last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.06473120367673237,
  "eval_steps": 5,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8840553760528564,
      "eval_runtime": 13.5126,
      "eval_samples_per_second": 3.256,
      "eval_steps_per_second": 0.222,
      "step": 0
    },
    {
      "epoch": 0.00012946240735346474,
      "grad_norm": 4.048585840442336,
      "learning_rate": 1.6170763260025876e-09,
      "loss": 7.5117,
      "step": 1
    },
    {
      "epoch": 0.00025892481470692947,
      "grad_norm": 1.4619548166976115,
      "learning_rate": 3.234152652005175e-09,
      "loss": 3.356,
      "step": 2
    },
    {
      "epoch": 0.00038838722206039424,
      "grad_norm": 1.9882128124669458,
      "learning_rate": 4.851228978007762e-09,
      "loss": 4.2322,
      "step": 3
    },
    {
      "epoch": 0.0005178496294138589,
      "grad_norm": 2.1398653569970243,
      "learning_rate": 6.46830530401035e-09,
      "loss": 3.7495,
      "step": 4
    },
    {
      "epoch": 0.0006473120367673237,
      "grad_norm": 2.558235727161999,
      "learning_rate": 8.085381630012938e-09,
      "loss": 4.8491,
      "step": 5
    },
    {
      "epoch": 0.0006473120367673237,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8840553760528564,
      "eval_runtime": 17.0556,
      "eval_samples_per_second": 2.58,
      "eval_steps_per_second": 0.176,
      "step": 5
    },
    {
      "epoch": 0.0007767744441207885,
      "grad_norm": 2.428011730931814,
      "learning_rate": 9.702457956015525e-09,
      "loss": 4.7065,
      "step": 6
    },
    {
      "epoch": 0.0009062368514742531,
      "grad_norm": 2.78377608662843,
      "learning_rate": 1.1319534282018111e-08,
      "loss": 5.9541,
      "step": 7
    },
    {
      "epoch": 0.0010356992588277179,
      "grad_norm": 1.7838311394350632,
      "learning_rate": 1.29366106080207e-08,
      "loss": 4.5327,
      "step": 8
    },
    {
      "epoch": 0.0011651616661811825,
      "grad_norm": 3.5127677525157206,
      "learning_rate": 1.4553686934023287e-08,
      "loss": 6.1851,
      "step": 9
    },
    {
      "epoch": 0.0012946240735346474,
      "grad_norm": 1.743780699692862,
      "learning_rate": 1.6170763260025877e-08,
      "loss": 3.6406,
      "step": 10
    },
    {
      "epoch": 0.0012946240735346474,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8851207494735718,
      "eval_runtime": 15.1683,
      "eval_samples_per_second": 2.901,
      "eval_steps_per_second": 0.198,
      "step": 10
    },
    {
      "epoch": 0.001424086480888112,
      "grad_norm": 2.059856514430833,
      "learning_rate": 1.778783958602846e-08,
      "loss": 4.5706,
      "step": 11
    },
    {
      "epoch": 0.001553548888241577,
      "grad_norm": 2.33399271055671,
      "learning_rate": 1.940491591203105e-08,
      "loss": 4.4399,
      "step": 12
    },
    {
      "epoch": 0.0016830112955950416,
      "grad_norm": 2.3905341429677143,
      "learning_rate": 2.1021992238033634e-08,
      "loss": 5.3672,
      "step": 13
    },
    {
      "epoch": 0.0018124737029485063,
      "grad_norm": 2.1359444235175746,
      "learning_rate": 2.2639068564036222e-08,
      "loss": 4.4424,
      "step": 14
    },
    {
      "epoch": 0.0019419361103019711,
      "grad_norm": 1.9099410879445384,
      "learning_rate": 2.425614489003881e-08,
      "loss": 4.1826,
      "step": 15
    },
    {
      "epoch": 0.0019419361103019711,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8869850635528564,
      "eval_runtime": 14.9328,
      "eval_samples_per_second": 2.947,
      "eval_steps_per_second": 0.201,
      "step": 15
    },
    {
      "epoch": 0.0020713985176554358,
      "grad_norm": 3.5697954412824697,
      "learning_rate": 2.58732212160414e-08,
      "loss": 6.3228,
      "step": 16
    },
    {
      "epoch": 0.0022008609250089004,
      "grad_norm": 2.709006358306772,
      "learning_rate": 2.749029754204399e-08,
      "loss": 5.1042,
      "step": 17
    },
    {
      "epoch": 0.002330323332362365,
      "grad_norm": 2.296208822101843,
      "learning_rate": 2.9107373868046574e-08,
      "loss": 4.438,
      "step": 18
    },
    {
      "epoch": 0.00245978573971583,
      "grad_norm": 2.1548811171938995,
      "learning_rate": 3.072445019404916e-08,
      "loss": 5.1499,
      "step": 19
    },
    {
      "epoch": 0.002589248147069295,
      "grad_norm": 2.268973635895758,
      "learning_rate": 3.2341526520051754e-08,
      "loss": 4.3889,
      "step": 20
    },
    {
      "epoch": 0.002589248147069295,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8849431872367859,
      "eval_runtime": 16.3184,
      "eval_samples_per_second": 2.696,
      "eval_steps_per_second": 0.184,
      "step": 20
    },
    {
      "epoch": 0.0027187105544227595,
      "grad_norm": 1.635931764205022,
      "learning_rate": 3.395860284605434e-08,
      "loss": 3.5371,
      "step": 21
    },
    {
      "epoch": 0.002848172961776224,
      "grad_norm": 1.6615743997623915,
      "learning_rate": 3.557567917205692e-08,
      "loss": 4.1558,
      "step": 22
    },
    {
      "epoch": 0.002977635369129689,
      "grad_norm": 2.056005996616881,
      "learning_rate": 3.7192755498059514e-08,
      "loss": 4.4595,
      "step": 23
    },
    {
      "epoch": 0.003107097776483154,
      "grad_norm": 3.753040242280733,
      "learning_rate": 3.88098318240621e-08,
      "loss": 5.7734,
      "step": 24
    },
    {
      "epoch": 0.0032365601838366185,
      "grad_norm": 2.366113463130707,
      "learning_rate": 4.042690815006469e-08,
      "loss": 3.8374,
      "step": 25
    },
    {
      "epoch": 0.0032365601838366185,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8881391882896423,
      "eval_runtime": 24.3124,
      "eval_samples_per_second": 1.81,
      "eval_steps_per_second": 0.123,
      "step": 25
    },
    {
      "epoch": 0.003366022591190083,
      "grad_norm": 2.3798748862140813,
      "learning_rate": 4.204398447606727e-08,
      "loss": 4.8389,
      "step": 26
    },
    {
      "epoch": 0.003495484998543548,
      "grad_norm": 1.9670885219929928,
      "learning_rate": 4.366106080206986e-08,
      "loss": 4.0776,
      "step": 27
    },
    {
      "epoch": 0.0036249474058970125,
      "grad_norm": 1.7473095176758717,
      "learning_rate": 4.5278137128072445e-08,
      "loss": 3.7896,
      "step": 28
    },
    {
      "epoch": 0.0037544098132504776,
      "grad_norm": 1.6502477708403551,
      "learning_rate": 4.6895213454075036e-08,
      "loss": 4.0874,
      "step": 29
    },
    {
      "epoch": 0.0038838722206039423,
      "grad_norm": 1.7502633563958625,
      "learning_rate": 4.851228978007762e-08,
      "loss": 4.7852,
      "step": 30
    },
    {
      "epoch": 0.0038838722206039423,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8831676244735718,
      "eval_runtime": 16.859,
      "eval_samples_per_second": 2.61,
      "eval_steps_per_second": 0.178,
      "step": 30
    },
    {
      "epoch": 0.004013334627957407,
      "grad_norm": 2.587615402257072,
      "learning_rate": 5.012936610608021e-08,
      "loss": 4.9688,
      "step": 31
    },
    {
      "epoch": 0.0041427970353108716,
      "grad_norm": 2.4101266831742256,
      "learning_rate": 5.17464424320828e-08,
      "loss": 4.6631,
      "step": 32
    },
    {
      "epoch": 0.004272259442664336,
      "grad_norm": 3.093327561925277,
      "learning_rate": 5.336351875808539e-08,
      "loss": 5.7959,
      "step": 33
    },
    {
      "epoch": 0.004401721850017801,
      "grad_norm": 3.3657109917022705,
      "learning_rate": 5.498059508408798e-08,
      "loss": 7.2173,
      "step": 34
    },
    {
      "epoch": 0.0045311842573712655,
      "grad_norm": 1.92332727613145,
      "learning_rate": 5.659767141009056e-08,
      "loss": 3.8652,
      "step": 35
    },
    {
      "epoch": 0.0045311842573712655,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8849431872367859,
      "eval_runtime": 15.8823,
      "eval_samples_per_second": 2.77,
      "eval_steps_per_second": 0.189,
      "step": 35
    },
    {
      "epoch": 0.00466064666472473,
      "grad_norm": 2.490260539009604,
      "learning_rate": 5.821474773609315e-08,
      "loss": 4.8418,
      "step": 36
    },
    {
      "epoch": 0.004790109072078196,
      "grad_norm": 3.273589100056769,
      "learning_rate": 5.983182406209573e-08,
      "loss": 6.3818,
      "step": 37
    },
    {
      "epoch": 0.00491957147943166,
      "grad_norm": 2.6670528339103154,
      "learning_rate": 6.144890038809832e-08,
      "loss": 6.3396,
      "step": 38
    },
    {
      "epoch": 0.005049033886785125,
      "grad_norm": 1.348631600436967,
      "learning_rate": 6.306597671410092e-08,
      "loss": 3.6206,
      "step": 39
    },
    {
      "epoch": 0.00517849629413859,
      "grad_norm": 2.2949071960950675,
      "learning_rate": 6.468305304010351e-08,
      "loss": 4.8369,
      "step": 40
    },
    {
      "epoch": 0.00517849629413859,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8851207494735718,
      "eval_runtime": 15.0616,
      "eval_samples_per_second": 2.921,
      "eval_steps_per_second": 0.199,
      "step": 40
    },
    {
      "epoch": 0.005307958701492054,
      "grad_norm": 2.5142130259666353,
      "learning_rate": 6.630012936610609e-08,
      "loss": 5.5762,
      "step": 41
    },
    {
      "epoch": 0.005437421108845519,
      "grad_norm": 2.5446434752413825,
      "learning_rate": 6.791720569210868e-08,
      "loss": 5.4185,
      "step": 42
    },
    {
      "epoch": 0.005566883516198984,
      "grad_norm": 2.821907647616337,
      "learning_rate": 6.953428201811125e-08,
      "loss": 5.5137,
      "step": 43
    },
    {
      "epoch": 0.005696345923552448,
      "grad_norm": 2.370354951113661,
      "learning_rate": 7.115135834411385e-08,
      "loss": 4.8589,
      "step": 44
    },
    {
      "epoch": 0.005825808330905913,
      "grad_norm": 2.829677972793875,
      "learning_rate": 7.276843467011644e-08,
      "loss": 5.0503,
      "step": 45
    },
    {
      "epoch": 0.005825808330905913,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8833451867103577,
      "eval_runtime": 14.8834,
      "eval_samples_per_second": 2.956,
      "eval_steps_per_second": 0.202,
      "step": 45
    },
    {
      "epoch": 0.005955270738259378,
      "grad_norm": 1.7827160663194768,
      "learning_rate": 7.438551099611903e-08,
      "loss": 4.282,
      "step": 46
    },
    {
      "epoch": 0.006084733145612842,
      "grad_norm": 2.5018263160341982,
      "learning_rate": 7.600258732212161e-08,
      "loss": 4.6943,
      "step": 47
    },
    {
      "epoch": 0.006214195552966308,
      "grad_norm": 2.718135384255652,
      "learning_rate": 7.76196636481242e-08,
      "loss": 5.6211,
      "step": 48
    },
    {
      "epoch": 0.006343657960319772,
      "grad_norm": 4.22821388187826,
      "learning_rate": 7.923673997412678e-08,
      "loss": 8.2871,
      "step": 49
    },
    {
      "epoch": 0.006473120367673237,
      "grad_norm": 1.6934454407957527,
      "learning_rate": 8.085381630012938e-08,
      "loss": 3.7959,
      "step": 50
    },
    {
      "epoch": 0.006473120367673237,
      "eval_PRM Accuracy": 0.4056603773584906,
      "eval_PRM F1": 0.4424778761061947,
      "eval_PRM F1 AUC": 0.5419067574646412,
      "eval_PRM F1 Neg": 0.36363636363636365,
      "eval_PRM NPV": 0.23684210526315788,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.30120481927710846,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8854758739471436,
      "eval_runtime": 15.1663,
      "eval_samples_per_second": 2.901,
      "eval_steps_per_second": 0.198,
      "step": 50
    },
    {
      "epoch": 0.006602582775026702,
      "grad_norm": 3.120989704267357,
      "learning_rate": 8.247089262613197e-08,
      "loss": 4.354,
      "step": 51
    },
    {
      "epoch": 0.006732045182380166,
      "grad_norm": 2.213437136965206,
      "learning_rate": 8.408796895213454e-08,
      "loss": 4.2153,
      "step": 52
    },
    {
      "epoch": 0.006861507589733631,
      "grad_norm": 3.187478732232664,
      "learning_rate": 8.570504527813713e-08,
      "loss": 6.2373,
      "step": 53
    },
    {
      "epoch": 0.006990969997087096,
      "grad_norm": 2.1448244699623036,
      "learning_rate": 8.732212160413972e-08,
      "loss": 5.0845,
      "step": 54
    },
    {
      "epoch": 0.00712043240444056,
      "grad_norm": 2.993483048913676,
      "learning_rate": 8.893919793014231e-08,
      "loss": 4.8887,
      "step": 55
    },
    {
      "epoch": 0.00712043240444056,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8858309388160706,
      "eval_runtime": 15.8335,
      "eval_samples_per_second": 2.779,
      "eval_steps_per_second": 0.189,
      "step": 55
    },
    {
      "epoch": 0.007249894811794025,
      "grad_norm": 3.5666108788118236,
      "learning_rate": 9.055627425614489e-08,
      "loss": 5.6782,
      "step": 56
    },
    {
      "epoch": 0.00737935721914749,
      "grad_norm": 2.2891689346612236,
      "learning_rate": 9.217335058214748e-08,
      "loss": 4.1877,
      "step": 57
    },
    {
      "epoch": 0.007508819626500955,
      "grad_norm": 2.5428379612016823,
      "learning_rate": 9.379042690815007e-08,
      "loss": 5.2524,
      "step": 58
    },
    {
      "epoch": 0.00763828203385442,
      "grad_norm": 3.8061609228898834,
      "learning_rate": 9.540750323415266e-08,
      "loss": 6.4536,
      "step": 59
    },
    {
      "epoch": 0.0077677444412078845,
      "grad_norm": 2.2419630022062567,
      "learning_rate": 9.702457956015524e-08,
      "loss": 4.6221,
      "step": 60
    },
    {
      "epoch": 0.0077677444412078845,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8837890625,
      "eval_runtime": 15.4702,
      "eval_samples_per_second": 2.844,
      "eval_steps_per_second": 0.194,
      "step": 60
    },
    {
      "epoch": 0.007897206848561348,
      "grad_norm": 1.8782241280158536,
      "learning_rate": 9.864165588615783e-08,
      "loss": 4.4683,
      "step": 61
    },
    {
      "epoch": 0.008026669255914814,
      "grad_norm": 3.570226272161863,
      "learning_rate": 1.0025873221216042e-07,
      "loss": 7.3521,
      "step": 62
    },
    {
      "epoch": 0.00815613166326828,
      "grad_norm": 2.51045834876466,
      "learning_rate": 1.0187580853816302e-07,
      "loss": 5.2793,
      "step": 63
    },
    {
      "epoch": 0.008285594070621743,
      "grad_norm": 1.5575108426738207,
      "learning_rate": 1.034928848641656e-07,
      "loss": 3.665,
      "step": 64
    },
    {
      "epoch": 0.008415056477975209,
      "grad_norm": 1.3322399221426193,
      "learning_rate": 1.0510996119016818e-07,
      "loss": 3.3931,
      "step": 65
    },
    {
      "epoch": 0.008415056477975209,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8874289989471436,
      "eval_runtime": 15.9089,
      "eval_samples_per_second": 2.766,
      "eval_steps_per_second": 0.189,
      "step": 65
    },
    {
      "epoch": 0.008544518885328672,
      "grad_norm": 2.492364990454832,
      "learning_rate": 1.0672703751617078e-07,
      "loss": 5.5532,
      "step": 66
    },
    {
      "epoch": 0.008673981292682138,
      "grad_norm": 3.7390863760396558,
      "learning_rate": 1.0834411384217337e-07,
      "loss": 6.2378,
      "step": 67
    },
    {
      "epoch": 0.008803443700035602,
      "grad_norm": 1.8473156925064862,
      "learning_rate": 1.0996119016817596e-07,
      "loss": 4.2271,
      "step": 68
    },
    {
      "epoch": 0.008932906107389067,
      "grad_norm": 1.8988855304261645,
      "learning_rate": 1.1157826649417852e-07,
      "loss": 4.6084,
      "step": 69
    },
    {
      "epoch": 0.009062368514742531,
      "grad_norm": 1.7803128991594623,
      "learning_rate": 1.1319534282018111e-07,
      "loss": 4.1704,
      "step": 70
    },
    {
      "epoch": 0.009062368514742531,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8866299986839294,
      "eval_runtime": 17.0252,
      "eval_samples_per_second": 2.584,
      "eval_steps_per_second": 0.176,
      "step": 70
    },
    {
      "epoch": 0.009191830922095997,
      "grad_norm": 3.5763120139801003,
      "learning_rate": 1.148124191461837e-07,
      "loss": 6.332,
      "step": 71
    },
    {
      "epoch": 0.00932129332944946,
      "grad_norm": 2.2141728938352876,
      "learning_rate": 1.164294954721863e-07,
      "loss": 4.2615,
      "step": 72
    },
    {
      "epoch": 0.009450755736802926,
      "grad_norm": 3.2073335886129595,
      "learning_rate": 1.1804657179818888e-07,
      "loss": 6.3149,
      "step": 73
    },
    {
      "epoch": 0.009580218144156391,
      "grad_norm": 1.3526098832419702,
      "learning_rate": 1.1966364812419147e-07,
      "loss": 3.5325,
      "step": 74
    },
    {
      "epoch": 0.009709680551509855,
      "grad_norm": 1.6799971849764905,
      "learning_rate": 1.2128072445019406e-07,
      "loss": 3.2056,
      "step": 75
    },
    {
      "epoch": 0.009709680551509855,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8876953125,
      "eval_runtime": 17.0093,
      "eval_samples_per_second": 2.587,
      "eval_steps_per_second": 0.176,
      "step": 75
    },
    {
      "epoch": 0.00983914295886332,
      "grad_norm": 2.924083764881815,
      "learning_rate": 1.2289780077619665e-07,
      "loss": 4.8882,
      "step": 76
    },
    {
      "epoch": 0.009968605366216785,
      "grad_norm": 1.5295160943757464,
      "learning_rate": 1.2451487710219921e-07,
      "loss": 3.3582,
      "step": 77
    },
    {
      "epoch": 0.01009806777357025,
      "grad_norm": 3.3154666333701295,
      "learning_rate": 1.2613195342820183e-07,
      "loss": 6.1145,
      "step": 78
    },
    {
      "epoch": 0.010227530180923714,
      "grad_norm": 3.8710027337937976,
      "learning_rate": 1.2774902975420442e-07,
      "loss": 7.8452,
      "step": 79
    },
    {
      "epoch": 0.01035699258827718,
      "grad_norm": 2.766212572050214,
      "learning_rate": 1.2936610608020701e-07,
      "loss": 5.6997,
      "step": 80
    },
    {
      "epoch": 0.01035699258827718,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.884765625,
      "eval_runtime": 15.54,
      "eval_samples_per_second": 2.831,
      "eval_steps_per_second": 0.193,
      "step": 80
    },
    {
      "epoch": 0.010486454995630643,
      "grad_norm": 3.1439318411520234,
      "learning_rate": 1.3098318240620958e-07,
      "loss": 5.9404,
      "step": 81
    },
    {
      "epoch": 0.010615917402984109,
      "grad_norm": 2.632291719511737,
      "learning_rate": 1.3260025873221217e-07,
      "loss": 4.7563,
      "step": 82
    },
    {
      "epoch": 0.010745379810337572,
      "grad_norm": 2.1378632862198614,
      "learning_rate": 1.3421733505821476e-07,
      "loss": 4.5176,
      "step": 83
    },
    {
      "epoch": 0.010874842217691038,
      "grad_norm": 3.0020895997866677,
      "learning_rate": 1.3583441138421735e-07,
      "loss": 5.7744,
      "step": 84
    },
    {
      "epoch": 0.011004304625044503,
      "grad_norm": 2.300999714939973,
      "learning_rate": 1.3745148771021992e-07,
      "loss": 4.9636,
      "step": 85
    },
    {
      "epoch": 0.011004304625044503,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8877840638160706,
      "eval_runtime": 15.4949,
      "eval_samples_per_second": 2.84,
      "eval_steps_per_second": 0.194,
      "step": 85
    },
    {
      "epoch": 0.011133767032397967,
      "grad_norm": 2.3778393912517926,
      "learning_rate": 1.390685640362225e-07,
      "loss": 5.46,
      "step": 86
    },
    {
      "epoch": 0.011263229439751433,
      "grad_norm": 2.265376177284017,
      "learning_rate": 1.406856403622251e-07,
      "loss": 4.312,
      "step": 87
    },
    {
      "epoch": 0.011392691847104897,
      "grad_norm": 1.975378983729391,
      "learning_rate": 1.423027166882277e-07,
      "loss": 3.9331,
      "step": 88
    },
    {
      "epoch": 0.011522154254458362,
      "grad_norm": 3.2796357543801578,
      "learning_rate": 1.4391979301423028e-07,
      "loss": 7.1753,
      "step": 89
    },
    {
      "epoch": 0.011651616661811826,
      "grad_norm": 2.537483324528456,
      "learning_rate": 1.4553686934023288e-07,
      "loss": 5.0693,
      "step": 90
    },
    {
      "epoch": 0.011651616661811826,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8862748742103577,
      "eval_runtime": 15.0123,
      "eval_samples_per_second": 2.931,
      "eval_steps_per_second": 0.2,
      "step": 90
    },
    {
      "epoch": 0.011781079069165291,
      "grad_norm": 1.5533230028334497,
      "learning_rate": 1.4715394566623547e-07,
      "loss": 4.3076,
      "step": 91
    },
    {
      "epoch": 0.011910541476518755,
      "grad_norm": 2.3953687079845194,
      "learning_rate": 1.4877102199223806e-07,
      "loss": 4.5845,
      "step": 92
    },
    {
      "epoch": 0.01204000388387222,
      "grad_norm": 1.6248421691564126,
      "learning_rate": 1.5038809831824065e-07,
      "loss": 3.7153,
      "step": 93
    },
    {
      "epoch": 0.012169466291225685,
      "grad_norm": 3.1074035832168847,
      "learning_rate": 1.5200517464424321e-07,
      "loss": 5.8594,
      "step": 94
    },
    {
      "epoch": 0.01229892869857915,
      "grad_norm": 1.6482149797758445,
      "learning_rate": 1.536222509702458e-07,
      "loss": 3.4629,
      "step": 95
    },
    {
      "epoch": 0.01229892869857915,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8864524364471436,
      "eval_runtime": 15.5412,
      "eval_samples_per_second": 2.831,
      "eval_steps_per_second": 0.193,
      "step": 95
    },
    {
      "epoch": 0.012428391105932616,
      "grad_norm": 3.6188297855501768,
      "learning_rate": 1.552393272962484e-07,
      "loss": 6.8918,
      "step": 96
    },
    {
      "epoch": 0.01255785351328608,
      "grad_norm": 2.7987644398773637,
      "learning_rate": 1.56856403622251e-07,
      "loss": 6.1279,
      "step": 97
    },
    {
      "epoch": 0.012687315920639545,
      "grad_norm": 2.3520379888616576,
      "learning_rate": 1.5847347994825355e-07,
      "loss": 4.769,
      "step": 98
    },
    {
      "epoch": 0.012816778327993009,
      "grad_norm": 3.246292126351779,
      "learning_rate": 1.6009055627425617e-07,
      "loss": 6.2231,
      "step": 99
    },
    {
      "epoch": 0.012946240735346474,
      "grad_norm": 2.6083710560500837,
      "learning_rate": 1.6170763260025876e-07,
      "loss": 5.2629,
      "step": 100
    },
    {
      "epoch": 0.012946240735346474,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8890269994735718,
      "eval_runtime": 14.9289,
      "eval_samples_per_second": 2.947,
      "eval_steps_per_second": 0.201,
      "step": 100
    },
    {
      "epoch": 0.013075703142699938,
      "grad_norm": 2.6541142347399296,
      "learning_rate": 1.6332470892626133e-07,
      "loss": 5.1086,
      "step": 101
    },
    {
      "epoch": 0.013205165550053403,
      "grad_norm": 1.6494228492552965,
      "learning_rate": 1.6494178525226394e-07,
      "loss": 3.917,
      "step": 102
    },
    {
      "epoch": 0.013334627957406867,
      "grad_norm": 2.921868684935309,
      "learning_rate": 1.665588615782665e-07,
      "loss": 6.1169,
      "step": 103
    },
    {
      "epoch": 0.013464090364760333,
      "grad_norm": 2.13241188421152,
      "learning_rate": 1.6817593790426907e-07,
      "loss": 4.29,
      "step": 104
    },
    {
      "epoch": 0.013593552772113798,
      "grad_norm": 1.1749121931211275,
      "learning_rate": 1.697930142302717e-07,
      "loss": 3.2705,
      "step": 105
    },
    {
      "epoch": 0.013593552772113798,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8862748742103577,
      "eval_runtime": 16.0015,
      "eval_samples_per_second": 2.75,
      "eval_steps_per_second": 0.187,
      "step": 105
    },
    {
      "epoch": 0.013723015179467262,
      "grad_norm": 1.575721321415035,
      "learning_rate": 1.7141009055627426e-07,
      "loss": 3.2231,
      "step": 106
    },
    {
      "epoch": 0.013852477586820728,
      "grad_norm": 3.1740074105498346,
      "learning_rate": 1.7302716688227687e-07,
      "loss": 6.1416,
      "step": 107
    },
    {
      "epoch": 0.013981939994174191,
      "grad_norm": 2.8235795914302306,
      "learning_rate": 1.7464424320827944e-07,
      "loss": 6.5957,
      "step": 108
    },
    {
      "epoch": 0.014111402401527657,
      "grad_norm": 2.7800697021429106,
      "learning_rate": 1.7626131953428203e-07,
      "loss": 5.5332,
      "step": 109
    },
    {
      "epoch": 0.01424086480888112,
      "grad_norm": 1.9166248786656255,
      "learning_rate": 1.7787839586028462e-07,
      "loss": 4.1089,
      "step": 110
    },
    {
      "epoch": 0.01424086480888112,
      "eval_PRM Accuracy": 0.4056603773584906,
      "eval_PRM F1": 0.4424778761061947,
      "eval_PRM F1 AUC": 0.5419067574646412,
      "eval_PRM F1 Neg": 0.36363636363636365,
      "eval_PRM NPV": 0.23684210526315788,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.30120481927710846,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8856534361839294,
      "eval_runtime": 14.6177,
      "eval_samples_per_second": 3.01,
      "eval_steps_per_second": 0.205,
      "step": 110
    },
    {
      "epoch": 0.014370327216234586,
      "grad_norm": 2.7418573899602325,
      "learning_rate": 1.7949547218628721e-07,
      "loss": 5.448,
      "step": 111
    },
    {
      "epoch": 0.01449978962358805,
      "grad_norm": 1.791130152898085,
      "learning_rate": 1.8111254851228978e-07,
      "loss": 3.7705,
      "step": 112
    },
    {
      "epoch": 0.014629252030941516,
      "grad_norm": 1.7712886879433232,
      "learning_rate": 1.827296248382924e-07,
      "loss": 4.0708,
      "step": 113
    },
    {
      "epoch": 0.01475871443829498,
      "grad_norm": 1.4011926721258692,
      "learning_rate": 1.8434670116429496e-07,
      "loss": 3.3662,
      "step": 114
    },
    {
      "epoch": 0.014888176845648445,
      "grad_norm": 2.5362570649868417,
      "learning_rate": 1.8596377749029758e-07,
      "loss": 4.9705,
      "step": 115
    },
    {
      "epoch": 0.014888176845648445,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8846768736839294,
      "eval_runtime": 16.422,
      "eval_samples_per_second": 2.679,
      "eval_steps_per_second": 0.183,
      "step": 115
    },
    {
      "epoch": 0.01501763925300191,
      "grad_norm": 2.0416731238959525,
      "learning_rate": 1.8758085381630014e-07,
      "loss": 3.9224,
      "step": 116
    },
    {
      "epoch": 0.015147101660355374,
      "grad_norm": 2.4303010252138733,
      "learning_rate": 1.891979301423027e-07,
      "loss": 5.5867,
      "step": 117
    },
    {
      "epoch": 0.01527656406770884,
      "grad_norm": 2.913544758821116,
      "learning_rate": 1.9081500646830533e-07,
      "loss": 5.1328,
      "step": 118
    },
    {
      "epoch": 0.015406026475062303,
      "grad_norm": 2.0386331121658685,
      "learning_rate": 1.924320827943079e-07,
      "loss": 4.9512,
      "step": 119
    },
    {
      "epoch": 0.015535488882415769,
      "grad_norm": 2.6467398937578515,
      "learning_rate": 1.9404915912031048e-07,
      "loss": 5.2549,
      "step": 120
    },
    {
      "epoch": 0.015535488882415769,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8829900622367859,
      "eval_runtime": 15.9024,
      "eval_samples_per_second": 2.767,
      "eval_steps_per_second": 0.189,
      "step": 120
    },
    {
      "epoch": 0.015664951289769233,
      "grad_norm": 2.4626541900261922,
      "learning_rate": 1.9566623544631307e-07,
      "loss": 4.1807,
      "step": 121
    },
    {
      "epoch": 0.015794413697122697,
      "grad_norm": 1.8878277717350793,
      "learning_rate": 1.9728331177231567e-07,
      "loss": 3.5337,
      "step": 122
    },
    {
      "epoch": 0.015923876104476164,
      "grad_norm": 2.8574871996520304,
      "learning_rate": 1.9890038809831828e-07,
      "loss": 5.0161,
      "step": 123
    },
    {
      "epoch": 0.016053338511829628,
      "grad_norm": 1.9082654564956048,
      "learning_rate": 2.0051746442432085e-07,
      "loss": 4.3105,
      "step": 124
    },
    {
      "epoch": 0.01618280091918309,
      "grad_norm": 2.210541489132059,
      "learning_rate": 2.021345407503234e-07,
      "loss": 4.6948,
      "step": 125
    },
    {
      "epoch": 0.01618280091918309,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8853870630264282,
      "eval_runtime": 15.0559,
      "eval_samples_per_second": 2.922,
      "eval_steps_per_second": 0.199,
      "step": 125
    },
    {
      "epoch": 0.01631226332653656,
      "grad_norm": 2.7336759960768435,
      "learning_rate": 2.0375161707632603e-07,
      "loss": 5.4995,
      "step": 126
    },
    {
      "epoch": 0.016441725733890022,
      "grad_norm": 2.244046155368536,
      "learning_rate": 2.053686934023286e-07,
      "loss": 4.2437,
      "step": 127
    },
    {
      "epoch": 0.016571188141243486,
      "grad_norm": 2.45262536092704,
      "learning_rate": 2.069857697283312e-07,
      "loss": 4.8989,
      "step": 128
    },
    {
      "epoch": 0.01670065054859695,
      "grad_norm": 3.3040287729974747,
      "learning_rate": 2.0860284605433378e-07,
      "loss": 6.6123,
      "step": 129
    },
    {
      "epoch": 0.016830112955950417,
      "grad_norm": 2.9261239742898564,
      "learning_rate": 2.1021992238033637e-07,
      "loss": 4.8569,
      "step": 130
    },
    {
      "epoch": 0.016830112955950417,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8856534361839294,
      "eval_runtime": 15.5434,
      "eval_samples_per_second": 2.831,
      "eval_steps_per_second": 0.193,
      "step": 130
    },
    {
      "epoch": 0.01695957536330388,
      "grad_norm": 1.8669972552610263,
      "learning_rate": 2.1183699870633896e-07,
      "loss": 4.4214,
      "step": 131
    },
    {
      "epoch": 0.017089037770657345,
      "grad_norm": 2.8355338022322005,
      "learning_rate": 2.1345407503234155e-07,
      "loss": 5.6006,
      "step": 132
    },
    {
      "epoch": 0.01721850017801081,
      "grad_norm": 3.8775566026344146,
      "learning_rate": 2.1507115135834412e-07,
      "loss": 6.9272,
      "step": 133
    },
    {
      "epoch": 0.017347962585364276,
      "grad_norm": 2.1906529658414975,
      "learning_rate": 2.1668822768434673e-07,
      "loss": 4.6504,
      "step": 134
    },
    {
      "epoch": 0.01747742499271774,
      "grad_norm": 1.7570863207219325,
      "learning_rate": 2.183053040103493e-07,
      "loss": 4.4053,
      "step": 135
    },
    {
      "epoch": 0.01747742499271774,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8849431872367859,
      "eval_runtime": 15.3918,
      "eval_samples_per_second": 2.859,
      "eval_steps_per_second": 0.195,
      "step": 135
    },
    {
      "epoch": 0.017606887400071203,
      "grad_norm": 1.3841515491996987,
      "learning_rate": 2.1992238033635192e-07,
      "loss": 3.7124,
      "step": 136
    },
    {
      "epoch": 0.01773634980742467,
      "grad_norm": 1.178526476409248,
      "learning_rate": 2.2153945666235448e-07,
      "loss": 3.4648,
      "step": 137
    },
    {
      "epoch": 0.017865812214778135,
      "grad_norm": 3.1071071039731044,
      "learning_rate": 2.2315653298835705e-07,
      "loss": 6.1064,
      "step": 138
    },
    {
      "epoch": 0.0179952746221316,
      "grad_norm": 2.5958202424020924,
      "learning_rate": 2.2477360931435966e-07,
      "loss": 5.0176,
      "step": 139
    },
    {
      "epoch": 0.018124737029485062,
      "grad_norm": 3.5329237791889043,
      "learning_rate": 2.2639068564036223e-07,
      "loss": 7.2939,
      "step": 140
    },
    {
      "epoch": 0.018124737029485062,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8851207494735718,
      "eval_runtime": 15.4656,
      "eval_samples_per_second": 2.845,
      "eval_steps_per_second": 0.194,
      "step": 140
    },
    {
      "epoch": 0.01825419943683853,
      "grad_norm": 1.698099498401497,
      "learning_rate": 2.2800776196636482e-07,
      "loss": 3.9492,
      "step": 141
    },
    {
      "epoch": 0.018383661844191993,
      "grad_norm": 2.0444575058001493,
      "learning_rate": 2.296248382923674e-07,
      "loss": 4.1685,
      "step": 142
    },
    {
      "epoch": 0.018513124251545457,
      "grad_norm": 2.420495782296046,
      "learning_rate": 2.3124191461837e-07,
      "loss": 5.0999,
      "step": 143
    },
    {
      "epoch": 0.01864258665889892,
      "grad_norm": 2.2734740240960436,
      "learning_rate": 2.328589909443726e-07,
      "loss": 4.312,
      "step": 144
    },
    {
      "epoch": 0.018772049066252388,
      "grad_norm": 2.5306624787049286,
      "learning_rate": 2.3447606727037519e-07,
      "loss": 4.9043,
      "step": 145
    },
    {
      "epoch": 0.018772049066252388,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8851207494735718,
      "eval_runtime": 16.3471,
      "eval_samples_per_second": 2.692,
      "eval_steps_per_second": 0.184,
      "step": 145
    },
    {
      "epoch": 0.018901511473605852,
      "grad_norm": 1.4008106167888073,
      "learning_rate": 2.3609314359637775e-07,
      "loss": 3.7585,
      "step": 146
    },
    {
      "epoch": 0.019030973880959316,
      "grad_norm": 3.7456861553981318,
      "learning_rate": 2.3771021992238037e-07,
      "loss": 7.1523,
      "step": 147
    },
    {
      "epoch": 0.019160436288312783,
      "grad_norm": 2.641442840623338,
      "learning_rate": 2.3932729624838293e-07,
      "loss": 5.2881,
      "step": 148
    },
    {
      "epoch": 0.019289898695666247,
      "grad_norm": 1.9569795328315067,
      "learning_rate": 2.409443725743855e-07,
      "loss": 4.395,
      "step": 149
    },
    {
      "epoch": 0.01941936110301971,
      "grad_norm": 2.3216166306121555,
      "learning_rate": 2.425614489003881e-07,
      "loss": 5.334,
      "step": 150
    },
    {
      "epoch": 0.01941936110301971,
      "eval_PRM Accuracy": 0.4056603773584906,
      "eval_PRM F1": 0.4424778761061947,
      "eval_PRM F1 AUC": 0.5419067574646412,
      "eval_PRM F1 Neg": 0.36363636363636365,
      "eval_PRM NPV": 0.23684210526315788,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.30120481927710846,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8857421875,
      "eval_runtime": 15.7514,
      "eval_samples_per_second": 2.793,
      "eval_steps_per_second": 0.19,
      "step": 150
    },
    {
      "epoch": 0.019548823510373174,
      "grad_norm": 2.1971180558555403,
      "learning_rate": 2.441785252263907e-07,
      "loss": 4.4736,
      "step": 151
    },
    {
      "epoch": 0.01967828591772664,
      "grad_norm": 3.150682034268968,
      "learning_rate": 2.457956015523933e-07,
      "loss": 6.6304,
      "step": 152
    },
    {
      "epoch": 0.019807748325080105,
      "grad_norm": 1.7781467557657258,
      "learning_rate": 2.4741267787839586e-07,
      "loss": 4.0156,
      "step": 153
    },
    {
      "epoch": 0.01993721073243357,
      "grad_norm": 1.9218504498599223,
      "learning_rate": 2.4902975420439843e-07,
      "loss": 4.6631,
      "step": 154
    },
    {
      "epoch": 0.020066673139787033,
      "grad_norm": 3.0912369033963416,
      "learning_rate": 2.5064683053040105e-07,
      "loss": 5.2312,
      "step": 155
    },
    {
      "epoch": 0.020066673139787033,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8844993114471436,
      "eval_runtime": 14.5707,
      "eval_samples_per_second": 3.02,
      "eval_steps_per_second": 0.206,
      "step": 155
    },
    {
      "epoch": 0.0201961355471405,
      "grad_norm": 3.374085982719268,
      "learning_rate": 2.5226390685640366e-07,
      "loss": 4.8442,
      "step": 156
    },
    {
      "epoch": 0.020325597954493964,
      "grad_norm": 2.6023678787454925,
      "learning_rate": 2.5388098318240623e-07,
      "loss": 5.2119,
      "step": 157
    },
    {
      "epoch": 0.020455060361847428,
      "grad_norm": 3.106650050358539,
      "learning_rate": 2.5549805950840885e-07,
      "loss": 5.978,
      "step": 158
    },
    {
      "epoch": 0.020584522769200895,
      "grad_norm": 2.240948031320514,
      "learning_rate": 2.571151358344114e-07,
      "loss": 4.7949,
      "step": 159
    },
    {
      "epoch": 0.02071398517655436,
      "grad_norm": 1.7837401909792663,
      "learning_rate": 2.5873221216041403e-07,
      "loss": 4.1484,
      "step": 160
    },
    {
      "epoch": 0.02071398517655436,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8870738744735718,
      "eval_runtime": 15.8666,
      "eval_samples_per_second": 2.773,
      "eval_steps_per_second": 0.189,
      "step": 160
    },
    {
      "epoch": 0.020843447583907822,
      "grad_norm": 2.772672736542775,
      "learning_rate": 2.603492884864166e-07,
      "loss": 4.7573,
      "step": 161
    },
    {
      "epoch": 0.020972909991261286,
      "grad_norm": 1.3714409296270307,
      "learning_rate": 2.6196636481241916e-07,
      "loss": 3.5115,
      "step": 162
    },
    {
      "epoch": 0.021102372398614754,
      "grad_norm": 2.031963096215658,
      "learning_rate": 2.635834411384218e-07,
      "loss": 4.1006,
      "step": 163
    },
    {
      "epoch": 0.021231834805968217,
      "grad_norm": 2.1993468910076066,
      "learning_rate": 2.6520051746442434e-07,
      "loss": 4.301,
      "step": 164
    },
    {
      "epoch": 0.02136129721332168,
      "grad_norm": 2.259418281777508,
      "learning_rate": 2.6681759379042696e-07,
      "loss": 4.8267,
      "step": 165
    },
    {
      "epoch": 0.02136129721332168,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8838778138160706,
      "eval_runtime": 15.4824,
      "eval_samples_per_second": 2.842,
      "eval_steps_per_second": 0.194,
      "step": 165
    },
    {
      "epoch": 0.021490759620675145,
      "grad_norm": 3.73088511507556,
      "learning_rate": 2.684346701164295e-07,
      "loss": 6.0649,
      "step": 166
    },
    {
      "epoch": 0.021620222028028612,
      "grad_norm": 2.6557645389732873,
      "learning_rate": 2.700517464424321e-07,
      "loss": 4.6919,
      "step": 167
    },
    {
      "epoch": 0.021749684435382076,
      "grad_norm": 1.997316671729813,
      "learning_rate": 2.716688227684347e-07,
      "loss": 4.7817,
      "step": 168
    },
    {
      "epoch": 0.02187914684273554,
      "grad_norm": 1.9657324745031033,
      "learning_rate": 2.7328589909443727e-07,
      "loss": 4.4263,
      "step": 169
    },
    {
      "epoch": 0.022008609250089007,
      "grad_norm": 2.4627709291744835,
      "learning_rate": 2.7490297542043984e-07,
      "loss": 5.0391,
      "step": 170
    },
    {
      "epoch": 0.022008609250089007,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8838778138160706,
      "eval_runtime": 14.1935,
      "eval_samples_per_second": 3.1,
      "eval_steps_per_second": 0.211,
      "step": 170
    },
    {
      "epoch": 0.02213807165744247,
      "grad_norm": 3.052788131039653,
      "learning_rate": 2.7652005174644245e-07,
      "loss": 4.7004,
      "step": 171
    },
    {
      "epoch": 0.022267534064795935,
      "grad_norm": 2.6186961622702554,
      "learning_rate": 2.78137128072445e-07,
      "loss": 4.7441,
      "step": 172
    },
    {
      "epoch": 0.0223969964721494,
      "grad_norm": 3.72505733347434,
      "learning_rate": 2.7975420439844764e-07,
      "loss": 7.8696,
      "step": 173
    },
    {
      "epoch": 0.022526458879502866,
      "grad_norm": 1.9464865151658428,
      "learning_rate": 2.813712807244502e-07,
      "loss": 3.6841,
      "step": 174
    },
    {
      "epoch": 0.02265592128685633,
      "grad_norm": 3.0311616691881302,
      "learning_rate": 2.8298835705045277e-07,
      "loss": 5.6128,
      "step": 175
    },
    {
      "epoch": 0.02265592128685633,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8851207494735718,
      "eval_runtime": 14.8654,
      "eval_samples_per_second": 2.96,
      "eval_steps_per_second": 0.202,
      "step": 175
    },
    {
      "epoch": 0.022785383694209793,
      "grad_norm": 3.2153799248538815,
      "learning_rate": 2.846054333764554e-07,
      "loss": 5.5913,
      "step": 176
    },
    {
      "epoch": 0.022914846101563257,
      "grad_norm": 3.1273926216154497,
      "learning_rate": 2.86222509702458e-07,
      "loss": 5.6323,
      "step": 177
    },
    {
      "epoch": 0.023044308508916724,
      "grad_norm": 2.723639596237429,
      "learning_rate": 2.8783958602846057e-07,
      "loss": 5.3838,
      "step": 178
    },
    {
      "epoch": 0.023173770916270188,
      "grad_norm": 2.7628886706248044,
      "learning_rate": 2.894566623544632e-07,
      "loss": 5.126,
      "step": 179
    },
    {
      "epoch": 0.023303233323623652,
      "grad_norm": 2.3460860198951368,
      "learning_rate": 2.9107373868046575e-07,
      "loss": 4.2046,
      "step": 180
    },
    {
      "epoch": 0.023303233323623652,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8838778138160706,
      "eval_runtime": 14.4704,
      "eval_samples_per_second": 3.041,
      "eval_steps_per_second": 0.207,
      "step": 180
    },
    {
      "epoch": 0.02343269573097712,
      "grad_norm": 1.8977281350603303,
      "learning_rate": 2.9269081500646837e-07,
      "loss": 3.7952,
      "step": 181
    },
    {
      "epoch": 0.023562158138330583,
      "grad_norm": 4.633373815126988,
      "learning_rate": 2.9430789133247093e-07,
      "loss": 8.4346,
      "step": 182
    },
    {
      "epoch": 0.023691620545684047,
      "grad_norm": 1.3213131050692126,
      "learning_rate": 2.959249676584735e-07,
      "loss": 3.7607,
      "step": 183
    },
    {
      "epoch": 0.02382108295303751,
      "grad_norm": 1.990277012223307,
      "learning_rate": 2.975420439844761e-07,
      "loss": 4.9785,
      "step": 184
    },
    {
      "epoch": 0.023950545360390978,
      "grad_norm": 2.239661015493053,
      "learning_rate": 2.991591203104787e-07,
      "loss": 5.7068,
      "step": 185
    },
    {
      "epoch": 0.023950545360390978,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8870738744735718,
      "eval_runtime": 14.5341,
      "eval_samples_per_second": 3.027,
      "eval_steps_per_second": 0.206,
      "step": 185
    },
    {
      "epoch": 0.02408000776774444,
      "grad_norm": 1.7206395570915416,
      "learning_rate": 3.007761966364813e-07,
      "loss": 4.1523,
      "step": 186
    },
    {
      "epoch": 0.024209470175097905,
      "grad_norm": 2.703519646145678,
      "learning_rate": 3.0239327296248386e-07,
      "loss": 5.522,
      "step": 187
    },
    {
      "epoch": 0.02433893258245137,
      "grad_norm": 1.8351375522927664,
      "learning_rate": 3.0401034928848643e-07,
      "loss": 4.0796,
      "step": 188
    },
    {
      "epoch": 0.024468394989804836,
      "grad_norm": 3.7534253262775192,
      "learning_rate": 3.0562742561448905e-07,
      "loss": 7.3333,
      "step": 189
    },
    {
      "epoch": 0.0245978573971583,
      "grad_norm": 3.6605853981413317,
      "learning_rate": 3.072445019404916e-07,
      "loss": 4.0422,
      "step": 190
    },
    {
      "epoch": 0.0245978573971583,
      "eval_PRM Accuracy": 0.4056603773584906,
      "eval_PRM F1": 0.4424778761061947,
      "eval_PRM F1 AUC": 0.5419067574646412,
      "eval_PRM F1 Neg": 0.36363636363636365,
      "eval_PRM NPV": 0.23684210526315788,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.30120481927710846,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8857421875,
      "eval_runtime": 15.5331,
      "eval_samples_per_second": 2.833,
      "eval_steps_per_second": 0.193,
      "step": 190
    },
    {
      "epoch": 0.024727319804511764,
      "grad_norm": 2.199660582380855,
      "learning_rate": 3.088615782664942e-07,
      "loss": 5.0649,
      "step": 191
    },
    {
      "epoch": 0.02485678221186523,
      "grad_norm": 2.023418667320133,
      "learning_rate": 3.104786545924968e-07,
      "loss": 3.8857,
      "step": 192
    },
    {
      "epoch": 0.024986244619218695,
      "grad_norm": 1.560125306308657,
      "learning_rate": 3.1209573091849936e-07,
      "loss": 3.4775,
      "step": 193
    },
    {
      "epoch": 0.02511570702657216,
      "grad_norm": 2.1055469781306724,
      "learning_rate": 3.13712807244502e-07,
      "loss": 4.6436,
      "step": 194
    },
    {
      "epoch": 0.025245169433925622,
      "grad_norm": 2.1623865960455886,
      "learning_rate": 3.1532988357050454e-07,
      "loss": 4.2874,
      "step": 195
    },
    {
      "epoch": 0.025245169433925622,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8839666247367859,
      "eval_runtime": 14.4616,
      "eval_samples_per_second": 3.043,
      "eval_steps_per_second": 0.207,
      "step": 195
    },
    {
      "epoch": 0.02537463184127909,
      "grad_norm": 2.0618961739343296,
      "learning_rate": 3.169469598965071e-07,
      "loss": 4.7808,
      "step": 196
    },
    {
      "epoch": 0.025504094248632554,
      "grad_norm": 2.5836308429390726,
      "learning_rate": 3.185640362225097e-07,
      "loss": 4.6035,
      "step": 197
    },
    {
      "epoch": 0.025633556655986017,
      "grad_norm": 1.9751794727575602,
      "learning_rate": 3.2018111254851234e-07,
      "loss": 3.7432,
      "step": 198
    },
    {
      "epoch": 0.02576301906333948,
      "grad_norm": 1.4354870173720242,
      "learning_rate": 3.217981888745149e-07,
      "loss": 3.8357,
      "step": 199
    },
    {
      "epoch": 0.02589248147069295,
      "grad_norm": 3.6734925154622338,
      "learning_rate": 3.234152652005175e-07,
      "loss": 6.9282,
      "step": 200
    },
    {
      "epoch": 0.02589248147069295,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8876065611839294,
      "eval_runtime": 14.5467,
      "eval_samples_per_second": 3.025,
      "eval_steps_per_second": 0.206,
      "step": 200
    },
    {
      "epoch": 0.026021943878046412,
      "grad_norm": 2.473074742494256,
      "learning_rate": 3.250323415265201e-07,
      "loss": 5.3984,
      "step": 201
    },
    {
      "epoch": 0.026151406285399876,
      "grad_norm": 2.5287299002672285,
      "learning_rate": 3.2664941785252265e-07,
      "loss": 5.5527,
      "step": 202
    },
    {
      "epoch": 0.026280868692753343,
      "grad_norm": 2.456362736987199,
      "learning_rate": 3.2826649417852527e-07,
      "loss": 4.6409,
      "step": 203
    },
    {
      "epoch": 0.026410331100106807,
      "grad_norm": 1.9215058395998181,
      "learning_rate": 3.298835705045279e-07,
      "loss": 4.1421,
      "step": 204
    },
    {
      "epoch": 0.02653979350746027,
      "grad_norm": 3.9408260148033922,
      "learning_rate": 3.315006468305304e-07,
      "loss": 7.333,
      "step": 205
    },
    {
      "epoch": 0.02653979350746027,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.884765625,
      "eval_runtime": 13.8341,
      "eval_samples_per_second": 3.181,
      "eval_steps_per_second": 0.217,
      "step": 205
    },
    {
      "epoch": 0.026669255914813735,
      "grad_norm": 3.5973768278716283,
      "learning_rate": 3.33117723156533e-07,
      "loss": 6.6558,
      "step": 206
    },
    {
      "epoch": 0.026798718322167202,
      "grad_norm": 2.7321593067095344,
      "learning_rate": 3.3473479948253564e-07,
      "loss": 4.6677,
      "step": 207
    },
    {
      "epoch": 0.026928180729520666,
      "grad_norm": 2.745785213443762,
      "learning_rate": 3.3635187580853815e-07,
      "loss": 4.5298,
      "step": 208
    },
    {
      "epoch": 0.02705764313687413,
      "grad_norm": 2.362963418655679,
      "learning_rate": 3.3796895213454077e-07,
      "loss": 4.4668,
      "step": 209
    },
    {
      "epoch": 0.027187105544227597,
      "grad_norm": 2.170885869291569,
      "learning_rate": 3.395860284605434e-07,
      "loss": 4.5439,
      "step": 210
    },
    {
      "epoch": 0.027187105544227597,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8869850635528564,
      "eval_runtime": 16.0254,
      "eval_samples_per_second": 2.746,
      "eval_steps_per_second": 0.187,
      "step": 210
    },
    {
      "epoch": 0.02731656795158106,
      "grad_norm": 2.8036859242240992,
      "learning_rate": 3.412031047865459e-07,
      "loss": 5.0664,
      "step": 211
    },
    {
      "epoch": 0.027446030358934524,
      "grad_norm": 1.9738146255572215,
      "learning_rate": 3.428201811125485e-07,
      "loss": 4.4531,
      "step": 212
    },
    {
      "epoch": 0.027575492766287988,
      "grad_norm": 3.7762534844913676,
      "learning_rate": 3.4443725743855113e-07,
      "loss": 6.1953,
      "step": 213
    },
    {
      "epoch": 0.027704955173641455,
      "grad_norm": 1.3922138159878588,
      "learning_rate": 3.4605433376455375e-07,
      "loss": 3.5786,
      "step": 214
    },
    {
      "epoch": 0.02783441758099492,
      "grad_norm": 1.1881268842261485,
      "learning_rate": 3.4767141009055626e-07,
      "loss": 2.936,
      "step": 215
    },
    {
      "epoch": 0.02783441758099492,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8840553760528564,
      "eval_runtime": 14.9814,
      "eval_samples_per_second": 2.937,
      "eval_steps_per_second": 0.2,
      "step": 215
    },
    {
      "epoch": 0.027963879988348383,
      "grad_norm": 2.518946321063278,
      "learning_rate": 3.492884864165589e-07,
      "loss": 5.0696,
      "step": 216
    },
    {
      "epoch": 0.028093342395701847,
      "grad_norm": 2.152032693026809,
      "learning_rate": 3.509055627425615e-07,
      "loss": 5.2261,
      "step": 217
    },
    {
      "epoch": 0.028222804803055314,
      "grad_norm": 2.9739441597144705,
      "learning_rate": 3.5252263906856406e-07,
      "loss": 6.2876,
      "step": 218
    },
    {
      "epoch": 0.028352267210408778,
      "grad_norm": 1.9749725553254946,
      "learning_rate": 3.541397153945666e-07,
      "loss": 4.2852,
      "step": 219
    },
    {
      "epoch": 0.02848172961776224,
      "grad_norm": 2.504483731096505,
      "learning_rate": 3.5575679172056924e-07,
      "loss": 4.562,
      "step": 220
    },
    {
      "epoch": 0.02848172961776224,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.884765625,
      "eval_runtime": 14.3789,
      "eval_samples_per_second": 3.06,
      "eval_steps_per_second": 0.209,
      "step": 220
    },
    {
      "epoch": 0.02861119202511571,
      "grad_norm": 3.237303782437478,
      "learning_rate": 3.573738680465718e-07,
      "loss": 6.4438,
      "step": 221
    },
    {
      "epoch": 0.028740654432469172,
      "grad_norm": 2.9563482106484487,
      "learning_rate": 3.5899094437257443e-07,
      "loss": 5.6177,
      "step": 222
    },
    {
      "epoch": 0.028870116839822636,
      "grad_norm": 2.169875290961729,
      "learning_rate": 3.6060802069857704e-07,
      "loss": 4.7012,
      "step": 223
    },
    {
      "epoch": 0.0289995792471761,
      "grad_norm": 3.097390281971058,
      "learning_rate": 3.6222509702457956e-07,
      "loss": 4.9597,
      "step": 224
    },
    {
      "epoch": 0.029129041654529567,
      "grad_norm": 1.1916093538628139,
      "learning_rate": 3.638421733505822e-07,
      "loss": 3.1731,
      "step": 225
    },
    {
      "epoch": 0.029129041654529567,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8854758739471436,
      "eval_runtime": 15.49,
      "eval_samples_per_second": 2.841,
      "eval_steps_per_second": 0.194,
      "step": 225
    },
    {
      "epoch": 0.02925850406188303,
      "grad_norm": 2.5146257457979346,
      "learning_rate": 3.654592496765848e-07,
      "loss": 4.9551,
      "step": 226
    },
    {
      "epoch": 0.029387966469236495,
      "grad_norm": 2.6612886213539113,
      "learning_rate": 3.670763260025874e-07,
      "loss": 4.5801,
      "step": 227
    },
    {
      "epoch": 0.02951742887658996,
      "grad_norm": 2.350864586600711,
      "learning_rate": 3.686934023285899e-07,
      "loss": 4.4111,
      "step": 228
    },
    {
      "epoch": 0.029646891283943426,
      "grad_norm": 2.9354493578714655,
      "learning_rate": 3.7031047865459254e-07,
      "loss": 4.6353,
      "step": 229
    },
    {
      "epoch": 0.02977635369129689,
      "grad_norm": 2.10405268997446,
      "learning_rate": 3.7192755498059516e-07,
      "loss": 3.8438,
      "step": 230
    },
    {
      "epoch": 0.02977635369129689,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8859197497367859,
      "eval_runtime": 14.6915,
      "eval_samples_per_second": 2.995,
      "eval_steps_per_second": 0.204,
      "step": 230
    },
    {
      "epoch": 0.029905816098650354,
      "grad_norm": 2.796776889785926,
      "learning_rate": 3.7354463130659767e-07,
      "loss": 5.1143,
      "step": 231
    },
    {
      "epoch": 0.03003527850600382,
      "grad_norm": 2.540196823573686,
      "learning_rate": 3.751617076326003e-07,
      "loss": 4.4585,
      "step": 232
    },
    {
      "epoch": 0.030164740913357285,
      "grad_norm": 1.9734948597315027,
      "learning_rate": 3.767787839586029e-07,
      "loss": 4.624,
      "step": 233
    },
    {
      "epoch": 0.03029420332071075,
      "grad_norm": 1.413122722618321,
      "learning_rate": 3.783958602846054e-07,
      "loss": 3.5017,
      "step": 234
    },
    {
      "epoch": 0.030423665728064212,
      "grad_norm": 1.6196061176180687,
      "learning_rate": 3.8001293661060803e-07,
      "loss": 3.5574,
      "step": 235
    },
    {
      "epoch": 0.030423665728064212,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8846768736839294,
      "eval_runtime": 14.4343,
      "eval_samples_per_second": 3.048,
      "eval_steps_per_second": 0.208,
      "step": 235
    },
    {
      "epoch": 0.03055312813541768,
      "grad_norm": 1.9251633527239032,
      "learning_rate": 3.8163001293661065e-07,
      "loss": 4.2485,
      "step": 236
    },
    {
      "epoch": 0.030682590542771143,
      "grad_norm": 2.86788432030648,
      "learning_rate": 3.832470892626132e-07,
      "loss": 5.1436,
      "step": 237
    },
    {
      "epoch": 0.030812052950124607,
      "grad_norm": 3.4081952293967217,
      "learning_rate": 3.848641655886158e-07,
      "loss": 5.4128,
      "step": 238
    },
    {
      "epoch": 0.03094151535747807,
      "grad_norm": 1.0069334503784966,
      "learning_rate": 3.864812419146184e-07,
      "loss": 2.7949,
      "step": 239
    },
    {
      "epoch": 0.031070977764831538,
      "grad_norm": 2.54531275256124,
      "learning_rate": 3.8809831824062096e-07,
      "loss": 5.3416,
      "step": 240
    },
    {
      "epoch": 0.031070977764831538,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8854758739471436,
      "eval_runtime": 15.2595,
      "eval_samples_per_second": 2.883,
      "eval_steps_per_second": 0.197,
      "step": 240
    },
    {
      "epoch": 0.031200440172185002,
      "grad_norm": 1.8770741090775689,
      "learning_rate": 3.897153945666236e-07,
      "loss": 3.8792,
      "step": 241
    },
    {
      "epoch": 0.031329902579538466,
      "grad_norm": 2.9652913430837136,
      "learning_rate": 3.9133247089262615e-07,
      "loss": 6.0073,
      "step": 242
    },
    {
      "epoch": 0.03145936498689193,
      "grad_norm": 4.106740705941837,
      "learning_rate": 3.9294954721862877e-07,
      "loss": 7.5737,
      "step": 243
    },
    {
      "epoch": 0.03158882739424539,
      "grad_norm": 2.0455585729583854,
      "learning_rate": 3.9456662354463133e-07,
      "loss": 4.6748,
      "step": 244
    },
    {
      "epoch": 0.03171828980159886,
      "grad_norm": 1.8645159203862933,
      "learning_rate": 3.9618369987063395e-07,
      "loss": 3.7271,
      "step": 245
    },
    {
      "epoch": 0.03171828980159886,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8874289989471436,
      "eval_runtime": 15.0412,
      "eval_samples_per_second": 2.925,
      "eval_steps_per_second": 0.199,
      "step": 245
    },
    {
      "epoch": 0.03184775220895233,
      "grad_norm": 2.0926513605013555,
      "learning_rate": 3.9780077619663657e-07,
      "loss": 4.0507,
      "step": 246
    },
    {
      "epoch": 0.03197721461630579,
      "grad_norm": 2.7296823858108135,
      "learning_rate": 3.994178525226391e-07,
      "loss": 5.2227,
      "step": 247
    },
    {
      "epoch": 0.032106677023659255,
      "grad_norm": 3.878143159601418,
      "learning_rate": 4.010349288486417e-07,
      "loss": 6.4014,
      "step": 248
    },
    {
      "epoch": 0.03223613943101272,
      "grad_norm": 2.309639696039142,
      "learning_rate": 4.026520051746443e-07,
      "loss": 5.0869,
      "step": 249
    },
    {
      "epoch": 0.03236560183836618,
      "grad_norm": 1.4112146486784303,
      "learning_rate": 4.042690815006468e-07,
      "loss": 3.1626,
      "step": 250
    },
    {
      "epoch": 0.03236560183836618,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8870738744735718,
      "eval_runtime": 14.9055,
      "eval_samples_per_second": 2.952,
      "eval_steps_per_second": 0.201,
      "step": 250
    },
    {
      "epoch": 0.03249506424571965,
      "grad_norm": 1.8751606104067502,
      "learning_rate": 4.0588615782664944e-07,
      "loss": 4.4971,
      "step": 251
    },
    {
      "epoch": 0.03262452665307312,
      "grad_norm": 2.579439703784705,
      "learning_rate": 4.0750323415265206e-07,
      "loss": 4.5474,
      "step": 252
    },
    {
      "epoch": 0.03275398906042658,
      "grad_norm": 1.9307201135550025,
      "learning_rate": 4.0912031047865457e-07,
      "loss": 3.6321,
      "step": 253
    },
    {
      "epoch": 0.032883451467780045,
      "grad_norm": 2.8048169313003037,
      "learning_rate": 4.107373868046572e-07,
      "loss": 6.1855,
      "step": 254
    },
    {
      "epoch": 0.033012913875133505,
      "grad_norm": 1.9406061976658442,
      "learning_rate": 4.123544631306598e-07,
      "loss": 4.2798,
      "step": 255
    },
    {
      "epoch": 0.033012913875133505,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8869850635528564,
      "eval_runtime": 14.5647,
      "eval_samples_per_second": 3.021,
      "eval_steps_per_second": 0.206,
      "step": 255
    },
    {
      "epoch": 0.03314237628248697,
      "grad_norm": 2.8163829206102613,
      "learning_rate": 4.139715394566624e-07,
      "loss": 4.9077,
      "step": 256
    },
    {
      "epoch": 0.03327183868984044,
      "grad_norm": 1.6884950389012772,
      "learning_rate": 4.1558861578266494e-07,
      "loss": 3.3872,
      "step": 257
    },
    {
      "epoch": 0.0334013010971939,
      "grad_norm": 1.4713171619014873,
      "learning_rate": 4.1720569210866756e-07,
      "loss": 3.6279,
      "step": 258
    },
    {
      "epoch": 0.03353076350454737,
      "grad_norm": 2.398750440671604,
      "learning_rate": 4.1882276843467017e-07,
      "loss": 4.675,
      "step": 259
    },
    {
      "epoch": 0.033660225911900835,
      "grad_norm": 2.325987124008959,
      "learning_rate": 4.2043984476067274e-07,
      "loss": 4.5254,
      "step": 260
    },
    {
      "epoch": 0.033660225911900835,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8865411877632141,
      "eval_runtime": 14.997,
      "eval_samples_per_second": 2.934,
      "eval_steps_per_second": 0.2,
      "step": 260
    },
    {
      "epoch": 0.033789688319254295,
      "grad_norm": 2.8100993157322263,
      "learning_rate": 4.220569210866753e-07,
      "loss": 4.8752,
      "step": 261
    },
    {
      "epoch": 0.03391915072660776,
      "grad_norm": 3.648392493445176,
      "learning_rate": 4.236739974126779e-07,
      "loss": 6.2065,
      "step": 262
    },
    {
      "epoch": 0.03404861313396123,
      "grad_norm": 2.6805319521377213,
      "learning_rate": 4.252910737386805e-07,
      "loss": 4.9004,
      "step": 263
    },
    {
      "epoch": 0.03417807554131469,
      "grad_norm": 2.3796110213603514,
      "learning_rate": 4.269081500646831e-07,
      "loss": 4.3872,
      "step": 264
    },
    {
      "epoch": 0.03430753794866816,
      "grad_norm": 2.40515345650876,
      "learning_rate": 4.2852522639068567e-07,
      "loss": 5.4297,
      "step": 265
    },
    {
      "epoch": 0.03430753794866816,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8852983117103577,
      "eval_runtime": 15.096,
      "eval_samples_per_second": 2.915,
      "eval_steps_per_second": 0.199,
      "step": 265
    },
    {
      "epoch": 0.03443700035602162,
      "grad_norm": 2.9258342379152245,
      "learning_rate": 4.3014230271668823e-07,
      "loss": 5.5503,
      "step": 266
    },
    {
      "epoch": 0.034566462763375085,
      "grad_norm": 3.61105782982967,
      "learning_rate": 4.3175937904269085e-07,
      "loss": 6.4419,
      "step": 267
    },
    {
      "epoch": 0.03469592517072855,
      "grad_norm": 2.78146236444209,
      "learning_rate": 4.3337645536869347e-07,
      "loss": 4.7144,
      "step": 268
    },
    {
      "epoch": 0.03482538757808201,
      "grad_norm": 2.9687116894361303,
      "learning_rate": 4.34993531694696e-07,
      "loss": 4.6682,
      "step": 269
    },
    {
      "epoch": 0.03495484998543548,
      "grad_norm": 2.4493788817536717,
      "learning_rate": 4.366106080206986e-07,
      "loss": 5.0649,
      "step": 270
    },
    {
      "epoch": 0.03495484998543548,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.884765625,
      "eval_runtime": 15.2162,
      "eval_samples_per_second": 2.892,
      "eval_steps_per_second": 0.197,
      "step": 270
    },
    {
      "epoch": 0.03508431239278895,
      "grad_norm": 3.6729121975196892,
      "learning_rate": 4.382276843467012e-07,
      "loss": 5.1895,
      "step": 271
    },
    {
      "epoch": 0.03521377480014241,
      "grad_norm": 2.2009377117963362,
      "learning_rate": 4.3984476067270383e-07,
      "loss": 4.1357,
      "step": 272
    },
    {
      "epoch": 0.035343237207495874,
      "grad_norm": 2.8980251483316244,
      "learning_rate": 4.4146183699870635e-07,
      "loss": 4.7271,
      "step": 273
    },
    {
      "epoch": 0.03547269961484934,
      "grad_norm": 2.649286538793547,
      "learning_rate": 4.4307891332470896e-07,
      "loss": 5.3291,
      "step": 274
    },
    {
      "epoch": 0.0356021620222028,
      "grad_norm": 2.407608486644509,
      "learning_rate": 4.446959896507116e-07,
      "loss": 4.6665,
      "step": 275
    },
    {
      "epoch": 0.0356021620222028,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8879616260528564,
      "eval_runtime": 15.3019,
      "eval_samples_per_second": 2.875,
      "eval_steps_per_second": 0.196,
      "step": 275
    },
    {
      "epoch": 0.03573162442955627,
      "grad_norm": 1.8412114012486007,
      "learning_rate": 4.463130659767141e-07,
      "loss": 3.5312,
      "step": 276
    },
    {
      "epoch": 0.03586108683690973,
      "grad_norm": 1.8829031229272333,
      "learning_rate": 4.479301423027167e-07,
      "loss": 4.104,
      "step": 277
    },
    {
      "epoch": 0.0359905492442632,
      "grad_norm": 3.318217480470199,
      "learning_rate": 4.4954721862871933e-07,
      "loss": 6.1147,
      "step": 278
    },
    {
      "epoch": 0.036120011651616664,
      "grad_norm": 3.349631812469815,
      "learning_rate": 4.511642949547219e-07,
      "loss": 5.5811,
      "step": 279
    },
    {
      "epoch": 0.036249474058970124,
      "grad_norm": 2.434116348743928,
      "learning_rate": 4.5278137128072446e-07,
      "loss": 4.1511,
      "step": 280
    },
    {
      "epoch": 0.036249474058970124,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.888671875,
      "eval_runtime": 13.3428,
      "eval_samples_per_second": 3.298,
      "eval_steps_per_second": 0.225,
      "step": 280
    },
    {
      "epoch": 0.03637893646632359,
      "grad_norm": 1.516175337782875,
      "learning_rate": 4.543984476067271e-07,
      "loss": 3.6128,
      "step": 281
    },
    {
      "epoch": 0.03650839887367706,
      "grad_norm": 2.622182149721442,
      "learning_rate": 4.5601552393272964e-07,
      "loss": 5.5483,
      "step": 282
    },
    {
      "epoch": 0.03663786128103052,
      "grad_norm": 2.0649261496070586,
      "learning_rate": 4.5763260025873226e-07,
      "loss": 4.5525,
      "step": 283
    },
    {
      "epoch": 0.036767323688383986,
      "grad_norm": 2.468057106361623,
      "learning_rate": 4.592496765847348e-07,
      "loss": 4.4458,
      "step": 284
    },
    {
      "epoch": 0.036896786095737454,
      "grad_norm": 2.4041084976851974,
      "learning_rate": 4.6086675291073744e-07,
      "loss": 4.571,
      "step": 285
    },
    {
      "epoch": 0.036896786095737454,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8911576867103577,
      "eval_runtime": 15.0977,
      "eval_samples_per_second": 2.914,
      "eval_steps_per_second": 0.199,
      "step": 285
    },
    {
      "epoch": 0.037026248503090914,
      "grad_norm": 2.421453764910423,
      "learning_rate": 4.6248382923674e-07,
      "loss": 4.3494,
      "step": 286
    },
    {
      "epoch": 0.03715571091044438,
      "grad_norm": 2.938594637147776,
      "learning_rate": 4.641009055627426e-07,
      "loss": 5.8699,
      "step": 287
    },
    {
      "epoch": 0.03728517331779784,
      "grad_norm": 3.3245404572673185,
      "learning_rate": 4.657179818887452e-07,
      "loss": 5.748,
      "step": 288
    },
    {
      "epoch": 0.03741463572515131,
      "grad_norm": 1.660595946323085,
      "learning_rate": 4.6733505821474775e-07,
      "loss": 3.3091,
      "step": 289
    },
    {
      "epoch": 0.037544098132504776,
      "grad_norm": 3.419572001471463,
      "learning_rate": 4.6895213454075037e-07,
      "loss": 6.2827,
      "step": 290
    },
    {
      "epoch": 0.037544098132504776,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8868963122367859,
      "eval_runtime": 14.2251,
      "eval_samples_per_second": 3.093,
      "eval_steps_per_second": 0.211,
      "step": 290
    },
    {
      "epoch": 0.037673560539858236,
      "grad_norm": 3.1209620215384737,
      "learning_rate": 4.70569210866753e-07,
      "loss": 5.6821,
      "step": 291
    },
    {
      "epoch": 0.037803022947211704,
      "grad_norm": 2.728390871733252,
      "learning_rate": 4.721862871927555e-07,
      "loss": 5.4678,
      "step": 292
    },
    {
      "epoch": 0.03793248535456517,
      "grad_norm": 1.1858990026523804,
      "learning_rate": 4.738033635187581e-07,
      "loss": 3.1748,
      "step": 293
    },
    {
      "epoch": 0.03806194776191863,
      "grad_norm": 1.0343287607700118,
      "learning_rate": 4.7542043984476074e-07,
      "loss": 2.7825,
      "step": 294
    },
    {
      "epoch": 0.0381914101692721,
      "grad_norm": 2.2100956196876402,
      "learning_rate": 4.770375161707633e-07,
      "loss": 4.1787,
      "step": 295
    },
    {
      "epoch": 0.0381914101692721,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8861860632896423,
      "eval_runtime": 14.5204,
      "eval_samples_per_second": 3.03,
      "eval_steps_per_second": 0.207,
      "step": 295
    },
    {
      "epoch": 0.038320872576625566,
      "grad_norm": 1.7395521864032368,
      "learning_rate": 4.786545924967659e-07,
      "loss": 3.9448,
      "step": 296
    },
    {
      "epoch": 0.038450334983979026,
      "grad_norm": 2.116504616762591,
      "learning_rate": 4.802716688227685e-07,
      "loss": 3.8701,
      "step": 297
    },
    {
      "epoch": 0.03857979739133249,
      "grad_norm": 1.980522519321597,
      "learning_rate": 4.81888745148771e-07,
      "loss": 4.3459,
      "step": 298
    },
    {
      "epoch": 0.038709259798685954,
      "grad_norm": 2.306067251236745,
      "learning_rate": 4.835058214747737e-07,
      "loss": 3.9229,
      "step": 299
    },
    {
      "epoch": 0.03883872220603942,
      "grad_norm": 2.52751944911705,
      "learning_rate": 4.851228978007762e-07,
      "loss": 5.0142,
      "step": 300
    },
    {
      "epoch": 0.03883872220603942,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8876065611839294,
      "eval_runtime": 14.5251,
      "eval_samples_per_second": 3.029,
      "eval_steps_per_second": 0.207,
      "step": 300
    },
    {
      "epoch": 0.03896818461339289,
      "grad_norm": 1.1084112549934337,
      "learning_rate": 4.867399741267789e-07,
      "loss": 3.595,
      "step": 301
    },
    {
      "epoch": 0.03909764702074635,
      "grad_norm": 2.555051046122174,
      "learning_rate": 4.883570504527814e-07,
      "loss": 5.9507,
      "step": 302
    },
    {
      "epoch": 0.039227109428099816,
      "grad_norm": 2.3812197403864244,
      "learning_rate": 4.89974126778784e-07,
      "loss": 4.7889,
      "step": 303
    },
    {
      "epoch": 0.03935657183545328,
      "grad_norm": 2.3063834650836927,
      "learning_rate": 4.915912031047866e-07,
      "loss": 4.5542,
      "step": 304
    },
    {
      "epoch": 0.03948603424280674,
      "grad_norm": 1.5953157628698709,
      "learning_rate": 4.932082794307892e-07,
      "loss": 3.4177,
      "step": 305
    },
    {
      "epoch": 0.03948603424280674,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8868963122367859,
      "eval_runtime": 15.031,
      "eval_samples_per_second": 2.927,
      "eval_steps_per_second": 0.2,
      "step": 305
    },
    {
      "epoch": 0.03961549665016021,
      "grad_norm": 2.52890823339106,
      "learning_rate": 4.948253557567917e-07,
      "loss": 4.3569,
      "step": 306
    },
    {
      "epoch": 0.03974495905751368,
      "grad_norm": 0.7325735914213573,
      "learning_rate": 4.964424320827944e-07,
      "loss": 2.3125,
      "step": 307
    },
    {
      "epoch": 0.03987442146486714,
      "grad_norm": 2.310293392821235,
      "learning_rate": 4.980595084087969e-07,
      "loss": 4.1499,
      "step": 308
    },
    {
      "epoch": 0.040003883872220605,
      "grad_norm": 1.402104868982115,
      "learning_rate": 4.996765847347995e-07,
      "loss": 3.123,
      "step": 309
    },
    {
      "epoch": 0.040133346279574066,
      "grad_norm": 1.894080574611163,
      "learning_rate": 5.012936610608021e-07,
      "loss": 4.0978,
      "step": 310
    },
    {
      "epoch": 0.040133346279574066,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8870738744735718,
      "eval_runtime": 14.8262,
      "eval_samples_per_second": 2.968,
      "eval_steps_per_second": 0.202,
      "step": 310
    },
    {
      "epoch": 0.04026280868692753,
      "grad_norm": 1.357102970182327,
      "learning_rate": 5.029107373868047e-07,
      "loss": 3.1089,
      "step": 311
    },
    {
      "epoch": 0.040392271094281,
      "grad_norm": 3.497980890858202,
      "learning_rate": 5.045278137128073e-07,
      "loss": 5.3418,
      "step": 312
    },
    {
      "epoch": 0.04052173350163446,
      "grad_norm": 1.0639089337277525,
      "learning_rate": 5.061448900388099e-07,
      "loss": 2.8896,
      "step": 313
    },
    {
      "epoch": 0.04065119590898793,
      "grad_norm": 3.300938656190228,
      "learning_rate": 5.077619663648125e-07,
      "loss": 5.9189,
      "step": 314
    },
    {
      "epoch": 0.040780658316341395,
      "grad_norm": 1.2413862644441418,
      "learning_rate": 5.09379042690815e-07,
      "loss": 3.4585,
      "step": 315
    },
    {
      "epoch": 0.040780658316341395,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8891157507896423,
      "eval_runtime": 15.4693,
      "eval_samples_per_second": 2.844,
      "eval_steps_per_second": 0.194,
      "step": 315
    },
    {
      "epoch": 0.040910120723694855,
      "grad_norm": 4.332081774178866,
      "learning_rate": 5.109961190168177e-07,
      "loss": 7.1943,
      "step": 316
    },
    {
      "epoch": 0.04103958313104832,
      "grad_norm": 3.573961749380484,
      "learning_rate": 5.126131953428203e-07,
      "loss": 5.5,
      "step": 317
    },
    {
      "epoch": 0.04116904553840179,
      "grad_norm": 3.266568775666242,
      "learning_rate": 5.142302716688228e-07,
      "loss": 5.0718,
      "step": 318
    },
    {
      "epoch": 0.04129850794575525,
      "grad_norm": 2.150736768179415,
      "learning_rate": 5.158473479948254e-07,
      "loss": 3.7144,
      "step": 319
    },
    {
      "epoch": 0.04142797035310872,
      "grad_norm": 2.9888130898725063,
      "learning_rate": 5.174644243208281e-07,
      "loss": 4.749,
      "step": 320
    },
    {
      "epoch": 0.04142797035310872,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8865411877632141,
      "eval_runtime": 16.5528,
      "eval_samples_per_second": 2.658,
      "eval_steps_per_second": 0.181,
      "step": 320
    },
    {
      "epoch": 0.04155743276046218,
      "grad_norm": 3.4246676283314756,
      "learning_rate": 5.190815006468305e-07,
      "loss": 5.1213,
      "step": 321
    },
    {
      "epoch": 0.041686895167815645,
      "grad_norm": 2.2618818130722693,
      "learning_rate": 5.206985769728332e-07,
      "loss": 4.0854,
      "step": 322
    },
    {
      "epoch": 0.04181635757516911,
      "grad_norm": 4.625362737261295,
      "learning_rate": 5.223156532988358e-07,
      "loss": 7.9058,
      "step": 323
    },
    {
      "epoch": 0.04194581998252257,
      "grad_norm": 1.9023520509429879,
      "learning_rate": 5.239327296248383e-07,
      "loss": 3.9753,
      "step": 324
    },
    {
      "epoch": 0.04207528238987604,
      "grad_norm": 2.43325767545436,
      "learning_rate": 5.255498059508409e-07,
      "loss": 4.7739,
      "step": 325
    },
    {
      "epoch": 0.04207528238987604,
      "eval_PRM Accuracy": 0.4056603773584906,
      "eval_PRM F1": 0.4424778761061947,
      "eval_PRM F1 AUC": 0.5419067574646412,
      "eval_PRM F1 Neg": 0.36363636363636365,
      "eval_PRM NPV": 0.23684210526315788,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.30120481927710846,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8883167505264282,
      "eval_runtime": 15.0397,
      "eval_samples_per_second": 2.926,
      "eval_steps_per_second": 0.199,
      "step": 325
    },
    {
      "epoch": 0.04220474479722951,
      "grad_norm": 2.207312548983323,
      "learning_rate": 5.271668822768436e-07,
      "loss": 4.5195,
      "step": 326
    },
    {
      "epoch": 0.04233420720458297,
      "grad_norm": 2.4928797281193606,
      "learning_rate": 5.28783958602846e-07,
      "loss": 4.6545,
      "step": 327
    },
    {
      "epoch": 0.042463669611936435,
      "grad_norm": 2.489344579365318,
      "learning_rate": 5.304010349288487e-07,
      "loss": 4.6489,
      "step": 328
    },
    {
      "epoch": 0.0425931320192899,
      "grad_norm": 4.57429433288173,
      "learning_rate": 5.320181112548512e-07,
      "loss": 7.3164,
      "step": 329
    },
    {
      "epoch": 0.04272259442664336,
      "grad_norm": 3.255168861943587,
      "learning_rate": 5.336351875808539e-07,
      "loss": 6.5586,
      "step": 330
    },
    {
      "epoch": 0.04272259442664336,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8872514367103577,
      "eval_runtime": 14.6803,
      "eval_samples_per_second": 2.997,
      "eval_steps_per_second": 0.204,
      "step": 330
    },
    {
      "epoch": 0.04285205683399683,
      "grad_norm": 3.9597125898952092,
      "learning_rate": 5.352522639068565e-07,
      "loss": 7.1958,
      "step": 331
    },
    {
      "epoch": 0.04298151924135029,
      "grad_norm": 2.5225364315179766,
      "learning_rate": 5.36869340232859e-07,
      "loss": 4.2236,
      "step": 332
    },
    {
      "epoch": 0.04311098164870376,
      "grad_norm": 2.459820107417454,
      "learning_rate": 5.384864165588616e-07,
      "loss": 4.5474,
      "step": 333
    },
    {
      "epoch": 0.043240444056057224,
      "grad_norm": 3.2767650864133366,
      "learning_rate": 5.401034928848642e-07,
      "loss": 5.583,
      "step": 334
    },
    {
      "epoch": 0.043369906463410685,
      "grad_norm": 2.7676992302367696,
      "learning_rate": 5.417205692108668e-07,
      "loss": 4.689,
      "step": 335
    },
    {
      "epoch": 0.043369906463410685,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.88671875,
      "eval_runtime": 14.3717,
      "eval_samples_per_second": 3.062,
      "eval_steps_per_second": 0.209,
      "step": 335
    },
    {
      "epoch": 0.04349936887076415,
      "grad_norm": 2.051096966942122,
      "learning_rate": 5.433376455368694e-07,
      "loss": 3.8599,
      "step": 336
    },
    {
      "epoch": 0.04362883127811762,
      "grad_norm": 3.844604070905217,
      "learning_rate": 5.44954721862872e-07,
      "loss": 6.0503,
      "step": 337
    },
    {
      "epoch": 0.04375829368547108,
      "grad_norm": 4.316281408572996,
      "learning_rate": 5.465717981888745e-07,
      "loss": 8.3369,
      "step": 338
    },
    {
      "epoch": 0.04388775609282455,
      "grad_norm": 3.821369370757319,
      "learning_rate": 5.481888745148772e-07,
      "loss": 5.4961,
      "step": 339
    },
    {
      "epoch": 0.044017218500178014,
      "grad_norm": 4.54905571010069,
      "learning_rate": 5.498059508408797e-07,
      "loss": 6.9441,
      "step": 340
    },
    {
      "epoch": 0.044017218500178014,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8871626257896423,
      "eval_runtime": 14.1195,
      "eval_samples_per_second": 3.116,
      "eval_steps_per_second": 0.212,
      "step": 340
    },
    {
      "epoch": 0.044146680907531474,
      "grad_norm": 4.952659576810068,
      "learning_rate": 5.514230271668823e-07,
      "loss": 8.1987,
      "step": 341
    },
    {
      "epoch": 0.04427614331488494,
      "grad_norm": 3.302548186599011,
      "learning_rate": 5.530401034928849e-07,
      "loss": 5.1035,
      "step": 342
    },
    {
      "epoch": 0.0444056057222384,
      "grad_norm": 4.000622890748476,
      "learning_rate": 5.546571798188876e-07,
      "loss": 6.5605,
      "step": 343
    },
    {
      "epoch": 0.04453506812959187,
      "grad_norm": 5.127080244551329,
      "learning_rate": 5.5627425614489e-07,
      "loss": 8.0862,
      "step": 344
    },
    {
      "epoch": 0.044664530536945336,
      "grad_norm": 1.7447122779201634,
      "learning_rate": 5.578913324708927e-07,
      "loss": 4.0793,
      "step": 345
    },
    {
      "epoch": 0.044664530536945336,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8879616260528564,
      "eval_runtime": 14.9009,
      "eval_samples_per_second": 2.953,
      "eval_steps_per_second": 0.201,
      "step": 345
    },
    {
      "epoch": 0.0447939929442988,
      "grad_norm": 4.038935042030998,
      "learning_rate": 5.595084087968953e-07,
      "loss": 6.3291,
      "step": 346
    },
    {
      "epoch": 0.044923455351652264,
      "grad_norm": 3.2818578613931506,
      "learning_rate": 5.611254851228978e-07,
      "loss": 5.6594,
      "step": 347
    },
    {
      "epoch": 0.04505291775900573,
      "grad_norm": 2.8882454187139217,
      "learning_rate": 5.627425614489004e-07,
      "loss": 4.877,
      "step": 348
    },
    {
      "epoch": 0.04518238016635919,
      "grad_norm": 2.0416649463880256,
      "learning_rate": 5.643596377749031e-07,
      "loss": 3.7373,
      "step": 349
    },
    {
      "epoch": 0.04531184257371266,
      "grad_norm": 1.6936810512434821,
      "learning_rate": 5.659767141009055e-07,
      "loss": 3.5723,
      "step": 350
    },
    {
      "epoch": 0.04531184257371266,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8869850635528564,
      "eval_runtime": 14.1331,
      "eval_samples_per_second": 3.113,
      "eval_steps_per_second": 0.212,
      "step": 350
    },
    {
      "epoch": 0.045441304981066126,
      "grad_norm": 2.0200771091713503,
      "learning_rate": 5.675937904269082e-07,
      "loss": 4.0376,
      "step": 351
    },
    {
      "epoch": 0.045570767388419586,
      "grad_norm": 2.1854125513454976,
      "learning_rate": 5.692108667529108e-07,
      "loss": 4.7383,
      "step": 352
    },
    {
      "epoch": 0.045700229795773054,
      "grad_norm": 3.1451764651800267,
      "learning_rate": 5.708279430789133e-07,
      "loss": 5.8213,
      "step": 353
    },
    {
      "epoch": 0.045829692203126514,
      "grad_norm": 2.1449333513952196,
      "learning_rate": 5.72445019404916e-07,
      "loss": 3.7407,
      "step": 354
    },
    {
      "epoch": 0.04595915461047998,
      "grad_norm": 3.9835954047834705,
      "learning_rate": 5.740620957309186e-07,
      "loss": 6.1538,
      "step": 355
    },
    {
      "epoch": 0.04595915461047998,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8890269994735718,
      "eval_runtime": 14.3354,
      "eval_samples_per_second": 3.069,
      "eval_steps_per_second": 0.209,
      "step": 355
    },
    {
      "epoch": 0.04608861701783345,
      "grad_norm": 2.8708628349479204,
      "learning_rate": 5.756791720569211e-07,
      "loss": 5.3345,
      "step": 356
    },
    {
      "epoch": 0.04621807942518691,
      "grad_norm": 4.249639661394875,
      "learning_rate": 5.772962483829237e-07,
      "loss": 6.1455,
      "step": 357
    },
    {
      "epoch": 0.046347541832540376,
      "grad_norm": 3.325582864238537,
      "learning_rate": 5.789133247089264e-07,
      "loss": 4.8584,
      "step": 358
    },
    {
      "epoch": 0.04647700423989384,
      "grad_norm": 4.466988684148305,
      "learning_rate": 5.805304010349289e-07,
      "loss": 7.502,
      "step": 359
    },
    {
      "epoch": 0.046606466647247304,
      "grad_norm": 2.4849567677341424,
      "learning_rate": 5.821474773609315e-07,
      "loss": 5.0854,
      "step": 360
    },
    {
      "epoch": 0.046606466647247304,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8874289989471436,
      "eval_runtime": 14.55,
      "eval_samples_per_second": 3.024,
      "eval_steps_per_second": 0.206,
      "step": 360
    },
    {
      "epoch": 0.04673592905460077,
      "grad_norm": 3.7224181988715532,
      "learning_rate": 5.837645536869341e-07,
      "loss": 6.3174,
      "step": 361
    },
    {
      "epoch": 0.04686539146195424,
      "grad_norm": 3.234172791958926,
      "learning_rate": 5.853816300129367e-07,
      "loss": 6.0649,
      "step": 362
    },
    {
      "epoch": 0.0469948538693077,
      "grad_norm": 2.588006383504166,
      "learning_rate": 5.869987063389392e-07,
      "loss": 4.3237,
      "step": 363
    },
    {
      "epoch": 0.047124316276661166,
      "grad_norm": 4.537494485701655,
      "learning_rate": 5.886157826649419e-07,
      "loss": 6.4592,
      "step": 364
    },
    {
      "epoch": 0.047253778684014626,
      "grad_norm": 1.8112966664258494,
      "learning_rate": 5.902328589909444e-07,
      "loss": 3.6387,
      "step": 365
    },
    {
      "epoch": 0.047253778684014626,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8860085010528564,
      "eval_runtime": 15.407,
      "eval_samples_per_second": 2.856,
      "eval_steps_per_second": 0.195,
      "step": 365
    },
    {
      "epoch": 0.04738324109136809,
      "grad_norm": 3.5146563433221965,
      "learning_rate": 5.91849935316947e-07,
      "loss": 4.7334,
      "step": 366
    },
    {
      "epoch": 0.04751270349872156,
      "grad_norm": 1.4379114753778197,
      "learning_rate": 5.934670116429496e-07,
      "loss": 3.2285,
      "step": 367
    },
    {
      "epoch": 0.04764216590607502,
      "grad_norm": 1.908983344446691,
      "learning_rate": 5.950840879689522e-07,
      "loss": 4.0811,
      "step": 368
    },
    {
      "epoch": 0.04777162831342849,
      "grad_norm": 2.169763911252574,
      "learning_rate": 5.967011642949547e-07,
      "loss": 4.6023,
      "step": 369
    },
    {
      "epoch": 0.047901090720781955,
      "grad_norm": 2.1575595312092926,
      "learning_rate": 5.983182406209574e-07,
      "loss": 3.9585,
      "step": 370
    },
    {
      "epoch": 0.047901090720781955,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8864524364471436,
      "eval_runtime": 15.3038,
      "eval_samples_per_second": 2.875,
      "eval_steps_per_second": 0.196,
      "step": 370
    },
    {
      "epoch": 0.048030553128135416,
      "grad_norm": 5.318161033192671,
      "learning_rate": 5.999353169469599e-07,
      "loss": 7.8088,
      "step": 371
    },
    {
      "epoch": 0.04816001553548888,
      "grad_norm": 1.9980890310244137,
      "learning_rate": 6.015523932729626e-07,
      "loss": 3.5575,
      "step": 372
    },
    {
      "epoch": 0.04828947794284235,
      "grad_norm": 3.949506127219697,
      "learning_rate": 6.031694695989651e-07,
      "loss": 6.4028,
      "step": 373
    },
    {
      "epoch": 0.04841894035019581,
      "grad_norm": 1.7781458866902489,
      "learning_rate": 6.047865459249677e-07,
      "loss": 3.4214,
      "step": 374
    },
    {
      "epoch": 0.04854840275754928,
      "grad_norm": 2.4135744708988938,
      "learning_rate": 6.064036222509703e-07,
      "loss": 4.8521,
      "step": 375
    },
    {
      "epoch": 0.04854840275754928,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8888494372367859,
      "eval_runtime": 16.1404,
      "eval_samples_per_second": 2.726,
      "eval_steps_per_second": 0.186,
      "step": 375
    },
    {
      "epoch": 0.04867786516490274,
      "grad_norm": 2.427114114429261,
      "learning_rate": 6.080206985769729e-07,
      "loss": 4.2141,
      "step": 376
    },
    {
      "epoch": 0.048807327572256205,
      "grad_norm": 2.206430166153848,
      "learning_rate": 6.096377749029755e-07,
      "loss": 3.7795,
      "step": 377
    },
    {
      "epoch": 0.04893678997960967,
      "grad_norm": 3.3553556761191574,
      "learning_rate": 6.112548512289781e-07,
      "loss": 4.9688,
      "step": 378
    },
    {
      "epoch": 0.04906625238696313,
      "grad_norm": 0.9161142075888727,
      "learning_rate": 6.128719275549807e-07,
      "loss": 3.0227,
      "step": 379
    },
    {
      "epoch": 0.0491957147943166,
      "grad_norm": 3.471482846055819,
      "learning_rate": 6.144890038809832e-07,
      "loss": 5.7178,
      "step": 380
    },
    {
      "epoch": 0.0491957147943166,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8895596861839294,
      "eval_runtime": 14.2623,
      "eval_samples_per_second": 3.085,
      "eval_steps_per_second": 0.21,
      "step": 380
    },
    {
      "epoch": 0.04932517720167007,
      "grad_norm": 1.9819082834705617,
      "learning_rate": 6.161060802069859e-07,
      "loss": 3.8608,
      "step": 381
    },
    {
      "epoch": 0.04945463960902353,
      "grad_norm": 2.903210725442833,
      "learning_rate": 6.177231565329884e-07,
      "loss": 5.5146,
      "step": 382
    },
    {
      "epoch": 0.049584102016376995,
      "grad_norm": 2.6708054607635168,
      "learning_rate": 6.19340232858991e-07,
      "loss": 4.4766,
      "step": 383
    },
    {
      "epoch": 0.04971356442373046,
      "grad_norm": 3.1183717764760837,
      "learning_rate": 6.209573091849936e-07,
      "loss": 5.4492,
      "step": 384
    },
    {
      "epoch": 0.04984302683108392,
      "grad_norm": 3.7400801881411736,
      "learning_rate": 6.225743855109962e-07,
      "loss": 6.1411,
      "step": 385
    },
    {
      "epoch": 0.04984302683108392,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8873401880264282,
      "eval_runtime": 16.4864,
      "eval_samples_per_second": 2.669,
      "eval_steps_per_second": 0.182,
      "step": 385
    },
    {
      "epoch": 0.04997248923843739,
      "grad_norm": 2.571588858217802,
      "learning_rate": 6.241914618369987e-07,
      "loss": 4.6377,
      "step": 386
    },
    {
      "epoch": 0.05010195164579085,
      "grad_norm": 1.905362147390013,
      "learning_rate": 6.258085381630014e-07,
      "loss": 3.9556,
      "step": 387
    },
    {
      "epoch": 0.05023141405314432,
      "grad_norm": 2.703486235188581,
      "learning_rate": 6.27425614489004e-07,
      "loss": 4.6113,
      "step": 388
    },
    {
      "epoch": 0.050360876460497785,
      "grad_norm": 2.0525631550788725,
      "learning_rate": 6.290426908150066e-07,
      "loss": 4.4429,
      "step": 389
    },
    {
      "epoch": 0.050490338867851245,
      "grad_norm": 3.215100360275304,
      "learning_rate": 6.306597671410091e-07,
      "loss": 5.5059,
      "step": 390
    },
    {
      "epoch": 0.050490338867851245,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8899147510528564,
      "eval_runtime": 18.5744,
      "eval_samples_per_second": 2.369,
      "eval_steps_per_second": 0.162,
      "step": 390
    },
    {
      "epoch": 0.05061980127520471,
      "grad_norm": 2.7883839598086984,
      "learning_rate": 6.322768434670116e-07,
      "loss": 5.1589,
      "step": 391
    },
    {
      "epoch": 0.05074926368255818,
      "grad_norm": 2.39681815806041,
      "learning_rate": 6.338939197930142e-07,
      "loss": 3.928,
      "step": 392
    },
    {
      "epoch": 0.05087872608991164,
      "grad_norm": 2.7225392239613098,
      "learning_rate": 6.355109961190169e-07,
      "loss": 4.7026,
      "step": 393
    },
    {
      "epoch": 0.05100818849726511,
      "grad_norm": 4.518284177510381,
      "learning_rate": 6.371280724450194e-07,
      "loss": 6.3408,
      "step": 394
    },
    {
      "epoch": 0.051137650904618574,
      "grad_norm": 2.5612781804229034,
      "learning_rate": 6.387451487710221e-07,
      "loss": 4.2725,
      "step": 395
    },
    {
      "epoch": 0.051137650904618574,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8902698755264282,
      "eval_runtime": 14.162,
      "eval_samples_per_second": 3.107,
      "eval_steps_per_second": 0.212,
      "step": 395
    },
    {
      "epoch": 0.051267113311972035,
      "grad_norm": 2.280496654996542,
      "learning_rate": 6.403622250970247e-07,
      "loss": 4.4756,
      "step": 396
    },
    {
      "epoch": 0.0513965757193255,
      "grad_norm": 3.021073527285268,
      "learning_rate": 6.419793014230271e-07,
      "loss": 5.7964,
      "step": 397
    },
    {
      "epoch": 0.05152603812667896,
      "grad_norm": 2.6373988360828675,
      "learning_rate": 6.435963777490298e-07,
      "loss": 5.1425,
      "step": 398
    },
    {
      "epoch": 0.05165550053403243,
      "grad_norm": 1.84905961264132,
      "learning_rate": 6.452134540750324e-07,
      "loss": 3.5947,
      "step": 399
    },
    {
      "epoch": 0.0517849629413859,
      "grad_norm": 2.277941994831721,
      "learning_rate": 6.46830530401035e-07,
      "loss": 4.4268,
      "step": 400
    },
    {
      "epoch": 0.0517849629413859,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8890269994735718,
      "eval_runtime": 14.6147,
      "eval_samples_per_second": 3.011,
      "eval_steps_per_second": 0.205,
      "step": 400
    },
    {
      "epoch": 0.05191442534873936,
      "grad_norm": 3.3556177810260066,
      "learning_rate": 6.484476067270376e-07,
      "loss": 5.3779,
      "step": 401
    },
    {
      "epoch": 0.052043887756092824,
      "grad_norm": 2.67169301328284,
      "learning_rate": 6.500646830530402e-07,
      "loss": 4.3022,
      "step": 402
    },
    {
      "epoch": 0.05217335016344629,
      "grad_norm": 2.7219829214054916,
      "learning_rate": 6.516817593790426e-07,
      "loss": 4.4153,
      "step": 403
    },
    {
      "epoch": 0.05230281257079975,
      "grad_norm": 2.5490216138266737,
      "learning_rate": 6.532988357050453e-07,
      "loss": 3.7578,
      "step": 404
    },
    {
      "epoch": 0.05243227497815322,
      "grad_norm": 1.788090552474802,
      "learning_rate": 6.549159120310479e-07,
      "loss": 3.8025,
      "step": 405
    },
    {
      "epoch": 0.05243227497815322,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8891157507896423,
      "eval_runtime": 14.8253,
      "eval_samples_per_second": 2.968,
      "eval_steps_per_second": 0.202,
      "step": 405
    },
    {
      "epoch": 0.052561737385506686,
      "grad_norm": 3.703355461870673,
      "learning_rate": 6.565329883570505e-07,
      "loss": 6.5244,
      "step": 406
    },
    {
      "epoch": 0.05269119979286015,
      "grad_norm": 3.380054763866634,
      "learning_rate": 6.581500646830531e-07,
      "loss": 5.3091,
      "step": 407
    },
    {
      "epoch": 0.052820662200213614,
      "grad_norm": 3.84788336684906,
      "learning_rate": 6.597671410090558e-07,
      "loss": 6.6704,
      "step": 408
    },
    {
      "epoch": 0.05295012460756708,
      "grad_norm": 2.375997572908149,
      "learning_rate": 6.613842173350583e-07,
      "loss": 3.1401,
      "step": 409
    },
    {
      "epoch": 0.05307958701492054,
      "grad_norm": 2.2528166969422854,
      "learning_rate": 6.630012936610608e-07,
      "loss": 3.981,
      "step": 410
    },
    {
      "epoch": 0.05307958701492054,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8892045617103577,
      "eval_runtime": 14.5997,
      "eval_samples_per_second": 3.014,
      "eval_steps_per_second": 0.205,
      "step": 410
    },
    {
      "epoch": 0.05320904942227401,
      "grad_norm": 2.7506030359883282,
      "learning_rate": 6.646183699870634e-07,
      "loss": 3.6577,
      "step": 411
    },
    {
      "epoch": 0.05333851182962747,
      "grad_norm": 2.7556404206790663,
      "learning_rate": 6.66235446313066e-07,
      "loss": 4.3684,
      "step": 412
    },
    {
      "epoch": 0.053467974236980936,
      "grad_norm": 2.6646633731112406,
      "learning_rate": 6.678525226390686e-07,
      "loss": 3.9849,
      "step": 413
    },
    {
      "epoch": 0.053597436644334404,
      "grad_norm": 3.212966981020102,
      "learning_rate": 6.694695989650713e-07,
      "loss": 6.1631,
      "step": 414
    },
    {
      "epoch": 0.053726899051687864,
      "grad_norm": 1.6691294126989227,
      "learning_rate": 6.710866752910738e-07,
      "loss": 4.3237,
      "step": 415
    },
    {
      "epoch": 0.053726899051687864,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8884943127632141,
      "eval_runtime": 14.796,
      "eval_samples_per_second": 2.974,
      "eval_steps_per_second": 0.203,
      "step": 415
    },
    {
      "epoch": 0.05385636145904133,
      "grad_norm": 2.3037675071234434,
      "learning_rate": 6.727037516170763e-07,
      "loss": 5.248,
      "step": 416
    },
    {
      "epoch": 0.0539858238663948,
      "grad_norm": 2.7338076925970465,
      "learning_rate": 6.74320827943079e-07,
      "loss": 5.0679,
      "step": 417
    },
    {
      "epoch": 0.05411528627374826,
      "grad_norm": 3.3411195413599852,
      "learning_rate": 6.759379042690815e-07,
      "loss": 5.0293,
      "step": 418
    },
    {
      "epoch": 0.054244748681101726,
      "grad_norm": 2.8988184361633214,
      "learning_rate": 6.775549805950841e-07,
      "loss": 4.6348,
      "step": 419
    },
    {
      "epoch": 0.05437421108845519,
      "grad_norm": 3.258555741636187,
      "learning_rate": 6.791720569210868e-07,
      "loss": 5.2798,
      "step": 420
    },
    {
      "epoch": 0.05437421108845519,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8904474377632141,
      "eval_runtime": 13.7417,
      "eval_samples_per_second": 3.202,
      "eval_steps_per_second": 0.218,
      "step": 420
    },
    {
      "epoch": 0.054503673495808654,
      "grad_norm": 2.845290843273376,
      "learning_rate": 6.807891332470893e-07,
      "loss": 5.0366,
      "step": 421
    },
    {
      "epoch": 0.05463313590316212,
      "grad_norm": 3.229948321320305,
      "learning_rate": 6.824062095730918e-07,
      "loss": 5.7407,
      "step": 422
    },
    {
      "epoch": 0.05476259831051558,
      "grad_norm": 3.986785020505741,
      "learning_rate": 6.840232858990945e-07,
      "loss": 6.0254,
      "step": 423
    },
    {
      "epoch": 0.05489206071786905,
      "grad_norm": 2.4422246084783574,
      "learning_rate": 6.85640362225097e-07,
      "loss": 4.2644,
      "step": 424
    },
    {
      "epoch": 0.055021523125222516,
      "grad_norm": 1.2793166510673597,
      "learning_rate": 6.872574385510997e-07,
      "loss": 2.6863,
      "step": 425
    },
    {
      "epoch": 0.055021523125222516,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8899147510528564,
      "eval_runtime": 15.7681,
      "eval_samples_per_second": 2.79,
      "eval_steps_per_second": 0.19,
      "step": 425
    },
    {
      "epoch": 0.055150985532575976,
      "grad_norm": 2.97927622169673,
      "learning_rate": 6.888745148771023e-07,
      "loss": 5.5723,
      "step": 426
    },
    {
      "epoch": 0.05528044793992944,
      "grad_norm": 3.6203735487325255,
      "learning_rate": 6.904915912031049e-07,
      "loss": 5.0869,
      "step": 427
    },
    {
      "epoch": 0.05540991034728291,
      "grad_norm": 1.9163894983398424,
      "learning_rate": 6.921086675291075e-07,
      "loss": 4.2043,
      "step": 428
    },
    {
      "epoch": 0.05553937275463637,
      "grad_norm": 3.4733572365302283,
      "learning_rate": 6.9372574385511e-07,
      "loss": 5.252,
      "step": 429
    },
    {
      "epoch": 0.05566883516198984,
      "grad_norm": 3.566197558885763,
      "learning_rate": 6.953428201811125e-07,
      "loss": 5.3286,
      "step": 430
    },
    {
      "epoch": 0.05566883516198984,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8918678760528564,
      "eval_runtime": 13.5644,
      "eval_samples_per_second": 3.244,
      "eval_steps_per_second": 0.221,
      "step": 430
    },
    {
      "epoch": 0.055798297569343305,
      "grad_norm": 2.035928802771165,
      "learning_rate": 6.969598965071152e-07,
      "loss": 4.3062,
      "step": 431
    },
    {
      "epoch": 0.055927759976696766,
      "grad_norm": 3.3999902557564505,
      "learning_rate": 6.985769728331178e-07,
      "loss": 5.4189,
      "step": 432
    },
    {
      "epoch": 0.05605722238405023,
      "grad_norm": 2.4809527375688374,
      "learning_rate": 7.001940491591204e-07,
      "loss": 4.335,
      "step": 433
    },
    {
      "epoch": 0.05618668479140369,
      "grad_norm": 1.5447436872047051,
      "learning_rate": 7.01811125485123e-07,
      "loss": 3.5166,
      "step": 434
    },
    {
      "epoch": 0.05631614719875716,
      "grad_norm": 2.1227145286062705,
      "learning_rate": 7.034282018111255e-07,
      "loss": 3.7466,
      "step": 435
    },
    {
      "epoch": 0.05631614719875716,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8898259997367859,
      "eval_runtime": 13.4876,
      "eval_samples_per_second": 3.262,
      "eval_steps_per_second": 0.222,
      "step": 435
    },
    {
      "epoch": 0.05644560960611063,
      "grad_norm": 2.8037983416166057,
      "learning_rate": 7.050452781371281e-07,
      "loss": 4.9492,
      "step": 436
    },
    {
      "epoch": 0.05657507201346409,
      "grad_norm": 2.618688269998642,
      "learning_rate": 7.066623544631307e-07,
      "loss": 4.6328,
      "step": 437
    },
    {
      "epoch": 0.056704534420817555,
      "grad_norm": 2.36505143201812,
      "learning_rate": 7.082794307891333e-07,
      "loss": 4.5103,
      "step": 438
    },
    {
      "epoch": 0.05683399682817102,
      "grad_norm": 2.525390782647627,
      "learning_rate": 7.098965071151359e-07,
      "loss": 5.0796,
      "step": 439
    },
    {
      "epoch": 0.05696345923552448,
      "grad_norm": 2.371315772725317,
      "learning_rate": 7.115135834411385e-07,
      "loss": 3.8354,
      "step": 440
    },
    {
      "epoch": 0.05696345923552448,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8881391882896423,
      "eval_runtime": 14.4353,
      "eval_samples_per_second": 3.048,
      "eval_steps_per_second": 0.208,
      "step": 440
    },
    {
      "epoch": 0.05709292164287795,
      "grad_norm": 2.389207964725016,
      "learning_rate": 7.131306597671412e-07,
      "loss": 4.4639,
      "step": 441
    },
    {
      "epoch": 0.05722238405023142,
      "grad_norm": 2.431164961294971,
      "learning_rate": 7.147477360931436e-07,
      "loss": 4.1638,
      "step": 442
    },
    {
      "epoch": 0.05735184645758488,
      "grad_norm": 1.8199313844536043,
      "learning_rate": 7.163648124191462e-07,
      "loss": 3.2183,
      "step": 443
    },
    {
      "epoch": 0.057481308864938345,
      "grad_norm": 2.624071289645593,
      "learning_rate": 7.179818887451489e-07,
      "loss": 5.0615,
      "step": 444
    },
    {
      "epoch": 0.057610771272291805,
      "grad_norm": 2.921808231280524,
      "learning_rate": 7.195989650711514e-07,
      "loss": 5.3948,
      "step": 445
    },
    {
      "epoch": 0.057610771272291805,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8890269994735718,
      "eval_runtime": 14.2128,
      "eval_samples_per_second": 3.096,
      "eval_steps_per_second": 0.211,
      "step": 445
    },
    {
      "epoch": 0.05774023367964527,
      "grad_norm": 2.4882204607921357,
      "learning_rate": 7.212160413971541e-07,
      "loss": 5.0454,
      "step": 446
    },
    {
      "epoch": 0.05786969608699874,
      "grad_norm": 5.292947670925015,
      "learning_rate": 7.228331177231567e-07,
      "loss": 7.4385,
      "step": 447
    },
    {
      "epoch": 0.0579991584943522,
      "grad_norm": 3.5914756621173463,
      "learning_rate": 7.244501940491591e-07,
      "loss": 5.6191,
      "step": 448
    },
    {
      "epoch": 0.05812862090170567,
      "grad_norm": 2.447688584948203,
      "learning_rate": 7.260672703751617e-07,
      "loss": 4.2495,
      "step": 449
    },
    {
      "epoch": 0.058258083309059135,
      "grad_norm": 2.1663943903015634,
      "learning_rate": 7.276843467011643e-07,
      "loss": 3.2456,
      "step": 450
    },
    {
      "epoch": 0.058258083309059135,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8911576867103577,
      "eval_runtime": 14.5949,
      "eval_samples_per_second": 3.015,
      "eval_steps_per_second": 0.206,
      "step": 450
    },
    {
      "epoch": 0.058387545716412595,
      "grad_norm": 2.8159884845122183,
      "learning_rate": 7.293014230271669e-07,
      "loss": 4.5798,
      "step": 451
    },
    {
      "epoch": 0.05851700812376606,
      "grad_norm": 3.1860772839931486,
      "learning_rate": 7.309184993531696e-07,
      "loss": 5.5576,
      "step": 452
    },
    {
      "epoch": 0.05864647053111953,
      "grad_norm": 3.1715276442448603,
      "learning_rate": 7.325355756791721e-07,
      "loss": 5.3296,
      "step": 453
    },
    {
      "epoch": 0.05877593293847299,
      "grad_norm": 1.8783448074542377,
      "learning_rate": 7.341526520051748e-07,
      "loss": 4.0239,
      "step": 454
    },
    {
      "epoch": 0.05890539534582646,
      "grad_norm": 1.6432187545216004,
      "learning_rate": 7.357697283311773e-07,
      "loss": 3.6694,
      "step": 455
    },
    {
      "epoch": 0.05890539534582646,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8937322497367859,
      "eval_runtime": 15.0472,
      "eval_samples_per_second": 2.924,
      "eval_steps_per_second": 0.199,
      "step": 455
    },
    {
      "epoch": 0.05903485775317992,
      "grad_norm": 3.9878701303834188,
      "learning_rate": 7.373868046571798e-07,
      "loss": 5.7021,
      "step": 456
    },
    {
      "epoch": 0.059164320160533385,
      "grad_norm": 2.795985084096421,
      "learning_rate": 7.390038809831824e-07,
      "loss": 4.4189,
      "step": 457
    },
    {
      "epoch": 0.05929378256788685,
      "grad_norm": 3.6794018234948163,
      "learning_rate": 7.406209573091851e-07,
      "loss": 4.9004,
      "step": 458
    },
    {
      "epoch": 0.05942324497524031,
      "grad_norm": 3.1117472837893887,
      "learning_rate": 7.422380336351876e-07,
      "loss": 5.7109,
      "step": 459
    },
    {
      "epoch": 0.05955270738259378,
      "grad_norm": 3.6666275759650735,
      "learning_rate": 7.438551099611903e-07,
      "loss": 5.7236,
      "step": 460
    },
    {
      "epoch": 0.05955270738259378,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8926668763160706,
      "eval_runtime": 14.9814,
      "eval_samples_per_second": 2.937,
      "eval_steps_per_second": 0.2,
      "step": 460
    },
    {
      "epoch": 0.05968216978994725,
      "grad_norm": 4.109836747246771,
      "learning_rate": 7.454721862871928e-07,
      "loss": 5.9663,
      "step": 461
    },
    {
      "epoch": 0.05981163219730071,
      "grad_norm": 2.9539081126626514,
      "learning_rate": 7.470892626131953e-07,
      "loss": 5.04,
      "step": 462
    },
    {
      "epoch": 0.059941094604654174,
      "grad_norm": 2.108616044894569,
      "learning_rate": 7.48706338939198e-07,
      "loss": 4.1074,
      "step": 463
    },
    {
      "epoch": 0.06007055701200764,
      "grad_norm": 2.3062645651100695,
      "learning_rate": 7.503234152652006e-07,
      "loss": 4.207,
      "step": 464
    },
    {
      "epoch": 0.0602000194193611,
      "grad_norm": 1.970504808382239,
      "learning_rate": 7.519404915912032e-07,
      "loss": 3.3633,
      "step": 465
    },
    {
      "epoch": 0.0602000194193611,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8894708752632141,
      "eval_runtime": 14.5718,
      "eval_samples_per_second": 3.02,
      "eval_steps_per_second": 0.206,
      "step": 465
    },
    {
      "epoch": 0.06032948182671457,
      "grad_norm": 3.7110334716004316,
      "learning_rate": 7.535575679172058e-07,
      "loss": 5.3782,
      "step": 466
    },
    {
      "epoch": 0.06045894423406803,
      "grad_norm": 3.2511972407968432,
      "learning_rate": 7.551746442432084e-07,
      "loss": 5.1436,
      "step": 467
    },
    {
      "epoch": 0.0605884066414215,
      "grad_norm": 4.045181351030494,
      "learning_rate": 7.567917205692108e-07,
      "loss": 5.3154,
      "step": 468
    },
    {
      "epoch": 0.060717869048774964,
      "grad_norm": 2.5219104109115245,
      "learning_rate": 7.584087968952135e-07,
      "loss": 4.4136,
      "step": 469
    },
    {
      "epoch": 0.060847331456128424,
      "grad_norm": 2.655383749644346,
      "learning_rate": 7.600258732212161e-07,
      "loss": 4.3418,
      "step": 470
    },
    {
      "epoch": 0.060847331456128424,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8932883739471436,
      "eval_runtime": 14.3407,
      "eval_samples_per_second": 3.068,
      "eval_steps_per_second": 0.209,
      "step": 470
    },
    {
      "epoch": 0.06097679386348189,
      "grad_norm": 3.162879797053235,
      "learning_rate": 7.616429495472187e-07,
      "loss": 5.5503,
      "step": 471
    },
    {
      "epoch": 0.06110625627083536,
      "grad_norm": 2.4947505686132256,
      "learning_rate": 7.632600258732213e-07,
      "loss": 4.1929,
      "step": 472
    },
    {
      "epoch": 0.06123571867818882,
      "grad_norm": 2.4022537273906552,
      "learning_rate": 7.64877102199224e-07,
      "loss": 4.6172,
      "step": 473
    },
    {
      "epoch": 0.061365181085542286,
      "grad_norm": 2.786311565932309,
      "learning_rate": 7.664941785252264e-07,
      "loss": 4.5874,
      "step": 474
    },
    {
      "epoch": 0.061494643492895754,
      "grad_norm": 2.7403569099140204,
      "learning_rate": 7.68111254851229e-07,
      "loss": 4.144,
      "step": 475
    },
    {
      "epoch": 0.061494643492895754,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8924005627632141,
      "eval_runtime": 15.4681,
      "eval_samples_per_second": 2.845,
      "eval_steps_per_second": 0.194,
      "step": 475
    },
    {
      "epoch": 0.061624105900249214,
      "grad_norm": 4.056895151035421,
      "learning_rate": 7.697283311772316e-07,
      "loss": 6.5996,
      "step": 476
    },
    {
      "epoch": 0.06175356830760268,
      "grad_norm": 3.019615547745032,
      "learning_rate": 7.713454075032342e-07,
      "loss": 4.4626,
      "step": 477
    },
    {
      "epoch": 0.06188303071495614,
      "grad_norm": 1.7076858671943458,
      "learning_rate": 7.729624838292368e-07,
      "loss": 3.6279,
      "step": 478
    },
    {
      "epoch": 0.06201249312230961,
      "grad_norm": 4.584219984959799,
      "learning_rate": 7.745795601552395e-07,
      "loss": 5.9634,
      "step": 479
    },
    {
      "epoch": 0.062141955529663076,
      "grad_norm": 2.588902587312457,
      "learning_rate": 7.761966364812419e-07,
      "loss": 4.2678,
      "step": 480
    },
    {
      "epoch": 0.062141955529663076,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8933771252632141,
      "eval_runtime": 15.4397,
      "eval_samples_per_second": 2.85,
      "eval_steps_per_second": 0.194,
      "step": 480
    },
    {
      "epoch": 0.062271417937016536,
      "grad_norm": 3.2694234740804493,
      "learning_rate": 7.778137128072445e-07,
      "loss": 4.5474,
      "step": 481
    },
    {
      "epoch": 0.062400880344370004,
      "grad_norm": 2.744101141601191,
      "learning_rate": 7.794307891332472e-07,
      "loss": 4.4507,
      "step": 482
    },
    {
      "epoch": 0.06253034275172346,
      "grad_norm": 2.6973642665625928,
      "learning_rate": 7.810478654592497e-07,
      "loss": 4.2295,
      "step": 483
    },
    {
      "epoch": 0.06265980515907693,
      "grad_norm": 2.8343522695658883,
      "learning_rate": 7.826649417852523e-07,
      "loss": 4.8569,
      "step": 484
    },
    {
      "epoch": 0.0627892675664304,
      "grad_norm": 1.5759981758527295,
      "learning_rate": 7.84282018111255e-07,
      "loss": 3.3433,
      "step": 485
    },
    {
      "epoch": 0.0627892675664304,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8951526880264282,
      "eval_runtime": 14.3716,
      "eval_samples_per_second": 3.062,
      "eval_steps_per_second": 0.209,
      "step": 485
    },
    {
      "epoch": 0.06291872997378387,
      "grad_norm": 3.065308612184375,
      "learning_rate": 7.858990944372575e-07,
      "loss": 4.9492,
      "step": 486
    },
    {
      "epoch": 0.06304819238113733,
      "grad_norm": 3.6197399144473685,
      "learning_rate": 7.8751617076326e-07,
      "loss": 5.8706,
      "step": 487
    },
    {
      "epoch": 0.06317765478849079,
      "grad_norm": 4.4138003218402035,
      "learning_rate": 7.891332470892627e-07,
      "loss": 5.9075,
      "step": 488
    },
    {
      "epoch": 0.06330711719584425,
      "grad_norm": 3.647047343731257,
      "learning_rate": 7.907503234152652e-07,
      "loss": 5.3857,
      "step": 489
    },
    {
      "epoch": 0.06343657960319772,
      "grad_norm": 2.087150238196096,
      "learning_rate": 7.923673997412679e-07,
      "loss": 3.5122,
      "step": 490
    },
    {
      "epoch": 0.06343657960319772,
      "eval_PRM Accuracy": 0.39622641509433965,
      "eval_PRM F1": 0.42857142857142855,
      "eval_PRM F1 AUC": 0.5358826610790991,
      "eval_PRM F1 Neg": 0.36,
      "eval_PRM NPV": 0.23376623376623376,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8950639367103577,
      "eval_runtime": 14.6726,
      "eval_samples_per_second": 2.999,
      "eval_steps_per_second": 0.204,
      "step": 490
    },
    {
      "epoch": 0.06356604201055119,
      "grad_norm": 2.915236854067916,
      "learning_rate": 7.939844760672705e-07,
      "loss": 4.5864,
      "step": 491
    },
    {
      "epoch": 0.06369550441790466,
      "grad_norm": 3.8377116990770133,
      "learning_rate": 7.956015523932731e-07,
      "loss": 5.5801,
      "step": 492
    },
    {
      "epoch": 0.06382496682525812,
      "grad_norm": 1.4409468802970715,
      "learning_rate": 7.972186287192756e-07,
      "loss": 3.3596,
      "step": 493
    },
    {
      "epoch": 0.06395442923261158,
      "grad_norm": 2.1640560193412863,
      "learning_rate": 7.988357050452782e-07,
      "loss": 4.2856,
      "step": 494
    },
    {
      "epoch": 0.06408389163996504,
      "grad_norm": 3.1133391337473446,
      "learning_rate": 8.004527813712807e-07,
      "loss": 4.3745,
      "step": 495
    },
    {
      "epoch": 0.06408389163996504,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8969282507896423,
      "eval_runtime": 13.4116,
      "eval_samples_per_second": 3.281,
      "eval_steps_per_second": 0.224,
      "step": 495
    },
    {
      "epoch": 0.06421335404731851,
      "grad_norm": 3.1589057385124737,
      "learning_rate": 8.020698576972834e-07,
      "loss": 4.1719,
      "step": 496
    },
    {
      "epoch": 0.06434281645467198,
      "grad_norm": 2.5352609093998444,
      "learning_rate": 8.03686934023286e-07,
      "loss": 4.3894,
      "step": 497
    },
    {
      "epoch": 0.06447227886202545,
      "grad_norm": 2.52113941438638,
      "learning_rate": 8.053040103492886e-07,
      "loss": 4.124,
      "step": 498
    },
    {
      "epoch": 0.0646017412693789,
      "grad_norm": 2.444114794921976,
      "learning_rate": 8.069210866752912e-07,
      "loss": 4.2605,
      "step": 499
    },
    {
      "epoch": 0.06473120367673237,
      "grad_norm": 1.8003741061422533,
      "learning_rate": 8.085381630012937e-07,
      "loss": 3.2925,
      "step": 500
    },
    {
      "epoch": 0.06473120367673237,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 Neg": 0.3564356435643564,
      "eval_PRM NPV": 0.23076923076923078,
      "eval_PRM Precision": 0.8214285714285714,
      "eval_PRM Recall": 0.27710843373493976,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8959516882896423,
      "eval_runtime": 15.069,
      "eval_samples_per_second": 2.92,
      "eval_steps_per_second": 0.199,
      "step": 500
    }
  ],
  "logging_steps": 1,
  "max_steps": 7724,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 419680048021504.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}