    {
      "epoch": 5.782608695652174,
      "grad_norm": 0.25577929615974426,
      "learning_rate": 1.115073123179128e-08,
      "loss": 0.0418,
      "step": 358
    },
    {
      "epoch": 5.798913043478261,
      "grad_norm": 0.25703752040863037,
      "learning_rate": 8.538767483325384e-09,
      "loss": 0.0382,
      "step": 359
    },
    {
      "epoch": 5.815217391304348,
      "grad_norm": 0.21773234009742737,
      "learning_rate": 6.274328339360702e-09,
      "loss": 0.0374,
      "step": 360
    },
    {
      "epoch": 5.831521739130435,
      "grad_norm": 0.24684062600135803,
      "learning_rate": 4.357729658039378e-09,
      "loss": 0.041,
      "step": 361
    },
    {
      "epoch": 5.8478260869565215,
      "grad_norm": 0.20334966480731964,
      "learning_rate": 2.789238778540537e-09,
      "loss": 0.0386,
      "step": 362
    },
    {
      "epoch": 5.864130434782608,
      "grad_norm": 0.2612401247024536,
      "learning_rate": 1.5690744837873473e-09,
      "loss": 0.0415,
      "step": 363
    },
    {
      "epoch": 5.880434782608695,
      "grad_norm": 0.2394614964723587,
      "learning_rate": 6.974069699314246e-10,
      "loss": 0.0376,
      "step": 364
    },
    {
      "epoch": 5.896739130434782,
      "grad_norm": 0.23148567974567413,
      "learning_rate": 1.743578226129361e-10,
      "loss": 0.0387,
      "step": 365
    },
    {
      "epoch": 5.913043478260869,
      "grad_norm": 0.24650397896766663,
      "learning_rate": 0.0,
      "loss": 0.0396,
      "step": 366
    }
  ],
  "logging_steps": 1,
  "max_steps": 366,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 61,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.114918032020603e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}