{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 6.850632911392405,
  "eval_steps": 500,
  "global_step": 224,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.030379746835443037,
      "grad_norm": 6.142133362846723,
      "learning_rate": 8.695652173913044e-07,
      "loss": 0.7916,
      "step": 1
    },
    {
      "epoch": 0.060759493670886074,
      "grad_norm": 6.357452301292556,
      "learning_rate": 1.7391304347826088e-06,
      "loss": 0.8104,
      "step": 2
    },
    {
      "epoch": 0.09113924050632911,
      "grad_norm": 6.198951704234916,
      "learning_rate": 2.6086956521739132e-06,
      "loss": 0.8154,
      "step": 3
    },
    {
      "epoch": 0.12151898734177215,
      "grad_norm": 5.65438381432537,
      "learning_rate": 3.4782608695652175e-06,
      "loss": 0.8006,
      "step": 4
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 4.4279853468183585,
      "learning_rate": 4.347826086956522e-06,
      "loss": 0.7842,
      "step": 5
    },
    {
      "epoch": 0.18227848101265823,
      "grad_norm": 2.4033605477843993,
      "learning_rate": 5.2173913043478265e-06,
      "loss": 0.7275,
      "step": 6
    },
    {
      "epoch": 0.21265822784810126,
      "grad_norm": 2.008389646266792,
      "learning_rate": 6.086956521739132e-06,
      "loss": 0.7078,
      "step": 7
    },
    {
      "epoch": 0.2430379746835443,
      "grad_norm": 3.856406389872771,
      "learning_rate": 6.956521739130435e-06,
      "loss": 0.7094,
      "step": 8
    },
    {
      "epoch": 0.27341772151898736,
      "grad_norm": 4.236402916502628,
      "learning_rate": 7.82608695652174e-06,
      "loss": 0.7278,
      "step": 9
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 4.000542971929127,
      "learning_rate": 8.695652173913044e-06,
      "loss": 0.6813,
      "step": 10
    },
    {
      "epoch": 0.3341772151898734,
      "grad_norm": 4.178694394396282,
      "learning_rate": 9.565217391304349e-06,
      "loss": 0.6878,
      "step": 11
    },
    {
      "epoch": 0.36455696202531646,
      "grad_norm": 3.2440423383286325,
      "learning_rate": 1.0434782608695653e-05,
      "loss": 0.668,
      "step": 12
    },
    {
      "epoch": 0.3949367088607595,
      "grad_norm": 1.9081921773489765,
      "learning_rate": 1.1304347826086957e-05,
      "loss": 0.6509,
      "step": 13
    },
    {
      "epoch": 0.4253164556962025,
      "grad_norm": 1.6856292551449288,
      "learning_rate": 1.2173913043478263e-05,
      "loss": 0.6206,
      "step": 14
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 1.9652445016067075,
      "learning_rate": 1.3043478260869566e-05,
      "loss": 0.5931,
      "step": 15
    },
    {
      "epoch": 0.4860759493670886,
      "grad_norm": 1.490317661212989,
      "learning_rate": 1.391304347826087e-05,
      "loss": 0.6108,
      "step": 16
    },
    {
      "epoch": 0.5164556962025316,
      "grad_norm": 1.0152980978340718,
      "learning_rate": 1.4782608695652174e-05,
      "loss": 0.5983,
      "step": 17
    },
    {
      "epoch": 0.5468354430379747,
      "grad_norm": 1.0762993241125118,
      "learning_rate": 1.565217391304348e-05,
      "loss": 0.5877,
      "step": 18
    },
    {
      "epoch": 0.5772151898734177,
      "grad_norm": 0.8925505589092044,
      "learning_rate": 1.6521739130434785e-05,
      "loss": 0.5633,
      "step": 19
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 0.9107043160755183,
      "learning_rate": 1.739130434782609e-05,
      "loss": 0.5642,
      "step": 20
    },
    {
      "epoch": 0.6379746835443038,
      "grad_norm": 0.7731829103121682,
      "learning_rate": 1.8260869565217393e-05,
      "loss": 0.5615,
      "step": 21
    },
    {
      "epoch": 0.6683544303797468,
      "grad_norm": 0.744129212592715,
      "learning_rate": 1.9130434782608697e-05,
      "loss": 0.5509,
      "step": 22
    },
    {
      "epoch": 0.6987341772151898,
      "grad_norm": 0.8955523452131373,
      "learning_rate": 2e-05,
      "loss": 0.5549,
      "step": 23
    },
    {
      "epoch": 0.7291139240506329,
      "grad_norm": 0.8183528477924088,
      "learning_rate": 1.999877856940653e-05,
      "loss": 0.5297,
      "step": 24
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.7240144765438188,
      "learning_rate": 1.999511457600466e-05,
      "loss": 0.5563,
      "step": 25
    },
    {
      "epoch": 0.789873417721519,
      "grad_norm": 0.8022538993822405,
      "learning_rate": 1.9989008914857115e-05,
      "loss": 0.5324,
      "step": 26
    },
    {
      "epoch": 0.8202531645569621,
      "grad_norm": 0.6940947230093911,
      "learning_rate": 1.998046307749216e-05,
      "loss": 0.5295,
      "step": 27
    },
    {
      "epoch": 0.850632911392405,
      "grad_norm": 0.7414931560011572,
      "learning_rate": 1.9969479151539238e-05,
      "loss": 0.5312,
      "step": 28
    },
    {
      "epoch": 0.8810126582278481,
      "grad_norm": 0.6914458295472939,
      "learning_rate": 1.9956059820218982e-05,
      "loss": 0.5265,
      "step": 29
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.6406228399495421,
      "learning_rate": 1.9940208361687762e-05,
      "loss": 0.5311,
      "step": 30
    },
    {
      "epoch": 0.9417721518987342,
      "grad_norm": 0.5010766427270893,
      "learning_rate": 1.9921928648236855e-05,
      "loss": 0.5279,
      "step": 31
    },
    {
      "epoch": 0.9721518987341772,
      "grad_norm": 0.5559353823675928,
      "learning_rate": 1.990122514534651e-05,
      "loss": 0.4971,
      "step": 32
    },
    {
      "epoch": 1.010126582278481,
      "grad_norm": 0.8149893932371196,
      "learning_rate": 1.9878102910595097e-05,
      "loss": 0.6966,
      "step": 33
    },
    {
      "epoch": 1.040506329113924,
      "grad_norm": 0.5294742502288636,
      "learning_rate": 1.985256759242359e-05,
      "loss": 0.4646,
      "step": 34
    },
    {
      "epoch": 1.070886075949367,
      "grad_norm": 0.6860122056058924,
      "learning_rate": 1.982462542875576e-05,
      "loss": 0.4899,
      "step": 35
    },
    {
      "epoch": 1.1012658227848102,
      "grad_norm": 0.57893128793477,
      "learning_rate": 1.979428324547432e-05,
      "loss": 0.4967,
      "step": 36
    },
    {
      "epoch": 1.1316455696202532,
      "grad_norm": 0.6199448384978387,
      "learning_rate": 1.9761548454753455e-05,
      "loss": 0.4847,
      "step": 37
    },
    {
      "epoch": 1.1620253164556962,
      "grad_norm": 0.5282551845339171,
      "learning_rate": 1.972642905324813e-05,
      "loss": 0.4804,
      "step": 38
    },
    {
      "epoch": 1.1924050632911392,
      "grad_norm": 0.582257344063008,
      "learning_rate": 1.9688933620140638e-05,
      "loss": 0.4881,
      "step": 39
    },
    {
      "epoch": 1.2227848101265824,
      "grad_norm": 0.5077564908323113,
      "learning_rate": 1.96490713150448e-05,
      "loss": 0.4699,
      "step": 40
    },
    {
      "epoch": 1.2531645569620253,
      "grad_norm": 0.5242356194926282,
      "learning_rate": 1.9606851875768404e-05,
      "loss": 0.4791,
      "step": 41
    },
    {
      "epoch": 1.2835443037974683,
      "grad_norm": 0.5024134383389819,
      "learning_rate": 1.956228561593441e-05,
      "loss": 0.4732,
      "step": 42
    },
    {
      "epoch": 1.3139240506329113,
      "grad_norm": 0.4483295211756338,
      "learning_rate": 1.9515383422461457e-05,
      "loss": 0.488,
      "step": 43
    },
    {
      "epoch": 1.3443037974683545,
      "grad_norm": 0.4483088881234574,
      "learning_rate": 1.9466156752904344e-05,
      "loss": 0.4618,
      "step": 44
    },
    {
      "epoch": 1.3746835443037975,
      "grad_norm": 0.3651600174872845,
      "learning_rate": 1.9414617632655114e-05,
      "loss": 0.4652,
      "step": 45
    },
    {
      "epoch": 1.4050632911392404,
      "grad_norm": 0.5024846295352086,
      "learning_rate": 1.9360778652005416e-05,
      "loss": 0.4986,
      "step": 46
    },
    {
      "epoch": 1.4354430379746836,
      "grad_norm": 0.3371982793352154,
      "learning_rate": 1.9304652963070868e-05,
      "loss": 0.4609,
      "step": 47
    },
    {
      "epoch": 1.4658227848101266,
      "grad_norm": 0.3490217572797855,
      "learning_rate": 1.9246254276578175e-05,
      "loss": 0.4671,
      "step": 48
    },
    {
      "epoch": 1.4962025316455696,
      "grad_norm": 0.38868867411004315,
      "learning_rate": 1.9185596858515797e-05,
      "loss": 0.4577,
      "step": 49
    },
    {
      "epoch": 1.5265822784810128,
      "grad_norm": 0.42412542915142576,
      "learning_rate": 1.9122695526648968e-05,
      "loss": 0.486,
      "step": 50
    },
    {
      "epoch": 1.5569620253164556,
      "grad_norm": 0.4088838516450633,
      "learning_rate": 1.905756564689991e-05,
      "loss": 0.4384,
      "step": 51
    },
    {
      "epoch": 1.5873417721518988,
      "grad_norm": 0.3833188082344623,
      "learning_rate": 1.8990223129594146e-05,
      "loss": 0.4443,
      "step": 52
    },
    {
      "epoch": 1.6177215189873417,
      "grad_norm": 0.4527768936233557,
      "learning_rate": 1.8920684425573865e-05,
      "loss": 0.4663,
      "step": 53
    },
    {
      "epoch": 1.6481012658227847,
      "grad_norm": 0.38073567725179286,
      "learning_rate": 1.884896652217917e-05,
      "loss": 0.4824,
      "step": 54
    },
    {
      "epoch": 1.678481012658228,
      "grad_norm": 0.385620749356317,
      "learning_rate": 1.877508693909831e-05,
      "loss": 0.4677,
      "step": 55
    },
    {
      "epoch": 1.7088607594936709,
      "grad_norm": 0.4071801379977593,
      "learning_rate": 1.8699063724087905e-05,
      "loss": 0.4631,
      "step": 56
    },
    {
      "epoch": 1.7392405063291139,
      "grad_norm": 0.42731940915469707,
      "learning_rate": 1.862091544856407e-05,
      "loss": 0.4568,
      "step": 57
    },
    {
      "epoch": 1.769620253164557,
      "grad_norm": 0.3388107941825912,
      "learning_rate": 1.854066120306571e-05,
      "loss": 0.4375,
      "step": 58
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.43264860692323054,
      "learning_rate": 1.8458320592590976e-05,
      "loss": 0.4795,
      "step": 59
    },
    {
      "epoch": 1.830379746835443,
      "grad_norm": 0.3823967868290758,
      "learning_rate": 1.837391373180801e-05,
      "loss": 0.4611,
      "step": 60
    },
    {
      "epoch": 1.8607594936708862,
      "grad_norm": 0.4186425213226183,
      "learning_rate": 1.8287461240141217e-05,
      "loss": 0.4669,
      "step": 61
    },
    {
      "epoch": 1.891139240506329,
      "grad_norm": 0.3703454303387632,
      "learning_rate": 1.8198984236734246e-05,
      "loss": 0.4557,
      "step": 62
    },
    {
      "epoch": 1.9215189873417722,
      "grad_norm": 0.3736106931482273,
      "learning_rate": 1.8108504335290852e-05,
      "loss": 0.4344,
      "step": 63
    },
    {
      "epoch": 1.9518987341772152,
      "grad_norm": 0.39212975675707035,
      "learning_rate": 1.8016043638794975e-05,
      "loss": 0.4554,
      "step": 64
    },
    {
      "epoch": 1.9822784810126581,
      "grad_norm": 0.4513720869681548,
      "learning_rate": 1.7921624734111292e-05,
      "loss": 0.5623,
      "step": 65
    },
    {
      "epoch": 2.020253164556962,
      "grad_norm": 0.5971726879383622,
      "learning_rate": 1.7825270686467567e-05,
      "loss": 0.517,
      "step": 66
    },
    {
      "epoch": 2.050632911392405,
      "grad_norm": 0.46712002103063616,
      "learning_rate": 1.7727005033820117e-05,
      "loss": 0.4231,
      "step": 67
    },
    {
      "epoch": 2.081012658227848,
      "grad_norm": 0.4189386682745267,
      "learning_rate": 1.762685178110382e-05,
      "loss": 0.405,
      "step": 68
    },
    {
      "epoch": 2.1113924050632913,
      "grad_norm": 0.5493762611884587,
      "learning_rate": 1.752483539436807e-05,
      "loss": 0.4388,
      "step": 69
    },
    {
      "epoch": 2.141772151898734,
      "grad_norm": 0.5178546231991467,
      "learning_rate": 1.7420980794800013e-05,
      "loss": 0.4169,
      "step": 70
    },
    {
      "epoch": 2.1721518987341772,
      "grad_norm": 0.40664079692056654,
      "learning_rate": 1.731531335263669e-05,
      "loss": 0.3984,
      "step": 71
    },
    {
      "epoch": 2.2025316455696204,
      "grad_norm": 0.49767340979477503,
      "learning_rate": 1.720785888096743e-05,
      "loss": 0.4132,
      "step": 72
    },
    {
      "epoch": 2.232911392405063,
      "grad_norm": 0.36495870737503733,
      "learning_rate": 1.7098643629428035e-05,
      "loss": 0.4062,
      "step": 73
    },
    {
      "epoch": 2.2632911392405064,
      "grad_norm": 0.48978226775847694,
      "learning_rate": 1.698769427778842e-05,
      "loss": 0.4129,
      "step": 74
    },
    {
      "epoch": 2.293670886075949,
      "grad_norm": 0.4112454996416562,
      "learning_rate": 1.687503792943506e-05,
      "loss": 0.4165,
      "step": 75
    },
    {
      "epoch": 2.3240506329113924,
      "grad_norm": 0.48272726495216617,
      "learning_rate": 1.6760702104750046e-05,
      "loss": 0.3878,
      "step": 76
    },
    {
      "epoch": 2.3544303797468356,
      "grad_norm": 0.4391820437847134,
      "learning_rate": 1.664471473438822e-05,
      "loss": 0.4001,
      "step": 77
    },
    {
      "epoch": 2.3848101265822783,
      "grad_norm": 0.4542156038994038,
      "learning_rate": 1.6527104152454096e-05,
      "loss": 0.3871,
      "step": 78
    },
    {
      "epoch": 2.4151898734177215,
      "grad_norm": 0.39387687088459095,
      "learning_rate": 1.6407899089580263e-05,
      "loss": 0.3907,
      "step": 79
    },
    {
      "epoch": 2.4455696202531647,
      "grad_norm": 0.3511312711326436,
      "learning_rate": 1.628712866590885e-05,
      "loss": 0.4172,
      "step": 80
    },
    {
      "epoch": 2.4759493670886075,
      "grad_norm": 0.41925040043693396,
      "learning_rate": 1.6164822383977912e-05,
      "loss": 0.4126,
      "step": 81
    },
    {
      "epoch": 2.5063291139240507,
      "grad_norm": 0.37811810641092347,
      "learning_rate": 1.604101012151436e-05,
      "loss": 0.406,
      "step": 82
    },
    {
      "epoch": 2.536708860759494,
      "grad_norm": 0.40554987345889937,
      "learning_rate": 1.5915722124135227e-05,
      "loss": 0.4078,
      "step": 83
    },
    {
      "epoch": 2.5670886075949366,
      "grad_norm": 0.37778791497437947,
      "learning_rate": 1.5788988997959115e-05,
      "loss": 0.3957,
      "step": 84
    },
    {
      "epoch": 2.59746835443038,
      "grad_norm": 0.33785598659802735,
      "learning_rate": 1.5660841702129533e-05,
      "loss": 0.4149,
      "step": 85
    },
    {
      "epoch": 2.6278481012658226,
      "grad_norm": 0.4206720029808489,
      "learning_rate": 1.5531311541251995e-05,
      "loss": 0.41,
      "step": 86
    },
    {
      "epoch": 2.6582278481012658,
      "grad_norm": 0.32357532653393634,
      "learning_rate": 1.540043015774676e-05,
      "loss": 0.4022,
      "step": 87
    },
    {
      "epoch": 2.688607594936709,
      "grad_norm": 0.3855413521175904,
      "learning_rate": 1.5268229524119007e-05,
      "loss": 0.3878,
      "step": 88
    },
    {
      "epoch": 2.7189873417721517,
      "grad_norm": 0.3606649896198645,
      "learning_rate": 1.513474193514842e-05,
      "loss": 0.4113,
      "step": 89
    },
    {
      "epoch": 2.749367088607595,
      "grad_norm": 0.380809979287942,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.3945,
      "step": 90
    },
    {
      "epoch": 2.779746835443038,
      "grad_norm": 0.3195477665445599,
      "learning_rate": 1.4864036634258112e-05,
      "loss": 0.3926,
      "step": 91
    },
    {
      "epoch": 2.810126582278481,
      "grad_norm": 0.4127265000543691,
      "learning_rate": 1.4726885051885654e-05,
      "loss": 0.4017,
      "step": 92
    },
    {
      "epoch": 2.840506329113924,
      "grad_norm": 0.37566148495430407,
      "learning_rate": 1.4588578757110359e-05,
      "loss": 0.4106,
      "step": 93
    },
    {
      "epoch": 2.8708860759493673,
      "grad_norm": 0.3202753523663073,
      "learning_rate": 1.4449151536240167e-05,
      "loss": 0.401,
      "step": 94
    },
    {
      "epoch": 2.90126582278481,
      "grad_norm": 0.35443238121748605,
      "learning_rate": 1.4308637449409705e-05,
      "loss": 0.3904,
      "step": 95
    },
    {
      "epoch": 2.9316455696202532,
      "grad_norm": 0.2999572051502914,
      "learning_rate": 1.4167070822259868e-05,
      "loss": 0.4236,
      "step": 96
    },
    {
      "epoch": 2.962025316455696,
      "grad_norm": 0.39881424254012526,
      "learning_rate": 1.402448623755254e-05,
      "loss": 0.4107,
      "step": 97
    },
    {
      "epoch": 2.992405063291139,
      "grad_norm": 0.467202088309694,
      "learning_rate": 1.3880918526722497e-05,
      "loss": 0.5364,
      "step": 98
    },
    {
      "epoch": 3.030379746835443,
      "grad_norm": 0.4785912096119782,
      "learning_rate": 1.3736402761368597e-05,
      "loss": 0.3627,
      "step": 99
    },
    {
      "epoch": 3.060759493670886,
      "grad_norm": 0.4123328521253836,
      "learning_rate": 1.3590974244686248e-05,
      "loss": 0.3557,
      "step": 100
    },
    {
      "epoch": 3.091139240506329,
      "grad_norm": 0.40421162698630236,
      "learning_rate": 1.344466850284333e-05,
      "loss": 0.3697,
      "step": 101
    },
    {
      "epoch": 3.1215189873417724,
      "grad_norm": 0.433984861645439,
      "learning_rate": 1.3297521276301666e-05,
      "loss": 0.3646,
      "step": 102
    },
    {
      "epoch": 3.151898734177215,
      "grad_norm": 0.4323593732774851,
      "learning_rate": 1.3149568511086104e-05,
      "loss": 0.3475,
      "step": 103
    },
    {
      "epoch": 3.1822784810126583,
      "grad_norm": 0.4282473502024761,
      "learning_rate": 1.300084635000341e-05,
      "loss": 0.3735,
      "step": 104
    },
    {
      "epoch": 3.212658227848101,
      "grad_norm": 0.450504914129418,
      "learning_rate": 1.2851391123813075e-05,
      "loss": 0.3667,
      "step": 105
    },
    {
      "epoch": 3.2430379746835443,
      "grad_norm": 0.3607456231305978,
      "learning_rate": 1.2701239342352223e-05,
      "loss": 0.3497,
      "step": 106
    },
    {
      "epoch": 3.2734177215189875,
      "grad_norm": 0.4006172961395924,
      "learning_rate": 1.2550427685616767e-05,
      "loss": 0.3627,
      "step": 107
    },
    {
      "epoch": 3.3037974683544302,
      "grad_norm": 0.3644163458288102,
      "learning_rate": 1.239899299480098e-05,
      "loss": 0.3482,
      "step": 108
    },
    {
      "epoch": 3.3341772151898734,
      "grad_norm": 0.3634646293817289,
      "learning_rate": 1.2246972263297718e-05,
      "loss": 0.3582,
      "step": 109
    },
    {
      "epoch": 3.3645569620253166,
      "grad_norm": 0.34636760729479316,
      "learning_rate": 1.2094402627661447e-05,
      "loss": 0.3469,
      "step": 110
    },
    {
      "epoch": 3.3949367088607594,
      "grad_norm": 0.350563977624681,
      "learning_rate": 1.1941321358536278e-05,
      "loss": 0.3578,
      "step": 111
    },
    {
      "epoch": 3.4253164556962026,
      "grad_norm": 0.37023856212680784,
      "learning_rate": 1.1787765851551296e-05,
      "loss": 0.3623,
      "step": 112
    },
    {
      "epoch": 3.4556962025316453,
      "grad_norm": 0.4226430510598916,
      "learning_rate": 1.1633773618185302e-05,
      "loss": 0.3601,
      "step": 113
    },
    {
      "epoch": 3.4860759493670885,
      "grad_norm": 0.34139990505608003,
      "learning_rate": 1.14793822766033e-05,
      "loss": 0.354,
      "step": 114
    },
    {
      "epoch": 3.5164556962025317,
      "grad_norm": 0.44156441256510204,
      "learning_rate": 1.132462954246688e-05,
      "loss": 0.3576,
      "step": 115
    },
    {
      "epoch": 3.546835443037975,
      "grad_norm": 0.35855259068342393,
      "learning_rate": 1.1169553219720828e-05,
      "loss": 0.3524,
      "step": 116
    },
    {
      "epoch": 3.5772151898734177,
      "grad_norm": 0.347968923156379,
      "learning_rate": 1.1014191191358118e-05,
      "loss": 0.3513,
      "step": 117
    },
    {
      "epoch": 3.607594936708861,
      "grad_norm": 0.44418621181905155,
      "learning_rate": 1.085858141016566e-05,
      "loss": 0.3641,
      "step": 118
    },
    {
      "epoch": 3.6379746835443036,
      "grad_norm": 0.31000630762886555,
      "learning_rate": 1.070276188945293e-05,
      "loss": 0.3495,
      "step": 119
    },
    {
      "epoch": 3.668354430379747,
      "grad_norm": 0.3800328620288808,
      "learning_rate": 1.0546770693765859e-05,
      "loss": 0.3466,
      "step": 120
    },
    {
      "epoch": 3.69873417721519,
      "grad_norm": 0.3529018173722371,
      "learning_rate": 1.0390645929588197e-05,
      "loss": 0.3515,
      "step": 121
    },
    {
      "epoch": 3.729113924050633,
      "grad_norm": 0.3037778673267576,
      "learning_rate": 1.0234425736032607e-05,
      "loss": 0.3636,
      "step": 122
    },
    {
      "epoch": 3.759493670886076,
      "grad_norm": 0.354487881138209,
      "learning_rate": 1.007814827552384e-05,
      "loss": 0.3526,
      "step": 123
    },
    {
      "epoch": 3.7898734177215188,
      "grad_norm": 0.2786469757823622,
      "learning_rate": 9.92185172447616e-06,
      "loss": 0.3393,
      "step": 124
    },
    {
      "epoch": 3.820253164556962,
      "grad_norm": 0.33617721823861874,
      "learning_rate": 9.765574263967397e-06,
      "loss": 0.3553,
      "step": 125
    },
    {
      "epoch": 3.850632911392405,
      "grad_norm": 0.2736871828499611,
      "learning_rate": 9.609354070411807e-06,
      "loss": 0.3602,
      "step": 126
    },
    {
      "epoch": 3.8810126582278484,
      "grad_norm": 0.323005194458033,
      "learning_rate": 9.453229306234143e-06,
      "loss": 0.3645,
      "step": 127
    },
    {
      "epoch": 3.911392405063291,
      "grad_norm": 0.2893454787886965,
      "learning_rate": 9.297238110547075e-06,
      "loss": 0.3523,
      "step": 128
    },
    {
      "epoch": 3.9417721518987343,
      "grad_norm": 0.30241738017140496,
      "learning_rate": 9.14141858983434e-06,
      "loss": 0.3565,
      "step": 129
    },
    {
      "epoch": 3.972151898734177,
      "grad_norm": 0.279249192979288,
      "learning_rate": 8.985808808641883e-06,
      "loss": 0.3531,
      "step": 130
    },
    {
      "epoch": 4.010126582278481,
      "grad_norm": 0.3994347972941027,
      "learning_rate": 8.830446780279175e-06,
      "loss": 0.4578,
      "step": 131
    },
    {
      "epoch": 4.040506329113924,
      "grad_norm": 0.3710764347791464,
      "learning_rate": 8.675370457533122e-06,
      "loss": 0.322,
      "step": 132
    },
    {
      "epoch": 4.0708860759493675,
      "grad_norm": 0.260181685428155,
      "learning_rate": 8.520617723396702e-06,
      "loss": 0.3249,
      "step": 133
    },
    {
      "epoch": 4.10126582278481,
      "grad_norm": 0.3506943517407662,
      "learning_rate": 8.366226381814698e-06,
      "loss": 0.3139,
      "step": 134
    },
    {
      "epoch": 4.131645569620253,
      "grad_norm": 0.3770019236558475,
      "learning_rate": 8.212234148448708e-06,
      "loss": 0.3174,
      "step": 135
    },
    {
      "epoch": 4.162025316455696,
      "grad_norm": 0.31703490391329475,
      "learning_rate": 8.058678641463724e-06,
      "loss": 0.3118,
      "step": 136
    },
    {
      "epoch": 4.192405063291139,
      "grad_norm": 0.29908697282259106,
      "learning_rate": 7.905597372338558e-06,
      "loss": 0.3161,
      "step": 137
    },
    {
      "epoch": 4.222784810126583,
      "grad_norm": 0.31704079268137586,
      "learning_rate": 7.753027736702283e-06,
      "loss": 0.3322,
      "step": 138
    },
    {
      "epoch": 4.253164556962025,
      "grad_norm": 0.30595858377507373,
      "learning_rate": 7.601007005199022e-06,
      "loss": 0.3239,
      "step": 139
    },
    {
      "epoch": 4.283544303797468,
      "grad_norm": 0.31903380828850986,
      "learning_rate": 7.449572314383237e-06,
      "loss": 0.3231,
      "step": 140
    },
    {
      "epoch": 4.313924050632911,
      "grad_norm": 0.26748898979731556,
      "learning_rate": 7.298760657647779e-06,
      "loss": 0.2972,
      "step": 141
    },
    {
      "epoch": 4.3443037974683545,
      "grad_norm": 0.3238521787842106,
      "learning_rate": 7.148608876186931e-06,
      "loss": 0.3232,
      "step": 142
    },
    {
      "epoch": 4.374683544303798,
      "grad_norm": 0.32648689746342563,
      "learning_rate": 6.999153649996595e-06,
      "loss": 0.3188,
      "step": 143
    },
    {
      "epoch": 4.405063291139241,
      "grad_norm": 0.26613832347863436,
      "learning_rate": 6.8504314889138956e-06,
      "loss": 0.3095,
      "step": 144
    },
    {
      "epoch": 4.435443037974683,
      "grad_norm": 0.29059459219393696,
      "learning_rate": 6.702478723698336e-06,
      "loss": 0.3095,
      "step": 145
    },
    {
      "epoch": 4.465822784810126,
      "grad_norm": 0.28272423574830846,
      "learning_rate": 6.555331497156671e-06,
      "loss": 0.3231,
      "step": 146
    },
    {
      "epoch": 4.49620253164557,
      "grad_norm": 0.27147630291281827,
      "learning_rate": 6.4090257553137566e-06,
      "loss": 0.3098,
      "step": 147
    },
    {
      "epoch": 4.526582278481013,
      "grad_norm": 0.27218085631978883,
      "learning_rate": 6.263597238631405e-06,
      "loss": 0.3041,
      "step": 148
    },
    {
      "epoch": 4.556962025316456,
      "grad_norm": 0.2907764850583011,
      "learning_rate": 6.119081473277502e-06,
      "loss": 0.3122,
      "step": 149
    },
    {
      "epoch": 4.587341772151898,
      "grad_norm": 0.2610491444711985,
      "learning_rate": 5.975513762447465e-06,
      "loss": 0.299,
      "step": 150
    },
    {
      "epoch": 4.6177215189873415,
      "grad_norm": 0.28102771407725835,
      "learning_rate": 5.832929177740134e-06,
      "loss": 0.3003,
      "step": 151
    },
    {
      "epoch": 4.648101265822785,
      "grad_norm": 0.2853109911022994,
      "learning_rate": 5.6913625505902966e-06,
      "loss": 0.3173,
      "step": 152
    },
    {
      "epoch": 4.678481012658228,
      "grad_norm": 0.25899162810004384,
      "learning_rate": 5.550848463759835e-06,
      "loss": 0.3037,
      "step": 153
    },
    {
      "epoch": 4.708860759493671,
      "grad_norm": 0.2856251013932412,
      "learning_rate": 5.411421242889643e-06,
      "loss": 0.312,
      "step": 154
    },
    {
      "epoch": 4.739240506329114,
      "grad_norm": 0.28563131510003925,
      "learning_rate": 5.273114948114346e-06,
      "loss": 0.3124,
      "step": 155
    },
    {
      "epoch": 4.769620253164557,
      "grad_norm": 0.2904879803329038,
      "learning_rate": 5.135963365741892e-06,
      "loss": 0.3056,
      "step": 156
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.27167234153720476,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.3114,
      "step": 157
    },
    {
      "epoch": 4.830379746835443,
      "grad_norm": 0.2781315527452187,
      "learning_rate": 4.865258064851579e-06,
      "loss": 0.3129,
      "step": 158
    },
    {
      "epoch": 4.860759493670886,
      "grad_norm": 0.2815845627503612,
      "learning_rate": 4.731770475880995e-06,
      "loss": 0.3112,
      "step": 159
    },
    {
      "epoch": 4.891139240506329,
      "grad_norm": 0.24314085632282065,
      "learning_rate": 4.599569842253244e-06,
      "loss": 0.2981,
      "step": 160
    },
    {
      "epoch": 4.921518987341772,
      "grad_norm": 0.2771334038160195,
      "learning_rate": 4.468688458748006e-06,
      "loss": 0.3217,
      "step": 161
    },
    {
      "epoch": 4.951898734177215,
      "grad_norm": 0.2778708149835448,
      "learning_rate": 4.339158297870469e-06,
      "loss": 0.3005,
      "step": 162
    },
    {
      "epoch": 4.982278481012658,
      "grad_norm": 0.3181763462270913,
      "learning_rate": 4.211011002040885e-06,
      "loss": 0.3692,
      "step": 163
    },
    {
      "epoch": 5.020253164556962,
      "grad_norm": 0.3380744967105364,
      "learning_rate": 4.084277875864776e-06,
      "loss": 0.3518,
      "step": 164
    },
    {
      "epoch": 5.050632911392405,
      "grad_norm": 0.3077116531141232,
      "learning_rate": 3.958989878485644e-06,
      "loss": 0.273,
      "step": 165
    },
    {
      "epoch": 5.0810126582278485,
      "grad_norm": 0.2817477362200768,
      "learning_rate": 3.83517761602209e-06,
      "loss": 0.2816,
      "step": 166
    },
    {
      "epoch": 5.111392405063291,
      "grad_norm": 0.25534644431709314,
      "learning_rate": 3.712871334091154e-06,
      "loss": 0.2676,
      "step": 167
    },
    {
      "epoch": 5.141772151898734,
      "grad_norm": 0.39733207815191407,
      "learning_rate": 3.592100910419738e-06,
      "loss": 0.2798,
      "step": 168
    },
    {
      "epoch": 5.172151898734177,
      "grad_norm": 0.35253652256958323,
      "learning_rate": 3.4728958475459052e-06,
      "loss": 0.2829,
      "step": 169
    },
    {
      "epoch": 5.2025316455696204,
      "grad_norm": 0.27761876273132563,
      "learning_rate": 3.355285265611784e-06,
      "loss": 0.2695,
      "step": 170
    },
    {
      "epoch": 5.232911392405064,
      "grad_norm": 0.27194117271731544,
      "learning_rate": 3.2392978952499553e-06,
      "loss": 0.287,
      "step": 171
    },
    {
      "epoch": 5.263291139240506,
      "grad_norm": 0.2638988503117021,
      "learning_rate": 3.1249620705649417e-06,
      "loss": 0.2743,
      "step": 172
    },
    {
      "epoch": 5.293670886075949,
      "grad_norm": 0.3254658079971883,
      "learning_rate": 3.0123057222115835e-06,
      "loss": 0.3088,
      "step": 173
    },
    {
      "epoch": 5.324050632911392,
      "grad_norm": 0.2630037510542853,
      "learning_rate": 2.9013563705719673e-06,
      "loss": 0.2873,
      "step": 174
    },
    {
      "epoch": 5.3544303797468356,
      "grad_norm": 0.2528822714263439,
      "learning_rate": 2.7921411190325753e-06,
      "loss": 0.2927,
      "step": 175
    },
    {
      "epoch": 5.384810126582279,
      "grad_norm": 0.2628939766078353,
      "learning_rate": 2.6846866473633126e-06,
      "loss": 0.2771,
      "step": 176
    },
    {
      "epoch": 5.415189873417722,
      "grad_norm": 0.26653726514044274,
      "learning_rate": 2.579019205199992e-06,
      "loss": 0.2753,
      "step": 177
    },
    {
      "epoch": 5.445569620253164,
      "grad_norm": 0.2743528017367845,
      "learning_rate": 2.4751646056319334e-06,
      "loss": 0.2903,
      "step": 178
    },
    {
      "epoch": 5.4759493670886075,
      "grad_norm": 0.2563760087561412,
      "learning_rate": 2.373148218896182e-06,
      "loss": 0.2743,
      "step": 179
    },
    {
      "epoch": 5.506329113924051,
      "grad_norm": 0.2563819896924902,
      "learning_rate": 2.2729949661798876e-06,
      "loss": 0.3062,
      "step": 180
    },
    {
      "epoch": 5.536708860759494,
      "grad_norm": 0.22888829009403563,
      "learning_rate": 2.174729313532433e-06,
      "loss": 0.265,
      "step": 181
    },
    {
      "epoch": 5.567088607594937,
      "grad_norm": 0.2641442442655395,
      "learning_rate": 2.078375265888707e-06,
      "loss": 0.2912,
      "step": 182
    },
    {
      "epoch": 5.597468354430379,
      "grad_norm": 0.24293783978668715,
      "learning_rate": 1.9839563612050273e-06,
      "loss": 0.2788,
      "step": 183
    },
    {
      "epoch": 5.627848101265823,
      "grad_norm": 0.234365720480904,
      "learning_rate": 1.8914956647091497e-06,
      "loss": 0.2738,
      "step": 184
    },
    {
      "epoch": 5.658227848101266,
      "grad_norm": 0.23733298634685637,
      "learning_rate": 1.8010157632657544e-06,
      "loss": 0.2886,
      "step": 185
    },
    {
      "epoch": 5.688607594936709,
      "grad_norm": 0.2320333532934732,
      "learning_rate": 1.7125387598587862e-06,
      "loss": 0.281,
      "step": 186
    },
    {
      "epoch": 5.718987341772152,
      "grad_norm": 0.2347008234627476,
      "learning_rate": 1.6260862681919965e-06,
      "loss": 0.2833,
      "step": 187
    },
    {
      "epoch": 5.749367088607595,
      "grad_norm": 0.23470344799587836,
      "learning_rate": 1.5416794074090258e-06,
      "loss": 0.2924,
      "step": 188
    },
    {
      "epoch": 5.779746835443038,
      "grad_norm": 0.22704919723280656,
      "learning_rate": 1.459338796934293e-06,
      "loss": 0.2824,
      "step": 189
    },
    {
      "epoch": 5.810126582278481,
      "grad_norm": 0.22649545986397604,
      "learning_rate": 1.3790845514359363e-06,
      "loss": 0.2829,
      "step": 190
    },
    {
      "epoch": 5.840506329113924,
      "grad_norm": 0.23366827962350076,
      "learning_rate": 1.300936275912098e-06,
      "loss": 0.274,
      "step": 191
    },
    {
      "epoch": 5.870886075949367,
      "grad_norm": 0.23325744323695088,
      "learning_rate": 1.224913060901688e-06,
      "loss": 0.2854,
      "step": 192
    },
    {
      "epoch": 5.9012658227848105,
      "grad_norm": 0.23452041472020713,
      "learning_rate": 1.1510334778208332e-06,
      "loss": 0.2862,
      "step": 193
    },
    {
      "epoch": 5.931645569620253,
      "grad_norm": 0.2128850574376154,
      "learning_rate": 1.0793155744261352e-06,
      "loss": 0.2716,
      "step": 194
    },
    {
      "epoch": 5.962025316455696,
      "grad_norm": 0.21756339881849357,
      "learning_rate": 1.0097768704058542e-06,
      "loss": 0.2769,
      "step": 195
    },
    {
      "epoch": 5.992405063291139,
      "grad_norm": 0.5731366409998244,
      "learning_rate": 9.424343531000968e-07,
      "loss": 0.3663,
      "step": 196
    },
    {
      "epoch": 6.030379746835443,
      "grad_norm": 0.2593936366293042,
      "learning_rate": 8.773044733510338e-07,
      "loss": 0.2701,
      "step": 197
    },
    {
      "epoch": 6.060759493670886,
      "grad_norm": 0.24007943216008748,
      "learning_rate": 8.144031414842012e-07,
      "loss": 0.271,
      "step": 198
    },
    {
      "epoch": 6.091139240506329,
      "grad_norm": 0.24480502414592123,
      "learning_rate": 7.537457234218271e-07,
      "loss": 0.2807,
      "step": 199
    },
    {
      "epoch": 6.121518987341772,
      "grad_norm": 0.22161317113429632,
      "learning_rate": 6.953470369291349e-07,
      "loss": 0.2718,
      "step": 200
    },
    {
      "epoch": 6.151898734177215,
      "grad_norm": 0.21942095908540368,
      "learning_rate": 6.392213479945852e-07,
      "loss": 0.2631,
      "step": 201
    },
    {
      "epoch": 6.182278481012658,
      "grad_norm": 0.22385509803686687,
      "learning_rate": 5.853823673448877e-07,
      "loss": 0.2736,
      "step": 202
    },
    {
      "epoch": 6.2126582278481015,
      "grad_norm": 0.21766702781637903,
      "learning_rate": 5.33843247095659e-07,
      "loss": 0.2682,
      "step": 203
    },
    {
      "epoch": 6.243037974683545,
      "grad_norm": 0.22207085212457575,
      "learning_rate": 4.846165775385459e-07,
      "loss": 0.2605,
      "step": 204
    },
    {
      "epoch": 6.273417721518987,
      "grad_norm": 0.21458752614863258,
      "learning_rate": 4.3771438406559173e-07,
      "loss": 0.2703,
      "step": 205
    },
    {
      "epoch": 6.30379746835443,
      "grad_norm": 0.2271561890743592,
      "learning_rate": 3.931481242315993e-07,
      "loss": 0.2604,
      "step": 206
    },
    {
      "epoch": 6.334177215189873,
      "grad_norm": 0.2305522662239651,
      "learning_rate": 3.5092868495520294e-07,
      "loss": 0.2732,
      "step": 207
    },
    {
      "epoch": 6.364556962025317,
      "grad_norm": 0.23576799695501874,
      "learning_rate": 3.110663798593616e-07,
      "loss": 0.2753,
      "step": 208
    },
    {
      "epoch": 6.39493670886076,
      "grad_norm": 0.21533008163072734,
      "learning_rate": 2.735709467518699e-07,
      "loss": 0.2767,
      "step": 209
    },
    {
      "epoch": 6.425316455696202,
      "grad_norm": 0.21999408184797783,
      "learning_rate": 2.384515452465475e-07,
      "loss": 0.2718,
      "step": 210
    },
    {
      "epoch": 6.455696202531645,
      "grad_norm": 0.2170059043336272,
      "learning_rate": 2.0571675452567997e-07,
      "loss": 0.2727,
      "step": 211
    },
    {
      "epoch": 6.4860759493670885,
      "grad_norm": 0.22180241918305427,
      "learning_rate": 1.7537457124423896e-07,
      "loss": 0.2674,
      "step": 212
    },
    {
      "epoch": 6.516455696202532,
      "grad_norm": 0.22975040085185996,
      "learning_rate": 1.474324075764111e-07,
      "loss": 0.2807,
      "step": 213
    },
    {
      "epoch": 6.546835443037975,
      "grad_norm": 0.2130343186081862,
      "learning_rate": 1.2189708940490653e-07,
      "loss": 0.2633,
      "step": 214
    },
    {
      "epoch": 6.577215189873417,
      "grad_norm": 0.21741201160529977,
      "learning_rate": 9.877485465349057e-08,
      "loss": 0.2715,
      "step": 215
    },
    {
      "epoch": 6.6075949367088604,
      "grad_norm": 0.21709840614985526,
      "learning_rate": 7.807135176314707e-08,
      "loss": 0.2759,
      "step": 216
    },
    {
      "epoch": 6.637974683544304,
      "grad_norm": 0.21754224290691965,
      "learning_rate": 5.979163831223988e-08,
      "loss": 0.2647,
      "step": 217
    },
    {
      "epoch": 6.668354430379747,
      "grad_norm": 0.21962783026780733,
      "learning_rate": 4.394017978101905e-08,
      "loss": 0.2679,
      "step": 218
    },
    {
      "epoch": 6.69873417721519,
      "grad_norm": 0.21340658097304008,
      "learning_rate": 3.0520848460765525e-08,
      "loss": 0.2739,
      "step": 219
    },
    {
      "epoch": 6.729113924050633,
      "grad_norm": 0.22462643761247802,
      "learning_rate": 1.9536922507841227e-08,
      "loss": 0.2761,
      "step": 220
    },
    {
      "epoch": 6.759493670886076,
      "grad_norm": 0.21469231815944442,
      "learning_rate": 1.099108514288627e-08,
      "loss": 0.2678,
      "step": 221
    },
    {
      "epoch": 6.789873417721519,
      "grad_norm": 0.2151131540272051,
      "learning_rate": 4.885423995341088e-09,
      "loss": 0.2651,
      "step": 222
    },
    {
      "epoch": 6.820253164556962,
      "grad_norm": 0.21049176928568308,
      "learning_rate": 1.2214305934699078e-09,
      "loss": 0.257,
      "step": 223
    },
    {
      "epoch": 6.850632911392405,
      "grad_norm": 0.21512970249569155,
      "learning_rate": 0.0,
      "loss": 0.2692,
      "step": 224
    },
    {
      "epoch": 6.850632911392405,
      "step": 224,
      "total_flos": 8.762232759008625e+17,
      "train_loss": 0.39511572915528503,
      "train_runtime": 26231.4222,
      "train_samples_per_second": 0.843,
      "train_steps_per_second": 0.009
    }
  ],
  "logging_steps": 1,
  "max_steps": 224,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.762232759008625e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}