{
  "best_metric": 1.3474754095077515,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.4282655246252677,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0021413276231263384,
      "grad_norm": 1.0215110778808594,
      "learning_rate": 1.0017e-05,
      "loss": 1.7563,
      "step": 1
    },
    {
      "epoch": 0.0021413276231263384,
      "eval_loss": 1.8309117555618286,
      "eval_runtime": 21.0014,
      "eval_samples_per_second": 9.38,
      "eval_steps_per_second": 2.381,
      "step": 1
    },
    {
      "epoch": 0.004282655246252677,
      "grad_norm": 1.1718658208847046,
      "learning_rate": 2.0034e-05,
      "loss": 1.6414,
      "step": 2
    },
    {
      "epoch": 0.006423982869379015,
      "grad_norm": 1.895135760307312,
      "learning_rate": 3.0050999999999997e-05,
      "loss": 1.6203,
      "step": 3
    },
    {
      "epoch": 0.008565310492505354,
      "grad_norm": 1.8952735662460327,
      "learning_rate": 4.0068e-05,
      "loss": 1.6905,
      "step": 4
    },
    {
      "epoch": 0.010706638115631691,
      "grad_norm": 0.9003539681434631,
      "learning_rate": 5.0085e-05,
      "loss": 1.5445,
      "step": 5
    },
    {
      "epoch": 0.01284796573875803,
      "grad_norm": 0.8659182786941528,
      "learning_rate": 6.0101999999999995e-05,
      "loss": 1.6224,
      "step": 6
    },
    {
      "epoch": 0.014989293361884369,
      "grad_norm": 0.9884287714958191,
      "learning_rate": 7.0119e-05,
      "loss": 1.5734,
      "step": 7
    },
    {
      "epoch": 0.017130620985010708,
      "grad_norm": 1.4739041328430176,
      "learning_rate": 8.0136e-05,
      "loss": 1.5098,
      "step": 8
    },
    {
      "epoch": 0.019271948608137045,
      "grad_norm": 0.956966757774353,
      "learning_rate": 9.0153e-05,
      "loss": 1.5619,
      "step": 9
    },
    {
      "epoch": 0.021413276231263382,
      "grad_norm": 0.809992790222168,
      "learning_rate": 0.00010017,
      "loss": 1.4481,
      "step": 10
    },
    {
      "epoch": 0.023554603854389723,
      "grad_norm": 0.9872943162918091,
      "learning_rate": 9.964278947368421e-05,
      "loss": 1.41,
      "step": 11
    },
    {
      "epoch": 0.02569593147751606,
      "grad_norm": 1.4342368841171265,
      "learning_rate": 9.911557894736841e-05,
      "loss": 1.4035,
      "step": 12
    },
    {
      "epoch": 0.027837259100642397,
      "grad_norm": 0.7947163581848145,
      "learning_rate": 9.858836842105263e-05,
      "loss": 1.4192,
      "step": 13
    },
    {
      "epoch": 0.029978586723768737,
      "grad_norm": 0.795746386051178,
      "learning_rate": 9.806115789473684e-05,
      "loss": 1.5559,
      "step": 14
    },
    {
      "epoch": 0.032119914346895075,
      "grad_norm": 0.6750279068946838,
      "learning_rate": 9.753394736842106e-05,
      "loss": 1.3721,
      "step": 15
    },
    {
      "epoch": 0.034261241970021415,
      "grad_norm": 0.666294276714325,
      "learning_rate": 9.700673684210526e-05,
      "loss": 1.3652,
      "step": 16
    },
    {
      "epoch": 0.03640256959314775,
      "grad_norm": 0.6185109615325928,
      "learning_rate": 9.647952631578948e-05,
      "loss": 1.3043,
      "step": 17
    },
    {
      "epoch": 0.03854389721627409,
      "grad_norm": 0.6206092238426208,
      "learning_rate": 9.595231578947368e-05,
      "loss": 1.2114,
      "step": 18
    },
    {
      "epoch": 0.04068522483940043,
      "grad_norm": 0.6620441675186157,
      "learning_rate": 9.542510526315789e-05,
      "loss": 1.3708,
      "step": 19
    },
    {
      "epoch": 0.042826552462526764,
      "grad_norm": 0.7349588871002197,
      "learning_rate": 9.48978947368421e-05,
      "loss": 1.4596,
      "step": 20
    },
    {
      "epoch": 0.044967880085653104,
      "grad_norm": 0.6277509927749634,
      "learning_rate": 9.437068421052632e-05,
      "loss": 1.3608,
      "step": 21
    },
    {
      "epoch": 0.047109207708779445,
      "grad_norm": 0.6911188960075378,
      "learning_rate": 9.384347368421052e-05,
      "loss": 1.2999,
      "step": 22
    },
    {
      "epoch": 0.04925053533190578,
      "grad_norm": 0.7883236408233643,
      "learning_rate": 9.331626315789474e-05,
      "loss": 1.6875,
      "step": 23
    },
    {
      "epoch": 0.05139186295503212,
      "grad_norm": 0.7931075096130371,
      "learning_rate": 9.278905263157894e-05,
      "loss": 1.3911,
      "step": 24
    },
    {
      "epoch": 0.05353319057815846,
      "grad_norm": 0.6752192974090576,
      "learning_rate": 9.226184210526316e-05,
      "loss": 1.2866,
      "step": 25
    },
    {
      "epoch": 0.055674518201284794,
      "grad_norm": 0.7393700480461121,
      "learning_rate": 9.173463157894736e-05,
      "loss": 1.4451,
      "step": 26
    },
    {
      "epoch": 0.057815845824411134,
      "grad_norm": 0.7108597159385681,
      "learning_rate": 9.120742105263159e-05,
      "loss": 1.415,
      "step": 27
    },
    {
      "epoch": 0.059957173447537475,
      "grad_norm": 0.7954888343811035,
      "learning_rate": 9.068021052631579e-05,
      "loss": 1.3438,
      "step": 28
    },
    {
      "epoch": 0.06209850107066381,
      "grad_norm": 0.765432596206665,
      "learning_rate": 9.0153e-05,
      "loss": 1.364,
      "step": 29
    },
    {
      "epoch": 0.06423982869379015,
      "grad_norm": 0.8585743308067322,
      "learning_rate": 8.96257894736842e-05,
      "loss": 1.4688,
      "step": 30
    },
    {
      "epoch": 0.06638115631691649,
      "grad_norm": 0.8325091004371643,
      "learning_rate": 8.909857894736842e-05,
      "loss": 1.595,
      "step": 31
    },
    {
      "epoch": 0.06852248394004283,
      "grad_norm": 0.7853144407272339,
      "learning_rate": 8.857136842105263e-05,
      "loss": 1.4109,
      "step": 32
    },
    {
      "epoch": 0.07066381156316917,
      "grad_norm": 0.8701279163360596,
      "learning_rate": 8.804415789473684e-05,
      "loss": 1.4184,
      "step": 33
    },
    {
      "epoch": 0.0728051391862955,
      "grad_norm": 0.8722823262214661,
      "learning_rate": 8.751694736842105e-05,
      "loss": 1.3662,
      "step": 34
    },
    {
      "epoch": 0.07494646680942184,
      "grad_norm": 1.042210340499878,
      "learning_rate": 8.698973684210527e-05,
      "loss": 1.523,
      "step": 35
    },
    {
      "epoch": 0.07708779443254818,
      "grad_norm": 1.1141560077667236,
      "learning_rate": 8.646252631578948e-05,
      "loss": 1.4057,
      "step": 36
    },
    {
      "epoch": 0.07922912205567452,
      "grad_norm": 1.5276129245758057,
      "learning_rate": 8.593531578947368e-05,
      "loss": 1.3753,
      "step": 37
    },
    {
      "epoch": 0.08137044967880086,
      "grad_norm": 1.219393014907837,
      "learning_rate": 8.54081052631579e-05,
      "loss": 1.2693,
      "step": 38
    },
    {
      "epoch": 0.0835117773019272,
      "grad_norm": 1.2096123695373535,
      "learning_rate": 8.48808947368421e-05,
      "loss": 1.2941,
      "step": 39
    },
    {
      "epoch": 0.08565310492505353,
      "grad_norm": 1.2960225343704224,
      "learning_rate": 8.435368421052631e-05,
      "loss": 1.4671,
      "step": 40
    },
    {
      "epoch": 0.08779443254817987,
      "grad_norm": 1.573539137840271,
      "learning_rate": 8.382647368421053e-05,
      "loss": 1.6669,
      "step": 41
    },
    {
      "epoch": 0.08993576017130621,
      "grad_norm": 1.4201273918151855,
      "learning_rate": 8.329926315789474e-05,
      "loss": 1.6799,
      "step": 42
    },
    {
      "epoch": 0.09207708779443255,
      "grad_norm": 1.5529521703720093,
      "learning_rate": 8.277205263157894e-05,
      "loss": 1.6815,
      "step": 43
    },
    {
      "epoch": 0.09421841541755889,
      "grad_norm": 1.4346344470977783,
      "learning_rate": 8.224484210526316e-05,
      "loss": 1.5518,
      "step": 44
    },
    {
      "epoch": 0.09635974304068523,
      "grad_norm": 1.3770591020584106,
      "learning_rate": 8.171763157894736e-05,
      "loss": 1.6801,
      "step": 45
    },
    {
      "epoch": 0.09850107066381156,
      "grad_norm": 1.6703461408615112,
      "learning_rate": 8.119042105263158e-05,
      "loss": 1.8867,
      "step": 46
    },
    {
      "epoch": 0.1006423982869379,
      "grad_norm": 1.524529218673706,
      "learning_rate": 8.066321052631578e-05,
      "loss": 1.7592,
      "step": 47
    },
    {
      "epoch": 0.10278372591006424,
      "grad_norm": 1.6096848249435425,
      "learning_rate": 8.0136e-05,
      "loss": 1.5122,
      "step": 48
    },
    {
      "epoch": 0.10492505353319058,
      "grad_norm": 2.0427606105804443,
      "learning_rate": 7.960878947368421e-05,
      "loss": 1.7678,
      "step": 49
    },
    {
      "epoch": 0.10706638115631692,
      "grad_norm": 1.9904823303222656,
      "learning_rate": 7.908157894736842e-05,
      "loss": 1.1802,
      "step": 50
    },
    {
      "epoch": 0.10706638115631692,
      "eval_loss": 1.4951374530792236,
      "eval_runtime": 20.9987,
      "eval_samples_per_second": 9.382,
      "eval_steps_per_second": 2.381,
      "step": 50
    },
    {
      "epoch": 0.10920770877944326,
      "grad_norm": 0.6053656339645386,
      "learning_rate": 7.855436842105262e-05,
      "loss": 1.3125,
      "step": 51
    },
    {
      "epoch": 0.11134903640256959,
      "grad_norm": 0.6838744878768921,
      "learning_rate": 7.802715789473684e-05,
      "loss": 1.4111,
      "step": 52
    },
    {
      "epoch": 0.11349036402569593,
      "grad_norm": 0.6398351192474365,
      "learning_rate": 7.749994736842104e-05,
      "loss": 1.3775,
      "step": 53
    },
    {
      "epoch": 0.11563169164882227,
      "grad_norm": 0.6202974319458008,
      "learning_rate": 7.697273684210526e-05,
      "loss": 1.3684,
      "step": 54
    },
    {
      "epoch": 0.11777301927194861,
      "grad_norm": 0.5496251583099365,
      "learning_rate": 7.644552631578947e-05,
      "loss": 1.3203,
      "step": 55
    },
    {
      "epoch": 0.11991434689507495,
      "grad_norm": 0.5528343319892883,
      "learning_rate": 7.591831578947369e-05,
      "loss": 1.3289,
      "step": 56
    },
    {
      "epoch": 0.12205567451820129,
      "grad_norm": 0.6197764277458191,
      "learning_rate": 7.539110526315789e-05,
      "loss": 1.3333,
      "step": 57
    },
    {
      "epoch": 0.12419700214132762,
      "grad_norm": 0.6319669485092163,
      "learning_rate": 7.48638947368421e-05,
      "loss": 1.4913,
      "step": 58
    },
    {
      "epoch": 0.12633832976445397,
      "grad_norm": 0.5520409345626831,
      "learning_rate": 7.433668421052632e-05,
      "loss": 1.3352,
      "step": 59
    },
    {
      "epoch": 0.1284796573875803,
      "grad_norm": 0.5273857116699219,
      "learning_rate": 7.380947368421052e-05,
      "loss": 1.3447,
      "step": 60
    },
    {
      "epoch": 0.13062098501070663,
      "grad_norm": 0.5762115716934204,
      "learning_rate": 7.328226315789473e-05,
      "loss": 1.2775,
      "step": 61
    },
    {
      "epoch": 0.13276231263383298,
      "grad_norm": 0.6039613485336304,
      "learning_rate": 7.275505263157895e-05,
      "loss": 1.1868,
      "step": 62
    },
    {
      "epoch": 0.1349036402569593,
      "grad_norm": 0.6341102719306946,
      "learning_rate": 7.222784210526316e-05,
      "loss": 1.4534,
      "step": 63
    },
    {
      "epoch": 0.13704496788008566,
      "grad_norm": 0.6609722971916199,
      "learning_rate": 7.170063157894737e-05,
      "loss": 1.5347,
      "step": 64
    },
    {
      "epoch": 0.139186295503212,
      "grad_norm": 0.5136928558349609,
      "learning_rate": 7.117342105263158e-05,
      "loss": 1.2483,
      "step": 65
    },
    {
      "epoch": 0.14132762312633834,
      "grad_norm": 0.5874619483947754,
      "learning_rate": 7.064621052631578e-05,
      "loss": 1.482,
      "step": 66
    },
    {
      "epoch": 0.14346895074946467,
      "grad_norm": 0.6109234094619751,
      "learning_rate": 7.0119e-05,
      "loss": 1.3627,
      "step": 67
    },
    {
      "epoch": 0.145610278372591,
      "grad_norm": 0.6875060796737671,
      "learning_rate": 6.959178947368421e-05,
      "loss": 1.3627,
      "step": 68
    },
    {
      "epoch": 0.14775160599571735,
      "grad_norm": 0.5841842889785767,
      "learning_rate": 6.906457894736843e-05,
      "loss": 1.1424,
      "step": 69
    },
    {
      "epoch": 0.14989293361884368,
      "grad_norm": 0.5467577576637268,
      "learning_rate": 6.853736842105263e-05,
      "loss": 1.0922,
      "step": 70
    },
    {
      "epoch": 0.15203426124197003,
      "grad_norm": 0.6267861723899841,
      "learning_rate": 6.801015789473684e-05,
      "loss": 1.3094,
      "step": 71
    },
    {
      "epoch": 0.15417558886509636,
      "grad_norm": 0.5981314778327942,
      "learning_rate": 6.748294736842105e-05,
      "loss": 1.3435,
      "step": 72
    },
    {
      "epoch": 0.15631691648822268,
      "grad_norm": 0.6704832911491394,
      "learning_rate": 6.695573684210526e-05,
      "loss": 1.4797,
      "step": 73
    },
    {
      "epoch": 0.15845824411134904,
      "grad_norm": 0.7358032464981079,
      "learning_rate": 6.642852631578946e-05,
      "loss": 1.4901,
      "step": 74
    },
    {
      "epoch": 0.16059957173447537,
      "grad_norm": 0.6578914523124695,
      "learning_rate": 6.590131578947369e-05,
      "loss": 1.3264,
      "step": 75
    },
    {
      "epoch": 0.16274089935760172,
      "grad_norm": 0.6772924065589905,
      "learning_rate": 6.537410526315789e-05,
      "loss": 1.3163,
      "step": 76
    },
    {
      "epoch": 0.16488222698072805,
      "grad_norm": 0.6964196562767029,
      "learning_rate": 6.484689473684211e-05,
      "loss": 1.3393,
      "step": 77
    },
    {
      "epoch": 0.1670235546038544,
      "grad_norm": 0.6746947765350342,
      "learning_rate": 6.431968421052631e-05,
      "loss": 1.4114,
      "step": 78
    },
    {
      "epoch": 0.16916488222698073,
      "grad_norm": 0.7291203737258911,
      "learning_rate": 6.379247368421052e-05,
      "loss": 1.3504,
      "step": 79
    },
    {
      "epoch": 0.17130620985010706,
      "grad_norm": 0.7056574821472168,
      "learning_rate": 6.326526315789474e-05,
      "loss": 1.276,
      "step": 80
    },
    {
      "epoch": 0.1734475374732334,
      "grad_norm": 0.7028514742851257,
      "learning_rate": 6.273805263157894e-05,
      "loss": 1.3409,
      "step": 81
    },
    {
      "epoch": 0.17558886509635974,
      "grad_norm": 0.8772363662719727,
      "learning_rate": 6.221084210526315e-05,
      "loss": 1.512,
      "step": 82
    },
    {
      "epoch": 0.1777301927194861,
      "grad_norm": 0.8027096390724182,
      "learning_rate": 6.168363157894737e-05,
      "loss": 1.3394,
      "step": 83
    },
    {
      "epoch": 0.17987152034261242,
      "grad_norm": 0.8053719997406006,
      "learning_rate": 6.115642105263159e-05,
      "loss": 1.3241,
      "step": 84
    },
    {
      "epoch": 0.18201284796573874,
      "grad_norm": 0.9479290843009949,
      "learning_rate": 6.0629210526315787e-05,
      "loss": 1.3755,
      "step": 85
    },
    {
      "epoch": 0.1841541755888651,
      "grad_norm": 0.9004842042922974,
      "learning_rate": 6.0101999999999995e-05,
      "loss": 1.2778,
      "step": 86
    },
    {
      "epoch": 0.18629550321199143,
      "grad_norm": 0.9118266701698303,
      "learning_rate": 5.95747894736842e-05,
      "loss": 1.2688,
      "step": 87
    },
    {
      "epoch": 0.18843683083511778,
      "grad_norm": 1.0496926307678223,
      "learning_rate": 5.904757894736841e-05,
      "loss": 1.682,
      "step": 88
    },
    {
      "epoch": 0.1905781584582441,
      "grad_norm": 1.1744533777236938,
      "learning_rate": 5.852036842105263e-05,
      "loss": 1.7398,
      "step": 89
    },
    {
      "epoch": 0.19271948608137046,
      "grad_norm": 1.1203596591949463,
      "learning_rate": 5.799315789473684e-05,
      "loss": 1.5468,
      "step": 90
    },
    {
      "epoch": 0.1948608137044968,
      "grad_norm": 1.1617553234100342,
      "learning_rate": 5.746594736842105e-05,
      "loss": 1.7715,
      "step": 91
    },
    {
      "epoch": 0.19700214132762311,
      "grad_norm": 1.3162297010421753,
      "learning_rate": 5.693873684210526e-05,
      "loss": 1.5453,
      "step": 92
    },
    {
      "epoch": 0.19914346895074947,
      "grad_norm": 1.465976595878601,
      "learning_rate": 5.641152631578947e-05,
      "loss": 1.6286,
      "step": 93
    },
    {
      "epoch": 0.2012847965738758,
      "grad_norm": 1.504904866218567,
      "learning_rate": 5.588431578947368e-05,
      "loss": 1.5897,
      "step": 94
    },
    {
      "epoch": 0.20342612419700215,
      "grad_norm": 1.3539937734603882,
      "learning_rate": 5.5357105263157896e-05,
      "loss": 1.3623,
      "step": 95
    },
    {
      "epoch": 0.20556745182012848,
      "grad_norm": 1.5463908910751343,
      "learning_rate": 5.482989473684211e-05,
      "loss": 1.5347,
      "step": 96
    },
    {
      "epoch": 0.20770877944325483,
      "grad_norm": 1.5721604824066162,
      "learning_rate": 5.430268421052632e-05,
      "loss": 1.4725,
      "step": 97
    },
    {
      "epoch": 0.20985010706638116,
      "grad_norm": 1.5885894298553467,
      "learning_rate": 5.377547368421053e-05,
      "loss": 1.509,
      "step": 98
    },
    {
      "epoch": 0.21199143468950749,
      "grad_norm": 1.7374659776687622,
      "learning_rate": 5.3248263157894736e-05,
      "loss": 1.4201,
      "step": 99
    },
    {
      "epoch": 0.21413276231263384,
      "grad_norm": 2.481282949447632,
      "learning_rate": 5.2721052631578944e-05,
      "loss": 1.3816,
      "step": 100
    },
    {
      "epoch": 0.21413276231263384,
      "eval_loss": 1.427242636680603,
      "eval_runtime": 21.0228,
      "eval_samples_per_second": 9.371,
      "eval_steps_per_second": 2.378,
      "step": 100
    },
    {
      "epoch": 0.21627408993576017,
      "grad_norm": 0.5142374634742737,
      "learning_rate": 5.219384210526315e-05,
      "loss": 1.2309,
      "step": 101
    },
    {
      "epoch": 0.21841541755888652,
      "grad_norm": 0.5063874125480652,
      "learning_rate": 5.1666631578947374e-05,
      "loss": 1.3589,
      "step": 102
    },
    {
      "epoch": 0.22055674518201285,
      "grad_norm": 0.496324360370636,
      "learning_rate": 5.113942105263158e-05,
      "loss": 1.4151,
      "step": 103
    },
    {
      "epoch": 0.22269807280513917,
      "grad_norm": 0.5609139204025269,
      "learning_rate": 5.061221052631579e-05,
      "loss": 1.479,
      "step": 104
    },
    {
      "epoch": 0.22483940042826553,
      "grad_norm": 0.533257246017456,
      "learning_rate": 5.0085e-05,
      "loss": 1.3388,
      "step": 105
    },
    {
      "epoch": 0.22698072805139186,
      "grad_norm": 0.6568904519081116,
      "learning_rate": 4.955778947368421e-05,
      "loss": 1.3423,
      "step": 106
    },
    {
      "epoch": 0.2291220556745182,
      "grad_norm": 0.5063288807868958,
      "learning_rate": 4.903057894736842e-05,
      "loss": 1.2765,
      "step": 107
    },
    {
      "epoch": 0.23126338329764454,
      "grad_norm": 0.5396547913551331,
      "learning_rate": 4.850336842105263e-05,
      "loss": 1.3108,
      "step": 108
    },
    {
      "epoch": 0.2334047109207709,
      "grad_norm": 0.5573462247848511,
      "learning_rate": 4.797615789473684e-05,
      "loss": 1.3893,
      "step": 109
    },
    {
      "epoch": 0.23554603854389722,
      "grad_norm": 0.6034949421882629,
      "learning_rate": 4.744894736842105e-05,
      "loss": 1.4021,
      "step": 110
    },
    {
      "epoch": 0.23768736616702354,
      "grad_norm": 0.5061251521110535,
      "learning_rate": 4.692173684210526e-05,
      "loss": 1.0391,
      "step": 111
    },
    {
      "epoch": 0.2398286937901499,
      "grad_norm": 0.5457649230957031,
      "learning_rate": 4.639452631578947e-05,
      "loss": 1.3074,
      "step": 112
    },
    {
      "epoch": 0.24197002141327623,
      "grad_norm": 0.5704851746559143,
      "learning_rate": 4.586731578947368e-05,
      "loss": 1.425,
      "step": 113
    },
    {
      "epoch": 0.24411134903640258,
      "grad_norm": 0.5784165263175964,
      "learning_rate": 4.5340105263157894e-05,
      "loss": 1.2314,
      "step": 114
    },
    {
      "epoch": 0.2462526766595289,
      "grad_norm": 0.5963638424873352,
      "learning_rate": 4.48128947368421e-05,
      "loss": 1.4618,
      "step": 115
    },
    {
      "epoch": 0.24839400428265523,
      "grad_norm": 0.5910683870315552,
      "learning_rate": 4.428568421052632e-05,
      "loss": 1.1647,
      "step": 116
    },
    {
      "epoch": 0.2505353319057816,
      "grad_norm": 0.6124658584594727,
      "learning_rate": 4.3758473684210525e-05,
      "loss": 1.4463,
      "step": 117
    },
    {
      "epoch": 0.25267665952890794,
      "grad_norm": 0.6125335693359375,
      "learning_rate": 4.323126315789474e-05,
      "loss": 1.2603,
      "step": 118
    },
    {
      "epoch": 0.25481798715203424,
      "grad_norm": 0.625206470489502,
      "learning_rate": 4.270405263157895e-05,
      "loss": 1.3451,
      "step": 119
    },
    {
      "epoch": 0.2569593147751606,
      "grad_norm": 0.56611168384552,
      "learning_rate": 4.217684210526316e-05,
      "loss": 1.0813,
      "step": 120
    },
    {
      "epoch": 0.25910064239828695,
      "grad_norm": 0.6412736177444458,
      "learning_rate": 4.164963157894737e-05,
      "loss": 1.2556,
      "step": 121
    },
    {
      "epoch": 0.26124197002141325,
      "grad_norm": 0.6445316076278687,
      "learning_rate": 4.112242105263158e-05,
      "loss": 1.3265,
      "step": 122
    },
    {
      "epoch": 0.2633832976445396,
      "grad_norm": 0.6097090840339661,
      "learning_rate": 4.059521052631579e-05,
      "loss": 1.2253,
      "step": 123
    },
    {
      "epoch": 0.26552462526766596,
      "grad_norm": 0.7497053146362305,
      "learning_rate": 4.0068e-05,
      "loss": 1.4526,
      "step": 124
    },
    {
      "epoch": 0.2676659528907923,
      "grad_norm": 0.6599008440971375,
      "learning_rate": 3.954078947368421e-05,
      "loss": 1.2398,
      "step": 125
    },
    {
      "epoch": 0.2698072805139186,
      "grad_norm": 0.6533190011978149,
      "learning_rate": 3.901357894736842e-05,
      "loss": 1.2709,
      "step": 126
    },
    {
      "epoch": 0.27194860813704497,
      "grad_norm": 0.7543668150901794,
      "learning_rate": 3.848636842105263e-05,
      "loss": 1.4877,
      "step": 127
    },
    {
      "epoch": 0.2740899357601713,
      "grad_norm": 0.7163957953453064,
      "learning_rate": 3.795915789473684e-05,
      "loss": 1.2075,
      "step": 128
    },
    {
      "epoch": 0.2762312633832976,
      "grad_norm": 0.6625785231590271,
      "learning_rate": 3.743194736842105e-05,
      "loss": 1.1437,
      "step": 129
    },
    {
      "epoch": 0.278372591006424,
      "grad_norm": 0.7296807169914246,
      "learning_rate": 3.690473684210526e-05,
      "loss": 1.1468,
      "step": 130
    },
    {
      "epoch": 0.28051391862955033,
      "grad_norm": 0.968973696231842,
      "learning_rate": 3.6377526315789475e-05,
      "loss": 1.3255,
      "step": 131
    },
    {
      "epoch": 0.2826552462526767,
      "grad_norm": 0.7866201996803284,
      "learning_rate": 3.585031578947368e-05,
      "loss": 1.4139,
      "step": 132
    },
    {
      "epoch": 0.284796573875803,
      "grad_norm": 0.7954109311103821,
      "learning_rate": 3.532310526315789e-05,
      "loss": 1.4214,
      "step": 133
    },
    {
      "epoch": 0.28693790149892934,
      "grad_norm": 0.8575154542922974,
      "learning_rate": 3.4795894736842106e-05,
      "loss": 1.2359,
      "step": 134
    },
    {
      "epoch": 0.2890792291220557,
      "grad_norm": 0.7624422311782837,
      "learning_rate": 3.4268684210526314e-05,
      "loss": 1.2425,
      "step": 135
    },
    {
      "epoch": 0.291220556745182,
      "grad_norm": 0.794262707233429,
      "learning_rate": 3.374147368421052e-05,
      "loss": 1.1578,
      "step": 136
    },
    {
      "epoch": 0.29336188436830835,
      "grad_norm": 0.9715974926948547,
      "learning_rate": 3.321426315789473e-05,
      "loss": 1.5022,
      "step": 137
    },
    {
      "epoch": 0.2955032119914347,
      "grad_norm": 0.9763919711112976,
      "learning_rate": 3.2687052631578946e-05,
      "loss": 1.4798,
      "step": 138
    },
    {
      "epoch": 0.29764453961456105,
      "grad_norm": 1.0104608535766602,
      "learning_rate": 3.2159842105263154e-05,
      "loss": 1.4237,
      "step": 139
    },
    {
      "epoch": 0.29978586723768735,
      "grad_norm": 1.0635695457458496,
      "learning_rate": 3.163263157894737e-05,
      "loss": 1.5483,
      "step": 140
    },
    {
      "epoch": 0.3019271948608137,
      "grad_norm": 1.1193994283676147,
      "learning_rate": 3.110542105263158e-05,
      "loss": 1.3314,
      "step": 141
    },
    {
      "epoch": 0.30406852248394006,
      "grad_norm": 1.1952261924743652,
      "learning_rate": 3.057821052631579e-05,
      "loss": 1.3942,
      "step": 142
    },
    {
      "epoch": 0.30620985010706636,
      "grad_norm": 1.217928171157837,
      "learning_rate": 3.0050999999999997e-05,
      "loss": 1.5372,
      "step": 143
    },
    {
      "epoch": 0.3083511777301927,
      "grad_norm": 1.791704773902893,
      "learning_rate": 2.9523789473684206e-05,
      "loss": 1.4163,
      "step": 144
    },
    {
      "epoch": 0.31049250535331907,
      "grad_norm": 1.3388066291809082,
      "learning_rate": 2.899657894736842e-05,
      "loss": 1.539,
      "step": 145
    },
    {
      "epoch": 0.31263383297644537,
      "grad_norm": 1.551537275314331,
      "learning_rate": 2.846936842105263e-05,
      "loss": 1.6248,
      "step": 146
    },
    {
      "epoch": 0.3147751605995717,
      "grad_norm": 1.4283239841461182,
      "learning_rate": 2.794215789473684e-05,
      "loss": 1.3883,
      "step": 147
    },
    {
      "epoch": 0.3169164882226981,
      "grad_norm": 1.760223150253296,
      "learning_rate": 2.7414947368421056e-05,
      "loss": 1.5218,
      "step": 148
    },
    {
      "epoch": 0.31905781584582443,
      "grad_norm": 1.4947000741958618,
      "learning_rate": 2.6887736842105264e-05,
      "loss": 1.2047,
      "step": 149
    },
    {
      "epoch": 0.32119914346895073,
      "grad_norm": 2.5351133346557617,
      "learning_rate": 2.6360526315789472e-05,
      "loss": 2.0579,
      "step": 150
    },
    {
      "epoch": 0.32119914346895073,
      "eval_loss": 1.3695226907730103,
      "eval_runtime": 21.0098,
      "eval_samples_per_second": 9.377,
      "eval_steps_per_second": 2.38,
      "step": 150
    },
    {
      "epoch": 0.3233404710920771,
      "grad_norm": 0.45036643743515015,
      "learning_rate": 2.5833315789473687e-05,
      "loss": 1.2182,
      "step": 151
    },
    {
      "epoch": 0.32548179871520344,
      "grad_norm": 0.4834808111190796,
      "learning_rate": 2.5306105263157895e-05,
      "loss": 1.3423,
      "step": 152
    },
    {
      "epoch": 0.32762312633832974,
      "grad_norm": 0.47053784132003784,
      "learning_rate": 2.4778894736842104e-05,
      "loss": 1.2392,
      "step": 153
    },
    {
      "epoch": 0.3297644539614561,
      "grad_norm": 0.48216691613197327,
      "learning_rate": 2.4251684210526315e-05,
      "loss": 1.3298,
      "step": 154
    },
    {
      "epoch": 0.33190578158458245,
      "grad_norm": 0.4759617745876312,
      "learning_rate": 2.3724473684210524e-05,
      "loss": 1.2354,
      "step": 155
    },
    {
      "epoch": 0.3340471092077088,
      "grad_norm": 0.47802191972732544,
      "learning_rate": 2.3197263157894735e-05,
      "loss": 1.2767,
      "step": 156
    },
    {
      "epoch": 0.3361884368308351,
      "grad_norm": 0.4797442555427551,
      "learning_rate": 2.2670052631578947e-05,
      "loss": 1.3062,
      "step": 157
    },
    {
      "epoch": 0.33832976445396146,
      "grad_norm": 0.4979493021965027,
      "learning_rate": 2.214284210526316e-05,
      "loss": 1.2686,
      "step": 158
    },
    {
      "epoch": 0.3404710920770878,
      "grad_norm": 0.5230771899223328,
      "learning_rate": 2.161563157894737e-05,
      "loss": 1.2563,
      "step": 159
    },
    {
      "epoch": 0.3426124197002141,
      "grad_norm": 0.5191596746444702,
      "learning_rate": 2.108842105263158e-05,
      "loss": 1.2641,
      "step": 160
    },
    {
      "epoch": 0.34475374732334046,
      "grad_norm": 0.5451028347015381,
      "learning_rate": 2.056121052631579e-05,
      "loss": 1.295,
      "step": 161
    },
    {
      "epoch": 0.3468950749464668,
      "grad_norm": 0.5517781972885132,
      "learning_rate": 2.0034e-05,
      "loss": 1.4134,
      "step": 162
    },
    {
      "epoch": 0.3490364025695932,
      "grad_norm": 0.5520772337913513,
      "learning_rate": 1.950678947368421e-05,
      "loss": 1.3434,
      "step": 163
    },
    {
      "epoch": 0.3511777301927195,
      "grad_norm": 0.5431661605834961,
      "learning_rate": 1.897957894736842e-05,
      "loss": 1.2326,
      "step": 164
    },
    {
      "epoch": 0.3533190578158458,
      "grad_norm": 0.5893025398254395,
      "learning_rate": 1.845236842105263e-05,
      "loss": 1.3092,
      "step": 165
    },
    {
      "epoch": 0.3554603854389722,
      "grad_norm": 0.5971054434776306,
      "learning_rate": 1.792515789473684e-05,
      "loss": 1.3007,
      "step": 166
    },
    {
      "epoch": 0.3576017130620985,
      "grad_norm": 0.5701837539672852,
      "learning_rate": 1.7397947368421053e-05,
      "loss": 1.273,
      "step": 167
    },
    {
      "epoch": 0.35974304068522484,
      "grad_norm": 0.6175338625907898,
      "learning_rate": 1.687073684210526e-05,
      "loss": 1.3302,
      "step": 168
    },
    {
      "epoch": 0.3618843683083512,
      "grad_norm": 0.6606188416481018,
      "learning_rate": 1.6343526315789473e-05,
      "loss": 1.3782,
      "step": 169
    },
    {
      "epoch": 0.3640256959314775,
      "grad_norm": 0.576826274394989,
      "learning_rate": 1.5816315789473685e-05,
      "loss": 1.2031,
      "step": 170
    },
    {
      "epoch": 0.36616702355460384,
      "grad_norm": 0.6040042638778687,
      "learning_rate": 1.5289105263157896e-05,
      "loss": 1.2756,
      "step": 171
    },
    {
      "epoch": 0.3683083511777302,
      "grad_norm": 0.6493667960166931,
      "learning_rate": 1.4761894736842103e-05,
      "loss": 1.3297,
      "step": 172
    },
    {
      "epoch": 0.37044967880085655,
      "grad_norm": 0.6507684588432312,
      "learning_rate": 1.4234684210526314e-05,
      "loss": 1.3311,
      "step": 173
    },
    {
      "epoch": 0.37259100642398285,
      "grad_norm": 0.6915395855903625,
      "learning_rate": 1.3707473684210528e-05,
      "loss": 1.2585,
      "step": 174
    },
    {
      "epoch": 0.3747323340471092,
      "grad_norm": 0.6682823300361633,
      "learning_rate": 1.3180263157894736e-05,
      "loss": 1.2924,
      "step": 175
    },
    {
      "epoch": 0.37687366167023556,
      "grad_norm": 0.7044462561607361,
      "learning_rate": 1.2653052631578948e-05,
      "loss": 1.4443,
      "step": 176
    },
    {
      "epoch": 0.37901498929336186,
      "grad_norm": 0.6910678744316101,
      "learning_rate": 1.2125842105263158e-05,
      "loss": 1.426,
      "step": 177
    },
    {
      "epoch": 0.3811563169164882,
      "grad_norm": 0.692883312702179,
      "learning_rate": 1.1598631578947368e-05,
      "loss": 1.3252,
      "step": 178
    },
    {
      "epoch": 0.38329764453961457,
      "grad_norm": 0.7135277986526489,
      "learning_rate": 1.107142105263158e-05,
      "loss": 1.2427,
      "step": 179
    },
    {
      "epoch": 0.3854389721627409,
      "grad_norm": 0.713554859161377,
      "learning_rate": 1.054421052631579e-05,
      "loss": 1.2957,
      "step": 180
    },
    {
      "epoch": 0.3875802997858672,
      "grad_norm": 0.7015721797943115,
      "learning_rate": 1.0017e-05,
      "loss": 1.22,
      "step": 181
    },
    {
      "epoch": 0.3897216274089936,
      "grad_norm": 0.7369879484176636,
      "learning_rate": 9.48978947368421e-06,
      "loss": 1.2632,
      "step": 182
    },
    {
      "epoch": 0.39186295503211993,
      "grad_norm": 0.7523634433746338,
      "learning_rate": 8.96257894736842e-06,
      "loss": 1.4292,
      "step": 183
    },
    {
      "epoch": 0.39400428265524623,
      "grad_norm": 0.810431182384491,
      "learning_rate": 8.43536842105263e-06,
      "loss": 1.3299,
      "step": 184
    },
    {
      "epoch": 0.3961456102783726,
      "grad_norm": 0.9314296245574951,
      "learning_rate": 7.908157894736842e-06,
      "loss": 1.231,
      "step": 185
    },
    {
      "epoch": 0.39828693790149894,
      "grad_norm": 0.8619545102119446,
      "learning_rate": 7.380947368421051e-06,
      "loss": 1.3958,
      "step": 186
    },
    {
      "epoch": 0.4004282655246253,
      "grad_norm": 1.000165343284607,
      "learning_rate": 6.853736842105264e-06,
      "loss": 1.4974,
      "step": 187
    },
    {
      "epoch": 0.4025695931477516,
      "grad_norm": 0.9065579771995544,
      "learning_rate": 6.326526315789474e-06,
      "loss": 1.4558,
      "step": 188
    },
    {
      "epoch": 0.40471092077087795,
      "grad_norm": 0.9769574403762817,
      "learning_rate": 5.799315789473684e-06,
      "loss": 1.4727,
      "step": 189
    },
    {
      "epoch": 0.4068522483940043,
      "grad_norm": 0.9996145367622375,
      "learning_rate": 5.272105263157895e-06,
      "loss": 1.3172,
      "step": 190
    },
    {
      "epoch": 0.4089935760171306,
      "grad_norm": 1.17989182472229,
      "learning_rate": 4.744894736842105e-06,
      "loss": 1.5406,
      "step": 191
    },
    {
      "epoch": 0.41113490364025695,
      "grad_norm": 1.2042046785354614,
      "learning_rate": 4.217684210526315e-06,
      "loss": 1.4914,
      "step": 192
    },
    {
      "epoch": 0.4132762312633833,
      "grad_norm": 1.4352355003356934,
      "learning_rate": 3.6904736842105257e-06,
      "loss": 1.809,
      "step": 193
    },
    {
      "epoch": 0.41541755888650966,
      "grad_norm": 1.3110612630844116,
      "learning_rate": 3.163263157894737e-06,
      "loss": 1.6551,
      "step": 194
    },
    {
      "epoch": 0.41755888650963596,
      "grad_norm": 1.3711544275283813,
      "learning_rate": 2.6360526315789473e-06,
      "loss": 1.486,
      "step": 195
    },
    {
      "epoch": 0.4197002141327623,
      "grad_norm": 1.410295844078064,
      "learning_rate": 2.1088421052631577e-06,
      "loss": 1.5721,
      "step": 196
    },
    {
      "epoch": 0.42184154175588867,
      "grad_norm": 1.5877814292907715,
      "learning_rate": 1.5816315789473685e-06,
      "loss": 1.8101,
      "step": 197
    },
    {
      "epoch": 0.42398286937901497,
      "grad_norm": 1.6766921281814575,
      "learning_rate": 1.0544210526315788e-06,
      "loss": 1.5397,
      "step": 198
    },
    {
      "epoch": 0.4261241970021413,
      "grad_norm": 1.6274172067642212,
      "learning_rate": 5.272105263157894e-07,
      "loss": 1.2012,
      "step": 199
    },
    {
      "epoch": 0.4282655246252677,
      "grad_norm": 1.9320127964019775,
      "learning_rate": 0.0,
      "loss": 1.2602,
      "step": 200
    },
    {
      "epoch": 0.4282655246252677,
      "eval_loss": 1.3474754095077515,
      "eval_runtime": 21.008,
      "eval_samples_per_second": 9.377,
      "eval_steps_per_second": 2.38,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 6.743758294587802e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}