{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.3883872220603942,
  "eval_steps": 5,
  "global_step": 3000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0,
      "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": 0.8880504369735718,
      "eval_runtime": 19.1231,
      "eval_samples_per_second": 2.301,
      "eval_steps_per_second": 0.157,
      "step": 0
    },
    {
      "epoch": 0.00012946240735346474,
      "grad_norm": 0.08478820147156459,
      "learning_rate": 1.6170763260025876e-09,
      "loss": 1.2262,
      "step": 1
    },
    {
      "epoch": 0.00025892481470692947,
      "grad_norm": 0.11997535582190841,
      "learning_rate": 3.234152652005175e-09,
      "loss": 1.214,
      "step": 2
    },
    {
      "epoch": 0.00038838722206039424,
      "grad_norm": 0.10999069946682373,
      "learning_rate": 4.851228978007762e-09,
      "loss": 1.2927,
      "step": 3
    },
    {
      "epoch": 0.0005178496294138589,
      "grad_norm": 0.10792249824043619,
      "learning_rate": 6.46830530401035e-09,
      "loss": 1.2,
      "step": 4
    },
    {
      "epoch": 0.0006473120367673237,
      "grad_norm": 0.09413050787203431,
      "learning_rate": 8.085381630012938e-09,
      "loss": 1.1707,
      "step": 5
    },
    {
      "epoch": 0.0006473120367673237,
      "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": 20.5842,
      "eval_samples_per_second": 2.138,
      "eval_steps_per_second": 0.146,
      "step": 5
    },
    {
      "epoch": 0.0007767744441207885,
      "grad_norm": 0.10511114915671642,
      "learning_rate": 9.702457956015525e-09,
      "loss": 1.1853,
      "step": 6
    },
    {
      "epoch": 0.0009062368514742531,
      "grad_norm": 0.09602856005634197,
      "learning_rate": 1.1319534282018111e-08,
      "loss": 1.202,
      "step": 7
    },
    {
      "epoch": 0.0010356992588277179,
      "grad_norm": 0.11937441591225446,
      "learning_rate": 1.29366106080207e-08,
      "loss": 1.2573,
      "step": 8
    },
    {
      "epoch": 0.0011651616661811825,
      "grad_norm": 0.08387530730385927,
      "learning_rate": 1.4553686934023287e-08,
      "loss": 1.1508,
      "step": 9
    },
    {
      "epoch": 0.0012946240735346474,
      "grad_norm": 0.10451065153628818,
      "learning_rate": 1.6170763260025877e-08,
      "loss": 1.1758,
      "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": 0.8863636255264282,
      "eval_runtime": 19.8202,
      "eval_samples_per_second": 2.22,
      "eval_steps_per_second": 0.151,
      "step": 10
    },
    {
      "epoch": 0.001424086480888112,
      "grad_norm": 0.11328759701012829,
      "learning_rate": 1.778783958602846e-08,
      "loss": 1.2809,
      "step": 11
    },
    {
      "epoch": 0.001553548888241577,
      "grad_norm": 0.08856579547518892,
      "learning_rate": 1.940491591203105e-08,
      "loss": 1.0582,
      "step": 12
    },
    {
      "epoch": 0.0016830112955950416,
      "grad_norm": 0.10025282389773865,
      "learning_rate": 2.1021992238033634e-08,
      "loss": 1.2438,
      "step": 13
    },
    {
      "epoch": 0.0018124737029485063,
      "grad_norm": 0.09709769598114013,
      "learning_rate": 2.2639068564036222e-08,
      "loss": 1.1768,
      "step": 14
    },
    {
      "epoch": 0.0019419361103019711,
      "grad_norm": 0.1084248047806943,
      "learning_rate": 2.425614489003881e-08,
      "loss": 1.2231,
      "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": 0.8845880627632141,
      "eval_runtime": 20.555,
      "eval_samples_per_second": 2.141,
      "eval_steps_per_second": 0.146,
      "step": 15
    },
    {
      "epoch": 0.0020713985176554358,
      "grad_norm": 0.08562392131185487,
      "learning_rate": 2.58732212160414e-08,
      "loss": 1.1553,
      "step": 16
    },
    {
      "epoch": 0.0022008609250089004,
      "grad_norm": 0.0819473962907359,
      "learning_rate": 2.749029754204399e-08,
      "loss": 1.0406,
      "step": 17
    },
    {
      "epoch": 0.002330323332362365,
      "grad_norm": 0.09796245151249723,
      "learning_rate": 2.9107373868046574e-08,
      "loss": 1.1268,
      "step": 18
    },
    {
      "epoch": 0.00245978573971583,
      "grad_norm": 0.1117574668766745,
      "learning_rate": 3.072445019404916e-08,
      "loss": 1.318,
      "step": 19
    },
    {
      "epoch": 0.002589248147069295,
      "grad_norm": 0.12029247530472317,
      "learning_rate": 3.2341526520051754e-08,
      "loss": 1.3162,
      "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": 0.8869850635528564,
      "eval_runtime": 20.2656,
      "eval_samples_per_second": 2.171,
      "eval_steps_per_second": 0.148,
      "step": 20
    },
    {
      "epoch": 0.0027187105544227595,
      "grad_norm": 0.1115751854829782,
      "learning_rate": 3.395860284605434e-08,
      "loss": 1.1694,
      "step": 21
    },
    {
      "epoch": 0.002848172961776224,
      "grad_norm": 0.12166177559759417,
      "learning_rate": 3.557567917205692e-08,
      "loss": 1.3344,
      "step": 22
    },
    {
      "epoch": 0.002977635369129689,
      "grad_norm": 0.12690149811518361,
      "learning_rate": 3.7192755498059514e-08,
      "loss": 1.4124,
      "step": 23
    },
    {
      "epoch": 0.003107097776483154,
      "grad_norm": 0.07506782524807051,
      "learning_rate": 3.88098318240621e-08,
      "loss": 1.0693,
      "step": 24
    },
    {
      "epoch": 0.0032365601838366185,
      "grad_norm": 0.08128129639128845,
      "learning_rate": 4.042690815006469e-08,
      "loss": 1.0026,
      "step": 25
    },
    {
      "epoch": 0.0032365601838366185,
      "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": 0.8846768736839294,
      "eval_runtime": 20.0996,
      "eval_samples_per_second": 2.189,
      "eval_steps_per_second": 0.149,
      "step": 25
    },
    {
      "epoch": 0.003366022591190083,
      "grad_norm": 0.09332234851084577,
      "learning_rate": 4.204398447606727e-08,
      "loss": 1.147,
      "step": 26
    },
    {
      "epoch": 0.003495484998543548,
      "grad_norm": 0.10621116984373306,
      "learning_rate": 4.366106080206986e-08,
      "loss": 1.1519,
      "step": 27
    },
    {
      "epoch": 0.0036249474058970125,
      "grad_norm": 0.11252425061948415,
      "learning_rate": 4.5278137128072445e-08,
      "loss": 1.2607,
      "step": 28
    },
    {
      "epoch": 0.0037544098132504776,
      "grad_norm": 0.11245264745971341,
      "learning_rate": 4.6895213454075036e-08,
      "loss": 1.2733,
      "step": 29
    },
    {
      "epoch": 0.0038838722206039423,
      "grad_norm": 0.1141344021779216,
      "learning_rate": 4.851228978007762e-08,
      "loss": 1.2542,
      "step": 30
    },
    {
      "epoch": 0.0038838722206039423,
      "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": 0.8828125,
      "eval_runtime": 20.1736,
      "eval_samples_per_second": 2.181,
      "eval_steps_per_second": 0.149,
      "step": 30
    },
    {
      "epoch": 0.004013334627957407,
      "grad_norm": 0.09950694175673312,
      "learning_rate": 5.012936610608021e-08,
      "loss": 1.1949,
      "step": 31
    },
    {
      "epoch": 0.0041427970353108716,
      "grad_norm": 0.10231813041556162,
      "learning_rate": 5.17464424320828e-08,
      "loss": 1.1417,
      "step": 32
    },
    {
      "epoch": 0.004272259442664336,
      "grad_norm": 0.07444023414621885,
      "learning_rate": 5.336351875808539e-08,
      "loss": 1.0649,
      "step": 33
    },
    {
      "epoch": 0.004401721850017801,
      "grad_norm": 0.0878701752276093,
      "learning_rate": 5.498059508408798e-08,
      "loss": 1.2594,
      "step": 34
    },
    {
      "epoch": 0.0045311842573712655,
      "grad_norm": 0.10751157047495194,
      "learning_rate": 5.659767141009056e-08,
      "loss": 1.2001,
      "step": 35
    },
    {
      "epoch": 0.0045311842573712655,
      "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": 20.5201,
      "eval_samples_per_second": 2.144,
      "eval_steps_per_second": 0.146,
      "step": 35
    },
    {
      "epoch": 0.00466064666472473,
      "grad_norm": 0.08922060222908158,
      "learning_rate": 5.821474773609315e-08,
      "loss": 1.0649,
      "step": 36
    },
    {
      "epoch": 0.004790109072078196,
      "grad_norm": 0.07379562505855507,
      "learning_rate": 5.983182406209573e-08,
      "loss": 1.0741,
      "step": 37
    },
    {
      "epoch": 0.00491957147943166,
      "grad_norm": 0.10846816441350182,
      "learning_rate": 6.144890038809832e-08,
      "loss": 1.3319,
      "step": 38
    },
    {
      "epoch": 0.005049033886785125,
      "grad_norm": 0.12094401826707907,
      "learning_rate": 6.306597671410092e-08,
      "loss": 1.3728,
      "step": 39
    },
    {
      "epoch": 0.00517849629413859,
      "grad_norm": 0.10934747102979701,
      "learning_rate": 6.468305304010351e-08,
      "loss": 1.3578,
      "step": 40
    },
    {
      "epoch": 0.00517849629413859,
      "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": 20.0654,
      "eval_samples_per_second": 2.193,
      "eval_steps_per_second": 0.15,
      "step": 40
    },
    {
      "epoch": 0.005307958701492054,
      "grad_norm": 0.10615453948153131,
      "learning_rate": 6.630012936610609e-08,
      "loss": 1.2599,
      "step": 41
    },
    {
      "epoch": 0.005437421108845519,
      "grad_norm": 0.11056036153366629,
      "learning_rate": 6.791720569210868e-08,
      "loss": 1.3488,
      "step": 42
    },
    {
      "epoch": 0.005566883516198984,
      "grad_norm": 0.09058223260043745,
      "learning_rate": 6.953428201811125e-08,
      "loss": 1.1667,
      "step": 43
    },
    {
      "epoch": 0.005696345923552448,
      "grad_norm": 0.10216779903545326,
      "learning_rate": 7.115135834411385e-08,
      "loss": 1.1344,
      "step": 44
    },
    {
      "epoch": 0.005825808330905913,
      "grad_norm": 0.06932051110481108,
      "learning_rate": 7.276843467011644e-08,
      "loss": 1.0135,
      "step": 45
    },
    {
      "epoch": 0.005825808330905913,
      "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": 0.8825461864471436,
      "eval_runtime": 20.2412,
      "eval_samples_per_second": 2.174,
      "eval_steps_per_second": 0.148,
      "step": 45
    },
    {
      "epoch": 0.005955270738259378,
      "grad_norm": 0.11523728703934258,
      "learning_rate": 7.438551099611903e-08,
      "loss": 1.3075,
      "step": 46
    },
    {
      "epoch": 0.006084733145612842,
      "grad_norm": 0.0984733810382846,
      "learning_rate": 7.600258732212161e-08,
      "loss": 1.2125,
      "step": 47
    },
    {
      "epoch": 0.006214195552966308,
      "grad_norm": 0.09237644777038281,
      "learning_rate": 7.76196636481242e-08,
      "loss": 1.1675,
      "step": 48
    },
    {
      "epoch": 0.006343657960319772,
      "grad_norm": 0.07478896592983698,
      "learning_rate": 7.923673997412678e-08,
      "loss": 1.1703,
      "step": 49
    },
    {
      "epoch": 0.006473120367673237,
      "grad_norm": 0.11855115146882207,
      "learning_rate": 8.085381630012938e-08,
      "loss": 1.3152,
      "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": 0.8862748742103577,
      "eval_runtime": 20.2464,
      "eval_samples_per_second": 2.173,
      "eval_steps_per_second": 0.148,
      "step": 50
    },
    {
      "epoch": 0.006602582775026702,
      "grad_norm": 0.11306115000527275,
      "learning_rate": 8.247089262613197e-08,
      "loss": 1.2427,
      "step": 51
    },
    {
      "epoch": 0.006732045182380166,
      "grad_norm": 0.10273504664211115,
      "learning_rate": 8.408796895213454e-08,
      "loss": 1.2511,
      "step": 52
    },
    {
      "epoch": 0.006861507589733631,
      "grad_norm": 0.09010794899306171,
      "learning_rate": 8.570504527813713e-08,
      "loss": 1.2772,
      "step": 53
    },
    {
      "epoch": 0.006990969997087096,
      "grad_norm": 0.11397035649023907,
      "learning_rate": 8.732212160413972e-08,
      "loss": 1.3699,
      "step": 54
    },
    {
      "epoch": 0.00712043240444056,
      "grad_norm": 0.08383378352842623,
      "learning_rate": 8.893919793014231e-08,
      "loss": 1.127,
      "step": 55
    },
    {
      "epoch": 0.00712043240444056,
      "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": 20.2884,
      "eval_samples_per_second": 2.169,
      "eval_steps_per_second": 0.148,
      "step": 55
    },
    {
      "epoch": 0.007249894811794025,
      "grad_norm": 0.07516516953134732,
      "learning_rate": 9.055627425614489e-08,
      "loss": 1.0452,
      "step": 56
    },
    {
      "epoch": 0.00737935721914749,
      "grad_norm": 0.11856008192807092,
      "learning_rate": 9.217335058214748e-08,
      "loss": 1.3323,
      "step": 57
    },
    {
      "epoch": 0.007508819626500955,
      "grad_norm": 0.09448177737493262,
      "learning_rate": 9.379042690815007e-08,
      "loss": 1.1937,
      "step": 58
    },
    {
      "epoch": 0.00763828203385442,
      "grad_norm": 0.06674465926141343,
      "learning_rate": 9.540750323415266e-08,
      "loss": 1.0464,
      "step": 59
    },
    {
      "epoch": 0.0077677444412078845,
      "grad_norm": 0.10245213113477966,
      "learning_rate": 9.702457956015524e-08,
      "loss": 1.1388,
      "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": 0.8860085010528564,
      "eval_runtime": 20.2681,
      "eval_samples_per_second": 2.171,
      "eval_steps_per_second": 0.148,
      "step": 60
    },
    {
      "epoch": 0.007897206848561348,
      "grad_norm": 0.13000310677651578,
      "learning_rate": 9.864165588615783e-08,
      "loss": 1.3938,
      "step": 61
    },
    {
      "epoch": 0.008026669255914814,
      "grad_norm": 0.08635161625163723,
      "learning_rate": 1.0025873221216042e-07,
      "loss": 1.1785,
      "step": 62
    },
    {
      "epoch": 0.00815613166326828,
      "grad_norm": 0.09060957968860127,
      "learning_rate": 1.0187580853816302e-07,
      "loss": 1.1366,
      "step": 63
    },
    {
      "epoch": 0.008285594070621743,
      "grad_norm": 0.12014586031798641,
      "learning_rate": 1.034928848641656e-07,
      "loss": 1.3383,
      "step": 64
    },
    {
      "epoch": 0.008415056477975209,
      "grad_norm": 0.12599238562069226,
      "learning_rate": 1.0510996119016818e-07,
      "loss": 1.3347,
      "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": 0.8860973119735718,
      "eval_runtime": 20.3059,
      "eval_samples_per_second": 2.167,
      "eval_steps_per_second": 0.148,
      "step": 65
    },
    {
      "epoch": 0.008544518885328672,
      "grad_norm": 0.11066436683839612,
      "learning_rate": 1.0672703751617078e-07,
      "loss": 1.2475,
      "step": 66
    },
    {
      "epoch": 0.008673981292682138,
      "grad_norm": 0.07545837744852751,
      "learning_rate": 1.0834411384217337e-07,
      "loss": 1.0618,
      "step": 67
    },
    {
      "epoch": 0.008803443700035602,
      "grad_norm": 0.12517435089416862,
      "learning_rate": 1.0996119016817596e-07,
      "loss": 1.2844,
      "step": 68
    },
    {
      "epoch": 0.008932906107389067,
      "grad_norm": 0.12454053197097449,
      "learning_rate": 1.1157826649417852e-07,
      "loss": 1.3711,
      "step": 69
    },
    {
      "epoch": 0.009062368514742531,
      "grad_norm": 0.11660093813063876,
      "learning_rate": 1.1319534282018111e-07,
      "loss": 1.2864,
      "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": 0.8854758739471436,
      "eval_runtime": 21.1441,
      "eval_samples_per_second": 2.081,
      "eval_steps_per_second": 0.142,
      "step": 70
    },
    {
      "epoch": 0.009191830922095997,
      "grad_norm": 0.07479576987889343,
      "learning_rate": 1.148124191461837e-07,
      "loss": 1.0665,
      "step": 71
    },
    {
      "epoch": 0.00932129332944946,
      "grad_norm": 0.09645615123636787,
      "learning_rate": 1.164294954721863e-07,
      "loss": 1.0367,
      "step": 72
    },
    {
      "epoch": 0.009450755736802926,
      "grad_norm": 0.09067405511936866,
      "learning_rate": 1.1804657179818888e-07,
      "loss": 1.2028,
      "step": 73
    },
    {
      "epoch": 0.009580218144156391,
      "grad_norm": 0.11564943494312371,
      "learning_rate": 1.1966364812419147e-07,
      "loss": 1.256,
      "step": 74
    },
    {
      "epoch": 0.009709680551509855,
      "grad_norm": 0.10836169635571818,
      "learning_rate": 1.2128072445019406e-07,
      "loss": 1.0905,
      "step": 75
    },
    {
      "epoch": 0.009709680551509855,
      "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": 0.8829900622367859,
      "eval_runtime": 20.3985,
      "eval_samples_per_second": 2.157,
      "eval_steps_per_second": 0.147,
      "step": 75
    },
    {
      "epoch": 0.00983914295886332,
      "grad_norm": 0.07584301973141086,
      "learning_rate": 1.2289780077619665e-07,
      "loss": 1.0189,
      "step": 76
    },
    {
      "epoch": 0.009968605366216785,
      "grad_norm": 0.117847899173823,
      "learning_rate": 1.2451487710219921e-07,
      "loss": 1.2346,
      "step": 77
    },
    {
      "epoch": 0.01009806777357025,
      "grad_norm": 0.10531149354765305,
      "learning_rate": 1.2613195342820183e-07,
      "loss": 1.3047,
      "step": 78
    },
    {
      "epoch": 0.010227530180923714,
      "grad_norm": 0.07952947755932173,
      "learning_rate": 1.2774902975420442e-07,
      "loss": 1.1699,
      "step": 79
    },
    {
      "epoch": 0.01035699258827718,
      "grad_norm": 0.09609598542285369,
      "learning_rate": 1.2936610608020701e-07,
      "loss": 1.196,
      "step": 80
    },
    {
      "epoch": 0.01035699258827718,
      "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": 0.8870738744735718,
      "eval_runtime": 20.1556,
      "eval_samples_per_second": 2.183,
      "eval_steps_per_second": 0.149,
      "step": 80
    },
    {
      "epoch": 0.010486454995630643,
      "grad_norm": 0.09054244024420528,
      "learning_rate": 1.3098318240620958e-07,
      "loss": 1.2297,
      "step": 81
    },
    {
      "epoch": 0.010615917402984109,
      "grad_norm": 0.08483946841297615,
      "learning_rate": 1.3260025873221217e-07,
      "loss": 1.0752,
      "step": 82
    },
    {
      "epoch": 0.010745379810337572,
      "grad_norm": 0.09905083310065368,
      "learning_rate": 1.3421733505821476e-07,
      "loss": 1.1501,
      "step": 83
    },
    {
      "epoch": 0.010874842217691038,
      "grad_norm": 0.09265993898604435,
      "learning_rate": 1.3583441138421735e-07,
      "loss": 1.1158,
      "step": 84
    },
    {
      "epoch": 0.011004304625044503,
      "grad_norm": 0.10181483376814672,
      "learning_rate": 1.3745148771021992e-07,
      "loss": 1.2386,
      "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": 0.884765625,
      "eval_runtime": 19.6997,
      "eval_samples_per_second": 2.234,
      "eval_steps_per_second": 0.152,
      "step": 85
    },
    {
      "epoch": 0.011133767032397967,
      "grad_norm": 0.10312905764108671,
      "learning_rate": 1.390685640362225e-07,
      "loss": 1.1814,
      "step": 86
    },
    {
      "epoch": 0.011263229439751433,
      "grad_norm": 0.1057201907117079,
      "learning_rate": 1.406856403622251e-07,
      "loss": 1.2407,
      "step": 87
    },
    {
      "epoch": 0.011392691847104897,
      "grad_norm": 0.11980798245318529,
      "learning_rate": 1.423027166882277e-07,
      "loss": 1.2464,
      "step": 88
    },
    {
      "epoch": 0.011522154254458362,
      "grad_norm": 0.09278733411176136,
      "learning_rate": 1.4391979301423028e-07,
      "loss": 1.3153,
      "step": 89
    },
    {
      "epoch": 0.011651616661811826,
      "grad_norm": 0.10575730686255778,
      "learning_rate": 1.4553686934023288e-07,
      "loss": 1.2704,
      "step": 90
    },
    {
      "epoch": 0.011651616661811826,
      "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": 0.8860973119735718,
      "eval_runtime": 20.6185,
      "eval_samples_per_second": 2.134,
      "eval_steps_per_second": 0.146,
      "step": 90
    },
    {
      "epoch": 0.011781079069165291,
      "grad_norm": 0.14058086621365165,
      "learning_rate": 1.4715394566623547e-07,
      "loss": 1.4902,
      "step": 91
    },
    {
      "epoch": 0.011910541476518755,
      "grad_norm": 0.08615073884517986,
      "learning_rate": 1.4877102199223806e-07,
      "loss": 1.0728,
      "step": 92
    },
    {
      "epoch": 0.01204000388387222,
      "grad_norm": 0.12507226495106538,
      "learning_rate": 1.5038809831824065e-07,
      "loss": 1.335,
      "step": 93
    },
    {
      "epoch": 0.012169466291225685,
      "grad_norm": 0.08305666623903088,
      "learning_rate": 1.5200517464424321e-07,
      "loss": 1.1194,
      "step": 94
    },
    {
      "epoch": 0.01229892869857915,
      "grad_norm": 0.11375931151205587,
      "learning_rate": 1.536222509702458e-07,
      "loss": 1.2041,
      "step": 95
    },
    {
      "epoch": 0.01229892869857915,
      "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": 0.8852983117103577,
      "eval_runtime": 20.5812,
      "eval_samples_per_second": 2.138,
      "eval_steps_per_second": 0.146,
      "step": 95
    },
    {
      "epoch": 0.012428391105932616,
      "grad_norm": 0.08253689217584612,
      "learning_rate": 1.552393272962484e-07,
      "loss": 1.2003,
      "step": 96
    },
    {
      "epoch": 0.01255785351328608,
      "grad_norm": 0.10806970834071163,
      "learning_rate": 1.56856403622251e-07,
      "loss": 1.3441,
      "step": 97
    },
    {
      "epoch": 0.012687315920639545,
      "grad_norm": 0.09914443401391931,
      "learning_rate": 1.5847347994825355e-07,
      "loss": 1.1819,
      "step": 98
    },
    {
      "epoch": 0.012816778327993009,
      "grad_norm": 0.07289540404594819,
      "learning_rate": 1.6009055627425617e-07,
      "loss": 1.0067,
      "step": 99
    },
    {
      "epoch": 0.012946240735346474,
      "grad_norm": 0.10448244167752356,
      "learning_rate": 1.6170763260025876e-07,
      "loss": 1.1703,
      "step": 100
    },
    {
      "epoch": 0.012946240735346474,
      "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": 0.8859197497367859,
      "eval_runtime": 20.3168,
      "eval_samples_per_second": 2.166,
      "eval_steps_per_second": 0.148,
      "step": 100
    },
    {
      "epoch": 0.013075703142699938,
      "grad_norm": 0.09147263332242413,
      "learning_rate": 1.6332470892626133e-07,
      "loss": 1.1608,
      "step": 101
    },
    {
      "epoch": 0.013205165550053403,
      "grad_norm": 0.12796359721874298,
      "learning_rate": 1.6494178525226394e-07,
      "loss": 1.3723,
      "step": 102
    },
    {
      "epoch": 0.013334627957406867,
      "grad_norm": 0.09686317035210135,
      "learning_rate": 1.665588615782665e-07,
      "loss": 1.177,
      "step": 103
    },
    {
      "epoch": 0.013464090364760333,
      "grad_norm": 0.11375885459238952,
      "learning_rate": 1.6817593790426907e-07,
      "loss": 1.33,
      "step": 104
    },
    {
      "epoch": 0.013593552772113798,
      "grad_norm": 0.13703539543151838,
      "learning_rate": 1.697930142302717e-07,
      "loss": 1.4099,
      "step": 105
    },
    {
      "epoch": 0.013593552772113798,
      "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": 0.8818359375,
      "eval_runtime": 20.9063,
      "eval_samples_per_second": 2.105,
      "eval_steps_per_second": 0.143,
      "step": 105
    },
    {
      "epoch": 0.013723015179467262,
      "grad_norm": 0.10634630000389733,
      "learning_rate": 1.7141009055627426e-07,
      "loss": 1.0785,
      "step": 106
    },
    {
      "epoch": 0.013852477586820728,
      "grad_norm": 0.08381174368678503,
      "learning_rate": 1.7302716688227687e-07,
      "loss": 1.1342,
      "step": 107
    },
    {
      "epoch": 0.013981939994174191,
      "grad_norm": 0.0973235401061372,
      "learning_rate": 1.7464424320827944e-07,
      "loss": 1.3323,
      "step": 108
    },
    {
      "epoch": 0.014111402401527657,
      "grad_norm": 0.09085491229240962,
      "learning_rate": 1.7626131953428203e-07,
      "loss": 1.1328,
      "step": 109
    },
    {
      "epoch": 0.01424086480888112,
      "grad_norm": 0.10582619691280479,
      "learning_rate": 1.7787839586028462e-07,
      "loss": 1.2716,
      "step": 110
    },
    {
      "epoch": 0.01424086480888112,
      "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": 0.8857421875,
      "eval_runtime": 21.4307,
      "eval_samples_per_second": 2.053,
      "eval_steps_per_second": 0.14,
      "step": 110
    },
    {
      "epoch": 0.014370327216234586,
      "grad_norm": 0.10386758070117412,
      "learning_rate": 1.7949547218628721e-07,
      "loss": 1.2521,
      "step": 111
    },
    {
      "epoch": 0.01449978962358805,
      "grad_norm": 0.10994280991393013,
      "learning_rate": 1.8111254851228978e-07,
      "loss": 1.1874,
      "step": 112
    },
    {
      "epoch": 0.014629252030941516,
      "grad_norm": 0.13644961308357315,
      "learning_rate": 1.827296248382924e-07,
      "loss": 1.3711,
      "step": 113
    },
    {
      "epoch": 0.01475871443829498,
      "grad_norm": 0.1308037942720227,
      "learning_rate": 1.8434670116429496e-07,
      "loss": 1.33,
      "step": 114
    },
    {
      "epoch": 0.014888176845648445,
      "grad_norm": 0.10698053418325305,
      "learning_rate": 1.8596377749029758e-07,
      "loss": 1.2362,
      "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": 0.8817471861839294,
      "eval_runtime": 20.661,
      "eval_samples_per_second": 2.13,
      "eval_steps_per_second": 0.145,
      "step": 115
    },
    {
      "epoch": 0.01501763925300191,
      "grad_norm": 0.10999851314389243,
      "learning_rate": 1.8758085381630014e-07,
      "loss": 1.2155,
      "step": 116
    },
    {
      "epoch": 0.015147101660355374,
      "grad_norm": 0.10084798210588232,
      "learning_rate": 1.891979301423027e-07,
      "loss": 1.2314,
      "step": 117
    },
    {
      "epoch": 0.01527656406770884,
      "grad_norm": 0.08214399336937146,
      "learning_rate": 1.9081500646830533e-07,
      "loss": 1.1076,
      "step": 118
    },
    {
      "epoch": 0.015406026475062303,
      "grad_norm": 0.11047401883926936,
      "learning_rate": 1.924320827943079e-07,
      "loss": 1.2844,
      "step": 119
    },
    {
      "epoch": 0.015535488882415769,
      "grad_norm": 0.11532925583583147,
      "learning_rate": 1.9404915912031048e-07,
      "loss": 1.3184,
      "step": 120
    },
    {
      "epoch": 0.015535488882415769,
      "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": 20.0268,
      "eval_samples_per_second": 2.197,
      "eval_steps_per_second": 0.15,
      "step": 120
    },
    {
      "epoch": 0.015664951289769233,
      "grad_norm": 0.08665120768487652,
      "learning_rate": 1.9566623544631307e-07,
      "loss": 1.0255,
      "step": 121
    },
    {
      "epoch": 0.015794413697122697,
      "grad_norm": 0.10049843276534641,
      "learning_rate": 1.9728331177231567e-07,
      "loss": 1.103,
      "step": 122
    },
    {
      "epoch": 0.015923876104476164,
      "grad_norm": 0.09048210269503958,
      "learning_rate": 1.9890038809831828e-07,
      "loss": 1.1517,
      "step": 123
    },
    {
      "epoch": 0.016053338511829628,
      "grad_norm": 0.10462884544935858,
      "learning_rate": 2.0051746442432085e-07,
      "loss": 1.1445,
      "step": 124
    },
    {
      "epoch": 0.01618280091918309,
      "grad_norm": 0.10662281387417837,
      "learning_rate": 2.021345407503234e-07,
      "loss": 1.2117,
      "step": 125
    },
    {
      "epoch": 0.01618280091918309,
      "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": 20.4419,
      "eval_samples_per_second": 2.152,
      "eval_steps_per_second": 0.147,
      "step": 125
    },
    {
      "epoch": 0.01631226332653656,
      "grad_norm": 0.09115046987556065,
      "learning_rate": 2.0375161707632603e-07,
      "loss": 1.196,
      "step": 126
    },
    {
      "epoch": 0.016441725733890022,
      "grad_norm": 0.11632428297536729,
      "learning_rate": 2.053686934023286e-07,
      "loss": 1.2394,
      "step": 127
    },
    {
      "epoch": 0.016571188141243486,
      "grad_norm": 0.10485496729903382,
      "learning_rate": 2.069857697283312e-07,
      "loss": 1.1929,
      "step": 128
    },
    {
      "epoch": 0.01670065054859695,
      "grad_norm": 0.09634706718889466,
      "learning_rate": 2.0860284605433378e-07,
      "loss": 1.2444,
      "step": 129
    },
    {
      "epoch": 0.016830112955950417,
      "grad_norm": 0.10355791111554799,
      "learning_rate": 2.1021992238033637e-07,
      "loss": 1.2194,
      "step": 130
    },
    {
      "epoch": 0.016830112955950417,
      "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": 0.8852983117103577,
      "eval_runtime": 20.8757,
      "eval_samples_per_second": 2.108,
      "eval_steps_per_second": 0.144,
      "step": 130
    },
    {
      "epoch": 0.01695957536330388,
      "grad_norm": 0.12484078159965999,
      "learning_rate": 2.1183699870633896e-07,
      "loss": 1.4118,
      "step": 131
    },
    {
      "epoch": 0.017089037770657345,
      "grad_norm": 0.10050898992299352,
      "learning_rate": 2.1345407503234155e-07,
      "loss": 1.2192,
      "step": 132
    },
    {
      "epoch": 0.01721850017801081,
      "grad_norm": 0.06844465902612236,
      "learning_rate": 2.1507115135834412e-07,
      "loss": 0.9901,
      "step": 133
    },
    {
      "epoch": 0.017347962585364276,
      "grad_norm": 0.09096242074062204,
      "learning_rate": 2.1668822768434673e-07,
      "loss": 1.0815,
      "step": 134
    },
    {
      "epoch": 0.01747742499271774,
      "grad_norm": 0.12122302707487999,
      "learning_rate": 2.183053040103493e-07,
      "loss": 1.3611,
      "step": 135
    },
    {
      "epoch": 0.01747742499271774,
      "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": 0.8866299986839294,
      "eval_runtime": 20.4893,
      "eval_samples_per_second": 2.147,
      "eval_steps_per_second": 0.146,
      "step": 135
    },
    {
      "epoch": 0.017606887400071203,
      "grad_norm": 0.11987113821834161,
      "learning_rate": 2.1992238033635192e-07,
      "loss": 1.2358,
      "step": 136
    },
    {
      "epoch": 0.01773634980742467,
      "grad_norm": 0.1263298110410404,
      "learning_rate": 2.2153945666235448e-07,
      "loss": 1.2943,
      "step": 137
    },
    {
      "epoch": 0.017865812214778135,
      "grad_norm": 0.0920443314472666,
      "learning_rate": 2.2315653298835705e-07,
      "loss": 1.1836,
      "step": 138
    },
    {
      "epoch": 0.0179952746221316,
      "grad_norm": 0.11191188650275118,
      "learning_rate": 2.2477360931435966e-07,
      "loss": 1.2073,
      "step": 139
    },
    {
      "epoch": 0.018124737029485062,
      "grad_norm": 0.08294395850075541,
      "learning_rate": 2.2639068564036223e-07,
      "loss": 1.1741,
      "step": 140
    },
    {
      "epoch": 0.018124737029485062,
      "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": 0.8855646252632141,
      "eval_runtime": 20.0043,
      "eval_samples_per_second": 2.2,
      "eval_steps_per_second": 0.15,
      "step": 140
    },
    {
      "epoch": 0.01825419943683853,
      "grad_norm": 0.11171080057835624,
      "learning_rate": 2.2800776196636482e-07,
      "loss": 1.2324,
      "step": 141
    },
    {
      "epoch": 0.018383661844191993,
      "grad_norm": 0.10796314469932673,
      "learning_rate": 2.296248382923674e-07,
      "loss": 1.1721,
      "step": 142
    },
    {
      "epoch": 0.018513124251545457,
      "grad_norm": 0.12712398362272836,
      "learning_rate": 2.3124191461837e-07,
      "loss": 1.4122,
      "step": 143
    },
    {
      "epoch": 0.01864258665889892,
      "grad_norm": 0.10355457326181436,
      "learning_rate": 2.328589909443726e-07,
      "loss": 1.1605,
      "step": 144
    },
    {
      "epoch": 0.018772049066252388,
      "grad_norm": 0.09328770288749962,
      "learning_rate": 2.3447606727037519e-07,
      "loss": 1.1057,
      "step": 145
    },
    {
      "epoch": 0.018772049066252388,
      "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": 0.8840553760528564,
      "eval_runtime": 19.9153,
      "eval_samples_per_second": 2.209,
      "eval_steps_per_second": 0.151,
      "step": 145
    },
    {
      "epoch": 0.018901511473605852,
      "grad_norm": 0.13310280044035516,
      "learning_rate": 2.3609314359637775e-07,
      "loss": 1.3058,
      "step": 146
    },
    {
      "epoch": 0.019030973880959316,
      "grad_norm": 0.07963672395336137,
      "learning_rate": 2.3771021992238037e-07,
      "loss": 1.1184,
      "step": 147
    },
    {
      "epoch": 0.019160436288312783,
      "grad_norm": 0.1111750252328608,
      "learning_rate": 2.3932729624838293e-07,
      "loss": 1.2821,
      "step": 148
    },
    {
      "epoch": 0.019289898695666247,
      "grad_norm": 0.11532975323041379,
      "learning_rate": 2.409443725743855e-07,
      "loss": 1.2628,
      "step": 149
    },
    {
      "epoch": 0.01941936110301971,
      "grad_norm": 0.10682092979339192,
      "learning_rate": 2.425614489003881e-07,
      "loss": 1.2635,
      "step": 150
    },
    {
      "epoch": 0.01941936110301971,
      "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": 0.8822798132896423,
      "eval_runtime": 19.7774,
      "eval_samples_per_second": 2.225,
      "eval_steps_per_second": 0.152,
      "step": 150
    },
    {
      "epoch": 0.019548823510373174,
      "grad_norm": 0.10463875736106931,
      "learning_rate": 2.441785252263907e-07,
      "loss": 1.173,
      "step": 151
    },
    {
      "epoch": 0.01967828591772664,
      "grad_norm": 0.09709167686807113,
      "learning_rate": 2.457956015523933e-07,
      "loss": 1.2463,
      "step": 152
    },
    {
      "epoch": 0.019807748325080105,
      "grad_norm": 0.10960151129651464,
      "learning_rate": 2.4741267787839586e-07,
      "loss": 1.1149,
      "step": 153
    },
    {
      "epoch": 0.01993721073243357,
      "grad_norm": 0.12752290392863128,
      "learning_rate": 2.4902975420439843e-07,
      "loss": 1.3616,
      "step": 154
    },
    {
      "epoch": 0.020066673139787033,
      "grad_norm": 0.09816483043952855,
      "learning_rate": 2.5064683053040105e-07,
      "loss": 1.1199,
      "step": 155
    },
    {
      "epoch": 0.020066673139787033,
      "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": 0.8838778138160706,
      "eval_runtime": 20.5984,
      "eval_samples_per_second": 2.136,
      "eval_steps_per_second": 0.146,
      "step": 155
    },
    {
      "epoch": 0.0201961355471405,
      "grad_norm": 0.09463984563935361,
      "learning_rate": 2.5226390685640366e-07,
      "loss": 1.1071,
      "step": 156
    },
    {
      "epoch": 0.020325597954493964,
      "grad_norm": 0.09168009370511442,
      "learning_rate": 2.5388098318240623e-07,
      "loss": 1.1381,
      "step": 157
    },
    {
      "epoch": 0.020455060361847428,
      "grad_norm": 0.09208257774573826,
      "learning_rate": 2.5549805950840885e-07,
      "loss": 1.1376,
      "step": 158
    },
    {
      "epoch": 0.020584522769200895,
      "grad_norm": 0.12427586657092204,
      "learning_rate": 2.571151358344114e-07,
      "loss": 1.3672,
      "step": 159
    },
    {
      "epoch": 0.02071398517655436,
      "grad_norm": 0.11573404557512999,
      "learning_rate": 2.5873221216041403e-07,
      "loss": 1.2316,
      "step": 160
    },
    {
      "epoch": 0.02071398517655436,
      "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": 20.0467,
      "eval_samples_per_second": 2.195,
      "eval_steps_per_second": 0.15,
      "step": 160
    },
    {
      "epoch": 0.020843447583907822,
      "grad_norm": 0.11007027110021322,
      "learning_rate": 2.603492884864166e-07,
      "loss": 1.2844,
      "step": 161
    },
    {
      "epoch": 0.020972909991261286,
      "grad_norm": 0.13709670419430806,
      "learning_rate": 2.6196636481241916e-07,
      "loss": 1.3548,
      "step": 162
    },
    {
      "epoch": 0.021102372398614754,
      "grad_norm": 0.11317611095748434,
      "learning_rate": 2.635834411384218e-07,
      "loss": 1.2104,
      "step": 163
    },
    {
      "epoch": 0.021231834805968217,
      "grad_norm": 0.10672103090068989,
      "learning_rate": 2.6520051746442434e-07,
      "loss": 1.1521,
      "step": 164
    },
    {
      "epoch": 0.02136129721332168,
      "grad_norm": 0.10834670899964745,
      "learning_rate": 2.6681759379042696e-07,
      "loss": 1.1912,
      "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": 0.8844105005264282,
      "eval_runtime": 19.3894,
      "eval_samples_per_second": 2.269,
      "eval_steps_per_second": 0.155,
      "step": 165
    },
    {
      "epoch": 0.021490759620675145,
      "grad_norm": 0.08450796503641816,
      "learning_rate": 2.684346701164295e-07,
      "loss": 1.1027,
      "step": 166
    },
    {
      "epoch": 0.021620222028028612,
      "grad_norm": 0.10570892962065266,
      "learning_rate": 2.700517464424321e-07,
      "loss": 1.1785,
      "step": 167
    },
    {
      "epoch": 0.021749684435382076,
      "grad_norm": 0.11777722316396518,
      "learning_rate": 2.716688227684347e-07,
      "loss": 1.3344,
      "step": 168
    },
    {
      "epoch": 0.02187914684273554,
      "grad_norm": 0.12069996743831052,
      "learning_rate": 2.7328589909443727e-07,
      "loss": 1.2674,
      "step": 169
    },
    {
      "epoch": 0.022008609250089007,
      "grad_norm": 0.1110754890407735,
      "learning_rate": 2.7490297542043984e-07,
      "loss": 1.243,
      "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": 0.8849431872367859,
      "eval_runtime": 19.8864,
      "eval_samples_per_second": 2.213,
      "eval_steps_per_second": 0.151,
      "step": 170
    },
    {
      "epoch": 0.02213807165744247,
      "grad_norm": 0.07630903696938833,
      "learning_rate": 2.7652005174644245e-07,
      "loss": 0.9263,
      "step": 171
    },
    {
      "epoch": 0.022267534064795935,
      "grad_norm": 0.10526605108944664,
      "learning_rate": 2.78137128072445e-07,
      "loss": 1.225,
      "step": 172
    },
    {
      "epoch": 0.0223969964721494,
      "grad_norm": 0.09651886186696854,
      "learning_rate": 2.7975420439844764e-07,
      "loss": 1.3138,
      "step": 173
    },
    {
      "epoch": 0.022526458879502866,
      "grad_norm": 0.10471496414422968,
      "learning_rate": 2.813712807244502e-07,
      "loss": 1.0485,
      "step": 174
    },
    {
      "epoch": 0.02265592128685633,
      "grad_norm": 0.09863242292167222,
      "learning_rate": 2.8298835705045277e-07,
      "loss": 1.2506,
      "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": 0.8832563757896423,
      "eval_runtime": 19.8236,
      "eval_samples_per_second": 2.22,
      "eval_steps_per_second": 0.151,
      "step": 175
    },
    {
      "epoch": 0.022785383694209793,
      "grad_norm": 0.09347477037272767,
      "learning_rate": 2.846054333764554e-07,
      "loss": 1.1108,
      "step": 176
    },
    {
      "epoch": 0.022914846101563257,
      "grad_norm": 0.11097995559327532,
      "learning_rate": 2.86222509702458e-07,
      "loss": 1.286,
      "step": 177
    },
    {
      "epoch": 0.023044308508916724,
      "grad_norm": 0.1000511341539029,
      "learning_rate": 2.8783958602846057e-07,
      "loss": 1.1928,
      "step": 178
    },
    {
      "epoch": 0.023173770916270188,
      "grad_norm": 0.10266864515715982,
      "learning_rate": 2.894566623544632e-07,
      "loss": 1.2,
      "step": 179
    },
    {
      "epoch": 0.023303233323623652,
      "grad_norm": 0.1137372586049092,
      "learning_rate": 2.9107373868046575e-07,
      "loss": 1.2178,
      "step": 180
    },
    {
      "epoch": 0.023303233323623652,
      "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": 0.8844105005264282,
      "eval_runtime": 20.0425,
      "eval_samples_per_second": 2.195,
      "eval_steps_per_second": 0.15,
      "step": 180
    },
    {
      "epoch": 0.02343269573097712,
      "grad_norm": 0.1162989138995753,
      "learning_rate": 2.9269081500646837e-07,
      "loss": 1.2488,
      "step": 181
    },
    {
      "epoch": 0.023562158138330583,
      "grad_norm": 0.07268273126639616,
      "learning_rate": 2.9430789133247093e-07,
      "loss": 1.0839,
      "step": 182
    },
    {
      "epoch": 0.023691620545684047,
      "grad_norm": 0.13490770235553373,
      "learning_rate": 2.959249676584735e-07,
      "loss": 1.3677,
      "step": 183
    },
    {
      "epoch": 0.02382108295303751,
      "grad_norm": 0.13904026697622657,
      "learning_rate": 2.975420439844761e-07,
      "loss": 1.4336,
      "step": 184
    },
    {
      "epoch": 0.023950545360390978,
      "grad_norm": 0.1305363808493564,
      "learning_rate": 2.991591203104787e-07,
      "loss": 1.3962,
      "step": 185
    },
    {
      "epoch": 0.023950545360390978,
      "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": 0.8872514367103577,
      "eval_runtime": 20.3875,
      "eval_samples_per_second": 2.158,
      "eval_steps_per_second": 0.147,
      "step": 185
    },
    {
      "epoch": 0.02408000776774444,
      "grad_norm": 0.13603228259075914,
      "learning_rate": 3.007761966364813e-07,
      "loss": 1.3553,
      "step": 186
    },
    {
      "epoch": 0.024209470175097905,
      "grad_norm": 0.10879136659648139,
      "learning_rate": 3.0239327296248386e-07,
      "loss": 1.3395,
      "step": 187
    },
    {
      "epoch": 0.02433893258245137,
      "grad_norm": 0.1180765829038625,
      "learning_rate": 3.0401034928848643e-07,
      "loss": 1.2233,
      "step": 188
    },
    {
      "epoch": 0.024468394989804836,
      "grad_norm": 0.08512528665542031,
      "learning_rate": 3.0562742561448905e-07,
      "loss": 1.123,
      "step": 189
    },
    {
      "epoch": 0.0245978573971583,
      "grad_norm": 0.10854412163899281,
      "learning_rate": 3.072445019404916e-07,
      "loss": 1.1178,
      "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": 0.8865411877632141,
      "eval_runtime": 19.6552,
      "eval_samples_per_second": 2.239,
      "eval_steps_per_second": 0.153,
      "step": 190
    },
    {
      "epoch": 0.024727319804511764,
      "grad_norm": 0.11244561721742004,
      "learning_rate": 3.088615782664942e-07,
      "loss": 1.257,
      "step": 191
    },
    {
      "epoch": 0.02485678221186523,
      "grad_norm": 0.10512136364338531,
      "learning_rate": 3.104786545924968e-07,
      "loss": 1.0677,
      "step": 192
    },
    {
      "epoch": 0.024986244619218695,
      "grad_norm": 0.12922699622988681,
      "learning_rate": 3.1209573091849936e-07,
      "loss": 1.2427,
      "step": 193
    },
    {
      "epoch": 0.02511570702657216,
      "grad_norm": 0.12101856386944265,
      "learning_rate": 3.13712807244502e-07,
      "loss": 1.2684,
      "step": 194
    },
    {
      "epoch": 0.025245169433925622,
      "grad_norm": 0.11243921103968564,
      "learning_rate": 3.1532988357050454e-07,
      "loss": 1.1877,
      "step": 195
    },
    {
      "epoch": 0.025245169433925622,
      "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": 0.8841441869735718,
      "eval_runtime": 20.5026,
      "eval_samples_per_second": 2.146,
      "eval_steps_per_second": 0.146,
      "step": 195
    },
    {
      "epoch": 0.02537463184127909,
      "grad_norm": 0.13153422145217109,
      "learning_rate": 3.169469598965071e-07,
      "loss": 1.3876,
      "step": 196
    },
    {
      "epoch": 0.025504094248632554,
      "grad_norm": 0.11533583861936048,
      "learning_rate": 3.185640362225097e-07,
      "loss": 1.2552,
      "step": 197
    },
    {
      "epoch": 0.025633556655986017,
      "grad_norm": 0.11062734361075345,
      "learning_rate": 3.2018111254851234e-07,
      "loss": 1.1499,
      "step": 198
    },
    {
      "epoch": 0.02576301906333948,
      "grad_norm": 0.1354563501043128,
      "learning_rate": 3.217981888745149e-07,
      "loss": 1.356,
      "step": 199
    },
    {
      "epoch": 0.02589248147069295,
      "grad_norm": 0.08278919974184297,
      "learning_rate": 3.234152652005175e-07,
      "loss": 1.1153,
      "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": 0.8857421875,
      "eval_runtime": 19.9572,
      "eval_samples_per_second": 2.205,
      "eval_steps_per_second": 0.15,
      "step": 200
    },
    {
      "epoch": 0.026021943878046412,
      "grad_norm": 0.11903759093528415,
      "learning_rate": 3.250323415265201e-07,
      "loss": 1.2701,
      "step": 201
    },
    {
      "epoch": 0.026151406285399876,
      "grad_norm": 0.12103130152600584,
      "learning_rate": 3.2664941785252265e-07,
      "loss": 1.251,
      "step": 202
    },
    {
      "epoch": 0.026280868692753343,
      "grad_norm": 0.10801775922276957,
      "learning_rate": 3.2826649417852527e-07,
      "loss": 1.1543,
      "step": 203
    },
    {
      "epoch": 0.026410331100106807,
      "grad_norm": 0.11638552337459787,
      "learning_rate": 3.298835705045279e-07,
      "loss": 1.208,
      "step": 204
    },
    {
      "epoch": 0.02653979350746027,
      "grad_norm": 0.0811122622520925,
      "learning_rate": 3.315006468305304e-07,
      "loss": 1.1199,
      "step": 205
    },
    {
      "epoch": 0.02653979350746027,
      "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": 0.8850319385528564,
      "eval_runtime": 20.5214,
      "eval_samples_per_second": 2.144,
      "eval_steps_per_second": 0.146,
      "step": 205
    },
    {
      "epoch": 0.026669255914813735,
      "grad_norm": 0.09076755365153202,
      "learning_rate": 3.33117723156533e-07,
      "loss": 1.1664,
      "step": 206
    },
    {
      "epoch": 0.026798718322167202,
      "grad_norm": 0.11955223362510094,
      "learning_rate": 3.3473479948253564e-07,
      "loss": 1.2611,
      "step": 207
    },
    {
      "epoch": 0.026928180729520666,
      "grad_norm": 0.08204219207738607,
      "learning_rate": 3.3635187580853815e-07,
      "loss": 0.99,
      "step": 208
    },
    {
      "epoch": 0.02705764313687413,
      "grad_norm": 0.10618499770378295,
      "learning_rate": 3.3796895213454077e-07,
      "loss": 1.212,
      "step": 209
    },
    {
      "epoch": 0.027187105544227597,
      "grad_norm": 0.11600312241676758,
      "learning_rate": 3.395860284605434e-07,
      "loss": 1.168,
      "step": 210
    },
    {
      "epoch": 0.027187105544227597,
      "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": 20.3787,
      "eval_samples_per_second": 2.159,
      "eval_steps_per_second": 0.147,
      "step": 210
    },
    {
      "epoch": 0.02731656795158106,
      "grad_norm": 0.08373439068298424,
      "learning_rate": 3.412031047865459e-07,
      "loss": 1.0891,
      "step": 211
    },
    {
      "epoch": 0.027446030358934524,
      "grad_norm": 0.12160660736718698,
      "learning_rate": 3.428201811125485e-07,
      "loss": 1.2379,
      "step": 212
    },
    {
      "epoch": 0.027575492766287988,
      "grad_norm": 0.10337743787399474,
      "learning_rate": 3.4443725743855113e-07,
      "loss": 1.2758,
      "step": 213
    },
    {
      "epoch": 0.027704955173641455,
      "grad_norm": 0.1373851676011821,
      "learning_rate": 3.4605433376455375e-07,
      "loss": 1.3553,
      "step": 214
    },
    {
      "epoch": 0.02783441758099492,
      "grad_norm": 0.13438629956130926,
      "learning_rate": 3.4767141009055626e-07,
      "loss": 1.1908,
      "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": 20.1881,
      "eval_samples_per_second": 2.179,
      "eval_steps_per_second": 0.149,
      "step": 215
    },
    {
      "epoch": 0.027963879988348383,
      "grad_norm": 0.10745292595086053,
      "learning_rate": 3.492884864165589e-07,
      "loss": 1.2008,
      "step": 216
    },
    {
      "epoch": 0.028093342395701847,
      "grad_norm": 0.11157408593952817,
      "learning_rate": 3.509055627425615e-07,
      "loss": 1.2206,
      "step": 217
    },
    {
      "epoch": 0.028222804803055314,
      "grad_norm": 0.10105404688870254,
      "learning_rate": 3.5252263906856406e-07,
      "loss": 1.3103,
      "step": 218
    },
    {
      "epoch": 0.028352267210408778,
      "grad_norm": 0.12938206645001452,
      "learning_rate": 3.541397153945666e-07,
      "loss": 1.2733,
      "step": 219
    },
    {
      "epoch": 0.02848172961776224,
      "grad_norm": 0.10460488037581166,
      "learning_rate": 3.5575679172056924e-07,
      "loss": 1.231,
      "step": 220
    },
    {
      "epoch": 0.02848172961776224,
      "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": 0.8846768736839294,
      "eval_runtime": 19.7603,
      "eval_samples_per_second": 2.227,
      "eval_steps_per_second": 0.152,
      "step": 220
    },
    {
      "epoch": 0.02861119202511571,
      "grad_norm": 0.10777671612310356,
      "learning_rate": 3.573738680465718e-07,
      "loss": 1.2766,
      "step": 221
    },
    {
      "epoch": 0.028740654432469172,
      "grad_norm": 0.09901243763603829,
      "learning_rate": 3.5899094437257443e-07,
      "loss": 1.1389,
      "step": 222
    },
    {
      "epoch": 0.028870116839822636,
      "grad_norm": 0.12373789936968146,
      "learning_rate": 3.6060802069857704e-07,
      "loss": 1.2894,
      "step": 223
    },
    {
      "epoch": 0.0289995792471761,
      "grad_norm": 0.09617280603094286,
      "learning_rate": 3.6222509702457956e-07,
      "loss": 1.0998,
      "step": 224
    },
    {
      "epoch": 0.029129041654529567,
      "grad_norm": 0.15472164560219914,
      "learning_rate": 3.638421733505822e-07,
      "loss": 1.4431,
      "step": 225
    },
    {
      "epoch": 0.029129041654529567,
      "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": 0.8830788135528564,
      "eval_runtime": 20.1146,
      "eval_samples_per_second": 2.187,
      "eval_steps_per_second": 0.149,
      "step": 225
    },
    {
      "epoch": 0.02925850406188303,
      "grad_norm": 0.10144618278229844,
      "learning_rate": 3.654592496765848e-07,
      "loss": 1.1799,
      "step": 226
    },
    {
      "epoch": 0.029387966469236495,
      "grad_norm": 0.09706405668414601,
      "learning_rate": 3.670763260025874e-07,
      "loss": 1.1628,
      "step": 227
    },
    {
      "epoch": 0.02951742887658996,
      "grad_norm": 0.11436343544698634,
      "learning_rate": 3.686934023285899e-07,
      "loss": 1.2128,
      "step": 228
    },
    {
      "epoch": 0.029646891283943426,
      "grad_norm": 0.10546259240586074,
      "learning_rate": 3.7031047865459254e-07,
      "loss": 1.1284,
      "step": 229
    },
    {
      "epoch": 0.02977635369129689,
      "grad_norm": 0.11117769533271506,
      "learning_rate": 3.7192755498059516e-07,
      "loss": 1.144,
      "step": 230
    },
    {
      "epoch": 0.02977635369129689,
      "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": 0.8844105005264282,
      "eval_runtime": 19.9099,
      "eval_samples_per_second": 2.21,
      "eval_steps_per_second": 0.151,
      "step": 230
    },
    {
      "epoch": 0.029905816098650354,
      "grad_norm": 0.11451468164389629,
      "learning_rate": 3.7354463130659767e-07,
      "loss": 1.2469,
      "step": 231
    },
    {
      "epoch": 0.03003527850600382,
      "grad_norm": 0.10609658031655758,
      "learning_rate": 3.751617076326003e-07,
      "loss": 1.2277,
      "step": 232
    },
    {
      "epoch": 0.030164740913357285,
      "grad_norm": 0.12049612842778328,
      "learning_rate": 3.767787839586029e-07,
      "loss": 1.2811,
      "step": 233
    },
    {
      "epoch": 0.03029420332071075,
      "grad_norm": 0.13490155770266238,
      "learning_rate": 3.783958602846054e-07,
      "loss": 1.2555,
      "step": 234
    },
    {
      "epoch": 0.030423665728064212,
      "grad_norm": 0.1324991023543262,
      "learning_rate": 3.8001293661060803e-07,
      "loss": 1.2751,
      "step": 235
    },
    {
      "epoch": 0.030423665728064212,
      "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": 0.8852095007896423,
      "eval_runtime": 20.9302,
      "eval_samples_per_second": 2.102,
      "eval_steps_per_second": 0.143,
      "step": 235
    },
    {
      "epoch": 0.03055312813541768,
      "grad_norm": 0.12017361471649342,
      "learning_rate": 3.8163001293661065e-07,
      "loss": 1.2979,
      "step": 236
    },
    {
      "epoch": 0.030682590542771143,
      "grad_norm": 0.08968222154017091,
      "learning_rate": 3.832470892626132e-07,
      "loss": 1.0392,
      "step": 237
    },
    {
      "epoch": 0.030812052950124607,
      "grad_norm": 0.09879437216597636,
      "learning_rate": 3.848641655886158e-07,
      "loss": 1.2058,
      "step": 238
    },
    {
      "epoch": 0.03094151535747807,
      "grad_norm": 0.14719583865774397,
      "learning_rate": 3.864812419146184e-07,
      "loss": 1.3021,
      "step": 239
    },
    {
      "epoch": 0.031070977764831538,
      "grad_norm": 0.1154091168736403,
      "learning_rate": 3.8809831824062096e-07,
      "loss": 1.2295,
      "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": 0.8852983117103577,
      "eval_runtime": 20.8603,
      "eval_samples_per_second": 2.109,
      "eval_steps_per_second": 0.144,
      "step": 240
    },
    {
      "epoch": 0.031200440172185002,
      "grad_norm": 0.12346178141874305,
      "learning_rate": 3.897153945666236e-07,
      "loss": 1.2118,
      "step": 241
    },
    {
      "epoch": 0.031329902579538466,
      "grad_norm": 0.10915858215201577,
      "learning_rate": 3.9133247089262615e-07,
      "loss": 1.2987,
      "step": 242
    },
    {
      "epoch": 0.03145936498689193,
      "grad_norm": 0.08663365247897703,
      "learning_rate": 3.9294954721862877e-07,
      "loss": 1.1378,
      "step": 243
    },
    {
      "epoch": 0.03158882739424539,
      "grad_norm": 0.11360562795109866,
      "learning_rate": 3.9456662354463133e-07,
      "loss": 1.2601,
      "step": 244
    },
    {
      "epoch": 0.03171828980159886,
      "grad_norm": 0.12769616880540022,
      "learning_rate": 3.9618369987063395e-07,
      "loss": 1.2339,
      "step": 245
    },
    {
      "epoch": 0.03171828980159886,
      "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": 0.8843217492103577,
      "eval_runtime": 20.2732,
      "eval_samples_per_second": 2.17,
      "eval_steps_per_second": 0.148,
      "step": 245
    },
    {
      "epoch": 0.03184775220895233,
      "grad_norm": 0.11567984810047752,
      "learning_rate": 3.9780077619663657e-07,
      "loss": 1.1334,
      "step": 246
    },
    {
      "epoch": 0.03197721461630579,
      "grad_norm": 0.10757786185212588,
      "learning_rate": 3.994178525226391e-07,
      "loss": 1.2585,
      "step": 247
    },
    {
      "epoch": 0.032106677023659255,
      "grad_norm": 0.09316561358533522,
      "learning_rate": 4.010349288486417e-07,
      "loss": 1.1821,
      "step": 248
    },
    {
      "epoch": 0.03223613943101272,
      "grad_norm": 0.11271688188658921,
      "learning_rate": 4.026520051746443e-07,
      "loss": 1.2122,
      "step": 249
    },
    {
      "epoch": 0.03236560183836618,
      "grad_norm": 0.13900940985307111,
      "learning_rate": 4.042690815006468e-07,
      "loss": 1.2889,
      "step": 250
    },
    {
      "epoch": 0.03236560183836618,
      "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": 19.8895,
      "eval_samples_per_second": 2.212,
      "eval_steps_per_second": 0.151,
      "step": 250
    },
    {
      "epoch": 0.03249506424571965,
      "grad_norm": 0.13826235572884446,
      "learning_rate": 4.0588615782664944e-07,
      "loss": 1.3893,
      "step": 251
    },
    {
      "epoch": 0.03262452665307312,
      "grad_norm": 0.12722394091217623,
      "learning_rate": 4.0750323415265206e-07,
      "loss": 1.2499,
      "step": 252
    },
    {
      "epoch": 0.03275398906042658,
      "grad_norm": 0.12250617534474276,
      "learning_rate": 4.0912031047865457e-07,
      "loss": 1.1841,
      "step": 253
    },
    {
      "epoch": 0.032883451467780045,
      "grad_norm": 0.11657045408627538,
      "learning_rate": 4.107373868046572e-07,
      "loss": 1.3517,
      "step": 254
    },
    {
      "epoch": 0.033012913875133505,
      "grad_norm": 0.14302437286275582,
      "learning_rate": 4.123544631306598e-07,
      "loss": 1.3724,
      "step": 255
    },
    {
      "epoch": 0.033012913875133505,
      "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": 0.8819246888160706,
      "eval_runtime": 19.9728,
      "eval_samples_per_second": 2.203,
      "eval_steps_per_second": 0.15,
      "step": 255
    },
    {
      "epoch": 0.03314237628248697,
      "grad_norm": 0.10333178847962524,
      "learning_rate": 4.139715394566624e-07,
      "loss": 1.1357,
      "step": 256
    },
    {
      "epoch": 0.03327183868984044,
      "grad_norm": 0.13427215155583572,
      "learning_rate": 4.1558861578266494e-07,
      "loss": 1.2397,
      "step": 257
    },
    {
      "epoch": 0.0334013010971939,
      "grad_norm": 0.14750450106329455,
      "learning_rate": 4.1720569210866756e-07,
      "loss": 1.3051,
      "step": 258
    },
    {
      "epoch": 0.03353076350454737,
      "grad_norm": 0.1127463261196524,
      "learning_rate": 4.1882276843467017e-07,
      "loss": 1.2086,
      "step": 259
    },
    {
      "epoch": 0.033660225911900835,
      "grad_norm": 0.12259595276170604,
      "learning_rate": 4.2043984476067274e-07,
      "loss": 1.2356,
      "step": 260
    },
    {
      "epoch": 0.033660225911900835,
      "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": 0.8812144994735718,
      "eval_runtime": 20.455,
      "eval_samples_per_second": 2.151,
      "eval_steps_per_second": 0.147,
      "step": 260
    },
    {
      "epoch": 0.033789688319254295,
      "grad_norm": 0.1081119083202823,
      "learning_rate": 4.220569210866753e-07,
      "loss": 1.1211,
      "step": 261
    },
    {
      "epoch": 0.03391915072660776,
      "grad_norm": 0.10094590392184564,
      "learning_rate": 4.236739974126779e-07,
      "loss": 1.2247,
      "step": 262
    },
    {
      "epoch": 0.03404861313396123,
      "grad_norm": 0.1080198081042654,
      "learning_rate": 4.252910737386805e-07,
      "loss": 1.2558,
      "step": 263
    },
    {
      "epoch": 0.03417807554131469,
      "grad_norm": 0.13531463379525158,
      "learning_rate": 4.269081500646831e-07,
      "loss": 1.3289,
      "step": 264
    },
    {
      "epoch": 0.03430753794866816,
      "grad_norm": 0.1192584266080463,
      "learning_rate": 4.2852522639068567e-07,
      "loss": 1.2001,
      "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": 0.8850319385528564,
      "eval_runtime": 20.5954,
      "eval_samples_per_second": 2.136,
      "eval_steps_per_second": 0.146,
      "step": 265
    },
    {
      "epoch": 0.03443700035602162,
      "grad_norm": 0.1145599867616478,
      "learning_rate": 4.3014230271668823e-07,
      "loss": 1.2288,
      "step": 266
    },
    {
      "epoch": 0.034566462763375085,
      "grad_norm": 0.09940885766013918,
      "learning_rate": 4.3175937904269085e-07,
      "loss": 1.1411,
      "step": 267
    },
    {
      "epoch": 0.03469592517072855,
      "grad_norm": 0.09954823206953185,
      "learning_rate": 4.3337645536869347e-07,
      "loss": 1.0541,
      "step": 268
    },
    {
      "epoch": 0.03482538757808201,
      "grad_norm": 0.10547474234904207,
      "learning_rate": 4.34993531694696e-07,
      "loss": 1.1302,
      "step": 269
    },
    {
      "epoch": 0.03495484998543548,
      "grad_norm": 0.12278132560468799,
      "learning_rate": 4.366106080206986e-07,
      "loss": 1.2271,
      "step": 270
    },
    {
      "epoch": 0.03495484998543548,
      "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": 20.9243,
      "eval_samples_per_second": 2.103,
      "eval_steps_per_second": 0.143,
      "step": 270
    },
    {
      "epoch": 0.03508431239278895,
      "grad_norm": 0.1063299717563165,
      "learning_rate": 4.382276843467012e-07,
      "loss": 1.1443,
      "step": 271
    },
    {
      "epoch": 0.03521377480014241,
      "grad_norm": 0.11361829984004645,
      "learning_rate": 4.3984476067270383e-07,
      "loss": 1.1493,
      "step": 272
    },
    {
      "epoch": 0.035343237207495874,
      "grad_norm": 0.11955734299685128,
      "learning_rate": 4.4146183699870635e-07,
      "loss": 1.1901,
      "step": 273
    },
    {
      "epoch": 0.03547269961484934,
      "grad_norm": 0.12216036897092306,
      "learning_rate": 4.4307891332470896e-07,
      "loss": 1.2062,
      "step": 274
    },
    {
      "epoch": 0.0356021620222028,
      "grad_norm": 0.1225220247541619,
      "learning_rate": 4.446959896507116e-07,
      "loss": 1.1893,
      "step": 275
    },
    {
      "epoch": 0.0356021620222028,
      "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": 0.8813920617103577,
      "eval_runtime": 20.9117,
      "eval_samples_per_second": 2.104,
      "eval_steps_per_second": 0.143,
      "step": 275
    },
    {
      "epoch": 0.03573162442955627,
      "grad_norm": 0.1234135077874815,
      "learning_rate": 4.463130659767141e-07,
      "loss": 1.187,
      "step": 276
    },
    {
      "epoch": 0.03586108683690973,
      "grad_norm": 0.12288652221032347,
      "learning_rate": 4.479301423027167e-07,
      "loss": 1.1403,
      "step": 277
    },
    {
      "epoch": 0.0359905492442632,
      "grad_norm": 0.1144718848603907,
      "learning_rate": 4.4954721862871933e-07,
      "loss": 1.2238,
      "step": 278
    },
    {
      "epoch": 0.036120011651616664,
      "grad_norm": 0.09889256847516971,
      "learning_rate": 4.511642949547219e-07,
      "loss": 1.0851,
      "step": 279
    },
    {
      "epoch": 0.036249474058970124,
      "grad_norm": 0.11883464252724685,
      "learning_rate": 4.5278137128072446e-07,
      "loss": 1.1859,
      "step": 280
    },
    {
      "epoch": 0.036249474058970124,
      "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": 0.8815696239471436,
      "eval_runtime": 20.3389,
      "eval_samples_per_second": 2.163,
      "eval_steps_per_second": 0.148,
      "step": 280
    },
    {
      "epoch": 0.03637893646632359,
      "grad_norm": 0.14855001737514048,
      "learning_rate": 4.543984476067271e-07,
      "loss": 1.3018,
      "step": 281
    },
    {
      "epoch": 0.03650839887367706,
      "grad_norm": 0.12852886271477076,
      "learning_rate": 4.5601552393272964e-07,
      "loss": 1.3441,
      "step": 282
    },
    {
      "epoch": 0.03663786128103052,
      "grad_norm": 0.13007480482549597,
      "learning_rate": 4.5763260025873226e-07,
      "loss": 1.2822,
      "step": 283
    },
    {
      "epoch": 0.036767323688383986,
      "grad_norm": 0.10654644929617305,
      "learning_rate": 4.592496765847348e-07,
      "loss": 1.0834,
      "step": 284
    },
    {
      "epoch": 0.036896786095737454,
      "grad_norm": 0.13364843480137337,
      "learning_rate": 4.6086675291073744e-07,
      "loss": 1.334,
      "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": 0.8832563757896423,
      "eval_runtime": 20.3181,
      "eval_samples_per_second": 2.166,
      "eval_steps_per_second": 0.148,
      "step": 285
    },
    {
      "epoch": 0.037026248503090914,
      "grad_norm": 0.13620762747530885,
      "learning_rate": 4.6248382923674e-07,
      "loss": 1.345,
      "step": 286
    },
    {
      "epoch": 0.03715571091044438,
      "grad_norm": 0.11273663654921359,
      "learning_rate": 4.641009055627426e-07,
      "loss": 1.2205,
      "step": 287
    },
    {
      "epoch": 0.03728517331779784,
      "grad_norm": 0.10093582755660459,
      "learning_rate": 4.657179818887452e-07,
      "loss": 1.1046,
      "step": 288
    },
    {
      "epoch": 0.03741463572515131,
      "grad_norm": 0.1325632753430021,
      "learning_rate": 4.6733505821474775e-07,
      "loss": 1.2281,
      "step": 289
    },
    {
      "epoch": 0.037544098132504776,
      "grad_norm": 0.1037411072875527,
      "learning_rate": 4.6895213454075037e-07,
      "loss": 1.1975,
      "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": 0.8821022510528564,
      "eval_runtime": 20.9274,
      "eval_samples_per_second": 2.103,
      "eval_steps_per_second": 0.143,
      "step": 290
    },
    {
      "epoch": 0.037673560539858236,
      "grad_norm": 0.11317078187027814,
      "learning_rate": 4.70569210866753e-07,
      "loss": 1.2573,
      "step": 291
    },
    {
      "epoch": 0.037803022947211704,
      "grad_norm": 0.1183053737061677,
      "learning_rate": 4.721862871927555e-07,
      "loss": 1.2319,
      "step": 292
    },
    {
      "epoch": 0.03793248535456517,
      "grad_norm": 0.152218664206153,
      "learning_rate": 4.738033635187581e-07,
      "loss": 1.2998,
      "step": 293
    },
    {
      "epoch": 0.03806194776191863,
      "grad_norm": 0.162411224930765,
      "learning_rate": 4.7542043984476074e-07,
      "loss": 1.4022,
      "step": 294
    },
    {
      "epoch": 0.0381914101692721,
      "grad_norm": 0.1362183683223164,
      "learning_rate": 4.770375161707633e-07,
      "loss": 1.3798,
      "step": 295
    },
    {
      "epoch": 0.0381914101692721,
      "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": 0.8844105005264282,
      "eval_runtime": 19.8525,
      "eval_samples_per_second": 2.216,
      "eval_steps_per_second": 0.151,
      "step": 295
    },
    {
      "epoch": 0.038320872576625566,
      "grad_norm": 0.14353690782712983,
      "learning_rate": 4.786545924967659e-07,
      "loss": 1.3505,
      "step": 296
    },
    {
      "epoch": 0.038450334983979026,
      "grad_norm": 0.11742868930017476,
      "learning_rate": 4.802716688227685e-07,
      "loss": 1.1918,
      "step": 297
    },
    {
      "epoch": 0.03857979739133249,
      "grad_norm": 0.14704953096556558,
      "learning_rate": 4.81888745148771e-07,
      "loss": 1.4049,
      "step": 298
    },
    {
      "epoch": 0.038709259798685954,
      "grad_norm": 0.11689292841354257,
      "learning_rate": 4.835058214747737e-07,
      "loss": 1.1422,
      "step": 299
    },
    {
      "epoch": 0.03883872220603942,
      "grad_norm": 0.11530321004659914,
      "learning_rate": 4.851228978007762e-07,
      "loss": 1.1746,
      "step": 300
    },
    {
      "epoch": 0.03883872220603942,
      "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": 0.8817471861839294,
      "eval_runtime": 19.9146,
      "eval_samples_per_second": 2.209,
      "eval_steps_per_second": 0.151,
      "step": 300
    },
    {
      "epoch": 0.03896818461339289,
      "grad_norm": 0.16219891372779602,
      "learning_rate": 4.867399741267789e-07,
      "loss": 1.3815,
      "step": 301
    },
    {
      "epoch": 0.03909764702074635,
      "grad_norm": 0.1393181611514879,
      "learning_rate": 4.883570504527814e-07,
      "loss": 1.4614,
      "step": 302
    },
    {
      "epoch": 0.039227109428099816,
      "grad_norm": 0.12972369032688627,
      "learning_rate": 4.89974126778784e-07,
      "loss": 1.2748,
      "step": 303
    },
    {
      "epoch": 0.03935657183545328,
      "grad_norm": 0.11847461526751804,
      "learning_rate": 4.915912031047866e-07,
      "loss": 1.2454,
      "step": 304
    },
    {
      "epoch": 0.03948603424280674,
      "grad_norm": 0.14299357478859223,
      "learning_rate": 4.932082794307892e-07,
      "loss": 1.2394,
      "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": 0.8813920617103577,
      "eval_runtime": 20.8297,
      "eval_samples_per_second": 2.112,
      "eval_steps_per_second": 0.144,
      "step": 305
    },
    {
      "epoch": 0.03961549665016021,
      "grad_norm": 0.10454935073906405,
      "learning_rate": 4.948253557567917e-07,
      "loss": 1.094,
      "step": 306
    },
    {
      "epoch": 0.03974495905751368,
      "grad_norm": 0.15163270224275624,
      "learning_rate": 4.964424320827944e-07,
      "loss": 1.1851,
      "step": 307
    },
    {
      "epoch": 0.03987442146486714,
      "grad_norm": 0.12717388215780231,
      "learning_rate": 4.980595084087969e-07,
      "loss": 1.2357,
      "step": 308
    },
    {
      "epoch": 0.040003883872220605,
      "grad_norm": 0.142934532955926,
      "learning_rate": 4.996765847347995e-07,
      "loss": 1.2128,
      "step": 309
    },
    {
      "epoch": 0.040133346279574066,
      "grad_norm": 0.13368423590152395,
      "learning_rate": 5.012936610608021e-07,
      "loss": 1.2325,
      "step": 310
    },
    {
      "epoch": 0.040133346279574066,
      "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": 0.8809481263160706,
      "eval_runtime": 20.2757,
      "eval_samples_per_second": 2.17,
      "eval_steps_per_second": 0.148,
      "step": 310
    },
    {
      "epoch": 0.04026280868692753,
      "grad_norm": 0.1471047972265692,
      "learning_rate": 5.029107373868047e-07,
      "loss": 1.1696,
      "step": 311
    },
    {
      "epoch": 0.040392271094281,
      "grad_norm": 0.0991633237938451,
      "learning_rate": 5.045278137128073e-07,
      "loss": 1.073,
      "step": 312
    },
    {
      "epoch": 0.04052173350163446,
      "grad_norm": 0.15717113005455266,
      "learning_rate": 5.061448900388099e-07,
      "loss": 1.3071,
      "step": 313
    },
    {
      "epoch": 0.04065119590898793,
      "grad_norm": 0.09831486488786917,
      "learning_rate": 5.077619663648125e-07,
      "loss": 1.1298,
      "step": 314
    },
    {
      "epoch": 0.040780658316341395,
      "grad_norm": 0.1531802070792968,
      "learning_rate": 5.09379042690815e-07,
      "loss": 1.2966,
      "step": 315
    },
    {
      "epoch": 0.040780658316341395,
      "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": 0.8836115002632141,
      "eval_runtime": 20.4679,
      "eval_samples_per_second": 2.15,
      "eval_steps_per_second": 0.147,
      "step": 315
    },
    {
      "epoch": 0.040910120723694855,
      "grad_norm": 0.09717805776965392,
      "learning_rate": 5.109961190168177e-07,
      "loss": 1.2307,
      "step": 316
    },
    {
      "epoch": 0.04103958313104832,
      "grad_norm": 0.09678226995270714,
      "learning_rate": 5.126131953428203e-07,
      "loss": 1.0577,
      "step": 317
    },
    {
      "epoch": 0.04116904553840179,
      "grad_norm": 0.09860290836300747,
      "learning_rate": 5.142302716688228e-07,
      "loss": 1.084,
      "step": 318
    },
    {
      "epoch": 0.04129850794575525,
      "grad_norm": 0.1105959019410525,
      "learning_rate": 5.158473479948254e-07,
      "loss": 1.0696,
      "step": 319
    },
    {
      "epoch": 0.04142797035310872,
      "grad_norm": 0.11452080292259209,
      "learning_rate": 5.174644243208281e-07,
      "loss": 1.0898,
      "step": 320
    },
    {
      "epoch": 0.04142797035310872,
      "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": 0.8822798132896423,
      "eval_runtime": 20.4222,
      "eval_samples_per_second": 2.155,
      "eval_steps_per_second": 0.147,
      "step": 320
    },
    {
      "epoch": 0.04155743276046218,
      "grad_norm": 0.10595506003969767,
      "learning_rate": 5.190815006468305e-07,
      "loss": 1.1566,
      "step": 321
    },
    {
      "epoch": 0.041686895167815645,
      "grad_norm": 0.12625343230519284,
      "learning_rate": 5.206985769728332e-07,
      "loss": 1.225,
      "step": 322
    },
    {
      "epoch": 0.04181635757516911,
      "grad_norm": 0.09053038076123116,
      "learning_rate": 5.223156532988358e-07,
      "loss": 1.1772,
      "step": 323
    },
    {
      "epoch": 0.04194581998252257,
      "grad_norm": 0.1451350070381464,
      "learning_rate": 5.239327296248383e-07,
      "loss": 1.2603,
      "step": 324
    },
    {
      "epoch": 0.04207528238987604,
      "grad_norm": 0.12810394869021174,
      "learning_rate": 5.255498059508409e-07,
      "loss": 1.2286,
      "step": 325
    },
    {
      "epoch": 0.04207528238987604,
      "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": 0.8813032507896423,
      "eval_runtime": 20.0637,
      "eval_samples_per_second": 2.193,
      "eval_steps_per_second": 0.15,
      "step": 325
    },
    {
      "epoch": 0.04220474479722951,
      "grad_norm": 0.13139674922314803,
      "learning_rate": 5.271668822768436e-07,
      "loss": 1.3004,
      "step": 326
    },
    {
      "epoch": 0.04233420720458297,
      "grad_norm": 0.1302589365045921,
      "learning_rate": 5.28783958602846e-07,
      "loss": 1.2689,
      "step": 327
    },
    {
      "epoch": 0.042463669611936435,
      "grad_norm": 0.12726417927327388,
      "learning_rate": 5.304010349288487e-07,
      "loss": 1.2448,
      "step": 328
    },
    {
      "epoch": 0.0425931320192899,
      "grad_norm": 0.07870398369107665,
      "learning_rate": 5.320181112548512e-07,
      "loss": 1.1431,
      "step": 329
    },
    {
      "epoch": 0.04272259442664336,
      "grad_norm": 0.12362368361822505,
      "learning_rate": 5.336351875808539e-07,
      "loss": 1.2786,
      "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": 0.8805042505264282,
      "eval_runtime": 19.5982,
      "eval_samples_per_second": 2.245,
      "eval_steps_per_second": 0.153,
      "step": 330
    },
    {
      "epoch": 0.04285205683399683,
      "grad_norm": 0.11612035242394682,
      "learning_rate": 5.352522639068565e-07,
      "loss": 1.2818,
      "step": 331
    },
    {
      "epoch": 0.04298151924135029,
      "grad_norm": 0.12388217133007283,
      "learning_rate": 5.36869340232859e-07,
      "loss": 1.1433,
      "step": 332
    },
    {
      "epoch": 0.04311098164870376,
      "grad_norm": 0.1343152929036638,
      "learning_rate": 5.384864165588616e-07,
      "loss": 1.2712,
      "step": 333
    },
    {
      "epoch": 0.043240444056057224,
      "grad_norm": 0.1065797417446008,
      "learning_rate": 5.401034928848642e-07,
      "loss": 1.2505,
      "step": 334
    },
    {
      "epoch": 0.043369906463410685,
      "grad_norm": 0.10643146140628564,
      "learning_rate": 5.417205692108668e-07,
      "loss": 1.1086,
      "step": 335
    },
    {
      "epoch": 0.043369906463410685,
      "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": 0.8820134997367859,
      "eval_runtime": 20.2813,
      "eval_samples_per_second": 2.169,
      "eval_steps_per_second": 0.148,
      "step": 335
    },
    {
      "epoch": 0.04349936887076415,
      "grad_norm": 0.1302056864087753,
      "learning_rate": 5.433376455368694e-07,
      "loss": 1.2192,
      "step": 336
    },
    {
      "epoch": 0.04362883127811762,
      "grad_norm": 0.1048013229557253,
      "learning_rate": 5.44954721862872e-07,
      "loss": 1.1385,
      "step": 337
    },
    {
      "epoch": 0.04375829368547108,
      "grad_norm": 0.12518448559534456,
      "learning_rate": 5.465717981888745e-07,
      "loss": 1.4475,
      "step": 338
    },
    {
      "epoch": 0.04388775609282455,
      "grad_norm": 0.09027833916000308,
      "learning_rate": 5.481888745148772e-07,
      "loss": 1.0989,
      "step": 339
    },
    {
      "epoch": 0.044017218500178014,
      "grad_norm": 0.08855964485419343,
      "learning_rate": 5.498059508408797e-07,
      "loss": 1.1508,
      "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": 0.8815696239471436,
      "eval_runtime": 20.7327,
      "eval_samples_per_second": 2.122,
      "eval_steps_per_second": 0.145,
      "step": 340
    },
    {
      "epoch": 0.044146680907531474,
      "grad_norm": 0.0865851993394446,
      "learning_rate": 5.514230271668823e-07,
      "loss": 1.1696,
      "step": 341
    },
    {
      "epoch": 0.04427614331488494,
      "grad_norm": 0.12897318715011416,
      "learning_rate": 5.530401034928849e-07,
      "loss": 1.3262,
      "step": 342
    },
    {
      "epoch": 0.0444056057222384,
      "grad_norm": 0.09660369608524746,
      "learning_rate": 5.546571798188876e-07,
      "loss": 1.1427,
      "step": 343
    },
    {
      "epoch": 0.04453506812959187,
      "grad_norm": 0.0736294215038434,
      "learning_rate": 5.5627425614489e-07,
      "loss": 1.066,
      "step": 344
    },
    {
      "epoch": 0.044664530536945336,
      "grad_norm": 0.17120448051393924,
      "learning_rate": 5.578913324708927e-07,
      "loss": 1.4232,
      "step": 345
    },
    {
      "epoch": 0.044664530536945336,
      "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": 0.8800603747367859,
      "eval_runtime": 19.7684,
      "eval_samples_per_second": 2.226,
      "eval_steps_per_second": 0.152,
      "step": 345
    },
    {
      "epoch": 0.0447939929442988,
      "grad_norm": 0.113777497929062,
      "learning_rate": 5.595084087968953e-07,
      "loss": 1.2468,
      "step": 346
    },
    {
      "epoch": 0.044923455351652264,
      "grad_norm": 0.1328291522395543,
      "learning_rate": 5.611254851228978e-07,
      "loss": 1.2592,
      "step": 347
    },
    {
      "epoch": 0.04505291775900573,
      "grad_norm": 0.11092466945061226,
      "learning_rate": 5.627425614489004e-07,
      "loss": 1.1082,
      "step": 348
    },
    {
      "epoch": 0.04518238016635919,
      "grad_norm": 0.13034881208321833,
      "learning_rate": 5.643596377749031e-07,
      "loss": 1.138,
      "step": 349
    },
    {
      "epoch": 0.04531184257371266,
      "grad_norm": 0.1574610938544109,
      "learning_rate": 5.659767141009055e-07,
      "loss": 1.3915,
      "step": 350
    },
    {
      "epoch": 0.04531184257371266,
      "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": 20.2778,
      "eval_samples_per_second": 2.17,
      "eval_steps_per_second": 0.148,
      "step": 350
    },
    {
      "epoch": 0.045441304981066126,
      "grad_norm": 0.13761873261758883,
      "learning_rate": 5.675937904269082e-07,
      "loss": 1.2703,
      "step": 351
    },
    {
      "epoch": 0.045570767388419586,
      "grad_norm": 0.1411450278923194,
      "learning_rate": 5.692108667529108e-07,
      "loss": 1.3245,
      "step": 352
    },
    {
      "epoch": 0.045700229795773054,
      "grad_norm": 0.11294136868516831,
      "learning_rate": 5.708279430789133e-07,
      "loss": 1.1745,
      "step": 353
    },
    {
      "epoch": 0.045829692203126514,
      "grad_norm": 0.14659842211253132,
      "learning_rate": 5.72445019404916e-07,
      "loss": 1.2466,
      "step": 354
    },
    {
      "epoch": 0.04595915461047998,
      "grad_norm": 0.09531640544352693,
      "learning_rate": 5.740620957309186e-07,
      "loss": 1.0896,
      "step": 355
    },
    {
      "epoch": 0.04595915461047998,
      "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": 0.8812144994735718,
      "eval_runtime": 19.8088,
      "eval_samples_per_second": 2.221,
      "eval_steps_per_second": 0.151,
      "step": 355
    },
    {
      "epoch": 0.04608861701783345,
      "grad_norm": 0.09844193801597724,
      "learning_rate": 5.756791720569211e-07,
      "loss": 1.0505,
      "step": 356
    },
    {
      "epoch": 0.04621807942518691,
      "grad_norm": 0.09358701051817164,
      "learning_rate": 5.772962483829237e-07,
      "loss": 1.0928,
      "step": 357
    },
    {
      "epoch": 0.046347541832540376,
      "grad_norm": 0.10694790932927545,
      "learning_rate": 5.789133247089264e-07,
      "loss": 1.0964,
      "step": 358
    },
    {
      "epoch": 0.04647700423989384,
      "grad_norm": 0.11122993907282537,
      "learning_rate": 5.805304010349289e-07,
      "loss": 1.3213,
      "step": 359
    },
    {
      "epoch": 0.046606466647247304,
      "grad_norm": 0.1518149515081859,
      "learning_rate": 5.821474773609315e-07,
      "loss": 1.327,
      "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": 0.8815696239471436,
      "eval_runtime": 19.7736,
      "eval_samples_per_second": 2.225,
      "eval_steps_per_second": 0.152,
      "step": 360
    },
    {
      "epoch": 0.04673592905460077,
      "grad_norm": 0.13292550832310077,
      "learning_rate": 5.837645536869341e-07,
      "loss": 1.2732,
      "step": 361
    },
    {
      "epoch": 0.04686539146195424,
      "grad_norm": 0.1284090371719686,
      "learning_rate": 5.853816300129367e-07,
      "loss": 1.2004,
      "step": 362
    },
    {
      "epoch": 0.0469948538693077,
      "grad_norm": 0.13113246028845052,
      "learning_rate": 5.869987063389392e-07,
      "loss": 1.2311,
      "step": 363
    },
    {
      "epoch": 0.047124316276661166,
      "grad_norm": 0.07766779807017206,
      "learning_rate": 5.886157826649419e-07,
      "loss": 1.0067,
      "step": 364
    },
    {
      "epoch": 0.047253778684014626,
      "grad_norm": 0.15900041244253268,
      "learning_rate": 5.902328589909444e-07,
      "loss": 1.3251,
      "step": 365
    },
    {
      "epoch": 0.047253778684014626,
      "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": 0.8805042505264282,
      "eval_runtime": 19.877,
      "eval_samples_per_second": 2.214,
      "eval_steps_per_second": 0.151,
      "step": 365
    },
    {
      "epoch": 0.04738324109136809,
      "grad_norm": 0.10570463312201128,
      "learning_rate": 5.91849935316947e-07,
      "loss": 1.0717,
      "step": 366
    },
    {
      "epoch": 0.04751270349872156,
      "grad_norm": 0.15500149932529037,
      "learning_rate": 5.934670116429496e-07,
      "loss": 1.2594,
      "step": 367
    },
    {
      "epoch": 0.04764216590607502,
      "grad_norm": 0.14845254317126025,
      "learning_rate": 5.950840879689522e-07,
      "loss": 1.2206,
      "step": 368
    },
    {
      "epoch": 0.04777162831342849,
      "grad_norm": 0.14842490613414192,
      "learning_rate": 5.967011642949547e-07,
      "loss": 1.2922,
      "step": 369
    },
    {
      "epoch": 0.047901090720781955,
      "grad_norm": 0.14815101313369738,
      "learning_rate": 5.983182406209574e-07,
      "loss": 1.2969,
      "step": 370
    },
    {
      "epoch": 0.047901090720781955,
      "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": 0.8797940611839294,
      "eval_runtime": 21.1049,
      "eval_samples_per_second": 2.085,
      "eval_steps_per_second": 0.142,
      "step": 370
    },
    {
      "epoch": 0.048030553128135416,
      "grad_norm": 0.09014402574016521,
      "learning_rate": 5.999353169469599e-07,
      "loss": 1.1739,
      "step": 371
    },
    {
      "epoch": 0.04816001553548888,
      "grad_norm": 0.15028461796070391,
      "learning_rate": 6.015523932729626e-07,
      "loss": 1.278,
      "step": 372
    },
    {
      "epoch": 0.04828947794284235,
      "grad_norm": 0.11005277253293279,
      "learning_rate": 6.031694695989651e-07,
      "loss": 1.2544,
      "step": 373
    },
    {
      "epoch": 0.04841894035019581,
      "grad_norm": 0.1554567308302687,
      "learning_rate": 6.047865459249677e-07,
      "loss": 1.2198,
      "step": 374
    },
    {
      "epoch": 0.04854840275754928,
      "grad_norm": 0.14678823627036489,
      "learning_rate": 6.064036222509703e-07,
      "loss": 1.3258,
      "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": 0.8821022510528564,
      "eval_runtime": 20.292,
      "eval_samples_per_second": 2.168,
      "eval_steps_per_second": 0.148,
      "step": 375
    },
    {
      "epoch": 0.04867786516490274,
      "grad_norm": 0.13299567128499856,
      "learning_rate": 6.080206985769729e-07,
      "loss": 1.1264,
      "step": 376
    },
    {
      "epoch": 0.048807327572256205,
      "grad_norm": 0.15433843982274714,
      "learning_rate": 6.096377749029755e-07,
      "loss": 1.2864,
      "step": 377
    },
    {
      "epoch": 0.04893678997960967,
      "grad_norm": 0.11992490712238417,
      "learning_rate": 6.112548512289781e-07,
      "loss": 1.1335,
      "step": 378
    },
    {
      "epoch": 0.04906625238696313,
      "grad_norm": 0.17686724801930648,
      "learning_rate": 6.128719275549807e-07,
      "loss": 1.2768,
      "step": 379
    },
    {
      "epoch": 0.0491957147943166,
      "grad_norm": 0.14373394534337985,
      "learning_rate": 6.144890038809832e-07,
      "loss": 1.2881,
      "step": 380
    },
    {
      "epoch": 0.0491957147943166,
      "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": 0.8779296875,
      "eval_runtime": 20.735,
      "eval_samples_per_second": 2.122,
      "eval_steps_per_second": 0.145,
      "step": 380
    },
    {
      "epoch": 0.04932517720167007,
      "grad_norm": 0.146918199055568,
      "learning_rate": 6.161060802069859e-07,
      "loss": 1.2343,
      "step": 381
    },
    {
      "epoch": 0.04945463960902353,
      "grad_norm": 0.1394916609074134,
      "learning_rate": 6.177231565329884e-07,
      "loss": 1.2329,
      "step": 382
    },
    {
      "epoch": 0.049584102016376995,
      "grad_norm": 0.14239529996801026,
      "learning_rate": 6.19340232858991e-07,
      "loss": 1.2391,
      "step": 383
    },
    {
      "epoch": 0.04971356442373046,
      "grad_norm": 0.125169912557683,
      "learning_rate": 6.209573091849936e-07,
      "loss": 1.174,
      "step": 384
    },
    {
      "epoch": 0.04984302683108392,
      "grad_norm": 0.13522945757896307,
      "learning_rate": 6.225743855109962e-07,
      "loss": 1.2832,
      "step": 385
    },
    {
      "epoch": 0.04984302683108392,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 Neg": 0.3673469387755102,
      "eval_PRM NPV": 0.24,
      "eval_PRM Precision": 0.8387096774193549,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8788174986839294,
      "eval_runtime": 20.892,
      "eval_samples_per_second": 2.106,
      "eval_steps_per_second": 0.144,
      "step": 385
    },
    {
      "epoch": 0.04997248923843739,
      "grad_norm": 0.14307903319718435,
      "learning_rate": 6.241914618369987e-07,
      "loss": 1.2855,
      "step": 386
    },
    {
      "epoch": 0.05010195164579085,
      "grad_norm": 0.13836819746174128,
      "learning_rate": 6.258085381630014e-07,
      "loss": 1.1495,
      "step": 387
    },
    {
      "epoch": 0.05023141405314432,
      "grad_norm": 0.13258405474821977,
      "learning_rate": 6.27425614489004e-07,
      "loss": 1.132,
      "step": 388
    },
    {
      "epoch": 0.050360876460497785,
      "grad_norm": 0.15948763660832313,
      "learning_rate": 6.290426908150066e-07,
      "loss": 1.3452,
      "step": 389
    },
    {
      "epoch": 0.050490338867851245,
      "grad_norm": 0.11548707492370487,
      "learning_rate": 6.306597671410091e-07,
      "loss": 1.1374,
      "step": 390
    },
    {
      "epoch": 0.050490338867851245,
      "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": 0.8797940611839294,
      "eval_runtime": 20.3188,
      "eval_samples_per_second": 2.165,
      "eval_steps_per_second": 0.148,
      "step": 390
    },
    {
      "epoch": 0.05061980127520471,
      "grad_norm": 0.144073559522834,
      "learning_rate": 6.322768434670116e-07,
      "loss": 1.3301,
      "step": 391
    },
    {
      "epoch": 0.05074926368255818,
      "grad_norm": 0.13283156565189616,
      "learning_rate": 6.338939197930142e-07,
      "loss": 1.1193,
      "step": 392
    },
    {
      "epoch": 0.05087872608991164,
      "grad_norm": 0.14398974067607204,
      "learning_rate": 6.355109961190169e-07,
      "loss": 1.2054,
      "step": 393
    },
    {
      "epoch": 0.05100818849726511,
      "grad_norm": 0.07340493188732004,
      "learning_rate": 6.371280724450194e-07,
      "loss": 0.918,
      "step": 394
    },
    {
      "epoch": 0.051137650904618574,
      "grad_norm": 0.13886635801019834,
      "learning_rate": 6.387451487710221e-07,
      "loss": 1.2516,
      "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": 0.8795276880264282,
      "eval_runtime": 20.6852,
      "eval_samples_per_second": 2.127,
      "eval_steps_per_second": 0.145,
      "step": 395
    },
    {
      "epoch": 0.051267113311972035,
      "grad_norm": 0.15417984316519673,
      "learning_rate": 6.403622250970247e-07,
      "loss": 1.2496,
      "step": 396
    },
    {
      "epoch": 0.0513965757193255,
      "grad_norm": 0.16030562833794265,
      "learning_rate": 6.419793014230271e-07,
      "loss": 1.4047,
      "step": 397
    },
    {
      "epoch": 0.05152603812667896,
      "grad_norm": 0.14054149622201578,
      "learning_rate": 6.435963777490298e-07,
      "loss": 1.2756,
      "step": 398
    },
    {
      "epoch": 0.05165550053403243,
      "grad_norm": 0.15046096364945252,
      "learning_rate": 6.452134540750324e-07,
      "loss": 1.2437,
      "step": 399
    },
    {
      "epoch": 0.0517849629413859,
      "grad_norm": 0.13401303816925894,
      "learning_rate": 6.46830530401035e-07,
      "loss": 1.1588,
      "step": 400
    },
    {
      "epoch": 0.0517849629413859,
      "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": 0.8780184388160706,
      "eval_runtime": 20.4887,
      "eval_samples_per_second": 2.148,
      "eval_steps_per_second": 0.146,
      "step": 400
    },
    {
      "epoch": 0.05191442534873936,
      "grad_norm": 0.10646339244494968,
      "learning_rate": 6.484476067270376e-07,
      "loss": 1.0908,
      "step": 401
    },
    {
      "epoch": 0.052043887756092824,
      "grad_norm": 0.12957481726997483,
      "learning_rate": 6.500646830530402e-07,
      "loss": 1.13,
      "step": 402
    },
    {
      "epoch": 0.05217335016344629,
      "grad_norm": 0.12895312523702518,
      "learning_rate": 6.516817593790426e-07,
      "loss": 1.1854,
      "step": 403
    },
    {
      "epoch": 0.05230281257079975,
      "grad_norm": 0.14055262206005995,
      "learning_rate": 6.532988357050453e-07,
      "loss": 1.185,
      "step": 404
    },
    {
      "epoch": 0.05243227497815322,
      "grad_norm": 0.1626415648007422,
      "learning_rate": 6.549159120310479e-07,
      "loss": 1.2705,
      "step": 405
    },
    {
      "epoch": 0.05243227497815322,
      "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": 0.8762428760528564,
      "eval_runtime": 20.5328,
      "eval_samples_per_second": 2.143,
      "eval_steps_per_second": 0.146,
      "step": 405
    },
    {
      "epoch": 0.052561737385506686,
      "grad_norm": 0.14141862679447156,
      "learning_rate": 6.565329883570505e-07,
      "loss": 1.3602,
      "step": 406
    },
    {
      "epoch": 0.05269119979286015,
      "grad_norm": 0.12262418841830709,
      "learning_rate": 6.581500646830531e-07,
      "loss": 1.1458,
      "step": 407
    },
    {
      "epoch": 0.052820662200213614,
      "grad_norm": 0.12389340578546944,
      "learning_rate": 6.597671410090558e-07,
      "loss": 1.2595,
      "step": 408
    },
    {
      "epoch": 0.05295012460756708,
      "grad_norm": 0.13691840105627476,
      "learning_rate": 6.613842173350583e-07,
      "loss": 1.1362,
      "step": 409
    },
    {
      "epoch": 0.05307958701492054,
      "grad_norm": 0.14869689357480315,
      "learning_rate": 6.630012936610608e-07,
      "loss": 1.2236,
      "step": 410
    },
    {
      "epoch": 0.05307958701492054,
      "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": 0.8762428760528564,
      "eval_runtime": 20.5261,
      "eval_samples_per_second": 2.144,
      "eval_steps_per_second": 0.146,
      "step": 410
    },
    {
      "epoch": 0.05320904942227401,
      "grad_norm": 0.12190313999854462,
      "learning_rate": 6.646183699870634e-07,
      "loss": 1.0384,
      "step": 411
    },
    {
      "epoch": 0.05333851182962747,
      "grad_norm": 0.13693348731156332,
      "learning_rate": 6.66235446313066e-07,
      "loss": 1.1571,
      "step": 412
    },
    {
      "epoch": 0.053467974236980936,
      "grad_norm": 0.1332508796066339,
      "learning_rate": 6.678525226390686e-07,
      "loss": 1.1186,
      "step": 413
    },
    {
      "epoch": 0.053597436644334404,
      "grad_norm": 0.12570462846313513,
      "learning_rate": 6.694695989650713e-07,
      "loss": 1.2213,
      "step": 414
    },
    {
      "epoch": 0.053726899051687864,
      "grad_norm": 0.18289109172787454,
      "learning_rate": 6.710866752910738e-07,
      "loss": 1.4254,
      "step": 415
    },
    {
      "epoch": 0.053726899051687864,
      "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": 0.8790838122367859,
      "eval_runtime": 20.1517,
      "eval_samples_per_second": 2.183,
      "eval_steps_per_second": 0.149,
      "step": 415
    },
    {
      "epoch": 0.05385636145904133,
      "grad_norm": 0.1622383451098016,
      "learning_rate": 6.727037516170763e-07,
      "loss": 1.4027,
      "step": 416
    },
    {
      "epoch": 0.0539858238663948,
      "grad_norm": 0.13062308569001535,
      "learning_rate": 6.74320827943079e-07,
      "loss": 1.1508,
      "step": 417
    },
    {
      "epoch": 0.05411528627374826,
      "grad_norm": 0.1152142702124476,
      "learning_rate": 6.759379042690815e-07,
      "loss": 1.0941,
      "step": 418
    },
    {
      "epoch": 0.054244748681101726,
      "grad_norm": 0.12292891950791789,
      "learning_rate": 6.775549805950841e-07,
      "loss": 1.1713,
      "step": 419
    },
    {
      "epoch": 0.05437421108845519,
      "grad_norm": 0.1330646694208271,
      "learning_rate": 6.791720569210868e-07,
      "loss": 1.264,
      "step": 420
    },
    {
      "epoch": 0.05437421108845519,
      "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": 0.8774858117103577,
      "eval_runtime": 19.5646,
      "eval_samples_per_second": 2.249,
      "eval_steps_per_second": 0.153,
      "step": 420
    },
    {
      "epoch": 0.054503673495808654,
      "grad_norm": 0.16645775159548504,
      "learning_rate": 6.807891332470893e-07,
      "loss": 1.3765,
      "step": 421
    },
    {
      "epoch": 0.05463313590316212,
      "grad_norm": 0.14400146098293068,
      "learning_rate": 6.824062095730918e-07,
      "loss": 1.303,
      "step": 422
    },
    {
      "epoch": 0.05476259831051558,
      "grad_norm": 0.10572663924299523,
      "learning_rate": 6.840232858990945e-07,
      "loss": 1.054,
      "step": 423
    },
    {
      "epoch": 0.05489206071786905,
      "grad_norm": 0.15326005790722444,
      "learning_rate": 6.85640362225097e-07,
      "loss": 1.2318,
      "step": 424
    },
    {
      "epoch": 0.055021523125222516,
      "grad_norm": 0.18391472032028477,
      "learning_rate": 6.872574385510997e-07,
      "loss": 1.2563,
      "step": 425
    },
    {
      "epoch": 0.055021523125222516,
      "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": 0.8776633739471436,
      "eval_runtime": 20.1954,
      "eval_samples_per_second": 2.179,
      "eval_steps_per_second": 0.149,
      "step": 425
    },
    {
      "epoch": 0.055150985532575976,
      "grad_norm": 0.1431641805343948,
      "learning_rate": 6.888745148771023e-07,
      "loss": 1.2361,
      "step": 426
    },
    {
      "epoch": 0.05528044793992944,
      "grad_norm": 0.11467355379889112,
      "learning_rate": 6.904915912031049e-07,
      "loss": 1.0933,
      "step": 427
    },
    {
      "epoch": 0.05540991034728291,
      "grad_norm": 0.18944597951569833,
      "learning_rate": 6.921086675291075e-07,
      "loss": 1.3757,
      "step": 428
    },
    {
      "epoch": 0.05553937275463637,
      "grad_norm": 0.13394735876815805,
      "learning_rate": 6.9372574385511e-07,
      "loss": 1.2459,
      "step": 429
    },
    {
      "epoch": 0.05566883516198984,
      "grad_norm": 0.10494145628880099,
      "learning_rate": 6.953428201811125e-07,
      "loss": 1.1139,
      "step": 430
    },
    {
      "epoch": 0.05566883516198984,
      "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": 0.8748224377632141,
      "eval_runtime": 20.1686,
      "eval_samples_per_second": 2.182,
      "eval_steps_per_second": 0.149,
      "step": 430
    },
    {
      "epoch": 0.055798297569343305,
      "grad_norm": 0.16425433015172597,
      "learning_rate": 6.969598965071152e-07,
      "loss": 1.2411,
      "step": 431
    },
    {
      "epoch": 0.055927759976696766,
      "grad_norm": 0.10752198074041201,
      "learning_rate": 6.985769728331178e-07,
      "loss": 1.0898,
      "step": 432
    },
    {
      "epoch": 0.05605722238405023,
      "grad_norm": 0.1651274315236705,
      "learning_rate": 7.001940491591204e-07,
      "loss": 1.3148,
      "step": 433
    },
    {
      "epoch": 0.05618668479140369,
      "grad_norm": 0.17009206014804526,
      "learning_rate": 7.01811125485123e-07,
      "loss": 1.2966,
      "step": 434
    },
    {
      "epoch": 0.05631614719875716,
      "grad_norm": 0.16381437351205527,
      "learning_rate": 7.034282018111255e-07,
      "loss": 1.2017,
      "step": 435
    },
    {
      "epoch": 0.05631614719875716,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 Neg": 0.3673469387755102,
      "eval_PRM NPV": 0.24,
      "eval_PRM Precision": 0.8387096774193549,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8767755627632141,
      "eval_runtime": 20.2346,
      "eval_samples_per_second": 2.174,
      "eval_steps_per_second": 0.148,
      "step": 435
    },
    {
      "epoch": 0.05644560960611063,
      "grad_norm": 0.16480158535186282,
      "learning_rate": 7.050452781371281e-07,
      "loss": 1.3282,
      "step": 436
    },
    {
      "epoch": 0.05657507201346409,
      "grad_norm": 0.15767547474888394,
      "learning_rate": 7.066623544631307e-07,
      "loss": 1.205,
      "step": 437
    },
    {
      "epoch": 0.056704534420817555,
      "grad_norm": 0.1781660657830749,
      "learning_rate": 7.082794307891333e-07,
      "loss": 1.3179,
      "step": 438
    },
    {
      "epoch": 0.05683399682817102,
      "grad_norm": 0.18620638329229608,
      "learning_rate": 7.098965071151359e-07,
      "loss": 1.4708,
      "step": 439
    },
    {
      "epoch": 0.05696345923552448,
      "grad_norm": 0.13380199637663936,
      "learning_rate": 7.115135834411385e-07,
      "loss": 1.142,
      "step": 440
    },
    {
      "epoch": 0.05696345923552448,
      "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": 0.8758878111839294,
      "eval_runtime": 20.2149,
      "eval_samples_per_second": 2.177,
      "eval_steps_per_second": 0.148,
      "step": 440
    },
    {
      "epoch": 0.05709292164287795,
      "grad_norm": 0.1469692237472956,
      "learning_rate": 7.131306597671412e-07,
      "loss": 1.2301,
      "step": 441
    },
    {
      "epoch": 0.05722238405023142,
      "grad_norm": 0.1512797018585946,
      "learning_rate": 7.147477360931436e-07,
      "loss": 1.1812,
      "step": 442
    },
    {
      "epoch": 0.05735184645758488,
      "grad_norm": 0.161925350794174,
      "learning_rate": 7.163648124191462e-07,
      "loss": 1.2089,
      "step": 443
    },
    {
      "epoch": 0.057481308864938345,
      "grad_norm": 0.1435075899868018,
      "learning_rate": 7.179818887451489e-07,
      "loss": 1.2444,
      "step": 444
    },
    {
      "epoch": 0.057610771272291805,
      "grad_norm": 0.17247849683048727,
      "learning_rate": 7.195989650711514e-07,
      "loss": 1.3109,
      "step": 445
    },
    {
      "epoch": 0.057610771272291805,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 Neg": 0.3673469387755102,
      "eval_PRM NPV": 0.24,
      "eval_PRM Precision": 0.8387096774193549,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8755326867103577,
      "eval_runtime": 19.8136,
      "eval_samples_per_second": 2.221,
      "eval_steps_per_second": 0.151,
      "step": 445
    },
    {
      "epoch": 0.05774023367964527,
      "grad_norm": 0.16678729232147582,
      "learning_rate": 7.212160413971541e-07,
      "loss": 1.3158,
      "step": 446
    },
    {
      "epoch": 0.05786969608699874,
      "grad_norm": 0.08647542766665041,
      "learning_rate": 7.228331177231567e-07,
      "loss": 1.0768,
      "step": 447
    },
    {
      "epoch": 0.0579991584943522,
      "grad_norm": 0.11466727649068152,
      "learning_rate": 7.244501940491591e-07,
      "loss": 1.1089,
      "step": 448
    },
    {
      "epoch": 0.05812862090170567,
      "grad_norm": 0.1507587636419148,
      "learning_rate": 7.260672703751617e-07,
      "loss": 1.1968,
      "step": 449
    },
    {
      "epoch": 0.058258083309059135,
      "grad_norm": 0.15686072776208848,
      "learning_rate": 7.276843467011643e-07,
      "loss": 1.1643,
      "step": 450
    },
    {
      "epoch": 0.058258083309059135,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 Neg": 0.3673469387755102,
      "eval_PRM NPV": 0.24,
      "eval_PRM Precision": 0.8387096774193549,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8743785619735718,
      "eval_runtime": 20.838,
      "eval_samples_per_second": 2.112,
      "eval_steps_per_second": 0.144,
      "step": 450
    },
    {
      "epoch": 0.058387545716412595,
      "grad_norm": 0.1445073686579041,
      "learning_rate": 7.293014230271669e-07,
      "loss": 1.1791,
      "step": 451
    },
    {
      "epoch": 0.05851700812376606,
      "grad_norm": 0.1398005129549432,
      "learning_rate": 7.309184993531696e-07,
      "loss": 1.199,
      "step": 452
    },
    {
      "epoch": 0.05864647053111953,
      "grad_norm": 0.15501125248077358,
      "learning_rate": 7.325355756791721e-07,
      "loss": 1.273,
      "step": 453
    },
    {
      "epoch": 0.05877593293847299,
      "grad_norm": 0.17304922806857126,
      "learning_rate": 7.341526520051748e-07,
      "loss": 1.2789,
      "step": 454
    },
    {
      "epoch": 0.05890539534582646,
      "grad_norm": 0.18428884322826286,
      "learning_rate": 7.357697283311773e-07,
      "loss": 1.3022,
      "step": 455
    },
    {
      "epoch": 0.05890539534582646,
      "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": 0.8700284361839294,
      "eval_runtime": 20.5993,
      "eval_samples_per_second": 2.136,
      "eval_steps_per_second": 0.146,
      "step": 455
    },
    {
      "epoch": 0.05903485775317992,
      "grad_norm": 0.11700940486297232,
      "learning_rate": 7.373868046571798e-07,
      "loss": 1.1996,
      "step": 456
    },
    {
      "epoch": 0.059164320160533385,
      "grad_norm": 0.14951891535595566,
      "learning_rate": 7.390038809831824e-07,
      "loss": 1.1924,
      "step": 457
    },
    {
      "epoch": 0.05929378256788685,
      "grad_norm": 0.10582286100498639,
      "learning_rate": 7.406209573091851e-07,
      "loss": 0.9731,
      "step": 458
    },
    {
      "epoch": 0.05942324497524031,
      "grad_norm": 0.16239779740262222,
      "learning_rate": 7.422380336351876e-07,
      "loss": 1.3309,
      "step": 459
    },
    {
      "epoch": 0.05955270738259378,
      "grad_norm": 0.1422983937760363,
      "learning_rate": 7.438551099611903e-07,
      "loss": 1.2321,
      "step": 460
    },
    {
      "epoch": 0.05955270738259378,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 Neg": 0.3673469387755102,
      "eval_PRM NPV": 0.24,
      "eval_PRM Precision": 0.8387096774193549,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8709161877632141,
      "eval_runtime": 20.5143,
      "eval_samples_per_second": 2.145,
      "eval_steps_per_second": 0.146,
      "step": 460
    },
    {
      "epoch": 0.05968216978994725,
      "grad_norm": 0.13105487867588464,
      "learning_rate": 7.454721862871928e-07,
      "loss": 1.2403,
      "step": 461
    },
    {
      "epoch": 0.05981163219730071,
      "grad_norm": 0.1463207465352042,
      "learning_rate": 7.470892626131953e-07,
      "loss": 1.2631,
      "step": 462
    },
    {
      "epoch": 0.059941094604654174,
      "grad_norm": 0.18155543515843833,
      "learning_rate": 7.48706338939198e-07,
      "loss": 1.3165,
      "step": 463
    },
    {
      "epoch": 0.06007055701200764,
      "grad_norm": 0.16772613369862782,
      "learning_rate": 7.503234152652006e-07,
      "loss": 1.2726,
      "step": 464
    },
    {
      "epoch": 0.0602000194193611,
      "grad_norm": 0.17221101565145364,
      "learning_rate": 7.519404915912032e-07,
      "loss": 1.241,
      "step": 465
    },
    {
      "epoch": 0.0602000194193611,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 Neg": 0.3673469387755102,
      "eval_PRM NPV": 0.24,
      "eval_PRM Precision": 0.8387096774193549,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8724254369735718,
      "eval_runtime": 20.7869,
      "eval_samples_per_second": 2.117,
      "eval_steps_per_second": 0.144,
      "step": 465
    },
    {
      "epoch": 0.06032948182671457,
      "grad_norm": 0.13884773459117258,
      "learning_rate": 7.535575679172058e-07,
      "loss": 1.2317,
      "step": 466
    },
    {
      "epoch": 0.06045894423406803,
      "grad_norm": 0.14601496643349082,
      "learning_rate": 7.551746442432084e-07,
      "loss": 1.1401,
      "step": 467
    },
    {
      "epoch": 0.0605884066414215,
      "grad_norm": 0.1171977745767822,
      "learning_rate": 7.567917205692108e-07,
      "loss": 1.0899,
      "step": 468
    },
    {
      "epoch": 0.060717869048774964,
      "grad_norm": 0.15898526091134496,
      "learning_rate": 7.584087968952135e-07,
      "loss": 1.244,
      "step": 469
    },
    {
      "epoch": 0.060847331456128424,
      "grad_norm": 0.14717091472252905,
      "learning_rate": 7.600258732212161e-07,
      "loss": 1.1522,
      "step": 470
    },
    {
      "epoch": 0.060847331456128424,
      "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": 0.8696733117103577,
      "eval_runtime": 19.786,
      "eval_samples_per_second": 2.224,
      "eval_steps_per_second": 0.152,
      "step": 470
    },
    {
      "epoch": 0.06097679386348189,
      "grad_norm": 0.1721605374486989,
      "learning_rate": 7.616429495472187e-07,
      "loss": 1.3363,
      "step": 471
    },
    {
      "epoch": 0.06110625627083536,
      "grad_norm": 0.15470832448277333,
      "learning_rate": 7.632600258732213e-07,
      "loss": 1.2373,
      "step": 472
    },
    {
      "epoch": 0.06123571867818882,
      "grad_norm": 0.1699166863524227,
      "learning_rate": 7.64877102199224e-07,
      "loss": 1.2898,
      "step": 473
    },
    {
      "epoch": 0.061365181085542286,
      "grad_norm": 0.15518315810260483,
      "learning_rate": 7.664941785252264e-07,
      "loss": 1.2178,
      "step": 474
    },
    {
      "epoch": 0.061494643492895754,
      "grad_norm": 0.1342887369088432,
      "learning_rate": 7.68111254851229e-07,
      "loss": 1.0675,
      "step": 475
    },
    {
      "epoch": 0.061494643492895754,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 Neg": 0.3673469387755102,
      "eval_PRM NPV": 0.24,
      "eval_PRM Precision": 0.8387096774193549,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8690518736839294,
      "eval_runtime": 20.8075,
      "eval_samples_per_second": 2.115,
      "eval_steps_per_second": 0.144,
      "step": 475
    },
    {
      "epoch": 0.061624105900249214,
      "grad_norm": 0.14001699136581958,
      "learning_rate": 7.697283311772316e-07,
      "loss": 1.261,
      "step": 476
    },
    {
      "epoch": 0.06175356830760268,
      "grad_norm": 0.15119983516027546,
      "learning_rate": 7.713454075032342e-07,
      "loss": 1.1957,
      "step": 477
    },
    {
      "epoch": 0.06188303071495614,
      "grad_norm": 0.18901478248968237,
      "learning_rate": 7.729624838292368e-07,
      "loss": 1.269,
      "step": 478
    },
    {
      "epoch": 0.06201249312230961,
      "grad_norm": 0.09125166166967802,
      "learning_rate": 7.745795601552395e-07,
      "loss": 0.9839,
      "step": 479
    },
    {
      "epoch": 0.062141955529663076,
      "grad_norm": 0.175511086783214,
      "learning_rate": 7.761966364812419e-07,
      "loss": 1.3175,
      "step": 480
    },
    {
      "epoch": 0.062141955529663076,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 Neg": 0.3673469387755102,
      "eval_PRM NPV": 0.24,
      "eval_PRM Precision": 0.8387096774193549,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8687855005264282,
      "eval_runtime": 20.3731,
      "eval_samples_per_second": 2.16,
      "eval_steps_per_second": 0.147,
      "step": 480
    },
    {
      "epoch": 0.062271417937016536,
      "grad_norm": 0.13721361177262167,
      "learning_rate": 7.778137128072445e-07,
      "loss": 1.1645,
      "step": 481
    },
    {
      "epoch": 0.062400880344370004,
      "grad_norm": 0.15056439471572244,
      "learning_rate": 7.794307891332472e-07,
      "loss": 1.1878,
      "step": 482
    },
    {
      "epoch": 0.06253034275172346,
      "grad_norm": 0.14421034351150194,
      "learning_rate": 7.810478654592497e-07,
      "loss": 1.0578,
      "step": 483
    },
    {
      "epoch": 0.06265980515907693,
      "grad_norm": 0.1625534004144903,
      "learning_rate": 7.826649417852523e-07,
      "loss": 1.2708,
      "step": 484
    },
    {
      "epoch": 0.0627892675664304,
      "grad_norm": 0.19825140696442986,
      "learning_rate": 7.84282018111255e-07,
      "loss": 1.333,
      "step": 485
    },
    {
      "epoch": 0.0627892675664304,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 Neg": 0.3673469387755102,
      "eval_PRM NPV": 0.24,
      "eval_PRM Precision": 0.8387096774193549,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8708274364471436,
      "eval_runtime": 20.0568,
      "eval_samples_per_second": 2.194,
      "eval_steps_per_second": 0.15,
      "step": 485
    },
    {
      "epoch": 0.06291872997378387,
      "grad_norm": 0.15990428714802346,
      "learning_rate": 7.858990944372575e-07,
      "loss": 1.262,
      "step": 486
    },
    {
      "epoch": 0.06304819238113733,
      "grad_norm": 0.13825763304629105,
      "learning_rate": 7.8751617076326e-07,
      "loss": 1.2025,
      "step": 487
    },
    {
      "epoch": 0.06317765478849079,
      "grad_norm": 0.11988274487175142,
      "learning_rate": 7.891332470892627e-07,
      "loss": 1.0953,
      "step": 488
    },
    {
      "epoch": 0.06330711719584425,
      "grad_norm": 0.1424916453284857,
      "learning_rate": 7.907503234152652e-07,
      "loss": 1.1959,
      "step": 489
    },
    {
      "epoch": 0.06343657960319772,
      "grad_norm": 0.1738545755948542,
      "learning_rate": 7.923673997412679e-07,
      "loss": 1.2179,
      "step": 490
    },
    {
      "epoch": 0.06343657960319772,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 Neg": 0.3673469387755102,
      "eval_PRM NPV": 0.24,
      "eval_PRM Precision": 0.8387096774193549,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8650568127632141,
      "eval_runtime": 20.1328,
      "eval_samples_per_second": 2.185,
      "eval_steps_per_second": 0.149,
      "step": 490
    },
    {
      "epoch": 0.06356604201055119,
      "grad_norm": 0.15924679731426286,
      "learning_rate": 7.939844760672705e-07,
      "loss": 1.2163,
      "step": 491
    },
    {
      "epoch": 0.06369550441790466,
      "grad_norm": 0.13504368322337154,
      "learning_rate": 7.956015523932731e-07,
      "loss": 1.1524,
      "step": 492
    },
    {
      "epoch": 0.06382496682525812,
      "grad_norm": 0.19516173571768566,
      "learning_rate": 7.972186287192756e-07,
      "loss": 1.2698,
      "step": 493
    },
    {
      "epoch": 0.06395442923261158,
      "grad_norm": 0.17296071460350795,
      "learning_rate": 7.988357050452782e-07,
      "loss": 1.2681,
      "step": 494
    },
    {
      "epoch": 0.06408389163996504,
      "grad_norm": 0.14507150919162592,
      "learning_rate": 8.004527813712807e-07,
      "loss": 1.2128,
      "step": 495
    },
    {
      "epoch": 0.06408389163996504,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 Neg": 0.3673469387755102,
      "eval_PRM NPV": 0.24,
      "eval_PRM Precision": 0.8387096774193549,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8668323755264282,
      "eval_runtime": 20.412,
      "eval_samples_per_second": 2.156,
      "eval_steps_per_second": 0.147,
      "step": 495
    },
    {
      "epoch": 0.06421335404731851,
      "grad_norm": 0.12947399462901077,
      "learning_rate": 8.020698576972834e-07,
      "loss": 1.0468,
      "step": 496
    },
    {
      "epoch": 0.06434281645467198,
      "grad_norm": 0.18382020007202413,
      "learning_rate": 8.03686934023286e-07,
      "loss": 1.321,
      "step": 497
    },
    {
      "epoch": 0.06447227886202545,
      "grad_norm": 0.17125290900300336,
      "learning_rate": 8.053040103492886e-07,
      "loss": 1.2533,
      "step": 498
    },
    {
      "epoch": 0.0646017412693789,
      "grad_norm": 0.18475349073549754,
      "learning_rate": 8.069210866752912e-07,
      "loss": 1.2334,
      "step": 499
    },
    {
      "epoch": 0.06473120367673237,
      "grad_norm": 0.1835365851532853,
      "learning_rate": 8.085381630012937e-07,
      "loss": 1.2693,
      "step": 500
    },
    {
      "epoch": 0.06473120367673237,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 Neg": 0.3673469387755102,
      "eval_PRM NPV": 0.24,
      "eval_PRM Precision": 0.8387096774193549,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8630149364471436,
      "eval_runtime": 20.2017,
      "eval_samples_per_second": 2.178,
      "eval_steps_per_second": 0.149,
      "step": 500
    },
    {
      "epoch": 0.06486066608408583,
      "grad_norm": 0.15533986164701288,
      "learning_rate": 8.101552393272963e-07,
      "loss": 1.2524,
      "step": 501
    },
    {
      "epoch": 0.0649901284914393,
      "grad_norm": 0.13272793025742835,
      "learning_rate": 8.117723156532989e-07,
      "loss": 1.2069,
      "step": 502
    },
    {
      "epoch": 0.06511959089879277,
      "grad_norm": 0.14307161885062755,
      "learning_rate": 8.133893919793015e-07,
      "loss": 1.1619,
      "step": 503
    },
    {
      "epoch": 0.06524905330614623,
      "grad_norm": 0.2049652065365769,
      "learning_rate": 8.150064683053041e-07,
      "loss": 1.4006,
      "step": 504
    },
    {
      "epoch": 0.06537851571349969,
      "grad_norm": 0.16082439520482802,
      "learning_rate": 8.166235446313067e-07,
      "loss": 1.2555,
      "step": 505
    },
    {
      "epoch": 0.06537851571349969,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 Neg": 0.3673469387755102,
      "eval_PRM NPV": 0.24,
      "eval_PRM Precision": 0.8387096774193549,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.865234375,
      "eval_runtime": 20.0475,
      "eval_samples_per_second": 2.195,
      "eval_steps_per_second": 0.15,
      "step": 505
    },
    {
      "epoch": 0.06550797812085316,
      "grad_norm": 0.19053121423386557,
      "learning_rate": 8.182406209573091e-07,
      "loss": 1.3206,
      "step": 506
    },
    {
      "epoch": 0.06563744052820662,
      "grad_norm": 0.13037440939866252,
      "learning_rate": 8.198576972833118e-07,
      "loss": 1.1543,
      "step": 507
    },
    {
      "epoch": 0.06576690293556009,
      "grad_norm": 0.16469524550826165,
      "learning_rate": 8.214747736093144e-07,
      "loss": 1.1472,
      "step": 508
    },
    {
      "epoch": 0.06589636534291356,
      "grad_norm": 0.15490494265034072,
      "learning_rate": 8.23091849935317e-07,
      "loss": 1.2928,
      "step": 509
    },
    {
      "epoch": 0.06602582775026701,
      "grad_norm": 0.15410454528969175,
      "learning_rate": 8.247089262613196e-07,
      "loss": 1.1617,
      "step": 510
    },
    {
      "epoch": 0.06602582775026701,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 Neg": 0.3673469387755102,
      "eval_PRM NPV": 0.24,
      "eval_PRM Precision": 0.8387096774193549,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8649680614471436,
      "eval_runtime": 19.8651,
      "eval_samples_per_second": 2.215,
      "eval_steps_per_second": 0.151,
      "step": 510
    },
    {
      "epoch": 0.06615529015762048,
      "grad_norm": 0.1422897460647621,
      "learning_rate": 8.263260025873223e-07,
      "loss": 1.1007,
      "step": 511
    },
    {
      "epoch": 0.06628475256497394,
      "grad_norm": 0.12994141541611715,
      "learning_rate": 8.279430789133249e-07,
      "loss": 1.231,
      "step": 512
    },
    {
      "epoch": 0.06641421497232741,
      "grad_norm": 0.13637163812302364,
      "learning_rate": 8.295601552393273e-07,
      "loss": 1.0859,
      "step": 513
    },
    {
      "epoch": 0.06654367737968088,
      "grad_norm": 0.18845660841968817,
      "learning_rate": 8.311772315653299e-07,
      "loss": 1.2546,
      "step": 514
    },
    {
      "epoch": 0.06667313978703435,
      "grad_norm": 0.19155304423346833,
      "learning_rate": 8.327943078913325e-07,
      "loss": 1.3252,
      "step": 515
    },
    {
      "epoch": 0.06667313978703435,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 Neg": 0.3673469387755102,
      "eval_PRM NPV": 0.24,
      "eval_PRM Precision": 0.8387096774193549,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8631036877632141,
      "eval_runtime": 20.4647,
      "eval_samples_per_second": 2.15,
      "eval_steps_per_second": 0.147,
      "step": 515
    },
    {
      "epoch": 0.0668026021943878,
      "grad_norm": 0.1354322038132936,
      "learning_rate": 8.344113842173351e-07,
      "loss": 1.1366,
      "step": 516
    },
    {
      "epoch": 0.06693206460174127,
      "grad_norm": 0.20426491263243218,
      "learning_rate": 8.360284605433378e-07,
      "loss": 1.3192,
      "step": 517
    },
    {
      "epoch": 0.06706152700909473,
      "grad_norm": 0.16293582348557498,
      "learning_rate": 8.376455368693403e-07,
      "loss": 1.114,
      "step": 518
    },
    {
      "epoch": 0.0671909894164482,
      "grad_norm": 0.16853350946740797,
      "learning_rate": 8.392626131953428e-07,
      "loss": 1.1948,
      "step": 519
    },
    {
      "epoch": 0.06732045182380167,
      "grad_norm": 0.18334128196239635,
      "learning_rate": 8.408796895213455e-07,
      "loss": 1.1876,
      "step": 520
    },
    {
      "epoch": 0.06732045182380167,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 Neg": 0.3673469387755102,
      "eval_PRM NPV": 0.24,
      "eval_PRM Precision": 0.8387096774193549,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8650568127632141,
      "eval_runtime": 19.9449,
      "eval_samples_per_second": 2.206,
      "eval_steps_per_second": 0.15,
      "step": 520
    },
    {
      "epoch": 0.06744991423115512,
      "grad_norm": 0.18358978489016634,
      "learning_rate": 8.42496765847348e-07,
      "loss": 1.3896,
      "step": 521
    },
    {
      "epoch": 0.06757937663850859,
      "grad_norm": 0.2038415529520888,
      "learning_rate": 8.441138421733506e-07,
      "loss": 1.2625,
      "step": 522
    },
    {
      "epoch": 0.06770883904586206,
      "grad_norm": 0.17070099910417433,
      "learning_rate": 8.457309184993533e-07,
      "loss": 1.1179,
      "step": 523
    },
    {
      "epoch": 0.06783830145321552,
      "grad_norm": 0.10916369039559713,
      "learning_rate": 8.473479948253558e-07,
      "loss": 1.0026,
      "step": 524
    },
    {
      "epoch": 0.06796776386056899,
      "grad_norm": 0.16343640191204784,
      "learning_rate": 8.489650711513583e-07,
      "loss": 1.2334,
      "step": 525
    },
    {
      "epoch": 0.06796776386056899,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 Neg": 0.3673469387755102,
      "eval_PRM NPV": 0.24,
      "eval_PRM Precision": 0.8387096774193549,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8608842492103577,
      "eval_runtime": 20.6875,
      "eval_samples_per_second": 2.127,
      "eval_steps_per_second": 0.145,
      "step": 525
    },
    {
      "epoch": 0.06809722626792246,
      "grad_norm": 0.17444180191071568,
      "learning_rate": 8.50582147477361e-07,
      "loss": 1.2553,
      "step": 526
    },
    {
      "epoch": 0.06822668867527591,
      "grad_norm": 0.18837570556921665,
      "learning_rate": 8.521992238033635e-07,
      "loss": 1.3142,
      "step": 527
    },
    {
      "epoch": 0.06835615108262938,
      "grad_norm": 0.1734288492759629,
      "learning_rate": 8.538163001293662e-07,
      "loss": 1.324,
      "step": 528
    },
    {
      "epoch": 0.06848561348998285,
      "grad_norm": 0.12331213871707122,
      "learning_rate": 8.554333764553688e-07,
      "loss": 1.1145,
      "step": 529
    },
    {
      "epoch": 0.06861507589733631,
      "grad_norm": 0.16831465606276777,
      "learning_rate": 8.570504527813713e-07,
      "loss": 1.13,
      "step": 530
    },
    {
      "epoch": 0.06861507589733631,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 Neg": 0.3673469387755102,
      "eval_PRM NPV": 0.24,
      "eval_PRM Precision": 0.8387096774193549,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8599964380264282,
      "eval_runtime": 20.8744,
      "eval_samples_per_second": 2.108,
      "eval_steps_per_second": 0.144,
      "step": 530
    },
    {
      "epoch": 0.06874453830468978,
      "grad_norm": 0.1445882247679938,
      "learning_rate": 8.58667529107374e-07,
      "loss": 1.158,
      "step": 531
    },
    {
      "epoch": 0.06887400071204323,
      "grad_norm": 0.19259645345371518,
      "learning_rate": 8.602846054333765e-07,
      "loss": 1.3007,
      "step": 532
    },
    {
      "epoch": 0.0690034631193967,
      "grad_norm": 0.18263105756733602,
      "learning_rate": 8.61901681759379e-07,
      "loss": 1.3082,
      "step": 533
    },
    {
      "epoch": 0.06913292552675017,
      "grad_norm": 0.1919604472818187,
      "learning_rate": 8.635187580853817e-07,
      "loss": 1.2994,
      "step": 534
    },
    {
      "epoch": 0.06926238793410364,
      "grad_norm": 0.16680511489009675,
      "learning_rate": 8.651358344113843e-07,
      "loss": 1.2211,
      "step": 535
    },
    {
      "epoch": 0.06926238793410364,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 Neg": 0.3673469387755102,
      "eval_PRM NPV": 0.24,
      "eval_PRM Precision": 0.8387096774193549,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8592862486839294,
      "eval_runtime": 20.7683,
      "eval_samples_per_second": 2.119,
      "eval_steps_per_second": 0.144,
      "step": 535
    },
    {
      "epoch": 0.0693918503414571,
      "grad_norm": 0.11888426697223828,
      "learning_rate": 8.667529107373869e-07,
      "loss": 1.101,
      "step": 536
    },
    {
      "epoch": 0.06952131274881057,
      "grad_norm": 0.13619872662400598,
      "learning_rate": 8.683699870633895e-07,
      "loss": 1.0791,
      "step": 537
    },
    {
      "epoch": 0.06965077515616402,
      "grad_norm": 0.21298827859890154,
      "learning_rate": 8.69987063389392e-07,
      "loss": 1.3262,
      "step": 538
    },
    {
      "epoch": 0.06978023756351749,
      "grad_norm": 0.19935232802501582,
      "learning_rate": 8.716041397153946e-07,
      "loss": 1.3511,
      "step": 539
    },
    {
      "epoch": 0.06990969997087096,
      "grad_norm": 0.16116761083285033,
      "learning_rate": 8.732212160413972e-07,
      "loss": 1.1763,
      "step": 540
    },
    {
      "epoch": 0.06990969997087096,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 Neg": 0.3673469387755102,
      "eval_PRM NPV": 0.24,
      "eval_PRM Precision": 0.8387096774193549,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8583984375,
      "eval_runtime": 19.7691,
      "eval_samples_per_second": 2.226,
      "eval_steps_per_second": 0.152,
      "step": 540
    },
    {
      "epoch": 0.07003916237822443,
      "grad_norm": 0.16606151506619116,
      "learning_rate": 8.748382923673998e-07,
      "loss": 1.248,
      "step": 541
    },
    {
      "epoch": 0.0701686247855779,
      "grad_norm": 0.19155037350559567,
      "learning_rate": 8.764553686934024e-07,
      "loss": 1.2234,
      "step": 542
    },
    {
      "epoch": 0.07029808719293135,
      "grad_norm": 0.17041558458586142,
      "learning_rate": 8.78072445019405e-07,
      "loss": 1.4387,
      "step": 543
    },
    {
      "epoch": 0.07042754960028481,
      "grad_norm": 0.16917436559163487,
      "learning_rate": 8.796895213454077e-07,
      "loss": 1.2923,
      "step": 544
    },
    {
      "epoch": 0.07055701200763828,
      "grad_norm": 0.1969624662563443,
      "learning_rate": 8.813065976714101e-07,
      "loss": 1.1927,
      "step": 545
    },
    {
      "epoch": 0.07055701200763828,
      "eval_PRM Accuracy": 0.4056603773584906,
      "eval_PRM F1": 0.45217391304347826,
      "eval_PRM F1 AUC": 0.5261917234154007,
      "eval_PRM F1 Neg": 0.35051546391752575,
      "eval_PRM NPV": 0.22972972972972974,
      "eval_PRM Precision": 0.8125,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.8569779992103577,
      "eval_runtime": 20.7431,
      "eval_samples_per_second": 2.121,
      "eval_steps_per_second": 0.145,
      "step": 545
    },
    {
      "epoch": 0.07068647441499175,
      "grad_norm": 0.16758299120956285,
      "learning_rate": 8.829236739974127e-07,
      "loss": 1.2706,
      "step": 546
    },
    {
      "epoch": 0.07081593682234522,
      "grad_norm": 0.1711082914098508,
      "learning_rate": 8.845407503234154e-07,
      "loss": 1.2469,
      "step": 547
    },
    {
      "epoch": 0.07094539922969868,
      "grad_norm": 0.20270607529372123,
      "learning_rate": 8.861578266494179e-07,
      "loss": 1.2977,
      "step": 548
    },
    {
      "epoch": 0.07107486163705214,
      "grad_norm": 0.1852464253854372,
      "learning_rate": 8.877749029754205e-07,
      "loss": 1.3358,
      "step": 549
    },
    {
      "epoch": 0.0712043240444056,
      "grad_norm": 0.19841295958610808,
      "learning_rate": 8.893919793014232e-07,
      "loss": 1.3088,
      "step": 550
    },
    {
      "epoch": 0.0712043240444056,
      "eval_PRM Accuracy": 0.4056603773584906,
      "eval_PRM F1": 0.45217391304347826,
      "eval_PRM F1 AUC": 0.5261917234154007,
      "eval_PRM F1 Neg": 0.35051546391752575,
      "eval_PRM NPV": 0.22972972972972974,
      "eval_PRM Precision": 0.8125,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.8559126257896423,
      "eval_runtime": 20.8278,
      "eval_samples_per_second": 2.113,
      "eval_steps_per_second": 0.144,
      "step": 550
    },
    {
      "epoch": 0.07133378645175907,
      "grad_norm": 0.18439949787344748,
      "learning_rate": 8.910090556274256e-07,
      "loss": 1.2147,
      "step": 551
    },
    {
      "epoch": 0.07146324885911254,
      "grad_norm": 0.1712580730880845,
      "learning_rate": 8.926261319534282e-07,
      "loss": 1.1318,
      "step": 552
    },
    {
      "epoch": 0.071592711266466,
      "grad_norm": 0.21128305636090958,
      "learning_rate": 8.942432082794309e-07,
      "loss": 1.3673,
      "step": 553
    },
    {
      "epoch": 0.07172217367381946,
      "grad_norm": 0.1613265149710803,
      "learning_rate": 8.958602846054334e-07,
      "loss": 1.1743,
      "step": 554
    },
    {
      "epoch": 0.07185163608117293,
      "grad_norm": 0.1457513959689151,
      "learning_rate": 8.974773609314361e-07,
      "loss": 1.14,
      "step": 555
    },
    {
      "epoch": 0.07185163608117293,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 Neg": 0.3673469387755102,
      "eval_PRM NPV": 0.24,
      "eval_PRM Precision": 0.8387096774193549,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8570667505264282,
      "eval_runtime": 19.5877,
      "eval_samples_per_second": 2.246,
      "eval_steps_per_second": 0.153,
      "step": 555
    },
    {
      "epoch": 0.0719810984885264,
      "grad_norm": 0.1847152662226901,
      "learning_rate": 8.990944372574387e-07,
      "loss": 1.2263,
      "step": 556
    },
    {
      "epoch": 0.07211056089587986,
      "grad_norm": 0.1155253478010054,
      "learning_rate": 9.007115135834413e-07,
      "loss": 1.0538,
      "step": 557
    },
    {
      "epoch": 0.07224002330323333,
      "grad_norm": 0.1385732996983949,
      "learning_rate": 9.023285899094438e-07,
      "loss": 1.0852,
      "step": 558
    },
    {
      "epoch": 0.0723694857105868,
      "grad_norm": 0.18683493876719523,
      "learning_rate": 9.039456662354464e-07,
      "loss": 1.2081,
      "step": 559
    },
    {
      "epoch": 0.07249894811794025,
      "grad_norm": 0.155354811949633,
      "learning_rate": 9.055627425614489e-07,
      "loss": 1.2561,
      "step": 560
    },
    {
      "epoch": 0.07249894811794025,
      "eval_PRM Accuracy": 0.4056603773584906,
      "eval_PRM F1": 0.45217391304347826,
      "eval_PRM F1 AUC": 0.5261917234154007,
      "eval_PRM F1 Neg": 0.35051546391752575,
      "eval_PRM NPV": 0.22972972972972974,
      "eval_PRM Precision": 0.8125,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.8552024364471436,
      "eval_runtime": 20.3271,
      "eval_samples_per_second": 2.165,
      "eval_steps_per_second": 0.148,
      "step": 560
    },
    {
      "epoch": 0.07262841052529372,
      "grad_norm": 0.1892510200193733,
      "learning_rate": 9.071798188874516e-07,
      "loss": 1.2211,
      "step": 561
    },
    {
      "epoch": 0.07275787293264718,
      "grad_norm": 0.21262039296717195,
      "learning_rate": 9.087968952134542e-07,
      "loss": 1.2981,
      "step": 562
    },
    {
      "epoch": 0.07288733534000065,
      "grad_norm": 0.17308827453310116,
      "learning_rate": 9.104139715394568e-07,
      "loss": 1.2527,
      "step": 563
    },
    {
      "epoch": 0.07301679774735412,
      "grad_norm": 0.16906490926984974,
      "learning_rate": 9.120310478654593e-07,
      "loss": 1.1633,
      "step": 564
    },
    {
      "epoch": 0.07314626015470757,
      "grad_norm": 0.11988652249069812,
      "learning_rate": 9.136481241914618e-07,
      "loss": 1.0525,
      "step": 565
    },
    {
      "epoch": 0.07314626015470757,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 Neg": 0.3673469387755102,
      "eval_PRM NPV": 0.24,
      "eval_PRM Precision": 0.8387096774193549,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8514737486839294,
      "eval_runtime": 20.3487,
      "eval_samples_per_second": 2.162,
      "eval_steps_per_second": 0.147,
      "step": 565
    },
    {
      "epoch": 0.07327572256206104,
      "grad_norm": 0.15296886676853855,
      "learning_rate": 9.152652005174645e-07,
      "loss": 1.1919,
      "step": 566
    },
    {
      "epoch": 0.0734051849694145,
      "grad_norm": 0.2066753964853689,
      "learning_rate": 9.168822768434671e-07,
      "loss": 1.2742,
      "step": 567
    },
    {
      "epoch": 0.07353464737676797,
      "grad_norm": 0.1497077319308681,
      "learning_rate": 9.184993531694696e-07,
      "loss": 1.0746,
      "step": 568
    },
    {
      "epoch": 0.07366410978412144,
      "grad_norm": 0.2173177327155543,
      "learning_rate": 9.201164294954723e-07,
      "loss": 1.3013,
      "step": 569
    },
    {
      "epoch": 0.07379357219147491,
      "grad_norm": 0.18223242906680462,
      "learning_rate": 9.217335058214749e-07,
      "loss": 1.1649,
      "step": 570
    },
    {
      "epoch": 0.07379357219147491,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 Neg": 0.3673469387755102,
      "eval_PRM NPV": 0.24,
      "eval_PRM Precision": 0.8387096774193549,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8512961864471436,
      "eval_runtime": 20.6008,
      "eval_samples_per_second": 2.136,
      "eval_steps_per_second": 0.146,
      "step": 570
    },
    {
      "epoch": 0.07392303459882836,
      "grad_norm": 0.22343647381701554,
      "learning_rate": 9.233505821474773e-07,
      "loss": 1.322,
      "step": 571
    },
    {
      "epoch": 0.07405249700618183,
      "grad_norm": 0.16184415259431117,
      "learning_rate": 9.2496765847348e-07,
      "loss": 1.1327,
      "step": 572
    },
    {
      "epoch": 0.0741819594135353,
      "grad_norm": 0.15616613926206077,
      "learning_rate": 9.265847347994826e-07,
      "loss": 1.2274,
      "step": 573
    },
    {
      "epoch": 0.07431142182088876,
      "grad_norm": 0.14240129436834736,
      "learning_rate": 9.282018111254852e-07,
      "loss": 1.2004,
      "step": 574
    },
    {
      "epoch": 0.07444088422824223,
      "grad_norm": 0.17199606502566842,
      "learning_rate": 9.298188874514878e-07,
      "loss": 1.2162,
      "step": 575
    },
    {
      "epoch": 0.07444088422824223,
      "eval_PRM Accuracy": 0.42452830188679247,
      "eval_PRM F1": 0.46956521739130436,
      "eval_PRM F1 AUC": 0.5539549502357256,
      "eval_PRM F1 Neg": 0.3711340206185567,
      "eval_PRM NPV": 0.24324324324324326,
      "eval_PRM Precision": 0.84375,
      "eval_PRM Recall": 0.3253012048192771,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8500532507896423,
      "eval_runtime": 20.1227,
      "eval_samples_per_second": 2.187,
      "eval_steps_per_second": 0.149,
      "step": 575
    },
    {
      "epoch": 0.07457034663559568,
      "grad_norm": 0.18013920228667082,
      "learning_rate": 9.314359637774904e-07,
      "loss": 1.1343,
      "step": 576
    },
    {
      "epoch": 0.07469980904294915,
      "grad_norm": 0.22027558714483902,
      "learning_rate": 9.330530401034929e-07,
      "loss": 1.3381,
      "step": 577
    },
    {
      "epoch": 0.07482927145030262,
      "grad_norm": 0.22043454340086804,
      "learning_rate": 9.346701164294955e-07,
      "loss": 1.2676,
      "step": 578
    },
    {
      "epoch": 0.07495873385765608,
      "grad_norm": 0.17957710018701944,
      "learning_rate": 9.362871927554981e-07,
      "loss": 1.2222,
      "step": 579
    },
    {
      "epoch": 0.07508819626500955,
      "grad_norm": 0.18252652262030405,
      "learning_rate": 9.379042690815007e-07,
      "loss": 1.1116,
      "step": 580
    },
    {
      "epoch": 0.07508819626500955,
      "eval_PRM Accuracy": 0.42452830188679247,
      "eval_PRM F1": 0.46956521739130436,
      "eval_PRM F1 AUC": 0.5539549502357256,
      "eval_PRM F1 Neg": 0.3711340206185567,
      "eval_PRM NPV": 0.24324324324324326,
      "eval_PRM Precision": 0.84375,
      "eval_PRM Recall": 0.3253012048192771,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8490766882896423,
      "eval_runtime": 19.6466,
      "eval_samples_per_second": 2.24,
      "eval_steps_per_second": 0.153,
      "step": 580
    },
    {
      "epoch": 0.07521765867236302,
      "grad_norm": 0.20980209175127562,
      "learning_rate": 9.395213454075033e-07,
      "loss": 1.2555,
      "step": 581
    },
    {
      "epoch": 0.07534712107971647,
      "grad_norm": 0.17727989607097341,
      "learning_rate": 9.41138421733506e-07,
      "loss": 1.2201,
      "step": 582
    },
    {
      "epoch": 0.07547658348706994,
      "grad_norm": 0.14146353776901985,
      "learning_rate": 9.427554980595084e-07,
      "loss": 1.177,
      "step": 583
    },
    {
      "epoch": 0.07560604589442341,
      "grad_norm": 0.1828278252702041,
      "learning_rate": 9.44372574385511e-07,
      "loss": 1.1968,
      "step": 584
    },
    {
      "epoch": 0.07573550830177687,
      "grad_norm": 0.20763803274849194,
      "learning_rate": 9.459896507115137e-07,
      "loss": 1.3428,
      "step": 585
    },
    {
      "epoch": 0.07573550830177687,
      "eval_PRM Accuracy": 0.42452830188679247,
      "eval_PRM F1": 0.46956521739130436,
      "eval_PRM F1 AUC": 0.5539549502357256,
      "eval_PRM F1 Neg": 0.3711340206185567,
      "eval_PRM NPV": 0.24324324324324326,
      "eval_PRM Precision": 0.84375,
      "eval_PRM Recall": 0.3253012048192771,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8464133739471436,
      "eval_runtime": 20.5487,
      "eval_samples_per_second": 2.141,
      "eval_steps_per_second": 0.146,
      "step": 585
    },
    {
      "epoch": 0.07586497070913034,
      "grad_norm": 0.18403077218642375,
      "learning_rate": 9.476067270375162e-07,
      "loss": 1.2616,
      "step": 586
    },
    {
      "epoch": 0.0759944331164838,
      "grad_norm": 0.1745292206717775,
      "learning_rate": 9.492238033635188e-07,
      "loss": 1.2581,
      "step": 587
    },
    {
      "epoch": 0.07612389552383726,
      "grad_norm": 0.21287042368677034,
      "learning_rate": 9.508408796895215e-07,
      "loss": 1.2311,
      "step": 588
    },
    {
      "epoch": 0.07625335793119073,
      "grad_norm": 0.18562649721453459,
      "learning_rate": 9.52457956015524e-07,
      "loss": 1.2653,
      "step": 589
    },
    {
      "epoch": 0.0763828203385442,
      "grad_norm": 0.15088666061786474,
      "learning_rate": 9.540750323415266e-07,
      "loss": 1.2665,
      "step": 590
    },
    {
      "epoch": 0.0763828203385442,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.46551724137931033,
      "eval_PRM F1 AUC": 0.5322158198009429,
      "eval_PRM F1 Neg": 0.3541666666666667,
      "eval_PRM NPV": 0.2328767123287671,
      "eval_PRM Precision": 0.8181818181818182,
      "eval_PRM Recall": 0.3253012048192771,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.8440163135528564,
      "eval_runtime": 20.7677,
      "eval_samples_per_second": 2.119,
      "eval_steps_per_second": 0.144,
      "step": 590
    },
    {
      "epoch": 0.07651228274589766,
      "grad_norm": 0.19810884025508352,
      "learning_rate": 9.55692108667529e-07,
      "loss": 1.2441,
      "step": 591
    },
    {
      "epoch": 0.07664174515325113,
      "grad_norm": 0.1538984229267891,
      "learning_rate": 9.573091849935317e-07,
      "loss": 1.0762,
      "step": 592
    },
    {
      "epoch": 0.07677120756060458,
      "grad_norm": 0.227282697618542,
      "learning_rate": 9.589262613195344e-07,
      "loss": 1.2837,
      "step": 593
    },
    {
      "epoch": 0.07690066996795805,
      "grad_norm": 0.14359725632220752,
      "learning_rate": 9.60543337645537e-07,
      "loss": 1.1511,
      "step": 594
    },
    {
      "epoch": 0.07703013237531152,
      "grad_norm": 0.2111230098409188,
      "learning_rate": 9.621604139715395e-07,
      "loss": 1.2981,
      "step": 595
    },
    {
      "epoch": 0.07703013237531152,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 Neg": 0.3673469387755102,
      "eval_PRM NPV": 0.24,
      "eval_PRM Precision": 0.8387096774193549,
      "eval_PRM Recall": 0.3132530120481928,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8430397510528564,
      "eval_runtime": 19.6605,
      "eval_samples_per_second": 2.238,
      "eval_steps_per_second": 0.153,
      "step": 595
    },
    {
      "epoch": 0.07715959478266499,
      "grad_norm": 0.1477765175861278,
      "learning_rate": 9.63777490297542e-07,
      "loss": 1.0883,
      "step": 596
    },
    {
      "epoch": 0.07728905719001845,
      "grad_norm": 0.1652592793015199,
      "learning_rate": 9.653945666235447e-07,
      "loss": 1.1495,
      "step": 597
    },
    {
      "epoch": 0.07741851959737191,
      "grad_norm": 0.2434964293895327,
      "learning_rate": 9.670116429495473e-07,
      "loss": 1.2626,
      "step": 598
    },
    {
      "epoch": 0.07754798200472537,
      "grad_norm": 0.17775056668002065,
      "learning_rate": 9.686287192755498e-07,
      "loss": 1.0856,
      "step": 599
    },
    {
      "epoch": 0.07767744441207884,
      "grad_norm": 0.18412335237074828,
      "learning_rate": 9.702457956015525e-07,
      "loss": 1.1545,
      "step": 600
    },
    {
      "epoch": 0.07767744441207884,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.46551724137931033,
      "eval_PRM F1 AUC": 0.5322158198009429,
      "eval_PRM F1 Neg": 0.3541666666666667,
      "eval_PRM NPV": 0.2328767123287671,
      "eval_PRM Precision": 0.8181818181818182,
      "eval_PRM Recall": 0.3253012048192771,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.8440163135528564,
      "eval_runtime": 20.5536,
      "eval_samples_per_second": 2.141,
      "eval_steps_per_second": 0.146,
      "step": 600
    },
    {
      "epoch": 0.07780690681943231,
      "grad_norm": 0.20451127740359148,
      "learning_rate": 9.718628719275551e-07,
      "loss": 1.2496,
      "step": 601
    },
    {
      "epoch": 0.07793636922678578,
      "grad_norm": 0.134688938014961,
      "learning_rate": 9.734799482535578e-07,
      "loss": 1.1478,
      "step": 602
    },
    {
      "epoch": 0.07806583163413924,
      "grad_norm": 0.18953121801844397,
      "learning_rate": 9.750970245795603e-07,
      "loss": 1.2623,
      "step": 603
    },
    {
      "epoch": 0.0781952940414927,
      "grad_norm": 0.20055597187343607,
      "learning_rate": 9.767141009055627e-07,
      "loss": 1.2225,
      "step": 604
    },
    {
      "epoch": 0.07832475644884616,
      "grad_norm": 0.1744283042399751,
      "learning_rate": 9.783311772315654e-07,
      "loss": 1.1193,
      "step": 605
    },
    {
      "epoch": 0.07832475644884616,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.46551724137931033,
      "eval_PRM F1 AUC": 0.5322158198009429,
      "eval_PRM F1 Neg": 0.3541666666666667,
      "eval_PRM NPV": 0.2328767123287671,
      "eval_PRM Precision": 0.8181818181818182,
      "eval_PRM Recall": 0.3253012048192771,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.8406427502632141,
      "eval_runtime": 20.005,
      "eval_samples_per_second": 2.199,
      "eval_steps_per_second": 0.15,
      "step": 605
    },
    {
      "epoch": 0.07845421885619963,
      "grad_norm": 0.15709069633705608,
      "learning_rate": 9.79948253557568e-07,
      "loss": 1.1813,
      "step": 606
    },
    {
      "epoch": 0.0785836812635531,
      "grad_norm": 0.12198167666990041,
      "learning_rate": 9.815653298835705e-07,
      "loss": 1.0535,
      "step": 607
    },
    {
      "epoch": 0.07871314367090657,
      "grad_norm": 0.16858140050121193,
      "learning_rate": 9.831824062095732e-07,
      "loss": 1.0499,
      "step": 608
    },
    {
      "epoch": 0.07884260607826002,
      "grad_norm": 0.150020677991825,
      "learning_rate": 9.847994825355757e-07,
      "loss": 1.2124,
      "step": 609
    },
    {
      "epoch": 0.07897206848561349,
      "grad_norm": 0.1305125673559483,
      "learning_rate": 9.864165588615783e-07,
      "loss": 1.1151,
      "step": 610
    },
    {
      "epoch": 0.07897206848561349,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.46551724137931033,
      "eval_PRM F1 AUC": 0.5322158198009429,
      "eval_PRM F1 Neg": 0.3541666666666667,
      "eval_PRM NPV": 0.2328767123287671,
      "eval_PRM Precision": 0.8181818181818182,
      "eval_PRM Recall": 0.3253012048192771,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.8380681872367859,
      "eval_runtime": 19.4751,
      "eval_samples_per_second": 2.259,
      "eval_steps_per_second": 0.154,
      "step": 610
    },
    {
      "epoch": 0.07910153089296695,
      "grad_norm": 0.19660200461069727,
      "learning_rate": 9.88033635187581e-07,
      "loss": 1.1732,
      "step": 611
    },
    {
      "epoch": 0.07923099330032042,
      "grad_norm": 0.19550512634253983,
      "learning_rate": 9.896507115135835e-07,
      "loss": 1.2325,
      "step": 612
    },
    {
      "epoch": 0.07936045570767389,
      "grad_norm": 0.20223104609724346,
      "learning_rate": 9.912677878395861e-07,
      "loss": 1.2853,
      "step": 613
    },
    {
      "epoch": 0.07948991811502736,
      "grad_norm": 0.16229259145988997,
      "learning_rate": 9.928848641655888e-07,
      "loss": 1.0121,
      "step": 614
    },
    {
      "epoch": 0.07961938052238081,
      "grad_norm": 0.17642818643864705,
      "learning_rate": 9.945019404915913e-07,
      "loss": 1.0804,
      "step": 615
    },
    {
      "epoch": 0.07961938052238081,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.46551724137931033,
      "eval_PRM F1 AUC": 0.5322158198009429,
      "eval_PRM F1 Neg": 0.3541666666666667,
      "eval_PRM NPV": 0.2328767123287671,
      "eval_PRM Precision": 0.8181818181818182,
      "eval_PRM Recall": 0.3253012048192771,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.8354048132896423,
      "eval_runtime": 20.224,
      "eval_samples_per_second": 2.176,
      "eval_steps_per_second": 0.148,
      "step": 615
    },
    {
      "epoch": 0.07974884292973428,
      "grad_norm": 0.15158698152541275,
      "learning_rate": 9.961190168175937e-07,
      "loss": 1.1259,
      "step": 616
    },
    {
      "epoch": 0.07987830533708774,
      "grad_norm": 0.21821063654265963,
      "learning_rate": 9.977360931435964e-07,
      "loss": 1.201,
      "step": 617
    },
    {
      "epoch": 0.08000776774444121,
      "grad_norm": 0.22009196919431773,
      "learning_rate": 9.99353169469599e-07,
      "loss": 1.199,
      "step": 618
    },
    {
      "epoch": 0.08013723015179468,
      "grad_norm": 0.19385089678883013,
      "learning_rate": 1.0009702457956017e-06,
      "loss": 1.1692,
      "step": 619
    },
    {
      "epoch": 0.08026669255914813,
      "grad_norm": 0.1920087048864159,
      "learning_rate": 1.0025873221216042e-06,
      "loss": 1.3226,
      "step": 620
    },
    {
      "epoch": 0.08026669255914813,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.46551724137931033,
      "eval_PRM F1 AUC": 0.5322158198009429,
      "eval_PRM F1 Neg": 0.3541666666666667,
      "eval_PRM NPV": 0.2328767123287671,
      "eval_PRM Precision": 0.8181818181818182,
      "eval_PRM Recall": 0.3253012048192771,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.8352272510528564,
      "eval_runtime": 20.905,
      "eval_samples_per_second": 2.105,
      "eval_steps_per_second": 0.144,
      "step": 620
    },
    {
      "epoch": 0.0803961549665016,
      "grad_norm": 0.2410325561559353,
      "learning_rate": 1.0042043984476069e-06,
      "loss": 1.2606,
      "step": 621
    },
    {
      "epoch": 0.08052561737385507,
      "grad_norm": 0.21663101423396477,
      "learning_rate": 1.0058214747736093e-06,
      "loss": 1.2263,
      "step": 622
    },
    {
      "epoch": 0.08065507978120853,
      "grad_norm": 0.19022487103327237,
      "learning_rate": 1.007438551099612e-06,
      "loss": 1.1016,
      "step": 623
    },
    {
      "epoch": 0.080784542188562,
      "grad_norm": 0.1859061257124025,
      "learning_rate": 1.0090556274256147e-06,
      "loss": 1.134,
      "step": 624
    },
    {
      "epoch": 0.08091400459591547,
      "grad_norm": 0.1750427098784179,
      "learning_rate": 1.0106727037516171e-06,
      "loss": 1.1778,
      "step": 625
    },
    {
      "epoch": 0.08091400459591547,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.46551724137931033,
      "eval_PRM F1 AUC": 0.5322158198009429,
      "eval_PRM F1 Neg": 0.3541666666666667,
      "eval_PRM NPV": 0.2328767123287671,
      "eval_PRM Precision": 0.8181818181818182,
      "eval_PRM Recall": 0.3253012048192771,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.833984375,
      "eval_runtime": 19.9809,
      "eval_samples_per_second": 2.202,
      "eval_steps_per_second": 0.15,
      "step": 625
    },
    {
      "epoch": 0.08104346700326892,
      "grad_norm": 0.20373155644029953,
      "learning_rate": 1.0122897800776198e-06,
      "loss": 1.3265,
      "step": 626
    },
    {
      "epoch": 0.08117292941062239,
      "grad_norm": 0.20544261412052614,
      "learning_rate": 1.0139068564036225e-06,
      "loss": 1.2438,
      "step": 627
    },
    {
      "epoch": 0.08130239181797586,
      "grad_norm": 0.25213857021720304,
      "learning_rate": 1.015523932729625e-06,
      "loss": 1.3451,
      "step": 628
    },
    {
      "epoch": 0.08143185422532932,
      "grad_norm": 0.1852596172036974,
      "learning_rate": 1.0171410090556274e-06,
      "loss": 1.2488,
      "step": 629
    },
    {
      "epoch": 0.08156131663268279,
      "grad_norm": 0.18492490755573396,
      "learning_rate": 1.01875808538163e-06,
      "loss": 1.0752,
      "step": 630
    },
    {
      "epoch": 0.08156131663268279,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.46551724137931033,
      "eval_PRM F1 AUC": 0.5322158198009429,
      "eval_PRM F1 Neg": 0.3541666666666667,
      "eval_PRM NPV": 0.2328767123287671,
      "eval_PRM Precision": 0.8181818181818182,
      "eval_PRM Recall": 0.3253012048192771,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.83203125,
      "eval_runtime": 20.6205,
      "eval_samples_per_second": 2.134,
      "eval_steps_per_second": 0.145,
      "step": 630
    },
    {
      "epoch": 0.08169077904003624,
      "grad_norm": 0.215974601433201,
      "learning_rate": 1.0203751617076327e-06,
      "loss": 1.2628,
      "step": 631
    },
    {
      "epoch": 0.08182024144738971,
      "grad_norm": 0.2082029106988089,
      "learning_rate": 1.0219922380336354e-06,
      "loss": 1.2837,
      "step": 632
    },
    {
      "epoch": 0.08194970385474318,
      "grad_norm": 0.24257661223325547,
      "learning_rate": 1.0236093143596378e-06,
      "loss": 1.2617,
      "step": 633
    },
    {
      "epoch": 0.08207916626209665,
      "grad_norm": 0.19125457895224893,
      "learning_rate": 1.0252263906856405e-06,
      "loss": 1.1785,
      "step": 634
    },
    {
      "epoch": 0.08220862866945011,
      "grad_norm": 0.18324450114320986,
      "learning_rate": 1.026843467011643e-06,
      "loss": 1.1764,
      "step": 635
    },
    {
      "epoch": 0.08220862866945011,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.46551724137931033,
      "eval_PRM F1 AUC": 0.5322158198009429,
      "eval_PRM F1 Neg": 0.3541666666666667,
      "eval_PRM NPV": 0.2328767123287671,
      "eval_PRM Precision": 0.8181818181818182,
      "eval_PRM Recall": 0.3253012048192771,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.8306108117103577,
      "eval_runtime": 19.7582,
      "eval_samples_per_second": 2.227,
      "eval_steps_per_second": 0.152,
      "step": 635
    },
    {
      "epoch": 0.08233809107680358,
      "grad_norm": 0.23910022887175347,
      "learning_rate": 1.0284605433376456e-06,
      "loss": 1.2593,
      "step": 636
    },
    {
      "epoch": 0.08246755348415703,
      "grad_norm": 0.17525638271870733,
      "learning_rate": 1.030077619663648e-06,
      "loss": 1.117,
      "step": 637
    },
    {
      "epoch": 0.0825970158915105,
      "grad_norm": 0.18554758396414467,
      "learning_rate": 1.0316946959896508e-06,
      "loss": 1.1935,
      "step": 638
    },
    {
      "epoch": 0.08272647829886397,
      "grad_norm": 0.2242274751047819,
      "learning_rate": 1.0333117723156534e-06,
      "loss": 1.2065,
      "step": 639
    },
    {
      "epoch": 0.08285594070621743,
      "grad_norm": 0.18594755502473304,
      "learning_rate": 1.0349288486416561e-06,
      "loss": 1.0703,
      "step": 640
    },
    {
      "epoch": 0.08285594070621743,
      "eval_PRM Accuracy": 0.42452830188679247,
      "eval_PRM F1": 0.47863247863247865,
      "eval_PRM F1 AUC": 0.538239916186485,
      "eval_PRM F1 Neg": 0.35789473684210527,
      "eval_PRM NPV": 0.2361111111111111,
      "eval_PRM Precision": 0.8235294117647058,
      "eval_PRM Recall": 0.3373493975903614,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.8270596861839294,
      "eval_runtime": 19.9843,
      "eval_samples_per_second": 2.202,
      "eval_steps_per_second": 0.15,
      "step": 640
    },
    {
      "epoch": 0.0829854031135709,
      "grad_norm": 0.15533079014938828,
      "learning_rate": 1.0365459249676586e-06,
      "loss": 1.1401,
      "step": 641
    },
    {
      "epoch": 0.08311486552092436,
      "grad_norm": 0.16403804886050868,
      "learning_rate": 1.038163001293661e-06,
      "loss": 1.1161,
      "step": 642
    },
    {
      "epoch": 0.08324432792827782,
      "grad_norm": 0.22604286299094503,
      "learning_rate": 1.0397800776196637e-06,
      "loss": 1.2501,
      "step": 643
    },
    {
      "epoch": 0.08337379033563129,
      "grad_norm": 0.2725461541421229,
      "learning_rate": 1.0413971539456664e-06,
      "loss": 1.3757,
      "step": 644
    },
    {
      "epoch": 0.08350325274298476,
      "grad_norm": 0.18481704246876085,
      "learning_rate": 1.0430142302716688e-06,
      "loss": 1.2492,
      "step": 645
    },
    {
      "epoch": 0.08350325274298476,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.46551724137931033,
      "eval_PRM F1 AUC": 0.5322158198009429,
      "eval_PRM F1 Neg": 0.3541666666666667,
      "eval_PRM NPV": 0.2328767123287671,
      "eval_PRM Precision": 0.8181818181818182,
      "eval_PRM Recall": 0.3253012048192771,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.8249289989471436,
      "eval_runtime": 20.1134,
      "eval_samples_per_second": 2.188,
      "eval_steps_per_second": 0.149,
      "step": 645
    },
    {
      "epoch": 0.08363271515033822,
      "grad_norm": 0.15664668168743495,
      "learning_rate": 1.0446313065976715e-06,
      "loss": 1.0685,
      "step": 646
    },
    {
      "epoch": 0.08376217755769169,
      "grad_norm": 0.19314453324977945,
      "learning_rate": 1.0462483829236742e-06,
      "loss": 1.2271,
      "step": 647
    },
    {
      "epoch": 0.08389163996504514,
      "grad_norm": 0.22042946699958396,
      "learning_rate": 1.0478654592496766e-06,
      "loss": 1.1367,
      "step": 648
    },
    {
      "epoch": 0.08402110237239861,
      "grad_norm": 0.1592804760009837,
      "learning_rate": 1.0494825355756793e-06,
      "loss": 1.1088,
      "step": 649
    },
    {
      "epoch": 0.08415056477975208,
      "grad_norm": 0.21381685893169836,
      "learning_rate": 1.0510996119016818e-06,
      "loss": 1.181,
      "step": 650
    },
    {
      "epoch": 0.08415056477975208,
      "eval_PRM Accuracy": 0.42452830188679247,
      "eval_PRM F1": 0.47863247863247865,
      "eval_PRM F1 AUC": 0.538239916186485,
      "eval_PRM F1 Neg": 0.35789473684210527,
      "eval_PRM NPV": 0.2361111111111111,
      "eval_PRM Precision": 0.8235294117647058,
      "eval_PRM Recall": 0.3373493975903614,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.8235973119735718,
      "eval_runtime": 20.4724,
      "eval_samples_per_second": 2.149,
      "eval_steps_per_second": 0.147,
      "step": 650
    },
    {
      "epoch": 0.08428002718710555,
      "grad_norm": 0.19722070171612469,
      "learning_rate": 1.0527166882276844e-06,
      "loss": 1.1498,
      "step": 651
    },
    {
      "epoch": 0.08440948959445901,
      "grad_norm": 0.18638623948992936,
      "learning_rate": 1.0543337645536871e-06,
      "loss": 1.14,
      "step": 652
    },
    {
      "epoch": 0.08453895200181247,
      "grad_norm": 0.17361589799056665,
      "learning_rate": 1.0559508408796896e-06,
      "loss": 1.0629,
      "step": 653
    },
    {
      "epoch": 0.08466841440916593,
      "grad_norm": 0.16058232555062704,
      "learning_rate": 1.057567917205692e-06,
      "loss": 1.0279,
      "step": 654
    },
    {
      "epoch": 0.0847978768165194,
      "grad_norm": 0.2594110366502841,
      "learning_rate": 1.0591849935316947e-06,
      "loss": 1.1725,
      "step": 655
    },
    {
      "epoch": 0.0847978768165194,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.46551724137931033,
      "eval_PRM F1 AUC": 0.5322158198009429,
      "eval_PRM F1 Neg": 0.3541666666666667,
      "eval_PRM NPV": 0.2328767123287671,
      "eval_PRM Precision": 0.8181818181818182,
      "eval_PRM Recall": 0.3253012048192771,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.8231534361839294,
      "eval_runtime": 20.0464,
      "eval_samples_per_second": 2.195,
      "eval_steps_per_second": 0.15,
      "step": 655
    },
    {
      "epoch": 0.08492733922387287,
      "grad_norm": 0.17828931812647264,
      "learning_rate": 1.0608020698576974e-06,
      "loss": 1.084,
      "step": 656
    },
    {
      "epoch": 0.08505680163122634,
      "grad_norm": 0.2292364665592309,
      "learning_rate": 1.0624191461837e-06,
      "loss": 1.2529,
      "step": 657
    },
    {
      "epoch": 0.0851862640385798,
      "grad_norm": 0.2535249597619893,
      "learning_rate": 1.0640362225097025e-06,
      "loss": 1.3447,
      "step": 658
    },
    {
      "epoch": 0.08531572644593326,
      "grad_norm": 0.22242542454789774,
      "learning_rate": 1.0656532988357052e-06,
      "loss": 1.218,
      "step": 659
    },
    {
      "epoch": 0.08544518885328672,
      "grad_norm": 0.20438897702859776,
      "learning_rate": 1.0672703751617078e-06,
      "loss": 1.0831,
      "step": 660
    },
    {
      "epoch": 0.08544518885328672,
      "eval_PRM Accuracy": 0.4339622641509434,
      "eval_PRM F1": 0.4915254237288136,
      "eval_PRM F1 AUC": 0.5442640125720272,
      "eval_PRM F1 Neg": 0.3617021276595745,
      "eval_PRM NPV": 0.23943661971830985,
      "eval_PRM Precision": 0.8285714285714286,
      "eval_PRM Recall": 0.3493975903614458,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.8225319385528564,
      "eval_runtime": 20.7566,
      "eval_samples_per_second": 2.12,
      "eval_steps_per_second": 0.145,
      "step": 660
    },
    {
      "epoch": 0.08557465126064019,
      "grad_norm": 0.17690767226793896,
      "learning_rate": 1.0688874514877103e-06,
      "loss": 1.0247,
      "step": 661
    },
    {
      "epoch": 0.08570411366799366,
      "grad_norm": 0.20176416990123527,
      "learning_rate": 1.070504527813713e-06,
      "loss": 1.184,
      "step": 662
    },
    {
      "epoch": 0.08583357607534713,
      "grad_norm": 0.18262888880229672,
      "learning_rate": 1.0721216041397154e-06,
      "loss": 1.2097,
      "step": 663
    },
    {
      "epoch": 0.08596303848270058,
      "grad_norm": 0.21244129132342518,
      "learning_rate": 1.073738680465718e-06,
      "loss": 1.2399,
      "step": 664
    },
    {
      "epoch": 0.08609250089005405,
      "grad_norm": 0.19529942397470798,
      "learning_rate": 1.0753557567917208e-06,
      "loss": 1.2333,
      "step": 665
    },
    {
      "epoch": 0.08609250089005405,
      "eval_PRM Accuracy": 0.4339622641509434,
      "eval_PRM F1": 0.4915254237288136,
      "eval_PRM F1 AUC": 0.5442640125720272,
      "eval_PRM F1 Neg": 0.3617021276595745,
      "eval_PRM NPV": 0.23943661971830985,
      "eval_PRM Precision": 0.8285714285714286,
      "eval_PRM Recall": 0.3493975903614458,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.8169389367103577,
      "eval_runtime": 20.8415,
      "eval_samples_per_second": 2.111,
      "eval_steps_per_second": 0.144,
      "step": 665
    },
    {
      "epoch": 0.08622196329740751,
      "grad_norm": 0.23585250344700395,
      "learning_rate": 1.0769728331177232e-06,
      "loss": 1.2513,
      "step": 666
    },
    {
      "epoch": 0.08635142570476098,
      "grad_norm": 0.14952656684145715,
      "learning_rate": 1.0785899094437257e-06,
      "loss": 1.0811,
      "step": 667
    },
    {
      "epoch": 0.08648088811211445,
      "grad_norm": 0.20062848072945358,
      "learning_rate": 1.0802069857697284e-06,
      "loss": 1.201,
      "step": 668
    },
    {
      "epoch": 0.08661035051946792,
      "grad_norm": 0.22367060793159224,
      "learning_rate": 1.081824062095731e-06,
      "loss": 1.2006,
      "step": 669
    },
    {
      "epoch": 0.08673981292682137,
      "grad_norm": 0.2040787667759277,
      "learning_rate": 1.0834411384217337e-06,
      "loss": 1.2313,
      "step": 670
    },
    {
      "epoch": 0.08673981292682137,
      "eval_PRM Accuracy": 0.42452830188679247,
      "eval_PRM F1": 0.47863247863247865,
      "eval_PRM F1 AUC": 0.538239916186485,
      "eval_PRM F1 Neg": 0.35789473684210527,
      "eval_PRM NPV": 0.2361111111111111,
      "eval_PRM Precision": 0.8235294117647058,
      "eval_PRM Recall": 0.3373493975903614,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.8178266882896423,
      "eval_runtime": 20.6776,
      "eval_samples_per_second": 2.128,
      "eval_steps_per_second": 0.145,
      "step": 670
    },
    {
      "epoch": 0.08686927533417484,
      "grad_norm": 0.25526686407750226,
      "learning_rate": 1.0850582147477362e-06,
      "loss": 1.219,
      "step": 671
    },
    {
      "epoch": 0.0869987377415283,
      "grad_norm": 0.13101939037668023,
      "learning_rate": 1.0866752910737388e-06,
      "loss": 1.0942,
      "step": 672
    },
    {
      "epoch": 0.08712820014888177,
      "grad_norm": 0.23585752002615576,
      "learning_rate": 1.0882923673997415e-06,
      "loss": 1.1882,
      "step": 673
    },
    {
      "epoch": 0.08725766255623524,
      "grad_norm": 0.1339957113599109,
      "learning_rate": 1.089909443725744e-06,
      "loss": 1.0079,
      "step": 674
    },
    {
      "epoch": 0.08738712496358869,
      "grad_norm": 0.1925221385017757,
      "learning_rate": 1.0915265200517464e-06,
      "loss": 1.0786,
      "step": 675
    },
    {
      "epoch": 0.08738712496358869,
      "eval_PRM Accuracy": 0.44339622641509435,
      "eval_PRM F1": 0.5042016806722689,
      "eval_PRM F1 AUC": 0.5502881089575693,
      "eval_PRM F1 Neg": 0.3655913978494624,
      "eval_PRM NPV": 0.24285714285714285,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.3614457831325301,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.8139204382896423,
      "eval_runtime": 20.2163,
      "eval_samples_per_second": 2.176,
      "eval_steps_per_second": 0.148,
      "step": 675
    },
    {
      "epoch": 0.08751658737094216,
      "grad_norm": 0.2267659111579892,
      "learning_rate": 1.093143596377749e-06,
      "loss": 1.3188,
      "step": 676
    },
    {
      "epoch": 0.08764604977829563,
      "grad_norm": 0.26228727436253496,
      "learning_rate": 1.0947606727037518e-06,
      "loss": 1.3037,
      "step": 677
    },
    {
      "epoch": 0.0877755121856491,
      "grad_norm": 0.17323508486735067,
      "learning_rate": 1.0963777490297544e-06,
      "loss": 1.2368,
      "step": 678
    },
    {
      "epoch": 0.08790497459300256,
      "grad_norm": 0.2338530487729072,
      "learning_rate": 1.0979948253557569e-06,
      "loss": 1.2469,
      "step": 679
    },
    {
      "epoch": 0.08803443700035603,
      "grad_norm": 0.2447686675850731,
      "learning_rate": 1.0996119016817593e-06,
      "loss": 1.2308,
      "step": 680
    },
    {
      "epoch": 0.08803443700035603,
      "eval_PRM Accuracy": 0.44339622641509435,
      "eval_PRM F1": 0.5042016806722689,
      "eval_PRM F1 AUC": 0.5502881089575693,
      "eval_PRM F1 Neg": 0.3655913978494624,
      "eval_PRM NPV": 0.24285714285714285,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.3614457831325301,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.8110795617103577,
      "eval_runtime": 20.7204,
      "eval_samples_per_second": 2.124,
      "eval_steps_per_second": 0.145,
      "step": 680
    },
    {
      "epoch": 0.08816389940770948,
      "grad_norm": 0.16836979000188007,
      "learning_rate": 1.101228978007762e-06,
      "loss": 1.1305,
      "step": 681
    },
    {
      "epoch": 0.08829336181506295,
      "grad_norm": 0.18045528434385183,
      "learning_rate": 1.1028460543337647e-06,
      "loss": 1.0947,
      "step": 682
    },
    {
      "epoch": 0.08842282422241642,
      "grad_norm": 0.14146749045737927,
      "learning_rate": 1.1044631306597671e-06,
      "loss": 1.0575,
      "step": 683
    },
    {
      "epoch": 0.08855228662976988,
      "grad_norm": 0.14012250655151107,
      "learning_rate": 1.1060802069857698e-06,
      "loss": 1.0856,
      "step": 684
    },
    {
      "epoch": 0.08868174903712335,
      "grad_norm": 0.2417463812326379,
      "learning_rate": 1.1076972833117725e-06,
      "loss": 1.2232,
      "step": 685
    },
    {
      "epoch": 0.08868174903712335,
      "eval_PRM Accuracy": 0.44339622641509435,
      "eval_PRM F1": 0.5042016806722689,
      "eval_PRM F1 AUC": 0.5502881089575693,
      "eval_PRM F1 Neg": 0.3655913978494624,
      "eval_PRM NPV": 0.24285714285714285,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.3614457831325301,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.8099254369735718,
      "eval_runtime": 20.6724,
      "eval_samples_per_second": 2.128,
      "eval_steps_per_second": 0.145,
      "step": 685
    },
    {
      "epoch": 0.0888112114444768,
      "grad_norm": 0.2575672945773416,
      "learning_rate": 1.1093143596377752e-06,
      "loss": 1.173,
      "step": 686
    },
    {
      "epoch": 0.08894067385183027,
      "grad_norm": 0.1739372476969271,
      "learning_rate": 1.1109314359637776e-06,
      "loss": 1.1282,
      "step": 687
    },
    {
      "epoch": 0.08907013625918374,
      "grad_norm": 0.22907972073670896,
      "learning_rate": 1.11254851228978e-06,
      "loss": 1.2397,
      "step": 688
    },
    {
      "epoch": 0.0891995986665372,
      "grad_norm": 0.1586394331462787,
      "learning_rate": 1.1141655886157827e-06,
      "loss": 0.943,
      "step": 689
    },
    {
      "epoch": 0.08932906107389067,
      "grad_norm": 0.17570396617790188,
      "learning_rate": 1.1157826649417854e-06,
      "loss": 1.1677,
      "step": 690
    },
    {
      "epoch": 0.08932906107389067,
      "eval_PRM Accuracy": 0.4339622641509434,
      "eval_PRM F1": 0.4915254237288136,
      "eval_PRM F1 AUC": 0.5442640125720272,
      "eval_PRM F1 Neg": 0.3617021276595745,
      "eval_PRM NPV": 0.23943661971830985,
      "eval_PRM Precision": 0.8285714285714286,
      "eval_PRM Recall": 0.3493975903614458,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.8084161877632141,
      "eval_runtime": 20.7112,
      "eval_samples_per_second": 2.124,
      "eval_steps_per_second": 0.145,
      "step": 690
    },
    {
      "epoch": 0.08945852348124414,
      "grad_norm": 0.2023680290120114,
      "learning_rate": 1.1173997412677879e-06,
      "loss": 1.0775,
      "step": 691
    },
    {
      "epoch": 0.0895879858885976,
      "grad_norm": 0.18025861010631436,
      "learning_rate": 1.1190168175937905e-06,
      "loss": 1.1085,
      "step": 692
    },
    {
      "epoch": 0.08971744829595106,
      "grad_norm": 0.25113789254616725,
      "learning_rate": 1.120633893919793e-06,
      "loss": 1.2981,
      "step": 693
    },
    {
      "epoch": 0.08984691070330453,
      "grad_norm": 0.22379107138037224,
      "learning_rate": 1.1222509702457957e-06,
      "loss": 1.1171,
      "step": 694
    },
    {
      "epoch": 0.089976373110658,
      "grad_norm": 0.21037997112404783,
      "learning_rate": 1.1238680465717983e-06,
      "loss": 1.1964,
      "step": 695
    },
    {
      "epoch": 0.089976373110658,
      "eval_PRM Accuracy": 0.4528301886792453,
      "eval_PRM F1": 0.5166666666666667,
      "eval_PRM F1 AUC": 0.5563122053431115,
      "eval_PRM F1 Neg": 0.3695652173913043,
      "eval_PRM NPV": 0.2463768115942029,
      "eval_PRM Precision": 0.8378378378378378,
      "eval_PRM Recall": 0.37349397590361444,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.8039772510528564,
      "eval_runtime": 20.7496,
      "eval_samples_per_second": 2.121,
      "eval_steps_per_second": 0.145,
      "step": 695
    },
    {
      "epoch": 0.09010583551801146,
      "grad_norm": 0.2606377649490679,
      "learning_rate": 1.1254851228978008e-06,
      "loss": 1.1665,
      "step": 696
    },
    {
      "epoch": 0.09023529792536492,
      "grad_norm": 0.2113235011341728,
      "learning_rate": 1.1271021992238035e-06,
      "loss": 1.2775,
      "step": 697
    },
    {
      "epoch": 0.09036476033271838,
      "grad_norm": 0.2486775979635233,
      "learning_rate": 1.1287192755498062e-06,
      "loss": 1.2161,
      "step": 698
    },
    {
      "epoch": 0.09049422274007185,
      "grad_norm": 0.2222114752489947,
      "learning_rate": 1.1303363518758086e-06,
      "loss": 1.1936,
      "step": 699
    },
    {
      "epoch": 0.09062368514742532,
      "grad_norm": 0.22642323162027564,
      "learning_rate": 1.131953428201811e-06,
      "loss": 1.2134,
      "step": 700
    },
    {
      "epoch": 0.09062368514742532,
      "eval_PRM Accuracy": 0.4528301886792453,
      "eval_PRM F1": 0.5166666666666667,
      "eval_PRM F1 AUC": 0.5563122053431115,
      "eval_PRM F1 Neg": 0.3695652173913043,
      "eval_PRM NPV": 0.2463768115942029,
      "eval_PRM Precision": 0.8378378378378378,
      "eval_PRM Recall": 0.37349397590361444,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.8037996888160706,
      "eval_runtime": 20.5456,
      "eval_samples_per_second": 2.142,
      "eval_steps_per_second": 0.146,
      "step": 700
    },
    {
      "epoch": 0.09075314755477878,
      "grad_norm": 0.2290038933782361,
      "learning_rate": 1.1335705045278137e-06,
      "loss": 1.1831,
      "step": 701
    },
    {
      "epoch": 0.09088260996213225,
      "grad_norm": 0.2025124937799278,
      "learning_rate": 1.1351875808538164e-06,
      "loss": 1.1666,
      "step": 702
    },
    {
      "epoch": 0.0910120723694857,
      "grad_norm": 0.1948554749755235,
      "learning_rate": 1.136804657179819e-06,
      "loss": 1.0779,
      "step": 703
    },
    {
      "epoch": 0.09114153477683917,
      "grad_norm": 0.18392465630639968,
      "learning_rate": 1.1384217335058215e-06,
      "loss": 1.0569,
      "step": 704
    },
    {
      "epoch": 0.09127099718419264,
      "grad_norm": 0.17716806451257086,
      "learning_rate": 1.1400388098318242e-06,
      "loss": 1.1206,
      "step": 705
    },
    {
      "epoch": 0.09127099718419264,
      "eval_PRM Accuracy": 0.44339622641509435,
      "eval_PRM F1": 0.5042016806722689,
      "eval_PRM F1 AUC": 0.5502881089575693,
      "eval_PRM F1 Neg": 0.3655913978494624,
      "eval_PRM NPV": 0.24285714285714285,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.3614457831325301,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.8018465638160706,
      "eval_runtime": 20.4005,
      "eval_samples_per_second": 2.157,
      "eval_steps_per_second": 0.147,
      "step": 705
    },
    {
      "epoch": 0.09140045959154611,
      "grad_norm": 0.16953282281538468,
      "learning_rate": 1.1416558861578267e-06,
      "loss": 1.1445,
      "step": 706
    },
    {
      "epoch": 0.09152992199889957,
      "grad_norm": 0.18977631120383,
      "learning_rate": 1.1432729624838293e-06,
      "loss": 1.0657,
      "step": 707
    },
    {
      "epoch": 0.09165938440625303,
      "grad_norm": 0.1708875913494527,
      "learning_rate": 1.144890038809832e-06,
      "loss": 1.1627,
      "step": 708
    },
    {
      "epoch": 0.0917888468136065,
      "grad_norm": 0.23252154332857472,
      "learning_rate": 1.1465071151358345e-06,
      "loss": 1.103,
      "step": 709
    },
    {
      "epoch": 0.09191830922095996,
      "grad_norm": 0.2165052273751677,
      "learning_rate": 1.1481241914618371e-06,
      "loss": 1.1841,
      "step": 710
    },
    {
      "epoch": 0.09191830922095996,
      "eval_PRM Accuracy": 0.46226415094339623,
      "eval_PRM F1": 0.5289256198347108,
      "eval_PRM F1 AUC": 0.5623363017286537,
      "eval_PRM F1 Neg": 0.37362637362637363,
      "eval_PRM NPV": 0.25,
      "eval_PRM Precision": 0.8421052631578947,
      "eval_PRM Recall": 0.3855421686746988,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.7997159361839294,
      "eval_runtime": 20.7977,
      "eval_samples_per_second": 2.116,
      "eval_steps_per_second": 0.144,
      "step": 710
    },
    {
      "epoch": 0.09204777162831343,
      "grad_norm": 0.21766521026264884,
      "learning_rate": 1.1497412677878398e-06,
      "loss": 1.2906,
      "step": 711
    },
    {
      "epoch": 0.0921772340356669,
      "grad_norm": 0.22197304320077468,
      "learning_rate": 1.1513583441138423e-06,
      "loss": 1.089,
      "step": 712
    },
    {
      "epoch": 0.09230669644302036,
      "grad_norm": 0.2238875762240465,
      "learning_rate": 1.1529754204398447e-06,
      "loss": 1.2035,
      "step": 713
    },
    {
      "epoch": 0.09243615885037382,
      "grad_norm": 0.1954242348039915,
      "learning_rate": 1.1545924967658474e-06,
      "loss": 1.1066,
      "step": 714
    },
    {
      "epoch": 0.09256562125772728,
      "grad_norm": 0.2232933614698844,
      "learning_rate": 1.15620957309185e-06,
      "loss": 1.1366,
      "step": 715
    },
    {
      "epoch": 0.09256562125772728,
      "eval_PRM Accuracy": 0.46226415094339623,
      "eval_PRM F1": 0.5289256198347108,
      "eval_PRM F1 AUC": 0.5623363017286537,
      "eval_PRM F1 Neg": 0.37362637362637363,
      "eval_PRM NPV": 0.25,
      "eval_PRM Precision": 0.8421052631578947,
      "eval_PRM Recall": 0.3855421686746988,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.7931463122367859,
      "eval_runtime": 20.7705,
      "eval_samples_per_second": 2.118,
      "eval_steps_per_second": 0.144,
      "step": 715
    },
    {
      "epoch": 0.09269508366508075,
      "grad_norm": 0.28404114622284343,
      "learning_rate": 1.1578266494178527e-06,
      "loss": 1.142,
      "step": 716
    },
    {
      "epoch": 0.09282454607243422,
      "grad_norm": 0.23132211829055413,
      "learning_rate": 1.1594437257438552e-06,
      "loss": 1.1134,
      "step": 717
    },
    {
      "epoch": 0.09295400847978769,
      "grad_norm": 0.254491916302133,
      "learning_rate": 1.1610608020698579e-06,
      "loss": 1.1635,
      "step": 718
    },
    {
      "epoch": 0.09308347088714114,
      "grad_norm": 0.22014051101893578,
      "learning_rate": 1.1626778783958603e-06,
      "loss": 1.2073,
      "step": 719
    },
    {
      "epoch": 0.09321293329449461,
      "grad_norm": 0.26117122499873874,
      "learning_rate": 1.164294954721863e-06,
      "loss": 1.2126,
      "step": 720
    },
    {
      "epoch": 0.09321293329449461,
      "eval_PRM Accuracy": 0.46226415094339623,
      "eval_PRM F1": 0.5289256198347108,
      "eval_PRM F1 AUC": 0.5623363017286537,
      "eval_PRM F1 Neg": 0.37362637362637363,
      "eval_PRM NPV": 0.25,
      "eval_PRM Precision": 0.8421052631578947,
      "eval_PRM Recall": 0.3855421686746988,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.7950994372367859,
      "eval_runtime": 20.3555,
      "eval_samples_per_second": 2.162,
      "eval_steps_per_second": 0.147,
      "step": 720
    },
    {
      "epoch": 0.09334239570184807,
      "grad_norm": 0.1762444227384742,
      "learning_rate": 1.1659120310478655e-06,
      "loss": 1.0873,
      "step": 721
    },
    {
      "epoch": 0.09347185810920154,
      "grad_norm": 0.2185467195206546,
      "learning_rate": 1.1675291073738681e-06,
      "loss": 1.2135,
      "step": 722
    },
    {
      "epoch": 0.09360132051655501,
      "grad_norm": 0.193551395954972,
      "learning_rate": 1.1691461836998708e-06,
      "loss": 1.0702,
      "step": 723
    },
    {
      "epoch": 0.09373078292390848,
      "grad_norm": 0.20484451560623293,
      "learning_rate": 1.1707632600258735e-06,
      "loss": 1.0972,
      "step": 724
    },
    {
      "epoch": 0.09386024533126193,
      "grad_norm": 0.20449170093654723,
      "learning_rate": 1.172380336351876e-06,
      "loss": 1.1573,
      "step": 725
    },
    {
      "epoch": 0.09386024533126193,
      "eval_PRM Accuracy": 0.46226415094339623,
      "eval_PRM F1": 0.5289256198347108,
      "eval_PRM F1 AUC": 0.5623363017286537,
      "eval_PRM F1 Neg": 0.37362637362637363,
      "eval_PRM NPV": 0.25,
      "eval_PRM Precision": 0.8421052631578947,
      "eval_PRM Recall": 0.3855421686746988,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.7913707494735718,
      "eval_runtime": 20.5034,
      "eval_samples_per_second": 2.146,
      "eval_steps_per_second": 0.146,
      "step": 725
    },
    {
      "epoch": 0.0939897077386154,
      "grad_norm": 0.23602814734352026,
      "learning_rate": 1.1739974126778784e-06,
      "loss": 1.1295,
      "step": 726
    },
    {
      "epoch": 0.09411917014596886,
      "grad_norm": 0.25020733076136553,
      "learning_rate": 1.175614489003881e-06,
      "loss": 1.2802,
      "step": 727
    },
    {
      "epoch": 0.09424863255332233,
      "grad_norm": 0.21298226598414483,
      "learning_rate": 1.1772315653298837e-06,
      "loss": 1.0381,
      "step": 728
    },
    {
      "epoch": 0.0943780949606758,
      "grad_norm": 0.20291897193281175,
      "learning_rate": 1.1788486416558862e-06,
      "loss": 1.1039,
      "step": 729
    },
    {
      "epoch": 0.09450755736802925,
      "grad_norm": 0.21533692228539952,
      "learning_rate": 1.1804657179818889e-06,
      "loss": 1.1329,
      "step": 730
    },
    {
      "epoch": 0.09450755736802925,
      "eval_PRM Accuracy": 0.46226415094339623,
      "eval_PRM F1": 0.5289256198347108,
      "eval_PRM F1 AUC": 0.5623363017286537,
      "eval_PRM F1 Neg": 0.37362637362637363,
      "eval_PRM NPV": 0.25,
      "eval_PRM Precision": 0.8421052631578947,
      "eval_PRM Recall": 0.3855421686746988,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.7887073755264282,
      "eval_runtime": 20.7739,
      "eval_samples_per_second": 2.118,
      "eval_steps_per_second": 0.144,
      "step": 730
    },
    {
      "epoch": 0.09463701977538272,
      "grad_norm": 0.22493749360228335,
      "learning_rate": 1.1820827943078915e-06,
      "loss": 1.1873,
      "step": 731
    },
    {
      "epoch": 0.09476648218273619,
      "grad_norm": 0.25258338541134534,
      "learning_rate": 1.183699870633894e-06,
      "loss": 1.2369,
      "step": 732
    },
    {
      "epoch": 0.09489594459008965,
      "grad_norm": 0.24502143906121532,
      "learning_rate": 1.1853169469598967e-06,
      "loss": 1.2261,
      "step": 733
    },
    {
      "epoch": 0.09502540699744312,
      "grad_norm": 0.19867196072302493,
      "learning_rate": 1.1869340232858991e-06,
      "loss": 1.0807,
      "step": 734
    },
    {
      "epoch": 0.09515486940479659,
      "grad_norm": 0.21560138528027784,
      "learning_rate": 1.1885510996119018e-06,
      "loss": 1.1527,
      "step": 735
    },
    {
      "epoch": 0.09515486940479659,
      "eval_PRM Accuracy": 0.46226415094339623,
      "eval_PRM F1": 0.5289256198347108,
      "eval_PRM F1 AUC": 0.5623363017286537,
      "eval_PRM F1 Neg": 0.37362637362637363,
      "eval_PRM NPV": 0.25,
      "eval_PRM Precision": 0.8421052631578947,
      "eval_PRM Recall": 0.3855421686746988,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.7887073755264282,
      "eval_runtime": 20.357,
      "eval_samples_per_second": 2.161,
      "eval_steps_per_second": 0.147,
      "step": 735
    },
    {
      "epoch": 0.09528433181215004,
      "grad_norm": 0.26988956826763044,
      "learning_rate": 1.1901681759379045e-06,
      "loss": 1.1559,
      "step": 736
    },
    {
      "epoch": 0.09541379421950351,
      "grad_norm": 0.23213832317126168,
      "learning_rate": 1.191785252263907e-06,
      "loss": 1.2421,
      "step": 737
    },
    {
      "epoch": 0.09554325662685698,
      "grad_norm": 0.2548441415698318,
      "learning_rate": 1.1934023285899094e-06,
      "loss": 1.1719,
      "step": 738
    },
    {
      "epoch": 0.09567271903421044,
      "grad_norm": 0.17831907994299862,
      "learning_rate": 1.195019404915912e-06,
      "loss": 1.1624,
      "step": 739
    },
    {
      "epoch": 0.09580218144156391,
      "grad_norm": 0.2118395048352656,
      "learning_rate": 1.1966364812419147e-06,
      "loss": 1.1143,
      "step": 740
    },
    {
      "epoch": 0.09580218144156391,
      "eval_PRM Accuracy": 0.46226415094339623,
      "eval_PRM F1": 0.5289256198347108,
      "eval_PRM F1 AUC": 0.5623363017286537,
      "eval_PRM F1 Neg": 0.37362637362637363,
      "eval_PRM NPV": 0.25,
      "eval_PRM Precision": 0.8421052631578947,
      "eval_PRM Recall": 0.3855421686746988,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.7819602489471436,
      "eval_runtime": 20.1834,
      "eval_samples_per_second": 2.18,
      "eval_steps_per_second": 0.149,
      "step": 740
    },
    {
      "epoch": 0.09593164384891736,
      "grad_norm": 0.22498359346494612,
      "learning_rate": 1.1982535575679174e-06,
      "loss": 1.1449,
      "step": 741
    },
    {
      "epoch": 0.09606110625627083,
      "grad_norm": 0.25679804483242924,
      "learning_rate": 1.1998706338939199e-06,
      "loss": 1.1763,
      "step": 742
    },
    {
      "epoch": 0.0961905686636243,
      "grad_norm": 0.21622985834388322,
      "learning_rate": 1.2014877102199225e-06,
      "loss": 1.0768,
      "step": 743
    },
    {
      "epoch": 0.09632003107097777,
      "grad_norm": 0.28229960043745805,
      "learning_rate": 1.2031047865459252e-06,
      "loss": 1.073,
      "step": 744
    },
    {
      "epoch": 0.09644949347833123,
      "grad_norm": 0.22607886227043467,
      "learning_rate": 1.2047218628719277e-06,
      "loss": 1.1025,
      "step": 745
    },
    {
      "epoch": 0.09644949347833123,
      "eval_PRM Accuracy": 0.4528301886792453,
      "eval_PRM F1": 0.5245901639344263,
      "eval_PRM F1 AUC": 0.5405971712938711,
      "eval_PRM F1 Neg": 0.35555555555555557,
      "eval_PRM NPV": 0.23880597014925373,
      "eval_PRM Precision": 0.8205128205128205,
      "eval_PRM Recall": 0.3855421686746988,
      "eval_PRM Specificty": 0.6956521739130435,
      "eval_loss": 0.7828480005264282,
      "eval_runtime": 20.8025,
      "eval_samples_per_second": 2.115,
      "eval_steps_per_second": 0.144,
      "step": 745
    },
    {
      "epoch": 0.0965789558856847,
      "grad_norm": 0.21547207796355666,
      "learning_rate": 1.2063389391979301e-06,
      "loss": 1.0645,
      "step": 746
    },
    {
      "epoch": 0.09670841829303815,
      "grad_norm": 0.23769599229538318,
      "learning_rate": 1.2079560155239328e-06,
      "loss": 1.0543,
      "step": 747
    },
    {
      "epoch": 0.09683788070039162,
      "grad_norm": 0.19792986678409308,
      "learning_rate": 1.2095730918499355e-06,
      "loss": 1.0785,
      "step": 748
    },
    {
      "epoch": 0.09696734310774509,
      "grad_norm": 0.2724932565635008,
      "learning_rate": 1.2111901681759381e-06,
      "loss": 1.2177,
      "step": 749
    },
    {
      "epoch": 0.09709680551509856,
      "grad_norm": 0.2754120410788316,
      "learning_rate": 1.2128072445019406e-06,
      "loss": 1.2132,
      "step": 750
    },
    {
      "epoch": 0.09709680551509856,
      "eval_PRM Accuracy": 0.46226415094339623,
      "eval_PRM F1": 0.5289256198347108,
      "eval_PRM F1 AUC": 0.5623363017286537,
      "eval_PRM F1 Neg": 0.37362637362637363,
      "eval_PRM NPV": 0.25,
      "eval_PRM Precision": 0.8421052631578947,
      "eval_PRM Recall": 0.3855421686746988,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.7798295617103577,
      "eval_runtime": 20.3,
      "eval_samples_per_second": 2.167,
      "eval_steps_per_second": 0.148,
      "step": 750
    },
    {
      "epoch": 0.09722626792245202,
      "grad_norm": 0.25180119882669694,
      "learning_rate": 1.214424320827943e-06,
      "loss": 1.1548,
      "step": 751
    },
    {
      "epoch": 0.09735573032980548,
      "grad_norm": 0.20404790639392936,
      "learning_rate": 1.2160413971539457e-06,
      "loss": 1.182,
      "step": 752
    },
    {
      "epoch": 0.09748519273715894,
      "grad_norm": 0.1735372404065468,
      "learning_rate": 1.2176584734799484e-06,
      "loss": 1.1332,
      "step": 753
    },
    {
      "epoch": 0.09761465514451241,
      "grad_norm": 0.23075240046720094,
      "learning_rate": 1.219275549805951e-06,
      "loss": 1.0579,
      "step": 754
    },
    {
      "epoch": 0.09774411755186588,
      "grad_norm": 0.23809237937129,
      "learning_rate": 1.2208926261319535e-06,
      "loss": 1.1243,
      "step": 755
    },
    {
      "epoch": 0.09774411755186588,
      "eval_PRM Accuracy": 0.46226415094339623,
      "eval_PRM F1": 0.5289256198347108,
      "eval_PRM F1 AUC": 0.5623363017286537,
      "eval_PRM F1 Neg": 0.37362637362637363,
      "eval_PRM NPV": 0.25,
      "eval_PRM Precision": 0.8421052631578947,
      "eval_PRM Recall": 0.3855421686746988,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.7787641882896423,
      "eval_runtime": 20.3263,
      "eval_samples_per_second": 2.165,
      "eval_steps_per_second": 0.148,
      "step": 755
    },
    {
      "epoch": 0.09787357995921935,
      "grad_norm": 0.26256074388894685,
      "learning_rate": 1.2225097024579562e-06,
      "loss": 1.2206,
      "step": 756
    },
    {
      "epoch": 0.09800304236657281,
      "grad_norm": 0.26459955124630286,
      "learning_rate": 1.2241267787839586e-06,
      "loss": 1.1464,
      "step": 757
    },
    {
      "epoch": 0.09813250477392627,
      "grad_norm": 0.17619686525138678,
      "learning_rate": 1.2257438551099613e-06,
      "loss": 1.1153,
      "step": 758
    },
    {
      "epoch": 0.09826196718127973,
      "grad_norm": 0.20591196202519804,
      "learning_rate": 1.2273609314359638e-06,
      "loss": 1.1368,
      "step": 759
    },
    {
      "epoch": 0.0983914295886332,
      "grad_norm": 0.2758261439083429,
      "learning_rate": 1.2289780077619664e-06,
      "loss": 1.2095,
      "step": 760
    },
    {
      "epoch": 0.0983914295886332,
      "eval_PRM Accuracy": 0.4716981132075472,
      "eval_PRM F1": 0.5409836065573771,
      "eval_PRM F1 AUC": 0.5683603981141959,
      "eval_PRM F1 Neg": 0.37777777777777777,
      "eval_PRM NPV": 0.2537313432835821,
      "eval_PRM Precision": 0.8461538461538461,
      "eval_PRM Recall": 0.39759036144578314,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.7729048132896423,
      "eval_runtime": 21.2037,
      "eval_samples_per_second": 2.075,
      "eval_steps_per_second": 0.141,
      "step": 760
    },
    {
      "epoch": 0.09852089199598667,
      "grad_norm": 0.22755765869153188,
      "learning_rate": 1.2305950840879691e-06,
      "loss": 1.1379,
      "step": 761
    },
    {
      "epoch": 0.09865035440334013,
      "grad_norm": 0.26708732476154073,
      "learning_rate": 1.2322121604139718e-06,
      "loss": 1.1426,
      "step": 762
    },
    {
      "epoch": 0.09877981681069359,
      "grad_norm": 0.16798937659854712,
      "learning_rate": 1.2338292367399742e-06,
      "loss": 1.0905,
      "step": 763
    },
    {
      "epoch": 0.09890927921804706,
      "grad_norm": 0.23179263637874012,
      "learning_rate": 1.2354463130659767e-06,
      "loss": 1.1682,
      "step": 764
    },
    {
      "epoch": 0.09903874162540052,
      "grad_norm": 0.2584907967765254,
      "learning_rate": 1.2370633893919794e-06,
      "loss": 1.1705,
      "step": 765
    },
    {
      "epoch": 0.09903874162540052,
      "eval_PRM Accuracy": 0.46226415094339623,
      "eval_PRM F1": 0.5365853658536586,
      "eval_PRM F1 AUC": 0.5466212676794133,
      "eval_PRM F1 Neg": 0.3595505617977528,
      "eval_PRM NPV": 0.24242424242424243,
      "eval_PRM Precision": 0.825,
      "eval_PRM Recall": 0.39759036144578314,
      "eval_PRM Specificty": 0.6956521739130435,
      "eval_loss": 0.7745028138160706,
      "eval_runtime": 20.6353,
      "eval_samples_per_second": 2.132,
      "eval_steps_per_second": 0.145,
      "step": 765
    },
    {
      "epoch": 0.09916820403275399,
      "grad_norm": 0.18909765975689427,
      "learning_rate": 1.238680465717982e-06,
      "loss": 1.163,
      "step": 766
    },
    {
      "epoch": 0.09929766644010746,
      "grad_norm": 0.21810232020049436,
      "learning_rate": 1.2402975420439845e-06,
      "loss": 1.0478,
      "step": 767
    },
    {
      "epoch": 0.09942712884746092,
      "grad_norm": 0.2143941439032801,
      "learning_rate": 1.2419146183699872e-06,
      "loss": 1.137,
      "step": 768
    },
    {
      "epoch": 0.09955659125481438,
      "grad_norm": 0.2206534494147366,
      "learning_rate": 1.2435316946959898e-06,
      "loss": 1.2112,
      "step": 769
    },
    {
      "epoch": 0.09968605366216785,
      "grad_norm": 0.11009506674702682,
      "learning_rate": 1.2451487710219923e-06,
      "loss": 0.9921,
      "step": 770
    },
    {
      "epoch": 0.09968605366216785,
      "eval_PRM Accuracy": 0.46226415094339623,
      "eval_PRM F1": 0.5365853658536586,
      "eval_PRM F1 AUC": 0.5466212676794133,
      "eval_PRM F1 Neg": 0.3595505617977528,
      "eval_PRM NPV": 0.24242424242424243,
      "eval_PRM Precision": 0.825,
      "eval_PRM Recall": 0.39759036144578314,
      "eval_PRM Specificty": 0.6956521739130435,
      "eval_loss": 0.7700639367103577,
      "eval_runtime": 20.7333,
      "eval_samples_per_second": 2.122,
      "eval_steps_per_second": 0.145,
      "step": 770
    },
    {
      "epoch": 0.09981551606952131,
      "grad_norm": 0.1726759440108128,
      "learning_rate": 1.246765847347995e-06,
      "loss": 1.1448,
      "step": 771
    },
    {
      "epoch": 0.09994497847687478,
      "grad_norm": 0.18023330118750003,
      "learning_rate": 1.2483829236739974e-06,
      "loss": 1.0341,
      "step": 772
    },
    {
      "epoch": 0.10007444088422825,
      "grad_norm": 0.17997573828682634,
      "learning_rate": 1.25e-06,
      "loss": 1.1392,
      "step": 773
    },
    {
      "epoch": 0.1002039032915817,
      "grad_norm": 0.2947517874316338,
      "learning_rate": 1.249999936165541e-06,
      "loss": 1.2645,
      "step": 774
    },
    {
      "epoch": 0.10033336569893517,
      "grad_norm": 0.1634906643517373,
      "learning_rate": 1.2499997446621767e-06,
      "loss": 1.1113,
      "step": 775
    },
    {
      "epoch": 0.10033336569893517,
      "eval_PRM Accuracy": 0.4716981132075472,
      "eval_PRM F1": 0.5483870967741935,
      "eval_PRM F1 AUC": 0.5526453640649555,
      "eval_PRM F1 Neg": 0.36363636363636365,
      "eval_PRM NPV": 0.24615384615384617,
      "eval_PRM Precision": 0.8292682926829268,
      "eval_PRM Recall": 0.40963855421686746,
      "eval_PRM Specificty": 0.6956521739130435,
      "eval_loss": 0.7691761255264282,
      "eval_runtime": 20.23,
      "eval_samples_per_second": 2.175,
      "eval_steps_per_second": 0.148,
      "step": 775
    },
    {
      "epoch": 0.10046282810628863,
      "grad_norm": 0.32602478038973365,
      "learning_rate": 1.2499994254899465e-06,
      "loss": 1.323,
      "step": 776
    },
    {
      "epoch": 0.1005922905136421,
      "grad_norm": 0.20179936293072492,
      "learning_rate": 1.2499989786489155e-06,
      "loss": 1.1453,
      "step": 777
    },
    {
      "epoch": 0.10072175292099557,
      "grad_norm": 0.20577352481423622,
      "learning_rate": 1.249998404139175e-06,
      "loss": 1.181,
      "step": 778
    },
    {
      "epoch": 0.10085121532834904,
      "grad_norm": 0.24750993568770668,
      "learning_rate": 1.2499977019608421e-06,
      "loss": 1.1901,
      "step": 779
    },
    {
      "epoch": 0.10098067773570249,
      "grad_norm": 0.11606933585077482,
      "learning_rate": 1.2499968721140606e-06,
      "loss": 0.9203,
      "step": 780
    },
    {
      "epoch": 0.10098067773570249,
      "eval_PRM Accuracy": 0.4716981132075472,
      "eval_PRM F1": 0.5483870967741935,
      "eval_PRM F1 AUC": 0.5526453640649555,
      "eval_PRM F1 Neg": 0.36363636363636365,
      "eval_PRM NPV": 0.24615384615384617,
      "eval_PRM Precision": 0.8292682926829268,
      "eval_PRM Recall": 0.40963855421686746,
      "eval_PRM Specificty": 0.6956521739130435,
      "eval_loss": 0.7666903138160706,
      "eval_runtime": 20.1493,
      "eval_samples_per_second": 2.184,
      "eval_steps_per_second": 0.149,
      "step": 780
    },
    {
      "epoch": 0.10111014014305596,
      "grad_norm": 0.27554214708355385,
      "learning_rate": 1.2499959145989998e-06,
      "loss": 1.1538,
      "step": 781
    },
    {
      "epoch": 0.10123960255040942,
      "grad_norm": 0.29854874437792533,
      "learning_rate": 1.2499948294158554e-06,
      "loss": 1.1133,
      "step": 782
    },
    {
      "epoch": 0.10136906495776289,
      "grad_norm": 0.1461415097520277,
      "learning_rate": 1.249993616564849e-06,
      "loss": 1.1016,
      "step": 783
    },
    {
      "epoch": 0.10149852736511636,
      "grad_norm": 0.1899700959275549,
      "learning_rate": 1.2499922760462283e-06,
      "loss": 1.0471,
      "step": 784
    },
    {
      "epoch": 0.10162798977246981,
      "grad_norm": 0.20130272206953717,
      "learning_rate": 1.2499908078602673e-06,
      "loss": 1.0713,
      "step": 785
    },
    {
      "epoch": 0.10162798977246981,
      "eval_PRM Accuracy": 0.4811320754716981,
      "eval_PRM F1": 0.56,
      "eval_PRM F1 AUC": 0.5586694604504976,
      "eval_PRM F1 Neg": 0.367816091954023,
      "eval_PRM NPV": 0.25,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.42168674698795183,
      "eval_PRM Specificty": 0.6956521739130435,
      "eval_loss": 0.7631391882896423,
      "eval_runtime": 19.6226,
      "eval_samples_per_second": 2.242,
      "eval_steps_per_second": 0.153,
      "step": 785
    },
    {
      "epoch": 0.10175745217982328,
      "grad_norm": 0.18990503731956368,
      "learning_rate": 1.249989212007266e-06,
      "loss": 1.0071,
      "step": 786
    },
    {
      "epoch": 0.10188691458717675,
      "grad_norm": 0.22835499308961776,
      "learning_rate": 1.2499874884875499e-06,
      "loss": 1.0748,
      "step": 787
    },
    {
      "epoch": 0.10201637699453021,
      "grad_norm": 0.21772889024471295,
      "learning_rate": 1.2499856373014715e-06,
      "loss": 1.1852,
      "step": 788
    },
    {
      "epoch": 0.10214583940188368,
      "grad_norm": 0.20171428516426992,
      "learning_rate": 1.2499836584494087e-06,
      "loss": 1.1459,
      "step": 789
    },
    {
      "epoch": 0.10227530180923715,
      "grad_norm": 0.26180551538176716,
      "learning_rate": 1.2499815519317659e-06,
      "loss": 1.0314,
      "step": 790
    },
    {
      "epoch": 0.10227530180923715,
      "eval_PRM Accuracy": 0.46226415094339623,
      "eval_PRM F1": 0.544,
      "eval_PRM F1 AUC": 0.5309062336301729,
      "eval_PRM F1 Neg": 0.3448275862068966,
      "eval_PRM NPV": 0.234375,
      "eval_PRM Precision": 0.8095238095238095,
      "eval_PRM Recall": 0.40963855421686746,
      "eval_PRM Specificty": 0.6521739130434783,
      "eval_loss": 0.7615411877632141,
      "eval_runtime": 19.8179,
      "eval_samples_per_second": 2.22,
      "eval_steps_per_second": 0.151,
      "step": 790
    },
    {
      "epoch": 0.1024047642165906,
      "grad_norm": 0.30427271253560895,
      "learning_rate": 1.2499793177489731e-06,
      "loss": 1.1116,
      "step": 791
    },
    {
      "epoch": 0.10253422662394407,
      "grad_norm": 0.2871261634016219,
      "learning_rate": 1.249976955901487e-06,
      "loss": 1.2562,
      "step": 792
    },
    {
      "epoch": 0.10266368903129754,
      "grad_norm": 0.2867264385576825,
      "learning_rate": 1.2499744663897902e-06,
      "loss": 1.1908,
      "step": 793
    },
    {
      "epoch": 0.102793151438651,
      "grad_norm": 0.2195857276311651,
      "learning_rate": 1.2499718492143908e-06,
      "loss": 1.1089,
      "step": 794
    },
    {
      "epoch": 0.10292261384600447,
      "grad_norm": 0.18002405653809045,
      "learning_rate": 1.2499691043758235e-06,
      "loss": 0.9922,
      "step": 795
    },
    {
      "epoch": 0.10292261384600447,
      "eval_PRM Accuracy": 0.4716981132075472,
      "eval_PRM F1": 0.5555555555555556,
      "eval_PRM F1 AUC": 0.536930330015715,
      "eval_PRM F1 Neg": 0.3488372093023256,
      "eval_PRM NPV": 0.23809523809523808,
      "eval_PRM Precision": 0.813953488372093,
      "eval_PRM Recall": 0.42168674698795183,
      "eval_PRM Specificty": 0.6521739130434783,
      "eval_loss": 0.755859375,
      "eval_runtime": 20.1458,
      "eval_samples_per_second": 2.184,
      "eval_steps_per_second": 0.149,
      "step": 795
    },
    {
      "epoch": 0.10305207625335792,
      "grad_norm": 0.26273441011096743,
      "learning_rate": 1.2499662318746493e-06,
      "loss": 1.1824,
      "step": 796
    },
    {
      "epoch": 0.10318153866071139,
      "grad_norm": 0.2035331448092164,
      "learning_rate": 1.2499632317114545e-06,
      "loss": 1.1109,
      "step": 797
    },
    {
      "epoch": 0.10331100106806486,
      "grad_norm": 0.23762720233613846,
      "learning_rate": 1.2499601038868525e-06,
      "loss": 1.1183,
      "step": 798
    },
    {
      "epoch": 0.10344046347541833,
      "grad_norm": 0.22516423721188877,
      "learning_rate": 1.2499568484014818e-06,
      "loss": 1.1132,
      "step": 799
    },
    {
      "epoch": 0.1035699258827718,
      "grad_norm": 0.23017321219911982,
      "learning_rate": 1.2499534652560074e-06,
      "loss": 1.2016,
      "step": 800
    },
    {
      "epoch": 0.1035699258827718,
      "eval_PRM Accuracy": 0.4716981132075472,
      "eval_PRM F1": 0.5555555555555556,
      "eval_PRM F1 AUC": 0.536930330015715,
      "eval_PRM F1 Neg": 0.3488372093023256,
      "eval_PRM NPV": 0.23809523809523808,
      "eval_PRM Precision": 0.813953488372093,
      "eval_PRM Recall": 0.42168674698795183,
      "eval_PRM Specificty": 0.6521739130434783,
      "eval_loss": 0.7574573755264282,
      "eval_runtime": 20.6817,
      "eval_samples_per_second": 2.127,
      "eval_steps_per_second": 0.145,
      "step": 800
    },
    {
      "epoch": 0.10369938829012526,
      "grad_norm": 0.2661833244040786,
      "learning_rate": 1.2499499544511205e-06,
      "loss": 1.1356,
      "step": 801
    },
    {
      "epoch": 0.10382885069747871,
      "grad_norm": 0.19677229037189647,
      "learning_rate": 1.2499463159875382e-06,
      "loss": 1.0425,
      "step": 802
    },
    {
      "epoch": 0.10395831310483218,
      "grad_norm": 0.23335311818799653,
      "learning_rate": 1.249942549866004e-06,
      "loss": 1.1335,
      "step": 803
    },
    {
      "epoch": 0.10408777551218565,
      "grad_norm": 0.16858199718321024,
      "learning_rate": 1.2499386560872869e-06,
      "loss": 1.028,
      "step": 804
    },
    {
      "epoch": 0.10421723791953912,
      "grad_norm": 0.12541411018143828,
      "learning_rate": 1.2499346346521823e-06,
      "loss": 0.9755,
      "step": 805
    },
    {
      "epoch": 0.10421723791953912,
      "eval_PRM Accuracy": 0.4811320754716981,
      "eval_PRM F1": 0.5669291338582677,
      "eval_PRM F1 AUC": 0.5429544264012572,
      "eval_PRM F1 Neg": 0.35294117647058826,
      "eval_PRM NPV": 0.24193548387096775,
      "eval_PRM Precision": 0.8181818181818182,
      "eval_PRM Recall": 0.43373493975903615,
      "eval_PRM Specificty": 0.6521739130434783,
      "eval_loss": 0.7537286877632141,
      "eval_runtime": 20.1429,
      "eval_samples_per_second": 2.184,
      "eval_steps_per_second": 0.149,
      "step": 805
    },
    {
      "epoch": 0.10434670032689258,
      "grad_norm": 0.27581313021921944,
      "learning_rate": 1.249930485561512e-06,
      "loss": 1.1312,
      "step": 806
    },
    {
      "epoch": 0.10447616273424604,
      "grad_norm": 0.21095757866507753,
      "learning_rate": 1.2499262088161227e-06,
      "loss": 0.9894,
      "step": 807
    },
    {
      "epoch": 0.1046056251415995,
      "grad_norm": 0.28978489914227834,
      "learning_rate": 1.249921804416889e-06,
      "loss": 1.1943,
      "step": 808
    },
    {
      "epoch": 0.10473508754895297,
      "grad_norm": 0.29046998993172607,
      "learning_rate": 1.24991727236471e-06,
      "loss": 1.2877,
      "step": 809
    },
    {
      "epoch": 0.10486454995630644,
      "grad_norm": 0.29162852816636675,
      "learning_rate": 1.2499126126605117e-06,
      "loss": 1.1349,
      "step": 810
    },
    {
      "epoch": 0.10486454995630644,
      "eval_PRM Accuracy": 0.4811320754716981,
      "eval_PRM F1": 0.5669291338582677,
      "eval_PRM F1 AUC": 0.5429544264012572,
      "eval_PRM F1 Neg": 0.35294117647058826,
      "eval_PRM NPV": 0.24193548387096775,
      "eval_PRM Precision": 0.8181818181818182,
      "eval_PRM Recall": 0.43373493975903615,
      "eval_PRM Specificty": 0.6521739130434783,
      "eval_loss": 0.7535511255264282,
      "eval_runtime": 20.7589,
      "eval_samples_per_second": 2.12,
      "eval_steps_per_second": 0.145,
      "step": 810
    },
    {
      "epoch": 0.1049940123636599,
      "grad_norm": 0.2749043529240586,
      "learning_rate": 1.2499078253052457e-06,
      "loss": 1.0829,
      "step": 811
    },
    {
      "epoch": 0.10512347477101337,
      "grad_norm": 0.2780284760895125,
      "learning_rate": 1.2499029102998902e-06,
      "loss": 1.1619,
      "step": 812
    },
    {
      "epoch": 0.10525293717836683,
      "grad_norm": 0.2734683706584518,
      "learning_rate": 1.2498978676454491e-06,
      "loss": 1.1861,
      "step": 813
    },
    {
      "epoch": 0.1053823995857203,
      "grad_norm": 0.21662488251312262,
      "learning_rate": 1.2498926973429524e-06,
      "loss": 1.1172,
      "step": 814
    },
    {
      "epoch": 0.10551186199307376,
      "grad_norm": 0.19336049556245302,
      "learning_rate": 1.2498873993934564e-06,
      "loss": 1.1609,
      "step": 815
    },
    {
      "epoch": 0.10551186199307376,
      "eval_PRM Accuracy": 0.4811320754716981,
      "eval_PRM F1": 0.5669291338582677,
      "eval_PRM F1 AUC": 0.5429544264012572,
      "eval_PRM F1 Neg": 0.35294117647058826,
      "eval_PRM NPV": 0.24193548387096775,
      "eval_PRM Precision": 0.8181818181818182,
      "eval_PRM Recall": 0.43373493975903615,
      "eval_PRM Specificty": 0.6521739130434783,
      "eval_loss": 0.7489346861839294,
      "eval_runtime": 20.3461,
      "eval_samples_per_second": 2.163,
      "eval_steps_per_second": 0.147,
      "step": 815
    },
    {
      "epoch": 0.10564132440042723,
      "grad_norm": 0.25441193174545484,
      "learning_rate": 1.2498819737980431e-06,
      "loss": 1.1665,
      "step": 816
    },
    {
      "epoch": 0.1057707868077807,
      "grad_norm": 0.2385243915678446,
      "learning_rate": 1.2498764205578206e-06,
      "loss": 1.1099,
      "step": 817
    },
    {
      "epoch": 0.10590024921513416,
      "grad_norm": 0.1927430459602009,
      "learning_rate": 1.249870739673924e-06,
      "loss": 0.9925,
      "step": 818
    },
    {
      "epoch": 0.10602971162248762,
      "grad_norm": 0.1462093731654415,
      "learning_rate": 1.249864931147513e-06,
      "loss": 1.0054,
      "step": 819
    },
    {
      "epoch": 0.10615917402984108,
      "grad_norm": 0.3495356342249866,
      "learning_rate": 1.2498589949797742e-06,
      "loss": 1.2188,
      "step": 820
    },
    {
      "epoch": 0.10615917402984108,
      "eval_PRM Accuracy": 0.49056603773584906,
      "eval_PRM F1": 0.578125,
      "eval_PRM F1 AUC": 0.5489785227867994,
      "eval_PRM F1 Neg": 0.35714285714285715,
      "eval_PRM NPV": 0.2459016393442623,
      "eval_PRM Precision": 0.8222222222222222,
      "eval_PRM Recall": 0.4457831325301205,
      "eval_PRM Specificty": 0.6521739130434783,
      "eval_loss": 0.7455610632896423,
      "eval_runtime": 20.5506,
      "eval_samples_per_second": 2.141,
      "eval_steps_per_second": 0.146,
      "step": 820
    },
    {
      "epoch": 0.10628863643719455,
      "grad_norm": 0.27414008357139674,
      "learning_rate": 1.2498529311719207e-06,
      "loss": 1.0292,
      "step": 821
    },
    {
      "epoch": 0.10641809884454802,
      "grad_norm": 0.19845022130664222,
      "learning_rate": 1.2498467397251908e-06,
      "loss": 1.1018,
      "step": 822
    },
    {
      "epoch": 0.10654756125190148,
      "grad_norm": 0.272713661848405,
      "learning_rate": 1.249840420640849e-06,
      "loss": 1.1055,
      "step": 823
    },
    {
      "epoch": 0.10667702365925494,
      "grad_norm": 0.09100060547831061,
      "learning_rate": 1.2498339739201865e-06,
      "loss": 0.9164,
      "step": 824
    },
    {
      "epoch": 0.1068064860666084,
      "grad_norm": 0.27746205802596247,
      "learning_rate": 1.24982739956452e-06,
      "loss": 1.0723,
      "step": 825
    },
    {
      "epoch": 0.1068064860666084,
      "eval_PRM Accuracy": 0.49056603773584906,
      "eval_PRM F1": 0.578125,
      "eval_PRM F1 AUC": 0.5489785227867994,
      "eval_PRM F1 Neg": 0.35714285714285715,
      "eval_PRM NPV": 0.2459016393442623,
      "eval_PRM Precision": 0.8222222222222222,
      "eval_PRM Recall": 0.4457831325301205,
      "eval_PRM Specificty": 0.6521739130434783,
      "eval_loss": 0.7446733117103577,
      "eval_runtime": 20.9275,
      "eval_samples_per_second": 2.102,
      "eval_steps_per_second": 0.143,
      "step": 825
    },
    {
      "epoch": 0.10693594847396187,
      "grad_norm": 0.2856954309225274,
      "learning_rate": 1.2498206975751924e-06,
      "loss": 1.1639,
      "step": 826
    },
    {
      "epoch": 0.10706541088131534,
      "grad_norm": 0.22483660343751327,
      "learning_rate": 1.2498138679535728e-06,
      "loss": 1.0824,
      "step": 827
    },
    {
      "epoch": 0.10719487328866881,
      "grad_norm": 0.28162454729901676,
      "learning_rate": 1.2498069107010563e-06,
      "loss": 1.0842,
      "step": 828
    },
    {
      "epoch": 0.10732433569602227,
      "grad_norm": 0.278988184754171,
      "learning_rate": 1.2497998258190641e-06,
      "loss": 1.1788,
      "step": 829
    },
    {
      "epoch": 0.10745379810337573,
      "grad_norm": 0.20258414245353767,
      "learning_rate": 1.2497926133090432e-06,
      "loss": 1.0306,
      "step": 830
    },
    {
      "epoch": 0.10745379810337573,
      "eval_PRM Accuracy": 0.4811320754716981,
      "eval_PRM F1": 0.5736434108527132,
      "eval_PRM F1 AUC": 0.5272393923520168,
      "eval_PRM F1 Neg": 0.3373493975903614,
      "eval_PRM NPV": 0.23333333333333334,
      "eval_PRM Precision": 0.8043478260869565,
      "eval_PRM Recall": 0.4457831325301205,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.7414772510528564,
      "eval_runtime": 20.1971,
      "eval_samples_per_second": 2.179,
      "eval_steps_per_second": 0.149,
      "step": 830
    },
    {
      "epoch": 0.1075832605107292,
      "grad_norm": 0.29352504982027333,
      "learning_rate": 1.2497852731724673e-06,
      "loss": 1.1937,
      "step": 831
    },
    {
      "epoch": 0.10771272291808266,
      "grad_norm": 0.1997199394712332,
      "learning_rate": 1.2497778054108353e-06,
      "loss": 1.0553,
      "step": 832
    },
    {
      "epoch": 0.10784218532543613,
      "grad_norm": 0.20385573810979046,
      "learning_rate": 1.2497702100256732e-06,
      "loss": 1.1356,
      "step": 833
    },
    {
      "epoch": 0.1079716477327896,
      "grad_norm": 0.2905650732021906,
      "learning_rate": 1.2497624870185319e-06,
      "loss": 1.1574,
      "step": 834
    },
    {
      "epoch": 0.10810111014014305,
      "grad_norm": 0.24331090516914258,
      "learning_rate": 1.2497546363909894e-06,
      "loss": 1.1205,
      "step": 835
    },
    {
      "epoch": 0.10810111014014305,
      "eval_PRM Accuracy": 0.4811320754716981,
      "eval_PRM F1": 0.5736434108527132,
      "eval_PRM F1 AUC": 0.5272393923520168,
      "eval_PRM F1 Neg": 0.3373493975903614,
      "eval_PRM NPV": 0.23333333333333334,
      "eval_PRM Precision": 0.8043478260869565,
      "eval_PRM Recall": 0.4457831325301205,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.7397016882896423,
      "eval_runtime": 20.496,
      "eval_samples_per_second": 2.147,
      "eval_steps_per_second": 0.146,
      "step": 835
    },
    {
      "epoch": 0.10823057254749652,
      "grad_norm": 0.22858775433161402,
      "learning_rate": 1.249746658144649e-06,
      "loss": 1.186,
      "step": 836
    },
    {
      "epoch": 0.10836003495484998,
      "grad_norm": 0.21026192530249205,
      "learning_rate": 1.2497385522811408e-06,
      "loss": 1.179,
      "step": 837
    },
    {
      "epoch": 0.10848949736220345,
      "grad_norm": 0.24474330888254284,
      "learning_rate": 1.2497303188021205e-06,
      "loss": 0.9609,
      "step": 838
    },
    {
      "epoch": 0.10861895976955692,
      "grad_norm": 0.25352881549065953,
      "learning_rate": 1.2497219577092696e-06,
      "loss": 1.0745,
      "step": 839
    },
    {
      "epoch": 0.10874842217691039,
      "grad_norm": 0.24550251550489754,
      "learning_rate": 1.2497134690042965e-06,
      "loss": 1.1367,
      "step": 840
    },
    {
      "epoch": 0.10874842217691039,
      "eval_PRM Accuracy": 0.4811320754716981,
      "eval_PRM F1": 0.5736434108527132,
      "eval_PRM F1 AUC": 0.5272393923520168,
      "eval_PRM F1 Neg": 0.3373493975903614,
      "eval_PRM NPV": 0.23333333333333334,
      "eval_PRM Precision": 0.8043478260869565,
      "eval_PRM Recall": 0.4457831325301205,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.7350852489471436,
      "eval_runtime": 20.6361,
      "eval_samples_per_second": 2.132,
      "eval_steps_per_second": 0.145,
      "step": 840
    },
    {
      "epoch": 0.10887788458426384,
      "grad_norm": 0.26855208114368556,
      "learning_rate": 1.249704852688935e-06,
      "loss": 1.0919,
      "step": 841
    },
    {
      "epoch": 0.10900734699161731,
      "grad_norm": 0.2451687749708207,
      "learning_rate": 1.249696108764945e-06,
      "loss": 1.0544,
      "step": 842
    },
    {
      "epoch": 0.10913680939897077,
      "grad_norm": 0.215728239693382,
      "learning_rate": 1.249687237234113e-06,
      "loss": 1.0809,
      "step": 843
    },
    {
      "epoch": 0.10926627180632424,
      "grad_norm": 0.174006739010851,
      "learning_rate": 1.2496782380982509e-06,
      "loss": 1.108,
      "step": 844
    },
    {
      "epoch": 0.10939573421367771,
      "grad_norm": 0.18644207525473072,
      "learning_rate": 1.249669111359197e-06,
      "loss": 0.9909,
      "step": 845
    },
    {
      "epoch": 0.10939573421367771,
      "eval_PRM Accuracy": 0.49056603773584906,
      "eval_PRM F1": 0.5846153846153846,
      "eval_PRM F1 AUC": 0.5332634887375589,
      "eval_PRM F1 Neg": 0.34146341463414637,
      "eval_PRM NPV": 0.23728813559322035,
      "eval_PRM Precision": 0.8085106382978723,
      "eval_PRM Recall": 0.4578313253012048,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.734375,
      "eval_runtime": 21.009,
      "eval_samples_per_second": 2.094,
      "eval_steps_per_second": 0.143,
      "step": 845
    },
    {
      "epoch": 0.10952519662103116,
      "grad_norm": 0.26419616421367637,
      "learning_rate": 1.2496598570188156e-06,
      "loss": 1.1373,
      "step": 846
    },
    {
      "epoch": 0.10965465902838463,
      "grad_norm": 0.20390310198513964,
      "learning_rate": 1.249650475078997e-06,
      "loss": 0.9963,
      "step": 847
    },
    {
      "epoch": 0.1097841214357381,
      "grad_norm": 0.29986196761336015,
      "learning_rate": 1.249640965541658e-06,
      "loss": 1.0739,
      "step": 848
    },
    {
      "epoch": 0.10991358384309156,
      "grad_norm": 0.20405694232259175,
      "learning_rate": 1.2496313284087408e-06,
      "loss": 1.0385,
      "step": 849
    },
    {
      "epoch": 0.11004304625044503,
      "grad_norm": 0.13756837749289258,
      "learning_rate": 1.249621563682214e-06,
      "loss": 0.9885,
      "step": 850
    },
    {
      "epoch": 0.11004304625044503,
      "eval_PRM Accuracy": 0.49056603773584906,
      "eval_PRM F1": 0.5846153846153846,
      "eval_PRM F1 AUC": 0.5332634887375589,
      "eval_PRM F1 Neg": 0.34146341463414637,
      "eval_PRM NPV": 0.23728813559322035,
      "eval_PRM Precision": 0.8085106382978723,
      "eval_PRM Recall": 0.4578313253012048,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.7320667505264282,
      "eval_runtime": 20.0761,
      "eval_samples_per_second": 2.192,
      "eval_steps_per_second": 0.149,
      "step": 850
    },
    {
      "epoch": 0.1101725086577985,
      "grad_norm": 0.273805488686973,
      "learning_rate": 1.2496116713640725e-06,
      "loss": 1.1967,
      "step": 851
    },
    {
      "epoch": 0.11030197106515195,
      "grad_norm": 0.3046452795490258,
      "learning_rate": 1.2496016514563367e-06,
      "loss": 1.1144,
      "step": 852
    },
    {
      "epoch": 0.11043143347250542,
      "grad_norm": 0.2497664350945725,
      "learning_rate": 1.2495915039610534e-06,
      "loss": 1.1184,
      "step": 853
    },
    {
      "epoch": 0.11056089587985889,
      "grad_norm": 0.18914970103280176,
      "learning_rate": 1.2495812288802959e-06,
      "loss": 1.0773,
      "step": 854
    },
    {
      "epoch": 0.11069035828721235,
      "grad_norm": 0.2892537061446785,
      "learning_rate": 1.2495708262161623e-06,
      "loss": 1.0602,
      "step": 855
    },
    {
      "epoch": 0.11069035828721235,
      "eval_PRM Accuracy": 0.49056603773584906,
      "eval_PRM F1": 0.5846153846153846,
      "eval_PRM F1 AUC": 0.5332634887375589,
      "eval_PRM F1 Neg": 0.34146341463414637,
      "eval_PRM NPV": 0.23728813559322035,
      "eval_PRM Precision": 0.8085106382978723,
      "eval_PRM Recall": 0.4578313253012048,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.7294034361839294,
      "eval_runtime": 20.1678,
      "eval_samples_per_second": 2.182,
      "eval_steps_per_second": 0.149,
      "step": 855
    },
    {
      "epoch": 0.11081982069456582,
      "grad_norm": 0.15588820046320429,
      "learning_rate": 1.249560295970778e-06,
      "loss": 1.0277,
      "step": 856
    },
    {
      "epoch": 0.11094928310191927,
      "grad_norm": 0.1633100810507839,
      "learning_rate": 1.2495496381462942e-06,
      "loss": 1.0233,
      "step": 857
    },
    {
      "epoch": 0.11107874550927274,
      "grad_norm": 0.22617649148387292,
      "learning_rate": 1.2495388527448875e-06,
      "loss": 1.0326,
      "step": 858
    },
    {
      "epoch": 0.11120820791662621,
      "grad_norm": 0.31475706897809314,
      "learning_rate": 1.2495279397687615e-06,
      "loss": 1.1859,
      "step": 859
    },
    {
      "epoch": 0.11133767032397968,
      "grad_norm": 0.18936347831859027,
      "learning_rate": 1.2495168992201454e-06,
      "loss": 1.0613,
      "step": 860
    },
    {
      "epoch": 0.11133767032397968,
      "eval_PRM Accuracy": 0.49056603773584906,
      "eval_PRM F1": 0.5846153846153846,
      "eval_PRM F1 AUC": 0.5332634887375589,
      "eval_PRM F1 Neg": 0.34146341463414637,
      "eval_PRM NPV": 0.23728813559322035,
      "eval_PRM Precision": 0.8085106382978723,
      "eval_PRM Recall": 0.4578313253012048,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.7286931872367859,
      "eval_runtime": 19.7793,
      "eval_samples_per_second": 2.225,
      "eval_steps_per_second": 0.152,
      "step": 860
    },
    {
      "epoch": 0.11146713273133314,
      "grad_norm": 0.35337764522004383,
      "learning_rate": 1.249505731101294e-06,
      "loss": 1.1591,
      "step": 861
    },
    {
      "epoch": 0.11159659513868661,
      "grad_norm": 0.14762819877439,
      "learning_rate": 1.249494435414489e-06,
      "loss": 1.0142,
      "step": 862
    },
    {
      "epoch": 0.11172605754604006,
      "grad_norm": 0.13807941149101205,
      "learning_rate": 1.2494830121620375e-06,
      "loss": 1.2063,
      "step": 863
    },
    {
      "epoch": 0.11185551995339353,
      "grad_norm": 0.224716573207089,
      "learning_rate": 1.249471461346273e-06,
      "loss": 0.9923,
      "step": 864
    },
    {
      "epoch": 0.111984982360747,
      "grad_norm": 0.24679079063618864,
      "learning_rate": 1.249459782969555e-06,
      "loss": 1.0436,
      "step": 865
    },
    {
      "epoch": 0.111984982360747,
      "eval_PRM Accuracy": 0.49056603773584906,
      "eval_PRM F1": 0.5846153846153846,
      "eval_PRM F1 AUC": 0.5332634887375589,
      "eval_PRM F1 Neg": 0.34146341463414637,
      "eval_PRM NPV": 0.23728813559322035,
      "eval_PRM Precision": 0.8085106382978723,
      "eval_PRM Recall": 0.4578313253012048,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.7262073755264282,
      "eval_runtime": 20.2977,
      "eval_samples_per_second": 2.168,
      "eval_steps_per_second": 0.148,
      "step": 865
    },
    {
      "epoch": 0.11211444476810047,
      "grad_norm": 0.15828680223488442,
      "learning_rate": 1.2494479770342694e-06,
      "loss": 0.9651,
      "step": 866
    },
    {
      "epoch": 0.11224390717545393,
      "grad_norm": 0.23652712546802734,
      "learning_rate": 1.2494360435428275e-06,
      "loss": 1.0587,
      "step": 867
    },
    {
      "epoch": 0.11237336958280739,
      "grad_norm": 0.14304318763758858,
      "learning_rate": 1.2494239824976668e-06,
      "loss": 0.96,
      "step": 868
    },
    {
      "epoch": 0.11250283199016085,
      "grad_norm": 0.2542128080504395,
      "learning_rate": 1.249411793901251e-06,
      "loss": 1.0457,
      "step": 869
    },
    {
      "epoch": 0.11263229439751432,
      "grad_norm": 0.2633721293475217,
      "learning_rate": 1.2493994777560703e-06,
      "loss": 1.0123,
      "step": 870
    },
    {
      "epoch": 0.11263229439751432,
      "eval_PRM Accuracy": 0.5,
      "eval_PRM F1": 0.5954198473282443,
      "eval_PRM F1 AUC": 0.5392875851231012,
      "eval_PRM F1 Neg": 0.345679012345679,
      "eval_PRM NPV": 0.2413793103448276,
      "eval_PRM Precision": 0.8125,
      "eval_PRM Recall": 0.46987951807228917,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.7219460010528564,
      "eval_runtime": 20.3132,
      "eval_samples_per_second": 2.166,
      "eval_steps_per_second": 0.148,
      "step": 870
    },
    {
      "epoch": 0.11276175680486779,
      "grad_norm": 0.24899244204212426,
      "learning_rate": 1.2493870340646403e-06,
      "loss": 1.0698,
      "step": 871
    },
    {
      "epoch": 0.11289121921222126,
      "grad_norm": 0.29978812674319827,
      "learning_rate": 1.2493744628295027e-06,
      "loss": 1.1271,
      "step": 872
    },
    {
      "epoch": 0.11302068161957472,
      "grad_norm": 0.19580760823087529,
      "learning_rate": 1.2493617640532257e-06,
      "loss": 1.036,
      "step": 873
    },
    {
      "epoch": 0.11315014402692818,
      "grad_norm": 0.2697965784255595,
      "learning_rate": 1.2493489377384029e-06,
      "loss": 1.1612,
      "step": 874
    },
    {
      "epoch": 0.11327960643428164,
      "grad_norm": 0.2119424022794168,
      "learning_rate": 1.249335983887655e-06,
      "loss": 1.0673,
      "step": 875
    },
    {
      "epoch": 0.11327960643428164,
      "eval_PRM Accuracy": 0.5,
      "eval_PRM F1": 0.5954198473282443,
      "eval_PRM F1 AUC": 0.5392875851231012,
      "eval_PRM F1 Neg": 0.345679012345679,
      "eval_PRM NPV": 0.2413793103448276,
      "eval_PRM Precision": 0.8125,
      "eval_PRM Recall": 0.46987951807228917,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.7215909361839294,
      "eval_runtime": 20.7457,
      "eval_samples_per_second": 2.121,
      "eval_steps_per_second": 0.145,
      "step": 875
    },
    {
      "epoch": 0.11340906884163511,
      "grad_norm": 0.24551085932339028,
      "learning_rate": 1.2493229025036275e-06,
      "loss": 1.0796,
      "step": 876
    },
    {
      "epoch": 0.11353853124898858,
      "grad_norm": 0.25908531934303725,
      "learning_rate": 1.2493096935889926e-06,
      "loss": 1.0904,
      "step": 877
    },
    {
      "epoch": 0.11366799365634205,
      "grad_norm": 0.28665239622251376,
      "learning_rate": 1.2492963571464487e-06,
      "loss": 1.1184,
      "step": 878
    },
    {
      "epoch": 0.1137974560636955,
      "grad_norm": 0.23391033561744506,
      "learning_rate": 1.2492828931787199e-06,
      "loss": 1.0857,
      "step": 879
    },
    {
      "epoch": 0.11392691847104897,
      "grad_norm": 0.21241735524506247,
      "learning_rate": 1.2492693016885564e-06,
      "loss": 1.024,
      "step": 880
    },
    {
      "epoch": 0.11392691847104897,
      "eval_PRM Accuracy": 0.5094339622641509,
      "eval_PRM F1": 0.6060606060606061,
      "eval_PRM F1 AUC": 0.5453116815086433,
      "eval_PRM F1 Neg": 0.35,
      "eval_PRM NPV": 0.24561403508771928,
      "eval_PRM Precision": 0.8163265306122449,
      "eval_PRM Recall": 0.4819277108433735,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.71875,
      "eval_runtime": 20.2351,
      "eval_samples_per_second": 2.174,
      "eval_steps_per_second": 0.148,
      "step": 880
    },
    {
      "epoch": 0.11405638087840243,
      "grad_norm": 0.11419247987681164,
      "learning_rate": 1.249255582678735e-06,
      "loss": 0.8253,
      "step": 881
    },
    {
      "epoch": 0.1141858432857559,
      "grad_norm": 0.30542636755880465,
      "learning_rate": 1.2492417361520575e-06,
      "loss": 1.1077,
      "step": 882
    },
    {
      "epoch": 0.11431530569310937,
      "grad_norm": 0.2056452779904224,
      "learning_rate": 1.2492277621113528e-06,
      "loss": 0.9579,
      "step": 883
    },
    {
      "epoch": 0.11444476810046283,
      "grad_norm": 0.10895236912331824,
      "learning_rate": 1.249213660559475e-06,
      "loss": 1.0045,
      "step": 884
    },
    {
      "epoch": 0.11457423050781629,
      "grad_norm": 0.24844699381788002,
      "learning_rate": 1.249199431499305e-06,
      "loss": 1.1223,
      "step": 885
    },
    {
      "epoch": 0.11457423050781629,
      "eval_PRM Accuracy": 0.5188679245283019,
      "eval_PRM F1": 0.6165413533834586,
      "eval_PRM F1 AUC": 0.5513357778941854,
      "eval_PRM F1 Neg": 0.35443037974683544,
      "eval_PRM NPV": 0.25,
      "eval_PRM Precision": 0.82,
      "eval_PRM Recall": 0.4939759036144578,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.7164417505264282,
      "eval_runtime": 20.2647,
      "eval_samples_per_second": 2.171,
      "eval_steps_per_second": 0.148,
      "step": 885
    },
    {
      "epoch": 0.11470369291516976,
      "grad_norm": 0.2318722761563911,
      "learning_rate": 1.249185074933749e-06,
      "loss": 0.9695,
      "step": 886
    },
    {
      "epoch": 0.11483315532252322,
      "grad_norm": 0.2771180603843615,
      "learning_rate": 1.2491705908657395e-06,
      "loss": 1.1239,
      "step": 887
    },
    {
      "epoch": 0.11496261772987669,
      "grad_norm": 0.26883115435178806,
      "learning_rate": 1.2491559792982358e-06,
      "loss": 1.0718,
      "step": 888
    },
    {
      "epoch": 0.11509208013723016,
      "grad_norm": 0.2549570288540099,
      "learning_rate": 1.2491412402342223e-06,
      "loss": 1.038,
      "step": 889
    },
    {
      "epoch": 0.11522154254458361,
      "grad_norm": 0.2112448201230687,
      "learning_rate": 1.2491263736767095e-06,
      "loss": 0.9882,
      "step": 890
    },
    {
      "epoch": 0.11522154254458361,
      "eval_PRM Accuracy": 0.5094339622641509,
      "eval_PRM F1": 0.6060606060606061,
      "eval_PRM F1 AUC": 0.5453116815086433,
      "eval_PRM F1 Neg": 0.35,
      "eval_PRM NPV": 0.24561403508771928,
      "eval_PRM Precision": 0.8163265306122449,
      "eval_PRM Recall": 0.4819277108433735,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.7144886255264282,
      "eval_runtime": 20.4399,
      "eval_samples_per_second": 2.153,
      "eval_steps_per_second": 0.147,
      "step": 890
    },
    {
      "epoch": 0.11535100495193708,
      "grad_norm": 0.2594065909746897,
      "learning_rate": 1.2491113796287347e-06,
      "loss": 1.1175,
      "step": 891
    },
    {
      "epoch": 0.11548046735929055,
      "grad_norm": 0.25961743714135654,
      "learning_rate": 1.2490962580933602e-06,
      "loss": 1.0705,
      "step": 892
    },
    {
      "epoch": 0.11560992976664401,
      "grad_norm": 0.17908845575530666,
      "learning_rate": 1.2490810090736752e-06,
      "loss": 1.0261,
      "step": 893
    },
    {
      "epoch": 0.11573939217399748,
      "grad_norm": 0.31613936347505417,
      "learning_rate": 1.2490656325727945e-06,
      "loss": 1.1121,
      "step": 894
    },
    {
      "epoch": 0.11586885458135095,
      "grad_norm": 0.2857287921852463,
      "learning_rate": 1.2490501285938592e-06,
      "loss": 1.1281,
      "step": 895
    },
    {
      "epoch": 0.11586885458135095,
      "eval_PRM Accuracy": 0.5188679245283019,
      "eval_PRM F1": 0.6165413533834586,
      "eval_PRM F1 AUC": 0.5513357778941854,
      "eval_PRM F1 Neg": 0.35443037974683544,
      "eval_PRM NPV": 0.25,
      "eval_PRM Precision": 0.82,
      "eval_PRM Recall": 0.4939759036144578,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.7132457494735718,
      "eval_runtime": 20.9549,
      "eval_samples_per_second": 2.1,
      "eval_steps_per_second": 0.143,
      "step": 895
    },
    {
      "epoch": 0.1159983169887044,
      "grad_norm": 0.30321862569801306,
      "learning_rate": 1.2490344971400363e-06,
      "loss": 1.1357,
      "step": 896
    },
    {
      "epoch": 0.11612777939605787,
      "grad_norm": 0.31183746712739363,
      "learning_rate": 1.2490187382145185e-06,
      "loss": 1.14,
      "step": 897
    },
    {
      "epoch": 0.11625724180341133,
      "grad_norm": 0.1755978736987123,
      "learning_rate": 1.2490028518205256e-06,
      "loss": 0.9612,
      "step": 898
    },
    {
      "epoch": 0.1163867042107648,
      "grad_norm": 0.3085330783048964,
      "learning_rate": 1.2489868379613017e-06,
      "loss": 0.9958,
      "step": 899
    },
    {
      "epoch": 0.11651616661811827,
      "grad_norm": 0.18416313133057224,
      "learning_rate": 1.248970696640119e-06,
      "loss": 1.0742,
      "step": 900
    },
    {
      "epoch": 0.11651616661811827,
      "eval_PRM Accuracy": 0.5377358490566038,
      "eval_PRM F1": 0.6370370370370371,
      "eval_PRM F1 AUC": 0.5633839706652698,
      "eval_PRM F1 Neg": 0.36363636363636365,
      "eval_PRM NPV": 0.25925925925925924,
      "eval_PRM Precision": 0.8269230769230769,
      "eval_PRM Recall": 0.5180722891566265,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.7082741260528564,
      "eval_runtime": 20.3598,
      "eval_samples_per_second": 2.161,
      "eval_steps_per_second": 0.147,
      "step": 900
    },
    {
      "epoch": 0.11664562902547172,
      "grad_norm": 0.12835841830653832,
      "learning_rate": 1.248954427860274e-06,
      "loss": 0.963,
      "step": 901
    },
    {
      "epoch": 0.11677509143282519,
      "grad_norm": 0.20071122181022247,
      "learning_rate": 1.2489380316250898e-06,
      "loss": 1.0999,
      "step": 902
    },
    {
      "epoch": 0.11690455384017866,
      "grad_norm": 0.2924281911983938,
      "learning_rate": 1.2489215079379164e-06,
      "loss": 0.9619,
      "step": 903
    },
    {
      "epoch": 0.11703401624753212,
      "grad_norm": 0.18459876402155906,
      "learning_rate": 1.2489048568021284e-06,
      "loss": 1.0217,
      "step": 904
    },
    {
      "epoch": 0.11716347865488559,
      "grad_norm": 0.24870341952160963,
      "learning_rate": 1.2488880782211277e-06,
      "loss": 0.9905,
      "step": 905
    },
    {
      "epoch": 0.11716347865488559,
      "eval_PRM Accuracy": 0.5377358490566038,
      "eval_PRM F1": 0.6370370370370371,
      "eval_PRM F1 AUC": 0.5633839706652698,
      "eval_PRM F1 Neg": 0.36363636363636365,
      "eval_PRM NPV": 0.25925925925925924,
      "eval_PRM Precision": 0.8269230769230769,
      "eval_PRM Recall": 0.5180722891566265,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.7073863744735718,
      "eval_runtime": 20.1407,
      "eval_samples_per_second": 2.185,
      "eval_steps_per_second": 0.149,
      "step": 905
    },
    {
      "epoch": 0.11729294106223906,
      "grad_norm": 0.2943022227518921,
      "learning_rate": 1.2488711721983413e-06,
      "loss": 1.1802,
      "step": 906
    },
    {
      "epoch": 0.11742240346959251,
      "grad_norm": 0.20927952420906956,
      "learning_rate": 1.2488541387372226e-06,
      "loss": 0.9427,
      "step": 907
    },
    {
      "epoch": 0.11755186587694598,
      "grad_norm": 0.20045053771377452,
      "learning_rate": 1.2488369778412513e-06,
      "loss": 1.0686,
      "step": 908
    },
    {
      "epoch": 0.11768132828429945,
      "grad_norm": 0.21925166112671532,
      "learning_rate": 1.2488196895139323e-06,
      "loss": 1.0447,
      "step": 909
    },
    {
      "epoch": 0.11781079069165291,
      "grad_norm": 0.22803737349362252,
      "learning_rate": 1.2488022737587978e-06,
      "loss": 1.042,
      "step": 910
    },
    {
      "epoch": 0.11781079069165291,
      "eval_PRM Accuracy": 0.5377358490566038,
      "eval_PRM F1": 0.6370370370370371,
      "eval_PRM F1 AUC": 0.5633839706652698,
      "eval_PRM F1 Neg": 0.36363636363636365,
      "eval_PRM NPV": 0.25925925925925924,
      "eval_PRM Precision": 0.8269230769230769,
      "eval_PRM Recall": 0.5180722891566265,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.7080965638160706,
      "eval_runtime": 20.2607,
      "eval_samples_per_second": 2.172,
      "eval_steps_per_second": 0.148,
      "step": 910
    },
    {
      "epoch": 0.11794025309900638,
      "grad_norm": 0.26257299652913973,
      "learning_rate": 1.248784730579405e-06,
      "loss": 1.1315,
      "step": 911
    },
    {
      "epoch": 0.11806971550635983,
      "grad_norm": 0.26153336312290165,
      "learning_rate": 1.2487670599793372e-06,
      "loss": 1.0878,
      "step": 912
    },
    {
      "epoch": 0.1181991779137133,
      "grad_norm": 0.22446569063354554,
      "learning_rate": 1.2487492619622045e-06,
      "loss": 1.0463,
      "step": 913
    },
    {
      "epoch": 0.11832864032106677,
      "grad_norm": 0.308313813282149,
      "learning_rate": 1.2487313365316419e-06,
      "loss": 0.9839,
      "step": 914
    },
    {
      "epoch": 0.11845810272842024,
      "grad_norm": 0.34442116697939235,
      "learning_rate": 1.2487132836913117e-06,
      "loss": 1.0446,
      "step": 915
    },
    {
      "epoch": 0.11845810272842024,
      "eval_PRM Accuracy": 0.5283018867924528,
      "eval_PRM F1": 0.6268656716417911,
      "eval_PRM F1 AUC": 0.5573598742797277,
      "eval_PRM F1 Neg": 0.358974358974359,
      "eval_PRM NPV": 0.2545454545454545,
      "eval_PRM Precision": 0.8235294117647058,
      "eval_PRM Recall": 0.5060240963855421,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.7052556872367859,
      "eval_runtime": 20.2169,
      "eval_samples_per_second": 2.176,
      "eval_steps_per_second": 0.148,
      "step": 915
    },
    {
      "epoch": 0.1185875651357737,
      "grad_norm": 0.20284590608333603,
      "learning_rate": 1.248695103444901e-06,
      "loss": 1.0165,
      "step": 916
    },
    {
      "epoch": 0.11871702754312717,
      "grad_norm": 0.24376948009314453,
      "learning_rate": 1.2486767957961237e-06,
      "loss": 1.0447,
      "step": 917
    },
    {
      "epoch": 0.11884648995048062,
      "grad_norm": 0.2675687967412616,
      "learning_rate": 1.2486583607487197e-06,
      "loss": 1.0378,
      "step": 918
    },
    {
      "epoch": 0.11897595235783409,
      "grad_norm": 0.1671036375348281,
      "learning_rate": 1.2486397983064543e-06,
      "loss": 1.0276,
      "step": 919
    },
    {
      "epoch": 0.11910541476518756,
      "grad_norm": 0.29111974867403917,
      "learning_rate": 1.2486211084731197e-06,
      "loss": 1.1477,
      "step": 920
    },
    {
      "epoch": 0.11910541476518756,
      "eval_PRM Accuracy": 0.5377358490566038,
      "eval_PRM F1": 0.6370370370370371,
      "eval_PRM F1 AUC": 0.5633839706652698,
      "eval_PRM F1 Neg": 0.36363636363636365,
      "eval_PRM NPV": 0.25925925925925924,
      "eval_PRM Precision": 0.8269230769230769,
      "eval_PRM Recall": 0.5180722891566265,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.7013494372367859,
      "eval_runtime": 20.6924,
      "eval_samples_per_second": 2.126,
      "eval_steps_per_second": 0.145,
      "step": 920
    },
    {
      "epoch": 0.11923487717254103,
      "grad_norm": 0.2927476002427883,
      "learning_rate": 1.2486022912525333e-06,
      "loss": 1.0867,
      "step": 921
    },
    {
      "epoch": 0.1193643395798945,
      "grad_norm": 0.30445920229169054,
      "learning_rate": 1.2485833466485392e-06,
      "loss": 1.1169,
      "step": 922
    },
    {
      "epoch": 0.11949380198724795,
      "grad_norm": 0.30580982272691565,
      "learning_rate": 1.248564274665007e-06,
      "loss": 1.0751,
      "step": 923
    },
    {
      "epoch": 0.11962326439460141,
      "grad_norm": 0.31166415065880815,
      "learning_rate": 1.2485450753058326e-06,
      "loss": 1.1058,
      "step": 924
    },
    {
      "epoch": 0.11975272680195488,
      "grad_norm": 0.11480287425920799,
      "learning_rate": 1.248525748574938e-06,
      "loss": 0.9757,
      "step": 925
    },
    {
      "epoch": 0.11975272680195488,
      "eval_PRM Accuracy": 0.5377358490566038,
      "eval_PRM F1": 0.6370370370370371,
      "eval_PRM F1 AUC": 0.5633839706652698,
      "eval_PRM F1 Neg": 0.36363636363636365,
      "eval_PRM NPV": 0.25925925925925924,
      "eval_PRM Precision": 0.8269230769230769,
      "eval_PRM Recall": 0.5180722891566265,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.7013494372367859,
      "eval_runtime": 20.3504,
      "eval_samples_per_second": 2.162,
      "eval_steps_per_second": 0.147,
      "step": 925
    },
    {
      "epoch": 0.11988218920930835,
      "grad_norm": 0.24314245870276266,
      "learning_rate": 1.2485062944762707e-06,
      "loss": 0.9986,
      "step": 926
    },
    {
      "epoch": 0.12001165161666182,
      "grad_norm": 0.2552476338883707,
      "learning_rate": 1.2484867130138051e-06,
      "loss": 1.0531,
      "step": 927
    },
    {
      "epoch": 0.12014111402401528,
      "grad_norm": 0.2359204746157435,
      "learning_rate": 1.2484670041915408e-06,
      "loss": 1.1035,
      "step": 928
    },
    {
      "epoch": 0.12027057643136874,
      "grad_norm": 0.3352930236383672,
      "learning_rate": 1.248447168013504e-06,
      "loss": 1.1707,
      "step": 929
    },
    {
      "epoch": 0.1204000388387222,
      "grad_norm": 0.34336983341594346,
      "learning_rate": 1.2484272044837462e-06,
      "loss": 1.181,
      "step": 930
    },
    {
      "epoch": 0.1204000388387222,
      "eval_PRM Accuracy": 0.5471698113207547,
      "eval_PRM F1": 0.6470588235294118,
      "eval_PRM F1 AUC": 0.5694080670508119,
      "eval_PRM F1 Neg": 0.3684210526315789,
      "eval_PRM NPV": 0.2641509433962264,
      "eval_PRM Precision": 0.8301886792452831,
      "eval_PRM Recall": 0.5301204819277109,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.7004616260528564,
      "eval_runtime": 19.7231,
      "eval_samples_per_second": 2.231,
      "eval_steps_per_second": 0.152,
      "step": 930
    },
    {
      "epoch": 0.12052950124607567,
      "grad_norm": 0.26507746551746175,
      "learning_rate": 1.2484071136063456e-06,
      "loss": 1.0639,
      "step": 931
    },
    {
      "epoch": 0.12065896365342914,
      "grad_norm": 0.15430756125265027,
      "learning_rate": 1.2483868953854061e-06,
      "loss": 0.9623,
      "step": 932
    },
    {
      "epoch": 0.1207884260607826,
      "grad_norm": 0.24730460456392214,
      "learning_rate": 1.248366549825058e-06,
      "loss": 1.0803,
      "step": 933
    },
    {
      "epoch": 0.12091788846813606,
      "grad_norm": 0.17759719029207735,
      "learning_rate": 1.248346076929457e-06,
      "loss": 0.9493,
      "step": 934
    },
    {
      "epoch": 0.12104735087548953,
      "grad_norm": 0.14805744054801356,
      "learning_rate": 1.2483254767027848e-06,
      "loss": 1.0398,
      "step": 935
    },
    {
      "epoch": 0.12104735087548953,
      "eval_PRM Accuracy": 0.5471698113207547,
      "eval_PRM F1": 0.6470588235294118,
      "eval_PRM F1 AUC": 0.5694080670508119,
      "eval_PRM F1 Neg": 0.3684210526315789,
      "eval_PRM NPV": 0.2641509433962264,
      "eval_PRM Precision": 0.8301886792452831,
      "eval_PRM Recall": 0.5301204819277109,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.6970880627632141,
      "eval_runtime": 20.4074,
      "eval_samples_per_second": 2.156,
      "eval_steps_per_second": 0.147,
      "step": 935
    },
    {
      "epoch": 0.121176813282843,
      "grad_norm": 0.0975216483398078,
      "learning_rate": 1.2483047491492501e-06,
      "loss": 0.8837,
      "step": 936
    },
    {
      "epoch": 0.12130627569019646,
      "grad_norm": 0.23511993554299532,
      "learning_rate": 1.2482838942730862e-06,
      "loss": 1.0808,
      "step": 937
    },
    {
      "epoch": 0.12143573809754993,
      "grad_norm": 0.10027602407600866,
      "learning_rate": 1.248262912078554e-06,
      "loss": 0.9457,
      "step": 938
    },
    {
      "epoch": 0.1215652005049034,
      "grad_norm": 0.3169258597939292,
      "learning_rate": 1.2482418025699386e-06,
      "loss": 1.0693,
      "step": 939
    },
    {
      "epoch": 0.12169466291225685,
      "grad_norm": 0.29996150877527894,
      "learning_rate": 1.2482205657515528e-06,
      "loss": 0.9786,
      "step": 940
    },
    {
      "epoch": 0.12169466291225685,
      "eval_PRM Accuracy": 0.5566037735849056,
      "eval_PRM F1": 0.656934306569343,
      "eval_PRM F1 AUC": 0.5754321634363542,
      "eval_PRM F1 Neg": 0.37333333333333335,
      "eval_PRM NPV": 0.2692307692307692,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.5421686746987951,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.6942471861839294,
      "eval_runtime": 19.9488,
      "eval_samples_per_second": 2.206,
      "eval_steps_per_second": 0.15,
      "step": 940
    },
    {
      "epoch": 0.12182412531961032,
      "grad_norm": 0.22453547614745933,
      "learning_rate": 1.2481992016277341e-06,
      "loss": 1.0378,
      "step": 941
    },
    {
      "epoch": 0.12195358772696378,
      "grad_norm": 0.21006845986473938,
      "learning_rate": 1.248177710202847e-06,
      "loss": 1.1132,
      "step": 942
    },
    {
      "epoch": 0.12208305013431725,
      "grad_norm": 0.15992988524757043,
      "learning_rate": 1.2481560914812812e-06,
      "loss": 1.101,
      "step": 943
    },
    {
      "epoch": 0.12221251254167072,
      "grad_norm": 0.22740313189887396,
      "learning_rate": 1.248134345467453e-06,
      "loss": 1.1594,
      "step": 944
    },
    {
      "epoch": 0.12234197494902417,
      "grad_norm": 0.28398594523294257,
      "learning_rate": 1.2481124721658043e-06,
      "loss": 0.9941,
      "step": 945
    },
    {
      "epoch": 0.12234197494902417,
      "eval_PRM Accuracy": 0.5566037735849056,
      "eval_PRM F1": 0.656934306569343,
      "eval_PRM F1 AUC": 0.5754321634363542,
      "eval_PRM F1 Neg": 0.37333333333333335,
      "eval_PRM NPV": 0.2692307692307692,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.5421686746987951,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.6922940611839294,
      "eval_runtime": 20.4976,
      "eval_samples_per_second": 2.147,
      "eval_steps_per_second": 0.146,
      "step": 945
    },
    {
      "epoch": 0.12247143735637764,
      "grad_norm": 0.28974589070997037,
      "learning_rate": 1.2480904715808035e-06,
      "loss": 1.1002,
      "step": 946
    },
    {
      "epoch": 0.1226008997637311,
      "grad_norm": 0.17185736104411076,
      "learning_rate": 1.2480683437169441e-06,
      "loss": 0.9274,
      "step": 947
    },
    {
      "epoch": 0.12273036217108457,
      "grad_norm": 0.3051407108944278,
      "learning_rate": 1.2480460885787465e-06,
      "loss": 1.0024,
      "step": 948
    },
    {
      "epoch": 0.12285982457843804,
      "grad_norm": 0.1938359421159141,
      "learning_rate": 1.248023706170757e-06,
      "loss": 1.0719,
      "step": 949
    },
    {
      "epoch": 0.12298928698579151,
      "grad_norm": 0.2642402308896572,
      "learning_rate": 1.2480011964975472e-06,
      "loss": 1.0945,
      "step": 950
    },
    {
      "epoch": 0.12298928698579151,
      "eval_PRM Accuracy": 0.5566037735849056,
      "eval_PRM F1": 0.656934306569343,
      "eval_PRM F1 AUC": 0.5754321634363542,
      "eval_PRM F1 Neg": 0.37333333333333335,
      "eval_PRM NPV": 0.2692307692307692,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.5421686746987951,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.6942471861839294,
      "eval_runtime": 20.4898,
      "eval_samples_per_second": 2.147,
      "eval_steps_per_second": 0.146,
      "step": 950
    },
    {
      "epoch": 0.12311874939314496,
      "grad_norm": 0.34714050386872264,
      "learning_rate": 1.2479785595637152e-06,
      "loss": 1.0513,
      "step": 951
    },
    {
      "epoch": 0.12324821180049843,
      "grad_norm": 0.26637826202424403,
      "learning_rate": 1.2479557953738854e-06,
      "loss": 0.9675,
      "step": 952
    },
    {
      "epoch": 0.1233776742078519,
      "grad_norm": 0.2010276621390886,
      "learning_rate": 1.2479329039327078e-06,
      "loss": 1.0109,
      "step": 953
    },
    {
      "epoch": 0.12350713661520536,
      "grad_norm": 0.2287827383786289,
      "learning_rate": 1.247909885244858e-06,
      "loss": 1.0419,
      "step": 954
    },
    {
      "epoch": 0.12363659902255883,
      "grad_norm": 0.1987488490791542,
      "learning_rate": 1.2478867393150387e-06,
      "loss": 0.8956,
      "step": 955
    },
    {
      "epoch": 0.12363659902255883,
      "eval_PRM Accuracy": 0.5566037735849056,
      "eval_PRM F1": 0.656934306569343,
      "eval_PRM F1 AUC": 0.5754321634363542,
      "eval_PRM F1 Neg": 0.37333333333333335,
      "eval_PRM NPV": 0.2692307692307692,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.5421686746987951,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.6905184388160706,
      "eval_runtime": 20.2499,
      "eval_samples_per_second": 2.173,
      "eval_steps_per_second": 0.148,
      "step": 955
    },
    {
      "epoch": 0.12376606142991228,
      "grad_norm": 0.25799023415568756,
      "learning_rate": 1.247863466147977e-06,
      "loss": 1.027,
      "step": 956
    },
    {
      "epoch": 0.12389552383726575,
      "grad_norm": 0.17114997717982633,
      "learning_rate": 1.247840065748428e-06,
      "loss": 1.0116,
      "step": 957
    },
    {
      "epoch": 0.12402498624461922,
      "grad_norm": 0.24403506533486202,
      "learning_rate": 1.247816538121171e-06,
      "loss": 1.0424,
      "step": 958
    },
    {
      "epoch": 0.12415444865197268,
      "grad_norm": 0.2641757358370001,
      "learning_rate": 1.247792883271012e-06,
      "loss": 1.0206,
      "step": 959
    },
    {
      "epoch": 0.12428391105932615,
      "grad_norm": 0.28572523781138776,
      "learning_rate": 1.2477691012027835e-06,
      "loss": 1.1404,
      "step": 960
    },
    {
      "epoch": 0.12428391105932615,
      "eval_PRM Accuracy": 0.5566037735849056,
      "eval_PRM F1": 0.656934306569343,
      "eval_PRM F1 AUC": 0.5754321634363542,
      "eval_PRM F1 Neg": 0.37333333333333335,
      "eval_PRM NPV": 0.2692307692307692,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.5421686746987951,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.6885653138160706,
      "eval_runtime": 20.36,
      "eval_samples_per_second": 2.161,
      "eval_steps_per_second": 0.147,
      "step": 960
    },
    {
      "epoch": 0.12441337346667962,
      "grad_norm": 0.16101329417132618,
      "learning_rate": 1.247745191921343e-06,
      "loss": 1.0167,
      "step": 961
    },
    {
      "epoch": 0.12454283587403307,
      "grad_norm": 0.23178472803206407,
      "learning_rate": 1.2477211554315744e-06,
      "loss": 0.9841,
      "step": 962
    },
    {
      "epoch": 0.12467229828138654,
      "grad_norm": 0.28990005171684535,
      "learning_rate": 1.247696991738388e-06,
      "loss": 1.0898,
      "step": 963
    },
    {
      "epoch": 0.12480176068874001,
      "grad_norm": 0.2406609874790822,
      "learning_rate": 1.2476727008467195e-06,
      "loss": 1.1318,
      "step": 964
    },
    {
      "epoch": 0.12493122309609347,
      "grad_norm": 0.23889792071952662,
      "learning_rate": 1.2476482827615308e-06,
      "loss": 1.1131,
      "step": 965
    },
    {
      "epoch": 0.12493122309609347,
      "eval_PRM Accuracy": 0.5566037735849056,
      "eval_PRM F1": 0.656934306569343,
      "eval_PRM F1 AUC": 0.5754321634363542,
      "eval_PRM F1 Neg": 0.37333333333333335,
      "eval_PRM NPV": 0.2692307692307692,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.5421686746987951,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.685546875,
      "eval_runtime": 20.0248,
      "eval_samples_per_second": 2.197,
      "eval_steps_per_second": 0.15,
      "step": 965
    },
    {
      "epoch": 0.12506068550344693,
      "grad_norm": 0.2867318832438925,
      "learning_rate": 1.2476237374878098e-06,
      "loss": 1.114,
      "step": 966
    },
    {
      "epoch": 0.1251901479108004,
      "grad_norm": 0.24665522425255476,
      "learning_rate": 1.2475990650305705e-06,
      "loss": 1.0485,
      "step": 967
    },
    {
      "epoch": 0.12531961031815386,
      "grad_norm": 0.2342570204432168,
      "learning_rate": 1.2475742653948527e-06,
      "loss": 1.048,
      "step": 968
    },
    {
      "epoch": 0.12544907272550734,
      "grad_norm": 0.2524333252167745,
      "learning_rate": 1.2475493385857222e-06,
      "loss": 0.9614,
      "step": 969
    },
    {
      "epoch": 0.1255785351328608,
      "grad_norm": 0.21165199025095557,
      "learning_rate": 1.2475242846082706e-06,
      "loss": 1.1005,
      "step": 970
    },
    {
      "epoch": 0.1255785351328608,
      "eval_PRM Accuracy": 0.5660377358490566,
      "eval_PRM F1": 0.6666666666666666,
      "eval_PRM F1 AUC": 0.5814562598218963,
      "eval_PRM F1 Neg": 0.3783783783783784,
      "eval_PRM NPV": 0.27450980392156865,
      "eval_PRM Precision": 0.8363636363636363,
      "eval_PRM Recall": 0.5542168674698795,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.6850141882896423,
      "eval_runtime": 20.3144,
      "eval_samples_per_second": 2.166,
      "eval_steps_per_second": 0.148,
      "step": 970
    },
    {
      "epoch": 0.12570799754021425,
      "grad_norm": 0.13222001672625228,
      "learning_rate": 1.247499103467616e-06,
      "loss": 0.9894,
      "step": 971
    },
    {
      "epoch": 0.12583745994756773,
      "grad_norm": 0.22312405691013296,
      "learning_rate": 1.247473795168902e-06,
      "loss": 1.0591,
      "step": 972
    },
    {
      "epoch": 0.12596692235492118,
      "grad_norm": 0.2653192637392524,
      "learning_rate": 1.2474483597172985e-06,
      "loss": 0.9784,
      "step": 973
    },
    {
      "epoch": 0.12609638476227467,
      "grad_norm": 0.24917799582714162,
      "learning_rate": 1.247422797118001e-06,
      "loss": 0.9749,
      "step": 974
    },
    {
      "epoch": 0.12622584716962812,
      "grad_norm": 0.11582279268731416,
      "learning_rate": 1.2473971073762312e-06,
      "loss": 0.9305,
      "step": 975
    },
    {
      "epoch": 0.12622584716962812,
      "eval_PRM Accuracy": 0.5566037735849056,
      "eval_PRM F1": 0.656934306569343,
      "eval_PRM F1 AUC": 0.5754321634363542,
      "eval_PRM F1 Neg": 0.37333333333333335,
      "eval_PRM NPV": 0.2692307692307692,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.5421686746987951,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.6825284361839294,
      "eval_runtime": 20.6381,
      "eval_samples_per_second": 2.132,
      "eval_steps_per_second": 0.145,
      "step": 975
    },
    {
      "epoch": 0.12635530957698157,
      "grad_norm": 0.2042436531713094,
      "learning_rate": 1.2473712904972369e-06,
      "loss": 0.9087,
      "step": 976
    },
    {
      "epoch": 0.12648477198433505,
      "grad_norm": 0.2325955386339354,
      "learning_rate": 1.2473453464862914e-06,
      "loss": 1.0846,
      "step": 977
    },
    {
      "epoch": 0.1266142343916885,
      "grad_norm": 0.2818198218864742,
      "learning_rate": 1.247319275348695e-06,
      "loss": 1.1457,
      "step": 978
    },
    {
      "epoch": 0.126743696799042,
      "grad_norm": 0.19345997360876083,
      "learning_rate": 1.2472930770897722e-06,
      "loss": 0.9977,
      "step": 979
    },
    {
      "epoch": 0.12687315920639544,
      "grad_norm": 0.21821019923485466,
      "learning_rate": 1.2472667517148754e-06,
      "loss": 0.9125,
      "step": 980
    },
    {
      "epoch": 0.12687315920639544,
      "eval_PRM Accuracy": 0.5754716981132075,
      "eval_PRM F1": 0.6762589928057554,
      "eval_PRM F1 AUC": 0.5874803562074384,
      "eval_PRM F1 Neg": 0.3835616438356164,
      "eval_PRM NPV": 0.28,
      "eval_PRM Precision": 0.8392857142857143,
      "eval_PRM Recall": 0.5662650602409639,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.6812855005264282,
      "eval_runtime": 20.7751,
      "eval_samples_per_second": 2.118,
      "eval_steps_per_second": 0.144,
      "step": 980
    },
    {
      "epoch": 0.1270026216137489,
      "grad_norm": 0.2541148069342123,
      "learning_rate": 1.2472402992293817e-06,
      "loss": 0.9821,
      "step": 981
    },
    {
      "epoch": 0.12713208402110238,
      "grad_norm": 0.23557301504909928,
      "learning_rate": 1.2472137196386946e-06,
      "loss": 0.9647,
      "step": 982
    },
    {
      "epoch": 0.12726154642845583,
      "grad_norm": 0.23328185838738508,
      "learning_rate": 1.2471870129482436e-06,
      "loss": 0.9785,
      "step": 983
    },
    {
      "epoch": 0.1273910088358093,
      "grad_norm": 0.16539467124796214,
      "learning_rate": 1.2471601791634843e-06,
      "loss": 0.9518,
      "step": 984
    },
    {
      "epoch": 0.12752047124316276,
      "grad_norm": 0.15360511168540106,
      "learning_rate": 1.2471332182898976e-06,
      "loss": 0.9041,
      "step": 985
    },
    {
      "epoch": 0.12752047124316276,
      "eval_PRM Accuracy": 0.5754716981132075,
      "eval_PRM F1": 0.6762589928057554,
      "eval_PRM F1 AUC": 0.5874803562074384,
      "eval_PRM F1 Neg": 0.3835616438356164,
      "eval_PRM NPV": 0.28,
      "eval_PRM Precision": 0.8392857142857143,
      "eval_PRM Recall": 0.5662650602409639,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.6780894994735718,
      "eval_runtime": 20.682,
      "eval_samples_per_second": 2.127,
      "eval_steps_per_second": 0.145,
      "step": 985
    },
    {
      "epoch": 0.12764993365051625,
      "grad_norm": 0.23043966158517099,
      "learning_rate": 1.247106130332991e-06,
      "loss": 1.0165,
      "step": 986
    },
    {
      "epoch": 0.1277793960578697,
      "grad_norm": 0.20981336403010664,
      "learning_rate": 1.2470789152982976e-06,
      "loss": 1.0427,
      "step": 987
    },
    {
      "epoch": 0.12790885846522315,
      "grad_norm": 0.27582215275835326,
      "learning_rate": 1.247051573191377e-06,
      "loss": 1.0491,
      "step": 988
    },
    {
      "epoch": 0.12803832087257663,
      "grad_norm": 0.2250831170100155,
      "learning_rate": 1.247024104017814e-06,
      "loss": 0.9861,
      "step": 989
    },
    {
      "epoch": 0.1281677832799301,
      "grad_norm": 0.18050711983185633,
      "learning_rate": 1.24699650778322e-06,
      "loss": 0.8724,
      "step": 990
    },
    {
      "epoch": 0.1281677832799301,
      "eval_PRM Accuracy": 0.5660377358490566,
      "eval_PRM F1": 0.6666666666666666,
      "eval_PRM F1 AUC": 0.5814562598218963,
      "eval_PRM F1 Neg": 0.3783783783783784,
      "eval_PRM NPV": 0.27450980392156865,
      "eval_PRM Precision": 0.8363636363636363,
      "eval_PRM Recall": 0.5542168674698795,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.6789772510528564,
      "eval_runtime": 20.6872,
      "eval_samples_per_second": 2.127,
      "eval_steps_per_second": 0.145,
      "step": 990
    },
    {
      "epoch": 0.12829724568728357,
      "grad_norm": 0.26754770001546774,
      "learning_rate": 1.246968784493232e-06,
      "loss": 1.0915,
      "step": 991
    },
    {
      "epoch": 0.12842670809463702,
      "grad_norm": 0.16966310322097386,
      "learning_rate": 1.2469409341535132e-06,
      "loss": 0.9379,
      "step": 992
    },
    {
      "epoch": 0.12855617050199047,
      "grad_norm": 0.2891491943508992,
      "learning_rate": 1.246912956769752e-06,
      "loss": 0.9877,
      "step": 993
    },
    {
      "epoch": 0.12868563290934396,
      "grad_norm": 0.26344659024845823,
      "learning_rate": 1.2468848523476642e-06,
      "loss": 1.0109,
      "step": 994
    },
    {
      "epoch": 0.1288150953166974,
      "grad_norm": 0.1274829380326075,
      "learning_rate": 1.24685662089299e-06,
      "loss": 0.9545,
      "step": 995
    },
    {
      "epoch": 0.1288150953166974,
      "eval_PRM Accuracy": 0.5754716981132075,
      "eval_PRM F1": 0.6762589928057554,
      "eval_PRM F1 AUC": 0.5874803562074384,
      "eval_PRM F1 Neg": 0.3835616438356164,
      "eval_PRM NPV": 0.28,
      "eval_PRM Precision": 0.8392857142857143,
      "eval_PRM Recall": 0.5662650602409639,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.6782670617103577,
      "eval_runtime": 20.7096,
      "eval_samples_per_second": 2.125,
      "eval_steps_per_second": 0.145,
      "step": 995
    },
    {
      "epoch": 0.1289445577240509,
      "grad_norm": 0.208178081251272,
      "learning_rate": 1.2468282624114965e-06,
      "loss": 1.0662,
      "step": 996
    },
    {
      "epoch": 0.12907402013140434,
      "grad_norm": 0.2814180283504166,
      "learning_rate": 1.2467997769089768e-06,
      "loss": 0.9773,
      "step": 997
    },
    {
      "epoch": 0.1292034825387578,
      "grad_norm": 0.2331440924789102,
      "learning_rate": 1.2467711643912493e-06,
      "loss": 0.967,
      "step": 998
    },
    {
      "epoch": 0.12933294494611128,
      "grad_norm": 0.2076851204570146,
      "learning_rate": 1.2467424248641585e-06,
      "loss": 1.017,
      "step": 999
    },
    {
      "epoch": 0.12946240735346473,
      "grad_norm": 0.21775231163710648,
      "learning_rate": 1.2467135583335757e-06,
      "loss": 1.0594,
      "step": 1000
    },
    {
      "epoch": 0.12946240735346473,
      "eval_PRM Accuracy": 0.5754716981132075,
      "eval_PRM F1": 0.6808510638297872,
      "eval_PRM F1 AUC": 0.5717653221581981,
      "eval_PRM F1 Neg": 0.36619718309859156,
      "eval_PRM NPV": 0.2708333333333333,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.5783132530120482,
      "eval_PRM Specificty": 0.5652173913043478,
      "eval_loss": 0.6761363744735718,
      "eval_runtime": 20.7606,
      "eval_samples_per_second": 2.119,
      "eval_steps_per_second": 0.145,
      "step": 1000
    },
    {
      "epoch": 0.1295918697608182,
      "grad_norm": 0.3096860149829295,
      "learning_rate": 1.2466845648053966e-06,
      "loss": 1.0576,
      "step": 1001
    },
    {
      "epoch": 0.12972133216817167,
      "grad_norm": 0.12943147025044327,
      "learning_rate": 1.2466554442855443e-06,
      "loss": 0.9316,
      "step": 1002
    },
    {
      "epoch": 0.12985079457552512,
      "grad_norm": 0.3346966536950126,
      "learning_rate": 1.2466261967799672e-06,
      "loss": 1.0526,
      "step": 1003
    },
    {
      "epoch": 0.1299802569828786,
      "grad_norm": 0.2634146419592761,
      "learning_rate": 1.2465968222946397e-06,
      "loss": 1.0956,
      "step": 1004
    },
    {
      "epoch": 0.13010971939023205,
      "grad_norm": 0.20436461744410558,
      "learning_rate": 1.2465673208355619e-06,
      "loss": 1.0601,
      "step": 1005
    },
    {
      "epoch": 0.13010971939023205,
      "eval_PRM Accuracy": 0.5754716981132075,
      "eval_PRM F1": 0.6808510638297872,
      "eval_PRM F1 AUC": 0.5717653221581981,
      "eval_PRM F1 Neg": 0.36619718309859156,
      "eval_PRM NPV": 0.2708333333333333,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.5783132530120482,
      "eval_PRM Specificty": 0.5652173913043478,
      "eval_loss": 0.6745383739471436,
      "eval_runtime": 20.4321,
      "eval_samples_per_second": 2.153,
      "eval_steps_per_second": 0.147,
      "step": 1005
    },
    {
      "epoch": 0.13023918179758553,
      "grad_norm": 0.23021855871720354,
      "learning_rate": 1.2465376924087603e-06,
      "loss": 1.0531,
      "step": 1006
    },
    {
      "epoch": 0.130368644204939,
      "grad_norm": 0.3422439005229899,
      "learning_rate": 1.246507937020287e-06,
      "loss": 1.0497,
      "step": 1007
    },
    {
      "epoch": 0.13049810661229247,
      "grad_norm": 0.2537515597961894,
      "learning_rate": 1.2464780546762201e-06,
      "loss": 0.9344,
      "step": 1008
    },
    {
      "epoch": 0.13062756901964592,
      "grad_norm": 0.2478210342260897,
      "learning_rate": 1.2464480453826638e-06,
      "loss": 1.0673,
      "step": 1009
    },
    {
      "epoch": 0.13075703142699938,
      "grad_norm": 0.2969920906971841,
      "learning_rate": 1.2464179091457482e-06,
      "loss": 1.0309,
      "step": 1010
    },
    {
      "epoch": 0.13075703142699938,
      "eval_PRM Accuracy": 0.5943396226415094,
      "eval_PRM F1": 0.6950354609929078,
      "eval_PRM F1 AUC": 0.5995285489785228,
      "eval_PRM F1 Neg": 0.39436619718309857,
      "eval_PRM NPV": 0.2916666666666667,
      "eval_PRM Precision": 0.8448275862068966,
      "eval_PRM Recall": 0.5903614457831325,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.673828125,
      "eval_runtime": 20.8577,
      "eval_samples_per_second": 2.11,
      "eval_steps_per_second": 0.144,
      "step": 1010
    },
    {
      "epoch": 0.13088649383435286,
      "grad_norm": 0.1456697221631524,
      "learning_rate": 1.246387645971629e-06,
      "loss": 0.9285,
      "step": 1011
    },
    {
      "epoch": 0.1310159562417063,
      "grad_norm": 0.22196114765510896,
      "learning_rate": 1.246357255866488e-06,
      "loss": 1.0524,
      "step": 1012
    },
    {
      "epoch": 0.1311454186490598,
      "grad_norm": 0.13778280734962478,
      "learning_rate": 1.2463267388365332e-06,
      "loss": 0.9579,
      "step": 1013
    },
    {
      "epoch": 0.13127488105641325,
      "grad_norm": 0.23098696901042143,
      "learning_rate": 1.2462960948879985e-06,
      "loss": 0.9567,
      "step": 1014
    },
    {
      "epoch": 0.1314043434637667,
      "grad_norm": 0.17946499246819086,
      "learning_rate": 1.246265324027143e-06,
      "loss": 0.9019,
      "step": 1015
    },
    {
      "epoch": 0.1314043434637667,
      "eval_PRM Accuracy": 0.5754716981132075,
      "eval_PRM F1": 0.6808510638297872,
      "eval_PRM F1 AUC": 0.5717653221581981,
      "eval_PRM F1 Neg": 0.36619718309859156,
      "eval_PRM NPV": 0.2708333333333333,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.5783132530120482,
      "eval_PRM Specificty": 0.5652173913043478,
      "eval_loss": 0.6695667505264282,
      "eval_runtime": 20.6988,
      "eval_samples_per_second": 2.126,
      "eval_steps_per_second": 0.145,
      "step": 1015
    },
    {
      "epoch": 0.13153380587112018,
      "grad_norm": 0.23314322469466636,
      "learning_rate": 1.2462344262602528e-06,
      "loss": 1.0779,
      "step": 1016
    },
    {
      "epoch": 0.13166326827847363,
      "grad_norm": 0.17306902653562214,
      "learning_rate": 1.2462034015936392e-06,
      "loss": 1.0325,
      "step": 1017
    },
    {
      "epoch": 0.13179273068582711,
      "grad_norm": 0.21514654605891248,
      "learning_rate": 1.2461722500336397e-06,
      "loss": 0.9264,
      "step": 1018
    },
    {
      "epoch": 0.13192219309318057,
      "grad_norm": 0.1431906512523626,
      "learning_rate": 1.2461409715866172e-06,
      "loss": 0.9675,
      "step": 1019
    },
    {
      "epoch": 0.13205165550053402,
      "grad_norm": 0.25929792973154747,
      "learning_rate": 1.2461095662589615e-06,
      "loss": 0.9178,
      "step": 1020
    },
    {
      "epoch": 0.13205165550053402,
      "eval_PRM Accuracy": 0.5849056603773585,
      "eval_PRM F1": 0.6901408450704225,
      "eval_PRM F1 AUC": 0.5777894185437402,
      "eval_PRM F1 Neg": 0.37142857142857144,
      "eval_PRM NPV": 0.2765957446808511,
      "eval_PRM Precision": 0.8305084745762712,
      "eval_PRM Recall": 0.5903614457831325,
      "eval_PRM Specificty": 0.5652173913043478,
      "eval_loss": 0.6693891882896423,
      "eval_runtime": 20.1923,
      "eval_samples_per_second": 2.179,
      "eval_steps_per_second": 0.149,
      "step": 1020
    },
    {
      "epoch": 0.1321811179078875,
      "grad_norm": 0.28890130395388486,
      "learning_rate": 1.2460780340570877e-06,
      "loss": 1.08,
      "step": 1021
    },
    {
      "epoch": 0.13231058031524096,
      "grad_norm": 0.2029391845698423,
      "learning_rate": 1.2460463749874366e-06,
      "loss": 1.0446,
      "step": 1022
    },
    {
      "epoch": 0.13244004272259444,
      "grad_norm": 0.2532782101717443,
      "learning_rate": 1.2460145890564754e-06,
      "loss": 0.9976,
      "step": 1023
    },
    {
      "epoch": 0.1325695051299479,
      "grad_norm": 0.2563246227815142,
      "learning_rate": 1.245982676270697e-06,
      "loss": 1.0104,
      "step": 1024
    },
    {
      "epoch": 0.13269896753730134,
      "grad_norm": 0.2145449015549503,
      "learning_rate": 1.24595063663662e-06,
      "loss": 1.0355,
      "step": 1025
    },
    {
      "epoch": 0.13269896753730134,
      "eval_PRM Accuracy": 0.5754716981132075,
      "eval_PRM F1": 0.6808510638297872,
      "eval_PRM F1 AUC": 0.5717653221581981,
      "eval_PRM F1 Neg": 0.36619718309859156,
      "eval_PRM NPV": 0.2708333333333333,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.5783132530120482,
      "eval_PRM Specificty": 0.5652173913043478,
      "eval_loss": 0.6697443127632141,
      "eval_runtime": 21.068,
      "eval_samples_per_second": 2.088,
      "eval_steps_per_second": 0.142,
      "step": 1025
    },
    {
      "epoch": 0.13282842994465482,
      "grad_norm": 0.3562448765953552,
      "learning_rate": 1.2459184701607896e-06,
      "loss": 1.0298,
      "step": 1026
    },
    {
      "epoch": 0.13295789235200828,
      "grad_norm": 0.1986236893307979,
      "learning_rate": 1.2458861768497763e-06,
      "loss": 0.9631,
      "step": 1027
    },
    {
      "epoch": 0.13308735475936176,
      "grad_norm": 0.16162216689499773,
      "learning_rate": 1.2458537567101765e-06,
      "loss": 0.9766,
      "step": 1028
    },
    {
      "epoch": 0.1332168171667152,
      "grad_norm": 0.1790228284040299,
      "learning_rate": 1.2458212097486128e-06,
      "loss": 1.0979,
      "step": 1029
    },
    {
      "epoch": 0.1333462795740687,
      "grad_norm": 0.27660569911587,
      "learning_rate": 1.2457885359717334e-06,
      "loss": 1.0488,
      "step": 1030
    },
    {
      "epoch": 0.1333462795740687,
      "eval_PRM Accuracy": 0.5849056603773585,
      "eval_PRM F1": 0.6901408450704225,
      "eval_PRM F1 AUC": 0.5777894185437402,
      "eval_PRM F1 Neg": 0.37142857142857144,
      "eval_PRM NPV": 0.2765957446808511,
      "eval_PRM Precision": 0.8305084745762712,
      "eval_PRM Recall": 0.5903614457831325,
      "eval_PRM Specificty": 0.5652173913043478,
      "eval_loss": 0.6690340638160706,
      "eval_runtime": 20.2919,
      "eval_samples_per_second": 2.168,
      "eval_steps_per_second": 0.148,
      "step": 1030
    },
    {
      "epoch": 0.13347574198142215,
      "grad_norm": 0.25695314009577197,
      "learning_rate": 1.245755735386213e-06,
      "loss": 1.0172,
      "step": 1031
    },
    {
      "epoch": 0.1336052043887756,
      "grad_norm": 0.29544455417263854,
      "learning_rate": 1.2457228079987512e-06,
      "loss": 1.0432,
      "step": 1032
    },
    {
      "epoch": 0.13373466679612908,
      "grad_norm": 0.21068471114535248,
      "learning_rate": 1.2456897538160745e-06,
      "loss": 0.8672,
      "step": 1033
    },
    {
      "epoch": 0.13386412920348253,
      "grad_norm": 0.22065171924668514,
      "learning_rate": 1.2456565728449349e-06,
      "loss": 0.9938,
      "step": 1034
    },
    {
      "epoch": 0.13399359161083602,
      "grad_norm": 0.1213019541307578,
      "learning_rate": 1.24562326509211e-06,
      "loss": 1.0112,
      "step": 1035
    },
    {
      "epoch": 0.13399359161083602,
      "eval_PRM Accuracy": 0.5943396226415094,
      "eval_PRM F1": 0.6993006993006993,
      "eval_PRM F1 AUC": 0.5838135149292825,
      "eval_PRM F1 Neg": 0.37681159420289856,
      "eval_PRM NPV": 0.2826086956521739,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.6024096385542169,
      "eval_PRM Specificty": 0.5652173913043478,
      "eval_loss": 0.6656605005264282,
      "eval_runtime": 20.589,
      "eval_samples_per_second": 2.137,
      "eval_steps_per_second": 0.146,
      "step": 1035
    },
    {
      "epoch": 0.13412305401818947,
      "grad_norm": 0.28540930465649333,
      "learning_rate": 1.2455898305644038e-06,
      "loss": 1.0156,
      "step": 1036
    },
    {
      "epoch": 0.13425251642554292,
      "grad_norm": 0.20180014069901442,
      "learning_rate": 1.2455562692686459e-06,
      "loss": 0.9778,
      "step": 1037
    },
    {
      "epoch": 0.1343819788328964,
      "grad_norm": 0.23203348616684075,
      "learning_rate": 1.245522581211692e-06,
      "loss": 1.0285,
      "step": 1038
    },
    {
      "epoch": 0.13451144124024986,
      "grad_norm": 0.34276355361359817,
      "learning_rate": 1.2454887664004233e-06,
      "loss": 1.0486,
      "step": 1039
    },
    {
      "epoch": 0.13464090364760334,
      "grad_norm": 0.3717587522722963,
      "learning_rate": 1.2454548248417475e-06,
      "loss": 1.1532,
      "step": 1040
    },
    {
      "epoch": 0.13464090364760334,
      "eval_PRM Accuracy": 0.5943396226415094,
      "eval_PRM F1": 0.6993006993006993,
      "eval_PRM F1 AUC": 0.5838135149292825,
      "eval_PRM F1 Neg": 0.37681159420289856,
      "eval_PRM NPV": 0.2826086956521739,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.6024096385542169,
      "eval_PRM Specificty": 0.5652173913043478,
      "eval_loss": 0.6658380627632141,
      "eval_runtime": 20.3046,
      "eval_samples_per_second": 2.167,
      "eval_steps_per_second": 0.148,
      "step": 1040
    },
    {
      "epoch": 0.1347703660549568,
      "grad_norm": 0.21517024204210905,
      "learning_rate": 1.2454207565425973e-06,
      "loss": 0.8998,
      "step": 1041
    },
    {
      "epoch": 0.13489982846231025,
      "grad_norm": 0.25120559654254815,
      "learning_rate": 1.2453865615099327e-06,
      "loss": 0.8973,
      "step": 1042
    },
    {
      "epoch": 0.13502929086966373,
      "grad_norm": 0.23235501137393036,
      "learning_rate": 1.245352239750738e-06,
      "loss": 0.9027,
      "step": 1043
    },
    {
      "epoch": 0.13515875327701718,
      "grad_norm": 0.3161268040685229,
      "learning_rate": 1.2453177912720244e-06,
      "loss": 0.9974,
      "step": 1044
    },
    {
      "epoch": 0.13528821568437066,
      "grad_norm": 0.26720379109445447,
      "learning_rate": 1.2452832160808286e-06,
      "loss": 0.8905,
      "step": 1045
    },
    {
      "epoch": 0.13528821568437066,
      "eval_PRM Accuracy": 0.5849056603773585,
      "eval_PRM F1": 0.6944444444444444,
      "eval_PRM F1 AUC": 0.5620743844944998,
      "eval_PRM F1 Neg": 0.35294117647058826,
      "eval_PRM NPV": 0.26666666666666666,
      "eval_PRM Precision": 0.819672131147541,
      "eval_PRM Recall": 0.6024096385542169,
      "eval_PRM Specificty": 0.5217391304347826,
      "eval_loss": 0.6638849377632141,
      "eval_runtime": 20.3888,
      "eval_samples_per_second": 2.158,
      "eval_steps_per_second": 0.147,
      "step": 1045
    },
    {
      "epoch": 0.13541767809172411,
      "grad_norm": 0.3040220123591935,
      "learning_rate": 1.2452485141842135e-06,
      "loss": 0.9663,
      "step": 1046
    },
    {
      "epoch": 0.13554714049907757,
      "grad_norm": 0.2704010975668842,
      "learning_rate": 1.2452136855892674e-06,
      "loss": 0.9642,
      "step": 1047
    },
    {
      "epoch": 0.13567660290643105,
      "grad_norm": 0.19828000738747453,
      "learning_rate": 1.2451787303031051e-06,
      "loss": 1.0009,
      "step": 1048
    },
    {
      "epoch": 0.1358060653137845,
      "grad_norm": 0.20772939139253724,
      "learning_rate": 1.2451436483328663e-06,
      "loss": 1.0367,
      "step": 1049
    },
    {
      "epoch": 0.13593552772113798,
      "grad_norm": 0.17308609144173692,
      "learning_rate": 1.2451084396857178e-06,
      "loss": 1.0103,
      "step": 1050
    },
    {
      "epoch": 0.13593552772113798,
      "eval_PRM Accuracy": 0.6037735849056604,
      "eval_PRM F1": 0.7083333333333334,
      "eval_PRM F1 AUC": 0.5898376113148246,
      "eval_PRM F1 Neg": 0.38235294117647056,
      "eval_PRM NPV": 0.28888888888888886,
      "eval_PRM Precision": 0.8360655737704918,
      "eval_PRM Recall": 0.6144578313253012,
      "eval_PRM Specificty": 0.5652173913043478,
      "eval_loss": 0.662109375,
      "eval_runtime": 20.3135,
      "eval_samples_per_second": 2.166,
      "eval_steps_per_second": 0.148,
      "step": 1050
    },
    {
      "epoch": 0.13606499012849144,
      "grad_norm": 0.1584196333045882,
      "learning_rate": 1.2450731043688515e-06,
      "loss": 0.9424,
      "step": 1051
    },
    {
      "epoch": 0.13619445253584492,
      "grad_norm": 0.23584540516887204,
      "learning_rate": 1.2450376423894853e-06,
      "loss": 1.0431,
      "step": 1052
    },
    {
      "epoch": 0.13632391494319837,
      "grad_norm": 0.22251339316286461,
      "learning_rate": 1.245002053754863e-06,
      "loss": 1.0635,
      "step": 1053
    },
    {
      "epoch": 0.13645337735055182,
      "grad_norm": 0.23109857776369042,
      "learning_rate": 1.2449663384722543e-06,
      "loss": 1.0686,
      "step": 1054
    },
    {
      "epoch": 0.1365828397579053,
      "grad_norm": 0.2412197232885547,
      "learning_rate": 1.244930496548955e-06,
      "loss": 0.9684,
      "step": 1055
    },
    {
      "epoch": 0.1365828397579053,
      "eval_PRM Accuracy": 0.6037735849056604,
      "eval_PRM F1": 0.7083333333333334,
      "eval_PRM F1 AUC": 0.5898376113148246,
      "eval_PRM F1 Neg": 0.38235294117647056,
      "eval_PRM NPV": 0.28888888888888886,
      "eval_PRM Precision": 0.8360655737704918,
      "eval_PRM Recall": 0.6144578313253012,
      "eval_PRM Specificty": 0.5652173913043478,
      "eval_loss": 0.6610440611839294,
      "eval_runtime": 20.2839,
      "eval_samples_per_second": 2.169,
      "eval_steps_per_second": 0.148,
      "step": 1055
    },
    {
      "epoch": 0.13671230216525876,
      "grad_norm": 0.2646179233825078,
      "learning_rate": 1.2448945279922863e-06,
      "loss": 1.0713,
      "step": 1056
    },
    {
      "epoch": 0.13684176457261224,
      "grad_norm": 0.21067864403224032,
      "learning_rate": 1.2448584328095955e-06,
      "loss": 0.9968,
      "step": 1057
    },
    {
      "epoch": 0.1369712269799657,
      "grad_norm": 0.2051303909570104,
      "learning_rate": 1.2448222110082557e-06,
      "loss": 0.9368,
      "step": 1058
    },
    {
      "epoch": 0.13710068938731915,
      "grad_norm": 0.1785550007935163,
      "learning_rate": 1.2447858625956663e-06,
      "loss": 0.9249,
      "step": 1059
    },
    {
      "epoch": 0.13723015179467263,
      "grad_norm": 0.28213780377358494,
      "learning_rate": 1.244749387579252e-06,
      "loss": 0.9948,
      "step": 1060
    },
    {
      "epoch": 0.13723015179467263,
      "eval_PRM Accuracy": 0.5943396226415094,
      "eval_PRM F1": 0.7034482758620689,
      "eval_PRM F1 AUC": 0.568098480880042,
      "eval_PRM F1 Neg": 0.3582089552238806,
      "eval_PRM NPV": 0.2727272727272727,
      "eval_PRM Precision": 0.8225806451612904,
      "eval_PRM Recall": 0.6144578313253012,
      "eval_PRM Specificty": 0.5217391304347826,
      "eval_loss": 0.658203125,
      "eval_runtime": 20.639,
      "eval_samples_per_second": 2.132,
      "eval_steps_per_second": 0.145,
      "step": 1060
    },
    {
      "epoch": 0.13735961420202608,
      "grad_norm": 0.17042696246103253,
      "learning_rate": 1.2447127859664633e-06,
      "loss": 1.0923,
      "step": 1061
    },
    {
      "epoch": 0.13748907660937956,
      "grad_norm": 0.14245086193038814,
      "learning_rate": 1.2446760577647774e-06,
      "loss": 0.8635,
      "step": 1062
    },
    {
      "epoch": 0.13761853901673302,
      "grad_norm": 0.14250062361768437,
      "learning_rate": 1.2446392029816962e-06,
      "loss": 0.939,
      "step": 1063
    },
    {
      "epoch": 0.13774800142408647,
      "grad_norm": 0.18936695518634158,
      "learning_rate": 1.2446022216247482e-06,
      "loss": 0.9949,
      "step": 1064
    },
    {
      "epoch": 0.13787746383143995,
      "grad_norm": 0.11826945092664617,
      "learning_rate": 1.244565113701488e-06,
      "loss": 1.0474,
      "step": 1065
    },
    {
      "epoch": 0.13787746383143995,
      "eval_PRM Accuracy": 0.6132075471698113,
      "eval_PRM F1": 0.7210884353741497,
      "eval_PRM F1 AUC": 0.5801466736511263,
      "eval_PRM F1 Neg": 0.36923076923076925,
      "eval_PRM NPV": 0.2857142857142857,
      "eval_PRM Precision": 0.828125,
      "eval_PRM Recall": 0.6385542168674698,
      "eval_PRM Specificty": 0.5217391304347826,
      "eval_loss": 0.6571378111839294,
      "eval_runtime": 19.8207,
      "eval_samples_per_second": 2.22,
      "eval_steps_per_second": 0.151,
      "step": 1065
    },
    {
      "epoch": 0.1380069262387934,
      "grad_norm": 0.2671162874062274,
      "learning_rate": 1.2445278792194952e-06,
      "loss": 0.9613,
      "step": 1066
    },
    {
      "epoch": 0.13813638864614688,
      "grad_norm": 0.2750132139664363,
      "learning_rate": 1.2444905181863756e-06,
      "loss": 0.9689,
      "step": 1067
    },
    {
      "epoch": 0.13826585105350034,
      "grad_norm": 0.15478729380022627,
      "learning_rate": 1.2444530306097612e-06,
      "loss": 1.0011,
      "step": 1068
    },
    {
      "epoch": 0.1383953134608538,
      "grad_norm": 0.2912392109241785,
      "learning_rate": 1.2444154164973097e-06,
      "loss": 1.0,
      "step": 1069
    },
    {
      "epoch": 0.13852477586820727,
      "grad_norm": 0.23857068592664332,
      "learning_rate": 1.2443776758567045e-06,
      "loss": 0.9635,
      "step": 1070
    },
    {
      "epoch": 0.13852477586820727,
      "eval_PRM Accuracy": 0.6037735849056604,
      "eval_PRM F1": 0.7123287671232876,
      "eval_PRM F1 AUC": 0.574122577265584,
      "eval_PRM F1 Neg": 0.36363636363636365,
      "eval_PRM NPV": 0.27906976744186046,
      "eval_PRM Precision": 0.8253968253968254,
      "eval_PRM Recall": 0.6265060240963856,
      "eval_PRM Specificty": 0.5217391304347826,
      "eval_loss": 0.65625,
      "eval_runtime": 20.1268,
      "eval_samples_per_second": 2.186,
      "eval_steps_per_second": 0.149,
      "step": 1070
    },
    {
      "epoch": 0.13865423827556073,
      "grad_norm": 0.32792641540608475,
      "learning_rate": 1.2443398086956545e-06,
      "loss": 1.0275,
      "step": 1071
    },
    {
      "epoch": 0.1387837006829142,
      "grad_norm": 0.14543973335668856,
      "learning_rate": 1.2443018150218954e-06,
      "loss": 1.038,
      "step": 1072
    },
    {
      "epoch": 0.13891316309026766,
      "grad_norm": 0.253594381927539,
      "learning_rate": 1.2442636948431877e-06,
      "loss": 0.8894,
      "step": 1073
    },
    {
      "epoch": 0.13904262549762114,
      "grad_norm": 0.19302270068739436,
      "learning_rate": 1.2442254481673185e-06,
      "loss": 1.0988,
      "step": 1074
    },
    {
      "epoch": 0.1391720879049746,
      "grad_norm": 0.21853701059011885,
      "learning_rate": 1.2441870750021003e-06,
      "loss": 0.9866,
      "step": 1075
    },
    {
      "epoch": 0.1391720879049746,
      "eval_PRM Accuracy": 0.6132075471698113,
      "eval_PRM F1": 0.7210884353741497,
      "eval_PRM F1 AUC": 0.5801466736511263,
      "eval_PRM F1 Neg": 0.36923076923076925,
      "eval_PRM NPV": 0.2857142857142857,
      "eval_PRM Precision": 0.828125,
      "eval_PRM Recall": 0.6385542168674698,
      "eval_PRM Specificty": 0.5217391304347826,
      "eval_loss": 0.6550071239471436,
      "eval_runtime": 20.7689,
      "eval_samples_per_second": 2.119,
      "eval_steps_per_second": 0.144,
      "step": 1075
    },
    {
      "epoch": 0.13930155031232805,
      "grad_norm": 0.1841902217872951,
      "learning_rate": 1.244148575355372e-06,
      "loss": 1.0803,
      "step": 1076
    },
    {
      "epoch": 0.13943101271968153,
      "grad_norm": 0.24699645704664566,
      "learning_rate": 1.2441099492349973e-06,
      "loss": 0.9568,
      "step": 1077
    },
    {
      "epoch": 0.13956047512703498,
      "grad_norm": 0.23450421472456093,
      "learning_rate": 1.2440711966488666e-06,
      "loss": 1.0083,
      "step": 1078
    },
    {
      "epoch": 0.13968993753438846,
      "grad_norm": 0.32027351722699277,
      "learning_rate": 1.2440323176048964e-06,
      "loss": 0.9396,
      "step": 1079
    },
    {
      "epoch": 0.13981939994174192,
      "grad_norm": 0.18195461221669762,
      "learning_rate": 1.2439933121110278e-06,
      "loss": 0.8784,
      "step": 1080
    },
    {
      "epoch": 0.13981939994174192,
      "eval_PRM Accuracy": 0.6320754716981132,
      "eval_PRM F1": 0.738255033557047,
      "eval_PRM F1 AUC": 0.5921948664222106,
      "eval_PRM F1 Neg": 0.38095238095238093,
      "eval_PRM NPV": 0.3,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.6626506024096386,
      "eval_PRM Specificty": 0.5217391304347826,
      "eval_loss": 0.6537641882896423,
      "eval_runtime": 20.3023,
      "eval_samples_per_second": 2.167,
      "eval_steps_per_second": 0.148,
      "step": 1080
    },
    {
      "epoch": 0.13994886234909537,
      "grad_norm": 0.2782958626827339,
      "learning_rate": 1.2439541801752288e-06,
      "loss": 1.0144,
      "step": 1081
    },
    {
      "epoch": 0.14007832475644885,
      "grad_norm": 0.14277695971202406,
      "learning_rate": 1.2439149218054928e-06,
      "loss": 0.9451,
      "step": 1082
    },
    {
      "epoch": 0.1402077871638023,
      "grad_norm": 0.20788048238522783,
      "learning_rate": 1.2438755370098393e-06,
      "loss": 1.0299,
      "step": 1083
    },
    {
      "epoch": 0.1403372495711558,
      "grad_norm": 0.22553031472889404,
      "learning_rate": 1.2438360257963134e-06,
      "loss": 0.9641,
      "step": 1084
    },
    {
      "epoch": 0.14046671197850924,
      "grad_norm": 0.20000495788314998,
      "learning_rate": 1.2437963881729858e-06,
      "loss": 1.0332,
      "step": 1085
    },
    {
      "epoch": 0.14046671197850924,
      "eval_PRM Accuracy": 0.6226415094339622,
      "eval_PRM F1": 0.7297297297297297,
      "eval_PRM F1 AUC": 0.5861707700366684,
      "eval_PRM F1 Neg": 0.375,
      "eval_PRM NPV": 0.2926829268292683,
      "eval_PRM Precision": 0.8307692307692308,
      "eval_PRM Recall": 0.6506024096385542,
      "eval_PRM Specificty": 0.5217391304347826,
      "eval_loss": 0.6512784361839294,
      "eval_runtime": 20.8809,
      "eval_samples_per_second": 2.107,
      "eval_steps_per_second": 0.144,
      "step": 1085
    },
    {
      "epoch": 0.1405961743858627,
      "grad_norm": 0.33025587766120834,
      "learning_rate": 1.2437566241479536e-06,
      "loss": 0.9672,
      "step": 1086
    },
    {
      "epoch": 0.14072563679321617,
      "grad_norm": 0.1812910575071967,
      "learning_rate": 1.2437167337293393e-06,
      "loss": 0.954,
      "step": 1087
    },
    {
      "epoch": 0.14085509920056963,
      "grad_norm": 0.19942063996669387,
      "learning_rate": 1.2436767169252915e-06,
      "loss": 0.9608,
      "step": 1088
    },
    {
      "epoch": 0.1409845616079231,
      "grad_norm": 0.23079586086775253,
      "learning_rate": 1.243636573743984e-06,
      "loss": 0.9846,
      "step": 1089
    },
    {
      "epoch": 0.14111402401527656,
      "grad_norm": 0.14866300838109334,
      "learning_rate": 1.2435963041936173e-06,
      "loss": 0.8832,
      "step": 1090
    },
    {
      "epoch": 0.14111402401527656,
      "eval_PRM Accuracy": 0.6320754716981132,
      "eval_PRM F1": 0.738255033557047,
      "eval_PRM F1 AUC": 0.5921948664222106,
      "eval_PRM F1 Neg": 0.38095238095238093,
      "eval_PRM NPV": 0.3,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.6626506024096386,
      "eval_PRM Specificty": 0.5217391304347826,
      "eval_loss": 0.6534090638160706,
      "eval_runtime": 20.3479,
      "eval_samples_per_second": 2.162,
      "eval_steps_per_second": 0.147,
      "step": 1090
    },
    {
      "epoch": 0.14124348642263002,
      "grad_norm": 0.23079814525451284,
      "learning_rate": 1.2435559082824169e-06,
      "loss": 0.9684,
      "step": 1091
    },
    {
      "epoch": 0.1413729488299835,
      "grad_norm": 0.23012922487658394,
      "learning_rate": 1.2435153860186347e-06,
      "loss": 0.953,
      "step": 1092
    },
    {
      "epoch": 0.14150241123733695,
      "grad_norm": 0.15945126759279368,
      "learning_rate": 1.2434747374105482e-06,
      "loss": 0.8815,
      "step": 1093
    },
    {
      "epoch": 0.14163187364469043,
      "grad_norm": 0.2309600948251421,
      "learning_rate": 1.2434339624664608e-06,
      "loss": 1.0181,
      "step": 1094
    },
    {
      "epoch": 0.14176133605204388,
      "grad_norm": 0.3146126630911263,
      "learning_rate": 1.2433930611947015e-06,
      "loss": 1.0894,
      "step": 1095
    },
    {
      "epoch": 0.14176133605204388,
      "eval_PRM Accuracy": 0.6320754716981132,
      "eval_PRM F1": 0.738255033557047,
      "eval_PRM F1 AUC": 0.5921948664222106,
      "eval_PRM F1 Neg": 0.38095238095238093,
      "eval_PRM NPV": 0.3,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.6626506024096386,
      "eval_PRM Specificty": 0.5217391304347826,
      "eval_loss": 0.6530539989471436,
      "eval_runtime": 20.2983,
      "eval_samples_per_second": 2.168,
      "eval_steps_per_second": 0.148,
      "step": 1095
    },
    {
      "epoch": 0.14189079845939737,
      "grad_norm": 0.2753802426544741,
      "learning_rate": 1.243352033603625e-06,
      "loss": 0.9553,
      "step": 1096
    },
    {
      "epoch": 0.14202026086675082,
      "grad_norm": 0.23687278613394785,
      "learning_rate": 1.2433108797016124e-06,
      "loss": 0.8784,
      "step": 1097
    },
    {
      "epoch": 0.14214972327410427,
      "grad_norm": 0.2956630598192226,
      "learning_rate": 1.2432695994970698e-06,
      "loss": 1.0517,
      "step": 1098
    },
    {
      "epoch": 0.14227918568145775,
      "grad_norm": 0.15684894444736966,
      "learning_rate": 1.24322819299843e-06,
      "loss": 0.9951,
      "step": 1099
    },
    {
      "epoch": 0.1424086480888112,
      "grad_norm": 0.2172842398770529,
      "learning_rate": 1.2431866602141506e-06,
      "loss": 1.0277,
      "step": 1100
    },
    {
      "epoch": 0.1424086480888112,
      "eval_PRM Accuracy": 0.6226415094339622,
      "eval_PRM F1": 0.7333333333333333,
      "eval_PRM F1 AUC": 0.5704557359874279,
      "eval_PRM F1 Neg": 0.3548387096774194,
      "eval_PRM NPV": 0.28205128205128205,
      "eval_PRM Precision": 0.8208955223880597,
      "eval_PRM Recall": 0.6626506024096386,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6539417505264282,
      "eval_runtime": 20.7973,
      "eval_samples_per_second": 2.116,
      "eval_steps_per_second": 0.144,
      "step": 1100
    },
    {
      "epoch": 0.1425381104961647,
      "grad_norm": 0.2700718674867675,
      "learning_rate": 1.243145001152716e-06,
      "loss": 0.9775,
      "step": 1101
    },
    {
      "epoch": 0.14266757290351814,
      "grad_norm": 0.3524104019445697,
      "learning_rate": 1.2431032158226356e-06,
      "loss": 0.989,
      "step": 1102
    },
    {
      "epoch": 0.1427970353108716,
      "grad_norm": 0.1702607741761378,
      "learning_rate": 1.243061304232445e-06,
      "loss": 1.0439,
      "step": 1103
    },
    {
      "epoch": 0.14292649771822508,
      "grad_norm": 0.1684005721650078,
      "learning_rate": 1.2430192663907054e-06,
      "loss": 0.9598,
      "step": 1104
    },
    {
      "epoch": 0.14305596012557853,
      "grad_norm": 0.1912667841555838,
      "learning_rate": 1.2429771023060042e-06,
      "loss": 0.981,
      "step": 1105
    },
    {
      "epoch": 0.14305596012557853,
      "eval_PRM Accuracy": 0.6320754716981132,
      "eval_PRM F1": 0.7417218543046358,
      "eval_PRM F1 AUC": 0.5764798323729701,
      "eval_PRM F1 Neg": 0.36065573770491804,
      "eval_PRM NPV": 0.2894736842105263,
      "eval_PRM Precision": 0.8235294117647058,
      "eval_PRM Recall": 0.6746987951807228,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6491477489471436,
      "eval_runtime": 20.4048,
      "eval_samples_per_second": 2.156,
      "eval_steps_per_second": 0.147,
      "step": 1105
    },
    {
      "epoch": 0.143185422532932,
      "grad_norm": 0.23613403596122787,
      "learning_rate": 1.2429348119869536e-06,
      "loss": 0.9326,
      "step": 1106
    },
    {
      "epoch": 0.14331488494028546,
      "grad_norm": 0.22717420536125202,
      "learning_rate": 1.2428923954421932e-06,
      "loss": 0.9415,
      "step": 1107
    },
    {
      "epoch": 0.14344434734763892,
      "grad_norm": 0.3005846779846251,
      "learning_rate": 1.2428498526803866e-06,
      "loss": 1.0387,
      "step": 1108
    },
    {
      "epoch": 0.1435738097549924,
      "grad_norm": 0.11950425559752326,
      "learning_rate": 1.2428071837102243e-06,
      "loss": 0.9906,
      "step": 1109
    },
    {
      "epoch": 0.14370327216234585,
      "grad_norm": 0.2183126388488348,
      "learning_rate": 1.2427643885404225e-06,
      "loss": 0.912,
      "step": 1110
    },
    {
      "epoch": 0.14370327216234585,
      "eval_PRM Accuracy": 0.6509433962264151,
      "eval_PRM F1": 0.7549668874172185,
      "eval_PRM F1 AUC": 0.6042430591932949,
      "eval_PRM F1 Neg": 0.39344262295081966,
      "eval_PRM NPV": 0.3157894736842105,
      "eval_PRM Precision": 0.8382352941176471,
      "eval_PRM Recall": 0.6867469879518072,
      "eval_PRM Specificty": 0.5217391304347826,
      "eval_loss": 0.6509233117103577,
      "eval_runtime": 20.3482,
      "eval_samples_per_second": 2.162,
      "eval_steps_per_second": 0.147,
      "step": 1110
    },
    {
      "epoch": 0.14383273456969933,
      "grad_norm": 0.17744916019756846,
      "learning_rate": 1.2427214671797228e-06,
      "loss": 0.9098,
      "step": 1111
    },
    {
      "epoch": 0.1439621969770528,
      "grad_norm": 0.24246746712156525,
      "learning_rate": 1.2426784196368927e-06,
      "loss": 0.9731,
      "step": 1112
    },
    {
      "epoch": 0.14409165938440624,
      "grad_norm": 0.22168722938360055,
      "learning_rate": 1.242635245920726e-06,
      "loss": 1.0226,
      "step": 1113
    },
    {
      "epoch": 0.14422112179175972,
      "grad_norm": 0.270987236674979,
      "learning_rate": 1.242591946040041e-06,
      "loss": 0.8828,
      "step": 1114
    },
    {
      "epoch": 0.14435058419911317,
      "grad_norm": 0.24876398263225982,
      "learning_rate": 1.242548520003683e-06,
      "loss": 0.9587,
      "step": 1115
    },
    {
      "epoch": 0.14435058419911317,
      "eval_PRM Accuracy": 0.6415094339622641,
      "eval_PRM F1": 0.75,
      "eval_PRM F1 AUC": 0.5825039287585123,
      "eval_PRM F1 Neg": 0.36666666666666664,
      "eval_PRM NPV": 0.2972972972972973,
      "eval_PRM Precision": 0.8260869565217391,
      "eval_PRM Recall": 0.6867469879518072,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6468394994735718,
      "eval_runtime": 20.2281,
      "eval_samples_per_second": 2.175,
      "eval_steps_per_second": 0.148,
      "step": 1115
    },
    {
      "epoch": 0.14448004660646666,
      "grad_norm": 0.1938556458352197,
      "learning_rate": 1.2425049678205228e-06,
      "loss": 1.0176,
      "step": 1116
    },
    {
      "epoch": 0.1446095090138201,
      "grad_norm": 0.23350119319067147,
      "learning_rate": 1.2424612894994566e-06,
      "loss": 1.0762,
      "step": 1117
    },
    {
      "epoch": 0.1447389714211736,
      "grad_norm": 0.1775481890015851,
      "learning_rate": 1.2424174850494068e-06,
      "loss": 1.0663,
      "step": 1118
    },
    {
      "epoch": 0.14486843382852704,
      "grad_norm": 0.21097317577771219,
      "learning_rate": 1.2423735544793209e-06,
      "loss": 1.0522,
      "step": 1119
    },
    {
      "epoch": 0.1449978962358805,
      "grad_norm": 0.3104672640607622,
      "learning_rate": 1.2423294977981732e-06,
      "loss": 0.9733,
      "step": 1120
    },
    {
      "epoch": 0.1449978962358805,
      "eval_PRM Accuracy": 0.6509433962264151,
      "eval_PRM F1": 0.7581699346405228,
      "eval_PRM F1 AUC": 0.5885280251440546,
      "eval_PRM F1 Neg": 0.3728813559322034,
      "eval_PRM NPV": 0.3055555555555556,
      "eval_PRM Precision": 0.8285714285714286,
      "eval_PRM Recall": 0.6987951807228916,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6457741260528564,
      "eval_runtime": 20.2755,
      "eval_samples_per_second": 2.17,
      "eval_steps_per_second": 0.148,
      "step": 1120
    },
    {
      "epoch": 0.14512735864323398,
      "grad_norm": 0.2554037544790402,
      "learning_rate": 1.2422853150149628e-06,
      "loss": 1.0018,
      "step": 1121
    },
    {
      "epoch": 0.14525682105058743,
      "grad_norm": 0.2653476801944702,
      "learning_rate": 1.2422410061387148e-06,
      "loss": 0.8772,
      "step": 1122
    },
    {
      "epoch": 0.1453862834579409,
      "grad_norm": 0.2299479736575289,
      "learning_rate": 1.2421965711784805e-06,
      "loss": 0.9036,
      "step": 1123
    },
    {
      "epoch": 0.14551574586529437,
      "grad_norm": 0.2609757048793732,
      "learning_rate": 1.2421520101433367e-06,
      "loss": 0.9426,
      "step": 1124
    },
    {
      "epoch": 0.14564520827264782,
      "grad_norm": 0.15295519773552457,
      "learning_rate": 1.2421073230423854e-06,
      "loss": 0.976,
      "step": 1125
    },
    {
      "epoch": 0.14564520827264782,
      "eval_PRM Accuracy": 0.6415094339622641,
      "eval_PRM F1": 0.75,
      "eval_PRM F1 AUC": 0.5825039287585123,
      "eval_PRM F1 Neg": 0.36666666666666664,
      "eval_PRM NPV": 0.2972972972972973,
      "eval_PRM Precision": 0.8260869565217391,
      "eval_PRM Recall": 0.6867469879518072,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6461292505264282,
      "eval_runtime": 20.4483,
      "eval_samples_per_second": 2.152,
      "eval_steps_per_second": 0.147,
      "step": 1125
    },
    {
      "epoch": 0.1457746706800013,
      "grad_norm": 0.20995254959442733,
      "learning_rate": 1.2420625098847555e-06,
      "loss": 0.926,
      "step": 1126
    },
    {
      "epoch": 0.14590413308735475,
      "grad_norm": 0.22832994659818454,
      "learning_rate": 1.2420175706796005e-06,
      "loss": 0.9348,
      "step": 1127
    },
    {
      "epoch": 0.14603359549470823,
      "grad_norm": 0.2707312544106785,
      "learning_rate": 1.2419725054361004e-06,
      "loss": 0.9698,
      "step": 1128
    },
    {
      "epoch": 0.1461630579020617,
      "grad_norm": 0.3298242876897229,
      "learning_rate": 1.2419273141634607e-06,
      "loss": 1.0163,
      "step": 1129
    },
    {
      "epoch": 0.14629252030941514,
      "grad_norm": 0.2043022960521275,
      "learning_rate": 1.2418819968709123e-06,
      "loss": 0.903,
      "step": 1130
    },
    {
      "epoch": 0.14629252030941514,
      "eval_PRM Accuracy": 0.6415094339622641,
      "eval_PRM F1": 0.75,
      "eval_PRM F1 AUC": 0.5825039287585123,
      "eval_PRM F1 Neg": 0.36666666666666664,
      "eval_PRM NPV": 0.2972972972972973,
      "eval_PRM Precision": 0.8260869565217391,
      "eval_PRM Recall": 0.6867469879518072,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.646484375,
      "eval_runtime": 19.8771,
      "eval_samples_per_second": 2.214,
      "eval_steps_per_second": 0.151,
      "step": 1130
    },
    {
      "epoch": 0.14642198271676862,
      "grad_norm": 0.26130171514096384,
      "learning_rate": 1.2418365535677127e-06,
      "loss": 1.0635,
      "step": 1131
    },
    {
      "epoch": 0.14655144512412208,
      "grad_norm": 0.23532230243927016,
      "learning_rate": 1.241790984263144e-06,
      "loss": 0.9934,
      "step": 1132
    },
    {
      "epoch": 0.14668090753147556,
      "grad_norm": 0.20931405859858235,
      "learning_rate": 1.2417452889665153e-06,
      "loss": 0.955,
      "step": 1133
    },
    {
      "epoch": 0.146810369938829,
      "grad_norm": 0.14838637913152816,
      "learning_rate": 1.2416994676871601e-06,
      "loss": 0.968,
      "step": 1134
    },
    {
      "epoch": 0.14693983234618246,
      "grad_norm": 0.1603067032203339,
      "learning_rate": 1.241653520434439e-06,
      "loss": 0.9318,
      "step": 1135
    },
    {
      "epoch": 0.14693983234618246,
      "eval_PRM Accuracy": 0.6509433962264151,
      "eval_PRM F1": 0.7581699346405228,
      "eval_PRM F1 AUC": 0.5885280251440546,
      "eval_PRM F1 Neg": 0.3728813559322034,
      "eval_PRM NPV": 0.3055555555555556,
      "eval_PRM Precision": 0.8285714285714286,
      "eval_PRM Recall": 0.6987951807228916,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6427556872367859,
      "eval_runtime": 20.1921,
      "eval_samples_per_second": 2.179,
      "eval_steps_per_second": 0.149,
      "step": 1135
    },
    {
      "epoch": 0.14706929475353595,
      "grad_norm": 0.2112575513086432,
      "learning_rate": 1.2416074472177371e-06,
      "loss": 0.9753,
      "step": 1136
    },
    {
      "epoch": 0.1471987571608894,
      "grad_norm": 0.3042693808272477,
      "learning_rate": 1.2415612480464663e-06,
      "loss": 1.0167,
      "step": 1137
    },
    {
      "epoch": 0.14732821956824288,
      "grad_norm": 0.25163994647635585,
      "learning_rate": 1.2415149229300632e-06,
      "loss": 1.002,
      "step": 1138
    },
    {
      "epoch": 0.14745768197559633,
      "grad_norm": 0.2086015451113152,
      "learning_rate": 1.2414684718779913e-06,
      "loss": 0.9711,
      "step": 1139
    },
    {
      "epoch": 0.14758714438294981,
      "grad_norm": 0.27625041813565593,
      "learning_rate": 1.2414218948997384e-06,
      "loss": 1.0404,
      "step": 1140
    },
    {
      "epoch": 0.14758714438294981,
      "eval_PRM Accuracy": 0.6509433962264151,
      "eval_PRM F1": 0.7581699346405228,
      "eval_PRM F1 AUC": 0.5885280251440546,
      "eval_PRM F1 Neg": 0.3728813559322034,
      "eval_PRM NPV": 0.3055555555555556,
      "eval_PRM Precision": 0.8285714285714286,
      "eval_PRM Recall": 0.6987951807228916,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6441761255264282,
      "eval_runtime": 19.8906,
      "eval_samples_per_second": 2.212,
      "eval_steps_per_second": 0.151,
      "step": 1140
    },
    {
      "epoch": 0.14771660679030327,
      "grad_norm": 0.20317518357281902,
      "learning_rate": 1.2413751920048193e-06,
      "loss": 0.9961,
      "step": 1141
    },
    {
      "epoch": 0.14784606919765672,
      "grad_norm": 0.16477324293602258,
      "learning_rate": 1.241328363202774e-06,
      "loss": 0.9927,
      "step": 1142
    },
    {
      "epoch": 0.1479755316050102,
      "grad_norm": 0.13844831679774353,
      "learning_rate": 1.2412814085031678e-06,
      "loss": 0.9045,
      "step": 1143
    },
    {
      "epoch": 0.14810499401236366,
      "grad_norm": 0.3317326522993117,
      "learning_rate": 1.2412343279155926e-06,
      "loss": 0.9714,
      "step": 1144
    },
    {
      "epoch": 0.14823445641971714,
      "grad_norm": 0.22483772556948292,
      "learning_rate": 1.2411871214496655e-06,
      "loss": 1.0619,
      "step": 1145
    },
    {
      "epoch": 0.14823445641971714,
      "eval_PRM Accuracy": 0.6509433962264151,
      "eval_PRM F1": 0.7581699346405228,
      "eval_PRM F1 AUC": 0.5885280251440546,
      "eval_PRM F1 Neg": 0.3728813559322034,
      "eval_PRM NPV": 0.3055555555555556,
      "eval_PRM Precision": 0.8285714285714286,
      "eval_PRM Recall": 0.6987951807228916,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6422230005264282,
      "eval_runtime": 20.2403,
      "eval_samples_per_second": 2.174,
      "eval_steps_per_second": 0.148,
      "step": 1145
    },
    {
      "epoch": 0.1483639188270706,
      "grad_norm": 0.15615368422567819,
      "learning_rate": 1.2411397891150293e-06,
      "loss": 0.9907,
      "step": 1146
    },
    {
      "epoch": 0.14849338123442404,
      "grad_norm": 0.23781313076755925,
      "learning_rate": 1.2410923309213525e-06,
      "loss": 0.9784,
      "step": 1147
    },
    {
      "epoch": 0.14862284364177752,
      "grad_norm": 0.1481688401315399,
      "learning_rate": 1.2410447468783295e-06,
      "loss": 0.9476,
      "step": 1148
    },
    {
      "epoch": 0.14875230604913098,
      "grad_norm": 0.23740933091061903,
      "learning_rate": 1.2409970369956804e-06,
      "loss": 0.9562,
      "step": 1149
    },
    {
      "epoch": 0.14888176845648446,
      "grad_norm": 0.14491110313333028,
      "learning_rate": 1.2409492012831505e-06,
      "loss": 1.01,
      "step": 1150
    },
    {
      "epoch": 0.14888176845648446,
      "eval_PRM Accuracy": 0.6509433962264151,
      "eval_PRM F1": 0.7581699346405228,
      "eval_PRM F1 AUC": 0.5885280251440546,
      "eval_PRM F1 Neg": 0.3728813559322034,
      "eval_PRM NPV": 0.3055555555555556,
      "eval_PRM Precision": 0.8285714285714286,
      "eval_PRM Recall": 0.6987951807228916,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6409801244735718,
      "eval_runtime": 20.4023,
      "eval_samples_per_second": 2.157,
      "eval_steps_per_second": 0.147,
      "step": 1150
    },
    {
      "epoch": 0.1490112308638379,
      "grad_norm": 0.1808433413766103,
      "learning_rate": 1.2409012397505118e-06,
      "loss": 0.8459,
      "step": 1151
    },
    {
      "epoch": 0.14914069327119137,
      "grad_norm": 0.19384349792101402,
      "learning_rate": 1.2408531524075613e-06,
      "loss": 0.9424,
      "step": 1152
    },
    {
      "epoch": 0.14927015567854485,
      "grad_norm": 0.2203720750601757,
      "learning_rate": 1.2408049392641215e-06,
      "loss": 0.9124,
      "step": 1153
    },
    {
      "epoch": 0.1493996180858983,
      "grad_norm": 0.20686120918593423,
      "learning_rate": 1.2407566003300408e-06,
      "loss": 0.9603,
      "step": 1154
    },
    {
      "epoch": 0.14952908049325178,
      "grad_norm": 0.19058116428446534,
      "learning_rate": 1.2407081356151937e-06,
      "loss": 0.9026,
      "step": 1155
    },
    {
      "epoch": 0.14952908049325178,
      "eval_PRM Accuracy": 0.6509433962264151,
      "eval_PRM F1": 0.7581699346405228,
      "eval_PRM F1 AUC": 0.5885280251440546,
      "eval_PRM F1 Neg": 0.3728813559322034,
      "eval_PRM NPV": 0.3055555555555556,
      "eval_PRM Precision": 0.8285714285714286,
      "eval_PRM Recall": 0.6987951807228916,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6415128111839294,
      "eval_runtime": 19.9462,
      "eval_samples_per_second": 2.206,
      "eval_steps_per_second": 0.15,
      "step": 1155
    },
    {
      "epoch": 0.14965854290060523,
      "grad_norm": 0.13641937832363665,
      "learning_rate": 1.2406595451294804e-06,
      "loss": 0.9327,
      "step": 1156
    },
    {
      "epoch": 0.1497880053079587,
      "grad_norm": 0.2184530702182258,
      "learning_rate": 1.240610828882826e-06,
      "loss": 0.9751,
      "step": 1157
    },
    {
      "epoch": 0.14991746771531217,
      "grad_norm": 0.20480547422914702,
      "learning_rate": 1.2405619868851818e-06,
      "loss": 0.9466,
      "step": 1158
    },
    {
      "epoch": 0.15004693012266562,
      "grad_norm": 0.2781929622240853,
      "learning_rate": 1.240513019146525e-06,
      "loss": 1.0227,
      "step": 1159
    },
    {
      "epoch": 0.1501763925300191,
      "grad_norm": 0.13373926321251356,
      "learning_rate": 1.2404639256768582e-06,
      "loss": 0.8794,
      "step": 1160
    },
    {
      "epoch": 0.1501763925300191,
      "eval_PRM Accuracy": 0.6509433962264151,
      "eval_PRM F1": 0.7581699346405228,
      "eval_PRM F1 AUC": 0.5885280251440546,
      "eval_PRM F1 Neg": 0.3728813559322034,
      "eval_PRM NPV": 0.3055555555555556,
      "eval_PRM Precision": 0.8285714285714286,
      "eval_PRM Recall": 0.6987951807228916,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6399147510528564,
      "eval_runtime": 19.5734,
      "eval_samples_per_second": 2.248,
      "eval_steps_per_second": 0.153,
      "step": 1160
    },
    {
      "epoch": 0.15030585493737256,
      "grad_norm": 0.2451526609272693,
      "learning_rate": 1.2404147064862096e-06,
      "loss": 0.913,
      "step": 1161
    },
    {
      "epoch": 0.15043531734472604,
      "grad_norm": 0.2933122240340928,
      "learning_rate": 1.2403653615846334e-06,
      "loss": 0.9858,
      "step": 1162
    },
    {
      "epoch": 0.1505647797520795,
      "grad_norm": 0.2563504990554291,
      "learning_rate": 1.240315890982209e-06,
      "loss": 0.8298,
      "step": 1163
    },
    {
      "epoch": 0.15069424215943295,
      "grad_norm": 0.250788970556738,
      "learning_rate": 1.2402662946890424e-06,
      "loss": 0.9435,
      "step": 1164
    },
    {
      "epoch": 0.15082370456678643,
      "grad_norm": 0.19635150354757944,
      "learning_rate": 1.2402165727152638e-06,
      "loss": 1.0032,
      "step": 1165
    },
    {
      "epoch": 0.15082370456678643,
      "eval_PRM Accuracy": 0.6509433962264151,
      "eval_PRM F1": 0.7581699346405228,
      "eval_PRM F1 AUC": 0.5885280251440546,
      "eval_PRM F1 Neg": 0.3728813559322034,
      "eval_PRM NPV": 0.3055555555555556,
      "eval_PRM Precision": 0.8285714285714286,
      "eval_PRM Recall": 0.6987951807228916,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6400923132896423,
      "eval_runtime": 19.8266,
      "eval_samples_per_second": 2.219,
      "eval_steps_per_second": 0.151,
      "step": 1165
    },
    {
      "epoch": 0.15095316697413988,
      "grad_norm": 0.11955777147895115,
      "learning_rate": 1.2401667250710306e-06,
      "loss": 1.0029,
      "step": 1166
    },
    {
      "epoch": 0.15108262938149336,
      "grad_norm": 0.3022190738226192,
      "learning_rate": 1.2401167517665249e-06,
      "loss": 0.9263,
      "step": 1167
    },
    {
      "epoch": 0.15121209178884681,
      "grad_norm": 0.17707775380319243,
      "learning_rate": 1.240066652811955e-06,
      "loss": 1.0165,
      "step": 1168
    },
    {
      "epoch": 0.15134155419620027,
      "grad_norm": 0.2031647626843248,
      "learning_rate": 1.2400164282175541e-06,
      "loss": 0.9179,
      "step": 1169
    },
    {
      "epoch": 0.15147101660355375,
      "grad_norm": 0.14634980569060102,
      "learning_rate": 1.239966077993582e-06,
      "loss": 0.9636,
      "step": 1170
    },
    {
      "epoch": 0.15147101660355375,
      "eval_PRM Accuracy": 0.6509433962264151,
      "eval_PRM F1": 0.7581699346405228,
      "eval_PRM F1 AUC": 0.5885280251440546,
      "eval_PRM F1 Neg": 0.3728813559322034,
      "eval_PRM NPV": 0.3055555555555556,
      "eval_PRM Precision": 0.8285714285714286,
      "eval_PRM Recall": 0.6987951807228916,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6390269994735718,
      "eval_runtime": 19.8363,
      "eval_samples_per_second": 2.218,
      "eval_steps_per_second": 0.151,
      "step": 1170
    },
    {
      "epoch": 0.1516004790109072,
      "grad_norm": 0.17307852446819663,
      "learning_rate": 1.2399156021503236e-06,
      "loss": 0.8715,
      "step": 1171
    },
    {
      "epoch": 0.15172994141826068,
      "grad_norm": 0.23309803272083393,
      "learning_rate": 1.23986500069809e-06,
      "loss": 1.0286,
      "step": 1172
    },
    {
      "epoch": 0.15185940382561414,
      "grad_norm": 0.13722858990127204,
      "learning_rate": 1.2398142736472168e-06,
      "loss": 0.9646,
      "step": 1173
    },
    {
      "epoch": 0.1519888662329676,
      "grad_norm": 0.2676900726738414,
      "learning_rate": 1.2397634210080669e-06,
      "loss": 0.8473,
      "step": 1174
    },
    {
      "epoch": 0.15211832864032107,
      "grad_norm": 0.21731716123685432,
      "learning_rate": 1.2397124427910275e-06,
      "loss": 0.9083,
      "step": 1175
    },
    {
      "epoch": 0.15211832864032107,
      "eval_PRM Accuracy": 0.6509433962264151,
      "eval_PRM F1": 0.7581699346405228,
      "eval_PRM F1 AUC": 0.5885280251440546,
      "eval_PRM F1 Neg": 0.3728813559322034,
      "eval_PRM NPV": 0.3055555555555556,
      "eval_PRM Precision": 0.8285714285714286,
      "eval_PRM Recall": 0.6987951807228916,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6379616260528564,
      "eval_runtime": 19.9783,
      "eval_samples_per_second": 2.202,
      "eval_steps_per_second": 0.15,
      "step": 1175
    },
    {
      "epoch": 0.15224779104767452,
      "grad_norm": 0.1273277947717622,
      "learning_rate": 1.2396613390065118e-06,
      "loss": 1.0093,
      "step": 1176
    },
    {
      "epoch": 0.152377253455028,
      "grad_norm": 0.1997031014388183,
      "learning_rate": 1.239610109664959e-06,
      "loss": 0.9499,
      "step": 1177
    },
    {
      "epoch": 0.15250671586238146,
      "grad_norm": 0.22222723830345983,
      "learning_rate": 1.2395587547768339e-06,
      "loss": 0.911,
      "step": 1178
    },
    {
      "epoch": 0.1526361782697349,
      "grad_norm": 0.24559222207820744,
      "learning_rate": 1.2395072743526267e-06,
      "loss": 0.8685,
      "step": 1179
    },
    {
      "epoch": 0.1527656406770884,
      "grad_norm": 0.22663242844801126,
      "learning_rate": 1.239455668402853e-06,
      "loss": 0.9938,
      "step": 1180
    },
    {
      "epoch": 0.1527656406770884,
      "eval_PRM Accuracy": 0.6698113207547169,
      "eval_PRM F1": 0.7741935483870968,
      "eval_PRM F1 AUC": 0.6005762179151388,
      "eval_PRM F1 Neg": 0.38596491228070173,
      "eval_PRM NPV": 0.3235294117647059,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.7228915662650602,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6397371888160706,
      "eval_runtime": 20.4754,
      "eval_samples_per_second": 2.149,
      "eval_steps_per_second": 0.147,
      "step": 1180
    },
    {
      "epoch": 0.15289510308444185,
      "grad_norm": 0.2942408480280235,
      "learning_rate": 1.2394039369380547e-06,
      "loss": 0.9442,
      "step": 1181
    },
    {
      "epoch": 0.15302456549179533,
      "grad_norm": 0.29605845253301505,
      "learning_rate": 1.2393520799687989e-06,
      "loss": 0.9392,
      "step": 1182
    },
    {
      "epoch": 0.15315402789914878,
      "grad_norm": 0.16767476786546262,
      "learning_rate": 1.2393000975056782e-06,
      "loss": 0.922,
      "step": 1183
    },
    {
      "epoch": 0.15328349030650226,
      "grad_norm": 0.11820550772244735,
      "learning_rate": 1.2392479895593116e-06,
      "loss": 0.9266,
      "step": 1184
    },
    {
      "epoch": 0.15341295271385572,
      "grad_norm": 0.12426008616773326,
      "learning_rate": 1.2391957561403425e-06,
      "loss": 0.9985,
      "step": 1185
    },
    {
      "epoch": 0.15341295271385572,
      "eval_PRM Accuracy": 0.660377358490566,
      "eval_PRM F1": 0.7662337662337663,
      "eval_PRM F1 AUC": 0.5945521215295966,
      "eval_PRM F1 Neg": 0.3793103448275862,
      "eval_PRM NPV": 0.3142857142857143,
      "eval_PRM Precision": 0.8309859154929577,
      "eval_PRM Recall": 0.7108433734939759,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6342329382896423,
      "eval_runtime": 19.9336,
      "eval_samples_per_second": 2.207,
      "eval_steps_per_second": 0.151,
      "step": 1185
    },
    {
      "epoch": 0.15354241512120917,
      "grad_norm": 0.1794905447532687,
      "learning_rate": 1.2391433972594414e-06,
      "loss": 0.9746,
      "step": 1186
    },
    {
      "epoch": 0.15367187752856265,
      "grad_norm": 0.1959839554262131,
      "learning_rate": 1.239090912927303e-06,
      "loss": 0.8878,
      "step": 1187
    },
    {
      "epoch": 0.1538013399359161,
      "grad_norm": 0.2538662783936271,
      "learning_rate": 1.2390383031546488e-06,
      "loss": 0.8492,
      "step": 1188
    },
    {
      "epoch": 0.15393080234326958,
      "grad_norm": 0.2660093252544258,
      "learning_rate": 1.238985567952225e-06,
      "loss": 0.9013,
      "step": 1189
    },
    {
      "epoch": 0.15406026475062304,
      "grad_norm": 0.16219455730207752,
      "learning_rate": 1.2389327073308043e-06,
      "loss": 1.0226,
      "step": 1190
    },
    {
      "epoch": 0.15406026475062304,
      "eval_PRM Accuracy": 0.660377358490566,
      "eval_PRM F1": 0.7662337662337663,
      "eval_PRM F1 AUC": 0.5945521215295966,
      "eval_PRM F1 Neg": 0.3793103448275862,
      "eval_PRM NPV": 0.3142857142857143,
      "eval_PRM Precision": 0.8309859154929577,
      "eval_PRM Recall": 0.7108433734939759,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6365411877632141,
      "eval_runtime": 19.5675,
      "eval_samples_per_second": 2.249,
      "eval_steps_per_second": 0.153,
      "step": 1190
    },
    {
      "epoch": 0.1541897271579765,
      "grad_norm": 0.20340216840519473,
      "learning_rate": 1.238879721301184e-06,
      "loss": 0.9598,
      "step": 1191
    },
    {
      "epoch": 0.15431918956532997,
      "grad_norm": 0.19454207460368061,
      "learning_rate": 1.238826609874188e-06,
      "loss": 1.1066,
      "step": 1192
    },
    {
      "epoch": 0.15444865197268343,
      "grad_norm": 0.2745083596121325,
      "learning_rate": 1.2387733730606651e-06,
      "loss": 0.9261,
      "step": 1193
    },
    {
      "epoch": 0.1545781143800369,
      "grad_norm": 0.27786226727288904,
      "learning_rate": 1.2387200108714902e-06,
      "loss": 0.961,
      "step": 1194
    },
    {
      "epoch": 0.15470757678739036,
      "grad_norm": 0.14735868879982847,
      "learning_rate": 1.2386665233175638e-06,
      "loss": 0.8885,
      "step": 1195
    },
    {
      "epoch": 0.15470757678739036,
      "eval_PRM Accuracy": 0.660377358490566,
      "eval_PRM F1": 0.7662337662337663,
      "eval_PRM F1 AUC": 0.5945521215295966,
      "eval_PRM F1 Neg": 0.3793103448275862,
      "eval_PRM NPV": 0.3142857142857143,
      "eval_PRM Precision": 0.8309859154929577,
      "eval_PRM Recall": 0.7108433734939759,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6354758739471436,
      "eval_runtime": 20.3082,
      "eval_samples_per_second": 2.167,
      "eval_steps_per_second": 0.148,
      "step": 1195
    },
    {
      "epoch": 0.15483703919474381,
      "grad_norm": 0.19017320492780895,
      "learning_rate": 1.2386129104098114e-06,
      "loss": 0.9171,
      "step": 1196
    },
    {
      "epoch": 0.1549665016020973,
      "grad_norm": 0.15964914528908863,
      "learning_rate": 1.2385591721591844e-06,
      "loss": 0.9103,
      "step": 1197
    },
    {
      "epoch": 0.15509596400945075,
      "grad_norm": 0.28209263351324493,
      "learning_rate": 1.2385053085766605e-06,
      "loss": 0.9143,
      "step": 1198
    },
    {
      "epoch": 0.15522542641680423,
      "grad_norm": 0.2056722999731579,
      "learning_rate": 1.238451319673242e-06,
      "loss": 1.0139,
      "step": 1199
    },
    {
      "epoch": 0.15535488882415768,
      "grad_norm": 0.21143472042470035,
      "learning_rate": 1.2383972054599575e-06,
      "loss": 1.0135,
      "step": 1200
    },
    {
      "epoch": 0.15535488882415768,
      "eval_PRM Accuracy": 0.660377358490566,
      "eval_PRM F1": 0.7662337662337663,
      "eval_PRM F1 AUC": 0.5945521215295966,
      "eval_PRM F1 Neg": 0.3793103448275862,
      "eval_PRM NPV": 0.3142857142857143,
      "eval_PRM Precision": 0.8309859154929577,
      "eval_PRM Recall": 0.7108433734939759,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6363636255264282,
      "eval_runtime": 19.9561,
      "eval_samples_per_second": 2.205,
      "eval_steps_per_second": 0.15,
      "step": 1200
    },
    {
      "epoch": 0.15548435123151114,
      "grad_norm": 0.19804577095483739,
      "learning_rate": 1.2383429659478608e-06,
      "loss": 0.9053,
      "step": 1201
    },
    {
      "epoch": 0.15561381363886462,
      "grad_norm": 0.17907280547346882,
      "learning_rate": 1.2382886011480312e-06,
      "loss": 0.8781,
      "step": 1202
    },
    {
      "epoch": 0.15574327604621807,
      "grad_norm": 0.1247129504257827,
      "learning_rate": 1.238234111071574e-06,
      "loss": 0.9591,
      "step": 1203
    },
    {
      "epoch": 0.15587273845357155,
      "grad_norm": 0.23431298607084353,
      "learning_rate": 1.2381794957296201e-06,
      "loss": 0.8444,
      "step": 1204
    },
    {
      "epoch": 0.156002200860925,
      "grad_norm": 0.13002140053146669,
      "learning_rate": 1.2381247551333255e-06,
      "loss": 0.9767,
      "step": 1205
    },
    {
      "epoch": 0.156002200860925,
      "eval_PRM Accuracy": 0.660377358490566,
      "eval_PRM F1": 0.7662337662337663,
      "eval_PRM F1 AUC": 0.5945521215295966,
      "eval_PRM F1 Neg": 0.3793103448275862,
      "eval_PRM NPV": 0.3142857142857143,
      "eval_PRM Precision": 0.8309859154929577,
      "eval_PRM Recall": 0.7108433734939759,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6363636255264282,
      "eval_runtime": 20.1307,
      "eval_samples_per_second": 2.186,
      "eval_steps_per_second": 0.149,
      "step": 1205
    },
    {
      "epoch": 0.1561316632682785,
      "grad_norm": 0.2604079545381218,
      "learning_rate": 1.2380698892938726e-06,
      "loss": 0.8953,
      "step": 1206
    },
    {
      "epoch": 0.15626112567563194,
      "grad_norm": 0.25151053767878584,
      "learning_rate": 1.238014898222468e-06,
      "loss": 1.0728,
      "step": 1207
    },
    {
      "epoch": 0.1563905880829854,
      "grad_norm": 0.2920941435425304,
      "learning_rate": 1.2379597819303454e-06,
      "loss": 0.9749,
      "step": 1208
    },
    {
      "epoch": 0.15652005049033887,
      "grad_norm": 0.18168517169032627,
      "learning_rate": 1.237904540428763e-06,
      "loss": 0.8827,
      "step": 1209
    },
    {
      "epoch": 0.15664951289769233,
      "grad_norm": 0.38519194590698935,
      "learning_rate": 1.2378491737290053e-06,
      "loss": 1.0259,
      "step": 1210
    },
    {
      "epoch": 0.15664951289769233,
      "eval_PRM Accuracy": 0.660377358490566,
      "eval_PRM F1": 0.7662337662337663,
      "eval_PRM F1 AUC": 0.5945521215295966,
      "eval_PRM F1 Neg": 0.3793103448275862,
      "eval_PRM NPV": 0.3142857142857143,
      "eval_PRM Precision": 0.8309859154929577,
      "eval_PRM Recall": 0.7108433734939759,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6326349377632141,
      "eval_runtime": 20.0041,
      "eval_samples_per_second": 2.2,
      "eval_steps_per_second": 0.15,
      "step": 1210
    },
    {
      "epoch": 0.1567789753050458,
      "grad_norm": 0.18582754592686593,
      "learning_rate": 1.237793681842382e-06,
      "loss": 0.8201,
      "step": 1211
    },
    {
      "epoch": 0.15690843771239926,
      "grad_norm": 0.30010454121097024,
      "learning_rate": 1.2377380647802283e-06,
      "loss": 0.9587,
      "step": 1212
    },
    {
      "epoch": 0.15703790011975272,
      "grad_norm": 0.23066153309944595,
      "learning_rate": 1.2376823225539056e-06,
      "loss": 0.8743,
      "step": 1213
    },
    {
      "epoch": 0.1571673625271062,
      "grad_norm": 0.24338807376320007,
      "learning_rate": 1.2376264551747998e-06,
      "loss": 0.9789,
      "step": 1214
    },
    {
      "epoch": 0.15729682493445965,
      "grad_norm": 0.15526680767498327,
      "learning_rate": 1.237570462654323e-06,
      "loss": 0.9309,
      "step": 1215
    },
    {
      "epoch": 0.15729682493445965,
      "eval_PRM Accuracy": 0.6698113207547169,
      "eval_PRM F1": 0.7741935483870968,
      "eval_PRM F1 AUC": 0.6005762179151388,
      "eval_PRM F1 Neg": 0.38596491228070173,
      "eval_PRM NPV": 0.3235294117647059,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.7228915662650602,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6319246888160706,
      "eval_runtime": 19.4951,
      "eval_samples_per_second": 2.257,
      "eval_steps_per_second": 0.154,
      "step": 1215
    },
    {
      "epoch": 0.15742628734181313,
      "grad_norm": 0.27084235854939354,
      "learning_rate": 1.2375143450039131e-06,
      "loss": 0.9901,
      "step": 1216
    },
    {
      "epoch": 0.15755574974916658,
      "grad_norm": 0.23664484423817284,
      "learning_rate": 1.2374581022350333e-06,
      "loss": 0.8358,
      "step": 1217
    },
    {
      "epoch": 0.15768521215652004,
      "grad_norm": 0.26516154484320137,
      "learning_rate": 1.237401734359172e-06,
      "loss": 1.1129,
      "step": 1218
    },
    {
      "epoch": 0.15781467456387352,
      "grad_norm": 0.12501324153041285,
      "learning_rate": 1.2373452413878438e-06,
      "loss": 0.9502,
      "step": 1219
    },
    {
      "epoch": 0.15794413697122697,
      "grad_norm": 0.19155435006311713,
      "learning_rate": 1.237288623332588e-06,
      "loss": 0.9639,
      "step": 1220
    },
    {
      "epoch": 0.15794413697122697,
      "eval_PRM Accuracy": 0.6698113207547169,
      "eval_PRM F1": 0.7741935483870968,
      "eval_PRM F1 AUC": 0.6005762179151388,
      "eval_PRM F1 Neg": 0.38596491228070173,
      "eval_PRM NPV": 0.3235294117647059,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.7228915662650602,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6324573755264282,
      "eval_runtime": 19.8869,
      "eval_samples_per_second": 2.213,
      "eval_steps_per_second": 0.151,
      "step": 1220
    },
    {
      "epoch": 0.15807359937858045,
      "grad_norm": 0.21628524142821845,
      "learning_rate": 1.2372318802049706e-06,
      "loss": 0.8588,
      "step": 1221
    },
    {
      "epoch": 0.1582030617859339,
      "grad_norm": 0.16386997400612602,
      "learning_rate": 1.2371750120165821e-06,
      "loss": 1.0151,
      "step": 1222
    },
    {
      "epoch": 0.15833252419328736,
      "grad_norm": 0.13375557762051188,
      "learning_rate": 1.2371180187790394e-06,
      "loss": 0.873,
      "step": 1223
    },
    {
      "epoch": 0.15846198660064084,
      "grad_norm": 0.26092773042580597,
      "learning_rate": 1.2370609005039844e-06,
      "loss": 0.8711,
      "step": 1224
    },
    {
      "epoch": 0.1585914490079943,
      "grad_norm": 0.3451634174443754,
      "learning_rate": 1.2370036572030844e-06,
      "loss": 0.8979,
      "step": 1225
    },
    {
      "epoch": 0.1585914490079943,
      "eval_PRM Accuracy": 0.6698113207547169,
      "eval_PRM F1": 0.7741935483870968,
      "eval_PRM F1 AUC": 0.6005762179151388,
      "eval_PRM F1 Neg": 0.38596491228070173,
      "eval_PRM NPV": 0.3235294117647059,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.7228915662650602,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6312144994735718,
      "eval_runtime": 20.6439,
      "eval_samples_per_second": 2.131,
      "eval_steps_per_second": 0.145,
      "step": 1225
    },
    {
      "epoch": 0.15872091141534778,
      "grad_norm": 0.12051125111758045,
      "learning_rate": 1.2369462888880327e-06,
      "loss": 0.8639,
      "step": 1226
    },
    {
      "epoch": 0.15885037382270123,
      "grad_norm": 0.24035370936560166,
      "learning_rate": 1.2368887955705477e-06,
      "loss": 0.9706,
      "step": 1227
    },
    {
      "epoch": 0.1589798362300547,
      "grad_norm": 0.17288088425293177,
      "learning_rate": 1.236831177262374e-06,
      "loss": 0.9919,
      "step": 1228
    },
    {
      "epoch": 0.15910929863740816,
      "grad_norm": 0.2232964280107207,
      "learning_rate": 1.236773433975281e-06,
      "loss": 0.8895,
      "step": 1229
    },
    {
      "epoch": 0.15923876104476162,
      "grad_norm": 0.30420257967616926,
      "learning_rate": 1.236715565721064e-06,
      "loss": 0.9672,
      "step": 1230
    },
    {
      "epoch": 0.15923876104476162,
      "eval_PRM Accuracy": 0.6698113207547169,
      "eval_PRM F1": 0.7741935483870968,
      "eval_PRM F1 AUC": 0.6005762179151388,
      "eval_PRM F1 Neg": 0.38596491228070173,
      "eval_PRM NPV": 0.3235294117647059,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.7228915662650602,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.630859375,
      "eval_runtime": 19.8771,
      "eval_samples_per_second": 2.214,
      "eval_steps_per_second": 0.151,
      "step": 1230
    },
    {
      "epoch": 0.1593682234521151,
      "grad_norm": 0.1736960564063056,
      "learning_rate": 1.236657572511544e-06,
      "loss": 0.8713,
      "step": 1231
    },
    {
      "epoch": 0.15949768585946855,
      "grad_norm": 0.23659996409806827,
      "learning_rate": 1.236599454358567e-06,
      "loss": 0.9294,
      "step": 1232
    },
    {
      "epoch": 0.15962714826682203,
      "grad_norm": 0.19250071785914116,
      "learning_rate": 1.2365412112740048e-06,
      "loss": 0.9783,
      "step": 1233
    },
    {
      "epoch": 0.1597566106741755,
      "grad_norm": 0.13621566085455186,
      "learning_rate": 1.2364828432697549e-06,
      "loss": 0.9122,
      "step": 1234
    },
    {
      "epoch": 0.15988607308152894,
      "grad_norm": 0.1261158171947266,
      "learning_rate": 1.23642435035774e-06,
      "loss": 0.8734,
      "step": 1235
    },
    {
      "epoch": 0.15988607308152894,
      "eval_PRM Accuracy": 0.6792452830188679,
      "eval_PRM F1": 0.782051282051282,
      "eval_PRM F1 AUC": 0.606600314300681,
      "eval_PRM F1 Neg": 0.39285714285714285,
      "eval_PRM NPV": 0.3333333333333333,
      "eval_PRM Precision": 0.8356164383561644,
      "eval_PRM Recall": 0.7349397590361446,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6319246888160706,
      "eval_runtime": 19.9671,
      "eval_samples_per_second": 2.204,
      "eval_steps_per_second": 0.15,
      "step": 1235
    },
    {
      "epoch": 0.16001553548888242,
      "grad_norm": 0.21615931992122617,
      "learning_rate": 1.2363657325499085e-06,
      "loss": 1.0001,
      "step": 1236
    },
    {
      "epoch": 0.16014499789623587,
      "grad_norm": 0.26891611950496735,
      "learning_rate": 1.2363069898582345e-06,
      "loss": 0.8907,
      "step": 1237
    },
    {
      "epoch": 0.16027446030358936,
      "grad_norm": 0.14602725609373735,
      "learning_rate": 1.236248122294717e-06,
      "loss": 0.9142,
      "step": 1238
    },
    {
      "epoch": 0.1604039227109428,
      "grad_norm": 0.17554251030796772,
      "learning_rate": 1.236189129871381e-06,
      "loss": 0.8644,
      "step": 1239
    },
    {
      "epoch": 0.16053338511829626,
      "grad_norm": 0.21078617866351543,
      "learning_rate": 1.2361300126002774e-06,
      "loss": 0.9459,
      "step": 1240
    },
    {
      "epoch": 0.16053338511829626,
      "eval_PRM Accuracy": 0.6886792452830188,
      "eval_PRM F1": 0.7898089171974523,
      "eval_PRM F1 AUC": 0.6126244106862232,
      "eval_PRM F1 Neg": 0.4,
      "eval_PRM NPV": 0.34375,
      "eval_PRM Precision": 0.8378378378378378,
      "eval_PRM Recall": 0.7469879518072289,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6319246888160706,
      "eval_runtime": 20.5137,
      "eval_samples_per_second": 2.145,
      "eval_steps_per_second": 0.146,
      "step": 1240
    },
    {
      "epoch": 0.16066284752564974,
      "grad_norm": 0.21610324540256529,
      "learning_rate": 1.2360707704934814e-06,
      "loss": 0.9442,
      "step": 1241
    },
    {
      "epoch": 0.1607923099330032,
      "grad_norm": 0.14112808275425917,
      "learning_rate": 1.2360114035630947e-06,
      "loss": 0.8616,
      "step": 1242
    },
    {
      "epoch": 0.16092177234035668,
      "grad_norm": 0.24447283677624748,
      "learning_rate": 1.235951911821244e-06,
      "loss": 0.8853,
      "step": 1243
    },
    {
      "epoch": 0.16105123474771013,
      "grad_norm": 0.1817359618143241,
      "learning_rate": 1.2358922952800822e-06,
      "loss": 0.8964,
      "step": 1244
    },
    {
      "epoch": 0.16118069715506358,
      "grad_norm": 0.24505122270315055,
      "learning_rate": 1.2358325539517867e-06,
      "loss": 0.917,
      "step": 1245
    },
    {
      "epoch": 0.16118069715506358,
      "eval_PRM Accuracy": 0.6792452830188679,
      "eval_PRM F1": 0.782051282051282,
      "eval_PRM F1 AUC": 0.606600314300681,
      "eval_PRM F1 Neg": 0.39285714285714285,
      "eval_PRM NPV": 0.3333333333333333,
      "eval_PRM Precision": 0.8356164383561644,
      "eval_PRM Recall": 0.7349397590361446,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6333451867103577,
      "eval_runtime": 20.2184,
      "eval_samples_per_second": 2.176,
      "eval_steps_per_second": 0.148,
      "step": 1245
    },
    {
      "epoch": 0.16131015956241707,
      "grad_norm": 0.20244816089783985,
      "learning_rate": 1.235772687848561e-06,
      "loss": 1.0082,
      "step": 1246
    },
    {
      "epoch": 0.16143962196977052,
      "grad_norm": 0.1947542528115675,
      "learning_rate": 1.2357126969826342e-06,
      "loss": 0.924,
      "step": 1247
    },
    {
      "epoch": 0.161569084377124,
      "grad_norm": 0.20138875782889068,
      "learning_rate": 1.2356525813662604e-06,
      "loss": 0.9067,
      "step": 1248
    },
    {
      "epoch": 0.16169854678447745,
      "grad_norm": 0.15669920024383371,
      "learning_rate": 1.2355923410117193e-06,
      "loss": 0.8992,
      "step": 1249
    },
    {
      "epoch": 0.16182800919183093,
      "grad_norm": 0.13761259399012377,
      "learning_rate": 1.2355319759313167e-06,
      "loss": 0.9167,
      "step": 1250
    },
    {
      "epoch": 0.16182800919183093,
      "eval_PRM Accuracy": 0.6981132075471698,
      "eval_PRM F1": 0.7974683544303798,
      "eval_PRM F1 AUC": 0.6186485070717653,
      "eval_PRM F1 Neg": 0.4074074074074074,
      "eval_PRM NPV": 0.3548387096774194,
      "eval_PRM Precision": 0.84,
      "eval_PRM Recall": 0.7590361445783133,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6280184388160706,
      "eval_runtime": 20.3571,
      "eval_samples_per_second": 2.161,
      "eval_steps_per_second": 0.147,
      "step": 1250
    },
    {
      "epoch": 0.1619574715991844,
      "grad_norm": 0.15986854557292204,
      "learning_rate": 1.2354714861373828e-06,
      "loss": 0.9388,
      "step": 1251
    },
    {
      "epoch": 0.16208693400653784,
      "grad_norm": 0.2959759650823562,
      "learning_rate": 1.2354108716422744e-06,
      "loss": 0.9229,
      "step": 1252
    },
    {
      "epoch": 0.16221639641389132,
      "grad_norm": 0.2727256561360072,
      "learning_rate": 1.235350132458373e-06,
      "loss": 0.9194,
      "step": 1253
    },
    {
      "epoch": 0.16234585882124478,
      "grad_norm": 0.18006938217338905,
      "learning_rate": 1.2352892685980856e-06,
      "loss": 1.0005,
      "step": 1254
    },
    {
      "epoch": 0.16247532122859826,
      "grad_norm": 0.2385885470193566,
      "learning_rate": 1.2352282800738451e-06,
      "loss": 0.8138,
      "step": 1255
    },
    {
      "epoch": 0.16247532122859826,
      "eval_PRM Accuracy": 0.6981132075471698,
      "eval_PRM F1": 0.7974683544303798,
      "eval_PRM F1 AUC": 0.6186485070717653,
      "eval_PRM F1 Neg": 0.4074074074074074,
      "eval_PRM NPV": 0.3548387096774194,
      "eval_PRM Precision": 0.84,
      "eval_PRM Recall": 0.7590361445783133,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6296164989471436,
      "eval_runtime": 20.4506,
      "eval_samples_per_second": 2.152,
      "eval_steps_per_second": 0.147,
      "step": 1255
    },
    {
      "epoch": 0.1626047836359517,
      "grad_norm": 0.1841966203458675,
      "learning_rate": 1.2351671668981098e-06,
      "loss": 0.9935,
      "step": 1256
    },
    {
      "epoch": 0.16273424604330516,
      "grad_norm": 0.11913594394218087,
      "learning_rate": 1.2351059290833631e-06,
      "loss": 0.88,
      "step": 1257
    },
    {
      "epoch": 0.16286370845065865,
      "grad_norm": 0.18350269474910014,
      "learning_rate": 1.235044566642114e-06,
      "loss": 0.9907,
      "step": 1258
    },
    {
      "epoch": 0.1629931708580121,
      "grad_norm": 0.2613839901549486,
      "learning_rate": 1.2349830795868972e-06,
      "loss": 0.9336,
      "step": 1259
    },
    {
      "epoch": 0.16312263326536558,
      "grad_norm": 0.29611014309178985,
      "learning_rate": 1.2349214679302726e-06,
      "loss": 0.847,
      "step": 1260
    },
    {
      "epoch": 0.16312263326536558,
      "eval_PRM Accuracy": 0.6886792452830188,
      "eval_PRM F1": 0.7898089171974523,
      "eval_PRM F1 AUC": 0.6126244106862232,
      "eval_PRM F1 Neg": 0.4,
      "eval_PRM NPV": 0.34375,
      "eval_PRM Precision": 0.8378378378378378,
      "eval_PRM Recall": 0.7469879518072289,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6294389367103577,
      "eval_runtime": 19.9181,
      "eval_samples_per_second": 2.209,
      "eval_steps_per_second": 0.151,
      "step": 1260
    },
    {
      "epoch": 0.16325209567271903,
      "grad_norm": 0.24083903801782308,
      "learning_rate": 1.2348597316848257e-06,
      "loss": 0.8684,
      "step": 1261
    },
    {
      "epoch": 0.1633815580800725,
      "grad_norm": 0.16001141169222954,
      "learning_rate": 1.2347978708631671e-06,
      "loss": 0.9414,
      "step": 1262
    },
    {
      "epoch": 0.16351102048742597,
      "grad_norm": 0.2936582015747638,
      "learning_rate": 1.2347358854779337e-06,
      "loss": 0.8871,
      "step": 1263
    },
    {
      "epoch": 0.16364048289477942,
      "grad_norm": 0.2524752713990941,
      "learning_rate": 1.2346737755417867e-06,
      "loss": 1.0067,
      "step": 1264
    },
    {
      "epoch": 0.1637699453021329,
      "grad_norm": 0.1865377339711969,
      "learning_rate": 1.2346115410674133e-06,
      "loss": 1.0234,
      "step": 1265
    },
    {
      "epoch": 0.1637699453021329,
      "eval_PRM Accuracy": 0.6981132075471698,
      "eval_PRM F1": 0.7974683544303798,
      "eval_PRM F1 AUC": 0.6186485070717653,
      "eval_PRM F1 Neg": 0.4074074074074074,
      "eval_PRM NPV": 0.3548387096774194,
      "eval_PRM Precision": 0.84,
      "eval_PRM Recall": 0.7590361445783133,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6294389367103577,
      "eval_runtime": 19.9298,
      "eval_samples_per_second": 2.208,
      "eval_steps_per_second": 0.151,
      "step": 1265
    },
    {
      "epoch": 0.16389940770948636,
      "grad_norm": 0.14917973705775253,
      "learning_rate": 1.2345491820675268e-06,
      "loss": 0.8728,
      "step": 1266
    },
    {
      "epoch": 0.1640288701168398,
      "grad_norm": 0.26000135200645685,
      "learning_rate": 1.2344866985548646e-06,
      "loss": 0.8322,
      "step": 1267
    },
    {
      "epoch": 0.1641583325241933,
      "grad_norm": 0.25693784419264637,
      "learning_rate": 1.2344240905421906e-06,
      "loss": 0.9456,
      "step": 1268
    },
    {
      "epoch": 0.16428779493154674,
      "grad_norm": 0.23538723924083907,
      "learning_rate": 1.2343613580422934e-06,
      "loss": 0.9384,
      "step": 1269
    },
    {
      "epoch": 0.16441725733890022,
      "grad_norm": 0.2570433288579732,
      "learning_rate": 1.2342985010679879e-06,
      "loss": 0.942,
      "step": 1270
    },
    {
      "epoch": 0.16441725733890022,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8125,
      "eval_PRM F1 AUC": 0.6306966998428496,
      "eval_PRM F1 Neg": 0.4230769230769231,
      "eval_PRM NPV": 0.3793103448275862,
      "eval_PRM Precision": 0.8441558441558441,
      "eval_PRM Recall": 0.7831325301204819,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6280184388160706,
      "eval_runtime": 20.4487,
      "eval_samples_per_second": 2.152,
      "eval_steps_per_second": 0.147,
      "step": 1270
    },
    {
      "epoch": 0.16454671974625368,
      "grad_norm": 0.14604782321647075,
      "learning_rate": 1.2342355196321135e-06,
      "loss": 0.918,
      "step": 1271
    },
    {
      "epoch": 0.16467618215360716,
      "grad_norm": 0.11809774312037648,
      "learning_rate": 1.2341724137475355e-06,
      "loss": 0.955,
      "step": 1272
    },
    {
      "epoch": 0.1648056445609606,
      "grad_norm": 0.11969291827722926,
      "learning_rate": 1.2341091834271449e-06,
      "loss": 0.9617,
      "step": 1273
    },
    {
      "epoch": 0.16493510696831407,
      "grad_norm": 0.24618717378883617,
      "learning_rate": 1.2340458286838573e-06,
      "loss": 0.8859,
      "step": 1274
    },
    {
      "epoch": 0.16506456937566755,
      "grad_norm": 0.2407329102984366,
      "learning_rate": 1.2339823495306145e-06,
      "loss": 1.0333,
      "step": 1275
    },
    {
      "epoch": 0.16506456937566755,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8125,
      "eval_PRM F1 AUC": 0.6306966998428496,
      "eval_PRM F1 Neg": 0.4230769230769231,
      "eval_PRM NPV": 0.3793103448275862,
      "eval_PRM Precision": 0.8441558441558441,
      "eval_PRM Recall": 0.7831325301204819,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6262428760528564,
      "eval_runtime": 20.0502,
      "eval_samples_per_second": 2.194,
      "eval_steps_per_second": 0.15,
      "step": 1275
    },
    {
      "epoch": 0.165194031783021,
      "grad_norm": 0.11603332679201331,
      "learning_rate": 1.2339187459803831e-06,
      "loss": 0.864,
      "step": 1276
    },
    {
      "epoch": 0.16532349419037448,
      "grad_norm": 0.24758857852195787,
      "learning_rate": 1.233855018046156e-06,
      "loss": 0.9578,
      "step": 1277
    },
    {
      "epoch": 0.16545295659772793,
      "grad_norm": 0.23230112358396754,
      "learning_rate": 1.23379116574095e-06,
      "loss": 0.9025,
      "step": 1278
    },
    {
      "epoch": 0.1655824190050814,
      "grad_norm": 0.22640221897072488,
      "learning_rate": 1.2337271890778093e-06,
      "loss": 0.845,
      "step": 1279
    },
    {
      "epoch": 0.16571188141243487,
      "grad_norm": 0.11584304810641602,
      "learning_rate": 1.2336630880698015e-06,
      "loss": 0.9022,
      "step": 1280
    },
    {
      "epoch": 0.16571188141243487,
      "eval_PRM Accuracy": 0.7075471698113207,
      "eval_PRM F1": 0.8050314465408805,
      "eval_PRM F1 AUC": 0.6246726034573076,
      "eval_PRM F1 Neg": 0.41509433962264153,
      "eval_PRM NPV": 0.36666666666666664,
      "eval_PRM Precision": 0.8421052631578947,
      "eval_PRM Recall": 0.7710843373493976,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6296164989471436,
      "eval_runtime": 19.9835,
      "eval_samples_per_second": 2.202,
      "eval_steps_per_second": 0.15,
      "step": 1280
    },
    {
      "epoch": 0.16584134381978832,
      "grad_norm": 0.3603144644545502,
      "learning_rate": 1.233598862730021e-06,
      "loss": 0.9429,
      "step": 1281
    },
    {
      "epoch": 0.1659708062271418,
      "grad_norm": 0.31768458179041537,
      "learning_rate": 1.2335345130715873e-06,
      "loss": 0.8895,
      "step": 1282
    },
    {
      "epoch": 0.16610026863449526,
      "grad_norm": 0.2333596434025405,
      "learning_rate": 1.2334700391076445e-06,
      "loss": 0.8221,
      "step": 1283
    },
    {
      "epoch": 0.1662297310418487,
      "grad_norm": 0.18306296919284373,
      "learning_rate": 1.2334054408513636e-06,
      "loss": 0.9044,
      "step": 1284
    },
    {
      "epoch": 0.1663591934492022,
      "grad_norm": 0.3212548225011112,
      "learning_rate": 1.233340718315939e-06,
      "loss": 0.9354,
      "step": 1285
    },
    {
      "epoch": 0.1663591934492022,
      "eval_PRM Accuracy": 0.7075471698113207,
      "eval_PRM F1": 0.8050314465408805,
      "eval_PRM F1 AUC": 0.6246726034573076,
      "eval_PRM F1 Neg": 0.41509433962264153,
      "eval_PRM NPV": 0.36666666666666664,
      "eval_PRM Precision": 0.8421052631578947,
      "eval_PRM Recall": 0.7710843373493976,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6294389367103577,
      "eval_runtime": 19.7914,
      "eval_samples_per_second": 2.223,
      "eval_steps_per_second": 0.152,
      "step": 1285
    },
    {
      "epoch": 0.16648865585655565,
      "grad_norm": 0.1977385477835639,
      "learning_rate": 1.2332758715145926e-06,
      "loss": 0.9321,
      "step": 1286
    },
    {
      "epoch": 0.16661811826390913,
      "grad_norm": 0.13299268797132124,
      "learning_rate": 1.23321090046057e-06,
      "loss": 0.9873,
      "step": 1287
    },
    {
      "epoch": 0.16674758067126258,
      "grad_norm": 0.20048853851854284,
      "learning_rate": 1.2331458051671432e-06,
      "loss": 1.0092,
      "step": 1288
    },
    {
      "epoch": 0.16687704307861603,
      "grad_norm": 0.2913165064215925,
      "learning_rate": 1.233080585647609e-06,
      "loss": 0.8706,
      "step": 1289
    },
    {
      "epoch": 0.16700650548596951,
      "grad_norm": 0.19304217591633832,
      "learning_rate": 1.2330152419152903e-06,
      "loss": 0.8675,
      "step": 1290
    },
    {
      "epoch": 0.16700650548596951,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8125,
      "eval_PRM F1 AUC": 0.6306966998428496,
      "eval_PRM F1 Neg": 0.4230769230769231,
      "eval_PRM NPV": 0.3793103448275862,
      "eval_PRM Precision": 0.8441558441558441,
      "eval_PRM Recall": 0.7831325301204819,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6292613744735718,
      "eval_runtime": 19.8504,
      "eval_samples_per_second": 2.217,
      "eval_steps_per_second": 0.151,
      "step": 1290
    },
    {
      "epoch": 0.16713596789332297,
      "grad_norm": 0.19561350955766074,
      "learning_rate": 1.2329497739835343e-06,
      "loss": 0.9534,
      "step": 1291
    },
    {
      "epoch": 0.16726543030067645,
      "grad_norm": 0.2136814290109197,
      "learning_rate": 1.2328841818657145e-06,
      "loss": 0.9811,
      "step": 1292
    },
    {
      "epoch": 0.1673948927080299,
      "grad_norm": 0.25787854445347946,
      "learning_rate": 1.2328184655752292e-06,
      "loss": 0.8452,
      "step": 1293
    },
    {
      "epoch": 0.16752435511538338,
      "grad_norm": 0.23401828296547045,
      "learning_rate": 1.2327526251255024e-06,
      "loss": 0.7946,
      "step": 1294
    },
    {
      "epoch": 0.16765381752273684,
      "grad_norm": 0.3212837434359875,
      "learning_rate": 1.2326866605299834e-06,
      "loss": 0.856,
      "step": 1295
    },
    {
      "epoch": 0.16765381752273684,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8125,
      "eval_PRM F1 AUC": 0.6306966998428496,
      "eval_PRM F1 Neg": 0.4230769230769231,
      "eval_PRM NPV": 0.3793103448275862,
      "eval_PRM Precision": 0.8441558441558441,
      "eval_PRM Recall": 0.7831325301204819,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6283735632896423,
      "eval_runtime": 19.2858,
      "eval_samples_per_second": 2.281,
      "eval_steps_per_second": 0.156,
      "step": 1295
    },
    {
      "epoch": 0.1677832799300903,
      "grad_norm": 0.21903713853477602,
      "learning_rate": 1.2326205718021466e-06,
      "loss": 0.8628,
      "step": 1296
    },
    {
      "epoch": 0.16791274233744377,
      "grad_norm": 0.24033135839307965,
      "learning_rate": 1.232554358955492e-06,
      "loss": 0.9024,
      "step": 1297
    },
    {
      "epoch": 0.16804220474479722,
      "grad_norm": 0.17623833780000686,
      "learning_rate": 1.2324880220035455e-06,
      "loss": 0.8972,
      "step": 1298
    },
    {
      "epoch": 0.1681716671521507,
      "grad_norm": 0.20117022500592038,
      "learning_rate": 1.2324215609598568e-06,
      "loss": 0.8969,
      "step": 1299
    },
    {
      "epoch": 0.16830112955950416,
      "grad_norm": 0.1802770194766478,
      "learning_rate": 1.2323549758380026e-06,
      "loss": 0.965,
      "step": 1300
    },
    {
      "epoch": 0.16830112955950416,
      "eval_PRM Accuracy": 0.7075471698113207,
      "eval_PRM F1": 0.8050314465408805,
      "eval_PRM F1 AUC": 0.6246726034573076,
      "eval_PRM F1 Neg": 0.41509433962264153,
      "eval_PRM NPV": 0.36666666666666664,
      "eval_PRM Precision": 0.8421052631578947,
      "eval_PRM Recall": 0.7710843373493976,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6274858117103577,
      "eval_runtime": 20.0348,
      "eval_samples_per_second": 2.196,
      "eval_steps_per_second": 0.15,
      "step": 1300
    },
    {
      "epoch": 0.1684305919668576,
      "grad_norm": 0.1748102089113948,
      "learning_rate": 1.232288266651584e-06,
      "loss": 0.8378,
      "step": 1301
    },
    {
      "epoch": 0.1685600543742111,
      "grad_norm": 0.18687683813959297,
      "learning_rate": 1.2322214334142279e-06,
      "loss": 0.8708,
      "step": 1302
    },
    {
      "epoch": 0.16868951678156455,
      "grad_norm": 0.17650978844639528,
      "learning_rate": 1.232154476139586e-06,
      "loss": 0.884,
      "step": 1303
    },
    {
      "epoch": 0.16881897918891803,
      "grad_norm": 0.2719780029987721,
      "learning_rate": 1.232087394841336e-06,
      "loss": 0.8885,
      "step": 1304
    },
    {
      "epoch": 0.16894844159627148,
      "grad_norm": 0.3430817141851236,
      "learning_rate": 1.2320201895331807e-06,
      "loss": 0.8575,
      "step": 1305
    },
    {
      "epoch": 0.16894844159627148,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8220858895705522,
      "eval_PRM F1 AUC": 0.6210057621791514,
      "eval_PRM F1 Neg": 0.40816326530612246,
      "eval_PRM NPV": 0.38461538461538464,
      "eval_PRM Precision": 0.8375,
      "eval_PRM Recall": 0.8072289156626506,
      "eval_PRM Specificty": 0.43478260869565216,
      "eval_loss": 0.6278409361839294,
      "eval_runtime": 20.2737,
      "eval_samples_per_second": 2.17,
      "eval_steps_per_second": 0.148,
      "step": 1305
    },
    {
      "epoch": 0.16907790400362493,
      "grad_norm": 0.1537638017303247,
      "learning_rate": 1.2319528602288477e-06,
      "loss": 0.955,
      "step": 1306
    },
    {
      "epoch": 0.16920736641097842,
      "grad_norm": 0.14347379733570925,
      "learning_rate": 1.2318854069420906e-06,
      "loss": 0.968,
      "step": 1307
    },
    {
      "epoch": 0.16933682881833187,
      "grad_norm": 0.16909183412527362,
      "learning_rate": 1.2318178296866881e-06,
      "loss": 0.8789,
      "step": 1308
    },
    {
      "epoch": 0.16946629122568535,
      "grad_norm": 0.22257753201105548,
      "learning_rate": 1.2317501284764443e-06,
      "loss": 0.8467,
      "step": 1309
    },
    {
      "epoch": 0.1695957536330388,
      "grad_norm": 0.1358940304463402,
      "learning_rate": 1.2316823033251887e-06,
      "loss": 0.9153,
      "step": 1310
    },
    {
      "epoch": 0.1695957536330388,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8198757763975155,
      "eval_PRM F1 AUC": 0.6367207962283918,
      "eval_PRM F1 Neg": 0.43137254901960786,
      "eval_PRM NPV": 0.39285714285714285,
      "eval_PRM Precision": 0.8461538461538461,
      "eval_PRM Recall": 0.7951807228915663,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6267755627632141,
      "eval_runtime": 20.3324,
      "eval_samples_per_second": 2.164,
      "eval_steps_per_second": 0.148,
      "step": 1310
    },
    {
      "epoch": 0.16972521604039228,
      "grad_norm": 0.23884469553572954,
      "learning_rate": 1.2316143542467756e-06,
      "loss": 0.8992,
      "step": 1311
    },
    {
      "epoch": 0.16985467844774574,
      "grad_norm": 0.20930617675003602,
      "learning_rate": 1.2315462812550853e-06,
      "loss": 0.797,
      "step": 1312
    },
    {
      "epoch": 0.1699841408550992,
      "grad_norm": 0.1103155132650564,
      "learning_rate": 1.2314780843640225e-06,
      "loss": 0.8331,
      "step": 1313
    },
    {
      "epoch": 0.17011360326245267,
      "grad_norm": 0.24965518991487626,
      "learning_rate": 1.2314097635875185e-06,
      "loss": 0.8396,
      "step": 1314
    },
    {
      "epoch": 0.17024306566980613,
      "grad_norm": 0.14616943182916875,
      "learning_rate": 1.2313413189395289e-06,
      "loss": 0.8856,
      "step": 1315
    },
    {
      "epoch": 0.17024306566980613,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8220858895705522,
      "eval_PRM F1 AUC": 0.6210057621791514,
      "eval_PRM F1 Neg": 0.40816326530612246,
      "eval_PRM NPV": 0.38461538461538464,
      "eval_PRM Precision": 0.8375,
      "eval_PRM Recall": 0.8072289156626506,
      "eval_PRM Specificty": 0.43478260869565216,
      "eval_loss": 0.62890625,
      "eval_runtime": 20.5082,
      "eval_samples_per_second": 2.145,
      "eval_steps_per_second": 0.146,
      "step": 1315
    },
    {
      "epoch": 0.1703725280771596,
      "grad_norm": 0.33842331696038447,
      "learning_rate": 1.231272750434035e-06,
      "loss": 0.9113,
      "step": 1316
    },
    {
      "epoch": 0.17050199048451306,
      "grad_norm": 0.19421326723908103,
      "learning_rate": 1.2312040580850432e-06,
      "loss": 0.9247,
      "step": 1317
    },
    {
      "epoch": 0.17063145289186651,
      "grad_norm": 0.21329271316868628,
      "learning_rate": 1.231135241906585e-06,
      "loss": 0.8918,
      "step": 1318
    },
    {
      "epoch": 0.17076091529922,
      "grad_norm": 0.14037823769347196,
      "learning_rate": 1.2310663019127184e-06,
      "loss": 1.1125,
      "step": 1319
    },
    {
      "epoch": 0.17089037770657345,
      "grad_norm": 0.11126798903393025,
      "learning_rate": 1.2309972381175249e-06,
      "loss": 0.9297,
      "step": 1320
    },
    {
      "epoch": 0.17089037770657345,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8292682926829268,
      "eval_PRM F1 AUC": 0.6270298585646936,
      "eval_PRM F1 Neg": 0.4166666666666667,
      "eval_PRM NPV": 0.4,
      "eval_PRM Precision": 0.8395061728395061,
      "eval_PRM Recall": 0.8192771084337349,
      "eval_PRM Specificty": 0.43478260869565216,
      "eval_loss": 0.6283735632896423,
      "eval_runtime": 19.5919,
      "eval_samples_per_second": 2.246,
      "eval_steps_per_second": 0.153,
      "step": 1320
    },
    {
      "epoch": 0.17101984011392693,
      "grad_norm": 0.21766922503075403,
      "learning_rate": 1.2309280505351127e-06,
      "loss": 0.9481,
      "step": 1321
    },
    {
      "epoch": 0.17114930252128038,
      "grad_norm": 0.1955301365625585,
      "learning_rate": 1.2308587391796143e-06,
      "loss": 0.9252,
      "step": 1322
    },
    {
      "epoch": 0.17127876492863384,
      "grad_norm": 0.21317075320585913,
      "learning_rate": 1.2307893040651884e-06,
      "loss": 0.9069,
      "step": 1323
    },
    {
      "epoch": 0.17140822733598732,
      "grad_norm": 0.3114349581957321,
      "learning_rate": 1.2307197452060182e-06,
      "loss": 0.8153,
      "step": 1324
    },
    {
      "epoch": 0.17153768974334077,
      "grad_norm": 0.2518772971429363,
      "learning_rate": 1.2306500626163129e-06,
      "loss": 0.9695,
      "step": 1325
    },
    {
      "epoch": 0.17153768974334077,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8292682926829268,
      "eval_PRM F1 AUC": 0.6270298585646936,
      "eval_PRM F1 Neg": 0.4166666666666667,
      "eval_PRM NPV": 0.4,
      "eval_PRM Precision": 0.8395061728395061,
      "eval_PRM Recall": 0.8192771084337349,
      "eval_PRM Specificty": 0.43478260869565216,
      "eval_loss": 0.6273082494735718,
      "eval_runtime": 19.8111,
      "eval_samples_per_second": 2.221,
      "eval_steps_per_second": 0.151,
      "step": 1325
    },
    {
      "epoch": 0.17166715215069425,
      "grad_norm": 0.28952595732916964,
      "learning_rate": 1.2305802563103062e-06,
      "loss": 0.9065,
      "step": 1326
    },
    {
      "epoch": 0.1717966145580477,
      "grad_norm": 0.20185663508033172,
      "learning_rate": 1.2305103263022577e-06,
      "loss": 0.8511,
      "step": 1327
    },
    {
      "epoch": 0.17192607696540116,
      "grad_norm": 0.28961312347737694,
      "learning_rate": 1.2304402726064517e-06,
      "loss": 0.8763,
      "step": 1328
    },
    {
      "epoch": 0.17205553937275464,
      "grad_norm": 0.24595767425933843,
      "learning_rate": 1.2303700952371986e-06,
      "loss": 0.9164,
      "step": 1329
    },
    {
      "epoch": 0.1721850017801081,
      "grad_norm": 0.2304767635990097,
      "learning_rate": 1.230299794208833e-06,
      "loss": 0.9429,
      "step": 1330
    },
    {
      "epoch": 0.1721850017801081,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8343558282208589,
      "eval_PRM F1 AUC": 0.6487689889994761,
      "eval_PRM F1 Neg": 0.4489795918367347,
      "eval_PRM NPV": 0.4230769230769231,
      "eval_PRM Precision": 0.85,
      "eval_PRM Recall": 0.8192771084337349,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.6244673132896423,
      "eval_runtime": 20.2399,
      "eval_samples_per_second": 2.174,
      "eval_steps_per_second": 0.148,
      "step": 1330
    },
    {
      "epoch": 0.17231446418746157,
      "grad_norm": 0.2831566260750998,
      "learning_rate": 1.2302293695357157e-06,
      "loss": 0.8506,
      "step": 1331
    },
    {
      "epoch": 0.17244392659481503,
      "grad_norm": 0.21306678553202021,
      "learning_rate": 1.2301588212322322e-06,
      "loss": 0.9773,
      "step": 1332
    },
    {
      "epoch": 0.1725733890021685,
      "grad_norm": 0.14301760749229445,
      "learning_rate": 1.2300881493127936e-06,
      "loss": 0.9398,
      "step": 1333
    },
    {
      "epoch": 0.17270285140952196,
      "grad_norm": 0.24630018896134054,
      "learning_rate": 1.2300173537918357e-06,
      "loss": 0.8602,
      "step": 1334
    },
    {
      "epoch": 0.17283231381687542,
      "grad_norm": 0.17710371007407089,
      "learning_rate": 1.2299464346838202e-06,
      "loss": 0.86,
      "step": 1335
    },
    {
      "epoch": 0.17283231381687542,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8292682926829268,
      "eval_PRM F1 AUC": 0.6270298585646936,
      "eval_PRM F1 Neg": 0.4166666666666667,
      "eval_PRM NPV": 0.4,
      "eval_PRM Precision": 0.8395061728395061,
      "eval_PRM Recall": 0.8192771084337349,
      "eval_PRM Specificty": 0.43478260869565216,
      "eval_loss": 0.6274858117103577,
      "eval_runtime": 20.075,
      "eval_samples_per_second": 2.192,
      "eval_steps_per_second": 0.149,
      "step": 1335
    },
    {
      "epoch": 0.1729617762242289,
      "grad_norm": 0.26907967783589065,
      "learning_rate": 1.2298753920032337e-06,
      "loss": 0.9644,
      "step": 1336
    },
    {
      "epoch": 0.17309123863158235,
      "grad_norm": 0.1329208419706331,
      "learning_rate": 1.229804225764588e-06,
      "loss": 0.9388,
      "step": 1337
    },
    {
      "epoch": 0.17322070103893583,
      "grad_norm": 0.1223327123939425,
      "learning_rate": 1.2297329359824206e-06,
      "loss": 0.8574,
      "step": 1338
    },
    {
      "epoch": 0.17335016344628928,
      "grad_norm": 0.18669845120707748,
      "learning_rate": 1.2296615226712938e-06,
      "loss": 0.9446,
      "step": 1339
    },
    {
      "epoch": 0.17347962585364274,
      "grad_norm": 0.27668555114406607,
      "learning_rate": 1.2295899858457946e-06,
      "loss": 0.963,
      "step": 1340
    },
    {
      "epoch": 0.17347962585364274,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8292682926829268,
      "eval_PRM F1 AUC": 0.6270298585646936,
      "eval_PRM F1 Neg": 0.4166666666666667,
      "eval_PRM NPV": 0.4,
      "eval_PRM Precision": 0.8395061728395061,
      "eval_PRM Recall": 0.8192771084337349,
      "eval_PRM Specificty": 0.43478260869565216,
      "eval_loss": 0.6241121888160706,
      "eval_runtime": 20.6168,
      "eval_samples_per_second": 2.134,
      "eval_steps_per_second": 0.146,
      "step": 1340
    },
    {
      "epoch": 0.17360908826099622,
      "grad_norm": 0.19119105716462811,
      "learning_rate": 1.2295183255205368e-06,
      "loss": 1.0338,
      "step": 1341
    },
    {
      "epoch": 0.17373855066834967,
      "grad_norm": 0.1714669764253873,
      "learning_rate": 1.2294465417101575e-06,
      "loss": 1.0151,
      "step": 1342
    },
    {
      "epoch": 0.17386801307570315,
      "grad_norm": 0.1594575258064153,
      "learning_rate": 1.229374634429321e-06,
      "loss": 0.8588,
      "step": 1343
    },
    {
      "epoch": 0.1739974754830566,
      "grad_norm": 0.11872216371265853,
      "learning_rate": 1.2293026036927146e-06,
      "loss": 0.7751,
      "step": 1344
    },
    {
      "epoch": 0.17412693789041006,
      "grad_norm": 0.1745939995182037,
      "learning_rate": 1.2292304495150535e-06,
      "loss": 0.9651,
      "step": 1345
    },
    {
      "epoch": 0.17412693789041006,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8292682926829268,
      "eval_PRM F1 AUC": 0.6270298585646936,
      "eval_PRM F1 Neg": 0.4166666666666667,
      "eval_PRM NPV": 0.4,
      "eval_PRM Precision": 0.8395061728395061,
      "eval_PRM Recall": 0.8192771084337349,
      "eval_PRM Specificty": 0.43478260869565216,
      "eval_loss": 0.6226917505264282,
      "eval_runtime": 20.2811,
      "eval_samples_per_second": 2.17,
      "eval_steps_per_second": 0.148,
      "step": 1345
    },
    {
      "epoch": 0.17425640029776354,
      "grad_norm": 0.14059727576662984,
      "learning_rate": 1.2291581719110754e-06,
      "loss": 0.8544,
      "step": 1346
    },
    {
      "epoch": 0.174385862705117,
      "grad_norm": 0.19651817505641844,
      "learning_rate": 1.2290857708955451e-06,
      "loss": 0.9423,
      "step": 1347
    },
    {
      "epoch": 0.17451532511247048,
      "grad_norm": 0.20634780922408394,
      "learning_rate": 1.229013246483252e-06,
      "loss": 0.8114,
      "step": 1348
    },
    {
      "epoch": 0.17464478751982393,
      "grad_norm": 0.1272213940238623,
      "learning_rate": 1.2289405986890102e-06,
      "loss": 0.9944,
      "step": 1349
    },
    {
      "epoch": 0.17477424992717738,
      "grad_norm": 0.2556150941501497,
      "learning_rate": 1.2288678275276599e-06,
      "loss": 0.933,
      "step": 1350
    },
    {
      "epoch": 0.17477424992717738,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8414634146341463,
      "eval_PRM F1 AUC": 0.6547930853850183,
      "eval_PRM F1 Neg": 0.4583333333333333,
      "eval_PRM NPV": 0.44,
      "eval_PRM Precision": 0.8518518518518519,
      "eval_PRM Recall": 0.8313253012048193,
      "eval_PRM Specificty": 0.4782608695652174,
      "eval_loss": 0.626953125,
      "eval_runtime": 20.7418,
      "eval_samples_per_second": 2.121,
      "eval_steps_per_second": 0.145,
      "step": 1350
    },
    {
      "epoch": 0.17490371233453086,
      "grad_norm": 0.19123205247189823,
      "learning_rate": 1.2287949330140662e-06,
      "loss": 0.9246,
      "step": 1351
    },
    {
      "epoch": 0.17503317474188432,
      "grad_norm": 0.2415831457427456,
      "learning_rate": 1.2287219151631186e-06,
      "loss": 0.9358,
      "step": 1352
    },
    {
      "epoch": 0.1751626371492378,
      "grad_norm": 0.23310488962215797,
      "learning_rate": 1.2286487739897334e-06,
      "loss": 0.9291,
      "step": 1353
    },
    {
      "epoch": 0.17529209955659125,
      "grad_norm": 0.2658958411745581,
      "learning_rate": 1.2285755095088506e-06,
      "loss": 0.839,
      "step": 1354
    },
    {
      "epoch": 0.17542156196394473,
      "grad_norm": 0.12865727499975024,
      "learning_rate": 1.2285021217354359e-06,
      "loss": 0.9794,
      "step": 1355
    },
    {
      "epoch": 0.17542156196394473,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8292682926829268,
      "eval_PRM F1 AUC": 0.6270298585646936,
      "eval_PRM F1 Neg": 0.4166666666666667,
      "eval_PRM NPV": 0.4,
      "eval_PRM Precision": 0.8395061728395061,
      "eval_PRM Recall": 0.8192771084337349,
      "eval_PRM Specificty": 0.43478260869565216,
      "eval_loss": 0.6264204382896423,
      "eval_runtime": 21.081,
      "eval_samples_per_second": 2.087,
      "eval_steps_per_second": 0.142,
      "step": 1355
    },
    {
      "epoch": 0.1755510243712982,
      "grad_norm": 0.1396093370858754,
      "learning_rate": 1.2284286106844805e-06,
      "loss": 0.8756,
      "step": 1356
    },
    {
      "epoch": 0.17568048677865164,
      "grad_norm": 0.15958839704320435,
      "learning_rate": 1.2283549763710005e-06,
      "loss": 0.8892,
      "step": 1357
    },
    {
      "epoch": 0.17580994918600512,
      "grad_norm": 0.12825537411255897,
      "learning_rate": 1.228281218810037e-06,
      "loss": 1.0026,
      "step": 1358
    },
    {
      "epoch": 0.17593941159335857,
      "grad_norm": 0.1998078206011593,
      "learning_rate": 1.2282073380166568e-06,
      "loss": 0.9194,
      "step": 1359
    },
    {
      "epoch": 0.17606887400071206,
      "grad_norm": 0.11589067714645511,
      "learning_rate": 1.2281333340059512e-06,
      "loss": 0.8916,
      "step": 1360
    },
    {
      "epoch": 0.17606887400071206,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8292682926829268,
      "eval_PRM F1 AUC": 0.6270298585646936,
      "eval_PRM F1 Neg": 0.4166666666666667,
      "eval_PRM NPV": 0.4,
      "eval_PRM Precision": 0.8395061728395061,
      "eval_PRM Recall": 0.8192771084337349,
      "eval_PRM Specificty": 0.43478260869565216,
      "eval_loss": 0.6264204382896423,
      "eval_runtime": 20.1038,
      "eval_samples_per_second": 2.189,
      "eval_steps_per_second": 0.149,
      "step": 1360
    },
    {
      "epoch": 0.1761983364080655,
      "grad_norm": 0.15452396096658158,
      "learning_rate": 1.2280592067930371e-06,
      "loss": 0.8699,
      "step": 1361
    },
    {
      "epoch": 0.17632779881541896,
      "grad_norm": 0.18628272987554184,
      "learning_rate": 1.2279849563930567e-06,
      "loss": 0.9246,
      "step": 1362
    },
    {
      "epoch": 0.17645726122277244,
      "grad_norm": 0.2072888300251209,
      "learning_rate": 1.227910582821177e-06,
      "loss": 0.8239,
      "step": 1363
    },
    {
      "epoch": 0.1765867236301259,
      "grad_norm": 0.16583698617715,
      "learning_rate": 1.2278360860925904e-06,
      "loss": 0.8555,
      "step": 1364
    },
    {
      "epoch": 0.17671618603747938,
      "grad_norm": 0.20640219789228176,
      "learning_rate": 1.2277614662225142e-06,
      "loss": 0.8544,
      "step": 1365
    },
    {
      "epoch": 0.17671618603747938,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8313253012048193,
      "eval_PRM F1 AUC": 0.6113148245154532,
      "eval_PRM F1 Neg": 0.391304347826087,
      "eval_PRM NPV": 0.391304347826087,
      "eval_PRM Precision": 0.8313253012048193,
      "eval_PRM Recall": 0.8313253012048193,
      "eval_PRM Specificty": 0.391304347826087,
      "eval_loss": 0.6239346861839294,
      "eval_runtime": 19.7941,
      "eval_samples_per_second": 2.223,
      "eval_steps_per_second": 0.152,
      "step": 1365
    },
    {
      "epoch": 0.17684564844483283,
      "grad_norm": 0.3099382694530306,
      "learning_rate": 1.227686723226191e-06,
      "loss": 0.8989,
      "step": 1366
    },
    {
      "epoch": 0.17697511085218628,
      "grad_norm": 0.21739507379481082,
      "learning_rate": 1.227611857118889e-06,
      "loss": 0.917,
      "step": 1367
    },
    {
      "epoch": 0.17710457325953977,
      "grad_norm": 0.13277343763542246,
      "learning_rate": 1.2275368679159002e-06,
      "loss": 0.8339,
      "step": 1368
    },
    {
      "epoch": 0.17723403566689322,
      "grad_norm": 0.14436188087042984,
      "learning_rate": 1.2274617556325438e-06,
      "loss": 0.9221,
      "step": 1369
    },
    {
      "epoch": 0.1773634980742467,
      "grad_norm": 0.17342385231138055,
      "learning_rate": 1.2273865202841623e-06,
      "loss": 0.8508,
      "step": 1370
    },
    {
      "epoch": 0.1773634980742467,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8363636363636363,
      "eval_PRM F1 AUC": 0.6330539549502358,
      "eval_PRM F1 Neg": 0.425531914893617,
      "eval_PRM NPV": 0.4166666666666667,
      "eval_PRM Precision": 0.8414634146341463,
      "eval_PRM Recall": 0.8313253012048193,
      "eval_PRM Specificty": 0.43478260869565216,
      "eval_loss": 0.6248224377632141,
      "eval_runtime": 20.2115,
      "eval_samples_per_second": 2.177,
      "eval_steps_per_second": 0.148,
      "step": 1370
    },
    {
      "epoch": 0.17749296048160015,
      "grad_norm": 0.20983822368087554,
      "learning_rate": 1.2273111618861242e-06,
      "loss": 0.9054,
      "step": 1371
    },
    {
      "epoch": 0.1776224228889536,
      "grad_norm": 0.15305817993104956,
      "learning_rate": 1.227235680453823e-06,
      "loss": 0.8993,
      "step": 1372
    },
    {
      "epoch": 0.1777518852963071,
      "grad_norm": 0.1176283225474841,
      "learning_rate": 1.2271600760026771e-06,
      "loss": 0.8926,
      "step": 1373
    },
    {
      "epoch": 0.17788134770366054,
      "grad_norm": 0.16436567478688732,
      "learning_rate": 1.2270843485481307e-06,
      "loss": 0.8142,
      "step": 1374
    },
    {
      "epoch": 0.17801081011101402,
      "grad_norm": 0.18662879618078787,
      "learning_rate": 1.2270084981056523e-06,
      "loss": 0.8795,
      "step": 1375
    },
    {
      "epoch": 0.17801081011101402,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8313253012048193,
      "eval_PRM F1 AUC": 0.6113148245154532,
      "eval_PRM F1 Neg": 0.391304347826087,
      "eval_PRM NPV": 0.391304347826087,
      "eval_PRM Precision": 0.8313253012048193,
      "eval_PRM Recall": 0.8313253012048193,
      "eval_PRM Specificty": 0.391304347826087,
      "eval_loss": 0.6262428760528564,
      "eval_runtime": 20.2021,
      "eval_samples_per_second": 2.178,
      "eval_steps_per_second": 0.148,
      "step": 1375
    },
    {
      "epoch": 0.17814027251836748,
      "grad_norm": 0.21902026561678822,
      "learning_rate": 1.2269325246907362e-06,
      "loss": 0.8528,
      "step": 1376
    },
    {
      "epoch": 0.17826973492572096,
      "grad_norm": 0.1375361542529811,
      "learning_rate": 1.226856428318901e-06,
      "loss": 1.0315,
      "step": 1377
    },
    {
      "epoch": 0.1783991973330744,
      "grad_norm": 0.16113478438336934,
      "learning_rate": 1.2267802090056916e-06,
      "loss": 1.0012,
      "step": 1378
    },
    {
      "epoch": 0.17852865974042786,
      "grad_norm": 0.1283762155241809,
      "learning_rate": 1.226703866766677e-06,
      "loss": 0.9171,
      "step": 1379
    },
    {
      "epoch": 0.17865812214778135,
      "grad_norm": 0.29455444132600606,
      "learning_rate": 1.2266274016174512e-06,
      "loss": 0.9323,
      "step": 1380
    },
    {
      "epoch": 0.17865812214778135,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8313253012048193,
      "eval_PRM F1 AUC": 0.6113148245154532,
      "eval_PRM F1 Neg": 0.391304347826087,
      "eval_PRM NPV": 0.391304347826087,
      "eval_PRM Precision": 0.8313253012048193,
      "eval_PRM Recall": 0.8313253012048193,
      "eval_PRM Specificty": 0.391304347826087,
      "eval_loss": 0.6246448755264282,
      "eval_runtime": 19.8297,
      "eval_samples_per_second": 2.219,
      "eval_steps_per_second": 0.151,
      "step": 1380
    },
    {
      "epoch": 0.1787875845551348,
      "grad_norm": 0.26600050576698653,
      "learning_rate": 1.2265508135736344e-06,
      "loss": 0.8776,
      "step": 1381
    },
    {
      "epoch": 0.17891704696248828,
      "grad_norm": 0.20439249131303666,
      "learning_rate": 1.2264741026508712e-06,
      "loss": 0.9095,
      "step": 1382
    },
    {
      "epoch": 0.17904650936984173,
      "grad_norm": 0.17474260251502893,
      "learning_rate": 1.226397268864831e-06,
      "loss": 0.9735,
      "step": 1383
    },
    {
      "epoch": 0.1791759717771952,
      "grad_norm": 0.16699083330272751,
      "learning_rate": 1.2263203122312089e-06,
      "loss": 0.8267,
      "step": 1384
    },
    {
      "epoch": 0.17930543418454867,
      "grad_norm": 0.11004138546509382,
      "learning_rate": 1.2262432327657251e-06,
      "loss": 0.9319,
      "step": 1385
    },
    {
      "epoch": 0.17930543418454867,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8313253012048193,
      "eval_PRM F1 AUC": 0.6113148245154532,
      "eval_PRM F1 Neg": 0.391304347826087,
      "eval_PRM NPV": 0.391304347826087,
      "eval_PRM Precision": 0.8313253012048193,
      "eval_PRM Recall": 0.8313253012048193,
      "eval_PRM Specificty": 0.391304347826087,
      "eval_loss": 0.6248224377632141,
      "eval_runtime": 19.5641,
      "eval_samples_per_second": 2.249,
      "eval_steps_per_second": 0.153,
      "step": 1385
    },
    {
      "epoch": 0.17943489659190212,
      "grad_norm": 0.25103410328265674,
      "learning_rate": 1.226166030484124e-06,
      "loss": 0.8706,
      "step": 1386
    },
    {
      "epoch": 0.1795643589992556,
      "grad_norm": 0.13751612175125358,
      "learning_rate": 1.2260887054021763e-06,
      "loss": 0.8905,
      "step": 1387
    },
    {
      "epoch": 0.17969382140660906,
      "grad_norm": 0.3170658515488868,
      "learning_rate": 1.226011257535677e-06,
      "loss": 0.9515,
      "step": 1388
    },
    {
      "epoch": 0.1798232838139625,
      "grad_norm": 0.17735346900451016,
      "learning_rate": 1.2259336869004462e-06,
      "loss": 0.8306,
      "step": 1389
    },
    {
      "epoch": 0.179952746221316,
      "grad_norm": 0.15650545159981655,
      "learning_rate": 1.2258559935123297e-06,
      "loss": 0.9452,
      "step": 1390
    },
    {
      "epoch": 0.179952746221316,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8383233532934131,
      "eval_PRM F1 AUC": 0.6173389209009953,
      "eval_PRM F1 Neg": 0.4,
      "eval_PRM NPV": 0.4090909090909091,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.8433734939759037,
      "eval_PRM Specificty": 0.391304347826087,
      "eval_loss": 0.6253551244735718,
      "eval_runtime": 20.5248,
      "eval_samples_per_second": 2.144,
      "eval_steps_per_second": 0.146,
      "step": 1390
    },
    {
      "epoch": 0.18008220862866944,
      "grad_norm": 0.17681393634264728,
      "learning_rate": 1.2257781773871977e-06,
      "loss": 0.923,
      "step": 1391
    },
    {
      "epoch": 0.18021167103602292,
      "grad_norm": 0.18953610391089057,
      "learning_rate": 1.2257002385409455e-06,
      "loss": 0.9026,
      "step": 1392
    },
    {
      "epoch": 0.18034113344337638,
      "grad_norm": 0.13717937727012017,
      "learning_rate": 1.2256221769894943e-06,
      "loss": 0.9285,
      "step": 1393
    },
    {
      "epoch": 0.18047059585072983,
      "grad_norm": 0.19107095375081048,
      "learning_rate": 1.225543992748789e-06,
      "loss": 0.8381,
      "step": 1394
    },
    {
      "epoch": 0.1806000582580833,
      "grad_norm": 0.25088910982763546,
      "learning_rate": 1.2254656858348013e-06,
      "loss": 0.847,
      "step": 1395
    },
    {
      "epoch": 0.1806000582580833,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8383233532934131,
      "eval_PRM F1 AUC": 0.6173389209009953,
      "eval_PRM F1 Neg": 0.4,
      "eval_PRM NPV": 0.4090909090909091,
      "eval_PRM Precision": 0.8333333333333334,
      "eval_PRM Recall": 0.8433734939759037,
      "eval_PRM Specificty": 0.391304347826087,
      "eval_loss": 0.6237571239471436,
      "eval_runtime": 20.4155,
      "eval_samples_per_second": 2.155,
      "eval_steps_per_second": 0.147,
      "step": 1395
    },
    {
      "epoch": 0.18072952066543677,
      "grad_norm": 0.20997587815147936,
      "learning_rate": 1.2253872562635258e-06,
      "loss": 0.8,
      "step": 1396
    },
    {
      "epoch": 0.18085898307279025,
      "grad_norm": 0.17600446715112597,
      "learning_rate": 1.2253087040509841e-06,
      "loss": 0.859,
      "step": 1397
    },
    {
      "epoch": 0.1809884454801437,
      "grad_norm": 0.18794158436870328,
      "learning_rate": 1.2252300292132222e-06,
      "loss": 0.8316,
      "step": 1398
    },
    {
      "epoch": 0.18111790788749718,
      "grad_norm": 0.1624620095126199,
      "learning_rate": 1.2251512317663102e-06,
      "loss": 0.9418,
      "step": 1399
    },
    {
      "epoch": 0.18124737029485063,
      "grad_norm": 0.13121583266354186,
      "learning_rate": 1.225072311726345e-06,
      "loss": 0.9675,
      "step": 1400
    },
    {
      "epoch": 0.18124737029485063,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8452380952380952,
      "eval_PRM F1 AUC": 0.6233630172865374,
      "eval_PRM F1 Neg": 0.4090909090909091,
      "eval_PRM NPV": 0.42857142857142855,
      "eval_PRM Precision": 0.8352941176470589,
      "eval_PRM Recall": 0.8554216867469879,
      "eval_PRM Specificty": 0.391304347826087,
      "eval_loss": 0.6262428760528564,
      "eval_runtime": 19.9408,
      "eval_samples_per_second": 2.207,
      "eval_steps_per_second": 0.15,
      "step": 1400
    },
    {
      "epoch": 0.1813768327022041,
      "grad_norm": 0.1639415665831153,
      "learning_rate": 1.2249932691094472e-06,
      "loss": 0.7927,
      "step": 1401
    },
    {
      "epoch": 0.18150629510955757,
      "grad_norm": 0.2878109644814213,
      "learning_rate": 1.2249141039317626e-06,
      "loss": 0.778,
      "step": 1402
    },
    {
      "epoch": 0.18163575751691102,
      "grad_norm": 0.11335652929466011,
      "learning_rate": 1.2248348162094628e-06,
      "loss": 0.9305,
      "step": 1403
    },
    {
      "epoch": 0.1817652199242645,
      "grad_norm": 0.19108861228613674,
      "learning_rate": 1.2247554059587435e-06,
      "loss": 0.916,
      "step": 1404
    },
    {
      "epoch": 0.18189468233161796,
      "grad_norm": 0.2532671818581398,
      "learning_rate": 1.2246758731958261e-06,
      "loss": 0.8518,
      "step": 1405
    },
    {
      "epoch": 0.18189468233161796,
      "eval_PRM Accuracy": 0.7641509433962265,
      "eval_PRM F1": 0.8520710059171598,
      "eval_PRM F1 AUC": 0.6293871136720797,
      "eval_PRM F1 Neg": 0.4186046511627907,
      "eval_PRM NPV": 0.45,
      "eval_PRM Precision": 0.8372093023255814,
      "eval_PRM Recall": 0.8674698795180723,
      "eval_PRM Specificty": 0.391304347826087,
      "eval_loss": 0.6244673132896423,
      "eval_runtime": 20.4399,
      "eval_samples_per_second": 2.153,
      "eval_steps_per_second": 0.147,
      "step": 1405
    },
    {
      "epoch": 0.1820241447389714,
      "grad_norm": 0.14413458059186252,
      "learning_rate": 1.224596217936957e-06,
      "loss": 0.8899,
      "step": 1406
    },
    {
      "epoch": 0.1821536071463249,
      "grad_norm": 0.14057455775674554,
      "learning_rate": 1.2245164401984067e-06,
      "loss": 0.824,
      "step": 1407
    },
    {
      "epoch": 0.18228306955367835,
      "grad_norm": 0.11235669979948194,
      "learning_rate": 1.2244365399964722e-06,
      "loss": 0.8901,
      "step": 1408
    },
    {
      "epoch": 0.18241253196103183,
      "grad_norm": 0.23112135121323385,
      "learning_rate": 1.224356517347474e-06,
      "loss": 0.8341,
      "step": 1409
    },
    {
      "epoch": 0.18254199436838528,
      "grad_norm": 0.20328138173664925,
      "learning_rate": 1.224276372267759e-06,
      "loss": 0.8237,
      "step": 1410
    },
    {
      "epoch": 0.18254199436838528,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8452380952380952,
      "eval_PRM F1 AUC": 0.6233630172865374,
      "eval_PRM F1 Neg": 0.4090909090909091,
      "eval_PRM NPV": 0.42857142857142855,
      "eval_PRM Precision": 0.8352941176470589,
      "eval_PRM Recall": 0.8554216867469879,
      "eval_PRM Specificty": 0.391304347826087,
      "eval_loss": 0.6237571239471436,
      "eval_runtime": 20.0536,
      "eval_samples_per_second": 2.194,
      "eval_steps_per_second": 0.15,
      "step": 1410
    },
    {
      "epoch": 0.18267145677573873,
      "grad_norm": 0.2637643281575498,
      "learning_rate": 1.2241961047736982e-06,
      "loss": 0.8433,
      "step": 1411
    },
    {
      "epoch": 0.18280091918309221,
      "grad_norm": 0.138009215775976,
      "learning_rate": 1.224115714881688e-06,
      "loss": 0.8849,
      "step": 1412
    },
    {
      "epoch": 0.18293038159044567,
      "grad_norm": 0.11810139900387447,
      "learning_rate": 1.2240352026081492e-06,
      "loss": 0.947,
      "step": 1413
    },
    {
      "epoch": 0.18305984399779915,
      "grad_norm": 0.27621869951612826,
      "learning_rate": 1.2239545679695284e-06,
      "loss": 0.9064,
      "step": 1414
    },
    {
      "epoch": 0.1831893064051526,
      "grad_norm": 0.1527385021867629,
      "learning_rate": 1.2238738109822968e-06,
      "loss": 0.8821,
      "step": 1415
    },
    {
      "epoch": 0.1831893064051526,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8470588235294118,
      "eval_PRM F1 AUC": 0.607647983237297,
      "eval_PRM F1 Neg": 0.38095238095238093,
      "eval_PRM NPV": 0.42105263157894735,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.8674698795180723,
      "eval_PRM Specificty": 0.34782608695652173,
      "eval_loss": 0.6265980005264282,
      "eval_runtime": 20.5776,
      "eval_samples_per_second": 2.138,
      "eval_steps_per_second": 0.146,
      "step": 1415
    },
    {
      "epoch": 0.18331876881250606,
      "grad_norm": 0.28192759382399984,
      "learning_rate": 1.223792931662951e-06,
      "loss": 0.9434,
      "step": 1416
    },
    {
      "epoch": 0.18344823121985954,
      "grad_norm": 0.1630413069563371,
      "learning_rate": 1.2237119300280117e-06,
      "loss": 0.8872,
      "step": 1417
    },
    {
      "epoch": 0.183577693627213,
      "grad_norm": 0.2228245746930564,
      "learning_rate": 1.2236308060940253e-06,
      "loss": 0.8068,
      "step": 1418
    },
    {
      "epoch": 0.18370715603456647,
      "grad_norm": 0.18971524666738618,
      "learning_rate": 1.223549559877563e-06,
      "loss": 0.8711,
      "step": 1419
    },
    {
      "epoch": 0.18383661844191992,
      "grad_norm": 0.22745728067083246,
      "learning_rate": 1.2234681913952213e-06,
      "loss": 0.7966,
      "step": 1420
    },
    {
      "epoch": 0.18383661844191992,
      "eval_PRM Accuracy": 0.7641509433962265,
      "eval_PRM F1": 0.8520710059171598,
      "eval_PRM F1 AUC": 0.6293871136720797,
      "eval_PRM F1 Neg": 0.4186046511627907,
      "eval_PRM NPV": 0.45,
      "eval_PRM Precision": 0.8372093023255814,
      "eval_PRM Recall": 0.8674698795180723,
      "eval_PRM Specificty": 0.391304347826087,
      "eval_loss": 0.6234019994735718,
      "eval_runtime": 20.4414,
      "eval_samples_per_second": 2.152,
      "eval_steps_per_second": 0.147,
      "step": 1420
    },
    {
      "epoch": 0.1839660808492734,
      "grad_norm": 0.22135151394564379,
      "learning_rate": 1.223386700663621e-06,
      "loss": 0.9252,
      "step": 1421
    },
    {
      "epoch": 0.18409554325662686,
      "grad_norm": 0.12845039096791888,
      "learning_rate": 1.2233050876994086e-06,
      "loss": 0.8319,
      "step": 1422
    },
    {
      "epoch": 0.1842250056639803,
      "grad_norm": 0.301307066036521,
      "learning_rate": 1.2232233525192545e-06,
      "loss": 0.788,
      "step": 1423
    },
    {
      "epoch": 0.1843544680713338,
      "grad_norm": 0.1621072252886301,
      "learning_rate": 1.2231414951398554e-06,
      "loss": 0.9,
      "step": 1424
    },
    {
      "epoch": 0.18448393047868725,
      "grad_norm": 0.1430598926574888,
      "learning_rate": 1.223059515577932e-06,
      "loss": 0.936,
      "step": 1425
    },
    {
      "epoch": 0.18448393047868725,
      "eval_PRM Accuracy": 0.7641509433962265,
      "eval_PRM F1": 0.8520710059171598,
      "eval_PRM F1 AUC": 0.6293871136720797,
      "eval_PRM F1 Neg": 0.4186046511627907,
      "eval_PRM NPV": 0.45,
      "eval_PRM Precision": 0.8372093023255814,
      "eval_PRM Recall": 0.8674698795180723,
      "eval_PRM Specificty": 0.391304347826087,
      "eval_loss": 0.6251775622367859,
      "eval_runtime": 19.55,
      "eval_samples_per_second": 2.251,
      "eval_steps_per_second": 0.153,
      "step": 1425
    },
    {
      "epoch": 0.18461339288604073,
      "grad_norm": 0.1842130784885946,
      "learning_rate": 1.2229774138502306e-06,
      "loss": 0.8618,
      "step": 1426
    },
    {
      "epoch": 0.18474285529339418,
      "grad_norm": 0.18031255904637608,
      "learning_rate": 1.222895189973522e-06,
      "loss": 0.9557,
      "step": 1427
    },
    {
      "epoch": 0.18487231770074763,
      "grad_norm": 0.20528644898378182,
      "learning_rate": 1.222812843964602e-06,
      "loss": 0.8534,
      "step": 1428
    },
    {
      "epoch": 0.18500178010810112,
      "grad_norm": 0.1478836146832837,
      "learning_rate": 1.2227303758402918e-06,
      "loss": 0.9144,
      "step": 1429
    },
    {
      "epoch": 0.18513124251545457,
      "grad_norm": 0.14214097339920825,
      "learning_rate": 1.2226477856174366e-06,
      "loss": 0.9352,
      "step": 1430
    },
    {
      "epoch": 0.18513124251545457,
      "eval_PRM Accuracy": 0.7641509433962265,
      "eval_PRM F1": 0.8520710059171598,
      "eval_PRM F1 AUC": 0.6293871136720797,
      "eval_PRM F1 Neg": 0.4186046511627907,
      "eval_PRM NPV": 0.45,
      "eval_PRM Precision": 0.8372093023255814,
      "eval_PRM Recall": 0.8674698795180723,
      "eval_PRM Specificty": 0.391304347826087,
      "eval_loss": 0.6242897510528564,
      "eval_runtime": 20.2049,
      "eval_samples_per_second": 2.178,
      "eval_steps_per_second": 0.148,
      "step": 1430
    },
    {
      "epoch": 0.18526070492280805,
      "grad_norm": 0.21898510286421038,
      "learning_rate": 1.2225650733129073e-06,
      "loss": 0.91,
      "step": 1431
    },
    {
      "epoch": 0.1853901673301615,
      "grad_norm": 0.30282827304784027,
      "learning_rate": 1.2224822389435997e-06,
      "loss": 0.7859,
      "step": 1432
    },
    {
      "epoch": 0.18551962973751496,
      "grad_norm": 0.2090557509523202,
      "learning_rate": 1.2223992825264347e-06,
      "loss": 0.8865,
      "step": 1433
    },
    {
      "epoch": 0.18564909214486844,
      "grad_norm": 0.15232493029770572,
      "learning_rate": 1.2223162040783572e-06,
      "loss": 0.8535,
      "step": 1434
    },
    {
      "epoch": 0.1857785545522219,
      "grad_norm": 0.12365460118713602,
      "learning_rate": 1.2222330036163381e-06,
      "loss": 0.9528,
      "step": 1435
    },
    {
      "epoch": 0.1857785545522219,
      "eval_PRM Accuracy": 0.7641509433962265,
      "eval_PRM F1": 0.8520710059171598,
      "eval_PRM F1 AUC": 0.6293871136720797,
      "eval_PRM F1 Neg": 0.4186046511627907,
      "eval_PRM NPV": 0.45,
      "eval_PRM Precision": 0.8372093023255814,
      "eval_PRM Recall": 0.8674698795180723,
      "eval_PRM Specificty": 0.391304347826087,
      "eval_loss": 0.6242897510528564,
      "eval_runtime": 20.3438,
      "eval_samples_per_second": 2.163,
      "eval_steps_per_second": 0.147,
      "step": 1435
    },
    {
      "epoch": 0.18590801695957537,
      "grad_norm": 0.20605387631023767,
      "learning_rate": 1.2221496811573726e-06,
      "loss": 0.8196,
      "step": 1436
    },
    {
      "epoch": 0.18603747936692883,
      "grad_norm": 0.15211691551952178,
      "learning_rate": 1.222066236718481e-06,
      "loss": 0.8788,
      "step": 1437
    },
    {
      "epoch": 0.18616694177428228,
      "grad_norm": 0.15900773448002722,
      "learning_rate": 1.2219826703167087e-06,
      "loss": 1.0223,
      "step": 1438
    },
    {
      "epoch": 0.18629640418163576,
      "grad_norm": 0.1845250726875403,
      "learning_rate": 1.2218989819691257e-06,
      "loss": 0.8444,
      "step": 1439
    },
    {
      "epoch": 0.18642586658898921,
      "grad_norm": 0.22557130773107648,
      "learning_rate": 1.221815171692827e-06,
      "loss": 0.978,
      "step": 1440
    },
    {
      "epoch": 0.18642586658898921,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8470588235294118,
      "eval_PRM F1 AUC": 0.607647983237297,
      "eval_PRM F1 Neg": 0.38095238095238093,
      "eval_PRM NPV": 0.42105263157894735,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.8674698795180723,
      "eval_PRM Specificty": 0.34782608695652173,
      "eval_loss": 0.6253551244735718,
      "eval_runtime": 19.9236,
      "eval_samples_per_second": 2.208,
      "eval_steps_per_second": 0.151,
      "step": 1440
    },
    {
      "epoch": 0.1865553289963427,
      "grad_norm": 0.20705631504527208,
      "learning_rate": 1.2217312395049324e-06,
      "loss": 0.8677,
      "step": 1441
    },
    {
      "epoch": 0.18668479140369615,
      "grad_norm": 0.15931006681734142,
      "learning_rate": 1.2216471854225873e-06,
      "loss": 0.9479,
      "step": 1442
    },
    {
      "epoch": 0.18681425381104963,
      "grad_norm": 0.16773671867371256,
      "learning_rate": 1.221563009462961e-06,
      "loss": 0.922,
      "step": 1443
    },
    {
      "epoch": 0.18694371621840308,
      "grad_norm": 0.12062727950962003,
      "learning_rate": 1.221478711643248e-06,
      "loss": 0.8813,
      "step": 1444
    },
    {
      "epoch": 0.18707317862575654,
      "grad_norm": 0.10980127115615827,
      "learning_rate": 1.221394291980668e-06,
      "loss": 0.8181,
      "step": 1445
    },
    {
      "epoch": 0.18707317862575654,
      "eval_PRM Accuracy": 0.7641509433962265,
      "eval_PRM F1": 0.8520710059171598,
      "eval_PRM F1 AUC": 0.6293871136720797,
      "eval_PRM F1 Neg": 0.4186046511627907,
      "eval_PRM NPV": 0.45,
      "eval_PRM Precision": 0.8372093023255814,
      "eval_PRM Recall": 0.8674698795180723,
      "eval_PRM Specificty": 0.391304347826087,
      "eval_loss": 0.6241121888160706,
      "eval_runtime": 20.3369,
      "eval_samples_per_second": 2.164,
      "eval_steps_per_second": 0.148,
      "step": 1445
    },
    {
      "epoch": 0.18720264103311002,
      "grad_norm": 0.21433236863414232,
      "learning_rate": 1.2213097504924659e-06,
      "loss": 0.8917,
      "step": 1446
    },
    {
      "epoch": 0.18733210344046347,
      "grad_norm": 0.13125910173560504,
      "learning_rate": 1.2212250871959106e-06,
      "loss": 0.8364,
      "step": 1447
    },
    {
      "epoch": 0.18746156584781695,
      "grad_norm": 0.15638372319442034,
      "learning_rate": 1.221140302108296e-06,
      "loss": 0.9613,
      "step": 1448
    },
    {
      "epoch": 0.1875910282551704,
      "grad_norm": 0.1280395998009669,
      "learning_rate": 1.2210553952469418e-06,
      "loss": 0.8844,
      "step": 1449
    },
    {
      "epoch": 0.18772049066252386,
      "grad_norm": 0.23914435644437665,
      "learning_rate": 1.2209703666291912e-06,
      "loss": 0.918,
      "step": 1450
    },
    {
      "epoch": 0.18772049066252386,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8470588235294118,
      "eval_PRM F1 AUC": 0.607647983237297,
      "eval_PRM F1 Neg": 0.38095238095238093,
      "eval_PRM NPV": 0.42105263157894735,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.8674698795180723,
      "eval_PRM Specificty": 0.34782608695652173,
      "eval_loss": 0.623046875,
      "eval_runtime": 19.8209,
      "eval_samples_per_second": 2.22,
      "eval_steps_per_second": 0.151,
      "step": 1450
    },
    {
      "epoch": 0.18784995306987734,
      "grad_norm": 0.23585186223120794,
      "learning_rate": 1.2208852162724138e-06,
      "loss": 0.8474,
      "step": 1451
    },
    {
      "epoch": 0.1879794154772308,
      "grad_norm": 0.17979559146684748,
      "learning_rate": 1.2207999441940028e-06,
      "loss": 0.9856,
      "step": 1452
    },
    {
      "epoch": 0.18810887788458427,
      "grad_norm": 0.1946278622453796,
      "learning_rate": 1.2207145504113769e-06,
      "loss": 0.9272,
      "step": 1453
    },
    {
      "epoch": 0.18823834029193773,
      "grad_norm": 0.14300810652043555,
      "learning_rate": 1.2206290349419795e-06,
      "loss": 0.8679,
      "step": 1454
    },
    {
      "epoch": 0.18836780269929118,
      "grad_norm": 0.23810946200519792,
      "learning_rate": 1.220543397803279e-06,
      "loss": 0.8606,
      "step": 1455
    },
    {
      "epoch": 0.18836780269929118,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8470588235294118,
      "eval_PRM F1 AUC": 0.607647983237297,
      "eval_PRM F1 Neg": 0.38095238095238093,
      "eval_PRM NPV": 0.42105263157894735,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.8674698795180723,
      "eval_PRM Specificty": 0.34782608695652173,
      "eval_loss": 0.6241121888160706,
      "eval_runtime": 19.4449,
      "eval_samples_per_second": 2.263,
      "eval_steps_per_second": 0.154,
      "step": 1455
    },
    {
      "epoch": 0.18849726510664466,
      "grad_norm": 0.19140065113585064,
      "learning_rate": 1.2204576390127681e-06,
      "loss": 0.8585,
      "step": 1456
    },
    {
      "epoch": 0.18862672751399812,
      "grad_norm": 0.16935532975588985,
      "learning_rate": 1.2203717585879652e-06,
      "loss": 0.9002,
      "step": 1457
    },
    {
      "epoch": 0.1887561899213516,
      "grad_norm": 0.158358257461577,
      "learning_rate": 1.220285756546413e-06,
      "loss": 0.8762,
      "step": 1458
    },
    {
      "epoch": 0.18888565232870505,
      "grad_norm": 0.16684657566372846,
      "learning_rate": 1.2201996329056793e-06,
      "loss": 1.0939,
      "step": 1459
    },
    {
      "epoch": 0.1890151147360585,
      "grad_norm": 0.16287433587580946,
      "learning_rate": 1.220113387683356e-06,
      "loss": 0.8739,
      "step": 1460
    },
    {
      "epoch": 0.1890151147360585,
      "eval_PRM Accuracy": 0.7641509433962265,
      "eval_PRM F1": 0.8520710059171598,
      "eval_PRM F1 AUC": 0.6293871136720797,
      "eval_PRM F1 Neg": 0.4186046511627907,
      "eval_PRM NPV": 0.45,
      "eval_PRM Precision": 0.8372093023255814,
      "eval_PRM Recall": 0.8674698795180723,
      "eval_PRM Specificty": 0.391304347826087,
      "eval_loss": 0.6212713122367859,
      "eval_runtime": 19.9583,
      "eval_samples_per_second": 2.205,
      "eval_steps_per_second": 0.15,
      "step": 1460
    },
    {
      "epoch": 0.18914457714341198,
      "grad_norm": 0.16280666596585452,
      "learning_rate": 1.2200270208970614e-06,
      "loss": 0.8569,
      "step": 1461
    },
    {
      "epoch": 0.18927403955076544,
      "grad_norm": 0.15869428901656932,
      "learning_rate": 1.2199405325644367e-06,
      "loss": 1.0715,
      "step": 1462
    },
    {
      "epoch": 0.18940350195811892,
      "grad_norm": 0.2452315745912076,
      "learning_rate": 1.2198539227031495e-06,
      "loss": 0.8842,
      "step": 1463
    },
    {
      "epoch": 0.18953296436547237,
      "grad_norm": 0.2127892218064515,
      "learning_rate": 1.2197671913308914e-06,
      "loss": 0.8519,
      "step": 1464
    },
    {
      "epoch": 0.18966242677282585,
      "grad_norm": 0.17121697486330492,
      "learning_rate": 1.2196803384653792e-06,
      "loss": 0.8716,
      "step": 1465
    },
    {
      "epoch": 0.18966242677282585,
      "eval_PRM Accuracy": 0.7641509433962265,
      "eval_PRM F1": 0.8520710059171598,
      "eval_PRM F1 AUC": 0.6293871136720797,
      "eval_PRM F1 Neg": 0.4186046511627907,
      "eval_PRM NPV": 0.45,
      "eval_PRM Precision": 0.8372093023255814,
      "eval_PRM Recall": 0.8674698795180723,
      "eval_PRM Specificty": 0.391304347826087,
      "eval_loss": 0.6244673132896423,
      "eval_runtime": 19.8212,
      "eval_samples_per_second": 2.22,
      "eval_steps_per_second": 0.151,
      "step": 1465
    },
    {
      "epoch": 0.1897918891801793,
      "grad_norm": 0.21947514416626934,
      "learning_rate": 1.219593364124354e-06,
      "loss": 0.8762,
      "step": 1466
    },
    {
      "epoch": 0.18992135158753276,
      "grad_norm": 0.2556303306578983,
      "learning_rate": 1.2195062683255825e-06,
      "loss": 0.7802,
      "step": 1467
    },
    {
      "epoch": 0.19005081399488624,
      "grad_norm": 0.2375757026756834,
      "learning_rate": 1.2194190510868555e-06,
      "loss": 0.8336,
      "step": 1468
    },
    {
      "epoch": 0.1901802764022397,
      "grad_norm": 0.18654497955407304,
      "learning_rate": 1.2193317124259894e-06,
      "loss": 0.8843,
      "step": 1469
    },
    {
      "epoch": 0.19030973880959318,
      "grad_norm": 0.1658593816492131,
      "learning_rate": 1.2192442523608239e-06,
      "loss": 1.0386,
      "step": 1470
    },
    {
      "epoch": 0.19030973880959318,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8470588235294118,
      "eval_PRM F1 AUC": 0.607647983237297,
      "eval_PRM F1 Neg": 0.38095238095238093,
      "eval_PRM NPV": 0.42105263157894735,
      "eval_PRM Precision": 0.8275862068965517,
      "eval_PRM Recall": 0.8674698795180723,
      "eval_PRM Specificty": 0.34782608695652173,
      "eval_loss": 0.6253551244735718,
      "eval_runtime": 20.0205,
      "eval_samples_per_second": 2.198,
      "eval_steps_per_second": 0.15,
      "step": 1470
    },
    {
      "epoch": 0.19043920121694663,
      "grad_norm": 0.214020331675023,
      "learning_rate": 1.2191566709092253e-06,
      "loss": 0.9362,
      "step": 1471
    },
    {
      "epoch": 0.19056866362430008,
      "grad_norm": 0.1430086762235561,
      "learning_rate": 1.2190689680890837e-06,
      "loss": 0.9207,
      "step": 1472
    },
    {
      "epoch": 0.19069812603165356,
      "grad_norm": 0.15451326712315716,
      "learning_rate": 1.2189811439183141e-06,
      "loss": 0.9193,
      "step": 1473
    },
    {
      "epoch": 0.19082758843900702,
      "grad_norm": 0.11966370352631486,
      "learning_rate": 1.2188931984148566e-06,
      "loss": 0.8242,
      "step": 1474
    },
    {
      "epoch": 0.1909570508463605,
      "grad_norm": 0.11524564767519589,
      "learning_rate": 1.2188051315966754e-06,
      "loss": 0.8721,
      "step": 1475
    },
    {
      "epoch": 0.1909570508463605,
      "eval_PRM Accuracy": 0.7641509433962265,
      "eval_PRM F1": 0.8538011695906432,
      "eval_PRM F1 AUC": 0.6136720796228392,
      "eval_PRM F1 Neg": 0.3902439024390244,
      "eval_PRM NPV": 0.4444444444444444,
      "eval_PRM Precision": 0.8295454545454546,
      "eval_PRM Recall": 0.8795180722891566,
      "eval_PRM Specificty": 0.34782608695652173,
      "eval_loss": 0.6246448755264282,
      "eval_runtime": 20.4005,
      "eval_samples_per_second": 2.157,
      "eval_steps_per_second": 0.147,
      "step": 1475
    },
    {
      "epoch": 0.19108651325371395,
      "grad_norm": 0.28563001176685554,
      "learning_rate": 1.2187169434817604e-06,
      "loss": 0.8685,
      "step": 1476
    },
    {
      "epoch": 0.1912159756610674,
      "grad_norm": 0.18862750637007747,
      "learning_rate": 1.2186286340881253e-06,
      "loss": 0.8716,
      "step": 1477
    },
    {
      "epoch": 0.1913454380684209,
      "grad_norm": 0.1793997933082881,
      "learning_rate": 1.2185402034338095e-06,
      "loss": 0.8966,
      "step": 1478
    },
    {
      "epoch": 0.19147490047577434,
      "grad_norm": 0.20200612281314445,
      "learning_rate": 1.218451651536877e-06,
      "loss": 0.8461,
      "step": 1479
    },
    {
      "epoch": 0.19160436288312782,
      "grad_norm": 0.20964874622311702,
      "learning_rate": 1.2183629784154155e-06,
      "loss": 0.8953,
      "step": 1480
    },
    {
      "epoch": 0.19160436288312782,
      "eval_PRM Accuracy": 0.7641509433962265,
      "eval_PRM F1": 0.8538011695906432,
      "eval_PRM F1 AUC": 0.6136720796228392,
      "eval_PRM F1 Neg": 0.3902439024390244,
      "eval_PRM NPV": 0.4444444444444444,
      "eval_PRM Precision": 0.8295454545454546,
      "eval_PRM Recall": 0.8795180722891566,
      "eval_PRM Specificty": 0.34782608695652173,
      "eval_loss": 0.6248224377632141,
      "eval_runtime": 19.9685,
      "eval_samples_per_second": 2.203,
      "eval_steps_per_second": 0.15,
      "step": 1480
    },
    {
      "epoch": 0.19173382529048127,
      "grad_norm": 0.1988129822498125,
      "learning_rate": 1.218274184087539e-06,
      "loss": 0.8177,
      "step": 1481
    },
    {
      "epoch": 0.19186328769783473,
      "grad_norm": 0.21679723675552984,
      "learning_rate": 1.218185268571385e-06,
      "loss": 0.8594,
      "step": 1482
    },
    {
      "epoch": 0.1919927501051882,
      "grad_norm": 0.17867288295399067,
      "learning_rate": 1.2180962318851168e-06,
      "loss": 0.7913,
      "step": 1483
    },
    {
      "epoch": 0.19212221251254166,
      "grad_norm": 0.17553514880151247,
      "learning_rate": 1.2180070740469217e-06,
      "loss": 0.8741,
      "step": 1484
    },
    {
      "epoch": 0.19225167491989514,
      "grad_norm": 0.1465332317531764,
      "learning_rate": 1.2179177950750118e-06,
      "loss": 0.855,
      "step": 1485
    },
    {
      "epoch": 0.19225167491989514,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8488372093023255,
      "eval_PRM F1 AUC": 0.5919329491880565,
      "eval_PRM F1 Neg": 0.35,
      "eval_PRM NPV": 0.4117647058823529,
      "eval_PRM Precision": 0.8202247191011236,
      "eval_PRM Recall": 0.8795180722891566,
      "eval_PRM Specificty": 0.30434782608695654,
      "eval_loss": 0.6239346861839294,
      "eval_runtime": 20.5719,
      "eval_samples_per_second": 2.139,
      "eval_steps_per_second": 0.146,
      "step": 1485
    },
    {
      "epoch": 0.1923811373272486,
      "grad_norm": 0.17321588041167113,
      "learning_rate": 1.2178283949876246e-06,
      "loss": 0.8705,
      "step": 1486
    },
    {
      "epoch": 0.19251059973460208,
      "grad_norm": 0.1312232559117139,
      "learning_rate": 1.2177388738030216e-06,
      "loss": 0.8971,
      "step": 1487
    },
    {
      "epoch": 0.19264006214195553,
      "grad_norm": 0.26388811428812864,
      "learning_rate": 1.2176492315394894e-06,
      "loss": 0.8824,
      "step": 1488
    },
    {
      "epoch": 0.19276952454930898,
      "grad_norm": 0.16543157600963063,
      "learning_rate": 1.217559468215339e-06,
      "loss": 0.8843,
      "step": 1489
    },
    {
      "epoch": 0.19289898695666247,
      "grad_norm": 0.1551365807462108,
      "learning_rate": 1.217469583848907e-06,
      "loss": 0.9177,
      "step": 1490
    },
    {
      "epoch": 0.19289898695666247,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8439306358381503,
      "eval_PRM F1 AUC": 0.570193818753274,
      "eval_PRM F1 Neg": 0.3076923076923077,
      "eval_PRM NPV": 0.375,
      "eval_PRM Precision": 0.8111111111111111,
      "eval_PRM Recall": 0.8795180722891566,
      "eval_PRM Specificty": 0.2608695652173913,
      "eval_loss": 0.6255326867103577,
      "eval_runtime": 19.5345,
      "eval_samples_per_second": 2.252,
      "eval_steps_per_second": 0.154,
      "step": 1490
    },
    {
      "epoch": 0.19302844936401592,
      "grad_norm": 0.21578447880394208,
      "learning_rate": 1.2173795784585534e-06,
      "loss": 0.8799,
      "step": 1491
    },
    {
      "epoch": 0.1931579117713694,
      "grad_norm": 0.2420838275074681,
      "learning_rate": 1.217289452062664e-06,
      "loss": 0.8538,
      "step": 1492
    },
    {
      "epoch": 0.19328737417872285,
      "grad_norm": 0.12899103428789666,
      "learning_rate": 1.217199204679649e-06,
      "loss": 0.9333,
      "step": 1493
    },
    {
      "epoch": 0.1934168365860763,
      "grad_norm": 0.1575044790775156,
      "learning_rate": 1.2171088363279426e-06,
      "loss": 0.8439,
      "step": 1494
    },
    {
      "epoch": 0.1935462989934298,
      "grad_norm": 0.16521142174900716,
      "learning_rate": 1.2170183470260053e-06,
      "loss": 0.8635,
      "step": 1495
    },
    {
      "epoch": 0.1935462989934298,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8372093023255814,
      "eval_PRM F1 AUC": 0.5641697223677318,
      "eval_PRM F1 Neg": 0.3,
      "eval_PRM NPV": 0.35294117647058826,
      "eval_PRM Precision": 0.8089887640449438,
      "eval_PRM Recall": 0.8674698795180723,
      "eval_PRM Specificty": 0.2608695652173913,
      "eval_loss": 0.6257102489471436,
      "eval_runtime": 19.8557,
      "eval_samples_per_second": 2.216,
      "eval_steps_per_second": 0.151,
      "step": 1495
    },
    {
      "epoch": 0.19367576140078324,
      "grad_norm": 0.25336287572368543,
      "learning_rate": 1.2169277367923207e-06,
      "loss": 0.8083,
      "step": 1496
    },
    {
      "epoch": 0.19380522380813672,
      "grad_norm": 0.1384116895443285,
      "learning_rate": 1.2168370056453983e-06,
      "loss": 0.9475,
      "step": 1497
    },
    {
      "epoch": 0.19393468621549018,
      "grad_norm": 0.12383439835723532,
      "learning_rate": 1.2167461536037712e-06,
      "loss": 0.8218,
      "step": 1498
    },
    {
      "epoch": 0.19406414862284363,
      "grad_norm": 0.19550464287812822,
      "learning_rate": 1.2166551806859984e-06,
      "loss": 0.8016,
      "step": 1499
    },
    {
      "epoch": 0.1941936110301971,
      "grad_norm": 0.1341300518085345,
      "learning_rate": 1.2165640869106624e-06,
      "loss": 1.0131,
      "step": 1500
    },
    {
      "epoch": 0.1941936110301971,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8439306358381503,
      "eval_PRM F1 AUC": 0.570193818753274,
      "eval_PRM F1 Neg": 0.3076923076923077,
      "eval_PRM NPV": 0.375,
      "eval_PRM Precision": 0.8111111111111111,
      "eval_PRM Recall": 0.8795180722891566,
      "eval_PRM Specificty": 0.2608695652173913,
      "eval_loss": 0.6244673132896423,
      "eval_runtime": 19.7358,
      "eval_samples_per_second": 2.229,
      "eval_steps_per_second": 0.152,
      "step": 1500
    },
    {
      "epoch": 0.19432307343755056,
      "grad_norm": 0.13052975270744205,
      "learning_rate": 1.2164728722963713e-06,
      "loss": 0.9156,
      "step": 1501
    },
    {
      "epoch": 0.19445253584490405,
      "grad_norm": 0.14072916232450214,
      "learning_rate": 1.2163815368617574e-06,
      "loss": 0.8717,
      "step": 1502
    },
    {
      "epoch": 0.1945819982522575,
      "grad_norm": 0.12458261388746868,
      "learning_rate": 1.2162900806254778e-06,
      "loss": 0.814,
      "step": 1503
    },
    {
      "epoch": 0.19471146065961095,
      "grad_norm": 0.16367818874654433,
      "learning_rate": 1.2161985036062146e-06,
      "loss": 0.8903,
      "step": 1504
    },
    {
      "epoch": 0.19484092306696443,
      "grad_norm": 0.23387189744720313,
      "learning_rate": 1.2161068058226737e-06,
      "loss": 0.975,
      "step": 1505
    },
    {
      "epoch": 0.19484092306696443,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8439306358381503,
      "eval_PRM F1 AUC": 0.570193818753274,
      "eval_PRM F1 Neg": 0.3076923076923077,
      "eval_PRM NPV": 0.375,
      "eval_PRM Precision": 0.8111111111111111,
      "eval_PRM Recall": 0.8795180722891566,
      "eval_PRM Specificty": 0.2608695652173913,
      "eval_loss": 0.6223366260528564,
      "eval_runtime": 19.708,
      "eval_samples_per_second": 2.233,
      "eval_steps_per_second": 0.152,
      "step": 1505
    },
    {
      "epoch": 0.1949703854743179,
      "grad_norm": 0.18029020563409295,
      "learning_rate": 1.2160149872935866e-06,
      "loss": 0.8647,
      "step": 1506
    },
    {
      "epoch": 0.19509984788167137,
      "grad_norm": 0.1844178992745953,
      "learning_rate": 1.215923048037709e-06,
      "loss": 0.8331,
      "step": 1507
    },
    {
      "epoch": 0.19522931028902482,
      "grad_norm": 0.2144744081910918,
      "learning_rate": 1.2158309880738215e-06,
      "loss": 0.8563,
      "step": 1508
    },
    {
      "epoch": 0.1953587726963783,
      "grad_norm": 0.2573212636995916,
      "learning_rate": 1.2157388074207288e-06,
      "loss": 0.8705,
      "step": 1509
    },
    {
      "epoch": 0.19548823510373176,
      "grad_norm": 0.18634567438631153,
      "learning_rate": 1.2156465060972612e-06,
      "loss": 0.7863,
      "step": 1510
    },
    {
      "epoch": 0.19548823510373176,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8439306358381503,
      "eval_PRM F1 AUC": 0.570193818753274,
      "eval_PRM F1 Neg": 0.3076923076923077,
      "eval_PRM NPV": 0.375,
      "eval_PRM Precision": 0.8111111111111111,
      "eval_PRM Recall": 0.8795180722891566,
      "eval_PRM Specificty": 0.2608695652173913,
      "eval_loss": 0.6239346861839294,
      "eval_runtime": 19.9058,
      "eval_samples_per_second": 2.21,
      "eval_steps_per_second": 0.151,
      "step": 1510
    },
    {
      "epoch": 0.1956176975110852,
      "grad_norm": 0.1472663757642275,
      "learning_rate": 1.2155540841222727e-06,
      "loss": 0.9034,
      "step": 1511
    },
    {
      "epoch": 0.1957471599184387,
      "grad_norm": 0.11607804311381055,
      "learning_rate": 1.2154615415146428e-06,
      "loss": 0.854,
      "step": 1512
    },
    {
      "epoch": 0.19587662232579214,
      "grad_norm": 0.16976877129100593,
      "learning_rate": 1.2153688782932746e-06,
      "loss": 0.8475,
      "step": 1513
    },
    {
      "epoch": 0.19600608473314562,
      "grad_norm": 0.13694575324728872,
      "learning_rate": 1.215276094477097e-06,
      "loss": 0.9114,
      "step": 1514
    },
    {
      "epoch": 0.19613554714049908,
      "grad_norm": 0.23889577306620396,
      "learning_rate": 1.2151831900850627e-06,
      "loss": 0.8826,
      "step": 1515
    },
    {
      "epoch": 0.19613554714049908,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8439306358381503,
      "eval_PRM F1 AUC": 0.570193818753274,
      "eval_PRM F1 Neg": 0.3076923076923077,
      "eval_PRM NPV": 0.375,
      "eval_PRM Precision": 0.8111111111111111,
      "eval_PRM Recall": 0.8795180722891566,
      "eval_PRM Specificty": 0.2608695652173913,
      "eval_loss": 0.6257102489471436,
      "eval_runtime": 20.6403,
      "eval_samples_per_second": 2.132,
      "eval_steps_per_second": 0.145,
      "step": 1515
    },
    {
      "epoch": 0.19626500954785253,
      "grad_norm": 0.2047004605603871,
      "learning_rate": 1.2150901651361493e-06,
      "loss": 0.8406,
      "step": 1516
    },
    {
      "epoch": 0.196394471955206,
      "grad_norm": 0.11453846322101759,
      "learning_rate": 1.2149970196493593e-06,
      "loss": 0.8597,
      "step": 1517
    },
    {
      "epoch": 0.19652393436255947,
      "grad_norm": 0.18248260253153736,
      "learning_rate": 1.2149037536437192e-06,
      "loss": 0.9385,
      "step": 1518
    },
    {
      "epoch": 0.19665339676991295,
      "grad_norm": 0.13676314851802718,
      "learning_rate": 1.2148103671382805e-06,
      "loss": 0.9036,
      "step": 1519
    },
    {
      "epoch": 0.1967828591772664,
      "grad_norm": 0.14380786441223595,
      "learning_rate": 1.2147168601521194e-06,
      "loss": 1.0725,
      "step": 1520
    },
    {
      "epoch": 0.1967828591772664,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8439306358381503,
      "eval_PRM F1 AUC": 0.570193818753274,
      "eval_PRM F1 Neg": 0.3076923076923077,
      "eval_PRM NPV": 0.375,
      "eval_PRM Precision": 0.8111111111111111,
      "eval_PRM Recall": 0.8795180722891566,
      "eval_PRM Specificty": 0.2608695652173913,
      "eval_loss": 0.625,
      "eval_runtime": 20.1181,
      "eval_samples_per_second": 2.187,
      "eval_steps_per_second": 0.149,
      "step": 1520
    },
    {
      "epoch": 0.19691232158461985,
      "grad_norm": 0.17463918802291536,
      "learning_rate": 1.2146232327043367e-06,
      "loss": 0.7628,
      "step": 1521
    },
    {
      "epoch": 0.19704178399197333,
      "grad_norm": 0.13517525619758083,
      "learning_rate": 1.2145294848140577e-06,
      "loss": 0.9646,
      "step": 1522
    },
    {
      "epoch": 0.1971712463993268,
      "grad_norm": 0.188857704691892,
      "learning_rate": 1.214435616500432e-06,
      "loss": 1.0177,
      "step": 1523
    },
    {
      "epoch": 0.19730070880668027,
      "grad_norm": 0.18367034632172918,
      "learning_rate": 1.2143416277826344e-06,
      "loss": 0.8076,
      "step": 1524
    },
    {
      "epoch": 0.19743017121403372,
      "grad_norm": 0.12301546315803834,
      "learning_rate": 1.214247518679864e-06,
      "loss": 0.8115,
      "step": 1525
    },
    {
      "epoch": 0.19743017121403372,
      "eval_PRM Accuracy": 0.7641509433962265,
      "eval_PRM F1": 0.8554913294797688,
      "eval_PRM F1 AUC": 0.5979570455735987,
      "eval_PRM F1 Neg": 0.358974358974359,
      "eval_PRM NPV": 0.4375,
      "eval_PRM Precision": 0.8222222222222222,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.30434782608695654,
      "eval_loss": 0.6235795617103577,
      "eval_runtime": 20.0972,
      "eval_samples_per_second": 2.189,
      "eval_steps_per_second": 0.149,
      "step": 1525
    },
    {
      "epoch": 0.19755963362138718,
      "grad_norm": 0.19283512954448337,
      "learning_rate": 1.2141532892113443e-06,
      "loss": 0.7791,
      "step": 1526
    },
    {
      "epoch": 0.19768909602874066,
      "grad_norm": 0.26700155774763557,
      "learning_rate": 1.2140589393963237e-06,
      "loss": 0.8768,
      "step": 1527
    },
    {
      "epoch": 0.1978185584360941,
      "grad_norm": 0.1508972091924051,
      "learning_rate": 1.213964469254075e-06,
      "loss": 1.0068,
      "step": 1528
    },
    {
      "epoch": 0.1979480208434476,
      "grad_norm": 0.22040804183628093,
      "learning_rate": 1.2138698788038958e-06,
      "loss": 0.9852,
      "step": 1529
    },
    {
      "epoch": 0.19807748325080105,
      "grad_norm": 0.24428903769620267,
      "learning_rate": 1.213775168065108e-06,
      "loss": 0.8591,
      "step": 1530
    },
    {
      "epoch": 0.19807748325080105,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8505747126436781,
      "eval_PRM F1 AUC": 0.5762179151388162,
      "eval_PRM F1 Neg": 0.3157894736842105,
      "eval_PRM NPV": 0.4,
      "eval_PRM Precision": 0.8131868131868132,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.2608695652173913,
      "eval_loss": 0.6232244372367859,
      "eval_runtime": 20.877,
      "eval_samples_per_second": 2.108,
      "eval_steps_per_second": 0.144,
      "step": 1530
    },
    {
      "epoch": 0.19820694565815453,
      "grad_norm": 0.18628252208328708,
      "learning_rate": 1.2136803370570583e-06,
      "loss": 0.759,
      "step": 1531
    },
    {
      "epoch": 0.19833640806550798,
      "grad_norm": 0.10885706677277951,
      "learning_rate": 1.2135853857991175e-06,
      "loss": 0.7986,
      "step": 1532
    },
    {
      "epoch": 0.19846587047286143,
      "grad_norm": 0.2381254474269766,
      "learning_rate": 1.2134903143106816e-06,
      "loss": 0.9197,
      "step": 1533
    },
    {
      "epoch": 0.19859533288021491,
      "grad_norm": 0.1996031887004159,
      "learning_rate": 1.2133951226111709e-06,
      "loss": 0.8341,
      "step": 1534
    },
    {
      "epoch": 0.19872479528756837,
      "grad_norm": 0.13837401381761152,
      "learning_rate": 1.2132998107200304e-06,
      "loss": 0.7312,
      "step": 1535
    },
    {
      "epoch": 0.19872479528756837,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8505747126436781,
      "eval_PRM F1 AUC": 0.5762179151388162,
      "eval_PRM F1 Neg": 0.3157894736842105,
      "eval_PRM NPV": 0.4,
      "eval_PRM Precision": 0.8131868131868132,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.2608695652173913,
      "eval_loss": 0.6260653138160706,
      "eval_runtime": 19.9196,
      "eval_samples_per_second": 2.209,
      "eval_steps_per_second": 0.151,
      "step": 1535
    },
    {
      "epoch": 0.19885425769492185,
      "grad_norm": 0.11936880371669605,
      "learning_rate": 1.213204378656729e-06,
      "loss": 0.816,
      "step": 1536
    },
    {
      "epoch": 0.1989837201022753,
      "grad_norm": 0.12455832123397904,
      "learning_rate": 1.213108826440761e-06,
      "loss": 0.8364,
      "step": 1537
    },
    {
      "epoch": 0.19911318250962876,
      "grad_norm": 0.16415738973021912,
      "learning_rate": 1.2130131540916447e-06,
      "loss": 0.7104,
      "step": 1538
    },
    {
      "epoch": 0.19924264491698224,
      "grad_norm": 0.2092732519883409,
      "learning_rate": 1.2129173616289233e-06,
      "loss": 0.8618,
      "step": 1539
    },
    {
      "epoch": 0.1993721073243357,
      "grad_norm": 0.1372179281715074,
      "learning_rate": 1.212821449072164e-06,
      "loss": 0.8546,
      "step": 1540
    },
    {
      "epoch": 0.1993721073243357,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8439306358381503,
      "eval_PRM F1 AUC": 0.570193818753274,
      "eval_PRM F1 Neg": 0.3076923076923077,
      "eval_PRM NPV": 0.375,
      "eval_PRM Precision": 0.8111111111111111,
      "eval_PRM Recall": 0.8795180722891566,
      "eval_PRM Specificty": 0.2608695652173913,
      "eval_loss": 0.6262428760528564,
      "eval_runtime": 19.4428,
      "eval_samples_per_second": 2.263,
      "eval_steps_per_second": 0.154,
      "step": 1540
    },
    {
      "epoch": 0.19950156973168917,
      "grad_norm": 0.12942096449903945,
      "learning_rate": 1.2127254164409593e-06,
      "loss": 0.9335,
      "step": 1541
    },
    {
      "epoch": 0.19963103213904262,
      "grad_norm": 0.13522747540437738,
      "learning_rate": 1.212629263754926e-06,
      "loss": 1.0117,
      "step": 1542
    },
    {
      "epoch": 0.19976049454639608,
      "grad_norm": 0.24356515335843182,
      "learning_rate": 1.2125329910337045e-06,
      "loss": 0.7337,
      "step": 1543
    },
    {
      "epoch": 0.19988995695374956,
      "grad_norm": 0.27170951989269226,
      "learning_rate": 1.2124365982969607e-06,
      "loss": 1.1162,
      "step": 1544
    },
    {
      "epoch": 0.200019419361103,
      "grad_norm": 0.13226609321275945,
      "learning_rate": 1.212340085564385e-06,
      "loss": 0.8494,
      "step": 1545
    },
    {
      "epoch": 0.200019419361103,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8505747126436781,
      "eval_PRM F1 AUC": 0.5762179151388162,
      "eval_PRM F1 Neg": 0.3157894736842105,
      "eval_PRM NPV": 0.4,
      "eval_PRM Precision": 0.8131868131868132,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.2608695652173913,
      "eval_loss": 0.6221590638160706,
      "eval_runtime": 19.9708,
      "eval_samples_per_second": 2.203,
      "eval_steps_per_second": 0.15,
      "step": 1545
    },
    {
      "epoch": 0.2001488817684565,
      "grad_norm": 0.11619693641881697,
      "learning_rate": 1.212243452855692e-06,
      "loss": 0.9149,
      "step": 1546
    },
    {
      "epoch": 0.20027834417580995,
      "grad_norm": 0.20244973962122417,
      "learning_rate": 1.212146700190621e-06,
      "loss": 0.8444,
      "step": 1547
    },
    {
      "epoch": 0.2004078065831634,
      "grad_norm": 0.16514647329761667,
      "learning_rate": 1.2120498275889357e-06,
      "loss": 0.9105,
      "step": 1548
    },
    {
      "epoch": 0.20053726899051688,
      "grad_norm": 0.242842086660808,
      "learning_rate": 1.2119528350704237e-06,
      "loss": 0.8248,
      "step": 1549
    },
    {
      "epoch": 0.20066673139787033,
      "grad_norm": 0.15488306908683416,
      "learning_rate": 1.2118557226548984e-06,
      "loss": 0.9443,
      "step": 1550
    },
    {
      "epoch": 0.20066673139787033,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8505747126436781,
      "eval_PRM F1 AUC": 0.5762179151388162,
      "eval_PRM F1 Neg": 0.3157894736842105,
      "eval_PRM NPV": 0.4,
      "eval_PRM Precision": 0.8131868131868132,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.2608695652173913,
      "eval_loss": 0.6237571239471436,
      "eval_runtime": 20.0954,
      "eval_samples_per_second": 2.19,
      "eval_steps_per_second": 0.149,
      "step": 1550
    },
    {
      "epoch": 0.20079619380522382,
      "grad_norm": 0.196169919226657,
      "learning_rate": 1.2117584903621968e-06,
      "loss": 0.8773,
      "step": 1551
    },
    {
      "epoch": 0.20092565621257727,
      "grad_norm": 0.17054639005776,
      "learning_rate": 1.2116611382121803e-06,
      "loss": 0.8887,
      "step": 1552
    },
    {
      "epoch": 0.20105511861993075,
      "grad_norm": 0.1321306491480727,
      "learning_rate": 1.2115636662247355e-06,
      "loss": 0.9053,
      "step": 1553
    },
    {
      "epoch": 0.2011845810272842,
      "grad_norm": 0.1343474242889647,
      "learning_rate": 1.2114660744197725e-06,
      "loss": 0.965,
      "step": 1554
    },
    {
      "epoch": 0.20131404343463766,
      "grad_norm": 0.1319088414890238,
      "learning_rate": 1.2113683628172269e-06,
      "loss": 0.9573,
      "step": 1555
    },
    {
      "epoch": 0.20131404343463766,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8505747126436781,
      "eval_PRM F1 AUC": 0.5762179151388162,
      "eval_PRM F1 Neg": 0.3157894736842105,
      "eval_PRM NPV": 0.4,
      "eval_PRM Precision": 0.8131868131868132,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.2608695652173913,
      "eval_loss": 0.6253551244735718,
      "eval_runtime": 19.9803,
      "eval_samples_per_second": 2.202,
      "eval_steps_per_second": 0.15,
      "step": 1555
    },
    {
      "epoch": 0.20144350584199114,
      "grad_norm": 0.2941162866688178,
      "learning_rate": 1.2112705314370579e-06,
      "loss": 0.8477,
      "step": 1556
    },
    {
      "epoch": 0.2015729682493446,
      "grad_norm": 0.1565228497978276,
      "learning_rate": 1.2111725802992498e-06,
      "loss": 0.8992,
      "step": 1557
    },
    {
      "epoch": 0.20170243065669807,
      "grad_norm": 0.22988756794947487,
      "learning_rate": 1.2110745094238107e-06,
      "loss": 0.7809,
      "step": 1558
    },
    {
      "epoch": 0.20183189306405153,
      "grad_norm": 0.20543238808099587,
      "learning_rate": 1.2109763188307741e-06,
      "loss": 0.8171,
      "step": 1559
    },
    {
      "epoch": 0.20196135547140498,
      "grad_norm": 0.1775243042627418,
      "learning_rate": 1.2108780085401971e-06,
      "loss": 0.8031,
      "step": 1560
    },
    {
      "epoch": 0.20196135547140498,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8505747126436781,
      "eval_PRM F1 AUC": 0.5762179151388162,
      "eval_PRM F1 Neg": 0.3157894736842105,
      "eval_PRM NPV": 0.4,
      "eval_PRM Precision": 0.8131868131868132,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.2608695652173913,
      "eval_loss": 0.6244673132896423,
      "eval_runtime": 19.941,
      "eval_samples_per_second": 2.207,
      "eval_steps_per_second": 0.15,
      "step": 1560
    },
    {
      "epoch": 0.20209081787875846,
      "grad_norm": 0.24211348048115597,
      "learning_rate": 1.2107795785721618e-06,
      "loss": 0.7687,
      "step": 1561
    },
    {
      "epoch": 0.20222028028611191,
      "grad_norm": 0.24804869407795624,
      "learning_rate": 1.2106810289467742e-06,
      "loss": 0.8571,
      "step": 1562
    },
    {
      "epoch": 0.2023497426934654,
      "grad_norm": 0.24306077280788405,
      "learning_rate": 1.2105823596841654e-06,
      "loss": 0.8899,
      "step": 1563
    },
    {
      "epoch": 0.20247920510081885,
      "grad_norm": 0.22091481898928056,
      "learning_rate": 1.2104835708044903e-06,
      "loss": 0.8024,
      "step": 1564
    },
    {
      "epoch": 0.2026086675081723,
      "grad_norm": 0.1455159498509513,
      "learning_rate": 1.2103846623279285e-06,
      "loss": 0.8088,
      "step": 1565
    },
    {
      "epoch": 0.2026086675081723,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8457142857142858,
      "eval_PRM F1 AUC": 0.5544787847040336,
      "eval_PRM F1 Neg": 0.2702702702702703,
      "eval_PRM NPV": 0.35714285714285715,
      "eval_PRM Precision": 0.8043478260869565,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.6248224377632141,
      "eval_runtime": 20.6518,
      "eval_samples_per_second": 2.131,
      "eval_steps_per_second": 0.145,
      "step": 1565
    },
    {
      "epoch": 0.20273812991552578,
      "grad_norm": 0.2202031079908931,
      "learning_rate": 1.2102856342746846e-06,
      "loss": 0.7515,
      "step": 1566
    },
    {
      "epoch": 0.20286759232287924,
      "grad_norm": 0.2756886096445566,
      "learning_rate": 1.2101864866649863e-06,
      "loss": 0.8577,
      "step": 1567
    },
    {
      "epoch": 0.20299705473023272,
      "grad_norm": 0.15147084632937635,
      "learning_rate": 1.2100872195190872e-06,
      "loss": 0.899,
      "step": 1568
    },
    {
      "epoch": 0.20312651713758617,
      "grad_norm": 0.14489090064892268,
      "learning_rate": 1.2099878328572642e-06,
      "loss": 0.9209,
      "step": 1569
    },
    {
      "epoch": 0.20325597954493962,
      "grad_norm": 0.1758676687728585,
      "learning_rate": 1.2098883266998193e-06,
      "loss": 0.8234,
      "step": 1570
    },
    {
      "epoch": 0.20325597954493962,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8505747126436781,
      "eval_PRM F1 AUC": 0.5762179151388162,
      "eval_PRM F1 Neg": 0.3157894736842105,
      "eval_PRM NPV": 0.4,
      "eval_PRM Precision": 0.8131868131868132,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.2608695652173913,
      "eval_loss": 0.6242897510528564,
      "eval_runtime": 20.1593,
      "eval_samples_per_second": 2.183,
      "eval_steps_per_second": 0.149,
      "step": 1570
    },
    {
      "epoch": 0.2033854419522931,
      "grad_norm": 0.1498618416260244,
      "learning_rate": 1.2097887010670783e-06,
      "loss": 0.9258,
      "step": 1571
    },
    {
      "epoch": 0.20351490435964656,
      "grad_norm": 0.10948602931564777,
      "learning_rate": 1.2096889559793921e-06,
      "loss": 0.8307,
      "step": 1572
    },
    {
      "epoch": 0.20364436676700004,
      "grad_norm": 0.18653487861509524,
      "learning_rate": 1.2095890914571356e-06,
      "loss": 0.9301,
      "step": 1573
    },
    {
      "epoch": 0.2037738291743535,
      "grad_norm": 0.13703302829833372,
      "learning_rate": 1.209489107520708e-06,
      "loss": 0.7942,
      "step": 1574
    },
    {
      "epoch": 0.20390329158170697,
      "grad_norm": 0.1955620750495807,
      "learning_rate": 1.2093890041905332e-06,
      "loss": 0.7502,
      "step": 1575
    },
    {
      "epoch": 0.20390329158170697,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8457142857142858,
      "eval_PRM F1 AUC": 0.5544787847040336,
      "eval_PRM F1 Neg": 0.2702702702702703,
      "eval_PRM NPV": 0.35714285714285715,
      "eval_PRM Precision": 0.8043478260869565,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.6232244372367859,
      "eval_runtime": 20.1555,
      "eval_samples_per_second": 2.183,
      "eval_steps_per_second": 0.149,
      "step": 1575
    },
    {
      "epoch": 0.20403275398906043,
      "grad_norm": 0.12719279267920622,
      "learning_rate": 1.2092887814870593e-06,
      "loss": 0.9412,
      "step": 1576
    },
    {
      "epoch": 0.20416221639641388,
      "grad_norm": 0.18610763048203205,
      "learning_rate": 1.2091884394307587e-06,
      "loss": 0.8206,
      "step": 1577
    },
    {
      "epoch": 0.20429167880376736,
      "grad_norm": 0.12891882993483156,
      "learning_rate": 1.2090879780421282e-06,
      "loss": 0.8757,
      "step": 1578
    },
    {
      "epoch": 0.20442114121112082,
      "grad_norm": 0.10916176066069223,
      "learning_rate": 1.2089873973416896e-06,
      "loss": 0.8531,
      "step": 1579
    },
    {
      "epoch": 0.2045506036184743,
      "grad_norm": 0.16964901923939799,
      "learning_rate": 1.2088866973499882e-06,
      "loss": 0.7671,
      "step": 1580
    },
    {
      "epoch": 0.2045506036184743,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8457142857142858,
      "eval_PRM F1 AUC": 0.5544787847040336,
      "eval_PRM F1 Neg": 0.2702702702702703,
      "eval_PRM NPV": 0.35714285714285715,
      "eval_PRM Precision": 0.8043478260869565,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.625,
      "eval_runtime": 20.8383,
      "eval_samples_per_second": 2.111,
      "eval_steps_per_second": 0.144,
      "step": 1580
    },
    {
      "epoch": 0.20468006602582775,
      "grad_norm": 0.1346530452139494,
      "learning_rate": 1.2087858780875938e-06,
      "loss": 0.988,
      "step": 1581
    },
    {
      "epoch": 0.2048095284331812,
      "grad_norm": 0.23779581833363456,
      "learning_rate": 1.2086849395751011e-06,
      "loss": 0.7836,
      "step": 1582
    },
    {
      "epoch": 0.20493899084053468,
      "grad_norm": 0.20651681893851645,
      "learning_rate": 1.208583881833129e-06,
      "loss": 0.9043,
      "step": 1583
    },
    {
      "epoch": 0.20506845324788814,
      "grad_norm": 0.20163406095638087,
      "learning_rate": 1.2084827048823198e-06,
      "loss": 0.9615,
      "step": 1584
    },
    {
      "epoch": 0.20519791565524162,
      "grad_norm": 0.13467576449701407,
      "learning_rate": 1.2083814087433418e-06,
      "loss": 0.7384,
      "step": 1585
    },
    {
      "epoch": 0.20519791565524162,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8457142857142858,
      "eval_PRM F1 AUC": 0.5544787847040336,
      "eval_PRM F1 Neg": 0.2702702702702703,
      "eval_PRM NPV": 0.35714285714285715,
      "eval_PRM Precision": 0.8043478260869565,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.6237571239471436,
      "eval_runtime": 20.5361,
      "eval_samples_per_second": 2.143,
      "eval_steps_per_second": 0.146,
      "step": 1585
    },
    {
      "epoch": 0.20532737806259507,
      "grad_norm": 0.1895209762572917,
      "learning_rate": 1.2082799934368864e-06,
      "loss": 0.7754,
      "step": 1586
    },
    {
      "epoch": 0.20545684046994853,
      "grad_norm": 0.18883349241424993,
      "learning_rate": 1.2081784589836699e-06,
      "loss": 0.7867,
      "step": 1587
    },
    {
      "epoch": 0.205586302877302,
      "grad_norm": 0.1612692585360582,
      "learning_rate": 1.2080768054044324e-06,
      "loss": 0.9226,
      "step": 1588
    },
    {
      "epoch": 0.20571576528465546,
      "grad_norm": 0.13303174178555616,
      "learning_rate": 1.207975032719939e-06,
      "loss": 0.7629,
      "step": 1589
    },
    {
      "epoch": 0.20584522769200894,
      "grad_norm": 0.2037710626083591,
      "learning_rate": 1.2078731409509792e-06,
      "loss": 0.81,
      "step": 1590
    },
    {
      "epoch": 0.20584522769200894,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8457142857142858,
      "eval_PRM F1 AUC": 0.5544787847040336,
      "eval_PRM F1 Neg": 0.2702702702702703,
      "eval_PRM NPV": 0.35714285714285715,
      "eval_PRM Precision": 0.8043478260869565,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.6241121888160706,
      "eval_runtime": 19.7642,
      "eval_samples_per_second": 2.226,
      "eval_steps_per_second": 0.152,
      "step": 1590
    },
    {
      "epoch": 0.2059746900993624,
      "grad_norm": 0.16426595682514808,
      "learning_rate": 1.2077711301183656e-06,
      "loss": 0.8038,
      "step": 1591
    },
    {
      "epoch": 0.20610415250671585,
      "grad_norm": 0.1921842484434961,
      "learning_rate": 1.2076690002429368e-06,
      "loss": 0.823,
      "step": 1592
    },
    {
      "epoch": 0.20623361491406933,
      "grad_norm": 0.1548305073680677,
      "learning_rate": 1.2075667513455542e-06,
      "loss": 0.7839,
      "step": 1593
    },
    {
      "epoch": 0.20636307732142278,
      "grad_norm": 0.14875560036727853,
      "learning_rate": 1.2074643834471048e-06,
      "loss": 0.9855,
      "step": 1594
    },
    {
      "epoch": 0.20649253972877626,
      "grad_norm": 0.2747014748592745,
      "learning_rate": 1.207361896568499e-06,
      "loss": 0.8872,
      "step": 1595
    },
    {
      "epoch": 0.20649253972877626,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8457142857142858,
      "eval_PRM F1 AUC": 0.5544787847040336,
      "eval_PRM F1 Neg": 0.2702702702702703,
      "eval_PRM NPV": 0.35714285714285715,
      "eval_PRM Precision": 0.8043478260869565,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.6246448755264282,
      "eval_runtime": 20.1819,
      "eval_samples_per_second": 2.18,
      "eval_steps_per_second": 0.149,
      "step": 1595
    },
    {
      "epoch": 0.20662200213612972,
      "grad_norm": 0.1561128334240305,
      "learning_rate": 1.2072592907306718e-06,
      "loss": 0.8591,
      "step": 1596
    },
    {
      "epoch": 0.2067514645434832,
      "grad_norm": 0.2521124820249253,
      "learning_rate": 1.2071565659545826e-06,
      "loss": 0.7778,
      "step": 1597
    },
    {
      "epoch": 0.20688092695083665,
      "grad_norm": 0.15820168378503868,
      "learning_rate": 1.2070537222612153e-06,
      "loss": 0.9658,
      "step": 1598
    },
    {
      "epoch": 0.2070103893581901,
      "grad_norm": 0.2874229339554774,
      "learning_rate": 1.2069507596715773e-06,
      "loss": 0.8182,
      "step": 1599
    },
    {
      "epoch": 0.2071398517655436,
      "grad_norm": 0.12423668190389105,
      "learning_rate": 1.206847678206701e-06,
      "loss": 0.8567,
      "step": 1600
    },
    {
      "epoch": 0.2071398517655436,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8457142857142858,
      "eval_PRM F1 AUC": 0.5544787847040336,
      "eval_PRM F1 Neg": 0.2702702702702703,
      "eval_PRM NPV": 0.35714285714285715,
      "eval_PRM Precision": 0.8043478260869565,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.6235795617103577,
      "eval_runtime": 20.3629,
      "eval_samples_per_second": 2.161,
      "eval_steps_per_second": 0.147,
      "step": 1600
    },
    {
      "epoch": 0.20726931417289704,
      "grad_norm": 0.1943232051377781,
      "learning_rate": 1.206744477887643e-06,
      "loss": 0.8463,
      "step": 1601
    },
    {
      "epoch": 0.20739877658025052,
      "grad_norm": 0.22373769645352284,
      "learning_rate": 1.206641158735484e-06,
      "loss": 0.7328,
      "step": 1602
    },
    {
      "epoch": 0.20752823898760397,
      "grad_norm": 0.2269676980008798,
      "learning_rate": 1.2065377207713288e-06,
      "loss": 0.8074,
      "step": 1603
    },
    {
      "epoch": 0.20765770139495743,
      "grad_norm": 0.13678190164639778,
      "learning_rate": 1.2064341640163071e-06,
      "loss": 0.8086,
      "step": 1604
    },
    {
      "epoch": 0.2077871638023109,
      "grad_norm": 0.15696847596627853,
      "learning_rate": 1.2063304884915722e-06,
      "loss": 0.7911,
      "step": 1605
    },
    {
      "epoch": 0.2077871638023109,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8457142857142858,
      "eval_PRM F1 AUC": 0.5544787847040336,
      "eval_PRM F1 Neg": 0.2702702702702703,
      "eval_PRM NPV": 0.35714285714285715,
      "eval_PRM Precision": 0.8043478260869565,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.6253551244735718,
      "eval_runtime": 20.3185,
      "eval_samples_per_second": 2.166,
      "eval_steps_per_second": 0.148,
      "step": 1605
    },
    {
      "epoch": 0.20791662620966436,
      "grad_norm": 0.13240466693604575,
      "learning_rate": 1.206226694218302e-06,
      "loss": 0.9292,
      "step": 1606
    },
    {
      "epoch": 0.20804608861701784,
      "grad_norm": 0.17778557793291358,
      "learning_rate": 1.2061227812176986e-06,
      "loss": 0.7954,
      "step": 1607
    },
    {
      "epoch": 0.2081755510243713,
      "grad_norm": 0.18644819040454264,
      "learning_rate": 1.2060187495109884e-06,
      "loss": 0.832,
      "step": 1608
    },
    {
      "epoch": 0.20830501343172475,
      "grad_norm": 0.1263814289395932,
      "learning_rate": 1.2059145991194218e-06,
      "loss": 0.8779,
      "step": 1609
    },
    {
      "epoch": 0.20843447583907823,
      "grad_norm": 0.14141713325381083,
      "learning_rate": 1.2058103300642737e-06,
      "loss": 0.9478,
      "step": 1610
    },
    {
      "epoch": 0.20843447583907823,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8457142857142858,
      "eval_PRM F1 AUC": 0.5544787847040336,
      "eval_PRM F1 Neg": 0.2702702702702703,
      "eval_PRM NPV": 0.35714285714285715,
      "eval_PRM Precision": 0.8043478260869565,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.6239346861839294,
      "eval_runtime": 20.2113,
      "eval_samples_per_second": 2.177,
      "eval_steps_per_second": 0.148,
      "step": 1610
    },
    {
      "epoch": 0.20856393824643168,
      "grad_norm": 0.1461677830465823,
      "learning_rate": 1.2057059423668433e-06,
      "loss": 0.8806,
      "step": 1611
    },
    {
      "epoch": 0.20869340065378517,
      "grad_norm": 0.16563366880289837,
      "learning_rate": 1.2056014360484535e-06,
      "loss": 0.8827,
      "step": 1612
    },
    {
      "epoch": 0.20882286306113862,
      "grad_norm": 0.19099635190589176,
      "learning_rate": 1.2054968111304525e-06,
      "loss": 0.8719,
      "step": 1613
    },
    {
      "epoch": 0.20895232546849207,
      "grad_norm": 0.14855435927405905,
      "learning_rate": 1.2053920676342115e-06,
      "loss": 0.8861,
      "step": 1614
    },
    {
      "epoch": 0.20908178787584555,
      "grad_norm": 0.14124825569879393,
      "learning_rate": 1.2052872055811268e-06,
      "loss": 0.9606,
      "step": 1615
    },
    {
      "epoch": 0.20908178787584555,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8457142857142858,
      "eval_PRM F1 AUC": 0.5544787847040336,
      "eval_PRM F1 Neg": 0.2702702702702703,
      "eval_PRM NPV": 0.35714285714285715,
      "eval_PRM Precision": 0.8043478260869565,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.6257102489471436,
      "eval_runtime": 20.7245,
      "eval_samples_per_second": 2.123,
      "eval_steps_per_second": 0.145,
      "step": 1615
    },
    {
      "epoch": 0.209211250283199,
      "grad_norm": 0.2314861669551217,
      "learning_rate": 1.2051822249926184e-06,
      "loss": 0.8156,
      "step": 1616
    },
    {
      "epoch": 0.2093407126905525,
      "grad_norm": 0.1733955664837983,
      "learning_rate": 1.2050771258901309e-06,
      "loss": 0.8668,
      "step": 1617
    },
    {
      "epoch": 0.20947017509790594,
      "grad_norm": 0.14233985224058066,
      "learning_rate": 1.2049719082951326e-06,
      "loss": 0.7931,
      "step": 1618
    },
    {
      "epoch": 0.20959963750525942,
      "grad_norm": 0.18436563654002758,
      "learning_rate": 1.2048665722291167e-06,
      "loss": 0.7699,
      "step": 1619
    },
    {
      "epoch": 0.20972909991261288,
      "grad_norm": 0.18214821030616316,
      "learning_rate": 1.2047611177135998e-06,
      "loss": 0.772,
      "step": 1620
    },
    {
      "epoch": 0.20972909991261288,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8457142857142858,
      "eval_PRM F1 AUC": 0.5544787847040336,
      "eval_PRM F1 Neg": 0.2702702702702703,
      "eval_PRM NPV": 0.35714285714285715,
      "eval_PRM Precision": 0.8043478260869565,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.623046875,
      "eval_runtime": 20.771,
      "eval_samples_per_second": 2.118,
      "eval_steps_per_second": 0.144,
      "step": 1620
    },
    {
      "epoch": 0.20985856231996633,
      "grad_norm": 0.14739770083286874,
      "learning_rate": 1.2046555447701236e-06,
      "loss": 0.9379,
      "step": 1621
    },
    {
      "epoch": 0.2099880247273198,
      "grad_norm": 0.11167110175006614,
      "learning_rate": 1.2045498534202532e-06,
      "loss": 0.7922,
      "step": 1622
    },
    {
      "epoch": 0.21011748713467326,
      "grad_norm": 0.1443638545314058,
      "learning_rate": 1.2044440436855786e-06,
      "loss": 0.9939,
      "step": 1623
    },
    {
      "epoch": 0.21024694954202675,
      "grad_norm": 0.11299289979267607,
      "learning_rate": 1.204338115587713e-06,
      "loss": 0.8971,
      "step": 1624
    },
    {
      "epoch": 0.2103764119493802,
      "grad_norm": 0.21022693713959548,
      "learning_rate": 1.2042320691482947e-06,
      "loss": 0.9445,
      "step": 1625
    },
    {
      "epoch": 0.2103764119493802,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8457142857142858,
      "eval_PRM F1 AUC": 0.5544787847040336,
      "eval_PRM F1 Neg": 0.2702702702702703,
      "eval_PRM NPV": 0.35714285714285715,
      "eval_PRM Precision": 0.8043478260869565,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.6253551244735718,
      "eval_runtime": 20.2156,
      "eval_samples_per_second": 2.177,
      "eval_steps_per_second": 0.148,
      "step": 1625
    },
    {
      "epoch": 0.21050587435673365,
      "grad_norm": 0.2972099726464034,
      "learning_rate": 1.2041259043889859e-06,
      "loss": 0.7909,
      "step": 1626
    },
    {
      "epoch": 0.21063533676408713,
      "grad_norm": 0.13232906838600983,
      "learning_rate": 1.2040196213314725e-06,
      "loss": 0.8514,
      "step": 1627
    },
    {
      "epoch": 0.2107647991714406,
      "grad_norm": 0.14863232072938184,
      "learning_rate": 1.2039132199974654e-06,
      "loss": 0.8855,
      "step": 1628
    },
    {
      "epoch": 0.21089426157879407,
      "grad_norm": 0.11047835354776975,
      "learning_rate": 1.203806700408699e-06,
      "loss": 0.8264,
      "step": 1629
    },
    {
      "epoch": 0.21102372398614752,
      "grad_norm": 0.17804988170646394,
      "learning_rate": 1.2037000625869322e-06,
      "loss": 0.8489,
      "step": 1630
    },
    {
      "epoch": 0.21102372398614752,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8457142857142858,
      "eval_PRM F1 AUC": 0.5544787847040336,
      "eval_PRM F1 Neg": 0.2702702702702703,
      "eval_PRM NPV": 0.35714285714285715,
      "eval_PRM Precision": 0.8043478260869565,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.6265980005264282,
      "eval_runtime": 20.2316,
      "eval_samples_per_second": 2.175,
      "eval_steps_per_second": 0.148,
      "step": 1630
    },
    {
      "epoch": 0.21115318639350097,
      "grad_norm": 0.1947043971396186,
      "learning_rate": 1.203593306553948e-06,
      "loss": 0.7358,
      "step": 1631
    },
    {
      "epoch": 0.21128264880085446,
      "grad_norm": 0.125221051711122,
      "learning_rate": 1.2034864323315534e-06,
      "loss": 0.8141,
      "step": 1632
    },
    {
      "epoch": 0.2114121112082079,
      "grad_norm": 0.16221257563566632,
      "learning_rate": 1.2033794399415797e-06,
      "loss": 1.0675,
      "step": 1633
    },
    {
      "epoch": 0.2115415736155614,
      "grad_norm": 0.20672253029314883,
      "learning_rate": 1.2032723294058818e-06,
      "loss": 0.8097,
      "step": 1634
    },
    {
      "epoch": 0.21167103602291484,
      "grad_norm": 0.12535106263704623,
      "learning_rate": 1.20316510074634e-06,
      "loss": 0.8927,
      "step": 1635
    },
    {
      "epoch": 0.21167103602291484,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8457142857142858,
      "eval_PRM F1 AUC": 0.5544787847040336,
      "eval_PRM F1 Neg": 0.2702702702702703,
      "eval_PRM NPV": 0.35714285714285715,
      "eval_PRM Precision": 0.8043478260869565,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.6228693127632141,
      "eval_runtime": 20.7962,
      "eval_samples_per_second": 2.116,
      "eval_steps_per_second": 0.144,
      "step": 1635
    },
    {
      "epoch": 0.21180049843026832,
      "grad_norm": 0.1254461214769376,
      "learning_rate": 1.2030577539848572e-06,
      "loss": 0.8713,
      "step": 1636
    },
    {
      "epoch": 0.21192996083762178,
      "grad_norm": 0.2370894967931306,
      "learning_rate": 1.2029502891433613e-06,
      "loss": 0.8047,
      "step": 1637
    },
    {
      "epoch": 0.21205942324497523,
      "grad_norm": 0.1916874023429982,
      "learning_rate": 1.2028427062438047e-06,
      "loss": 0.8459,
      "step": 1638
    },
    {
      "epoch": 0.2121888856523287,
      "grad_norm": 0.18891311316561454,
      "learning_rate": 1.202735005308163e-06,
      "loss": 0.799,
      "step": 1639
    },
    {
      "epoch": 0.21231834805968217,
      "grad_norm": 0.2095059780428372,
      "learning_rate": 1.202627186358436e-06,
      "loss": 0.7596,
      "step": 1640
    },
    {
      "epoch": 0.21231834805968217,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8457142857142858,
      "eval_PRM F1 AUC": 0.5544787847040336,
      "eval_PRM F1 Neg": 0.2702702702702703,
      "eval_PRM NPV": 0.35714285714285715,
      "eval_PRM Precision": 0.8043478260869565,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.626953125,
      "eval_runtime": 19.4047,
      "eval_samples_per_second": 2.267,
      "eval_steps_per_second": 0.155,
      "step": 1640
    },
    {
      "epoch": 0.21244781046703565,
      "grad_norm": 0.14401665611756415,
      "learning_rate": 1.2025192494166485e-06,
      "loss": 0.8323,
      "step": 1641
    },
    {
      "epoch": 0.2125772728743891,
      "grad_norm": 0.1914654167099316,
      "learning_rate": 1.2024111945048484e-06,
      "loss": 0.8562,
      "step": 1642
    },
    {
      "epoch": 0.21270673528174255,
      "grad_norm": 0.20437245954539726,
      "learning_rate": 1.2023030216451083e-06,
      "loss": 0.7755,
      "step": 1643
    },
    {
      "epoch": 0.21283619768909603,
      "grad_norm": 0.1921470674500222,
      "learning_rate": 1.2021947308595246e-06,
      "loss": 0.8109,
      "step": 1644
    },
    {
      "epoch": 0.2129656600964495,
      "grad_norm": 0.13516926309539465,
      "learning_rate": 1.2020863221702179e-06,
      "loss": 0.8058,
      "step": 1645
    },
    {
      "epoch": 0.2129656600964495,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8522727272727273,
      "eval_PRM F1 AUC": 0.5605028810895757,
      "eval_PRM F1 Neg": 0.2777777777777778,
      "eval_PRM NPV": 0.38461538461538464,
      "eval_PRM Precision": 0.8064516129032258,
      "eval_PRM Recall": 0.9036144578313253,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.6234019994735718,
      "eval_runtime": 20.575,
      "eval_samples_per_second": 2.139,
      "eval_steps_per_second": 0.146,
      "step": 1645
    },
    {
      "epoch": 0.21309512250380297,
      "grad_norm": 0.13291223014977083,
      "learning_rate": 1.201977795599333e-06,
      "loss": 0.9065,
      "step": 1646
    },
    {
      "epoch": 0.21322458491115642,
      "grad_norm": 0.12257918192241055,
      "learning_rate": 1.2018691511690384e-06,
      "loss": 0.8722,
      "step": 1647
    },
    {
      "epoch": 0.21335404731850988,
      "grad_norm": 0.2308924430351409,
      "learning_rate": 1.201760388901527e-06,
      "loss": 0.8982,
      "step": 1648
    },
    {
      "epoch": 0.21348350972586336,
      "grad_norm": 0.17942523211789718,
      "learning_rate": 1.2016515088190158e-06,
      "loss": 0.8846,
      "step": 1649
    },
    {
      "epoch": 0.2136129721332168,
      "grad_norm": 0.14939012663899884,
      "learning_rate": 1.201542510943746e-06,
      "loss": 0.9879,
      "step": 1650
    },
    {
      "epoch": 0.2136129721332168,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8457142857142858,
      "eval_PRM F1 AUC": 0.5544787847040336,
      "eval_PRM F1 Neg": 0.2702702702702703,
      "eval_PRM NPV": 0.35714285714285715,
      "eval_PRM Precision": 0.8043478260869565,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.6248224377632141,
      "eval_runtime": 20.5794,
      "eval_samples_per_second": 2.138,
      "eval_steps_per_second": 0.146,
      "step": 1650
    },
    {
      "epoch": 0.2137424345405703,
      "grad_norm": 0.1862565728832567,
      "learning_rate": 1.2014333952979821e-06,
      "loss": 0.8593,
      "step": 1651
    },
    {
      "epoch": 0.21387189694792375,
      "grad_norm": 0.12152028410130804,
      "learning_rate": 1.2013241619040134e-06,
      "loss": 0.8519,
      "step": 1652
    },
    {
      "epoch": 0.2140013593552772,
      "grad_norm": 0.1808115077844761,
      "learning_rate": 1.2012148107841532e-06,
      "loss": 0.8532,
      "step": 1653
    },
    {
      "epoch": 0.21413082176263068,
      "grad_norm": 0.11184599205969432,
      "learning_rate": 1.2011053419607385e-06,
      "loss": 0.825,
      "step": 1654
    },
    {
      "epoch": 0.21426028416998413,
      "grad_norm": 0.22434610530103707,
      "learning_rate": 1.2009957554561308e-06,
      "loss": 0.6998,
      "step": 1655
    },
    {
      "epoch": 0.21426028416998413,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8457142857142858,
      "eval_PRM F1 AUC": 0.5544787847040336,
      "eval_PRM F1 Neg": 0.2702702702702703,
      "eval_PRM NPV": 0.35714285714285715,
      "eval_PRM Precision": 0.8043478260869565,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.6253551244735718,
      "eval_runtime": 19.7923,
      "eval_samples_per_second": 2.223,
      "eval_steps_per_second": 0.152,
      "step": 1655
    },
    {
      "epoch": 0.21438974657733761,
      "grad_norm": 0.2508112866640438,
      "learning_rate": 1.2008860512927148e-06,
      "loss": 0.868,
      "step": 1656
    },
    {
      "epoch": 0.21451920898469107,
      "grad_norm": 0.1798881412718787,
      "learning_rate": 1.2007762294929004e-06,
      "loss": 0.95,
      "step": 1657
    },
    {
      "epoch": 0.21464867139204455,
      "grad_norm": 0.1495826044868666,
      "learning_rate": 1.2006662900791204e-06,
      "loss": 0.8508,
      "step": 1658
    },
    {
      "epoch": 0.214778133799398,
      "grad_norm": 0.21937029945787437,
      "learning_rate": 1.2005562330738326e-06,
      "loss": 0.8474,
      "step": 1659
    },
    {
      "epoch": 0.21490759620675146,
      "grad_norm": 0.10731102479485896,
      "learning_rate": 1.2004460584995184e-06,
      "loss": 0.772,
      "step": 1660
    },
    {
      "epoch": 0.21490759620675146,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8457142857142858,
      "eval_PRM F1 AUC": 0.5544787847040336,
      "eval_PRM F1 Neg": 0.2702702702702703,
      "eval_PRM NPV": 0.35714285714285715,
      "eval_PRM Precision": 0.8043478260869565,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.6246448755264282,
      "eval_runtime": 19.5746,
      "eval_samples_per_second": 2.248,
      "eval_steps_per_second": 0.153,
      "step": 1660
    },
    {
      "epoch": 0.21503705861410494,
      "grad_norm": 0.12278249698706925,
      "learning_rate": 1.2003357663786827e-06,
      "loss": 0.8357,
      "step": 1661
    },
    {
      "epoch": 0.2151665210214584,
      "grad_norm": 0.25242030792120174,
      "learning_rate": 1.2002253567338554e-06,
      "loss": 0.8511,
      "step": 1662
    },
    {
      "epoch": 0.21529598342881187,
      "grad_norm": 0.18731625607428962,
      "learning_rate": 1.2001148295875894e-06,
      "loss": 0.8112,
      "step": 1663
    },
    {
      "epoch": 0.21542544583616532,
      "grad_norm": 0.1500816866573872,
      "learning_rate": 1.2000041849624627e-06,
      "loss": 0.9492,
      "step": 1664
    },
    {
      "epoch": 0.21555490824351878,
      "grad_norm": 0.18323257143464702,
      "learning_rate": 1.1998934228810763e-06,
      "loss": 0.8246,
      "step": 1665
    },
    {
      "epoch": 0.21555490824351878,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8457142857142858,
      "eval_PRM F1 AUC": 0.5544787847040336,
      "eval_PRM F1 Neg": 0.2702702702702703,
      "eval_PRM NPV": 0.35714285714285715,
      "eval_PRM Precision": 0.8043478260869565,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.6232244372367859,
      "eval_runtime": 20.4959,
      "eval_samples_per_second": 2.147,
      "eval_steps_per_second": 0.146,
      "step": 1665
    },
    {
      "epoch": 0.21568437065087226,
      "grad_norm": 0.14279905629221976,
      "learning_rate": 1.1997825433660557e-06,
      "loss": 0.9572,
      "step": 1666
    },
    {
      "epoch": 0.2158138330582257,
      "grad_norm": 0.16025741467419216,
      "learning_rate": 1.1996715464400504e-06,
      "loss": 0.8777,
      "step": 1667
    },
    {
      "epoch": 0.2159432954655792,
      "grad_norm": 0.21440297287959514,
      "learning_rate": 1.1995604321257335e-06,
      "loss": 0.8152,
      "step": 1668
    },
    {
      "epoch": 0.21607275787293265,
      "grad_norm": 0.1542638084206063,
      "learning_rate": 1.1994492004458028e-06,
      "loss": 0.9006,
      "step": 1669
    },
    {
      "epoch": 0.2162022202802861,
      "grad_norm": 0.13872701097827905,
      "learning_rate": 1.1993378514229791e-06,
      "loss": 0.9041,
      "step": 1670
    },
    {
      "epoch": 0.2162022202802861,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8409090909090909,
      "eval_PRM F1 AUC": 0.5327396542692509,
      "eval_PRM F1 Neg": 0.2222222222222222,
      "eval_PRM NPV": 0.3076923076923077,
      "eval_PRM Precision": 0.7956989247311828,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6253551244735718,
      "eval_runtime": 19.9474,
      "eval_samples_per_second": 2.206,
      "eval_steps_per_second": 0.15,
      "step": 1670
    },
    {
      "epoch": 0.21633168268763958,
      "grad_norm": 0.1632965224768131,
      "learning_rate": 1.199226385080008e-06,
      "loss": 0.8638,
      "step": 1671
    },
    {
      "epoch": 0.21646114509499303,
      "grad_norm": 0.2826474756013327,
      "learning_rate": 1.1991148014396587e-06,
      "loss": 0.8386,
      "step": 1672
    },
    {
      "epoch": 0.21659060750234652,
      "grad_norm": 0.15415166475719874,
      "learning_rate": 1.1990031005247247e-06,
      "loss": 0.8418,
      "step": 1673
    },
    {
      "epoch": 0.21672006990969997,
      "grad_norm": 0.21833036390345384,
      "learning_rate": 1.1988912823580226e-06,
      "loss": 0.7814,
      "step": 1674
    },
    {
      "epoch": 0.21684953231705342,
      "grad_norm": 0.14128440178807034,
      "learning_rate": 1.1987793469623939e-06,
      "loss": 0.8196,
      "step": 1675
    },
    {
      "epoch": 0.21684953231705342,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8409090909090909,
      "eval_PRM F1 AUC": 0.5327396542692509,
      "eval_PRM F1 Neg": 0.2222222222222222,
      "eval_PRM NPV": 0.3076923076923077,
      "eval_PRM Precision": 0.7956989247311828,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6234019994735718,
      "eval_runtime": 20.0473,
      "eval_samples_per_second": 2.195,
      "eval_steps_per_second": 0.15,
      "step": 1675
    },
    {
      "epoch": 0.2169789947244069,
      "grad_norm": 0.14409077807884502,
      "learning_rate": 1.1986672943607037e-06,
      "loss": 0.8181,
      "step": 1676
    },
    {
      "epoch": 0.21710845713176036,
      "grad_norm": 0.27499071832332506,
      "learning_rate": 1.1985551245758408e-06,
      "loss": 0.8329,
      "step": 1677
    },
    {
      "epoch": 0.21723791953911384,
      "grad_norm": 0.16935102762160953,
      "learning_rate": 1.1984428376307184e-06,
      "loss": 0.8549,
      "step": 1678
    },
    {
      "epoch": 0.2173673819464673,
      "grad_norm": 0.10410841387239507,
      "learning_rate": 1.198330433548273e-06,
      "loss": 0.7771,
      "step": 1679
    },
    {
      "epoch": 0.21749684435382077,
      "grad_norm": 0.24479215586209097,
      "learning_rate": 1.198217912351466e-06,
      "loss": 0.7433,
      "step": 1680
    },
    {
      "epoch": 0.21749684435382077,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8457142857142858,
      "eval_PRM F1 AUC": 0.5544787847040336,
      "eval_PRM F1 Neg": 0.2702702702702703,
      "eval_PRM NPV": 0.35714285714285715,
      "eval_PRM Precision": 0.8043478260869565,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.6241121888160706,
      "eval_runtime": 20.4886,
      "eval_samples_per_second": 2.148,
      "eval_steps_per_second": 0.146,
      "step": 1680
    },
    {
      "epoch": 0.21762630676117423,
      "grad_norm": 0.164736031134444,
      "learning_rate": 1.1981052740632815e-06,
      "loss": 0.7551,
      "step": 1681
    },
    {
      "epoch": 0.21775576916852768,
      "grad_norm": 0.15051288093278598,
      "learning_rate": 1.1979925187067287e-06,
      "loss": 0.903,
      "step": 1682
    },
    {
      "epoch": 0.21788523157588116,
      "grad_norm": 0.14688326983219724,
      "learning_rate": 1.1978796463048398e-06,
      "loss": 0.8751,
      "step": 1683
    },
    {
      "epoch": 0.21801469398323461,
      "grad_norm": 0.13484116756469,
      "learning_rate": 1.1977666568806715e-06,
      "loss": 0.8292,
      "step": 1684
    },
    {
      "epoch": 0.2181441563905881,
      "grad_norm": 0.164572572582006,
      "learning_rate": 1.1976535504573038e-06,
      "loss": 0.8232,
      "step": 1685
    },
    {
      "epoch": 0.2181441563905881,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8522727272727273,
      "eval_PRM F1 AUC": 0.5605028810895757,
      "eval_PRM F1 Neg": 0.2777777777777778,
      "eval_PRM NPV": 0.38461538461538464,
      "eval_PRM Precision": 0.8064516129032258,
      "eval_PRM Recall": 0.9036144578313253,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.6265980005264282,
      "eval_runtime": 20.5037,
      "eval_samples_per_second": 2.146,
      "eval_steps_per_second": 0.146,
      "step": 1685
    },
    {
      "epoch": 0.21827361879794155,
      "grad_norm": 0.1420144393644705,
      "learning_rate": 1.1975403270578416e-06,
      "loss": 0.9579,
      "step": 1686
    },
    {
      "epoch": 0.218403081205295,
      "grad_norm": 0.12959176286340576,
      "learning_rate": 1.1974269867054124e-06,
      "loss": 0.8641,
      "step": 1687
    },
    {
      "epoch": 0.21853254361264848,
      "grad_norm": 0.1828063718717352,
      "learning_rate": 1.197313529423169e-06,
      "loss": 0.8423,
      "step": 1688
    },
    {
      "epoch": 0.21866200602000194,
      "grad_norm": 0.11448863268146442,
      "learning_rate": 1.1971999552342867e-06,
      "loss": 0.8044,
      "step": 1689
    },
    {
      "epoch": 0.21879146842735542,
      "grad_norm": 0.12367408696126257,
      "learning_rate": 1.1970862641619657e-06,
      "loss": 0.8397,
      "step": 1690
    },
    {
      "epoch": 0.21879146842735542,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.847457627118644,
      "eval_PRM F1 AUC": 0.5387637506547931,
      "eval_PRM F1 Neg": 0.22857142857142856,
      "eval_PRM NPV": 0.3333333333333333,
      "eval_PRM Precision": 0.7978723404255319,
      "eval_PRM Recall": 0.9036144578313253,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6244673132896423,
      "eval_runtime": 19.6269,
      "eval_samples_per_second": 2.242,
      "eval_steps_per_second": 0.153,
      "step": 1690
    },
    {
      "epoch": 0.21892093083470887,
      "grad_norm": 0.13924431126545844,
      "learning_rate": 1.1969724562294294e-06,
      "loss": 0.8289,
      "step": 1691
    },
    {
      "epoch": 0.21905039324206232,
      "grad_norm": 0.2037121778311325,
      "learning_rate": 1.196858531459926e-06,
      "loss": 0.8203,
      "step": 1692
    },
    {
      "epoch": 0.2191798556494158,
      "grad_norm": 0.14367440034863393,
      "learning_rate": 1.196744489876726e-06,
      "loss": 0.928,
      "step": 1693
    },
    {
      "epoch": 0.21930931805676926,
      "grad_norm": 0.16046475795142845,
      "learning_rate": 1.1966303315031254e-06,
      "loss": 0.9891,
      "step": 1694
    },
    {
      "epoch": 0.21943878046412274,
      "grad_norm": 0.14527017345819146,
      "learning_rate": 1.1965160563624432e-06,
      "loss": 0.9127,
      "step": 1695
    },
    {
      "epoch": 0.21943878046412274,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.847457627118644,
      "eval_PRM F1 AUC": 0.5387637506547931,
      "eval_PRM F1 Neg": 0.22857142857142856,
      "eval_PRM NPV": 0.3333333333333333,
      "eval_PRM Precision": 0.7978723404255319,
      "eval_PRM Recall": 0.9036144578313253,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.625,
      "eval_runtime": 19.9273,
      "eval_samples_per_second": 2.208,
      "eval_steps_per_second": 0.151,
      "step": 1695
    },
    {
      "epoch": 0.2195682428714762,
      "grad_norm": 0.33740388214780936,
      "learning_rate": 1.1964016644780222e-06,
      "loss": 0.7666,
      "step": 1696
    },
    {
      "epoch": 0.21969770527882965,
      "grad_norm": 0.1368367002980863,
      "learning_rate": 1.1962871558732297e-06,
      "loss": 0.988,
      "step": 1697
    },
    {
      "epoch": 0.21982716768618313,
      "grad_norm": 0.16949220764596365,
      "learning_rate": 1.196172530571456e-06,
      "loss": 0.8641,
      "step": 1698
    },
    {
      "epoch": 0.21995663009353658,
      "grad_norm": 0.11737498642837646,
      "learning_rate": 1.1960577885961158e-06,
      "loss": 0.824,
      "step": 1699
    },
    {
      "epoch": 0.22008609250089006,
      "grad_norm": 0.1481792538915297,
      "learning_rate": 1.1959429299706477e-06,
      "loss": 0.8198,
      "step": 1700
    },
    {
      "epoch": 0.22008609250089006,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8457142857142858,
      "eval_PRM F1 AUC": 0.5544787847040336,
      "eval_PRM F1 Neg": 0.2702702702702703,
      "eval_PRM NPV": 0.35714285714285715,
      "eval_PRM Precision": 0.8043478260869565,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.6278409361839294,
      "eval_runtime": 19.996,
      "eval_samples_per_second": 2.2,
      "eval_steps_per_second": 0.15,
      "step": 1700
    },
    {
      "epoch": 0.22021555490824352,
      "grad_norm": 0.1517566462425552,
      "learning_rate": 1.1958279547185133e-06,
      "loss": 0.8871,
      "step": 1701
    },
    {
      "epoch": 0.220345017315597,
      "grad_norm": 0.24706193971049054,
      "learning_rate": 1.1957128628631991e-06,
      "loss": 0.772,
      "step": 1702
    },
    {
      "epoch": 0.22047447972295045,
      "grad_norm": 0.11146643411150035,
      "learning_rate": 1.195597654428215e-06,
      "loss": 0.8358,
      "step": 1703
    },
    {
      "epoch": 0.2206039421303039,
      "grad_norm": 0.2357967318405864,
      "learning_rate": 1.1954823294370944e-06,
      "loss": 0.8434,
      "step": 1704
    },
    {
      "epoch": 0.22073340453765739,
      "grad_norm": 0.12906258640709964,
      "learning_rate": 1.1953668879133949e-06,
      "loss": 0.8978,
      "step": 1705
    },
    {
      "epoch": 0.22073340453765739,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8522727272727273,
      "eval_PRM F1 AUC": 0.5605028810895757,
      "eval_PRM F1 Neg": 0.2777777777777778,
      "eval_PRM NPV": 0.38461538461538464,
      "eval_PRM Precision": 0.8064516129032258,
      "eval_PRM Recall": 0.9036144578313253,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.6251775622367859,
      "eval_runtime": 20.2968,
      "eval_samples_per_second": 2.168,
      "eval_steps_per_second": 0.148,
      "step": 1705
    },
    {
      "epoch": 0.22086286694501084,
      "grad_norm": 0.15325436137732495,
      "learning_rate": 1.1952513298806976e-06,
      "loss": 1.001,
      "step": 1706
    },
    {
      "epoch": 0.22099232935236432,
      "grad_norm": 0.1363836911355042,
      "learning_rate": 1.1951356553626079e-06,
      "loss": 0.9563,
      "step": 1707
    },
    {
      "epoch": 0.22112179175971777,
      "grad_norm": 0.13330227511963907,
      "learning_rate": 1.1950198643827543e-06,
      "loss": 0.8975,
      "step": 1708
    },
    {
      "epoch": 0.22125125416707123,
      "grad_norm": 0.15537782313082532,
      "learning_rate": 1.1949039569647897e-06,
      "loss": 0.795,
      "step": 1709
    },
    {
      "epoch": 0.2213807165744247,
      "grad_norm": 0.14935694125276952,
      "learning_rate": 1.1947879331323905e-06,
      "loss": 0.8396,
      "step": 1710
    },
    {
      "epoch": 0.2213807165744247,
      "eval_PRM Accuracy": 0.7641509433962265,
      "eval_PRM F1": 0.8587570621468926,
      "eval_PRM F1 AUC": 0.5665269774751178,
      "eval_PRM F1 Neg": 0.2857142857142857,
      "eval_PRM NPV": 0.4166666666666667,
      "eval_PRM Precision": 0.8085106382978723,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.623046875,
      "eval_runtime": 20.3325,
      "eval_samples_per_second": 2.164,
      "eval_steps_per_second": 0.148,
      "step": 1710
    },
    {
      "epoch": 0.22151017898177816,
      "grad_norm": 0.2364117935885425,
      "learning_rate": 1.194671792909257e-06,
      "loss": 0.8497,
      "step": 1711
    },
    {
      "epoch": 0.22163964138913164,
      "grad_norm": 0.1293844038959051,
      "learning_rate": 1.1945555363191127e-06,
      "loss": 0.8481,
      "step": 1712
    },
    {
      "epoch": 0.2217691037964851,
      "grad_norm": 0.12234719232355831,
      "learning_rate": 1.194439163385706e-06,
      "loss": 0.8597,
      "step": 1713
    },
    {
      "epoch": 0.22189856620383855,
      "grad_norm": 0.16220760531786516,
      "learning_rate": 1.1943226741328082e-06,
      "loss": 1.0197,
      "step": 1714
    },
    {
      "epoch": 0.22202802861119203,
      "grad_norm": 0.23123262992725765,
      "learning_rate": 1.1942060685842144e-06,
      "loss": 0.8405,
      "step": 1715
    },
    {
      "epoch": 0.22202802861119203,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8457142857142858,
      "eval_PRM F1 AUC": 0.5544787847040336,
      "eval_PRM F1 Neg": 0.2702702702702703,
      "eval_PRM NPV": 0.35714285714285715,
      "eval_PRM Precision": 0.8043478260869565,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.6244673132896423,
      "eval_runtime": 19.9614,
      "eval_samples_per_second": 2.204,
      "eval_steps_per_second": 0.15,
      "step": 1715
    },
    {
      "epoch": 0.22215749101854548,
      "grad_norm": 0.13786283463520654,
      "learning_rate": 1.194089346763744e-06,
      "loss": 0.8387,
      "step": 1716
    },
    {
      "epoch": 0.22228695342589896,
      "grad_norm": 0.2911777670952419,
      "learning_rate": 1.1939725086952395e-06,
      "loss": 0.6647,
      "step": 1717
    },
    {
      "epoch": 0.22241641583325242,
      "grad_norm": 0.2789160348878563,
      "learning_rate": 1.1938555544025675e-06,
      "loss": 0.7331,
      "step": 1718
    },
    {
      "epoch": 0.22254587824060587,
      "grad_norm": 0.11622751450140378,
      "learning_rate": 1.1937384839096184e-06,
      "loss": 0.8601,
      "step": 1719
    },
    {
      "epoch": 0.22267534064795935,
      "grad_norm": 0.18485767915318838,
      "learning_rate": 1.1936212972403062e-06,
      "loss": 0.8207,
      "step": 1720
    },
    {
      "epoch": 0.22267534064795935,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8522727272727273,
      "eval_PRM F1 AUC": 0.5605028810895757,
      "eval_PRM F1 Neg": 0.2777777777777778,
      "eval_PRM NPV": 0.38461538461538464,
      "eval_PRM Precision": 0.8064516129032258,
      "eval_PRM Recall": 0.9036144578313253,
      "eval_PRM Specificty": 0.21739130434782608,
      "eval_loss": 0.6234019994735718,
      "eval_runtime": 19.9955,
      "eval_samples_per_second": 2.2,
      "eval_steps_per_second": 0.15,
      "step": 1720
    },
    {
      "epoch": 0.2228048030553128,
      "grad_norm": 0.15071602505248993,
      "learning_rate": 1.1935039944185685e-06,
      "loss": 0.8666,
      "step": 1721
    },
    {
      "epoch": 0.2229342654626663,
      "grad_norm": 0.19862425621084578,
      "learning_rate": 1.193386575468367e-06,
      "loss": 0.7986,
      "step": 1722
    },
    {
      "epoch": 0.22306372787001974,
      "grad_norm": 0.11246223062524208,
      "learning_rate": 1.1932690404136869e-06,
      "loss": 0.8478,
      "step": 1723
    },
    {
      "epoch": 0.22319319027737322,
      "grad_norm": 0.11624361428035429,
      "learning_rate": 1.1931513892785366e-06,
      "loss": 0.7629,
      "step": 1724
    },
    {
      "epoch": 0.22332265268472667,
      "grad_norm": 0.12695460066536254,
      "learning_rate": 1.1930336220869495e-06,
      "loss": 0.7853,
      "step": 1725
    },
    {
      "epoch": 0.22332265268472667,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.847457627118644,
      "eval_PRM F1 AUC": 0.5387637506547931,
      "eval_PRM F1 Neg": 0.22857142857142856,
      "eval_PRM NPV": 0.3333333333333333,
      "eval_PRM Precision": 0.7978723404255319,
      "eval_PRM Recall": 0.9036144578313253,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6260653138160706,
      "eval_runtime": 19.7995,
      "eval_samples_per_second": 2.222,
      "eval_steps_per_second": 0.152,
      "step": 1725
    },
    {
      "epoch": 0.22345211509208013,
      "grad_norm": 0.11960642488928618,
      "learning_rate": 1.1929157388629815e-06,
      "loss": 0.8911,
      "step": 1726
    },
    {
      "epoch": 0.2235815774994336,
      "grad_norm": 0.11898548936540317,
      "learning_rate": 1.1927977396307126e-06,
      "loss": 0.8701,
      "step": 1727
    },
    {
      "epoch": 0.22371103990678706,
      "grad_norm": 0.21225890812949744,
      "learning_rate": 1.1926796244142466e-06,
      "loss": 0.7762,
      "step": 1728
    },
    {
      "epoch": 0.22384050231414054,
      "grad_norm": 0.19942902511098987,
      "learning_rate": 1.1925613932377112e-06,
      "loss": 0.7502,
      "step": 1729
    },
    {
      "epoch": 0.223969964721494,
      "grad_norm": 0.15863478020397725,
      "learning_rate": 1.1924430461252572e-06,
      "loss": 0.8478,
      "step": 1730
    },
    {
      "epoch": 0.223969964721494,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8409090909090909,
      "eval_PRM F1 AUC": 0.5327396542692509,
      "eval_PRM F1 Neg": 0.2222222222222222,
      "eval_PRM NPV": 0.3076923076923077,
      "eval_PRM Precision": 0.7956989247311828,
      "eval_PRM Recall": 0.891566265060241,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6267755627632141,
      "eval_runtime": 19.7041,
      "eval_samples_per_second": 2.233,
      "eval_steps_per_second": 0.152,
      "step": 1730
    },
    {
      "epoch": 0.22409942712884745,
      "grad_norm": 0.1275234533581469,
      "learning_rate": 1.1923245831010594e-06,
      "loss": 0.8967,
      "step": 1731
    },
    {
      "epoch": 0.22422888953620093,
      "grad_norm": 0.15637258403487483,
      "learning_rate": 1.1922060041893163e-06,
      "loss": 0.7505,
      "step": 1732
    },
    {
      "epoch": 0.22435835194355438,
      "grad_norm": 0.2080380012813621,
      "learning_rate": 1.1920873094142504e-06,
      "loss": 0.7789,
      "step": 1733
    },
    {
      "epoch": 0.22448781435090787,
      "grad_norm": 0.19248997176905078,
      "learning_rate": 1.191968498800107e-06,
      "loss": 0.9619,
      "step": 1734
    },
    {
      "epoch": 0.22461727675826132,
      "grad_norm": 0.11380333197359294,
      "learning_rate": 1.1918495723711557e-06,
      "loss": 0.7661,
      "step": 1735
    },
    {
      "epoch": 0.22461727675826132,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.847457627118644,
      "eval_PRM F1 AUC": 0.5387637506547931,
      "eval_PRM F1 Neg": 0.22857142857142856,
      "eval_PRM NPV": 0.3333333333333333,
      "eval_PRM Precision": 0.7978723404255319,
      "eval_PRM Recall": 0.9036144578313253,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6262428760528564,
      "eval_runtime": 20.4309,
      "eval_samples_per_second": 2.154,
      "eval_steps_per_second": 0.147,
      "step": 1735
    },
    {
      "epoch": 0.22474673916561477,
      "grad_norm": 0.2452206456040745,
      "learning_rate": 1.1917305301516898e-06,
      "loss": 0.7737,
      "step": 1736
    },
    {
      "epoch": 0.22487620157296825,
      "grad_norm": 0.16282027374936228,
      "learning_rate": 1.191611372166026e-06,
      "loss": 0.8047,
      "step": 1737
    },
    {
      "epoch": 0.2250056639803217,
      "grad_norm": 0.1366226534503337,
      "learning_rate": 1.191492098438505e-06,
      "loss": 0.8229,
      "step": 1738
    },
    {
      "epoch": 0.2251351263876752,
      "grad_norm": 0.14124776251815885,
      "learning_rate": 1.1913727089934905e-06,
      "loss": 0.9152,
      "step": 1739
    },
    {
      "epoch": 0.22526458879502864,
      "grad_norm": 0.11766385657008495,
      "learning_rate": 1.19125320385537e-06,
      "loss": 0.8478,
      "step": 1740
    },
    {
      "epoch": 0.22526458879502864,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.847457627118644,
      "eval_PRM F1 AUC": 0.5387637506547931,
      "eval_PRM F1 Neg": 0.22857142857142856,
      "eval_PRM NPV": 0.3333333333333333,
      "eval_PRM Precision": 0.7978723404255319,
      "eval_PRM Recall": 0.9036144578313253,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6248224377632141,
      "eval_runtime": 19.8971,
      "eval_samples_per_second": 2.211,
      "eval_steps_per_second": 0.151,
      "step": 1740
    },
    {
      "epoch": 0.2253940512023821,
      "grad_norm": 0.1472799117470293,
      "learning_rate": 1.1911335830485556e-06,
      "loss": 0.9547,
      "step": 1741
    },
    {
      "epoch": 0.22552351360973558,
      "grad_norm": 0.17286513023787456,
      "learning_rate": 1.1910138465974815e-06,
      "loss": 0.8009,
      "step": 1742
    },
    {
      "epoch": 0.22565297601708903,
      "grad_norm": 0.16087009994463852,
      "learning_rate": 1.190893994526607e-06,
      "loss": 0.7745,
      "step": 1743
    },
    {
      "epoch": 0.2257824384244425,
      "grad_norm": 0.1719381301226881,
      "learning_rate": 1.1907740268604139e-06,
      "loss": 0.7974,
      "step": 1744
    },
    {
      "epoch": 0.22591190083179596,
      "grad_norm": 0.14111506493994722,
      "learning_rate": 1.190653943623408e-06,
      "loss": 0.9188,
      "step": 1745
    },
    {
      "epoch": 0.22591190083179596,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.847457627118644,
      "eval_PRM F1 AUC": 0.5387637506547931,
      "eval_PRM F1 Neg": 0.22857142857142856,
      "eval_PRM NPV": 0.3333333333333333,
      "eval_PRM Precision": 0.7978723404255319,
      "eval_PRM Recall": 0.9036144578313253,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6253551244735718,
      "eval_runtime": 19.9471,
      "eval_samples_per_second": 2.206,
      "eval_steps_per_second": 0.15,
      "step": 1745
    },
    {
      "epoch": 0.22604136323914945,
      "grad_norm": 0.13821163695640404,
      "learning_rate": 1.190533744840119e-06,
      "loss": 0.8809,
      "step": 1746
    },
    {
      "epoch": 0.2261708256465029,
      "grad_norm": 0.16963104588890457,
      "learning_rate": 1.1904134305350996e-06,
      "loss": 0.8494,
      "step": 1747
    },
    {
      "epoch": 0.22630028805385635,
      "grad_norm": 0.13321366826458353,
      "learning_rate": 1.1902930007329265e-06,
      "loss": 0.8107,
      "step": 1748
    },
    {
      "epoch": 0.22642975046120983,
      "grad_norm": 0.18367944989577217,
      "learning_rate": 1.1901724554582001e-06,
      "loss": 0.7952,
      "step": 1749
    },
    {
      "epoch": 0.2265592128685633,
      "grad_norm": 0.1261922477096692,
      "learning_rate": 1.1900517947355442e-06,
      "loss": 0.8062,
      "step": 1750
    },
    {
      "epoch": 0.2265592128685633,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8539325842696629,
      "eval_PRM F1 AUC": 0.5447878470403352,
      "eval_PRM F1 Neg": 0.23529411764705882,
      "eval_PRM NPV": 0.36363636363636365,
      "eval_PRM Precision": 0.8,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6248224377632141,
      "eval_runtime": 20.3905,
      "eval_samples_per_second": 2.158,
      "eval_steps_per_second": 0.147,
      "step": 1750
    },
    {
      "epoch": 0.22668867527591677,
      "grad_norm": 0.16426536158141405,
      "learning_rate": 1.1899310185896062e-06,
      "loss": 0.865,
      "step": 1751
    },
    {
      "epoch": 0.22681813768327022,
      "grad_norm": 0.3101477480392658,
      "learning_rate": 1.1898101270450569e-06,
      "loss": 0.7104,
      "step": 1752
    },
    {
      "epoch": 0.22694760009062367,
      "grad_norm": 0.17937816623785363,
      "learning_rate": 1.1896891201265911e-06,
      "loss": 0.7563,
      "step": 1753
    },
    {
      "epoch": 0.22707706249797716,
      "grad_norm": 0.12808462120402053,
      "learning_rate": 1.1895679978589266e-06,
      "loss": 0.9115,
      "step": 1754
    },
    {
      "epoch": 0.2272065249053306,
      "grad_norm": 0.21925166773462007,
      "learning_rate": 1.1894467602668056e-06,
      "loss": 0.8105,
      "step": 1755
    },
    {
      "epoch": 0.2272065249053306,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8539325842696629,
      "eval_PRM F1 AUC": 0.5447878470403352,
      "eval_PRM F1 Neg": 0.23529411764705882,
      "eval_PRM NPV": 0.36363636363636365,
      "eval_PRM Precision": 0.8,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6271306872367859,
      "eval_runtime": 19.9777,
      "eval_samples_per_second": 2.202,
      "eval_steps_per_second": 0.15,
      "step": 1755
    },
    {
      "epoch": 0.2273359873126841,
      "grad_norm": 0.12773139224955699,
      "learning_rate": 1.1893254073749926e-06,
      "loss": 0.9425,
      "step": 1756
    },
    {
      "epoch": 0.22746544972003754,
      "grad_norm": 0.13432159207580402,
      "learning_rate": 1.1892039392082771e-06,
      "loss": 0.8628,
      "step": 1757
    },
    {
      "epoch": 0.227594912127391,
      "grad_norm": 0.26082923261861224,
      "learning_rate": 1.189082355791471e-06,
      "loss": 0.6685,
      "step": 1758
    },
    {
      "epoch": 0.22772437453474448,
      "grad_norm": 0.14911836887335536,
      "learning_rate": 1.1889606571494103e-06,
      "loss": 0.924,
      "step": 1759
    },
    {
      "epoch": 0.22785383694209793,
      "grad_norm": 0.13910765171115858,
      "learning_rate": 1.1888388433069545e-06,
      "loss": 0.9263,
      "step": 1760
    },
    {
      "epoch": 0.22785383694209793,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8539325842696629,
      "eval_PRM F1 AUC": 0.5447878470403352,
      "eval_PRM F1 Neg": 0.23529411764705882,
      "eval_PRM NPV": 0.36363636363636365,
      "eval_PRM Precision": 0.8,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6246448755264282,
      "eval_runtime": 20.3186,
      "eval_samples_per_second": 2.166,
      "eval_steps_per_second": 0.148,
      "step": 1760
    },
    {
      "epoch": 0.2279832993494514,
      "grad_norm": 0.21151097441312383,
      "learning_rate": 1.1887169142889866e-06,
      "loss": 0.8148,
      "step": 1761
    },
    {
      "epoch": 0.22811276175680487,
      "grad_norm": 0.25241606431079533,
      "learning_rate": 1.1885948701204125e-06,
      "loss": 0.875,
      "step": 1762
    },
    {
      "epoch": 0.22824222416415832,
      "grad_norm": 0.18327299128800523,
      "learning_rate": 1.188472710826163e-06,
      "loss": 0.8688,
      "step": 1763
    },
    {
      "epoch": 0.2283716865715118,
      "grad_norm": 0.13500759656672187,
      "learning_rate": 1.188350436431191e-06,
      "loss": 0.8102,
      "step": 1764
    },
    {
      "epoch": 0.22850114897886525,
      "grad_norm": 0.15903535672223595,
      "learning_rate": 1.188228046960474e-06,
      "loss": 0.8608,
      "step": 1765
    },
    {
      "epoch": 0.22850114897886525,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8539325842696629,
      "eval_PRM F1 AUC": 0.5447878470403352,
      "eval_PRM F1 Neg": 0.23529411764705882,
      "eval_PRM NPV": 0.36363636363636365,
      "eval_PRM Precision": 0.8,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6271306872367859,
      "eval_runtime": 20.3822,
      "eval_samples_per_second": 2.159,
      "eval_steps_per_second": 0.147,
      "step": 1765
    },
    {
      "epoch": 0.22863061138621874,
      "grad_norm": 0.2028864386196964,
      "learning_rate": 1.188105542439012e-06,
      "loss": 0.765,
      "step": 1766
    },
    {
      "epoch": 0.2287600737935722,
      "grad_norm": 0.16092025977352375,
      "learning_rate": 1.1879829228918295e-06,
      "loss": 0.7293,
      "step": 1767
    },
    {
      "epoch": 0.22888953620092567,
      "grad_norm": 0.17640450406985428,
      "learning_rate": 1.1878601883439738e-06,
      "loss": 0.7433,
      "step": 1768
    },
    {
      "epoch": 0.22901899860827912,
      "grad_norm": 0.11738087226597035,
      "learning_rate": 1.1877373388205162e-06,
      "loss": 0.9241,
      "step": 1769
    },
    {
      "epoch": 0.22914846101563258,
      "grad_norm": 0.1834339110741731,
      "learning_rate": 1.1876143743465507e-06,
      "loss": 0.7742,
      "step": 1770
    },
    {
      "epoch": 0.22914846101563258,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8539325842696629,
      "eval_PRM F1 AUC": 0.5447878470403352,
      "eval_PRM F1 Neg": 0.23529411764705882,
      "eval_PRM NPV": 0.36363636363636365,
      "eval_PRM Precision": 0.8,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6257102489471436,
      "eval_runtime": 19.9729,
      "eval_samples_per_second": 2.203,
      "eval_steps_per_second": 0.15,
      "step": 1770
    },
    {
      "epoch": 0.22927792342298606,
      "grad_norm": 0.15654355700586792,
      "learning_rate": 1.1874912949471957e-06,
      "loss": 0.7645,
      "step": 1771
    },
    {
      "epoch": 0.2294073858303395,
      "grad_norm": 0.14303435850685806,
      "learning_rate": 1.1873681006475924e-06,
      "loss": 0.8237,
      "step": 1772
    },
    {
      "epoch": 0.229536848237693,
      "grad_norm": 0.18792105540430876,
      "learning_rate": 1.187244791472906e-06,
      "loss": 0.7994,
      "step": 1773
    },
    {
      "epoch": 0.22966631064504645,
      "grad_norm": 0.13564823286211622,
      "learning_rate": 1.1871213674483247e-06,
      "loss": 0.952,
      "step": 1774
    },
    {
      "epoch": 0.2297957730523999,
      "grad_norm": 0.14064075662402373,
      "learning_rate": 1.1869978285990604e-06,
      "loss": 0.8861,
      "step": 1775
    },
    {
      "epoch": 0.2297957730523999,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8539325842696629,
      "eval_PRM F1 AUC": 0.5447878470403352,
      "eval_PRM F1 Neg": 0.23529411764705882,
      "eval_PRM NPV": 0.36363636363636365,
      "eval_PRM Precision": 0.8,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6283735632896423,
      "eval_runtime": 19.9298,
      "eval_samples_per_second": 2.208,
      "eval_steps_per_second": 0.151,
      "step": 1775
    },
    {
      "epoch": 0.22992523545975338,
      "grad_norm": 0.18478124629474976,
      "learning_rate": 1.1868741749503485e-06,
      "loss": 0.8305,
      "step": 1776
    },
    {
      "epoch": 0.23005469786710683,
      "grad_norm": 0.16865925728245776,
      "learning_rate": 1.1867504065274477e-06,
      "loss": 0.8317,
      "step": 1777
    },
    {
      "epoch": 0.23018416027446031,
      "grad_norm": 0.10788789562624643,
      "learning_rate": 1.18662652335564e-06,
      "loss": 0.7626,
      "step": 1778
    },
    {
      "epoch": 0.23031362268181377,
      "grad_norm": 0.12336278931809924,
      "learning_rate": 1.1865025254602317e-06,
      "loss": 0.8617,
      "step": 1779
    },
    {
      "epoch": 0.23044308508916722,
      "grad_norm": 0.24277335238214004,
      "learning_rate": 1.1863784128665512e-06,
      "loss": 0.768,
      "step": 1780
    },
    {
      "epoch": 0.23044308508916722,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.847457627118644,
      "eval_PRM F1 AUC": 0.5387637506547931,
      "eval_PRM F1 Neg": 0.22857142857142856,
      "eval_PRM NPV": 0.3333333333333333,
      "eval_PRM Precision": 0.7978723404255319,
      "eval_PRM Recall": 0.9036144578313253,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6260653138160706,
      "eval_runtime": 19.3493,
      "eval_samples_per_second": 2.274,
      "eval_steps_per_second": 0.155,
      "step": 1780
    },
    {
      "epoch": 0.2305725474965207,
      "grad_norm": 0.1573619877301961,
      "learning_rate": 1.186254185599951e-06,
      "loss": 0.8468,
      "step": 1781
    },
    {
      "epoch": 0.23070200990387416,
      "grad_norm": 0.2390429300648234,
      "learning_rate": 1.1861298436858075e-06,
      "loss": 0.8251,
      "step": 1782
    },
    {
      "epoch": 0.23083147231122764,
      "grad_norm": 0.1608955765379953,
      "learning_rate": 1.1860053871495199e-06,
      "loss": 0.8904,
      "step": 1783
    },
    {
      "epoch": 0.2309609347185811,
      "grad_norm": 0.10752615400957437,
      "learning_rate": 1.1858808160165108e-06,
      "loss": 0.8042,
      "step": 1784
    },
    {
      "epoch": 0.23109039712593454,
      "grad_norm": 0.16067744543997137,
      "learning_rate": 1.1857561303122266e-06,
      "loss": 0.9741,
      "step": 1785
    },
    {
      "epoch": 0.23109039712593454,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8539325842696629,
      "eval_PRM F1 AUC": 0.5447878470403352,
      "eval_PRM F1 Neg": 0.23529411764705882,
      "eval_PRM NPV": 0.36363636363636365,
      "eval_PRM Precision": 0.8,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6287286877632141,
      "eval_runtime": 19.532,
      "eval_samples_per_second": 2.253,
      "eval_steps_per_second": 0.154,
      "step": 1785
    },
    {
      "epoch": 0.23121985953328802,
      "grad_norm": 0.1588514480756209,
      "learning_rate": 1.1856313300621367e-06,
      "loss": 0.7771,
      "step": 1786
    },
    {
      "epoch": 0.23134932194064148,
      "grad_norm": 0.13695531226003063,
      "learning_rate": 1.1855064152917342e-06,
      "loss": 0.843,
      "step": 1787
    },
    {
      "epoch": 0.23147878434799496,
      "grad_norm": 0.23576559825130003,
      "learning_rate": 1.1853813860265354e-06,
      "loss": 0.7286,
      "step": 1788
    },
    {
      "epoch": 0.2316082467553484,
      "grad_norm": 0.189076902532477,
      "learning_rate": 1.18525624229208e-06,
      "loss": 0.8159,
      "step": 1789
    },
    {
      "epoch": 0.2317377091627019,
      "grad_norm": 0.16255828624174729,
      "learning_rate": 1.1851309841139313e-06,
      "loss": 0.8466,
      "step": 1790
    },
    {
      "epoch": 0.2317377091627019,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8539325842696629,
      "eval_PRM F1 AUC": 0.5447878470403352,
      "eval_PRM F1 Neg": 0.23529411764705882,
      "eval_PRM NPV": 0.36363636363636365,
      "eval_PRM Precision": 0.8,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6285511255264282,
      "eval_runtime": 19.9589,
      "eval_samples_per_second": 2.205,
      "eval_steps_per_second": 0.15,
      "step": 1790
    },
    {
      "epoch": 0.23186717157005535,
      "grad_norm": 0.1851588432926368,
      "learning_rate": 1.1850056115176758e-06,
      "loss": 0.7753,
      "step": 1791
    },
    {
      "epoch": 0.2319966339774088,
      "grad_norm": 0.19681409370502465,
      "learning_rate": 1.1848801245289236e-06,
      "loss": 0.7872,
      "step": 1792
    },
    {
      "epoch": 0.23212609638476228,
      "grad_norm": 0.24060857455084902,
      "learning_rate": 1.1847545231733071e-06,
      "loss": 0.7664,
      "step": 1793
    },
    {
      "epoch": 0.23225555879211573,
      "grad_norm": 0.2578686442224168,
      "learning_rate": 1.184628807476484e-06,
      "loss": 0.7772,
      "step": 1794
    },
    {
      "epoch": 0.23238502119946922,
      "grad_norm": 0.1242718676691956,
      "learning_rate": 1.1845029774641338e-06,
      "loss": 0.8136,
      "step": 1795
    },
    {
      "epoch": 0.23238502119946922,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8539325842696629,
      "eval_PRM F1 AUC": 0.5447878470403352,
      "eval_PRM F1 Neg": 0.23529411764705882,
      "eval_PRM NPV": 0.36363636363636365,
      "eval_PRM Precision": 0.8,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6292613744735718,
      "eval_runtime": 20.4033,
      "eval_samples_per_second": 2.157,
      "eval_steps_per_second": 0.147,
      "step": 1795
    },
    {
      "epoch": 0.23251448360682267,
      "grad_norm": 0.14151360950463232,
      "learning_rate": 1.1843770331619597e-06,
      "loss": 0.8998,
      "step": 1796
    },
    {
      "epoch": 0.23264394601417612,
      "grad_norm": 0.22669112615492323,
      "learning_rate": 1.1842509745956885e-06,
      "loss": 0.7439,
      "step": 1797
    },
    {
      "epoch": 0.2327734084215296,
      "grad_norm": 0.16093081292976,
      "learning_rate": 1.1841248017910702e-06,
      "loss": 0.7324,
      "step": 1798
    },
    {
      "epoch": 0.23290287082888306,
      "grad_norm": 0.1651526645288737,
      "learning_rate": 1.1839985147738784e-06,
      "loss": 0.8291,
      "step": 1799
    },
    {
      "epoch": 0.23303233323623654,
      "grad_norm": 0.13132052566037938,
      "learning_rate": 1.1838721135699094e-06,
      "loss": 0.8022,
      "step": 1800
    },
    {
      "epoch": 0.23303233323623654,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8539325842696629,
      "eval_PRM F1 AUC": 0.5447878470403352,
      "eval_PRM F1 Neg": 0.23529411764705882,
      "eval_PRM NPV": 0.36363636363636365,
      "eval_PRM Precision": 0.8,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6294389367103577,
      "eval_runtime": 20.2638,
      "eval_samples_per_second": 2.171,
      "eval_steps_per_second": 0.148,
      "step": 1800
    },
    {
      "epoch": 0.23316179564359,
      "grad_norm": 0.14107680017603527,
      "learning_rate": 1.1837455982049835e-06,
      "loss": 0.8627,
      "step": 1801
    },
    {
      "epoch": 0.23329125805094345,
      "grad_norm": 0.26659937212700924,
      "learning_rate": 1.183618968704944e-06,
      "loss": 0.7881,
      "step": 1802
    },
    {
      "epoch": 0.23342072045829693,
      "grad_norm": 0.28101516016343514,
      "learning_rate": 1.1834922250956572e-06,
      "loss": 0.7598,
      "step": 1803
    },
    {
      "epoch": 0.23355018286565038,
      "grad_norm": 0.18202143653646077,
      "learning_rate": 1.1833653674030134e-06,
      "loss": 0.9363,
      "step": 1804
    },
    {
      "epoch": 0.23367964527300386,
      "grad_norm": 0.15839459003478543,
      "learning_rate": 1.1832383956529258e-06,
      "loss": 0.7965,
      "step": 1805
    },
    {
      "epoch": 0.23367964527300386,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8539325842696629,
      "eval_PRM F1 AUC": 0.5447878470403352,
      "eval_PRM F1 Neg": 0.23529411764705882,
      "eval_PRM NPV": 0.36363636363636365,
      "eval_PRM Precision": 0.8,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6278409361839294,
      "eval_runtime": 20.7618,
      "eval_samples_per_second": 2.119,
      "eval_steps_per_second": 0.144,
      "step": 1805
    },
    {
      "epoch": 0.23380910768035731,
      "grad_norm": 0.2509483336923381,
      "learning_rate": 1.1831113098713307e-06,
      "loss": 0.7484,
      "step": 1806
    },
    {
      "epoch": 0.23393857008771077,
      "grad_norm": 0.1251604486837487,
      "learning_rate": 1.1829841100841884e-06,
      "loss": 0.8302,
      "step": 1807
    },
    {
      "epoch": 0.23406803249506425,
      "grad_norm": 0.12232554131209387,
      "learning_rate": 1.1828567963174816e-06,
      "loss": 0.8544,
      "step": 1808
    },
    {
      "epoch": 0.2341974949024177,
      "grad_norm": 0.22773923772503424,
      "learning_rate": 1.1827293685972172e-06,
      "loss": 0.6698,
      "step": 1809
    },
    {
      "epoch": 0.23432695730977118,
      "grad_norm": 0.11833420605256383,
      "learning_rate": 1.1826018269494244e-06,
      "loss": 0.7906,
      "step": 1810
    },
    {
      "epoch": 0.23432695730977118,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8539325842696629,
      "eval_PRM F1 AUC": 0.5447878470403352,
      "eval_PRM F1 Neg": 0.23529411764705882,
      "eval_PRM NPV": 0.36363636363636365,
      "eval_PRM Precision": 0.8,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6294389367103577,
      "eval_runtime": 20.2237,
      "eval_samples_per_second": 2.176,
      "eval_steps_per_second": 0.148,
      "step": 1810
    },
    {
      "epoch": 0.23445641971712464,
      "grad_norm": 0.12772543860403307,
      "learning_rate": 1.1824741714001565e-06,
      "loss": 0.7994,
      "step": 1811
    },
    {
      "epoch": 0.23458588212447812,
      "grad_norm": 0.1416951088510125,
      "learning_rate": 1.1823464019754895e-06,
      "loss": 0.7587,
      "step": 1812
    },
    {
      "epoch": 0.23471534453183157,
      "grad_norm": 0.13214575223552885,
      "learning_rate": 1.182218518701523e-06,
      "loss": 0.85,
      "step": 1813
    },
    {
      "epoch": 0.23484480693918502,
      "grad_norm": 0.13088030040635665,
      "learning_rate": 1.1820905216043797e-06,
      "loss": 0.879,
      "step": 1814
    },
    {
      "epoch": 0.2349742693465385,
      "grad_norm": 0.19879970798830487,
      "learning_rate": 1.1819624107102057e-06,
      "loss": 0.8173,
      "step": 1815
    },
    {
      "epoch": 0.2349742693465385,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8539325842696629,
      "eval_PRM F1 AUC": 0.5447878470403352,
      "eval_PRM F1 Neg": 0.23529411764705882,
      "eval_PRM NPV": 0.36363636363636365,
      "eval_PRM Precision": 0.8,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6280184388160706,
      "eval_runtime": 20.1916,
      "eval_samples_per_second": 2.179,
      "eval_steps_per_second": 0.149,
      "step": 1815
    },
    {
      "epoch": 0.23510373175389196,
      "grad_norm": 0.17196504483505806,
      "learning_rate": 1.1818341860451703e-06,
      "loss": 0.7467,
      "step": 1816
    },
    {
      "epoch": 0.23523319416124544,
      "grad_norm": 0.1259184647125946,
      "learning_rate": 1.1817058476354657e-06,
      "loss": 0.9033,
      "step": 1817
    },
    {
      "epoch": 0.2353626565685989,
      "grad_norm": 0.17545063090234747,
      "learning_rate": 1.181577395507308e-06,
      "loss": 0.831,
      "step": 1818
    },
    {
      "epoch": 0.23549211897595235,
      "grad_norm": 0.12282888225220605,
      "learning_rate": 1.181448829686936e-06,
      "loss": 0.886,
      "step": 1819
    },
    {
      "epoch": 0.23562158138330583,
      "grad_norm": 0.16235943348964024,
      "learning_rate": 1.1813201502006116e-06,
      "loss": 0.9128,
      "step": 1820
    },
    {
      "epoch": 0.23562158138330583,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8539325842696629,
      "eval_PRM F1 AUC": 0.5447878470403352,
      "eval_PRM F1 Neg": 0.23529411764705882,
      "eval_PRM NPV": 0.36363636363636365,
      "eval_PRM Precision": 0.8,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.630859375,
      "eval_runtime": 20.9271,
      "eval_samples_per_second": 2.103,
      "eval_steps_per_second": 0.143,
      "step": 1820
    },
    {
      "epoch": 0.23575104379065928,
      "grad_norm": 0.16102320183197288,
      "learning_rate": 1.1811913570746205e-06,
      "loss": 0.7423,
      "step": 1821
    },
    {
      "epoch": 0.23588050619801276,
      "grad_norm": 0.11843223779657591,
      "learning_rate": 1.1810624503352713e-06,
      "loss": 0.816,
      "step": 1822
    },
    {
      "epoch": 0.23600996860536622,
      "grad_norm": 0.18567524513477238,
      "learning_rate": 1.1809334300088957e-06,
      "loss": 1.0291,
      "step": 1823
    },
    {
      "epoch": 0.23613943101271967,
      "grad_norm": 0.16042977153317808,
      "learning_rate": 1.1808042961218488e-06,
      "loss": 0.7544,
      "step": 1824
    },
    {
      "epoch": 0.23626889342007315,
      "grad_norm": 0.12991248114041043,
      "learning_rate": 1.1806750487005086e-06,
      "loss": 0.9534,
      "step": 1825
    },
    {
      "epoch": 0.23626889342007315,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8491620111731844,
      "eval_PRM F1 AUC": 0.5230487166055526,
      "eval_PRM F1 Neg": 0.18181818181818182,
      "eval_PRM NPV": 0.3,
      "eval_PRM Precision": 0.7916666666666666,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.13043478260869565,
      "eval_loss": 0.630859375,
      "eval_runtime": 20.5693,
      "eval_samples_per_second": 2.139,
      "eval_steps_per_second": 0.146,
      "step": 1825
    },
    {
      "epoch": 0.2363983558274266,
      "grad_norm": 0.1535947630232259,
      "learning_rate": 1.1805456877712769e-06,
      "loss": 0.8206,
      "step": 1826
    },
    {
      "epoch": 0.23652781823478009,
      "grad_norm": 0.15235910423146162,
      "learning_rate": 1.1804162133605781e-06,
      "loss": 0.7856,
      "step": 1827
    },
    {
      "epoch": 0.23665728064213354,
      "grad_norm": 0.15350653359172325,
      "learning_rate": 1.1802866254948598e-06,
      "loss": 0.678,
      "step": 1828
    },
    {
      "epoch": 0.236786743049487,
      "grad_norm": 0.14482814114367037,
      "learning_rate": 1.180156924200593e-06,
      "loss": 0.9011,
      "step": 1829
    },
    {
      "epoch": 0.23691620545684047,
      "grad_norm": 0.13572543477434887,
      "learning_rate": 1.180027109504272e-06,
      "loss": 0.8615,
      "step": 1830
    },
    {
      "epoch": 0.23691620545684047,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8539325842696629,
      "eval_PRM F1 AUC": 0.5447878470403352,
      "eval_PRM F1 Neg": 0.23529411764705882,
      "eval_PRM NPV": 0.36363636363636365,
      "eval_PRM Precision": 0.8,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6264204382896423,
      "eval_runtime": 20.7281,
      "eval_samples_per_second": 2.123,
      "eval_steps_per_second": 0.145,
      "step": 1830
    },
    {
      "epoch": 0.23704566786419393,
      "grad_norm": 0.18365251480839614,
      "learning_rate": 1.1798971814324143e-06,
      "loss": 0.7512,
      "step": 1831
    },
    {
      "epoch": 0.2371751302715474,
      "grad_norm": 0.13994397387070429,
      "learning_rate": 1.1797671400115596e-06,
      "loss": 0.8199,
      "step": 1832
    },
    {
      "epoch": 0.23730459267890086,
      "grad_norm": 0.26736150842162837,
      "learning_rate": 1.1796369852682722e-06,
      "loss": 0.7456,
      "step": 1833
    },
    {
      "epoch": 0.23743405508625434,
      "grad_norm": 0.121693570689154,
      "learning_rate": 1.1795067172291385e-06,
      "loss": 0.7371,
      "step": 1834
    },
    {
      "epoch": 0.2375635174936078,
      "grad_norm": 0.17122536788119297,
      "learning_rate": 1.1793763359207686e-06,
      "loss": 0.7729,
      "step": 1835
    },
    {
      "epoch": 0.2375635174936078,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8491620111731844,
      "eval_PRM F1 AUC": 0.5230487166055526,
      "eval_PRM F1 Neg": 0.18181818181818182,
      "eval_PRM NPV": 0.3,
      "eval_PRM Precision": 0.7916666666666666,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.13043478260869565,
      "eval_loss": 0.6303266882896423,
      "eval_runtime": 20.3828,
      "eval_samples_per_second": 2.159,
      "eval_steps_per_second": 0.147,
      "step": 1835
    },
    {
      "epoch": 0.23769297990096125,
      "grad_norm": 0.13125926880202782,
      "learning_rate": 1.1792458413697951e-06,
      "loss": 0.7241,
      "step": 1836
    },
    {
      "epoch": 0.23782244230831473,
      "grad_norm": 0.14609515435836865,
      "learning_rate": 1.1791152336028746e-06,
      "loss": 0.8457,
      "step": 1837
    },
    {
      "epoch": 0.23795190471566818,
      "grad_norm": 0.2050518186298441,
      "learning_rate": 1.1789845126466864e-06,
      "loss": 0.7365,
      "step": 1838
    },
    {
      "epoch": 0.23808136712302166,
      "grad_norm": 0.14630482574605344,
      "learning_rate": 1.1788536785279325e-06,
      "loss": 0.7441,
      "step": 1839
    },
    {
      "epoch": 0.23821082953037512,
      "grad_norm": 0.16137748175666597,
      "learning_rate": 1.1787227312733388e-06,
      "loss": 0.8771,
      "step": 1840
    },
    {
      "epoch": 0.23821082953037512,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8539325842696629,
      "eval_PRM F1 AUC": 0.5447878470403352,
      "eval_PRM F1 Neg": 0.23529411764705882,
      "eval_PRM NPV": 0.36363636363636365,
      "eval_PRM Precision": 0.8,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6301491260528564,
      "eval_runtime": 20.51,
      "eval_samples_per_second": 2.145,
      "eval_steps_per_second": 0.146,
      "step": 1840
    },
    {
      "epoch": 0.23834029193772857,
      "grad_norm": 0.14299094504189705,
      "learning_rate": 1.1785916709096538e-06,
      "loss": 0.8059,
      "step": 1841
    },
    {
      "epoch": 0.23846975434508205,
      "grad_norm": 0.13561275967004827,
      "learning_rate": 1.178460497463649e-06,
      "loss": 0.7421,
      "step": 1842
    },
    {
      "epoch": 0.2385992167524355,
      "grad_norm": 0.11755249085745938,
      "learning_rate": 1.1783292109621196e-06,
      "loss": 0.83,
      "step": 1843
    },
    {
      "epoch": 0.238728679159789,
      "grad_norm": 0.18331797580105738,
      "learning_rate": 1.1781978114318835e-06,
      "loss": 0.7808,
      "step": 1844
    },
    {
      "epoch": 0.23885814156714244,
      "grad_norm": 0.1111462179020677,
      "learning_rate": 1.1780662988997814e-06,
      "loss": 0.8239,
      "step": 1845
    },
    {
      "epoch": 0.23885814156714244,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8491620111731844,
      "eval_PRM F1 AUC": 0.5230487166055526,
      "eval_PRM F1 Neg": 0.18181818181818182,
      "eval_PRM NPV": 0.3,
      "eval_PRM Precision": 0.7916666666666666,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.13043478260869565,
      "eval_loss": 0.6324573755264282,
      "eval_runtime": 20.7294,
      "eval_samples_per_second": 2.123,
      "eval_steps_per_second": 0.145,
      "step": 1845
    },
    {
      "epoch": 0.2389876039744959,
      "grad_norm": 0.24728669392137842,
      "learning_rate": 1.1779346733926778e-06,
      "loss": 0.7131,
      "step": 1846
    },
    {
      "epoch": 0.23911706638184937,
      "grad_norm": 0.11829461483571425,
      "learning_rate": 1.1778029349374597e-06,
      "loss": 0.8586,
      "step": 1847
    },
    {
      "epoch": 0.23924652878920283,
      "grad_norm": 0.17090982074740754,
      "learning_rate": 1.1776710835610371e-06,
      "loss": 0.7546,
      "step": 1848
    },
    {
      "epoch": 0.2393759911965563,
      "grad_norm": 0.16943332427184796,
      "learning_rate": 1.1775391192903437e-06,
      "loss": 0.8501,
      "step": 1849
    },
    {
      "epoch": 0.23950545360390976,
      "grad_norm": 0.1345507880445574,
      "learning_rate": 1.1774070421523358e-06,
      "loss": 0.8285,
      "step": 1850
    },
    {
      "epoch": 0.23950545360390976,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8539325842696629,
      "eval_PRM F1 AUC": 0.5447878470403352,
      "eval_PRM F1 Neg": 0.23529411764705882,
      "eval_PRM NPV": 0.36363636363636365,
      "eval_PRM Precision": 0.8,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6283735632896423,
      "eval_runtime": 20.3233,
      "eval_samples_per_second": 2.165,
      "eval_steps_per_second": 0.148,
      "step": 1850
    },
    {
      "epoch": 0.23963491601126322,
      "grad_norm": 0.13157758398076128,
      "learning_rate": 1.1772748521739926e-06,
      "loss": 0.7733,
      "step": 1851
    },
    {
      "epoch": 0.2397643784186167,
      "grad_norm": 0.11229137403990017,
      "learning_rate": 1.1771425493823168e-06,
      "loss": 0.8042,
      "step": 1852
    },
    {
      "epoch": 0.23989384082597015,
      "grad_norm": 0.14388152590874392,
      "learning_rate": 1.177010133804334e-06,
      "loss": 0.8736,
      "step": 1853
    },
    {
      "epoch": 0.24002330323332363,
      "grad_norm": 0.13102705911807844,
      "learning_rate": 1.1768776054670924e-06,
      "loss": 0.8312,
      "step": 1854
    },
    {
      "epoch": 0.24015276564067708,
      "grad_norm": 0.20810776149869276,
      "learning_rate": 1.1767449643976641e-06,
      "loss": 0.8401,
      "step": 1855
    },
    {
      "epoch": 0.24015276564067708,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8491620111731844,
      "eval_PRM F1 AUC": 0.5230487166055526,
      "eval_PRM F1 Neg": 0.18181818181818182,
      "eval_PRM NPV": 0.3,
      "eval_PRM Precision": 0.7916666666666666,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.13043478260869565,
      "eval_loss": 0.630859375,
      "eval_runtime": 20.7469,
      "eval_samples_per_second": 2.121,
      "eval_steps_per_second": 0.145,
      "step": 1855
    },
    {
      "epoch": 0.24028222804803057,
      "grad_norm": 0.19820069562939113,
      "learning_rate": 1.1766122106231432e-06,
      "loss": 0.9747,
      "step": 1856
    },
    {
      "epoch": 0.24041169045538402,
      "grad_norm": 0.16800154306606135,
      "learning_rate": 1.1764793441706478e-06,
      "loss": 0.8962,
      "step": 1857
    },
    {
      "epoch": 0.24054115286273747,
      "grad_norm": 0.152323137575858,
      "learning_rate": 1.1763463650673182e-06,
      "loss": 0.7462,
      "step": 1858
    },
    {
      "epoch": 0.24067061527009095,
      "grad_norm": 0.13772616546244507,
      "learning_rate": 1.1762132733403184e-06,
      "loss": 0.9456,
      "step": 1859
    },
    {
      "epoch": 0.2408000776774444,
      "grad_norm": 0.1483638176131778,
      "learning_rate": 1.1760800690168349e-06,
      "loss": 0.8228,
      "step": 1860
    },
    {
      "epoch": 0.2408000776774444,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8491620111731844,
      "eval_PRM F1 AUC": 0.5230487166055526,
      "eval_PRM F1 Neg": 0.18181818181818182,
      "eval_PRM NPV": 0.3,
      "eval_PRM Precision": 0.7916666666666666,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.13043478260869565,
      "eval_loss": 0.630859375,
      "eval_runtime": 20.2558,
      "eval_samples_per_second": 2.172,
      "eval_steps_per_second": 0.148,
      "step": 1860
    },
    {
      "epoch": 0.2409295400847979,
      "grad_norm": 0.1271104991796403,
      "learning_rate": 1.1759467521240774e-06,
      "loss": 0.9099,
      "step": 1861
    },
    {
      "epoch": 0.24105900249215134,
      "grad_norm": 0.18363482930251043,
      "learning_rate": 1.1758133226892786e-06,
      "loss": 0.7885,
      "step": 1862
    },
    {
      "epoch": 0.2411884648995048,
      "grad_norm": 0.11566043422126357,
      "learning_rate": 1.1756797807396941e-06,
      "loss": 0.7983,
      "step": 1863
    },
    {
      "epoch": 0.24131792730685828,
      "grad_norm": 0.11424498786452805,
      "learning_rate": 1.1755461263026028e-06,
      "loss": 0.8563,
      "step": 1864
    },
    {
      "epoch": 0.24144738971421173,
      "grad_norm": 0.20333840549329754,
      "learning_rate": 1.175412359405306e-06,
      "loss": 0.735,
      "step": 1865
    },
    {
      "epoch": 0.24144738971421173,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8491620111731844,
      "eval_PRM F1 AUC": 0.5230487166055526,
      "eval_PRM F1 Neg": 0.18181818181818182,
      "eval_PRM NPV": 0.3,
      "eval_PRM Precision": 0.7916666666666666,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.13043478260869565,
      "eval_loss": 0.6313920617103577,
      "eval_runtime": 20.5497,
      "eval_samples_per_second": 2.141,
      "eval_steps_per_second": 0.146,
      "step": 1865
    },
    {
      "epoch": 0.2415768521215652,
      "grad_norm": 0.18653261835698837,
      "learning_rate": 1.1752784800751285e-06,
      "loss": 0.6794,
      "step": 1866
    },
    {
      "epoch": 0.24170631452891866,
      "grad_norm": 0.13254590824630008,
      "learning_rate": 1.1751444883394177e-06,
      "loss": 0.7876,
      "step": 1867
    },
    {
      "epoch": 0.24183577693627212,
      "grad_norm": 0.11808434437372467,
      "learning_rate": 1.1750103842255443e-06,
      "loss": 0.78,
      "step": 1868
    },
    {
      "epoch": 0.2419652393436256,
      "grad_norm": 0.2614421336116862,
      "learning_rate": 1.1748761677609018e-06,
      "loss": 0.7228,
      "step": 1869
    },
    {
      "epoch": 0.24209470175097905,
      "grad_norm": 0.2337658627078837,
      "learning_rate": 1.1747418389729066e-06,
      "loss": 0.8651,
      "step": 1870
    },
    {
      "epoch": 0.24209470175097905,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8491620111731844,
      "eval_PRM F1 AUC": 0.5230487166055526,
      "eval_PRM F1 Neg": 0.18181818181818182,
      "eval_PRM NPV": 0.3,
      "eval_PRM Precision": 0.7916666666666666,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.13043478260869565,
      "eval_loss": 0.62890625,
      "eval_runtime": 19.5969,
      "eval_samples_per_second": 2.245,
      "eval_steps_per_second": 0.153,
      "step": 1870
    },
    {
      "epoch": 0.24222416415833253,
      "grad_norm": 0.15406008200654658,
      "learning_rate": 1.1746073978889977e-06,
      "loss": 0.8461,
      "step": 1871
    },
    {
      "epoch": 0.242353626565686,
      "grad_norm": 0.1217665189240172,
      "learning_rate": 1.174472844536638e-06,
      "loss": 0.9454,
      "step": 1872
    },
    {
      "epoch": 0.24248308897303944,
      "grad_norm": 0.15186200804887565,
      "learning_rate": 1.1743381789433126e-06,
      "loss": 0.8528,
      "step": 1873
    },
    {
      "epoch": 0.24261255138039292,
      "grad_norm": 0.1159995400177423,
      "learning_rate": 1.1742034011365293e-06,
      "loss": 0.84,
      "step": 1874
    },
    {
      "epoch": 0.24274201378774637,
      "grad_norm": 0.12025484458177316,
      "learning_rate": 1.1740685111438196e-06,
      "loss": 0.7976,
      "step": 1875
    },
    {
      "epoch": 0.24274201378774637,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8491620111731844,
      "eval_PRM F1 AUC": 0.5230487166055526,
      "eval_PRM F1 Neg": 0.18181818181818182,
      "eval_PRM NPV": 0.3,
      "eval_PRM Precision": 0.7916666666666666,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.13043478260869565,
      "eval_loss": 0.6303266882896423,
      "eval_runtime": 19.9133,
      "eval_samples_per_second": 2.21,
      "eval_steps_per_second": 0.151,
      "step": 1875
    },
    {
      "epoch": 0.24287147619509986,
      "grad_norm": 0.13825805919481454,
      "learning_rate": 1.1739335089927373e-06,
      "loss": 0.9054,
      "step": 1876
    },
    {
      "epoch": 0.2430009386024533,
      "grad_norm": 0.1466259351275708,
      "learning_rate": 1.1737983947108595e-06,
      "loss": 0.8455,
      "step": 1877
    },
    {
      "epoch": 0.2431304010098068,
      "grad_norm": 0.1559630794936154,
      "learning_rate": 1.173663168325786e-06,
      "loss": 0.8262,
      "step": 1878
    },
    {
      "epoch": 0.24325986341716024,
      "grad_norm": 0.1495907836450684,
      "learning_rate": 1.1735278298651395e-06,
      "loss": 0.8142,
      "step": 1879
    },
    {
      "epoch": 0.2433893258245137,
      "grad_norm": 0.11416223101436763,
      "learning_rate": 1.1733923793565656e-06,
      "loss": 0.811,
      "step": 1880
    },
    {
      "epoch": 0.2433893258245137,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8491620111731844,
      "eval_PRM F1 AUC": 0.5230487166055526,
      "eval_PRM F1 Neg": 0.18181818181818182,
      "eval_PRM NPV": 0.3,
      "eval_PRM Precision": 0.7916666666666666,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.13043478260869565,
      "eval_loss": 0.6313920617103577,
      "eval_runtime": 20.3743,
      "eval_samples_per_second": 2.16,
      "eval_steps_per_second": 0.147,
      "step": 1880
    },
    {
      "epoch": 0.24351878823186718,
      "grad_norm": 0.14223075331524934,
      "learning_rate": 1.1732568168277326e-06,
      "loss": 0.9103,
      "step": 1881
    },
    {
      "epoch": 0.24364825063922063,
      "grad_norm": 0.14380182402701014,
      "learning_rate": 1.1731211423063323e-06,
      "loss": 0.8213,
      "step": 1882
    },
    {
      "epoch": 0.2437777130465741,
      "grad_norm": 0.13127206246595616,
      "learning_rate": 1.1729853558200786e-06,
      "loss": 0.9028,
      "step": 1883
    },
    {
      "epoch": 0.24390717545392757,
      "grad_norm": 0.14105446734621277,
      "learning_rate": 1.172849457396709e-06,
      "loss": 0.9062,
      "step": 1884
    },
    {
      "epoch": 0.24403663786128102,
      "grad_norm": 0.19666911046286312,
      "learning_rate": 1.1727134470639832e-06,
      "loss": 0.7893,
      "step": 1885
    },
    {
      "epoch": 0.24403663786128102,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8491620111731844,
      "eval_PRM F1 AUC": 0.5230487166055526,
      "eval_PRM F1 Neg": 0.18181818181818182,
      "eval_PRM NPV": 0.3,
      "eval_PRM Precision": 0.7916666666666666,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.13043478260869565,
      "eval_loss": 0.6281960010528564,
      "eval_runtime": 20.0044,
      "eval_samples_per_second": 2.2,
      "eval_steps_per_second": 0.15,
      "step": 1885
    },
    {
      "epoch": 0.2441661002686345,
      "grad_norm": 0.12081576908909004,
      "learning_rate": 1.172577324849684e-06,
      "loss": 0.801,
      "step": 1886
    },
    {
      "epoch": 0.24429556267598795,
      "grad_norm": 0.1315469307922582,
      "learning_rate": 1.1724410907816178e-06,
      "loss": 0.9128,
      "step": 1887
    },
    {
      "epoch": 0.24442502508334144,
      "grad_norm": 0.12983658208243617,
      "learning_rate": 1.1723047448876124e-06,
      "loss": 0.8103,
      "step": 1888
    },
    {
      "epoch": 0.2445544874906949,
      "grad_norm": 0.14950108014044786,
      "learning_rate": 1.1721682871955196e-06,
      "loss": 0.8629,
      "step": 1889
    },
    {
      "epoch": 0.24468394989804834,
      "grad_norm": 0.12400723928261714,
      "learning_rate": 1.1720317177332132e-06,
      "loss": 0.9178,
      "step": 1890
    },
    {
      "epoch": 0.24468394989804834,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8491620111731844,
      "eval_PRM F1 AUC": 0.5230487166055526,
      "eval_PRM F1 Neg": 0.18181818181818182,
      "eval_PRM NPV": 0.3,
      "eval_PRM Precision": 0.7916666666666666,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.13043478260869565,
      "eval_loss": 0.6292613744735718,
      "eval_runtime": 19.7361,
      "eval_samples_per_second": 2.229,
      "eval_steps_per_second": 0.152,
      "step": 1890
    },
    {
      "epoch": 0.24481341230540182,
      "grad_norm": 0.11454180760409764,
      "learning_rate": 1.171895036528591e-06,
      "loss": 0.8713,
      "step": 1891
    },
    {
      "epoch": 0.24494287471275528,
      "grad_norm": 0.1181452421915348,
      "learning_rate": 1.1717582436095725e-06,
      "loss": 0.766,
      "step": 1892
    },
    {
      "epoch": 0.24507233712010876,
      "grad_norm": 0.15625864298243242,
      "learning_rate": 1.1716213390041004e-06,
      "loss": 0.8259,
      "step": 1893
    },
    {
      "epoch": 0.2452017995274622,
      "grad_norm": 0.12919708469671992,
      "learning_rate": 1.1714843227401404e-06,
      "loss": 0.91,
      "step": 1894
    },
    {
      "epoch": 0.24533126193481566,
      "grad_norm": 0.13616183143241703,
      "learning_rate": 1.1713471948456804e-06,
      "loss": 0.7852,
      "step": 1895
    },
    {
      "epoch": 0.24533126193481566,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8491620111731844,
      "eval_PRM F1 AUC": 0.5230487166055526,
      "eval_PRM F1 Neg": 0.18181818181818182,
      "eval_PRM NPV": 0.3,
      "eval_PRM Precision": 0.7916666666666666,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.13043478260869565,
      "eval_loss": 0.6285511255264282,
      "eval_runtime": 20.611,
      "eval_samples_per_second": 2.135,
      "eval_steps_per_second": 0.146,
      "step": 1895
    },
    {
      "epoch": 0.24546072434216915,
      "grad_norm": 0.1685824944146491,
      "learning_rate": 1.1712099553487323e-06,
      "loss": 0.8571,
      "step": 1896
    },
    {
      "epoch": 0.2455901867495226,
      "grad_norm": 0.13130685031058717,
      "learning_rate": 1.1710726042773294e-06,
      "loss": 0.8468,
      "step": 1897
    },
    {
      "epoch": 0.24571964915687608,
      "grad_norm": 0.16396116056150337,
      "learning_rate": 1.1709351416595288e-06,
      "loss": 0.9182,
      "step": 1898
    },
    {
      "epoch": 0.24584911156422953,
      "grad_norm": 0.15343920979491008,
      "learning_rate": 1.1707975675234099e-06,
      "loss": 0.8783,
      "step": 1899
    },
    {
      "epoch": 0.24597857397158301,
      "grad_norm": 0.13391605982570756,
      "learning_rate": 1.170659881897075e-06,
      "loss": 0.842,
      "step": 1900
    },
    {
      "epoch": 0.24597857397158301,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8491620111731844,
      "eval_PRM F1 AUC": 0.5230487166055526,
      "eval_PRM F1 Neg": 0.18181818181818182,
      "eval_PRM NPV": 0.3,
      "eval_PRM Precision": 0.7916666666666666,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.13043478260869565,
      "eval_loss": 0.6278409361839294,
      "eval_runtime": 20.0485,
      "eval_samples_per_second": 2.195,
      "eval_steps_per_second": 0.15,
      "step": 1900
    },
    {
      "epoch": 0.24610803637893647,
      "grad_norm": 0.12745409106475744,
      "learning_rate": 1.1705220848086491e-06,
      "loss": 0.8203,
      "step": 1901
    },
    {
      "epoch": 0.24623749878628992,
      "grad_norm": 0.17190148364036936,
      "learning_rate": 1.1703841762862802e-06,
      "loss": 0.9288,
      "step": 1902
    },
    {
      "epoch": 0.2463669611936434,
      "grad_norm": 0.14575056117941482,
      "learning_rate": 1.170246156358139e-06,
      "loss": 0.8606,
      "step": 1903
    },
    {
      "epoch": 0.24649642360099686,
      "grad_norm": 0.1503698892090548,
      "learning_rate": 1.1701080250524187e-06,
      "loss": 0.7275,
      "step": 1904
    },
    {
      "epoch": 0.24662588600835034,
      "grad_norm": 0.1761589462314195,
      "learning_rate": 1.1699697823973354e-06,
      "loss": 0.6475,
      "step": 1905
    },
    {
      "epoch": 0.24662588600835034,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8491620111731844,
      "eval_PRM F1 AUC": 0.5230487166055526,
      "eval_PRM F1 Neg": 0.18181818181818182,
      "eval_PRM NPV": 0.3,
      "eval_PRM Precision": 0.7916666666666666,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.13043478260869565,
      "eval_loss": 0.6274858117103577,
      "eval_runtime": 19.7772,
      "eval_samples_per_second": 2.225,
      "eval_steps_per_second": 0.152,
      "step": 1905
    },
    {
      "epoch": 0.2467553484157038,
      "grad_norm": 0.1537244271269392,
      "learning_rate": 1.169831428421128e-06,
      "loss": 0.8531,
      "step": 1906
    },
    {
      "epoch": 0.24688481082305724,
      "grad_norm": 0.19059233458713676,
      "learning_rate": 1.169692963152058e-06,
      "loss": 0.9774,
      "step": 1907
    },
    {
      "epoch": 0.24701427323041072,
      "grad_norm": 0.14640292153669124,
      "learning_rate": 1.16955438661841e-06,
      "loss": 0.8226,
      "step": 1908
    },
    {
      "epoch": 0.24714373563776418,
      "grad_norm": 0.13083946388456033,
      "learning_rate": 1.1694156988484909e-06,
      "loss": 0.8602,
      "step": 1909
    },
    {
      "epoch": 0.24727319804511766,
      "grad_norm": 0.13557339362593737,
      "learning_rate": 1.1692768998706306e-06,
      "loss": 0.838,
      "step": 1910
    },
    {
      "epoch": 0.24727319804511766,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8539325842696629,
      "eval_PRM F1 AUC": 0.5447878470403352,
      "eval_PRM F1 Neg": 0.23529411764705882,
      "eval_PRM NPV": 0.36363636363636365,
      "eval_PRM Precision": 0.8,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6296164989471436,
      "eval_runtime": 20.5417,
      "eval_samples_per_second": 2.142,
      "eval_steps_per_second": 0.146,
      "step": 1910
    },
    {
      "epoch": 0.2474026604524711,
      "grad_norm": 0.1438321596147449,
      "learning_rate": 1.1691379897131814e-06,
      "loss": 0.8737,
      "step": 1911
    },
    {
      "epoch": 0.24753212285982457,
      "grad_norm": 0.14466766031724226,
      "learning_rate": 1.1689989684045186e-06,
      "loss": 0.796,
      "step": 1912
    },
    {
      "epoch": 0.24766158526717805,
      "grad_norm": 0.13730639463113248,
      "learning_rate": 1.1688598359730401e-06,
      "loss": 0.7905,
      "step": 1913
    },
    {
      "epoch": 0.2477910476745315,
      "grad_norm": 0.21301195243820503,
      "learning_rate": 1.1687205924471668e-06,
      "loss": 0.7709,
      "step": 1914
    },
    {
      "epoch": 0.24792051008188498,
      "grad_norm": 0.10796768730262489,
      "learning_rate": 1.1685812378553416e-06,
      "loss": 0.7975,
      "step": 1915
    },
    {
      "epoch": 0.24792051008188498,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8491620111731844,
      "eval_PRM F1 AUC": 0.5230487166055526,
      "eval_PRM F1 Neg": 0.18181818181818182,
      "eval_PRM NPV": 0.3,
      "eval_PRM Precision": 0.7916666666666666,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.13043478260869565,
      "eval_loss": 0.62890625,
      "eval_runtime": 19.901,
      "eval_samples_per_second": 2.211,
      "eval_steps_per_second": 0.151,
      "step": 1915
    },
    {
      "epoch": 0.24804997248923844,
      "grad_norm": 0.13814601332758988,
      "learning_rate": 1.1684417722260306e-06,
      "loss": 0.8136,
      "step": 1916
    },
    {
      "epoch": 0.2481794348965919,
      "grad_norm": 0.14743771419003096,
      "learning_rate": 1.1683021955877226e-06,
      "loss": 0.8306,
      "step": 1917
    },
    {
      "epoch": 0.24830889730394537,
      "grad_norm": 0.13497541757763676,
      "learning_rate": 1.168162507968929e-06,
      "loss": 0.7094,
      "step": 1918
    },
    {
      "epoch": 0.24843835971129882,
      "grad_norm": 0.16623568950263984,
      "learning_rate": 1.1680227093981838e-06,
      "loss": 0.7772,
      "step": 1919
    },
    {
      "epoch": 0.2485678221186523,
      "grad_norm": 0.13342026051396888,
      "learning_rate": 1.1678827999040434e-06,
      "loss": 0.9084,
      "step": 1920
    },
    {
      "epoch": 0.2485678221186523,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8444444444444444,
      "eval_PRM F1 AUC": 0.50130958617077,
      "eval_PRM F1 Neg": 0.125,
      "eval_PRM NPV": 0.2222222222222222,
      "eval_PRM Precision": 0.7835051546391752,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.08695652173913043,
      "eval_loss": 0.6287286877632141,
      "eval_runtime": 19.9891,
      "eval_samples_per_second": 2.201,
      "eval_steps_per_second": 0.15,
      "step": 1920
    },
    {
      "epoch": 0.24869728452600576,
      "grad_norm": 0.17691767627759009,
      "learning_rate": 1.1677427795150874e-06,
      "loss": 0.7793,
      "step": 1921
    },
    {
      "epoch": 0.24882674693335924,
      "grad_norm": 0.15975313416844006,
      "learning_rate": 1.1676026482599181e-06,
      "loss": 0.8608,
      "step": 1922
    },
    {
      "epoch": 0.2489562093407127,
      "grad_norm": 0.19979229779840876,
      "learning_rate": 1.1674624061671597e-06,
      "loss": 0.7374,
      "step": 1923
    },
    {
      "epoch": 0.24908567174806615,
      "grad_norm": 0.13782916082841962,
      "learning_rate": 1.1673220532654596e-06,
      "loss": 0.8584,
      "step": 1924
    },
    {
      "epoch": 0.24921513415541963,
      "grad_norm": 0.13957738933418853,
      "learning_rate": 1.1671815895834878e-06,
      "loss": 0.8494,
      "step": 1925
    },
    {
      "epoch": 0.24921513415541963,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8444444444444444,
      "eval_PRM F1 AUC": 0.50130958617077,
      "eval_PRM F1 Neg": 0.125,
      "eval_PRM NPV": 0.2222222222222222,
      "eval_PRM Precision": 0.7835051546391752,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.08695652173913043,
      "eval_loss": 0.6276633739471436,
      "eval_runtime": 19.9314,
      "eval_samples_per_second": 2.208,
      "eval_steps_per_second": 0.151,
      "step": 1925
    },
    {
      "epoch": 0.24934459656277308,
      "grad_norm": 0.23466602778212164,
      "learning_rate": 1.1670410151499367e-06,
      "loss": 0.7162,
      "step": 1926
    },
    {
      "epoch": 0.24947405897012656,
      "grad_norm": 0.1102736473903305,
      "learning_rate": 1.1669003299935216e-06,
      "loss": 0.7488,
      "step": 1927
    },
    {
      "epoch": 0.24960352137748001,
      "grad_norm": 0.15518533184826852,
      "learning_rate": 1.1667595341429804e-06,
      "loss": 0.8014,
      "step": 1928
    },
    {
      "epoch": 0.24973298378483347,
      "grad_norm": 0.12014354000434208,
      "learning_rate": 1.1666186276270732e-06,
      "loss": 0.7405,
      "step": 1929
    },
    {
      "epoch": 0.24986244619218695,
      "grad_norm": 0.11027979041053852,
      "learning_rate": 1.1664776104745833e-06,
      "loss": 0.821,
      "step": 1930
    },
    {
      "epoch": 0.24986244619218695,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8444444444444444,
      "eval_PRM F1 AUC": 0.50130958617077,
      "eval_PRM F1 Neg": 0.125,
      "eval_PRM NPV": 0.2222222222222222,
      "eval_PRM Precision": 0.7835051546391752,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.08695652173913043,
      "eval_loss": 0.6283735632896423,
      "eval_runtime": 20.4103,
      "eval_samples_per_second": 2.156,
      "eval_steps_per_second": 0.147,
      "step": 1930
    },
    {
      "epoch": 0.2499919085995404,
      "grad_norm": 0.1258801105018145,
      "learning_rate": 1.1663364827143164e-06,
      "loss": 0.8099,
      "step": 1931
    },
    {
      "epoch": 0.25012137100689386,
      "grad_norm": 0.22609744499809295,
      "learning_rate": 1.1661952443751003e-06,
      "loss": 0.7563,
      "step": 1932
    },
    {
      "epoch": 0.25025083341424736,
      "grad_norm": 0.1457928132315668,
      "learning_rate": 1.166053895485786e-06,
      "loss": 0.9554,
      "step": 1933
    },
    {
      "epoch": 0.2503802958216008,
      "grad_norm": 0.1397408403340529,
      "learning_rate": 1.165912436075247e-06,
      "loss": 0.7216,
      "step": 1934
    },
    {
      "epoch": 0.25050975822895427,
      "grad_norm": 0.15632853495430424,
      "learning_rate": 1.165770866172379e-06,
      "loss": 0.7695,
      "step": 1935
    },
    {
      "epoch": 0.25050975822895427,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8491620111731844,
      "eval_PRM F1 AUC": 0.5230487166055526,
      "eval_PRM F1 Neg": 0.18181818181818182,
      "eval_PRM NPV": 0.3,
      "eval_PRM Precision": 0.7916666666666666,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.13043478260869565,
      "eval_loss": 0.6264204382896423,
      "eval_runtime": 19.8885,
      "eval_samples_per_second": 2.212,
      "eval_steps_per_second": 0.151,
      "step": 1935
    },
    {
      "epoch": 0.2506392206363077,
      "grad_norm": 0.1328345179356024,
      "learning_rate": 1.1656291858061007e-06,
      "loss": 0.8441,
      "step": 1936
    },
    {
      "epoch": 0.2507686830436612,
      "grad_norm": 0.12524298648514604,
      "learning_rate": 1.165487395005353e-06,
      "loss": 0.8485,
      "step": 1937
    },
    {
      "epoch": 0.2508981454510147,
      "grad_norm": 0.1385508428901087,
      "learning_rate": 1.1653454937990998e-06,
      "loss": 0.8708,
      "step": 1938
    },
    {
      "epoch": 0.25102760785836814,
      "grad_norm": 0.10963315542092797,
      "learning_rate": 1.1652034822163274e-06,
      "loss": 0.7264,
      "step": 1939
    },
    {
      "epoch": 0.2511570702657216,
      "grad_norm": 0.13749896673468692,
      "learning_rate": 1.1650613602860442e-06,
      "loss": 0.8892,
      "step": 1940
    },
    {
      "epoch": 0.2511570702657216,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8491620111731844,
      "eval_PRM F1 AUC": 0.5230487166055526,
      "eval_PRM F1 Neg": 0.18181818181818182,
      "eval_PRM NPV": 0.3,
      "eval_PRM Precision": 0.7916666666666666,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.13043478260869565,
      "eval_loss": 0.6285511255264282,
      "eval_runtime": 20.2615,
      "eval_samples_per_second": 2.172,
      "eval_steps_per_second": 0.148,
      "step": 1940
    },
    {
      "epoch": 0.25128653267307505,
      "grad_norm": 0.12067764994621238,
      "learning_rate": 1.1649191280372815e-06,
      "loss": 0.8275,
      "step": 1941
    },
    {
      "epoch": 0.2514159950804285,
      "grad_norm": 0.12879866961021263,
      "learning_rate": 1.1647767854990934e-06,
      "loss": 0.8312,
      "step": 1942
    },
    {
      "epoch": 0.251545457487782,
      "grad_norm": 0.1551316365218683,
      "learning_rate": 1.164634332700556e-06,
      "loss": 0.9441,
      "step": 1943
    },
    {
      "epoch": 0.25167491989513546,
      "grad_norm": 0.1295722672623929,
      "learning_rate": 1.1644917696707683e-06,
      "loss": 0.7772,
      "step": 1944
    },
    {
      "epoch": 0.2518043823024889,
      "grad_norm": 0.177198669864036,
      "learning_rate": 1.1643490964388516e-06,
      "loss": 0.8585,
      "step": 1945
    },
    {
      "epoch": 0.2518043823024889,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8539325842696629,
      "eval_PRM F1 AUC": 0.5447878470403352,
      "eval_PRM F1 Neg": 0.23529411764705882,
      "eval_PRM NPV": 0.36363636363636365,
      "eval_PRM Precision": 0.8,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.17391304347826086,
      "eval_loss": 0.6273082494735718,
      "eval_runtime": 19.5362,
      "eval_samples_per_second": 2.252,
      "eval_steps_per_second": 0.154,
      "step": 1945
    },
    {
      "epoch": 0.25193384470984237,
      "grad_norm": 0.1345102600657231,
      "learning_rate": 1.1642063130339499e-06,
      "loss": 0.8682,
      "step": 1946
    },
    {
      "epoch": 0.2520633071171958,
      "grad_norm": 0.2047847795220941,
      "learning_rate": 1.1640634194852296e-06,
      "loss": 1.0645,
      "step": 1947
    },
    {
      "epoch": 0.25219276952454933,
      "grad_norm": 0.2195214606581273,
      "learning_rate": 1.1639204158218793e-06,
      "loss": 0.7411,
      "step": 1948
    },
    {
      "epoch": 0.2523222319319028,
      "grad_norm": 0.11288568737342215,
      "learning_rate": 1.1637773020731108e-06,
      "loss": 0.7531,
      "step": 1949
    },
    {
      "epoch": 0.25245169433925624,
      "grad_norm": 0.14116998189704644,
      "learning_rate": 1.1636340782681578e-06,
      "loss": 0.8275,
      "step": 1950
    },
    {
      "epoch": 0.25245169433925624,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8444444444444444,
      "eval_PRM F1 AUC": 0.50130958617077,
      "eval_PRM F1 Neg": 0.125,
      "eval_PRM NPV": 0.2222222222222222,
      "eval_PRM Precision": 0.7835051546391752,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.08695652173913043,
      "eval_loss": 0.62890625,
      "eval_runtime": 21.0097,
      "eval_samples_per_second": 2.094,
      "eval_steps_per_second": 0.143,
      "step": 1950
    },
    {
      "epoch": 0.2525811567466097,
      "grad_norm": 0.1383304300698129,
      "learning_rate": 1.1634907444362767e-06,
      "loss": 0.7773,
      "step": 1951
    },
    {
      "epoch": 0.25271061915396315,
      "grad_norm": 0.1457834679757666,
      "learning_rate": 1.1633473006067464e-06,
      "loss": 0.8063,
      "step": 1952
    },
    {
      "epoch": 0.25284008156131665,
      "grad_norm": 0.25865396019684334,
      "learning_rate": 1.1632037468088682e-06,
      "loss": 0.7639,
      "step": 1953
    },
    {
      "epoch": 0.2529695439686701,
      "grad_norm": 0.20332543550105217,
      "learning_rate": 1.1630600830719657e-06,
      "loss": 0.7823,
      "step": 1954
    },
    {
      "epoch": 0.25309900637602356,
      "grad_norm": 0.18564462251824457,
      "learning_rate": 1.1629163094253853e-06,
      "loss": 0.7411,
      "step": 1955
    },
    {
      "epoch": 0.25309900637602356,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8444444444444444,
      "eval_PRM F1 AUC": 0.50130958617077,
      "eval_PRM F1 Neg": 0.125,
      "eval_PRM NPV": 0.2222222222222222,
      "eval_PRM Precision": 0.7835051546391752,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.08695652173913043,
      "eval_loss": 0.6290838122367859,
      "eval_runtime": 20.3608,
      "eval_samples_per_second": 2.161,
      "eval_steps_per_second": 0.147,
      "step": 1955
    },
    {
      "epoch": 0.253228468783377,
      "grad_norm": 0.14371204515009353,
      "learning_rate": 1.1627724258984956e-06,
      "loss": 0.7889,
      "step": 1956
    },
    {
      "epoch": 0.25335793119073047,
      "grad_norm": 0.1685862147201397,
      "learning_rate": 1.1626284325206878e-06,
      "loss": 0.8308,
      "step": 1957
    },
    {
      "epoch": 0.253487393598084,
      "grad_norm": 0.13757083399447004,
      "learning_rate": 1.1624843293213754e-06,
      "loss": 0.8192,
      "step": 1958
    },
    {
      "epoch": 0.25361685600543743,
      "grad_norm": 0.1350975658473077,
      "learning_rate": 1.1623401163299944e-06,
      "loss": 0.8981,
      "step": 1959
    },
    {
      "epoch": 0.2537463184127909,
      "grad_norm": 0.13676378667988795,
      "learning_rate": 1.1621957935760033e-06,
      "loss": 0.724,
      "step": 1960
    },
    {
      "epoch": 0.2537463184127909,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8491620111731844,
      "eval_PRM F1 AUC": 0.5230487166055526,
      "eval_PRM F1 Neg": 0.18181818181818182,
      "eval_PRM NPV": 0.3,
      "eval_PRM Precision": 0.7916666666666666,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.13043478260869565,
      "eval_loss": 0.625,
      "eval_runtime": 20.2894,
      "eval_samples_per_second": 2.169,
      "eval_steps_per_second": 0.148,
      "step": 1960
    },
    {
      "epoch": 0.25387578082014434,
      "grad_norm": 0.20379127207600814,
      "learning_rate": 1.1620513610888829e-06,
      "loss": 0.8256,
      "step": 1961
    },
    {
      "epoch": 0.2540052432274978,
      "grad_norm": 0.14884051463798825,
      "learning_rate": 1.1619068188981366e-06,
      "loss": 0.7762,
      "step": 1962
    },
    {
      "epoch": 0.2541347056348513,
      "grad_norm": 0.1774721747457146,
      "learning_rate": 1.1617621670332896e-06,
      "loss": 0.8191,
      "step": 1963
    },
    {
      "epoch": 0.25426416804220475,
      "grad_norm": 0.12951076722980323,
      "learning_rate": 1.1616174055238907e-06,
      "loss": 0.8615,
      "step": 1964
    },
    {
      "epoch": 0.2543936304495582,
      "grad_norm": 0.117439304680866,
      "learning_rate": 1.1614725343995096e-06,
      "loss": 0.7192,
      "step": 1965
    },
    {
      "epoch": 0.2543936304495582,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8491620111731844,
      "eval_PRM F1 AUC": 0.5230487166055526,
      "eval_PRM F1 Neg": 0.18181818181818182,
      "eval_PRM NPV": 0.3,
      "eval_PRM Precision": 0.7916666666666666,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.13043478260869565,
      "eval_loss": 0.6276633739471436,
      "eval_runtime": 19.9246,
      "eval_samples_per_second": 2.208,
      "eval_steps_per_second": 0.151,
      "step": 1965
    },
    {
      "epoch": 0.25452309285691166,
      "grad_norm": 0.18686221677762116,
      "learning_rate": 1.1613275536897397e-06,
      "loss": 0.8351,
      "step": 1966
    },
    {
      "epoch": 0.2546525552642651,
      "grad_norm": 0.26747967319661453,
      "learning_rate": 1.1611824634241962e-06,
      "loss": 0.7561,
      "step": 1967
    },
    {
      "epoch": 0.2547820176716186,
      "grad_norm": 0.15266231858165313,
      "learning_rate": 1.1610372636325165e-06,
      "loss": 0.7654,
      "step": 1968
    },
    {
      "epoch": 0.2549114800789721,
      "grad_norm": 0.09633311846900564,
      "learning_rate": 1.1608919543443608e-06,
      "loss": 0.756,
      "step": 1969
    },
    {
      "epoch": 0.25504094248632553,
      "grad_norm": 0.1669779577070707,
      "learning_rate": 1.1607465355894114e-06,
      "loss": 0.838,
      "step": 1970
    },
    {
      "epoch": 0.25504094248632553,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8444444444444444,
      "eval_PRM F1 AUC": 0.50130958617077,
      "eval_PRM F1 Neg": 0.125,
      "eval_PRM NPV": 0.2222222222222222,
      "eval_PRM Precision": 0.7835051546391752,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.08695652173913043,
      "eval_loss": 0.6297940611839294,
      "eval_runtime": 20.4147,
      "eval_samples_per_second": 2.155,
      "eval_steps_per_second": 0.147,
      "step": 1970
    },
    {
      "epoch": 0.255170404893679,
      "grad_norm": 0.1280438109150198,
      "learning_rate": 1.160601007397373e-06,
      "loss": 0.8872,
      "step": 1971
    },
    {
      "epoch": 0.2552998673010325,
      "grad_norm": 0.1700861056924668,
      "learning_rate": 1.1604553697979725e-06,
      "loss": 0.7496,
      "step": 1972
    },
    {
      "epoch": 0.25542932970838594,
      "grad_norm": 0.12788202806754967,
      "learning_rate": 1.1603096228209599e-06,
      "loss": 0.8213,
      "step": 1973
    },
    {
      "epoch": 0.2555587921157394,
      "grad_norm": 0.14914762419672875,
      "learning_rate": 1.1601637664961063e-06,
      "loss": 0.7936,
      "step": 1974
    },
    {
      "epoch": 0.25568825452309285,
      "grad_norm": 0.14486025302854463,
      "learning_rate": 1.1600178008532062e-06,
      "loss": 0.897,
      "step": 1975
    },
    {
      "epoch": 0.25568825452309285,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8444444444444444,
      "eval_PRM F1 AUC": 0.50130958617077,
      "eval_PRM F1 Neg": 0.125,
      "eval_PRM NPV": 0.2222222222222222,
      "eval_PRM Precision": 0.7835051546391752,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.08695652173913043,
      "eval_loss": 0.6264204382896423,
      "eval_runtime": 19.4115,
      "eval_samples_per_second": 2.267,
      "eval_steps_per_second": 0.155,
      "step": 1975
    },
    {
      "epoch": 0.2558177169304463,
      "grad_norm": 0.14448073866641478,
      "learning_rate": 1.1598717259220762e-06,
      "loss": 0.8989,
      "step": 1976
    },
    {
      "epoch": 0.2559471793377998,
      "grad_norm": 0.16650844444345914,
      "learning_rate": 1.1597255417325544e-06,
      "loss": 0.7148,
      "step": 1977
    },
    {
      "epoch": 0.25607664174515327,
      "grad_norm": 0.19984075756293565,
      "learning_rate": 1.1595792483145027e-06,
      "loss": 0.7509,
      "step": 1978
    },
    {
      "epoch": 0.2562061041525067,
      "grad_norm": 0.17914739476276764,
      "learning_rate": 1.1594328456978038e-06,
      "loss": 0.7656,
      "step": 1979
    },
    {
      "epoch": 0.2563355665598602,
      "grad_norm": 0.17559943522595292,
      "learning_rate": 1.159286333912364e-06,
      "loss": 0.7195,
      "step": 1980
    },
    {
      "epoch": 0.2563355665598602,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8444444444444444,
      "eval_PRM F1 AUC": 0.50130958617077,
      "eval_PRM F1 Neg": 0.125,
      "eval_PRM NPV": 0.2222222222222222,
      "eval_PRM Precision": 0.7835051546391752,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.08695652173913043,
      "eval_loss": 0.6287286877632141,
      "eval_runtime": 19.9417,
      "eval_samples_per_second": 2.206,
      "eval_steps_per_second": 0.15,
      "step": 1980
    },
    {
      "epoch": 0.2564650289672136,
      "grad_norm": 0.21065962521271814,
      "learning_rate": 1.1591397129881106e-06,
      "loss": 0.6821,
      "step": 1981
    },
    {
      "epoch": 0.25659449137456714,
      "grad_norm": 0.22790785927320714,
      "learning_rate": 1.1589929829549947e-06,
      "loss": 1.0652,
      "step": 1982
    },
    {
      "epoch": 0.2567239537819206,
      "grad_norm": 0.17819035827049462,
      "learning_rate": 1.1588461438429882e-06,
      "loss": 0.9248,
      "step": 1983
    },
    {
      "epoch": 0.25685341618927404,
      "grad_norm": 0.1432936682786671,
      "learning_rate": 1.1586991956820864e-06,
      "loss": 0.8076,
      "step": 1984
    },
    {
      "epoch": 0.2569828785966275,
      "grad_norm": 0.11361100891638747,
      "learning_rate": 1.1585521385023062e-06,
      "loss": 0.8118,
      "step": 1985
    },
    {
      "epoch": 0.2569828785966275,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8397790055248618,
      "eval_PRM F1 AUC": 0.4795704557359874,
      "eval_PRM F1 Neg": 0.06451612903225806,
      "eval_PRM NPV": 0.125,
      "eval_PRM Precision": 0.7755102040816326,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.043478260869565216,
      "eval_loss": 0.6305042505264282,
      "eval_runtime": 20.4932,
      "eval_samples_per_second": 2.147,
      "eval_steps_per_second": 0.146,
      "step": 1985
    },
    {
      "epoch": 0.25711234100398095,
      "grad_norm": 0.12093654504791841,
      "learning_rate": 1.1584049723336874e-06,
      "loss": 0.874,
      "step": 1986
    },
    {
      "epoch": 0.25724180341133446,
      "grad_norm": 0.08885153782673692,
      "learning_rate": 1.1582576972062912e-06,
      "loss": 0.688,
      "step": 1987
    },
    {
      "epoch": 0.2573712658186879,
      "grad_norm": 0.1263723801323602,
      "learning_rate": 1.1581103131502016e-06,
      "loss": 0.8427,
      "step": 1988
    },
    {
      "epoch": 0.25750072822604136,
      "grad_norm": 0.21484206695682037,
      "learning_rate": 1.1579628201955248e-06,
      "loss": 0.7504,
      "step": 1989
    },
    {
      "epoch": 0.2576301906333948,
      "grad_norm": 0.11136936028856535,
      "learning_rate": 1.1578152183723895e-06,
      "loss": 0.849,
      "step": 1990
    },
    {
      "epoch": 0.2576301906333948,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8444444444444444,
      "eval_PRM F1 AUC": 0.50130958617077,
      "eval_PRM F1 Neg": 0.125,
      "eval_PRM NPV": 0.2222222222222222,
      "eval_PRM Precision": 0.7835051546391752,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.08695652173913043,
      "eval_loss": 0.626953125,
      "eval_runtime": 20.2047,
      "eval_samples_per_second": 2.178,
      "eval_steps_per_second": 0.148,
      "step": 1990
    },
    {
      "epoch": 0.25775965304074827,
      "grad_norm": 0.12440594821660243,
      "learning_rate": 1.1576675077109462e-06,
      "loss": 0.8154,
      "step": 1991
    },
    {
      "epoch": 0.2578891154481018,
      "grad_norm": 0.16061753672431187,
      "learning_rate": 1.1575196882413677e-06,
      "loss": 0.751,
      "step": 1992
    },
    {
      "epoch": 0.25801857785545523,
      "grad_norm": 0.28965185343139443,
      "learning_rate": 1.1573717599938492e-06,
      "loss": 0.6744,
      "step": 1993
    },
    {
      "epoch": 0.2581480402628087,
      "grad_norm": 0.14114242764852597,
      "learning_rate": 1.157223722998608e-06,
      "loss": 0.8048,
      "step": 1994
    },
    {
      "epoch": 0.25827750267016214,
      "grad_norm": 0.1922661978490616,
      "learning_rate": 1.1570755772858837e-06,
      "loss": 0.7499,
      "step": 1995
    },
    {
      "epoch": 0.25827750267016214,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8444444444444444,
      "eval_PRM F1 AUC": 0.50130958617077,
      "eval_PRM F1 Neg": 0.125,
      "eval_PRM NPV": 0.2222222222222222,
      "eval_PRM Precision": 0.7835051546391752,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.08695652173913043,
      "eval_loss": 0.6280184388160706,
      "eval_runtime": 20.086,
      "eval_samples_per_second": 2.191,
      "eval_steps_per_second": 0.149,
      "step": 1995
    },
    {
      "epoch": 0.2584069650775156,
      "grad_norm": 0.14664055352824307,
      "learning_rate": 1.1569273228859381e-06,
      "loss": 0.7541,
      "step": 1996
    },
    {
      "epoch": 0.2585364274848691,
      "grad_norm": 0.188801091711494,
      "learning_rate": 1.1567789598290552e-06,
      "loss": 0.884,
      "step": 1997
    },
    {
      "epoch": 0.25866588989222256,
      "grad_norm": 0.11536965995853421,
      "learning_rate": 1.156630488145541e-06,
      "loss": 0.7991,
      "step": 1998
    },
    {
      "epoch": 0.258795352299576,
      "grad_norm": 0.14128483309740617,
      "learning_rate": 1.1564819078657238e-06,
      "loss": 0.8816,
      "step": 1999
    },
    {
      "epoch": 0.25892481470692946,
      "grad_norm": 0.13007428616774513,
      "learning_rate": 1.1563332190199544e-06,
      "loss": 0.8108,
      "step": 2000
    },
    {
      "epoch": 0.25892481470692946,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8444444444444444,
      "eval_PRM F1 AUC": 0.50130958617077,
      "eval_PRM F1 Neg": 0.125,
      "eval_PRM NPV": 0.2222222222222222,
      "eval_PRM Precision": 0.7835051546391752,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.08695652173913043,
      "eval_loss": 0.6278409361839294,
      "eval_runtime": 20.3895,
      "eval_samples_per_second": 2.158,
      "eval_steps_per_second": 0.147,
      "step": 2000
    },
    {
      "epoch": 0.2590542771142829,
      "grad_norm": 0.137201624662328,
      "learning_rate": 1.1561844216386054e-06,
      "loss": 0.7937,
      "step": 2001
    },
    {
      "epoch": 0.2591837395216364,
      "grad_norm": 0.14382572682434164,
      "learning_rate": 1.1560355157520717e-06,
      "loss": 0.8229,
      "step": 2002
    },
    {
      "epoch": 0.2593132019289899,
      "grad_norm": 0.17435310405244686,
      "learning_rate": 1.1558865013907703e-06,
      "loss": 0.826,
      "step": 2003
    },
    {
      "epoch": 0.25944266433634333,
      "grad_norm": 0.18439542268748568,
      "learning_rate": 1.1557373785851404e-06,
      "loss": 0.7404,
      "step": 2004
    },
    {
      "epoch": 0.2595721267436968,
      "grad_norm": 0.14401459283363124,
      "learning_rate": 1.1555881473656433e-06,
      "loss": 0.7562,
      "step": 2005
    },
    {
      "epoch": 0.2595721267436968,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8397790055248618,
      "eval_PRM F1 AUC": 0.4795704557359874,
      "eval_PRM F1 Neg": 0.06451612903225806,
      "eval_PRM NPV": 0.125,
      "eval_PRM Precision": 0.7755102040816326,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.043478260869565216,
      "eval_loss": 0.6285511255264282,
      "eval_runtime": 20.8918,
      "eval_samples_per_second": 2.106,
      "eval_steps_per_second": 0.144,
      "step": 2005
    },
    {
      "epoch": 0.25970158915105024,
      "grad_norm": 0.15757788765910422,
      "learning_rate": 1.1554388077627625e-06,
      "loss": 0.8573,
      "step": 2006
    },
    {
      "epoch": 0.25983105155840375,
      "grad_norm": 0.1139284466918151,
      "learning_rate": 1.1552893598070039e-06,
      "loss": 0.8081,
      "step": 2007
    },
    {
      "epoch": 0.2599605139657572,
      "grad_norm": 0.19899656660259354,
      "learning_rate": 1.155139803528895e-06,
      "loss": 0.7479,
      "step": 2008
    },
    {
      "epoch": 0.26008997637311065,
      "grad_norm": 0.1397981526918433,
      "learning_rate": 1.1549901389589857e-06,
      "loss": 0.8324,
      "step": 2009
    },
    {
      "epoch": 0.2602194387804641,
      "grad_norm": 0.18278413775738017,
      "learning_rate": 1.1548403661278482e-06,
      "loss": 0.6917,
      "step": 2010
    },
    {
      "epoch": 0.2602194387804641,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8397790055248618,
      "eval_PRM F1 AUC": 0.4795704557359874,
      "eval_PRM F1 Neg": 0.06451612903225806,
      "eval_PRM NPV": 0.125,
      "eval_PRM Precision": 0.7755102040816326,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.043478260869565216,
      "eval_loss": 0.6292613744735718,
      "eval_runtime": 20.2268,
      "eval_samples_per_second": 2.175,
      "eval_steps_per_second": 0.148,
      "step": 2010
    },
    {
      "epoch": 0.26034890118781756,
      "grad_norm": 0.13999189876492063,
      "learning_rate": 1.1546904850660764e-06,
      "loss": 0.778,
      "step": 2011
    },
    {
      "epoch": 0.26047836359517107,
      "grad_norm": 0.31321386310656163,
      "learning_rate": 1.154540495804287e-06,
      "loss": 0.7316,
      "step": 2012
    },
    {
      "epoch": 0.2606078260025245,
      "grad_norm": 0.1299608005216472,
      "learning_rate": 1.1543903983731178e-06,
      "loss": 0.8265,
      "step": 2013
    },
    {
      "epoch": 0.260737288409878,
      "grad_norm": 0.12087701962897705,
      "learning_rate": 1.1542401928032295e-06,
      "loss": 0.803,
      "step": 2014
    },
    {
      "epoch": 0.26086675081723143,
      "grad_norm": 0.20484797472002392,
      "learning_rate": 1.1540898791253047e-06,
      "loss": 0.7745,
      "step": 2015
    },
    {
      "epoch": 0.26086675081723143,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8397790055248618,
      "eval_PRM F1 AUC": 0.4795704557359874,
      "eval_PRM F1 Neg": 0.06451612903225806,
      "eval_PRM NPV": 0.125,
      "eval_PRM Precision": 0.7755102040816326,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.043478260869565216,
      "eval_loss": 0.6274858117103577,
      "eval_runtime": 20.3723,
      "eval_samples_per_second": 2.16,
      "eval_steps_per_second": 0.147,
      "step": 2015
    },
    {
      "epoch": 0.26099621322458494,
      "grad_norm": 0.13584156249337234,
      "learning_rate": 1.153939457370048e-06,
      "loss": 0.8384,
      "step": 2016
    },
    {
      "epoch": 0.2611256756319384,
      "grad_norm": 0.11028620413761461,
      "learning_rate": 1.1537889275681857e-06,
      "loss": 0.7418,
      "step": 2017
    },
    {
      "epoch": 0.26125513803929185,
      "grad_norm": 0.15783218685077557,
      "learning_rate": 1.153638289750467e-06,
      "loss": 0.7241,
      "step": 2018
    },
    {
      "epoch": 0.2613846004466453,
      "grad_norm": 0.1436463184272526,
      "learning_rate": 1.1534875439476629e-06,
      "loss": 0.8899,
      "step": 2019
    },
    {
      "epoch": 0.26151406285399875,
      "grad_norm": 0.12750033879260042,
      "learning_rate": 1.1533366901905656e-06,
      "loss": 0.7988,
      "step": 2020
    },
    {
      "epoch": 0.26151406285399875,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8397790055248618,
      "eval_PRM F1 AUC": 0.4795704557359874,
      "eval_PRM F1 Neg": 0.06451612903225806,
      "eval_PRM NPV": 0.125,
      "eval_PRM Precision": 0.7755102040816326,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.043478260869565216,
      "eval_loss": 0.6278409361839294,
      "eval_runtime": 20.0601,
      "eval_samples_per_second": 2.193,
      "eval_steps_per_second": 0.15,
      "step": 2020
    },
    {
      "epoch": 0.26164352526135226,
      "grad_norm": 0.2540584018694295,
      "learning_rate": 1.1531857285099906e-06,
      "loss": 0.6709,
      "step": 2021
    },
    {
      "epoch": 0.2617729876687057,
      "grad_norm": 0.1747971447866238,
      "learning_rate": 1.1530346589367748e-06,
      "loss": 0.8379,
      "step": 2022
    },
    {
      "epoch": 0.26190245007605917,
      "grad_norm": 0.18075509115225027,
      "learning_rate": 1.152883481501777e-06,
      "loss": 0.9508,
      "step": 2023
    },
    {
      "epoch": 0.2620319124834126,
      "grad_norm": 0.1532988905863353,
      "learning_rate": 1.1527321962358785e-06,
      "loss": 0.757,
      "step": 2024
    },
    {
      "epoch": 0.2621613748907661,
      "grad_norm": 0.10273897501664732,
      "learning_rate": 1.152580803169982e-06,
      "loss": 0.7821,
      "step": 2025
    },
    {
      "epoch": 0.2621613748907661,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8397790055248618,
      "eval_PRM F1 AUC": 0.4795704557359874,
      "eval_PRM F1 Neg": 0.06451612903225806,
      "eval_PRM NPV": 0.125,
      "eval_PRM Precision": 0.7755102040816326,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.043478260869565216,
      "eval_loss": 0.6287286877632141,
      "eval_runtime": 21.0503,
      "eval_samples_per_second": 2.09,
      "eval_steps_per_second": 0.143,
      "step": 2025
    },
    {
      "epoch": 0.2622908372981196,
      "grad_norm": 0.14550023581490018,
      "learning_rate": 1.1524293023350133e-06,
      "loss": 0.7924,
      "step": 2026
    },
    {
      "epoch": 0.26242029970547304,
      "grad_norm": 0.2436683445551189,
      "learning_rate": 1.1522776937619189e-06,
      "loss": 0.7759,
      "step": 2027
    },
    {
      "epoch": 0.2625497621128265,
      "grad_norm": 0.14193629941167152,
      "learning_rate": 1.152125977481668e-06,
      "loss": 0.6581,
      "step": 2028
    },
    {
      "epoch": 0.26267922452017994,
      "grad_norm": 0.1275687661550096,
      "learning_rate": 1.151974153525252e-06,
      "loss": 0.8688,
      "step": 2029
    },
    {
      "epoch": 0.2628086869275334,
      "grad_norm": 0.1464015221492921,
      "learning_rate": 1.1518222219236839e-06,
      "loss": 0.7217,
      "step": 2030
    },
    {
      "epoch": 0.2628086869275334,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6290838122367859,
      "eval_runtime": 20.3334,
      "eval_samples_per_second": 2.164,
      "eval_steps_per_second": 0.148,
      "step": 2030
    },
    {
      "epoch": 0.2629381493348869,
      "grad_norm": 0.17790912853679774,
      "learning_rate": 1.1516701827079986e-06,
      "loss": 0.8655,
      "step": 2031
    },
    {
      "epoch": 0.26306761174224036,
      "grad_norm": 0.12056052874758104,
      "learning_rate": 1.1515180359092534e-06,
      "loss": 0.7825,
      "step": 2032
    },
    {
      "epoch": 0.2631970741495938,
      "grad_norm": 0.13417653505407964,
      "learning_rate": 1.1513657815585276e-06,
      "loss": 0.7643,
      "step": 2033
    },
    {
      "epoch": 0.26332653655694727,
      "grad_norm": 0.19444922059745784,
      "learning_rate": 1.1512134196869214e-06,
      "loss": 1.0112,
      "step": 2034
    },
    {
      "epoch": 0.2634559989643007,
      "grad_norm": 0.1522841990011029,
      "learning_rate": 1.151060950325559e-06,
      "loss": 0.8878,
      "step": 2035
    },
    {
      "epoch": 0.2634559989643007,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6317471861839294,
      "eval_runtime": 19.5914,
      "eval_samples_per_second": 2.246,
      "eval_steps_per_second": 0.153,
      "step": 2035
    },
    {
      "epoch": 0.26358546137165423,
      "grad_norm": 0.13578549393755698,
      "learning_rate": 1.1509083735055843e-06,
      "loss": 0.7694,
      "step": 2036
    },
    {
      "epoch": 0.2637149237790077,
      "grad_norm": 0.13854295204570605,
      "learning_rate": 1.1507556892581648e-06,
      "loss": 0.7623,
      "step": 2037
    },
    {
      "epoch": 0.26384438618636114,
      "grad_norm": 0.14994794738488412,
      "learning_rate": 1.1506028976144892e-06,
      "loss": 0.7786,
      "step": 2038
    },
    {
      "epoch": 0.2639738485937146,
      "grad_norm": 0.10843924933641072,
      "learning_rate": 1.1504499986057682e-06,
      "loss": 0.7245,
      "step": 2039
    },
    {
      "epoch": 0.26410331100106804,
      "grad_norm": 0.16390675222756576,
      "learning_rate": 1.1502969922632345e-06,
      "loss": 0.798,
      "step": 2040
    },
    {
      "epoch": 0.26410331100106804,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8397790055248618,
      "eval_PRM F1 AUC": 0.4795704557359874,
      "eval_PRM F1 Neg": 0.06451612903225806,
      "eval_PRM NPV": 0.125,
      "eval_PRM Precision": 0.7755102040816326,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.043478260869565216,
      "eval_loss": 0.6294389367103577,
      "eval_runtime": 20.4459,
      "eval_samples_per_second": 2.152,
      "eval_steps_per_second": 0.147,
      "step": 2040
    },
    {
      "epoch": 0.26423277340842155,
      "grad_norm": 0.13787470234240212,
      "learning_rate": 1.1501438786181432e-06,
      "loss": 0.9595,
      "step": 2041
    },
    {
      "epoch": 0.264362235815775,
      "grad_norm": 0.21742547403424145,
      "learning_rate": 1.1499906577017703e-06,
      "loss": 0.7463,
      "step": 2042
    },
    {
      "epoch": 0.26449169822312846,
      "grad_norm": 0.15585781149643402,
      "learning_rate": 1.1498373295454142e-06,
      "loss": 0.807,
      "step": 2043
    },
    {
      "epoch": 0.2646211606304819,
      "grad_norm": 0.14063825696141274,
      "learning_rate": 1.149683894180396e-06,
      "loss": 0.9086,
      "step": 2044
    },
    {
      "epoch": 0.26475062303783536,
      "grad_norm": 0.13734097806472706,
      "learning_rate": 1.1495303516380575e-06,
      "loss": 0.7955,
      "step": 2045
    },
    {
      "epoch": 0.26475062303783536,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8397790055248618,
      "eval_PRM F1 AUC": 0.4795704557359874,
      "eval_PRM F1 Neg": 0.06451612903225806,
      "eval_PRM NPV": 0.125,
      "eval_PRM Precision": 0.7755102040816326,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.043478260869565216,
      "eval_loss": 0.6290838122367859,
      "eval_runtime": 19.7683,
      "eval_samples_per_second": 2.226,
      "eval_steps_per_second": 0.152,
      "step": 2045
    },
    {
      "epoch": 0.2648800854451889,
      "grad_norm": 0.14094162893931325,
      "learning_rate": 1.1493767019497628e-06,
      "loss": 0.7335,
      "step": 2046
    },
    {
      "epoch": 0.2650095478525423,
      "grad_norm": 0.14225046027056987,
      "learning_rate": 1.149222945146898e-06,
      "loss": 0.7166,
      "step": 2047
    },
    {
      "epoch": 0.2651390102598958,
      "grad_norm": 0.19976704310314186,
      "learning_rate": 1.1490690812608714e-06,
      "loss": 0.7103,
      "step": 2048
    },
    {
      "epoch": 0.26526847266724923,
      "grad_norm": 0.2434710822445155,
      "learning_rate": 1.1489151103231125e-06,
      "loss": 0.6697,
      "step": 2049
    },
    {
      "epoch": 0.2653979350746027,
      "grad_norm": 0.13861505650283112,
      "learning_rate": 1.148761032365073e-06,
      "loss": 0.8972,
      "step": 2050
    },
    {
      "epoch": 0.2653979350746027,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.62890625,
      "eval_runtime": 20.0381,
      "eval_samples_per_second": 2.196,
      "eval_steps_per_second": 0.15,
      "step": 2050
    },
    {
      "epoch": 0.2655273974819562,
      "grad_norm": 0.243835942786741,
      "learning_rate": 1.1486068474182264e-06,
      "loss": 0.7112,
      "step": 2051
    },
    {
      "epoch": 0.26565685988930965,
      "grad_norm": 0.17563423103450296,
      "learning_rate": 1.1484525555140682e-06,
      "loss": 0.9773,
      "step": 2052
    },
    {
      "epoch": 0.2657863222966631,
      "grad_norm": 0.13972882853763716,
      "learning_rate": 1.1482981566841156e-06,
      "loss": 0.8161,
      "step": 2053
    },
    {
      "epoch": 0.26591578470401656,
      "grad_norm": 0.24629559453165226,
      "learning_rate": 1.148143650959908e-06,
      "loss": 0.7847,
      "step": 2054
    },
    {
      "epoch": 0.26604524711137,
      "grad_norm": 0.12580338552201048,
      "learning_rate": 1.1479890383730059e-06,
      "loss": 0.922,
      "step": 2055
    },
    {
      "epoch": 0.26604524711137,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8397790055248618,
      "eval_PRM F1 AUC": 0.4795704557359874,
      "eval_PRM F1 Neg": 0.06451612903225806,
      "eval_PRM NPV": 0.125,
      "eval_PRM Precision": 0.7755102040816326,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.043478260869565216,
      "eval_loss": 0.62890625,
      "eval_runtime": 20.0951,
      "eval_samples_per_second": 2.19,
      "eval_steps_per_second": 0.149,
      "step": 2055
    },
    {
      "epoch": 0.2661747095187235,
      "grad_norm": 0.21113541841364264,
      "learning_rate": 1.147834318954992e-06,
      "loss": 0.6624,
      "step": 2056
    },
    {
      "epoch": 0.26630417192607697,
      "grad_norm": 0.17868179002842696,
      "learning_rate": 1.1476794927374713e-06,
      "loss": 0.8654,
      "step": 2057
    },
    {
      "epoch": 0.2664336343334304,
      "grad_norm": 0.22243953984346007,
      "learning_rate": 1.14752455975207e-06,
      "loss": 0.7401,
      "step": 2058
    },
    {
      "epoch": 0.2665630967407839,
      "grad_norm": 0.1562921321633234,
      "learning_rate": 1.1473695200304362e-06,
      "loss": 0.6879,
      "step": 2059
    },
    {
      "epoch": 0.2666925591481374,
      "grad_norm": 0.18476061907075264,
      "learning_rate": 1.1472143736042398e-06,
      "loss": 0.7897,
      "step": 2060
    },
    {
      "epoch": 0.2666925591481374,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8397790055248618,
      "eval_PRM F1 AUC": 0.4795704557359874,
      "eval_PRM F1 Neg": 0.06451612903225806,
      "eval_PRM NPV": 0.125,
      "eval_PRM Precision": 0.7755102040816326,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.043478260869565216,
      "eval_loss": 0.6271306872367859,
      "eval_runtime": 20.7885,
      "eval_samples_per_second": 2.117,
      "eval_steps_per_second": 0.144,
      "step": 2060
    },
    {
      "epoch": 0.26682202155549084,
      "grad_norm": 0.11284110118497762,
      "learning_rate": 1.1470591205051731e-06,
      "loss": 0.9042,
      "step": 2061
    },
    {
      "epoch": 0.2669514839628443,
      "grad_norm": 0.1315748169681488,
      "learning_rate": 1.1469037607649494e-06,
      "loss": 0.6456,
      "step": 2062
    },
    {
      "epoch": 0.26708094637019775,
      "grad_norm": 0.15834110318139566,
      "learning_rate": 1.1467482944153039e-06,
      "loss": 0.908,
      "step": 2063
    },
    {
      "epoch": 0.2672104087775512,
      "grad_norm": 0.11799634584539614,
      "learning_rate": 1.1465927214879938e-06,
      "loss": 0.771,
      "step": 2064
    },
    {
      "epoch": 0.2673398711849047,
      "grad_norm": 0.16888478458417452,
      "learning_rate": 1.1464370420147984e-06,
      "loss": 0.7589,
      "step": 2065
    },
    {
      "epoch": 0.2673398711849047,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6303266882896423,
      "eval_runtime": 19.6744,
      "eval_samples_per_second": 2.236,
      "eval_steps_per_second": 0.152,
      "step": 2065
    },
    {
      "epoch": 0.26746933359225816,
      "grad_norm": 0.13034443792198333,
      "learning_rate": 1.146281256027518e-06,
      "loss": 0.925,
      "step": 2066
    },
    {
      "epoch": 0.2675987959996116,
      "grad_norm": 0.13851462771680342,
      "learning_rate": 1.146125363557975e-06,
      "loss": 0.799,
      "step": 2067
    },
    {
      "epoch": 0.26772825840696507,
      "grad_norm": 0.15862172829061857,
      "learning_rate": 1.145969364638014e-06,
      "loss": 0.929,
      "step": 2068
    },
    {
      "epoch": 0.2678577208143185,
      "grad_norm": 0.16293507712891653,
      "learning_rate": 1.1458132592995005e-06,
      "loss": 0.937,
      "step": 2069
    },
    {
      "epoch": 0.26798718322167203,
      "grad_norm": 0.13826581638690566,
      "learning_rate": 1.1456570475743225e-06,
      "loss": 0.8674,
      "step": 2070
    },
    {
      "epoch": 0.26798718322167203,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6283735632896423,
      "eval_runtime": 20.8278,
      "eval_samples_per_second": 2.113,
      "eval_steps_per_second": 0.144,
      "step": 2070
    },
    {
      "epoch": 0.2681166456290255,
      "grad_norm": 0.13469938592642924,
      "learning_rate": 1.1455007294943895e-06,
      "loss": 0.8718,
      "step": 2071
    },
    {
      "epoch": 0.26824610803637894,
      "grad_norm": 0.11913643616582607,
      "learning_rate": 1.1453443050916322e-06,
      "loss": 0.8238,
      "step": 2072
    },
    {
      "epoch": 0.2683755704437324,
      "grad_norm": 0.12451029738114792,
      "learning_rate": 1.1451877743980036e-06,
      "loss": 0.8052,
      "step": 2073
    },
    {
      "epoch": 0.26850503285108585,
      "grad_norm": 0.11661993246999278,
      "learning_rate": 1.1450311374454786e-06,
      "loss": 0.7682,
      "step": 2074
    },
    {
      "epoch": 0.26863449525843935,
      "grad_norm": 0.11806072916892774,
      "learning_rate": 1.1448743942660532e-06,
      "loss": 0.7954,
      "step": 2075
    },
    {
      "epoch": 0.26863449525843935,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6278409361839294,
      "eval_runtime": 20.0346,
      "eval_samples_per_second": 2.196,
      "eval_steps_per_second": 0.15,
      "step": 2075
    },
    {
      "epoch": 0.2687639576657928,
      "grad_norm": 0.17452596634115977,
      "learning_rate": 1.1447175448917452e-06,
      "loss": 0.8247,
      "step": 2076
    },
    {
      "epoch": 0.26889342007314626,
      "grad_norm": 0.24841548939942049,
      "learning_rate": 1.1445605893545946e-06,
      "loss": 0.7986,
      "step": 2077
    },
    {
      "epoch": 0.2690228824804997,
      "grad_norm": 0.1492423260526689,
      "learning_rate": 1.1444035276866625e-06,
      "loss": 0.7856,
      "step": 2078
    },
    {
      "epoch": 0.26915234488785317,
      "grad_norm": 0.15051238127958882,
      "learning_rate": 1.144246359920032e-06,
      "loss": 0.7749,
      "step": 2079
    },
    {
      "epoch": 0.2692818072952067,
      "grad_norm": 0.1364043029930431,
      "learning_rate": 1.1440890860868082e-06,
      "loss": 0.8076,
      "step": 2080
    },
    {
      "epoch": 0.2692818072952067,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6264204382896423,
      "eval_runtime": 19.7757,
      "eval_samples_per_second": 2.225,
      "eval_steps_per_second": 0.152,
      "step": 2080
    },
    {
      "epoch": 0.26941126970256013,
      "grad_norm": 0.13353098695003038,
      "learning_rate": 1.1439317062191168e-06,
      "loss": 0.9014,
      "step": 2081
    },
    {
      "epoch": 0.2695407321099136,
      "grad_norm": 0.15019240341274687,
      "learning_rate": 1.1437742203491062e-06,
      "loss": 0.8453,
      "step": 2082
    },
    {
      "epoch": 0.26967019451726704,
      "grad_norm": 0.19669012615416093,
      "learning_rate": 1.1436166285089458e-06,
      "loss": 0.7369,
      "step": 2083
    },
    {
      "epoch": 0.2697996569246205,
      "grad_norm": 0.128352566024135,
      "learning_rate": 1.1434589307308275e-06,
      "loss": 0.8274,
      "step": 2084
    },
    {
      "epoch": 0.269929119331974,
      "grad_norm": 0.11852638320446639,
      "learning_rate": 1.1433011270469638e-06,
      "loss": 0.8528,
      "step": 2085
    },
    {
      "epoch": 0.269929119331974,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.62890625,
      "eval_runtime": 20.4541,
      "eval_samples_per_second": 2.151,
      "eval_steps_per_second": 0.147,
      "step": 2085
    },
    {
      "epoch": 0.27005858173932745,
      "grad_norm": 0.12286183889937305,
      "learning_rate": 1.1431432174895894e-06,
      "loss": 0.7715,
      "step": 2086
    },
    {
      "epoch": 0.2701880441466809,
      "grad_norm": 0.189585778305281,
      "learning_rate": 1.1429852020909603e-06,
      "loss": 0.9607,
      "step": 2087
    },
    {
      "epoch": 0.27031750655403436,
      "grad_norm": 0.14444807234281665,
      "learning_rate": 1.142827080883355e-06,
      "loss": 0.7231,
      "step": 2088
    },
    {
      "epoch": 0.2704469689613878,
      "grad_norm": 0.13215926277744616,
      "learning_rate": 1.1426688538990724e-06,
      "loss": 0.7935,
      "step": 2089
    },
    {
      "epoch": 0.2705764313687413,
      "grad_norm": 0.17935968612034753,
      "learning_rate": 1.1425105211704338e-06,
      "loss": 0.8237,
      "step": 2090
    },
    {
      "epoch": 0.2705764313687413,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6278409361839294,
      "eval_runtime": 20.4283,
      "eval_samples_per_second": 2.154,
      "eval_steps_per_second": 0.147,
      "step": 2090
    },
    {
      "epoch": 0.2707058937760948,
      "grad_norm": 0.1523108667693704,
      "learning_rate": 1.1423520827297814e-06,
      "loss": 0.7833,
      "step": 2091
    },
    {
      "epoch": 0.27083535618344823,
      "grad_norm": 0.12134428215595013,
      "learning_rate": 1.1421935386094803e-06,
      "loss": 0.8484,
      "step": 2092
    },
    {
      "epoch": 0.2709648185908017,
      "grad_norm": 0.18506670224088292,
      "learning_rate": 1.1420348888419158e-06,
      "loss": 0.6644,
      "step": 2093
    },
    {
      "epoch": 0.27109428099815513,
      "grad_norm": 0.12407500949135952,
      "learning_rate": 1.1418761334594954e-06,
      "loss": 0.7015,
      "step": 2094
    },
    {
      "epoch": 0.27122374340550864,
      "grad_norm": 0.1709362721159291,
      "learning_rate": 1.1417172724946482e-06,
      "loss": 0.7601,
      "step": 2095
    },
    {
      "epoch": 0.27122374340550864,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.62890625,
      "eval_runtime": 20.1812,
      "eval_samples_per_second": 2.18,
      "eval_steps_per_second": 0.149,
      "step": 2095
    },
    {
      "epoch": 0.2713532058128621,
      "grad_norm": 0.1614746521877732,
      "learning_rate": 1.1415583059798247e-06,
      "loss": 0.9598,
      "step": 2096
    },
    {
      "epoch": 0.27148266822021555,
      "grad_norm": 0.15712883319962734,
      "learning_rate": 1.1413992339474968e-06,
      "loss": 0.8398,
      "step": 2097
    },
    {
      "epoch": 0.271612130627569,
      "grad_norm": 0.14111020022794962,
      "learning_rate": 1.1412400564301587e-06,
      "loss": 0.899,
      "step": 2098
    },
    {
      "epoch": 0.27174159303492246,
      "grad_norm": 0.16266019139291651,
      "learning_rate": 1.1410807734603251e-06,
      "loss": 0.7275,
      "step": 2099
    },
    {
      "epoch": 0.27187105544227597,
      "grad_norm": 0.1447843382277028,
      "learning_rate": 1.1409213850705334e-06,
      "loss": 0.8463,
      "step": 2100
    },
    {
      "epoch": 0.27187105544227597,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6271306872367859,
      "eval_runtime": 20.4168,
      "eval_samples_per_second": 2.155,
      "eval_steps_per_second": 0.147,
      "step": 2100
    },
    {
      "epoch": 0.2720005178496294,
      "grad_norm": 0.16816794297321921,
      "learning_rate": 1.1407618912933414e-06,
      "loss": 0.798,
      "step": 2101
    },
    {
      "epoch": 0.2721299802569829,
      "grad_norm": 0.13118467959444624,
      "learning_rate": 1.140602292161329e-06,
      "loss": 0.8379,
      "step": 2102
    },
    {
      "epoch": 0.2722594426643363,
      "grad_norm": 0.12813423602124158,
      "learning_rate": 1.1404425877070977e-06,
      "loss": 0.8442,
      "step": 2103
    },
    {
      "epoch": 0.27238890507168984,
      "grad_norm": 0.1590888270183565,
      "learning_rate": 1.1402827779632702e-06,
      "loss": 0.7507,
      "step": 2104
    },
    {
      "epoch": 0.2725183674790433,
      "grad_norm": 0.13425098678688283,
      "learning_rate": 1.1401228629624912e-06,
      "loss": 0.7766,
      "step": 2105
    },
    {
      "epoch": 0.2725183674790433,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6278409361839294,
      "eval_runtime": 19.3347,
      "eval_samples_per_second": 2.276,
      "eval_steps_per_second": 0.155,
      "step": 2105
    },
    {
      "epoch": 0.27264782988639674,
      "grad_norm": 0.11918205337250133,
      "learning_rate": 1.1399628427374262e-06,
      "loss": 0.7654,
      "step": 2106
    },
    {
      "epoch": 0.2727772922937502,
      "grad_norm": 0.11007301258573662,
      "learning_rate": 1.139802717320763e-06,
      "loss": 0.7507,
      "step": 2107
    },
    {
      "epoch": 0.27290675470110365,
      "grad_norm": 0.15291913592859657,
      "learning_rate": 1.1396424867452099e-06,
      "loss": 0.9532,
      "step": 2108
    },
    {
      "epoch": 0.27303621710845716,
      "grad_norm": 0.1489892642537092,
      "learning_rate": 1.139482151043498e-06,
      "loss": 0.847,
      "step": 2109
    },
    {
      "epoch": 0.2731656795158106,
      "grad_norm": 0.1479702947644868,
      "learning_rate": 1.139321710248378e-06,
      "loss": 0.7797,
      "step": 2110
    },
    {
      "epoch": 0.2731656795158106,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6310369372367859,
      "eval_runtime": 19.8801,
      "eval_samples_per_second": 2.213,
      "eval_steps_per_second": 0.151,
      "step": 2110
    },
    {
      "epoch": 0.27329514192316406,
      "grad_norm": 0.1310554951012864,
      "learning_rate": 1.1391611643926243e-06,
      "loss": 0.7784,
      "step": 2111
    },
    {
      "epoch": 0.2734246043305175,
      "grad_norm": 0.1230669224562225,
      "learning_rate": 1.1390005135090313e-06,
      "loss": 0.8855,
      "step": 2112
    },
    {
      "epoch": 0.27355406673787097,
      "grad_norm": 0.11268676488856726,
      "learning_rate": 1.1388397576304148e-06,
      "loss": 0.7441,
      "step": 2113
    },
    {
      "epoch": 0.2736835291452245,
      "grad_norm": 0.17806787128522342,
      "learning_rate": 1.138678896789613e-06,
      "loss": 0.9554,
      "step": 2114
    },
    {
      "epoch": 0.27381299155257793,
      "grad_norm": 0.1252606649075309,
      "learning_rate": 1.1385179310194844e-06,
      "loss": 0.7615,
      "step": 2115
    },
    {
      "epoch": 0.27381299155257793,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.626953125,
      "eval_runtime": 19.8693,
      "eval_samples_per_second": 2.214,
      "eval_steps_per_second": 0.151,
      "step": 2115
    },
    {
      "epoch": 0.2739424539599314,
      "grad_norm": 0.12318046596180698,
      "learning_rate": 1.1383568603529101e-06,
      "loss": 0.8226,
      "step": 2116
    },
    {
      "epoch": 0.27407191636728484,
      "grad_norm": 0.13410334863288612,
      "learning_rate": 1.138195684822792e-06,
      "loss": 0.857,
      "step": 2117
    },
    {
      "epoch": 0.2742013787746383,
      "grad_norm": 0.11551928944667572,
      "learning_rate": 1.1380344044620529e-06,
      "loss": 0.749,
      "step": 2118
    },
    {
      "epoch": 0.2743308411819918,
      "grad_norm": 0.11630817450828106,
      "learning_rate": 1.137873019303638e-06,
      "loss": 0.8356,
      "step": 2119
    },
    {
      "epoch": 0.27446030358934526,
      "grad_norm": 0.12761558831240527,
      "learning_rate": 1.1377115293805136e-06,
      "loss": 0.7997,
      "step": 2120
    },
    {
      "epoch": 0.27446030358934526,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6285511255264282,
      "eval_runtime": 20.557,
      "eval_samples_per_second": 2.14,
      "eval_steps_per_second": 0.146,
      "step": 2120
    },
    {
      "epoch": 0.2745897659966987,
      "grad_norm": 0.14464603936749176,
      "learning_rate": 1.137549934725667e-06,
      "loss": 0.9263,
      "step": 2121
    },
    {
      "epoch": 0.27471922840405216,
      "grad_norm": 0.20626137519668034,
      "learning_rate": 1.1373882353721074e-06,
      "loss": 0.8785,
      "step": 2122
    },
    {
      "epoch": 0.2748486908114056,
      "grad_norm": 0.17515953480208737,
      "learning_rate": 1.1372264313528652e-06,
      "loss": 0.7586,
      "step": 2123
    },
    {
      "epoch": 0.2749781532187591,
      "grad_norm": 0.18864761037051608,
      "learning_rate": 1.137064522700992e-06,
      "loss": 0.7776,
      "step": 2124
    },
    {
      "epoch": 0.2751076156261126,
      "grad_norm": 0.18644543618586593,
      "learning_rate": 1.1369025094495606e-06,
      "loss": 0.7802,
      "step": 2125
    },
    {
      "epoch": 0.2751076156261126,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6296164989471436,
      "eval_runtime": 19.884,
      "eval_samples_per_second": 2.213,
      "eval_steps_per_second": 0.151,
      "step": 2125
    },
    {
      "epoch": 0.27523707803346603,
      "grad_norm": 0.139575319031462,
      "learning_rate": 1.1367403916316663e-06,
      "loss": 0.7444,
      "step": 2126
    },
    {
      "epoch": 0.2753665404408195,
      "grad_norm": 0.21021141106371222,
      "learning_rate": 1.1365781692804242e-06,
      "loss": 0.8224,
      "step": 2127
    },
    {
      "epoch": 0.27549600284817294,
      "grad_norm": 0.14418676206369596,
      "learning_rate": 1.1364158424289721e-06,
      "loss": 0.7671,
      "step": 2128
    },
    {
      "epoch": 0.27562546525552645,
      "grad_norm": 0.11783695977888403,
      "learning_rate": 1.136253411110468e-06,
      "loss": 0.7949,
      "step": 2129
    },
    {
      "epoch": 0.2757549276628799,
      "grad_norm": 0.13305132599151867,
      "learning_rate": 1.1360908753580921e-06,
      "loss": 0.8169,
      "step": 2130
    },
    {
      "epoch": 0.2757549276628799,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6306818127632141,
      "eval_runtime": 19.7787,
      "eval_samples_per_second": 2.225,
      "eval_steps_per_second": 0.152,
      "step": 2130
    },
    {
      "epoch": 0.27588439007023335,
      "grad_norm": 0.18931445841393957,
      "learning_rate": 1.1359282352050458e-06,
      "loss": 0.8027,
      "step": 2131
    },
    {
      "epoch": 0.2760138524775868,
      "grad_norm": 0.172932780389384,
      "learning_rate": 1.135765490684551e-06,
      "loss": 0.9933,
      "step": 2132
    },
    {
      "epoch": 0.27614331488494026,
      "grad_norm": 0.2830729318960173,
      "learning_rate": 1.1356026418298525e-06,
      "loss": 0.7217,
      "step": 2133
    },
    {
      "epoch": 0.27627277729229377,
      "grad_norm": 0.14090690594538988,
      "learning_rate": 1.1354396886742144e-06,
      "loss": 0.8502,
      "step": 2134
    },
    {
      "epoch": 0.2764022396996472,
      "grad_norm": 0.15150992761790058,
      "learning_rate": 1.1352766312509242e-06,
      "loss": 0.8988,
      "step": 2135
    },
    {
      "epoch": 0.2764022396996472,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8397790055248618,
      "eval_PRM F1 AUC": 0.4795704557359874,
      "eval_PRM F1 Neg": 0.06451612903225806,
      "eval_PRM NPV": 0.125,
      "eval_PRM Precision": 0.7755102040816326,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.043478260869565216,
      "eval_loss": 0.6244673132896423,
      "eval_runtime": 19.9879,
      "eval_samples_per_second": 2.201,
      "eval_steps_per_second": 0.15,
      "step": 2135
    },
    {
      "epoch": 0.2765317021070007,
      "grad_norm": 0.14325901277506736,
      "learning_rate": 1.135113469593289e-06,
      "loss": 0.77,
      "step": 2136
    },
    {
      "epoch": 0.27666116451435413,
      "grad_norm": 0.12717405834219583,
      "learning_rate": 1.134950203734638e-06,
      "loss": 0.8557,
      "step": 2137
    },
    {
      "epoch": 0.2767906269217076,
      "grad_norm": 0.1229603672018342,
      "learning_rate": 1.134786833708322e-06,
      "loss": 0.7648,
      "step": 2138
    },
    {
      "epoch": 0.2769200893290611,
      "grad_norm": 0.17853372293399467,
      "learning_rate": 1.134623359547712e-06,
      "loss": 0.7151,
      "step": 2139
    },
    {
      "epoch": 0.27704955173641455,
      "grad_norm": 0.13628628776695148,
      "learning_rate": 1.1344597812862016e-06,
      "loss": 0.9323,
      "step": 2140
    },
    {
      "epoch": 0.27704955173641455,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6271306872367859,
      "eval_runtime": 19.979,
      "eval_samples_per_second": 2.202,
      "eval_steps_per_second": 0.15,
      "step": 2140
    },
    {
      "epoch": 0.277179014143768,
      "grad_norm": 0.13210845553674672,
      "learning_rate": 1.1342960989572041e-06,
      "loss": 0.7729,
      "step": 2141
    },
    {
      "epoch": 0.27730847655112145,
      "grad_norm": 0.1941621639977534,
      "learning_rate": 1.1341323125941559e-06,
      "loss": 0.7979,
      "step": 2142
    },
    {
      "epoch": 0.2774379389584749,
      "grad_norm": 0.14408261098565486,
      "learning_rate": 1.1339684222305127e-06,
      "loss": 0.827,
      "step": 2143
    },
    {
      "epoch": 0.2775674013658284,
      "grad_norm": 0.1298448844184219,
      "learning_rate": 1.1338044278997535e-06,
      "loss": 0.8137,
      "step": 2144
    },
    {
      "epoch": 0.27769686377318187,
      "grad_norm": 0.19007673074960652,
      "learning_rate": 1.1336403296353765e-06,
      "loss": 0.7045,
      "step": 2145
    },
    {
      "epoch": 0.27769686377318187,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6292613744735718,
      "eval_runtime": 20.1222,
      "eval_samples_per_second": 2.187,
      "eval_steps_per_second": 0.149,
      "step": 2145
    },
    {
      "epoch": 0.2778263261805353,
      "grad_norm": 0.13988829026117985,
      "learning_rate": 1.1334761274709026e-06,
      "loss": 0.9229,
      "step": 2146
    },
    {
      "epoch": 0.2779557885878888,
      "grad_norm": 0.20459886729658713,
      "learning_rate": 1.1333118214398733e-06,
      "loss": 0.723,
      "step": 2147
    },
    {
      "epoch": 0.2780852509952423,
      "grad_norm": 0.17585070843642192,
      "learning_rate": 1.1331474115758517e-06,
      "loss": 0.6464,
      "step": 2148
    },
    {
      "epoch": 0.27821471340259574,
      "grad_norm": 0.12977116199598965,
      "learning_rate": 1.1329828979124212e-06,
      "loss": 0.7557,
      "step": 2149
    },
    {
      "epoch": 0.2783441758099492,
      "grad_norm": 0.15032108047815024,
      "learning_rate": 1.1328182804831878e-06,
      "loss": 0.6954,
      "step": 2150
    },
    {
      "epoch": 0.2783441758099492,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6274858117103577,
      "eval_runtime": 20.9644,
      "eval_samples_per_second": 2.099,
      "eval_steps_per_second": 0.143,
      "step": 2150
    },
    {
      "epoch": 0.27847363821730264,
      "grad_norm": 0.11203669081268536,
      "learning_rate": 1.1326535593217774e-06,
      "loss": 0.8064,
      "step": 2151
    },
    {
      "epoch": 0.2786031006246561,
      "grad_norm": 0.13088902478203138,
      "learning_rate": 1.132488734461838e-06,
      "loss": 0.6674,
      "step": 2152
    },
    {
      "epoch": 0.2787325630320096,
      "grad_norm": 0.14364781859839337,
      "learning_rate": 1.132323805937038e-06,
      "loss": 0.8586,
      "step": 2153
    },
    {
      "epoch": 0.27886202543936306,
      "grad_norm": 0.16676564603853666,
      "learning_rate": 1.132158773781068e-06,
      "loss": 0.8877,
      "step": 2154
    },
    {
      "epoch": 0.2789914878467165,
      "grad_norm": 0.17855427563203882,
      "learning_rate": 1.1319936380276386e-06,
      "loss": 0.8345,
      "step": 2155
    },
    {
      "epoch": 0.2789914878467165,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6287286877632141,
      "eval_runtime": 20.1106,
      "eval_samples_per_second": 2.188,
      "eval_steps_per_second": 0.149,
      "step": 2155
    },
    {
      "epoch": 0.27912095025406997,
      "grad_norm": 0.19282127315013745,
      "learning_rate": 1.1318283987104826e-06,
      "loss": 0.6667,
      "step": 2156
    },
    {
      "epoch": 0.2792504126614234,
      "grad_norm": 0.18264951488084624,
      "learning_rate": 1.131663055863353e-06,
      "loss": 0.7747,
      "step": 2157
    },
    {
      "epoch": 0.27937987506877693,
      "grad_norm": 0.1530118424558895,
      "learning_rate": 1.1314976095200244e-06,
      "loss": 0.7247,
      "step": 2158
    },
    {
      "epoch": 0.2795093374761304,
      "grad_norm": 0.2111613810261186,
      "learning_rate": 1.1313320597142932e-06,
      "loss": 0.7174,
      "step": 2159
    },
    {
      "epoch": 0.27963879988348384,
      "grad_norm": 0.20867011386356343,
      "learning_rate": 1.1311664064799756e-06,
      "loss": 0.9348,
      "step": 2160
    },
    {
      "epoch": 0.27963879988348384,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6296164989471436,
      "eval_runtime": 20.1737,
      "eval_samples_per_second": 2.181,
      "eval_steps_per_second": 0.149,
      "step": 2160
    },
    {
      "epoch": 0.2797682622908373,
      "grad_norm": 0.17592507751543,
      "learning_rate": 1.1310006498509102e-06,
      "loss": 0.7705,
      "step": 2161
    },
    {
      "epoch": 0.27989772469819074,
      "grad_norm": 0.11146912573378448,
      "learning_rate": 1.130834789860956e-06,
      "loss": 0.7518,
      "step": 2162
    },
    {
      "epoch": 0.28002718710554425,
      "grad_norm": 0.14252157624724704,
      "learning_rate": 1.130668826543993e-06,
      "loss": 0.7271,
      "step": 2163
    },
    {
      "epoch": 0.2801566495128977,
      "grad_norm": 0.12085950576452345,
      "learning_rate": 1.1305027599339228e-06,
      "loss": 0.8284,
      "step": 2164
    },
    {
      "epoch": 0.28028611192025116,
      "grad_norm": 0.23426118582449756,
      "learning_rate": 1.1303365900646675e-06,
      "loss": 0.6704,
      "step": 2165
    },
    {
      "epoch": 0.28028611192025116,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6274858117103577,
      "eval_runtime": 20.0088,
      "eval_samples_per_second": 2.199,
      "eval_steps_per_second": 0.15,
      "step": 2165
    },
    {
      "epoch": 0.2804155743276046,
      "grad_norm": 0.15291588149530186,
      "learning_rate": 1.1301703169701713e-06,
      "loss": 0.7692,
      "step": 2166
    },
    {
      "epoch": 0.28054503673495806,
      "grad_norm": 0.14980326104750438,
      "learning_rate": 1.1300039406843985e-06,
      "loss": 0.7037,
      "step": 2167
    },
    {
      "epoch": 0.2806744991423116,
      "grad_norm": 0.17625111940719115,
      "learning_rate": 1.129837461241335e-06,
      "loss": 0.8285,
      "step": 2168
    },
    {
      "epoch": 0.280803961549665,
      "grad_norm": 0.1512848661727698,
      "learning_rate": 1.1296708786749872e-06,
      "loss": 0.8796,
      "step": 2169
    },
    {
      "epoch": 0.2809334239570185,
      "grad_norm": 0.13198573961035262,
      "learning_rate": 1.129504193019383e-06,
      "loss": 0.8059,
      "step": 2170
    },
    {
      "epoch": 0.2809334239570185,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6285511255264282,
      "eval_runtime": 20.549,
      "eval_samples_per_second": 2.141,
      "eval_steps_per_second": 0.146,
      "step": 2170
    },
    {
      "epoch": 0.28106288636437193,
      "grad_norm": 0.17015552710486695,
      "learning_rate": 1.129337404308572e-06,
      "loss": 0.9099,
      "step": 2171
    },
    {
      "epoch": 0.2811923487717254,
      "grad_norm": 0.17010425945712626,
      "learning_rate": 1.1291705125766235e-06,
      "loss": 0.8105,
      "step": 2172
    },
    {
      "epoch": 0.2813218111790789,
      "grad_norm": 0.16460499535033168,
      "learning_rate": 1.1290035178576288e-06,
      "loss": 0.7623,
      "step": 2173
    },
    {
      "epoch": 0.28145127358643235,
      "grad_norm": 0.14629761036833386,
      "learning_rate": 1.1288364201856998e-06,
      "loss": 0.8517,
      "step": 2174
    },
    {
      "epoch": 0.2815807359937858,
      "grad_norm": 0.14684758360977557,
      "learning_rate": 1.1286692195949696e-06,
      "loss": 0.936,
      "step": 2175
    },
    {
      "epoch": 0.2815807359937858,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6290838122367859,
      "eval_runtime": 20.2676,
      "eval_samples_per_second": 2.171,
      "eval_steps_per_second": 0.148,
      "step": 2175
    },
    {
      "epoch": 0.28171019840113926,
      "grad_norm": 0.1268981011739878,
      "learning_rate": 1.1285019161195926e-06,
      "loss": 0.7357,
      "step": 2176
    },
    {
      "epoch": 0.2818396608084927,
      "grad_norm": 0.2072503193227775,
      "learning_rate": 1.1283345097937438e-06,
      "loss": 0.9722,
      "step": 2177
    },
    {
      "epoch": 0.2819691232158462,
      "grad_norm": 0.11052142550498867,
      "learning_rate": 1.128167000651619e-06,
      "loss": 0.7484,
      "step": 2178
    },
    {
      "epoch": 0.28209858562319967,
      "grad_norm": 0.14107822277105744,
      "learning_rate": 1.1279993887274356e-06,
      "loss": 0.7186,
      "step": 2179
    },
    {
      "epoch": 0.2822280480305531,
      "grad_norm": 0.13689293294583718,
      "learning_rate": 1.1278316740554319e-06,
      "loss": 0.7076,
      "step": 2180
    },
    {
      "epoch": 0.2822280480305531,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6297940611839294,
      "eval_runtime": 19.8631,
      "eval_samples_per_second": 2.215,
      "eval_steps_per_second": 0.151,
      "step": 2180
    },
    {
      "epoch": 0.2823575104379066,
      "grad_norm": 0.12711909025832288,
      "learning_rate": 1.1276638566698667e-06,
      "loss": 0.8727,
      "step": 2181
    },
    {
      "epoch": 0.28248697284526003,
      "grad_norm": 0.11389936009925926,
      "learning_rate": 1.1274959366050203e-06,
      "loss": 0.8318,
      "step": 2182
    },
    {
      "epoch": 0.28261643525261354,
      "grad_norm": 0.1898025408612317,
      "learning_rate": 1.1273279138951937e-06,
      "loss": 0.7837,
      "step": 2183
    },
    {
      "epoch": 0.282745897659967,
      "grad_norm": 0.19257348676507174,
      "learning_rate": 1.127159788574709e-06,
      "loss": 0.7052,
      "step": 2184
    },
    {
      "epoch": 0.28287536006732045,
      "grad_norm": 0.13371177627722391,
      "learning_rate": 1.1269915606779091e-06,
      "loss": 0.9369,
      "step": 2185
    },
    {
      "epoch": 0.28287536006732045,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6287286877632141,
      "eval_runtime": 20.2606,
      "eval_samples_per_second": 2.172,
      "eval_steps_per_second": 0.148,
      "step": 2185
    },
    {
      "epoch": 0.2830048224746739,
      "grad_norm": 0.12178438083270308,
      "learning_rate": 1.126823230239158e-06,
      "loss": 0.6426,
      "step": 2186
    },
    {
      "epoch": 0.28313428488202735,
      "grad_norm": 0.12545452180136973,
      "learning_rate": 1.1266547972928407e-06,
      "loss": 0.7635,
      "step": 2187
    },
    {
      "epoch": 0.28326374728938086,
      "grad_norm": 0.16351322556007397,
      "learning_rate": 1.126486261873363e-06,
      "loss": 0.75,
      "step": 2188
    },
    {
      "epoch": 0.2833932096967343,
      "grad_norm": 0.13663874726596317,
      "learning_rate": 1.1263176240151516e-06,
      "loss": 0.6873,
      "step": 2189
    },
    {
      "epoch": 0.28352267210408777,
      "grad_norm": 0.20613703087689805,
      "learning_rate": 1.1261488837526543e-06,
      "loss": 0.7031,
      "step": 2190
    },
    {
      "epoch": 0.28352267210408777,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6287286877632141,
      "eval_runtime": 20.0487,
      "eval_samples_per_second": 2.195,
      "eval_steps_per_second": 0.15,
      "step": 2190
    },
    {
      "epoch": 0.2836521345114412,
      "grad_norm": 0.11321940022544584,
      "learning_rate": 1.1259800411203395e-06,
      "loss": 0.7563,
      "step": 2191
    },
    {
      "epoch": 0.28378159691879473,
      "grad_norm": 0.12902730882522143,
      "learning_rate": 1.1258110961526973e-06,
      "loss": 0.7975,
      "step": 2192
    },
    {
      "epoch": 0.2839110593261482,
      "grad_norm": 0.11864422893235974,
      "learning_rate": 1.1256420488842373e-06,
      "loss": 0.788,
      "step": 2193
    },
    {
      "epoch": 0.28404052173350164,
      "grad_norm": 0.12583645928117654,
      "learning_rate": 1.1254728993494914e-06,
      "loss": 0.807,
      "step": 2194
    },
    {
      "epoch": 0.2841699841408551,
      "grad_norm": 0.11070742464950291,
      "learning_rate": 1.1253036475830116e-06,
      "loss": 0.8044,
      "step": 2195
    },
    {
      "epoch": 0.2841699841408551,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6262428760528564,
      "eval_runtime": 19.5699,
      "eval_samples_per_second": 2.248,
      "eval_steps_per_second": 0.153,
      "step": 2195
    },
    {
      "epoch": 0.28429944654820855,
      "grad_norm": 0.20373197375006635,
      "learning_rate": 1.1251342936193713e-06,
      "loss": 0.7139,
      "step": 2196
    },
    {
      "epoch": 0.28442890895556205,
      "grad_norm": 0.19654326772008868,
      "learning_rate": 1.124964837493164e-06,
      "loss": 0.6697,
      "step": 2197
    },
    {
      "epoch": 0.2845583713629155,
      "grad_norm": 0.11738856692371957,
      "learning_rate": 1.124795279239005e-06,
      "loss": 0.8017,
      "step": 2198
    },
    {
      "epoch": 0.28468783377026896,
      "grad_norm": 0.2144788988084152,
      "learning_rate": 1.12462561889153e-06,
      "loss": 0.7608,
      "step": 2199
    },
    {
      "epoch": 0.2848172961776224,
      "grad_norm": 0.17335662749691996,
      "learning_rate": 1.1244558564853951e-06,
      "loss": 0.8682,
      "step": 2200
    },
    {
      "epoch": 0.2848172961776224,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6281960010528564,
      "eval_runtime": 19.8948,
      "eval_samples_per_second": 2.212,
      "eval_steps_per_second": 0.151,
      "step": 2200
    },
    {
      "epoch": 0.28494675858497587,
      "grad_norm": 0.14155104249345568,
      "learning_rate": 1.1242859920552782e-06,
      "loss": 0.7178,
      "step": 2201
    },
    {
      "epoch": 0.2850762209923294,
      "grad_norm": 0.13952159037160428,
      "learning_rate": 1.1241160256358771e-06,
      "loss": 0.858,
      "step": 2202
    },
    {
      "epoch": 0.28520568339968283,
      "grad_norm": 0.2234007599938507,
      "learning_rate": 1.1239459572619114e-06,
      "loss": 0.6837,
      "step": 2203
    },
    {
      "epoch": 0.2853351458070363,
      "grad_norm": 0.19479316833492794,
      "learning_rate": 1.1237757869681207e-06,
      "loss": 0.7759,
      "step": 2204
    },
    {
      "epoch": 0.28546460821438974,
      "grad_norm": 0.14248034273017918,
      "learning_rate": 1.1236055147892655e-06,
      "loss": 0.7576,
      "step": 2205
    },
    {
      "epoch": 0.28546460821438974,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6283735632896423,
      "eval_runtime": 20.2075,
      "eval_samples_per_second": 2.177,
      "eval_steps_per_second": 0.148,
      "step": 2205
    },
    {
      "epoch": 0.2855940706217432,
      "grad_norm": 0.12654043060331294,
      "learning_rate": 1.123435140760128e-06,
      "loss": 0.7458,
      "step": 2206
    },
    {
      "epoch": 0.2857235330290967,
      "grad_norm": 0.1666770178619673,
      "learning_rate": 1.12326466491551e-06,
      "loss": 0.7512,
      "step": 2207
    },
    {
      "epoch": 0.28585299543645015,
      "grad_norm": 0.159291610749143,
      "learning_rate": 1.123094087290235e-06,
      "loss": 0.9246,
      "step": 2208
    },
    {
      "epoch": 0.2859824578438036,
      "grad_norm": 0.10790514766520116,
      "learning_rate": 1.1229234079191465e-06,
      "loss": 0.7791,
      "step": 2209
    },
    {
      "epoch": 0.28611192025115706,
      "grad_norm": 0.16261399511341934,
      "learning_rate": 1.1227526268371097e-06,
      "loss": 0.8734,
      "step": 2210
    },
    {
      "epoch": 0.28611192025115706,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6280184388160706,
      "eval_runtime": 19.9502,
      "eval_samples_per_second": 2.205,
      "eval_steps_per_second": 0.15,
      "step": 2210
    },
    {
      "epoch": 0.2862413826585105,
      "grad_norm": 0.1926128592915257,
      "learning_rate": 1.1225817440790099e-06,
      "loss": 0.6973,
      "step": 2211
    },
    {
      "epoch": 0.286370845065864,
      "grad_norm": 0.1495962382821001,
      "learning_rate": 1.1224107596797533e-06,
      "loss": 0.7405,
      "step": 2212
    },
    {
      "epoch": 0.2865003074732175,
      "grad_norm": 0.1094761989890774,
      "learning_rate": 1.1222396736742667e-06,
      "loss": 0.7916,
      "step": 2213
    },
    {
      "epoch": 0.28662976988057093,
      "grad_norm": 0.15878228375440784,
      "learning_rate": 1.1220684860974984e-06,
      "loss": 0.8135,
      "step": 2214
    },
    {
      "epoch": 0.2867592322879244,
      "grad_norm": 0.15208042967516577,
      "learning_rate": 1.1218971969844168e-06,
      "loss": 0.7046,
      "step": 2215
    },
    {
      "epoch": 0.2867592322879244,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6265980005264282,
      "eval_runtime": 20.505,
      "eval_samples_per_second": 2.146,
      "eval_steps_per_second": 0.146,
      "step": 2215
    },
    {
      "epoch": 0.28688869469527783,
      "grad_norm": 0.19451286801802606,
      "learning_rate": 1.1217258063700109e-06,
      "loss": 0.7767,
      "step": 2216
    },
    {
      "epoch": 0.28701815710263134,
      "grad_norm": 0.1858468740017975,
      "learning_rate": 1.121554314289291e-06,
      "loss": 0.7402,
      "step": 2217
    },
    {
      "epoch": 0.2871476195099848,
      "grad_norm": 0.12709260116872906,
      "learning_rate": 1.1213827207772876e-06,
      "loss": 0.715,
      "step": 2218
    },
    {
      "epoch": 0.28727708191733825,
      "grad_norm": 0.20558913300815634,
      "learning_rate": 1.1212110258690522e-06,
      "loss": 0.731,
      "step": 2219
    },
    {
      "epoch": 0.2874065443246917,
      "grad_norm": 0.18254328322275684,
      "learning_rate": 1.1210392295996572e-06,
      "loss": 0.8053,
      "step": 2220
    },
    {
      "epoch": 0.2874065443246917,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8415300546448088,
      "eval_PRM F1 AUC": 0.463855421686747,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.77,
      "eval_PRM Recall": 0.927710843373494,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.630859375,
      "eval_runtime": 20.43,
      "eval_samples_per_second": 2.154,
      "eval_steps_per_second": 0.147,
      "step": 2220
    },
    {
      "epoch": 0.28753600673204516,
      "grad_norm": 0.15078859857257096,
      "learning_rate": 1.1208673320041954e-06,
      "loss": 0.6869,
      "step": 2221
    },
    {
      "epoch": 0.28766546913939867,
      "grad_norm": 0.1634969589266697,
      "learning_rate": 1.12069533311778e-06,
      "loss": 0.7275,
      "step": 2222
    },
    {
      "epoch": 0.2877949315467521,
      "grad_norm": 0.10776252231805807,
      "learning_rate": 1.1205232329755455e-06,
      "loss": 0.7059,
      "step": 2223
    },
    {
      "epoch": 0.2879243939541056,
      "grad_norm": 0.13844318358050542,
      "learning_rate": 1.120351031612647e-06,
      "loss": 0.8542,
      "step": 2224
    },
    {
      "epoch": 0.288053856361459,
      "grad_norm": 0.13395784855418844,
      "learning_rate": 1.1201787290642603e-06,
      "loss": 0.7534,
      "step": 2225
    },
    {
      "epoch": 0.288053856361459,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8415300546448088,
      "eval_PRM F1 AUC": 0.463855421686747,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.77,
      "eval_PRM Recall": 0.927710843373494,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6299715638160706,
      "eval_runtime": 19.4332,
      "eval_samples_per_second": 2.264,
      "eval_steps_per_second": 0.154,
      "step": 2225
    },
    {
      "epoch": 0.2881833187688125,
      "grad_norm": 0.15000003276966598,
      "learning_rate": 1.120006325365581e-06,
      "loss": 0.8018,
      "step": 2226
    },
    {
      "epoch": 0.288312781176166,
      "grad_norm": 0.2022778305673234,
      "learning_rate": 1.1198338205518264e-06,
      "loss": 0.769,
      "step": 2227
    },
    {
      "epoch": 0.28844224358351944,
      "grad_norm": 0.1472555584151396,
      "learning_rate": 1.1196612146582341e-06,
      "loss": 0.7577,
      "step": 2228
    },
    {
      "epoch": 0.2885717059908729,
      "grad_norm": 0.11818907560428629,
      "learning_rate": 1.1194885077200625e-06,
      "loss": 0.7791,
      "step": 2229
    },
    {
      "epoch": 0.28870116839822635,
      "grad_norm": 0.14979106928655028,
      "learning_rate": 1.1193156997725905e-06,
      "loss": 0.7607,
      "step": 2230
    },
    {
      "epoch": 0.28870116839822635,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6297940611839294,
      "eval_runtime": 19.9822,
      "eval_samples_per_second": 2.202,
      "eval_steps_per_second": 0.15,
      "step": 2230
    },
    {
      "epoch": 0.2888306308055798,
      "grad_norm": 0.13382412255274342,
      "learning_rate": 1.1191427908511171e-06,
      "loss": 0.923,
      "step": 2231
    },
    {
      "epoch": 0.2889600932129333,
      "grad_norm": 0.1605542724112372,
      "learning_rate": 1.1189697809909632e-06,
      "loss": 0.8962,
      "step": 2232
    },
    {
      "epoch": 0.28908955562028676,
      "grad_norm": 0.1370650271084169,
      "learning_rate": 1.1187966702274692e-06,
      "loss": 0.9114,
      "step": 2233
    },
    {
      "epoch": 0.2892190180276402,
      "grad_norm": 0.2191160569901304,
      "learning_rate": 1.1186234585959962e-06,
      "loss": 0.9175,
      "step": 2234
    },
    {
      "epoch": 0.28934848043499367,
      "grad_norm": 0.18844533348771192,
      "learning_rate": 1.1184501461319268e-06,
      "loss": 0.8046,
      "step": 2235
    },
    {
      "epoch": 0.28934848043499367,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6283735632896423,
      "eval_runtime": 19.9772,
      "eval_samples_per_second": 2.203,
      "eval_steps_per_second": 0.15,
      "step": 2235
    },
    {
      "epoch": 0.2894779428423472,
      "grad_norm": 0.14178739546086197,
      "learning_rate": 1.1182767328706633e-06,
      "loss": 0.9009,
      "step": 2236
    },
    {
      "epoch": 0.28960740524970063,
      "grad_norm": 0.15948650313606783,
      "learning_rate": 1.1181032188476288e-06,
      "loss": 0.8771,
      "step": 2237
    },
    {
      "epoch": 0.2897368676570541,
      "grad_norm": 0.15041620587627608,
      "learning_rate": 1.117929604098267e-06,
      "loss": 0.7785,
      "step": 2238
    },
    {
      "epoch": 0.28986633006440754,
      "grad_norm": 0.17743992944228004,
      "learning_rate": 1.1177558886580423e-06,
      "loss": 0.7389,
      "step": 2239
    },
    {
      "epoch": 0.289995792471761,
      "grad_norm": 0.21107179272812568,
      "learning_rate": 1.1175820725624397e-06,
      "loss": 0.9839,
      "step": 2240
    },
    {
      "epoch": 0.289995792471761,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6294389367103577,
      "eval_runtime": 20.0616,
      "eval_samples_per_second": 2.193,
      "eval_steps_per_second": 0.15,
      "step": 2240
    },
    {
      "epoch": 0.2901252548791145,
      "grad_norm": 0.13602941795550466,
      "learning_rate": 1.1174081558469647e-06,
      "loss": 0.8522,
      "step": 2241
    },
    {
      "epoch": 0.29025471728646796,
      "grad_norm": 0.12146567823587738,
      "learning_rate": 1.117234138547143e-06,
      "loss": 0.7673,
      "step": 2242
    },
    {
      "epoch": 0.2903841796938214,
      "grad_norm": 0.1210737667519944,
      "learning_rate": 1.1170600206985217e-06,
      "loss": 0.7692,
      "step": 2243
    },
    {
      "epoch": 0.29051364210117486,
      "grad_norm": 0.14852502455914487,
      "learning_rate": 1.1168858023366672e-06,
      "loss": 0.7568,
      "step": 2244
    },
    {
      "epoch": 0.2906431045085283,
      "grad_norm": 0.12867718867052472,
      "learning_rate": 1.1167114834971676e-06,
      "loss": 0.8464,
      "step": 2245
    },
    {
      "epoch": 0.2906431045085283,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6260653138160706,
      "eval_runtime": 19.8906,
      "eval_samples_per_second": 2.212,
      "eval_steps_per_second": 0.151,
      "step": 2245
    },
    {
      "epoch": 0.2907725669158818,
      "grad_norm": 0.12401458568125362,
      "learning_rate": 1.116537064215631e-06,
      "loss": 0.8161,
      "step": 2246
    },
    {
      "epoch": 0.2909020293232353,
      "grad_norm": 0.21164857438336815,
      "learning_rate": 1.116362544527686e-06,
      "loss": 0.7479,
      "step": 2247
    },
    {
      "epoch": 0.29103149173058873,
      "grad_norm": 0.22354385407044602,
      "learning_rate": 1.1161879244689818e-06,
      "loss": 0.6387,
      "step": 2248
    },
    {
      "epoch": 0.2911609541379422,
      "grad_norm": 0.1668599362413456,
      "learning_rate": 1.1160132040751882e-06,
      "loss": 0.9221,
      "step": 2249
    },
    {
      "epoch": 0.29129041654529564,
      "grad_norm": 0.19616831145599412,
      "learning_rate": 1.1158383833819953e-06,
      "loss": 0.644,
      "step": 2250
    },
    {
      "epoch": 0.29129041654529564,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6276633739471436,
      "eval_runtime": 20.0127,
      "eval_samples_per_second": 2.199,
      "eval_steps_per_second": 0.15,
      "step": 2250
    },
    {
      "epoch": 0.29141987895264915,
      "grad_norm": 0.1381241334361546,
      "learning_rate": 1.1156634624251135e-06,
      "loss": 0.8836,
      "step": 2251
    },
    {
      "epoch": 0.2915493413600026,
      "grad_norm": 0.15915037403418603,
      "learning_rate": 1.1154884412402747e-06,
      "loss": 0.7886,
      "step": 2252
    },
    {
      "epoch": 0.29167880376735605,
      "grad_norm": 0.14384205946867865,
      "learning_rate": 1.1153133198632296e-06,
      "loss": 0.7525,
      "step": 2253
    },
    {
      "epoch": 0.2918082661747095,
      "grad_norm": 0.14060396113490253,
      "learning_rate": 1.1151380983297508e-06,
      "loss": 0.805,
      "step": 2254
    },
    {
      "epoch": 0.29193772858206296,
      "grad_norm": 0.15097751955090635,
      "learning_rate": 1.114962776675631e-06,
      "loss": 0.973,
      "step": 2255
    },
    {
      "epoch": 0.29193772858206296,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6276633739471436,
      "eval_runtime": 20.7938,
      "eval_samples_per_second": 2.116,
      "eval_steps_per_second": 0.144,
      "step": 2255
    },
    {
      "epoch": 0.29206719098941647,
      "grad_norm": 0.11599080480387099,
      "learning_rate": 1.1147873549366827e-06,
      "loss": 0.7706,
      "step": 2256
    },
    {
      "epoch": 0.2921966533967699,
      "grad_norm": 0.11723877767979292,
      "learning_rate": 1.1146118331487398e-06,
      "loss": 0.7367,
      "step": 2257
    },
    {
      "epoch": 0.2923261158041234,
      "grad_norm": 0.11268144855624392,
      "learning_rate": 1.1144362113476558e-06,
      "loss": 0.7596,
      "step": 2258
    },
    {
      "epoch": 0.29245557821147683,
      "grad_norm": 0.12011468278729495,
      "learning_rate": 1.1142604895693055e-06,
      "loss": 0.8129,
      "step": 2259
    },
    {
      "epoch": 0.2925850406188303,
      "grad_norm": 0.1173171778100709,
      "learning_rate": 1.114084667849583e-06,
      "loss": 0.6335,
      "step": 2260
    },
    {
      "epoch": 0.2925850406188303,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6273082494735718,
      "eval_runtime": 19.7017,
      "eval_samples_per_second": 2.233,
      "eval_steps_per_second": 0.152,
      "step": 2260
    },
    {
      "epoch": 0.2927145030261838,
      "grad_norm": 0.1502884336227462,
      "learning_rate": 1.1139087462244038e-06,
      "loss": 0.7533,
      "step": 2261
    },
    {
      "epoch": 0.29284396543353725,
      "grad_norm": 0.16870770265075222,
      "learning_rate": 1.113732724729704e-06,
      "loss": 0.8656,
      "step": 2262
    },
    {
      "epoch": 0.2929734278408907,
      "grad_norm": 0.14577756413947285,
      "learning_rate": 1.1135566034014384e-06,
      "loss": 0.7753,
      "step": 2263
    },
    {
      "epoch": 0.29310289024824415,
      "grad_norm": 0.16494727023287545,
      "learning_rate": 1.1133803822755843e-06,
      "loss": 0.6733,
      "step": 2264
    },
    {
      "epoch": 0.2932323526555976,
      "grad_norm": 0.1891419644219371,
      "learning_rate": 1.1132040613881378e-06,
      "loss": 0.7161,
      "step": 2265
    },
    {
      "epoch": 0.2932323526555976,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6273082494735718,
      "eval_runtime": 20.9916,
      "eval_samples_per_second": 2.096,
      "eval_steps_per_second": 0.143,
      "step": 2265
    },
    {
      "epoch": 0.2933618150629511,
      "grad_norm": 0.15982174031467763,
      "learning_rate": 1.1130276407751167e-06,
      "loss": 0.7157,
      "step": 2266
    },
    {
      "epoch": 0.29349127747030457,
      "grad_norm": 0.15552362062858358,
      "learning_rate": 1.1128511204725579e-06,
      "loss": 0.6818,
      "step": 2267
    },
    {
      "epoch": 0.293620739877658,
      "grad_norm": 0.14551231150783425,
      "learning_rate": 1.1126745005165194e-06,
      "loss": 0.708,
      "step": 2268
    },
    {
      "epoch": 0.2937502022850115,
      "grad_norm": 0.2872478676721363,
      "learning_rate": 1.1124977809430794e-06,
      "loss": 0.9343,
      "step": 2269
    },
    {
      "epoch": 0.29387966469236493,
      "grad_norm": 0.1600824825946363,
      "learning_rate": 1.1123209617883368e-06,
      "loss": 0.7456,
      "step": 2270
    },
    {
      "epoch": 0.29387966469236493,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6281960010528564,
      "eval_runtime": 19.9892,
      "eval_samples_per_second": 2.201,
      "eval_steps_per_second": 0.15,
      "step": 2270
    },
    {
      "epoch": 0.29400912709971844,
      "grad_norm": 0.18500535832451254,
      "learning_rate": 1.11214404308841e-06,
      "loss": 0.8473,
      "step": 2271
    },
    {
      "epoch": 0.2941385895070719,
      "grad_norm": 0.13970488425017122,
      "learning_rate": 1.1119670248794384e-06,
      "loss": 0.6896,
      "step": 2272
    },
    {
      "epoch": 0.29426805191442534,
      "grad_norm": 0.11201045006978583,
      "learning_rate": 1.1117899071975819e-06,
      "loss": 0.7737,
      "step": 2273
    },
    {
      "epoch": 0.2943975143217788,
      "grad_norm": 0.16655384200376577,
      "learning_rate": 1.1116126900790197e-06,
      "loss": 0.8516,
      "step": 2274
    },
    {
      "epoch": 0.29452697672913225,
      "grad_norm": 0.17191008424789045,
      "learning_rate": 1.1114353735599525e-06,
      "loss": 1.0581,
      "step": 2275
    },
    {
      "epoch": 0.29452697672913225,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6253551244735718,
      "eval_runtime": 20.4672,
      "eval_samples_per_second": 2.15,
      "eval_steps_per_second": 0.147,
      "step": 2275
    },
    {
      "epoch": 0.29465643913648576,
      "grad_norm": 0.17027768987789363,
      "learning_rate": 1.111257957676601e-06,
      "loss": 0.9572,
      "step": 2276
    },
    {
      "epoch": 0.2947859015438392,
      "grad_norm": 0.11105436891909527,
      "learning_rate": 1.111080442465205e-06,
      "loss": 0.7619,
      "step": 2277
    },
    {
      "epoch": 0.29491536395119267,
      "grad_norm": 0.11111687533974614,
      "learning_rate": 1.1109028279620266e-06,
      "loss": 0.7632,
      "step": 2278
    },
    {
      "epoch": 0.2950448263585461,
      "grad_norm": 0.14776330183581557,
      "learning_rate": 1.1107251142033468e-06,
      "loss": 0.8441,
      "step": 2279
    },
    {
      "epoch": 0.29517428876589963,
      "grad_norm": 0.1448318485899028,
      "learning_rate": 1.110547301225467e-06,
      "loss": 0.8693,
      "step": 2280
    },
    {
      "epoch": 0.29517428876589963,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6258878111839294,
      "eval_runtime": 21.119,
      "eval_samples_per_second": 2.083,
      "eval_steps_per_second": 0.142,
      "step": 2280
    },
    {
      "epoch": 0.2953037511732531,
      "grad_norm": 0.1667268661584801,
      "learning_rate": 1.1103693890647097e-06,
      "loss": 0.7415,
      "step": 2281
    },
    {
      "epoch": 0.29543321358060654,
      "grad_norm": 0.1600395960809429,
      "learning_rate": 1.1101913777574164e-06,
      "loss": 0.6768,
      "step": 2282
    },
    {
      "epoch": 0.29556267598796,
      "grad_norm": 0.12969906275075574,
      "learning_rate": 1.1100132673399498e-06,
      "loss": 0.8918,
      "step": 2283
    },
    {
      "epoch": 0.29569213839531344,
      "grad_norm": 0.19220819747881462,
      "learning_rate": 1.1098350578486927e-06,
      "loss": 0.8638,
      "step": 2284
    },
    {
      "epoch": 0.29582160080266695,
      "grad_norm": 0.13775382384746251,
      "learning_rate": 1.1096567493200477e-06,
      "loss": 0.8468,
      "step": 2285
    },
    {
      "epoch": 0.29582160080266695,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6287286877632141,
      "eval_runtime": 20.0838,
      "eval_samples_per_second": 2.191,
      "eval_steps_per_second": 0.149,
      "step": 2285
    },
    {
      "epoch": 0.2959510632100204,
      "grad_norm": 0.11424983410266093,
      "learning_rate": 1.1094783417904385e-06,
      "loss": 0.769,
      "step": 2286
    },
    {
      "epoch": 0.29608052561737386,
      "grad_norm": 0.11094950989821305,
      "learning_rate": 1.1092998352963078e-06,
      "loss": 0.7513,
      "step": 2287
    },
    {
      "epoch": 0.2962099880247273,
      "grad_norm": 0.12382845342051162,
      "learning_rate": 1.1091212298741195e-06,
      "loss": 0.8311,
      "step": 2288
    },
    {
      "epoch": 0.29633945043208076,
      "grad_norm": 0.11393722144113118,
      "learning_rate": 1.108942525560357e-06,
      "loss": 0.7522,
      "step": 2289
    },
    {
      "epoch": 0.2964689128394343,
      "grad_norm": 0.15434090656090005,
      "learning_rate": 1.1087637223915249e-06,
      "loss": 0.8799,
      "step": 2290
    },
    {
      "epoch": 0.2964689128394343,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6264204382896423,
      "eval_runtime": 19.5399,
      "eval_samples_per_second": 2.252,
      "eval_steps_per_second": 0.154,
      "step": 2290
    },
    {
      "epoch": 0.2965983752467877,
      "grad_norm": 0.13706481051710268,
      "learning_rate": 1.1085848204041468e-06,
      "loss": 0.7611,
      "step": 2291
    },
    {
      "epoch": 0.2967278376541412,
      "grad_norm": 0.13502303456862697,
      "learning_rate": 1.1084058196347676e-06,
      "loss": 0.8537,
      "step": 2292
    },
    {
      "epoch": 0.29685730006149463,
      "grad_norm": 0.18915451022944493,
      "learning_rate": 1.1082267201199515e-06,
      "loss": 0.8669,
      "step": 2293
    },
    {
      "epoch": 0.2969867624688481,
      "grad_norm": 0.12876034007122264,
      "learning_rate": 1.1080475218962832e-06,
      "loss": 0.8167,
      "step": 2294
    },
    {
      "epoch": 0.2971162248762016,
      "grad_norm": 0.13896672831082954,
      "learning_rate": 1.1078682250003677e-06,
      "loss": 0.8616,
      "step": 2295
    },
    {
      "epoch": 0.2971162248762016,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6287286877632141,
      "eval_runtime": 20.4453,
      "eval_samples_per_second": 2.152,
      "eval_steps_per_second": 0.147,
      "step": 2295
    },
    {
      "epoch": 0.29724568728355505,
      "grad_norm": 0.14329714218181866,
      "learning_rate": 1.1076888294688298e-06,
      "loss": 0.8184,
      "step": 2296
    },
    {
      "epoch": 0.2973751496909085,
      "grad_norm": 0.23178602564368408,
      "learning_rate": 1.1075093353383148e-06,
      "loss": 0.7755,
      "step": 2297
    },
    {
      "epoch": 0.29750461209826196,
      "grad_norm": 0.12025665627366647,
      "learning_rate": 1.1073297426454883e-06,
      "loss": 0.7466,
      "step": 2298
    },
    {
      "epoch": 0.2976340745056154,
      "grad_norm": 0.1589236593947832,
      "learning_rate": 1.1071500514270354e-06,
      "loss": 0.9423,
      "step": 2299
    },
    {
      "epoch": 0.2977635369129689,
      "grad_norm": 0.1685103709159684,
      "learning_rate": 1.1069702617196616e-06,
      "loss": 0.7643,
      "step": 2300
    },
    {
      "epoch": 0.2977635369129689,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6265980005264282,
      "eval_runtime": 20.1458,
      "eval_samples_per_second": 2.184,
      "eval_steps_per_second": 0.149,
      "step": 2300
    },
    {
      "epoch": 0.29789299932032237,
      "grad_norm": 0.18076775397118913,
      "learning_rate": 1.1067903735600928e-06,
      "loss": 0.7753,
      "step": 2301
    },
    {
      "epoch": 0.2980224617276758,
      "grad_norm": 0.1836985441796816,
      "learning_rate": 1.1066103869850749e-06,
      "loss": 0.9723,
      "step": 2302
    },
    {
      "epoch": 0.2981519241350293,
      "grad_norm": 0.1483873696694304,
      "learning_rate": 1.1064303020313736e-06,
      "loss": 0.9464,
      "step": 2303
    },
    {
      "epoch": 0.29828138654238273,
      "grad_norm": 0.28924734550478154,
      "learning_rate": 1.1062501187357747e-06,
      "loss": 0.7033,
      "step": 2304
    },
    {
      "epoch": 0.29841084894973624,
      "grad_norm": 0.14372797087057437,
      "learning_rate": 1.1060698371350849e-06,
      "loss": 0.8167,
      "step": 2305
    },
    {
      "epoch": 0.29841084894973624,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6283735632896423,
      "eval_runtime": 20.0371,
      "eval_samples_per_second": 2.196,
      "eval_steps_per_second": 0.15,
      "step": 2305
    },
    {
      "epoch": 0.2985403113570897,
      "grad_norm": 0.16331265711808524,
      "learning_rate": 1.10588945726613e-06,
      "loss": 0.759,
      "step": 2306
    },
    {
      "epoch": 0.29866977376444315,
      "grad_norm": 0.10595970602721703,
      "learning_rate": 1.105708979165756e-06,
      "loss": 0.7711,
      "step": 2307
    },
    {
      "epoch": 0.2987992361717966,
      "grad_norm": 0.12750504536975163,
      "learning_rate": 1.1055284028708297e-06,
      "loss": 0.775,
      "step": 2308
    },
    {
      "epoch": 0.29892869857915005,
      "grad_norm": 0.1412813126893143,
      "learning_rate": 1.1053477284182373e-06,
      "loss": 0.7041,
      "step": 2309
    },
    {
      "epoch": 0.29905816098650356,
      "grad_norm": 0.1138221569895431,
      "learning_rate": 1.105166955844885e-06,
      "loss": 0.8184,
      "step": 2310
    },
    {
      "epoch": 0.29905816098650356,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6260653138160706,
      "eval_runtime": 19.973,
      "eval_samples_per_second": 2.203,
      "eval_steps_per_second": 0.15,
      "step": 2310
    },
    {
      "epoch": 0.299187623393857,
      "grad_norm": 0.16097573931329026,
      "learning_rate": 1.1049860851876996e-06,
      "loss": 0.9327,
      "step": 2311
    },
    {
      "epoch": 0.29931708580121047,
      "grad_norm": 0.1323167739180932,
      "learning_rate": 1.1048051164836271e-06,
      "loss": 0.8395,
      "step": 2312
    },
    {
      "epoch": 0.2994465482085639,
      "grad_norm": 0.13092654915986918,
      "learning_rate": 1.1046240497696345e-06,
      "loss": 0.8298,
      "step": 2313
    },
    {
      "epoch": 0.2995760106159174,
      "grad_norm": 0.15400575935986827,
      "learning_rate": 1.1044428850827083e-06,
      "loss": 0.9059,
      "step": 2314
    },
    {
      "epoch": 0.2997054730232709,
      "grad_norm": 0.13522662200340263,
      "learning_rate": 1.104261622459855e-06,
      "loss": 0.8555,
      "step": 2315
    },
    {
      "epoch": 0.2997054730232709,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6271306872367859,
      "eval_runtime": 20.2117,
      "eval_samples_per_second": 2.177,
      "eval_steps_per_second": 0.148,
      "step": 2315
    },
    {
      "epoch": 0.29983493543062434,
      "grad_norm": 0.1317784273991651,
      "learning_rate": 1.104080261938101e-06,
      "loss": 0.828,
      "step": 2316
    },
    {
      "epoch": 0.2999643978379778,
      "grad_norm": 0.2128661276012086,
      "learning_rate": 1.103898803554493e-06,
      "loss": 0.7483,
      "step": 2317
    },
    {
      "epoch": 0.30009386024533125,
      "grad_norm": 0.1540754421292361,
      "learning_rate": 1.1037172473460976e-06,
      "loss": 0.8845,
      "step": 2318
    },
    {
      "epoch": 0.30022332265268475,
      "grad_norm": 0.1453401310772915,
      "learning_rate": 1.1035355933500012e-06,
      "loss": 0.877,
      "step": 2319
    },
    {
      "epoch": 0.3003527850600382,
      "grad_norm": 0.10859124327341281,
      "learning_rate": 1.1033538416033105e-06,
      "loss": 0.7734,
      "step": 2320
    },
    {
      "epoch": 0.3003527850600382,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6274858117103577,
      "eval_runtime": 19.9018,
      "eval_samples_per_second": 2.211,
      "eval_steps_per_second": 0.151,
      "step": 2320
    },
    {
      "epoch": 0.30048224746739166,
      "grad_norm": 0.17046717952101975,
      "learning_rate": 1.1031719921431517e-06,
      "loss": 0.7845,
      "step": 2321
    },
    {
      "epoch": 0.3006117098747451,
      "grad_norm": 0.12732415632276328,
      "learning_rate": 1.1029900450066715e-06,
      "loss": 0.7617,
      "step": 2322
    },
    {
      "epoch": 0.30074117228209857,
      "grad_norm": 0.10707953451962489,
      "learning_rate": 1.1028080002310361e-06,
      "loss": 0.7499,
      "step": 2323
    },
    {
      "epoch": 0.3008706346894521,
      "grad_norm": 0.19322089131518028,
      "learning_rate": 1.1026258578534322e-06,
      "loss": 0.7384,
      "step": 2324
    },
    {
      "epoch": 0.30100009709680553,
      "grad_norm": 0.11788457996274594,
      "learning_rate": 1.1024436179110654e-06,
      "loss": 0.8129,
      "step": 2325
    },
    {
      "epoch": 0.30100009709680553,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6267755627632141,
      "eval_runtime": 20.6491,
      "eval_samples_per_second": 2.131,
      "eval_steps_per_second": 0.145,
      "step": 2325
    },
    {
      "epoch": 0.301129559504159,
      "grad_norm": 0.15764670488117122,
      "learning_rate": 1.1022612804411625e-06,
      "loss": 0.7623,
      "step": 2326
    },
    {
      "epoch": 0.30125902191151244,
      "grad_norm": 0.21742802775179212,
      "learning_rate": 1.1020788454809695e-06,
      "loss": 0.9379,
      "step": 2327
    },
    {
      "epoch": 0.3013884843188659,
      "grad_norm": 0.12259532536813243,
      "learning_rate": 1.1018963130677523e-06,
      "loss": 0.8342,
      "step": 2328
    },
    {
      "epoch": 0.3015179467262194,
      "grad_norm": 0.12291926317995507,
      "learning_rate": 1.1017136832387967e-06,
      "loss": 0.7505,
      "step": 2329
    },
    {
      "epoch": 0.30164740913357285,
      "grad_norm": 0.14000965791378558,
      "learning_rate": 1.101530956031409e-06,
      "loss": 0.7764,
      "step": 2330
    },
    {
      "epoch": 0.30164740913357285,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6260653138160706,
      "eval_runtime": 20.3175,
      "eval_samples_per_second": 2.166,
      "eval_steps_per_second": 0.148,
      "step": 2330
    },
    {
      "epoch": 0.3017768715409263,
      "grad_norm": 0.11047345655036277,
      "learning_rate": 1.1013481314829146e-06,
      "loss": 0.7808,
      "step": 2331
    },
    {
      "epoch": 0.30190633394827976,
      "grad_norm": 0.17585288151322376,
      "learning_rate": 1.1011652096306593e-06,
      "loss": 0.6268,
      "step": 2332
    },
    {
      "epoch": 0.3020357963556332,
      "grad_norm": 0.10634303996483258,
      "learning_rate": 1.1009821905120087e-06,
      "loss": 0.7686,
      "step": 2333
    },
    {
      "epoch": 0.3021652587629867,
      "grad_norm": 0.12150726122710424,
      "learning_rate": 1.1007990741643476e-06,
      "loss": 0.7723,
      "step": 2334
    },
    {
      "epoch": 0.3022947211703402,
      "grad_norm": 0.12004779779109133,
      "learning_rate": 1.1006158606250816e-06,
      "loss": 0.8083,
      "step": 2335
    },
    {
      "epoch": 0.3022947211703402,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6258878111839294,
      "eval_runtime": 20.5465,
      "eval_samples_per_second": 2.141,
      "eval_steps_per_second": 0.146,
      "step": 2335
    },
    {
      "epoch": 0.30242418357769363,
      "grad_norm": 0.11392785379694016,
      "learning_rate": 1.100432549931636e-06,
      "loss": 0.6949,
      "step": 2336
    },
    {
      "epoch": 0.3025536459850471,
      "grad_norm": 0.1767088288940512,
      "learning_rate": 1.1002491421214552e-06,
      "loss": 0.6579,
      "step": 2337
    },
    {
      "epoch": 0.30268310839240054,
      "grad_norm": 0.20516309049668333,
      "learning_rate": 1.1000656372320045e-06,
      "loss": 0.7292,
      "step": 2338
    },
    {
      "epoch": 0.30281257079975404,
      "grad_norm": 0.17097943415888148,
      "learning_rate": 1.0998820353007683e-06,
      "loss": 0.843,
      "step": 2339
    },
    {
      "epoch": 0.3029420332071075,
      "grad_norm": 0.16556838509784919,
      "learning_rate": 1.099698336365251e-06,
      "loss": 0.7931,
      "step": 2340
    },
    {
      "epoch": 0.3029420332071075,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6278409361839294,
      "eval_runtime": 20.9966,
      "eval_samples_per_second": 2.096,
      "eval_steps_per_second": 0.143,
      "step": 2340
    },
    {
      "epoch": 0.30307149561446095,
      "grad_norm": 0.1886961282167186,
      "learning_rate": 1.0995145404629764e-06,
      "loss": 0.7535,
      "step": 2341
    },
    {
      "epoch": 0.3032009580218144,
      "grad_norm": 0.1279514543001204,
      "learning_rate": 1.0993306476314892e-06,
      "loss": 0.712,
      "step": 2342
    },
    {
      "epoch": 0.30333042042916786,
      "grad_norm": 0.12696425730599323,
      "learning_rate": 1.099146657908353e-06,
      "loss": 0.7739,
      "step": 2343
    },
    {
      "epoch": 0.30345988283652137,
      "grad_norm": 0.11151166196258999,
      "learning_rate": 1.0989625713311514e-06,
      "loss": 0.6631,
      "step": 2344
    },
    {
      "epoch": 0.3035893452438748,
      "grad_norm": 0.11068702426580258,
      "learning_rate": 1.0987783879374878e-06,
      "loss": 0.7122,
      "step": 2345
    },
    {
      "epoch": 0.3035893452438748,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6242897510528564,
      "eval_runtime": 20.2789,
      "eval_samples_per_second": 2.17,
      "eval_steps_per_second": 0.148,
      "step": 2345
    },
    {
      "epoch": 0.3037188076512283,
      "grad_norm": 0.14805452498923785,
      "learning_rate": 1.0985941077649854e-06,
      "loss": 0.7129,
      "step": 2346
    },
    {
      "epoch": 0.3038482700585817,
      "grad_norm": 0.12726239784420756,
      "learning_rate": 1.098409730851287e-06,
      "loss": 0.788,
      "step": 2347
    },
    {
      "epoch": 0.3039777324659352,
      "grad_norm": 0.13460361933280068,
      "learning_rate": 1.098225257234056e-06,
      "loss": 0.8197,
      "step": 2348
    },
    {
      "epoch": 0.3041071948732887,
      "grad_norm": 0.11016610685294498,
      "learning_rate": 1.0980406869509737e-06,
      "loss": 0.7356,
      "step": 2349
    },
    {
      "epoch": 0.30423665728064214,
      "grad_norm": 0.1867607137937224,
      "learning_rate": 1.0978560200397434e-06,
      "loss": 0.7433,
      "step": 2350
    },
    {
      "epoch": 0.30423665728064214,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6265980005264282,
      "eval_runtime": 20.7661,
      "eval_samples_per_second": 2.119,
      "eval_steps_per_second": 0.144,
      "step": 2350
    },
    {
      "epoch": 0.3043661196879956,
      "grad_norm": 0.10892952931430959,
      "learning_rate": 1.0976712565380868e-06,
      "loss": 0.8015,
      "step": 2351
    },
    {
      "epoch": 0.30449558209534905,
      "grad_norm": 0.16294318617867853,
      "learning_rate": 1.0974863964837452e-06,
      "loss": 0.7224,
      "step": 2352
    },
    {
      "epoch": 0.3046250445027025,
      "grad_norm": 0.14526788464360893,
      "learning_rate": 1.0973014399144804e-06,
      "loss": 0.8043,
      "step": 2353
    },
    {
      "epoch": 0.304754506910056,
      "grad_norm": 0.15790908362050413,
      "learning_rate": 1.0971163868680732e-06,
      "loss": 0.6747,
      "step": 2354
    },
    {
      "epoch": 0.30488396931740946,
      "grad_norm": 0.1339211979374657,
      "learning_rate": 1.0969312373823246e-06,
      "loss": 0.7355,
      "step": 2355
    },
    {
      "epoch": 0.30488396931740946,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6232244372367859,
      "eval_runtime": 20.8261,
      "eval_samples_per_second": 2.113,
      "eval_steps_per_second": 0.144,
      "step": 2355
    },
    {
      "epoch": 0.3050134317247629,
      "grad_norm": 0.11058592283370441,
      "learning_rate": 1.0967459914950555e-06,
      "loss": 0.7238,
      "step": 2356
    },
    {
      "epoch": 0.30514289413211637,
      "grad_norm": 0.19153756512899792,
      "learning_rate": 1.0965606492441055e-06,
      "loss": 0.7032,
      "step": 2357
    },
    {
      "epoch": 0.3052723565394698,
      "grad_norm": 0.12765904540151063,
      "learning_rate": 1.0963752106673348e-06,
      "loss": 0.7694,
      "step": 2358
    },
    {
      "epoch": 0.30540181894682333,
      "grad_norm": 0.162323790159088,
      "learning_rate": 1.0961896758026231e-06,
      "loss": 0.7761,
      "step": 2359
    },
    {
      "epoch": 0.3055312813541768,
      "grad_norm": 0.12695199265416757,
      "learning_rate": 1.0960040446878691e-06,
      "loss": 0.6781,
      "step": 2360
    },
    {
      "epoch": 0.3055312813541768,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6223366260528564,
      "eval_runtime": 20.4124,
      "eval_samples_per_second": 2.156,
      "eval_steps_per_second": 0.147,
      "step": 2360
    },
    {
      "epoch": 0.30566074376153024,
      "grad_norm": 0.123741300365718,
      "learning_rate": 1.0958183173609927e-06,
      "loss": 0.8228,
      "step": 2361
    },
    {
      "epoch": 0.3057902061688837,
      "grad_norm": 0.13641726838161325,
      "learning_rate": 1.0956324938599317e-06,
      "loss": 0.6631,
      "step": 2362
    },
    {
      "epoch": 0.3059196685762372,
      "grad_norm": 0.1401015221260348,
      "learning_rate": 1.0954465742226446e-06,
      "loss": 0.767,
      "step": 2363
    },
    {
      "epoch": 0.30604913098359066,
      "grad_norm": 0.1257888607254981,
      "learning_rate": 1.0952605584871092e-06,
      "loss": 0.7765,
      "step": 2364
    },
    {
      "epoch": 0.3061785933909441,
      "grad_norm": 0.19173584869420307,
      "learning_rate": 1.095074446691323e-06,
      "loss": 0.877,
      "step": 2365
    },
    {
      "epoch": 0.3061785933909441,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6234019994735718,
      "eval_runtime": 20.2016,
      "eval_samples_per_second": 2.178,
      "eval_steps_per_second": 0.149,
      "step": 2365
    },
    {
      "epoch": 0.30630805579829756,
      "grad_norm": 0.127009700474598,
      "learning_rate": 1.0948882388733028e-06,
      "loss": 0.8192,
      "step": 2366
    },
    {
      "epoch": 0.306437518205651,
      "grad_norm": 0.1182373472696294,
      "learning_rate": 1.094701935071086e-06,
      "loss": 0.6792,
      "step": 2367
    },
    {
      "epoch": 0.3065669806130045,
      "grad_norm": 0.16870857588576713,
      "learning_rate": 1.0945155353227284e-06,
      "loss": 0.6837,
      "step": 2368
    },
    {
      "epoch": 0.306696443020358,
      "grad_norm": 0.216160413708548,
      "learning_rate": 1.094329039666306e-06,
      "loss": 0.8872,
      "step": 2369
    },
    {
      "epoch": 0.30682590542771143,
      "grad_norm": 0.1781498915270634,
      "learning_rate": 1.0941424481399145e-06,
      "loss": 1.0031,
      "step": 2370
    },
    {
      "epoch": 0.30682590542771143,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.626953125,
      "eval_runtime": 20.2462,
      "eval_samples_per_second": 2.173,
      "eval_steps_per_second": 0.148,
      "step": 2370
    },
    {
      "epoch": 0.3069553678350649,
      "grad_norm": 0.17323909042268537,
      "learning_rate": 1.0939557607816687e-06,
      "loss": 0.7318,
      "step": 2371
    },
    {
      "epoch": 0.30708483024241834,
      "grad_norm": 0.11564557303406463,
      "learning_rate": 1.0937689776297036e-06,
      "loss": 0.8069,
      "step": 2372
    },
    {
      "epoch": 0.30721429264977185,
      "grad_norm": 0.2386186871973314,
      "learning_rate": 1.0935820987221734e-06,
      "loss": 0.7001,
      "step": 2373
    },
    {
      "epoch": 0.3073437550571253,
      "grad_norm": 0.143525313633731,
      "learning_rate": 1.0933951240972517e-06,
      "loss": 0.831,
      "step": 2374
    },
    {
      "epoch": 0.30747321746447875,
      "grad_norm": 0.13137951918154678,
      "learning_rate": 1.0932080537931319e-06,
      "loss": 0.7593,
      "step": 2375
    },
    {
      "epoch": 0.30747321746447875,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6274858117103577,
      "eval_runtime": 20.8102,
      "eval_samples_per_second": 2.114,
      "eval_steps_per_second": 0.144,
      "step": 2375
    },
    {
      "epoch": 0.3076026798718322,
      "grad_norm": 0.12264172859392415,
      "learning_rate": 1.093020887848027e-06,
      "loss": 0.7329,
      "step": 2376
    },
    {
      "epoch": 0.30773214227918566,
      "grad_norm": 0.1380140885986046,
      "learning_rate": 1.0928336263001694e-06,
      "loss": 0.8601,
      "step": 2377
    },
    {
      "epoch": 0.30786160468653917,
      "grad_norm": 0.15928615400347804,
      "learning_rate": 1.0926462691878113e-06,
      "loss": 0.7384,
      "step": 2378
    },
    {
      "epoch": 0.3079910670938926,
      "grad_norm": 0.1117880894679808,
      "learning_rate": 1.0924588165492236e-06,
      "loss": 0.7772,
      "step": 2379
    },
    {
      "epoch": 0.3081205295012461,
      "grad_norm": 0.17725074676448552,
      "learning_rate": 1.0922712684226977e-06,
      "loss": 0.7825,
      "step": 2380
    },
    {
      "epoch": 0.3081205295012461,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6258878111839294,
      "eval_runtime": 20.1227,
      "eval_samples_per_second": 2.187,
      "eval_steps_per_second": 0.149,
      "step": 2380
    },
    {
      "epoch": 0.30824999190859953,
      "grad_norm": 0.13147092144491432,
      "learning_rate": 1.0920836248465444e-06,
      "loss": 0.7947,
      "step": 2381
    },
    {
      "epoch": 0.308379454315953,
      "grad_norm": 0.13415532589304818,
      "learning_rate": 1.091895885859093e-06,
      "loss": 0.827,
      "step": 2382
    },
    {
      "epoch": 0.3085089167233065,
      "grad_norm": 0.13983548958931974,
      "learning_rate": 1.0917080514986936e-06,
      "loss": 0.7432,
      "step": 2383
    },
    {
      "epoch": 0.30863837913065995,
      "grad_norm": 0.11800238209817694,
      "learning_rate": 1.091520121803715e-06,
      "loss": 0.7808,
      "step": 2384
    },
    {
      "epoch": 0.3087678415380134,
      "grad_norm": 0.1412218791037719,
      "learning_rate": 1.0913320968125454e-06,
      "loss": 0.8599,
      "step": 2385
    },
    {
      "epoch": 0.3087678415380134,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6258878111839294,
      "eval_runtime": 20.1724,
      "eval_samples_per_second": 2.181,
      "eval_steps_per_second": 0.149,
      "step": 2385
    },
    {
      "epoch": 0.30889730394536685,
      "grad_norm": 0.13637071136676548,
      "learning_rate": 1.0911439765635927e-06,
      "loss": 0.6896,
      "step": 2386
    },
    {
      "epoch": 0.3090267663527203,
      "grad_norm": 0.1965029459446922,
      "learning_rate": 1.090955761095285e-06,
      "loss": 0.6342,
      "step": 2387
    },
    {
      "epoch": 0.3091562287600738,
      "grad_norm": 0.13285299897872893,
      "learning_rate": 1.0907674504460684e-06,
      "loss": 0.8219,
      "step": 2388
    },
    {
      "epoch": 0.30928569116742727,
      "grad_norm": 0.11854097382009306,
      "learning_rate": 1.0905790446544094e-06,
      "loss": 0.8309,
      "step": 2389
    },
    {
      "epoch": 0.3094151535747807,
      "grad_norm": 0.14279753362572947,
      "learning_rate": 1.0903905437587934e-06,
      "loss": 0.772,
      "step": 2390
    },
    {
      "epoch": 0.3094151535747807,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6241121888160706,
      "eval_runtime": 20.2711,
      "eval_samples_per_second": 2.171,
      "eval_steps_per_second": 0.148,
      "step": 2390
    },
    {
      "epoch": 0.3095446159821342,
      "grad_norm": 0.14154672534822169,
      "learning_rate": 1.090201947797726e-06,
      "loss": 0.8936,
      "step": 2391
    },
    {
      "epoch": 0.30967407838948763,
      "grad_norm": 0.24601490375108462,
      "learning_rate": 1.0900132568097315e-06,
      "loss": 0.6388,
      "step": 2392
    },
    {
      "epoch": 0.30980354079684114,
      "grad_norm": 0.12811331376666152,
      "learning_rate": 1.089824470833354e-06,
      "loss": 0.7068,
      "step": 2393
    },
    {
      "epoch": 0.3099330032041946,
      "grad_norm": 0.16243446567511755,
      "learning_rate": 1.0896355899071568e-06,
      "loss": 0.6927,
      "step": 2394
    },
    {
      "epoch": 0.31006246561154804,
      "grad_norm": 0.17294656126538524,
      "learning_rate": 1.0894466140697225e-06,
      "loss": 0.8124,
      "step": 2395
    },
    {
      "epoch": 0.31006246561154804,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6255326867103577,
      "eval_runtime": 20.4351,
      "eval_samples_per_second": 2.153,
      "eval_steps_per_second": 0.147,
      "step": 2395
    },
    {
      "epoch": 0.3101919280189015,
      "grad_norm": 0.1588789557123746,
      "learning_rate": 1.0892575433596533e-06,
      "loss": 0.9244,
      "step": 2396
    },
    {
      "epoch": 0.31032139042625495,
      "grad_norm": 0.1519750869998078,
      "learning_rate": 1.0890683778155708e-06,
      "loss": 0.8002,
      "step": 2397
    },
    {
      "epoch": 0.31045085283360846,
      "grad_norm": 0.14395770152680884,
      "learning_rate": 1.0888791174761162e-06,
      "loss": 0.9391,
      "step": 2398
    },
    {
      "epoch": 0.3105803152409619,
      "grad_norm": 0.12899141327667735,
      "learning_rate": 1.088689762379949e-06,
      "loss": 0.8105,
      "step": 2399
    },
    {
      "epoch": 0.31070977764831537,
      "grad_norm": 0.11421980465823432,
      "learning_rate": 1.0885003125657497e-06,
      "loss": 0.8068,
      "step": 2400
    },
    {
      "epoch": 0.31070977764831537,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6241121888160706,
      "eval_runtime": 20.2517,
      "eval_samples_per_second": 2.173,
      "eval_steps_per_second": 0.148,
      "step": 2400
    },
    {
      "epoch": 0.3108392400556688,
      "grad_norm": 0.14128568778720565,
      "learning_rate": 1.0883107680722167e-06,
      "loss": 0.8636,
      "step": 2401
    },
    {
      "epoch": 0.3109687024630223,
      "grad_norm": 0.1348041071772375,
      "learning_rate": 1.0881211289380682e-06,
      "loss": 0.8284,
      "step": 2402
    },
    {
      "epoch": 0.3110981648703758,
      "grad_norm": 0.11367378828275754,
      "learning_rate": 1.0879313952020424e-06,
      "loss": 0.8009,
      "step": 2403
    },
    {
      "epoch": 0.31122762727772924,
      "grad_norm": 0.1494416559780264,
      "learning_rate": 1.087741566902896e-06,
      "loss": 0.7142,
      "step": 2404
    },
    {
      "epoch": 0.3113570896850827,
      "grad_norm": 0.1697207007093198,
      "learning_rate": 1.0875516440794048e-06,
      "loss": 0.5991,
      "step": 2405
    },
    {
      "epoch": 0.3113570896850827,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6253551244735718,
      "eval_runtime": 20.5303,
      "eval_samples_per_second": 2.143,
      "eval_steps_per_second": 0.146,
      "step": 2405
    },
    {
      "epoch": 0.31148655209243614,
      "grad_norm": 0.23362716276082,
      "learning_rate": 1.0873616267703651e-06,
      "loss": 0.8896,
      "step": 2406
    },
    {
      "epoch": 0.31161601449978965,
      "grad_norm": 0.12656348504533868,
      "learning_rate": 1.0871715150145915e-06,
      "loss": 0.7656,
      "step": 2407
    },
    {
      "epoch": 0.3117454769071431,
      "grad_norm": 0.1696778363874934,
      "learning_rate": 1.086981308850918e-06,
      "loss": 0.8855,
      "step": 2408
    },
    {
      "epoch": 0.31187493931449656,
      "grad_norm": 0.1449596253714716,
      "learning_rate": 1.0867910083181984e-06,
      "loss": 0.8165,
      "step": 2409
    },
    {
      "epoch": 0.31200440172185,
      "grad_norm": 0.15561534988218603,
      "learning_rate": 1.0866006134553053e-06,
      "loss": 0.7048,
      "step": 2410
    },
    {
      "epoch": 0.31200440172185,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6262428760528564,
      "eval_runtime": 20.7046,
      "eval_samples_per_second": 2.125,
      "eval_steps_per_second": 0.145,
      "step": 2410
    },
    {
      "epoch": 0.31213386412920346,
      "grad_norm": 0.16699553091292135,
      "learning_rate": 1.0864101243011306e-06,
      "loss": 0.8679,
      "step": 2411
    },
    {
      "epoch": 0.312263326536557,
      "grad_norm": 0.11455335113243467,
      "learning_rate": 1.0862195408945856e-06,
      "loss": 0.8037,
      "step": 2412
    },
    {
      "epoch": 0.3123927889439104,
      "grad_norm": 0.11721509771364215,
      "learning_rate": 1.0860288632746007e-06,
      "loss": 0.7762,
      "step": 2413
    },
    {
      "epoch": 0.3125222513512639,
      "grad_norm": 0.20851404529417966,
      "learning_rate": 1.085838091480126e-06,
      "loss": 0.9738,
      "step": 2414
    },
    {
      "epoch": 0.31265171375861733,
      "grad_norm": 0.15246792306740253,
      "learning_rate": 1.0856472255501307e-06,
      "loss": 0.8342,
      "step": 2415
    },
    {
      "epoch": 0.31265171375861733,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6232244372367859,
      "eval_runtime": 20.8021,
      "eval_samples_per_second": 2.115,
      "eval_steps_per_second": 0.144,
      "step": 2415
    },
    {
      "epoch": 0.3127811761659708,
      "grad_norm": 0.12595539965791425,
      "learning_rate": 1.0854562655236022e-06,
      "loss": 0.7455,
      "step": 2416
    },
    {
      "epoch": 0.3129106385733243,
      "grad_norm": 0.2513386840670701,
      "learning_rate": 1.0852652114395488e-06,
      "loss": 0.6619,
      "step": 2417
    },
    {
      "epoch": 0.31304010098067775,
      "grad_norm": 0.206389039364769,
      "learning_rate": 1.0850740633369965e-06,
      "loss": 0.8282,
      "step": 2418
    },
    {
      "epoch": 0.3131695633880312,
      "grad_norm": 0.147462904680677,
      "learning_rate": 1.0848828212549917e-06,
      "loss": 0.9102,
      "step": 2419
    },
    {
      "epoch": 0.31329902579538466,
      "grad_norm": 0.11636639475543832,
      "learning_rate": 1.0846914852325994e-06,
      "loss": 0.7513,
      "step": 2420
    },
    {
      "epoch": 0.31329902579538466,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6241121888160706,
      "eval_runtime": 20.4129,
      "eval_samples_per_second": 2.156,
      "eval_steps_per_second": 0.147,
      "step": 2420
    },
    {
      "epoch": 0.3134284882027381,
      "grad_norm": 0.21944931309898624,
      "learning_rate": 1.0845000553089032e-06,
      "loss": 0.9314,
      "step": 2421
    },
    {
      "epoch": 0.3135579506100916,
      "grad_norm": 0.14989995705349987,
      "learning_rate": 1.0843085315230076e-06,
      "loss": 0.869,
      "step": 2422
    },
    {
      "epoch": 0.31368741301744507,
      "grad_norm": 0.11324771294561456,
      "learning_rate": 1.0841169139140345e-06,
      "loss": 0.7586,
      "step": 2423
    },
    {
      "epoch": 0.3138168754247985,
      "grad_norm": 0.1175135524489197,
      "learning_rate": 1.0839252025211259e-06,
      "loss": 0.7544,
      "step": 2424
    },
    {
      "epoch": 0.313946337832152,
      "grad_norm": 0.22843264810536018,
      "learning_rate": 1.0837333973834423e-06,
      "loss": 0.6852,
      "step": 2425
    },
    {
      "epoch": 0.313946337832152,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6255326867103577,
      "eval_runtime": 20.596,
      "eval_samples_per_second": 2.136,
      "eval_steps_per_second": 0.146,
      "step": 2425
    },
    {
      "epoch": 0.31407580023950543,
      "grad_norm": 0.1133576042026987,
      "learning_rate": 1.0835414985401646e-06,
      "loss": 0.7772,
      "step": 2426
    },
    {
      "epoch": 0.31420526264685894,
      "grad_norm": 0.12344473419501979,
      "learning_rate": 1.0833495060304916e-06,
      "loss": 0.7924,
      "step": 2427
    },
    {
      "epoch": 0.3143347250542124,
      "grad_norm": 0.17055630904999222,
      "learning_rate": 1.0831574198936414e-06,
      "loss": 0.8132,
      "step": 2428
    },
    {
      "epoch": 0.31446418746156585,
      "grad_norm": 0.12108417325256941,
      "learning_rate": 1.082965240168852e-06,
      "loss": 0.819,
      "step": 2429
    },
    {
      "epoch": 0.3145936498689193,
      "grad_norm": 0.19136421255909758,
      "learning_rate": 1.08277296689538e-06,
      "loss": 0.7444,
      "step": 2430
    },
    {
      "epoch": 0.3145936498689193,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6274858117103577,
      "eval_runtime": 20.8315,
      "eval_samples_per_second": 2.112,
      "eval_steps_per_second": 0.144,
      "step": 2430
    },
    {
      "epoch": 0.31472311227627275,
      "grad_norm": 0.19447654317245303,
      "learning_rate": 1.0825806001125003e-06,
      "loss": 0.6525,
      "step": 2431
    },
    {
      "epoch": 0.31485257468362626,
      "grad_norm": 0.10919488053020052,
      "learning_rate": 1.0823881398595085e-06,
      "loss": 0.8058,
      "step": 2432
    },
    {
      "epoch": 0.3149820370909797,
      "grad_norm": 0.15328182934543524,
      "learning_rate": 1.0821955861757185e-06,
      "loss": 0.7195,
      "step": 2433
    },
    {
      "epoch": 0.31511149949833317,
      "grad_norm": 0.17644974553811982,
      "learning_rate": 1.082002939100463e-06,
      "loss": 0.6459,
      "step": 2434
    },
    {
      "epoch": 0.3152409619056866,
      "grad_norm": 0.17958907685763298,
      "learning_rate": 1.0818101986730942e-06,
      "loss": 0.6782,
      "step": 2435
    },
    {
      "epoch": 0.3152409619056866,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6248224377632141,
      "eval_runtime": 20.1852,
      "eval_samples_per_second": 2.18,
      "eval_steps_per_second": 0.149,
      "step": 2435
    },
    {
      "epoch": 0.3153704243130401,
      "grad_norm": 0.14365298264518006,
      "learning_rate": 1.0816173649329831e-06,
      "loss": 0.7203,
      "step": 2436
    },
    {
      "epoch": 0.3154998867203936,
      "grad_norm": 0.1860063922777146,
      "learning_rate": 1.0814244379195199e-06,
      "loss": 0.6689,
      "step": 2437
    },
    {
      "epoch": 0.31562934912774704,
      "grad_norm": 0.15243656286744653,
      "learning_rate": 1.0812314176721141e-06,
      "loss": 0.728,
      "step": 2438
    },
    {
      "epoch": 0.3157588115351005,
      "grad_norm": 0.16201627482915845,
      "learning_rate": 1.0810383042301938e-06,
      "loss": 0.675,
      "step": 2439
    },
    {
      "epoch": 0.31588827394245395,
      "grad_norm": 0.16722093974913427,
      "learning_rate": 1.0808450976332064e-06,
      "loss": 0.735,
      "step": 2440
    },
    {
      "epoch": 0.31588827394245395,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6265980005264282,
      "eval_runtime": 20.3194,
      "eval_samples_per_second": 2.165,
      "eval_steps_per_second": 0.148,
      "step": 2440
    },
    {
      "epoch": 0.3160177363498074,
      "grad_norm": 0.1260594987228566,
      "learning_rate": 1.0806517979206183e-06,
      "loss": 0.7733,
      "step": 2441
    },
    {
      "epoch": 0.3161471987571609,
      "grad_norm": 0.12781273635823043,
      "learning_rate": 1.0804584051319146e-06,
      "loss": 0.8438,
      "step": 2442
    },
    {
      "epoch": 0.31627666116451436,
      "grad_norm": 0.28187305994241185,
      "learning_rate": 1.0802649193066003e-06,
      "loss": 0.6313,
      "step": 2443
    },
    {
      "epoch": 0.3164061235718678,
      "grad_norm": 0.1210581233509803,
      "learning_rate": 1.0800713404841982e-06,
      "loss": 0.8229,
      "step": 2444
    },
    {
      "epoch": 0.31653558597922127,
      "grad_norm": 0.13805959164885548,
      "learning_rate": 1.079877668704251e-06,
      "loss": 0.7589,
      "step": 2445
    },
    {
      "epoch": 0.31653558597922127,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6264204382896423,
      "eval_runtime": 20.419,
      "eval_samples_per_second": 2.155,
      "eval_steps_per_second": 0.147,
      "step": 2445
    },
    {
      "epoch": 0.3166650483865747,
      "grad_norm": 0.15219453675341646,
      "learning_rate": 1.07968390400632e-06,
      "loss": 0.8573,
      "step": 2446
    },
    {
      "epoch": 0.31679451079392823,
      "grad_norm": 0.15652528770201463,
      "learning_rate": 1.0794900464299856e-06,
      "loss": 0.6385,
      "step": 2447
    },
    {
      "epoch": 0.3169239732012817,
      "grad_norm": 0.11285373458633965,
      "learning_rate": 1.0792960960148472e-06,
      "loss": 0.7483,
      "step": 2448
    },
    {
      "epoch": 0.31705343560863514,
      "grad_norm": 0.19604446344088078,
      "learning_rate": 1.0791020528005228e-06,
      "loss": 0.7889,
      "step": 2449
    },
    {
      "epoch": 0.3171828980159886,
      "grad_norm": 0.14197912294169385,
      "learning_rate": 1.07890791682665e-06,
      "loss": 0.8691,
      "step": 2450
    },
    {
      "epoch": 0.3171828980159886,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6264204382896423,
      "eval_runtime": 20.7449,
      "eval_samples_per_second": 2.121,
      "eval_steps_per_second": 0.145,
      "step": 2450
    },
    {
      "epoch": 0.3173123604233421,
      "grad_norm": 0.1135029975719742,
      "learning_rate": 1.078713688132885e-06,
      "loss": 0.7386,
      "step": 2451
    },
    {
      "epoch": 0.31744182283069555,
      "grad_norm": 0.19473571463984898,
      "learning_rate": 1.0785193667589029e-06,
      "loss": 0.6719,
      "step": 2452
    },
    {
      "epoch": 0.317571285238049,
      "grad_norm": 0.13223201176252414,
      "learning_rate": 1.0783249527443973e-06,
      "loss": 0.6902,
      "step": 2453
    },
    {
      "epoch": 0.31770074764540246,
      "grad_norm": 0.11480326318575491,
      "learning_rate": 1.0781304461290821e-06,
      "loss": 0.7468,
      "step": 2454
    },
    {
      "epoch": 0.3178302100527559,
      "grad_norm": 0.11728375652171937,
      "learning_rate": 1.0779358469526886e-06,
      "loss": 0.7374,
      "step": 2455
    },
    {
      "epoch": 0.3178302100527559,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6265980005264282,
      "eval_runtime": 20.4486,
      "eval_samples_per_second": 2.152,
      "eval_steps_per_second": 0.147,
      "step": 2455
    },
    {
      "epoch": 0.3179596724601094,
      "grad_norm": 0.2310218870627307,
      "learning_rate": 1.0777411552549675e-06,
      "loss": 0.6417,
      "step": 2456
    },
    {
      "epoch": 0.3180891348674629,
      "grad_norm": 0.12471416841325392,
      "learning_rate": 1.077546371075689e-06,
      "loss": 0.8192,
      "step": 2457
    },
    {
      "epoch": 0.31821859727481633,
      "grad_norm": 0.16260024590665795,
      "learning_rate": 1.0773514944546414e-06,
      "loss": 0.8214,
      "step": 2458
    },
    {
      "epoch": 0.3183480596821698,
      "grad_norm": 0.1300818935710419,
      "learning_rate": 1.0771565254316327e-06,
      "loss": 0.7616,
      "step": 2459
    },
    {
      "epoch": 0.31847752208952324,
      "grad_norm": 0.1475805416729426,
      "learning_rate": 1.0769614640464883e-06,
      "loss": 0.7361,
      "step": 2460
    },
    {
      "epoch": 0.31847752208952324,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6264204382896423,
      "eval_runtime": 20.1023,
      "eval_samples_per_second": 2.189,
      "eval_steps_per_second": 0.149,
      "step": 2460
    },
    {
      "epoch": 0.31860698449687674,
      "grad_norm": 0.12953339284946241,
      "learning_rate": 1.0767663103390546e-06,
      "loss": 0.7563,
      "step": 2461
    },
    {
      "epoch": 0.3187364469042302,
      "grad_norm": 0.15123777450693882,
      "learning_rate": 1.0765710643491947e-06,
      "loss": 0.7839,
      "step": 2462
    },
    {
      "epoch": 0.31886590931158365,
      "grad_norm": 0.12911037215700114,
      "learning_rate": 1.0763757261167923e-06,
      "loss": 0.7806,
      "step": 2463
    },
    {
      "epoch": 0.3189953717189371,
      "grad_norm": 0.12832399398836775,
      "learning_rate": 1.0761802956817486e-06,
      "loss": 0.8024,
      "step": 2464
    },
    {
      "epoch": 0.31912483412629056,
      "grad_norm": 0.1216095020922706,
      "learning_rate": 1.0759847730839847e-06,
      "loss": 0.7419,
      "step": 2465
    },
    {
      "epoch": 0.31912483412629056,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6292613744735718,
      "eval_runtime": 20.5583,
      "eval_samples_per_second": 2.14,
      "eval_steps_per_second": 0.146,
      "step": 2465
    },
    {
      "epoch": 0.31925429653364407,
      "grad_norm": 0.2211082840634109,
      "learning_rate": 1.0757891583634398e-06,
      "loss": 0.6272,
      "step": 2466
    },
    {
      "epoch": 0.3193837589409975,
      "grad_norm": 0.11886178777936064,
      "learning_rate": 1.0755934515600721e-06,
      "loss": 0.6996,
      "step": 2467
    },
    {
      "epoch": 0.319513221348351,
      "grad_norm": 0.1403622052508912,
      "learning_rate": 1.075397652713859e-06,
      "loss": 0.882,
      "step": 2468
    },
    {
      "epoch": 0.3196426837557044,
      "grad_norm": 0.11835818613404944,
      "learning_rate": 1.0752017618647959e-06,
      "loss": 0.7822,
      "step": 2469
    },
    {
      "epoch": 0.3197721461630579,
      "grad_norm": 0.12539832021361638,
      "learning_rate": 1.075005779052898e-06,
      "loss": 0.8015,
      "step": 2470
    },
    {
      "epoch": 0.3197721461630579,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.626953125,
      "eval_runtime": 20.2676,
      "eval_samples_per_second": 2.171,
      "eval_steps_per_second": 0.148,
      "step": 2470
    },
    {
      "epoch": 0.3199016085704114,
      "grad_norm": 0.10325655472534062,
      "learning_rate": 1.0748097043181984e-06,
      "loss": 0.7822,
      "step": 2471
    },
    {
      "epoch": 0.32003107097776484,
      "grad_norm": 0.20021690188519256,
      "learning_rate": 1.0746135377007496e-06,
      "loss": 0.6887,
      "step": 2472
    },
    {
      "epoch": 0.3201605333851183,
      "grad_norm": 0.1521183170462301,
      "learning_rate": 1.0744172792406225e-06,
      "loss": 0.6973,
      "step": 2473
    },
    {
      "epoch": 0.32028999579247175,
      "grad_norm": 0.10686817205555298,
      "learning_rate": 1.0742209289779069e-06,
      "loss": 0.7064,
      "step": 2474
    },
    {
      "epoch": 0.3204194581998252,
      "grad_norm": 0.13390972553763222,
      "learning_rate": 1.074024486952711e-06,
      "loss": 0.8514,
      "step": 2475
    },
    {
      "epoch": 0.3204194581998252,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6253551244735718,
      "eval_runtime": 20.1372,
      "eval_samples_per_second": 2.185,
      "eval_steps_per_second": 0.149,
      "step": 2475
    },
    {
      "epoch": 0.3205489206071787,
      "grad_norm": 0.1564191678507911,
      "learning_rate": 1.0738279532051625e-06,
      "loss": 0.8361,
      "step": 2476
    },
    {
      "epoch": 0.32067838301453216,
      "grad_norm": 0.12165235012654953,
      "learning_rate": 1.073631327775407e-06,
      "loss": 0.7055,
      "step": 2477
    },
    {
      "epoch": 0.3208078454218856,
      "grad_norm": 0.10300619134574486,
      "learning_rate": 1.0734346107036097e-06,
      "loss": 0.6387,
      "step": 2478
    },
    {
      "epoch": 0.32093730782923907,
      "grad_norm": 0.1556586680309464,
      "learning_rate": 1.073237802029954e-06,
      "loss": 0.739,
      "step": 2479
    },
    {
      "epoch": 0.3210667702365925,
      "grad_norm": 0.1301825687195915,
      "learning_rate": 1.0730409017946414e-06,
      "loss": 0.7598,
      "step": 2480
    },
    {
      "epoch": 0.3210667702365925,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6244673132896423,
      "eval_runtime": 19.9729,
      "eval_samples_per_second": 2.203,
      "eval_steps_per_second": 0.15,
      "step": 2480
    },
    {
      "epoch": 0.32119623264394603,
      "grad_norm": 0.13373025798100974,
      "learning_rate": 1.0728439100378935e-06,
      "loss": 0.7823,
      "step": 2481
    },
    {
      "epoch": 0.3213256950512995,
      "grad_norm": 0.12729423145872187,
      "learning_rate": 1.0726468267999495e-06,
      "loss": 0.6782,
      "step": 2482
    },
    {
      "epoch": 0.32145515745865294,
      "grad_norm": 0.10611060005692165,
      "learning_rate": 1.072449652121068e-06,
      "loss": 0.7751,
      "step": 2483
    },
    {
      "epoch": 0.3215846198660064,
      "grad_norm": 0.1629717760593082,
      "learning_rate": 1.0722523860415251e-06,
      "loss": 0.7025,
      "step": 2484
    },
    {
      "epoch": 0.32171408227335985,
      "grad_norm": 0.14341375075703797,
      "learning_rate": 1.0720550286016174e-06,
      "loss": 0.7701,
      "step": 2485
    },
    {
      "epoch": 0.32171408227335985,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6251775622367859,
      "eval_runtime": 20.157,
      "eval_samples_per_second": 2.183,
      "eval_steps_per_second": 0.149,
      "step": 2485
    },
    {
      "epoch": 0.32184354468071336,
      "grad_norm": 0.15068433202544396,
      "learning_rate": 1.0718575798416585e-06,
      "loss": 0.8724,
      "step": 2486
    },
    {
      "epoch": 0.3219730070880668,
      "grad_norm": 0.14840189690235997,
      "learning_rate": 1.0716600398019818e-06,
      "loss": 0.85,
      "step": 2487
    },
    {
      "epoch": 0.32210246949542026,
      "grad_norm": 0.1651713590394895,
      "learning_rate": 1.0714624085229383e-06,
      "loss": 0.8965,
      "step": 2488
    },
    {
      "epoch": 0.3222319319027737,
      "grad_norm": 0.18375415345682763,
      "learning_rate": 1.0712646860448985e-06,
      "loss": 0.8892,
      "step": 2489
    },
    {
      "epoch": 0.32236139431012717,
      "grad_norm": 0.13981888969736853,
      "learning_rate": 1.0710668724082516e-06,
      "loss": 0.8622,
      "step": 2490
    },
    {
      "epoch": 0.32236139431012717,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6262428760528564,
      "eval_runtime": 19.6665,
      "eval_samples_per_second": 2.237,
      "eval_steps_per_second": 0.153,
      "step": 2490
    },
    {
      "epoch": 0.3224908567174807,
      "grad_norm": 0.13384676555874972,
      "learning_rate": 1.0708689676534044e-06,
      "loss": 0.7653,
      "step": 2491
    },
    {
      "epoch": 0.32262031912483413,
      "grad_norm": 0.18204938949654942,
      "learning_rate": 1.070670971820783e-06,
      "loss": 0.6777,
      "step": 2492
    },
    {
      "epoch": 0.3227497815321876,
      "grad_norm": 0.09860610865302899,
      "learning_rate": 1.0704728849508324e-06,
      "loss": 0.6593,
      "step": 2493
    },
    {
      "epoch": 0.32287924393954104,
      "grad_norm": 0.23062032075749622,
      "learning_rate": 1.070274707084016e-06,
      "loss": 0.6664,
      "step": 2494
    },
    {
      "epoch": 0.32300870634689455,
      "grad_norm": 0.11341688837989083,
      "learning_rate": 1.0700764382608148e-06,
      "loss": 0.8231,
      "step": 2495
    },
    {
      "epoch": 0.32300870634689455,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.625,
      "eval_runtime": 19.7113,
      "eval_samples_per_second": 2.232,
      "eval_steps_per_second": 0.152,
      "step": 2495
    },
    {
      "epoch": 0.323138168754248,
      "grad_norm": 0.1877055894285442,
      "learning_rate": 1.0698780785217302e-06,
      "loss": 0.9774,
      "step": 2496
    },
    {
      "epoch": 0.32326763116160145,
      "grad_norm": 0.1149182870690813,
      "learning_rate": 1.0696796279072807e-06,
      "loss": 0.6481,
      "step": 2497
    },
    {
      "epoch": 0.3233970935689549,
      "grad_norm": 0.10487655201497655,
      "learning_rate": 1.0694810864580041e-06,
      "loss": 0.6829,
      "step": 2498
    },
    {
      "epoch": 0.32352655597630836,
      "grad_norm": 0.13366939281534124,
      "learning_rate": 1.0692824542144562e-06,
      "loss": 0.7439,
      "step": 2499
    },
    {
      "epoch": 0.32365601838366187,
      "grad_norm": 0.12615596288868197,
      "learning_rate": 1.0690837312172117e-06,
      "loss": 0.7559,
      "step": 2500
    },
    {
      "epoch": 0.32365601838366187,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6273082494735718,
      "eval_runtime": 20.3494,
      "eval_samples_per_second": 2.162,
      "eval_steps_per_second": 0.147,
      "step": 2500
    },
    {
      "epoch": 0.3237854807910153,
      "grad_norm": 0.1286293053646549,
      "learning_rate": 1.0688849175068642e-06,
      "loss": 0.7847,
      "step": 2501
    },
    {
      "epoch": 0.3239149431983688,
      "grad_norm": 0.10727493358640168,
      "learning_rate": 1.0686860131240251e-06,
      "loss": 0.78,
      "step": 2502
    },
    {
      "epoch": 0.32404440560572223,
      "grad_norm": 0.2976472329155784,
      "learning_rate": 1.0684870181093246e-06,
      "loss": 1.0571,
      "step": 2503
    },
    {
      "epoch": 0.3241738680130757,
      "grad_norm": 0.12325642751187146,
      "learning_rate": 1.0682879325034119e-06,
      "loss": 0.8109,
      "step": 2504
    },
    {
      "epoch": 0.3243033304204292,
      "grad_norm": 0.12365696973617851,
      "learning_rate": 1.0680887563469537e-06,
      "loss": 0.7683,
      "step": 2505
    },
    {
      "epoch": 0.3243033304204292,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6237571239471436,
      "eval_runtime": 19.9083,
      "eval_samples_per_second": 2.21,
      "eval_steps_per_second": 0.151,
      "step": 2505
    },
    {
      "epoch": 0.32443279282778265,
      "grad_norm": 0.12209497530594306,
      "learning_rate": 1.0678894896806357e-06,
      "loss": 0.7454,
      "step": 2506
    },
    {
      "epoch": 0.3245622552351361,
      "grad_norm": 0.14609822840491737,
      "learning_rate": 1.0676901325451628e-06,
      "loss": 0.8041,
      "step": 2507
    },
    {
      "epoch": 0.32469171764248955,
      "grad_norm": 0.13991505378466332,
      "learning_rate": 1.0674906849812574e-06,
      "loss": 0.7755,
      "step": 2508
    },
    {
      "epoch": 0.324821180049843,
      "grad_norm": 0.12336661808318582,
      "learning_rate": 1.0672911470296609e-06,
      "loss": 0.7726,
      "step": 2509
    },
    {
      "epoch": 0.3249506424571965,
      "grad_norm": 0.16165952779730758,
      "learning_rate": 1.0670915187311326e-06,
      "loss": 0.7122,
      "step": 2510
    },
    {
      "epoch": 0.3249506424571965,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6232244372367859,
      "eval_runtime": 19.912,
      "eval_samples_per_second": 2.21,
      "eval_steps_per_second": 0.151,
      "step": 2510
    },
    {
      "epoch": 0.32508010486454997,
      "grad_norm": 0.15164086770151672,
      "learning_rate": 1.0668918001264508e-06,
      "loss": 0.8653,
      "step": 2511
    },
    {
      "epoch": 0.3252095672719034,
      "grad_norm": 0.1445976304509436,
      "learning_rate": 1.0666919912564119e-06,
      "loss": 0.7969,
      "step": 2512
    },
    {
      "epoch": 0.3253390296792569,
      "grad_norm": 0.20763170818268722,
      "learning_rate": 1.0664920921618313e-06,
      "loss": 0.8655,
      "step": 2513
    },
    {
      "epoch": 0.32546849208661033,
      "grad_norm": 0.12818776808593144,
      "learning_rate": 1.0662921028835424e-06,
      "loss": 0.7578,
      "step": 2514
    },
    {
      "epoch": 0.32559795449396384,
      "grad_norm": 0.18409897859995156,
      "learning_rate": 1.0660920234623968e-06,
      "loss": 0.7402,
      "step": 2515
    },
    {
      "epoch": 0.32559795449396384,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8415300546448088,
      "eval_PRM F1 AUC": 0.463855421686747,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.77,
      "eval_PRM Recall": 0.927710843373494,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6237571239471436,
      "eval_runtime": 20.4397,
      "eval_samples_per_second": 2.153,
      "eval_steps_per_second": 0.147,
      "step": 2515
    },
    {
      "epoch": 0.3257274169013173,
      "grad_norm": 0.2316620601513271,
      "learning_rate": 1.0658918539392645e-06,
      "loss": 0.5781,
      "step": 2516
    },
    {
      "epoch": 0.32585687930867074,
      "grad_norm": 0.1974934235022006,
      "learning_rate": 1.065691594355035e-06,
      "loss": 0.7312,
      "step": 2517
    },
    {
      "epoch": 0.3259863417160242,
      "grad_norm": 0.16313638175543904,
      "learning_rate": 1.065491244750615e-06,
      "loss": 0.712,
      "step": 2518
    },
    {
      "epoch": 0.32611580412337765,
      "grad_norm": 0.11093318196282526,
      "learning_rate": 1.0652908051669296e-06,
      "loss": 0.7704,
      "step": 2519
    },
    {
      "epoch": 0.32624526653073116,
      "grad_norm": 0.10669963210051349,
      "learning_rate": 1.065090275644923e-06,
      "loss": 0.7666,
      "step": 2520
    },
    {
      "epoch": 0.32624526653073116,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6273082494735718,
      "eval_runtime": 20.2669,
      "eval_samples_per_second": 2.171,
      "eval_steps_per_second": 0.148,
      "step": 2520
    },
    {
      "epoch": 0.3263747289380846,
      "grad_norm": 0.11546550226086605,
      "learning_rate": 1.0648896562255576e-06,
      "loss": 0.7648,
      "step": 2521
    },
    {
      "epoch": 0.32650419134543807,
      "grad_norm": 0.11850389204885863,
      "learning_rate": 1.0646889469498137e-06,
      "loss": 0.7339,
      "step": 2522
    },
    {
      "epoch": 0.3266336537527915,
      "grad_norm": 0.13210643746083273,
      "learning_rate": 1.06448814785869e-06,
      "loss": 0.8257,
      "step": 2523
    },
    {
      "epoch": 0.326763116160145,
      "grad_norm": 0.16349367600614878,
      "learning_rate": 1.0642872589932044e-06,
      "loss": 0.8529,
      "step": 2524
    },
    {
      "epoch": 0.3268925785674985,
      "grad_norm": 0.2049814657162059,
      "learning_rate": 1.0640862803943922e-06,
      "loss": 0.6689,
      "step": 2525
    },
    {
      "epoch": 0.3268925785674985,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6239346861839294,
      "eval_runtime": 19.9404,
      "eval_samples_per_second": 2.207,
      "eval_steps_per_second": 0.15,
      "step": 2525
    },
    {
      "epoch": 0.32702204097485194,
      "grad_norm": 0.17167872861434297,
      "learning_rate": 1.0638852121033071e-06,
      "loss": 0.7202,
      "step": 2526
    },
    {
      "epoch": 0.3271515033822054,
      "grad_norm": 0.12325501954976645,
      "learning_rate": 1.0636840541610216e-06,
      "loss": 0.788,
      "step": 2527
    },
    {
      "epoch": 0.32728096578955884,
      "grad_norm": 0.10114746938236344,
      "learning_rate": 1.0634828066086266e-06,
      "loss": 0.6918,
      "step": 2528
    },
    {
      "epoch": 0.3274104281969123,
      "grad_norm": 0.14420876282554568,
      "learning_rate": 1.0632814694872306e-06,
      "loss": 0.7489,
      "step": 2529
    },
    {
      "epoch": 0.3275398906042658,
      "grad_norm": 0.12981954808787974,
      "learning_rate": 1.0630800428379609e-06,
      "loss": 0.7265,
      "step": 2530
    },
    {
      "epoch": 0.3275398906042658,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6258878111839294,
      "eval_runtime": 19.9699,
      "eval_samples_per_second": 2.203,
      "eval_steps_per_second": 0.15,
      "step": 2530
    },
    {
      "epoch": 0.32766935301161926,
      "grad_norm": 0.11207075342553162,
      "learning_rate": 1.0628785267019631e-06,
      "loss": 0.7076,
      "step": 2531
    },
    {
      "epoch": 0.3277988154189727,
      "grad_norm": 0.17207524813827438,
      "learning_rate": 1.0626769211204007e-06,
      "loss": 0.9155,
      "step": 2532
    },
    {
      "epoch": 0.32792827782632616,
      "grad_norm": 0.2221344887549533,
      "learning_rate": 1.062475226134456e-06,
      "loss": 0.5692,
      "step": 2533
    },
    {
      "epoch": 0.3280577402336796,
      "grad_norm": 0.12633453353727667,
      "learning_rate": 1.062273441785329e-06,
      "loss": 0.7148,
      "step": 2534
    },
    {
      "epoch": 0.3281872026410331,
      "grad_norm": 0.1499964540412509,
      "learning_rate": 1.0620715681142382e-06,
      "loss": 0.9048,
      "step": 2535
    },
    {
      "epoch": 0.3281872026410331,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6265980005264282,
      "eval_runtime": 20.5061,
      "eval_samples_per_second": 2.146,
      "eval_steps_per_second": 0.146,
      "step": 2535
    },
    {
      "epoch": 0.3283166650483866,
      "grad_norm": 0.22298546756014595,
      "learning_rate": 1.0618696051624208e-06,
      "loss": 1.0857,
      "step": 2536
    },
    {
      "epoch": 0.32844612745574003,
      "grad_norm": 0.13535827907204878,
      "learning_rate": 1.0616675529711317e-06,
      "loss": 0.7972,
      "step": 2537
    },
    {
      "epoch": 0.3285755898630935,
      "grad_norm": 0.15485772805981585,
      "learning_rate": 1.0614654115816439e-06,
      "loss": 0.67,
      "step": 2538
    },
    {
      "epoch": 0.328705052270447,
      "grad_norm": 0.12569407768417754,
      "learning_rate": 1.0612631810352491e-06,
      "loss": 0.8186,
      "step": 2539
    },
    {
      "epoch": 0.32883451467780045,
      "grad_norm": 0.12500376543993366,
      "learning_rate": 1.0610608613732568e-06,
      "loss": 0.7719,
      "step": 2540
    },
    {
      "epoch": 0.32883451467780045,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8415300546448088,
      "eval_PRM F1 AUC": 0.463855421686747,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.77,
      "eval_PRM Recall": 0.927710843373494,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6248224377632141,
      "eval_runtime": 20.0515,
      "eval_samples_per_second": 2.194,
      "eval_steps_per_second": 0.15,
      "step": 2540
    },
    {
      "epoch": 0.3289639770851539,
      "grad_norm": 0.10653074587312804,
      "learning_rate": 1.060858452636995e-06,
      "loss": 0.7354,
      "step": 2541
    },
    {
      "epoch": 0.32909343949250736,
      "grad_norm": 0.12250437988012987,
      "learning_rate": 1.06065595486781e-06,
      "loss": 0.6868,
      "step": 2542
    },
    {
      "epoch": 0.3292229018998608,
      "grad_norm": 0.14660267004828095,
      "learning_rate": 1.0604533681070657e-06,
      "loss": 0.7292,
      "step": 2543
    },
    {
      "epoch": 0.3293523643072143,
      "grad_norm": 0.1408864943970288,
      "learning_rate": 1.0602506923961447e-06,
      "loss": 0.8792,
      "step": 2544
    },
    {
      "epoch": 0.32948182671456777,
      "grad_norm": 0.13038386632171534,
      "learning_rate": 1.060047927776448e-06,
      "loss": 0.7838,
      "step": 2545
    },
    {
      "epoch": 0.32948182671456777,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8415300546448088,
      "eval_PRM F1 AUC": 0.463855421686747,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.77,
      "eval_PRM Recall": 0.927710843373494,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6235795617103577,
      "eval_runtime": 20.4845,
      "eval_samples_per_second": 2.148,
      "eval_steps_per_second": 0.146,
      "step": 2545
    },
    {
      "epoch": 0.3296112891219212,
      "grad_norm": 0.1388242810052516,
      "learning_rate": 1.0598450742893937e-06,
      "loss": 0.8229,
      "step": 2546
    },
    {
      "epoch": 0.3297407515292747,
      "grad_norm": 0.12030531003783426,
      "learning_rate": 1.0596421319764192e-06,
      "loss": 0.7744,
      "step": 2547
    },
    {
      "epoch": 0.32987021393662813,
      "grad_norm": 0.1236488932397587,
      "learning_rate": 1.0594391008789791e-06,
      "loss": 0.8347,
      "step": 2548
    },
    {
      "epoch": 0.32999967634398164,
      "grad_norm": 0.12531832342752372,
      "learning_rate": 1.0592359810385475e-06,
      "loss": 0.8184,
      "step": 2549
    },
    {
      "epoch": 0.3301291387513351,
      "grad_norm": 0.13981178835270341,
      "learning_rate": 1.059032772496615e-06,
      "loss": 0.8821,
      "step": 2550
    },
    {
      "epoch": 0.3301291387513351,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6241121888160706,
      "eval_runtime": 19.907,
      "eval_samples_per_second": 2.21,
      "eval_steps_per_second": 0.151,
      "step": 2550
    },
    {
      "epoch": 0.33025860115868855,
      "grad_norm": 0.20600820404196976,
      "learning_rate": 1.0588294752946912e-06,
      "loss": 0.6261,
      "step": 2551
    },
    {
      "epoch": 0.330388063566042,
      "grad_norm": 0.16700591586790886,
      "learning_rate": 1.0586260894743037e-06,
      "loss": 0.7218,
      "step": 2552
    },
    {
      "epoch": 0.33051752597339545,
      "grad_norm": 0.14946712208579255,
      "learning_rate": 1.0584226150769983e-06,
      "loss": 0.701,
      "step": 2553
    },
    {
      "epoch": 0.33064698838074896,
      "grad_norm": 0.15750378536137175,
      "learning_rate": 1.0582190521443388e-06,
      "loss": 0.6202,
      "step": 2554
    },
    {
      "epoch": 0.3307764507881024,
      "grad_norm": 0.10569303380423906,
      "learning_rate": 1.0580154007179068e-06,
      "loss": 0.718,
      "step": 2555
    },
    {
      "epoch": 0.3307764507881024,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6262428760528564,
      "eval_runtime": 20.4008,
      "eval_samples_per_second": 2.157,
      "eval_steps_per_second": 0.147,
      "step": 2555
    },
    {
      "epoch": 0.33090591319545587,
      "grad_norm": 0.26722142187332865,
      "learning_rate": 1.0578116608393025e-06,
      "loss": 0.6171,
      "step": 2556
    },
    {
      "epoch": 0.3310353756028093,
      "grad_norm": 0.2208170882345125,
      "learning_rate": 1.0576078325501437e-06,
      "loss": 0.9171,
      "step": 2557
    },
    {
      "epoch": 0.3311648380101628,
      "grad_norm": 0.15997326260438854,
      "learning_rate": 1.0574039158920665e-06,
      "loss": 0.6835,
      "step": 2558
    },
    {
      "epoch": 0.3312943004175163,
      "grad_norm": 0.15284642047355682,
      "learning_rate": 1.0571999109067252e-06,
      "loss": 0.8569,
      "step": 2559
    },
    {
      "epoch": 0.33142376282486974,
      "grad_norm": 0.16872619052578008,
      "learning_rate": 1.0569958176357916e-06,
      "loss": 0.6519,
      "step": 2560
    },
    {
      "epoch": 0.33142376282486974,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6255326867103577,
      "eval_runtime": 19.4266,
      "eval_samples_per_second": 2.265,
      "eval_steps_per_second": 0.154,
      "step": 2560
    },
    {
      "epoch": 0.3315532252322232,
      "grad_norm": 0.17243617796642635,
      "learning_rate": 1.0567916361209562e-06,
      "loss": 0.7642,
      "step": 2561
    },
    {
      "epoch": 0.33168268763957665,
      "grad_norm": 0.12301840863130908,
      "learning_rate": 1.0565873664039272e-06,
      "loss": 0.8629,
      "step": 2562
    },
    {
      "epoch": 0.3318121500469301,
      "grad_norm": 0.12936303240228877,
      "learning_rate": 1.0563830085264307e-06,
      "loss": 0.6534,
      "step": 2563
    },
    {
      "epoch": 0.3319416124542836,
      "grad_norm": 0.16195616596317314,
      "learning_rate": 1.056178562530211e-06,
      "loss": 0.6738,
      "step": 2564
    },
    {
      "epoch": 0.33207107486163706,
      "grad_norm": 0.11355581873748197,
      "learning_rate": 1.0559740284570301e-06,
      "loss": 0.7377,
      "step": 2565
    },
    {
      "epoch": 0.33207107486163706,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6274858117103577,
      "eval_runtime": 20.9385,
      "eval_samples_per_second": 2.101,
      "eval_steps_per_second": 0.143,
      "step": 2565
    },
    {
      "epoch": 0.3322005372689905,
      "grad_norm": 0.13996208584183203,
      "learning_rate": 1.0557694063486685e-06,
      "loss": 0.8141,
      "step": 2566
    },
    {
      "epoch": 0.33232999967634397,
      "grad_norm": 0.13054174299374852,
      "learning_rate": 1.0555646962469247e-06,
      "loss": 0.6982,
      "step": 2567
    },
    {
      "epoch": 0.3324594620836974,
      "grad_norm": 0.11780819763068695,
      "learning_rate": 1.0553598981936143e-06,
      "loss": 0.8246,
      "step": 2568
    },
    {
      "epoch": 0.33258892449105093,
      "grad_norm": 0.1192339483173683,
      "learning_rate": 1.0551550122305714e-06,
      "loss": 0.6584,
      "step": 2569
    },
    {
      "epoch": 0.3327183868984044,
      "grad_norm": 0.15337537785786137,
      "learning_rate": 1.0549500383996487e-06,
      "loss": 0.7429,
      "step": 2570
    },
    {
      "epoch": 0.3327183868984044,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8415300546448088,
      "eval_PRM F1 AUC": 0.463855421686747,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.77,
      "eval_PRM Recall": 0.927710843373494,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6248224377632141,
      "eval_runtime": 19.9542,
      "eval_samples_per_second": 2.205,
      "eval_steps_per_second": 0.15,
      "step": 2570
    },
    {
      "epoch": 0.33284784930575784,
      "grad_norm": 0.1121123409602909,
      "learning_rate": 1.054744976742716e-06,
      "loss": 0.7423,
      "step": 2571
    },
    {
      "epoch": 0.3329773117131113,
      "grad_norm": 0.12116299423593384,
      "learning_rate": 1.0545398273016612e-06,
      "loss": 0.7389,
      "step": 2572
    },
    {
      "epoch": 0.33310677412046474,
      "grad_norm": 0.12222673939487311,
      "learning_rate": 1.0543345901183903e-06,
      "loss": 0.7754,
      "step": 2573
    },
    {
      "epoch": 0.33323623652781825,
      "grad_norm": 0.20383055813797735,
      "learning_rate": 1.0541292652348273e-06,
      "loss": 0.7181,
      "step": 2574
    },
    {
      "epoch": 0.3333656989351717,
      "grad_norm": 0.19542916785418335,
      "learning_rate": 1.0539238526929135e-06,
      "loss": 0.6307,
      "step": 2575
    },
    {
      "epoch": 0.3333656989351717,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8415300546448088,
      "eval_PRM F1 AUC": 0.463855421686747,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.77,
      "eval_PRM Recall": 0.927710843373494,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6248224377632141,
      "eval_runtime": 19.911,
      "eval_samples_per_second": 2.21,
      "eval_steps_per_second": 0.151,
      "step": 2575
    },
    {
      "epoch": 0.33349516134252516,
      "grad_norm": 0.1327357600189503,
      "learning_rate": 1.0537183525346093e-06,
      "loss": 0.6935,
      "step": 2576
    },
    {
      "epoch": 0.3336246237498786,
      "grad_norm": 0.10650846776750161,
      "learning_rate": 1.0535127648018915e-06,
      "loss": 0.7089,
      "step": 2577
    },
    {
      "epoch": 0.33375408615723207,
      "grad_norm": 0.16704426988904944,
      "learning_rate": 1.0533070895367562e-06,
      "loss": 0.8016,
      "step": 2578
    },
    {
      "epoch": 0.3338835485645856,
      "grad_norm": 0.21838144382994687,
      "learning_rate": 1.0531013267812162e-06,
      "loss": 0.6487,
      "step": 2579
    },
    {
      "epoch": 0.33401301097193903,
      "grad_norm": 0.1271165463063533,
      "learning_rate": 1.0528954765773032e-06,
      "loss": 0.6818,
      "step": 2580
    },
    {
      "epoch": 0.33401301097193903,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8415300546448088,
      "eval_PRM F1 AUC": 0.463855421686747,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.77,
      "eval_PRM Recall": 0.927710843373494,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.623046875,
      "eval_runtime": 20.4295,
      "eval_samples_per_second": 2.154,
      "eval_steps_per_second": 0.147,
      "step": 2580
    },
    {
      "epoch": 0.3341424733792925,
      "grad_norm": 0.12195052851066047,
      "learning_rate": 1.052689538967066e-06,
      "loss": 0.7354,
      "step": 2581
    },
    {
      "epoch": 0.33427193578664594,
      "grad_norm": 0.17628339737773374,
      "learning_rate": 1.0524835139925715e-06,
      "loss": 0.9652,
      "step": 2582
    },
    {
      "epoch": 0.33440139819399944,
      "grad_norm": 0.10918984453422255,
      "learning_rate": 1.0522774016959048e-06,
      "loss": 0.7352,
      "step": 2583
    },
    {
      "epoch": 0.3345308606013529,
      "grad_norm": 0.14462155888792924,
      "learning_rate": 1.0520712021191682e-06,
      "loss": 0.6984,
      "step": 2584
    },
    {
      "epoch": 0.33466032300870635,
      "grad_norm": 0.09970203797892369,
      "learning_rate": 1.0518649153044822e-06,
      "loss": 0.7046,
      "step": 2585
    },
    {
      "epoch": 0.33466032300870635,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8415300546448088,
      "eval_PRM F1 AUC": 0.463855421686747,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.77,
      "eval_PRM Recall": 0.927710843373494,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6255326867103577,
      "eval_runtime": 20.0041,
      "eval_samples_per_second": 2.2,
      "eval_steps_per_second": 0.15,
      "step": 2585
    },
    {
      "epoch": 0.3347897854160598,
      "grad_norm": 0.1486874324039095,
      "learning_rate": 1.0516585412939851e-06,
      "loss": 0.8538,
      "step": 2586
    },
    {
      "epoch": 0.33491924782341326,
      "grad_norm": 0.20030081755227566,
      "learning_rate": 1.0514520801298328e-06,
      "loss": 0.8746,
      "step": 2587
    },
    {
      "epoch": 0.33504871023076677,
      "grad_norm": 0.09382219697559281,
      "learning_rate": 1.0512455318541996e-06,
      "loss": 0.6876,
      "step": 2588
    },
    {
      "epoch": 0.3351781726381202,
      "grad_norm": 0.16940404576028745,
      "learning_rate": 1.0510388965092767e-06,
      "loss": 0.7438,
      "step": 2589
    },
    {
      "epoch": 0.3353076350454737,
      "grad_norm": 0.19741779144205993,
      "learning_rate": 1.0508321741372738e-06,
      "loss": 0.635,
      "step": 2590
    },
    {
      "epoch": 0.3353076350454737,
      "eval_PRM Accuracy": 0.7169811320754716,
      "eval_PRM F1": 0.8351648351648352,
      "eval_PRM F1 AUC": 0.4578313253012048,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7676767676767676,
      "eval_PRM Recall": 0.9156626506024096,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6239346861839294,
      "eval_runtime": 19.9553,
      "eval_samples_per_second": 2.205,
      "eval_steps_per_second": 0.15,
      "step": 2590
    },
    {
      "epoch": 0.3354370974528271,
      "grad_norm": 0.11687311729562885,
      "learning_rate": 1.0506253647804183e-06,
      "loss": 0.8301,
      "step": 2591
    },
    {
      "epoch": 0.3355665598601806,
      "grad_norm": 0.12279553823303907,
      "learning_rate": 1.0504184684809548e-06,
      "loss": 0.7532,
      "step": 2592
    },
    {
      "epoch": 0.3356960222675341,
      "grad_norm": 0.19674474560532282,
      "learning_rate": 1.0502114852811463e-06,
      "loss": 0.8757,
      "step": 2593
    },
    {
      "epoch": 0.33582548467488754,
      "grad_norm": 0.1186047804146949,
      "learning_rate": 1.0500044152232734e-06,
      "loss": 0.7018,
      "step": 2594
    },
    {
      "epoch": 0.335954947082241,
      "grad_norm": 0.16272681071442963,
      "learning_rate": 1.0497972583496341e-06,
      "loss": 0.881,
      "step": 2595
    },
    {
      "epoch": 0.335954947082241,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8415300546448088,
      "eval_PRM F1 AUC": 0.463855421686747,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.77,
      "eval_PRM Recall": 0.927710843373494,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6265980005264282,
      "eval_runtime": 20.0495,
      "eval_samples_per_second": 2.195,
      "eval_steps_per_second": 0.15,
      "step": 2595
    },
    {
      "epoch": 0.33608440948959445,
      "grad_norm": 0.2139245702166506,
      "learning_rate": 1.0495900147025449e-06,
      "loss": 0.9753,
      "step": 2596
    },
    {
      "epoch": 0.3362138718969479,
      "grad_norm": 0.15071971663123399,
      "learning_rate": 1.0493826843243388e-06,
      "loss": 0.6392,
      "step": 2597
    },
    {
      "epoch": 0.3363433343043014,
      "grad_norm": 0.13654563750288123,
      "learning_rate": 1.0491752672573676e-06,
      "loss": 0.7286,
      "step": 2598
    },
    {
      "epoch": 0.33647279671165486,
      "grad_norm": 0.20109159622083803,
      "learning_rate": 1.0489677635440004e-06,
      "loss": 0.8362,
      "step": 2599
    },
    {
      "epoch": 0.3366022591190083,
      "grad_norm": 0.22332801691996676,
      "learning_rate": 1.0487601732266242e-06,
      "loss": 0.7119,
      "step": 2600
    },
    {
      "epoch": 0.3366022591190083,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8415300546448088,
      "eval_PRM F1 AUC": 0.463855421686747,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.77,
      "eval_PRM Recall": 0.927710843373494,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6273082494735718,
      "eval_runtime": 20.2782,
      "eval_samples_per_second": 2.17,
      "eval_steps_per_second": 0.148,
      "step": 2600
    },
    {
      "epoch": 0.33673172152636177,
      "grad_norm": 0.1925383087652379,
      "learning_rate": 1.0485524963476433e-06,
      "loss": 0.7003,
      "step": 2601
    },
    {
      "epoch": 0.3368611839337152,
      "grad_norm": 0.15134354594418042,
      "learning_rate": 1.04834473294948e-06,
      "loss": 0.7604,
      "step": 2602
    },
    {
      "epoch": 0.33699064634106873,
      "grad_norm": 0.14167714467495984,
      "learning_rate": 1.0481368830745743e-06,
      "loss": 0.7626,
      "step": 2603
    },
    {
      "epoch": 0.3371201087484222,
      "grad_norm": 0.16759947980786186,
      "learning_rate": 1.0479289467653836e-06,
      "loss": 0.7953,
      "step": 2604
    },
    {
      "epoch": 0.33724957115577564,
      "grad_norm": 0.1544954716106123,
      "learning_rate": 1.047720924064383e-06,
      "loss": 0.8048,
      "step": 2605
    },
    {
      "epoch": 0.33724957115577564,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8415300546448088,
      "eval_PRM F1 AUC": 0.463855421686747,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.77,
      "eval_PRM Recall": 0.927710843373494,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6260653138160706,
      "eval_runtime": 20.1955,
      "eval_samples_per_second": 2.179,
      "eval_steps_per_second": 0.149,
      "step": 2605
    },
    {
      "epoch": 0.3373790335631291,
      "grad_norm": 0.1524190436451572,
      "learning_rate": 1.0475128150140658e-06,
      "loss": 0.7749,
      "step": 2606
    },
    {
      "epoch": 0.33750849597048255,
      "grad_norm": 0.17363936514132744,
      "learning_rate": 1.0473046196569419e-06,
      "loss": 0.9148,
      "step": 2607
    },
    {
      "epoch": 0.33763795837783606,
      "grad_norm": 0.1937717003106368,
      "learning_rate": 1.0470963380355398e-06,
      "loss": 0.6096,
      "step": 2608
    },
    {
      "epoch": 0.3377674207851895,
      "grad_norm": 0.14336489022879448,
      "learning_rate": 1.046887970192405e-06,
      "loss": 0.6964,
      "step": 2609
    },
    {
      "epoch": 0.33789688319254296,
      "grad_norm": 0.20991428917306634,
      "learning_rate": 1.0466795161701013e-06,
      "loss": 0.6235,
      "step": 2610
    },
    {
      "epoch": 0.33789688319254296,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6248224377632141,
      "eval_runtime": 20.36,
      "eval_samples_per_second": 2.161,
      "eval_steps_per_second": 0.147,
      "step": 2610
    },
    {
      "epoch": 0.3380263455998964,
      "grad_norm": 0.12879521811422606,
      "learning_rate": 1.0464709760112092e-06,
      "loss": 0.8732,
      "step": 2611
    },
    {
      "epoch": 0.33815580800724987,
      "grad_norm": 0.16017792598991287,
      "learning_rate": 1.0462623497583274e-06,
      "loss": 0.6876,
      "step": 2612
    },
    {
      "epoch": 0.3382852704146034,
      "grad_norm": 0.1250461625102018,
      "learning_rate": 1.046053637454072e-06,
      "loss": 0.8083,
      "step": 2613
    },
    {
      "epoch": 0.33841473282195683,
      "grad_norm": 0.1277852383755882,
      "learning_rate": 1.045844839141077e-06,
      "loss": 0.7278,
      "step": 2614
    },
    {
      "epoch": 0.3385441952293103,
      "grad_norm": 0.16342899037129738,
      "learning_rate": 1.0456359548619934e-06,
      "loss": 0.7871,
      "step": 2615
    },
    {
      "epoch": 0.3385441952293103,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8415300546448088,
      "eval_PRM F1 AUC": 0.463855421686747,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.77,
      "eval_PRM Recall": 0.927710843373494,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6257102489471436,
      "eval_runtime": 19.9022,
      "eval_samples_per_second": 2.211,
      "eval_steps_per_second": 0.151,
      "step": 2615
    },
    {
      "epoch": 0.33867365763666374,
      "grad_norm": 0.12268385230203797,
      "learning_rate": 1.04542698465949e-06,
      "loss": 0.806,
      "step": 2616
    },
    {
      "epoch": 0.3388031200440172,
      "grad_norm": 0.12212168645413116,
      "learning_rate": 1.0452179285762534e-06,
      "loss": 0.7314,
      "step": 2617
    },
    {
      "epoch": 0.3389325824513707,
      "grad_norm": 0.18553815474421625,
      "learning_rate": 1.0450087866549876e-06,
      "loss": 0.6898,
      "step": 2618
    },
    {
      "epoch": 0.33906204485872415,
      "grad_norm": 0.12153516795325979,
      "learning_rate": 1.0447995589384136e-06,
      "loss": 0.7773,
      "step": 2619
    },
    {
      "epoch": 0.3391915072660776,
      "grad_norm": 0.20135277789850037,
      "learning_rate": 1.044590245469271e-06,
      "loss": 0.8654,
      "step": 2620
    },
    {
      "epoch": 0.3391915072660776,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8415300546448088,
      "eval_PRM F1 AUC": 0.463855421686747,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.77,
      "eval_PRM Recall": 0.927710843373494,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6235795617103577,
      "eval_runtime": 19.9248,
      "eval_samples_per_second": 2.208,
      "eval_steps_per_second": 0.151,
      "step": 2620
    },
    {
      "epoch": 0.33932096967343106,
      "grad_norm": 0.1918013471562724,
      "learning_rate": 1.044380846290316e-06,
      "loss": 0.834,
      "step": 2621
    },
    {
      "epoch": 0.33945043208078457,
      "grad_norm": 0.14007894213389382,
      "learning_rate": 1.0441713614443226e-06,
      "loss": 0.7924,
      "step": 2622
    },
    {
      "epoch": 0.339579894488138,
      "grad_norm": 0.11536178092358225,
      "learning_rate": 1.0439617909740826e-06,
      "loss": 0.8242,
      "step": 2623
    },
    {
      "epoch": 0.3397093568954915,
      "grad_norm": 0.19881756739596487,
      "learning_rate": 1.0437521349224048e-06,
      "loss": 0.5793,
      "step": 2624
    },
    {
      "epoch": 0.33983881930284493,
      "grad_norm": 0.17362701791786683,
      "learning_rate": 1.0435423933321157e-06,
      "loss": 0.8368,
      "step": 2625
    },
    {
      "epoch": 0.33983881930284493,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8415300546448088,
      "eval_PRM F1 AUC": 0.463855421686747,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.77,
      "eval_PRM Recall": 0.927710843373494,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6242897510528564,
      "eval_runtime": 20.2982,
      "eval_samples_per_second": 2.168,
      "eval_steps_per_second": 0.148,
      "step": 2625
    },
    {
      "epoch": 0.3399682817101984,
      "grad_norm": 0.12759920017620227,
      "learning_rate": 1.043332566246059e-06,
      "loss": 0.7639,
      "step": 2626
    },
    {
      "epoch": 0.3400977441175519,
      "grad_norm": 0.09105113202688621,
      "learning_rate": 1.0431226537070967e-06,
      "loss": 0.6909,
      "step": 2627
    },
    {
      "epoch": 0.34022720652490535,
      "grad_norm": 0.12903831865269513,
      "learning_rate": 1.0429126557581073e-06,
      "loss": 0.7563,
      "step": 2628
    },
    {
      "epoch": 0.3403566689322588,
      "grad_norm": 0.11269634428287907,
      "learning_rate": 1.0427025724419873e-06,
      "loss": 0.8113,
      "step": 2629
    },
    {
      "epoch": 0.34048613133961225,
      "grad_norm": 0.1334134667935212,
      "learning_rate": 1.0424924038016505e-06,
      "loss": 0.7981,
      "step": 2630
    },
    {
      "epoch": 0.34048613133961225,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8415300546448088,
      "eval_PRM F1 AUC": 0.463855421686747,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.77,
      "eval_PRM Recall": 0.927710843373494,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6239346861839294,
      "eval_runtime": 20.2484,
      "eval_samples_per_second": 2.173,
      "eval_steps_per_second": 0.148,
      "step": 2630
    },
    {
      "epoch": 0.3406155937469657,
      "grad_norm": 0.14078639195361398,
      "learning_rate": 1.0422821498800282e-06,
      "loss": 0.7513,
      "step": 2631
    },
    {
      "epoch": 0.3407450561543192,
      "grad_norm": 0.13295704302215192,
      "learning_rate": 1.0420718107200686e-06,
      "loss": 0.8292,
      "step": 2632
    },
    {
      "epoch": 0.34087451856167267,
      "grad_norm": 0.14475051624638857,
      "learning_rate": 1.041861386364738e-06,
      "loss": 0.8287,
      "step": 2633
    },
    {
      "epoch": 0.3410039809690261,
      "grad_norm": 0.18888318855888467,
      "learning_rate": 1.04165087685702e-06,
      "loss": 0.5947,
      "step": 2634
    },
    {
      "epoch": 0.3411334433763796,
      "grad_norm": 0.1705580010733914,
      "learning_rate": 1.041440282239915e-06,
      "loss": 0.7481,
      "step": 2635
    },
    {
      "epoch": 0.3411334433763796,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.62109375,
      "eval_runtime": 20.8497,
      "eval_samples_per_second": 2.11,
      "eval_steps_per_second": 0.144,
      "step": 2635
    },
    {
      "epoch": 0.34126290578373303,
      "grad_norm": 0.15600534856508574,
      "learning_rate": 1.0412296025564417e-06,
      "loss": 0.7549,
      "step": 2636
    },
    {
      "epoch": 0.34139236819108654,
      "grad_norm": 0.21404375217749666,
      "learning_rate": 1.0410188378496356e-06,
      "loss": 0.9884,
      "step": 2637
    },
    {
      "epoch": 0.34152183059844,
      "grad_norm": 0.11357799376180283,
      "learning_rate": 1.0408079881625495e-06,
      "loss": 0.7446,
      "step": 2638
    },
    {
      "epoch": 0.34165129300579344,
      "grad_norm": 0.11406328782372947,
      "learning_rate": 1.0405970535382535e-06,
      "loss": 0.7598,
      "step": 2639
    },
    {
      "epoch": 0.3417807554131469,
      "grad_norm": 0.1143277614097742,
      "learning_rate": 1.0403860340198358e-06,
      "loss": 0.6803,
      "step": 2640
    },
    {
      "epoch": 0.3417807554131469,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6251775622367859,
      "eval_runtime": 19.9878,
      "eval_samples_per_second": 2.201,
      "eval_steps_per_second": 0.15,
      "step": 2640
    },
    {
      "epoch": 0.34191021782050035,
      "grad_norm": 0.1325834694575373,
      "learning_rate": 1.040174929650401e-06,
      "loss": 0.7645,
      "step": 2641
    },
    {
      "epoch": 0.34203968022785386,
      "grad_norm": 0.23137663727013474,
      "learning_rate": 1.0399637404730717e-06,
      "loss": 0.6006,
      "step": 2642
    },
    {
      "epoch": 0.3421691426352073,
      "grad_norm": 0.09997056987710319,
      "learning_rate": 1.0397524665309875e-06,
      "loss": 0.7209,
      "step": 2643
    },
    {
      "epoch": 0.34229860504256077,
      "grad_norm": 0.13317083402570234,
      "learning_rate": 1.0395411078673053e-06,
      "loss": 0.8613,
      "step": 2644
    },
    {
      "epoch": 0.3424280674499142,
      "grad_norm": 0.14026283989799987,
      "learning_rate": 1.0393296645251996e-06,
      "loss": 0.7736,
      "step": 2645
    },
    {
      "epoch": 0.3424280674499142,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8415300546448088,
      "eval_PRM F1 AUC": 0.463855421686747,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.77,
      "eval_PRM Recall": 0.927710843373494,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6251775622367859,
      "eval_runtime": 20.4582,
      "eval_samples_per_second": 2.151,
      "eval_steps_per_second": 0.147,
      "step": 2645
    },
    {
      "epoch": 0.3425575298572677,
      "grad_norm": 0.14405732501478088,
      "learning_rate": 1.0391181365478616e-06,
      "loss": 0.7401,
      "step": 2646
    },
    {
      "epoch": 0.3426869922646212,
      "grad_norm": 0.16614393788319257,
      "learning_rate": 1.0389065239785005e-06,
      "loss": 0.637,
      "step": 2647
    },
    {
      "epoch": 0.34281645467197464,
      "grad_norm": 0.09485729293980032,
      "learning_rate": 1.0386948268603426e-06,
      "loss": 0.7268,
      "step": 2648
    },
    {
      "epoch": 0.3429459170793281,
      "grad_norm": 0.22875285712804944,
      "learning_rate": 1.0384830452366306e-06,
      "loss": 1.0294,
      "step": 2649
    },
    {
      "epoch": 0.34307537948668154,
      "grad_norm": 0.23790018258796006,
      "learning_rate": 1.0382711791506262e-06,
      "loss": 0.6724,
      "step": 2650
    },
    {
      "epoch": 0.34307537948668154,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6232244372367859,
      "eval_runtime": 19.3721,
      "eval_samples_per_second": 2.271,
      "eval_steps_per_second": 0.155,
      "step": 2650
    },
    {
      "epoch": 0.343204841894035,
      "grad_norm": 0.13175074564086442,
      "learning_rate": 1.0380592286456066e-06,
      "loss": 0.6892,
      "step": 2651
    },
    {
      "epoch": 0.3433343043013885,
      "grad_norm": 0.14110833892011498,
      "learning_rate": 1.0378471937648674e-06,
      "loss": 0.7354,
      "step": 2652
    },
    {
      "epoch": 0.34346376670874196,
      "grad_norm": 0.1429568980715398,
      "learning_rate": 1.0376350745517206e-06,
      "loss": 0.8157,
      "step": 2653
    },
    {
      "epoch": 0.3435932291160954,
      "grad_norm": 0.11968003507531476,
      "learning_rate": 1.0374228710494962e-06,
      "loss": 0.7716,
      "step": 2654
    },
    {
      "epoch": 0.34372269152344886,
      "grad_norm": 0.1617965113197315,
      "learning_rate": 1.037210583301541e-06,
      "loss": 0.7498,
      "step": 2655
    },
    {
      "epoch": 0.34372269152344886,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6281960010528564,
      "eval_runtime": 20.052,
      "eval_samples_per_second": 2.194,
      "eval_steps_per_second": 0.15,
      "step": 2655
    },
    {
      "epoch": 0.3438521539308023,
      "grad_norm": 0.20831998808657423,
      "learning_rate": 1.036998211351219e-06,
      "loss": 0.6411,
      "step": 2656
    },
    {
      "epoch": 0.3439816163381558,
      "grad_norm": 0.12279500439729611,
      "learning_rate": 1.0367857552419114e-06,
      "loss": 0.7384,
      "step": 2657
    },
    {
      "epoch": 0.3441110787455093,
      "grad_norm": 0.1709773300643467,
      "learning_rate": 1.0365732150170168e-06,
      "loss": 0.9459,
      "step": 2658
    },
    {
      "epoch": 0.34424054115286273,
      "grad_norm": 0.2040132352795188,
      "learning_rate": 1.036360590719951e-06,
      "loss": 0.6287,
      "step": 2659
    },
    {
      "epoch": 0.3443700035602162,
      "grad_norm": 0.12573185782788923,
      "learning_rate": 1.0361478823941464e-06,
      "loss": 0.7859,
      "step": 2660
    },
    {
      "epoch": 0.3443700035602162,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8415300546448088,
      "eval_PRM F1 AUC": 0.463855421686747,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.77,
      "eval_PRM Recall": 0.927710843373494,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6228693127632141,
      "eval_runtime": 20.3388,
      "eval_samples_per_second": 2.163,
      "eval_steps_per_second": 0.148,
      "step": 2660
    },
    {
      "epoch": 0.34449946596756964,
      "grad_norm": 0.13621373743432413,
      "learning_rate": 1.0359350900830534e-06,
      "loss": 0.6208,
      "step": 2661
    },
    {
      "epoch": 0.34462892837492315,
      "grad_norm": 0.16741725297647234,
      "learning_rate": 1.035722213830139e-06,
      "loss": 0.687,
      "step": 2662
    },
    {
      "epoch": 0.3447583907822766,
      "grad_norm": 0.1156327315027558,
      "learning_rate": 1.0355092536788875e-06,
      "loss": 0.5619,
      "step": 2663
    },
    {
      "epoch": 0.34488785318963006,
      "grad_norm": 0.1212208220445449,
      "learning_rate": 1.0352962096728001e-06,
      "loss": 0.798,
      "step": 2664
    },
    {
      "epoch": 0.3450173155969835,
      "grad_norm": 0.11820724338405045,
      "learning_rate": 1.0350830818553957e-06,
      "loss": 0.6941,
      "step": 2665
    },
    {
      "epoch": 0.3450173155969835,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8415300546448088,
      "eval_PRM F1 AUC": 0.463855421686747,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.77,
      "eval_PRM Recall": 0.927710843373494,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6241121888160706,
      "eval_runtime": 19.897,
      "eval_samples_per_second": 2.211,
      "eval_steps_per_second": 0.151,
      "step": 2665
    },
    {
      "epoch": 0.345146778004337,
      "grad_norm": 0.15255628485005296,
      "learning_rate": 1.0348698702702099e-06,
      "loss": 0.7039,
      "step": 2666
    },
    {
      "epoch": 0.34527624041169047,
      "grad_norm": 0.12916764880859496,
      "learning_rate": 1.0346565749607952e-06,
      "loss": 0.7878,
      "step": 2667
    },
    {
      "epoch": 0.3454057028190439,
      "grad_norm": 0.10036332260030756,
      "learning_rate": 1.0344431959707219e-06,
      "loss": 0.7538,
      "step": 2668
    },
    {
      "epoch": 0.3455351652263974,
      "grad_norm": 0.10269489481516424,
      "learning_rate": 1.0342297333435766e-06,
      "loss": 0.6985,
      "step": 2669
    },
    {
      "epoch": 0.34566462763375083,
      "grad_norm": 0.12987353301378946,
      "learning_rate": 1.0340161871229637e-06,
      "loss": 0.8119,
      "step": 2670
    },
    {
      "epoch": 0.34566462763375083,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8415300546448088,
      "eval_PRM F1 AUC": 0.463855421686747,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.77,
      "eval_PRM Recall": 0.927710843373494,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6234019994735718,
      "eval_runtime": 20.5336,
      "eval_samples_per_second": 2.143,
      "eval_steps_per_second": 0.146,
      "step": 2670
    },
    {
      "epoch": 0.34579409004110434,
      "grad_norm": 0.11580370508488996,
      "learning_rate": 1.033802557352504e-06,
      "loss": 0.7507,
      "step": 2671
    },
    {
      "epoch": 0.3459235524484578,
      "grad_norm": 0.14533680127135218,
      "learning_rate": 1.033588844075836e-06,
      "loss": 0.7997,
      "step": 2672
    },
    {
      "epoch": 0.34605301485581125,
      "grad_norm": 0.20721079088658484,
      "learning_rate": 1.033375047336615e-06,
      "loss": 0.6609,
      "step": 2673
    },
    {
      "epoch": 0.3461824772631647,
      "grad_norm": 0.24362859962560357,
      "learning_rate": 1.0331611671785132e-06,
      "loss": 0.6876,
      "step": 2674
    },
    {
      "epoch": 0.34631193967051815,
      "grad_norm": 0.1431266209570801,
      "learning_rate": 1.0329472036452198e-06,
      "loss": 0.6792,
      "step": 2675
    },
    {
      "epoch": 0.34631193967051815,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8415300546448088,
      "eval_PRM F1 AUC": 0.463855421686747,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.77,
      "eval_PRM Recall": 0.927710843373494,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6265980005264282,
      "eval_runtime": 20.5805,
      "eval_samples_per_second": 2.138,
      "eval_steps_per_second": 0.146,
      "step": 2675
    },
    {
      "epoch": 0.34644140207787166,
      "grad_norm": 0.2415249924537078,
      "learning_rate": 1.0327331567804414e-06,
      "loss": 0.6059,
      "step": 2676
    },
    {
      "epoch": 0.3465708644852251,
      "grad_norm": 0.11246736554411921,
      "learning_rate": 1.0325190266279013e-06,
      "loss": 0.7438,
      "step": 2677
    },
    {
      "epoch": 0.34670032689257857,
      "grad_norm": 0.23526935635338145,
      "learning_rate": 1.0323048132313402e-06,
      "loss": 0.8691,
      "step": 2678
    },
    {
      "epoch": 0.346829789299932,
      "grad_norm": 0.12992693319037193,
      "learning_rate": 1.0320905166345153e-06,
      "loss": 0.8116,
      "step": 2679
    },
    {
      "epoch": 0.3469592517072855,
      "grad_norm": 0.1429023472760469,
      "learning_rate": 1.0318761368812008e-06,
      "loss": 0.8479,
      "step": 2680
    },
    {
      "epoch": 0.3469592517072855,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6237571239471436,
      "eval_runtime": 19.5882,
      "eval_samples_per_second": 2.246,
      "eval_steps_per_second": 0.153,
      "step": 2680
    },
    {
      "epoch": 0.347088714114639,
      "grad_norm": 0.1528258583526236,
      "learning_rate": 1.0316616740151885e-06,
      "loss": 0.6842,
      "step": 2681
    },
    {
      "epoch": 0.34721817652199244,
      "grad_norm": 0.16575898010334725,
      "learning_rate": 1.0314471280802865e-06,
      "loss": 0.6744,
      "step": 2682
    },
    {
      "epoch": 0.3473476389293459,
      "grad_norm": 0.11636198998472357,
      "learning_rate": 1.0312324991203202e-06,
      "loss": 0.6772,
      "step": 2683
    },
    {
      "epoch": 0.34747710133669935,
      "grad_norm": 0.11412395454855213,
      "learning_rate": 1.0310177871791322e-06,
      "loss": 0.7563,
      "step": 2684
    },
    {
      "epoch": 0.3476065637440528,
      "grad_norm": 0.1628665221110542,
      "learning_rate": 1.0308029923005815e-06,
      "loss": 0.6613,
      "step": 2685
    },
    {
      "epoch": 0.3476065637440528,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8415300546448088,
      "eval_PRM F1 AUC": 0.463855421686747,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.77,
      "eval_PRM Recall": 0.927710843373494,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6242897510528564,
      "eval_runtime": 20.3174,
      "eval_samples_per_second": 2.166,
      "eval_steps_per_second": 0.148,
      "step": 2685
    },
    {
      "epoch": 0.3477360261514063,
      "grad_norm": 0.32074752419295516,
      "learning_rate": 1.0305881145285443e-06,
      "loss": 0.9639,
      "step": 2686
    },
    {
      "epoch": 0.34786548855875976,
      "grad_norm": 0.16925582037262216,
      "learning_rate": 1.030373153906914e-06,
      "loss": 0.8098,
      "step": 2687
    },
    {
      "epoch": 0.3479949509661132,
      "grad_norm": 0.20625428743496452,
      "learning_rate": 1.0301581104796002e-06,
      "loss": 0.769,
      "step": 2688
    },
    {
      "epoch": 0.34812441337346667,
      "grad_norm": 0.14538778546930728,
      "learning_rate": 1.0299429842905303e-06,
      "loss": 0.6874,
      "step": 2689
    },
    {
      "epoch": 0.3482538757808201,
      "grad_norm": 0.1918668567156437,
      "learning_rate": 1.0297277753836482e-06,
      "loss": 0.9476,
      "step": 2690
    },
    {
      "epoch": 0.3482538757808201,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6214488744735718,
      "eval_runtime": 20.1661,
      "eval_samples_per_second": 2.182,
      "eval_steps_per_second": 0.149,
      "step": 2690
    },
    {
      "epoch": 0.34838333818817363,
      "grad_norm": 0.14259114546137175,
      "learning_rate": 1.0295124838029143e-06,
      "loss": 0.7056,
      "step": 2691
    },
    {
      "epoch": 0.3485128005955271,
      "grad_norm": 0.17073909321017092,
      "learning_rate": 1.0292971095923065e-06,
      "loss": 0.7964,
      "step": 2692
    },
    {
      "epoch": 0.34864226300288054,
      "grad_norm": 0.1826840704302061,
      "learning_rate": 1.0290816527958194e-06,
      "loss": 0.7531,
      "step": 2693
    },
    {
      "epoch": 0.348771725410234,
      "grad_norm": 0.11257348838214089,
      "learning_rate": 1.0288661134574643e-06,
      "loss": 0.7407,
      "step": 2694
    },
    {
      "epoch": 0.34890118781758744,
      "grad_norm": 0.11395645096633086,
      "learning_rate": 1.0286504916212697e-06,
      "loss": 0.738,
      "step": 2695
    },
    {
      "epoch": 0.34890118781758744,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6232244372367859,
      "eval_runtime": 20.2687,
      "eval_samples_per_second": 2.171,
      "eval_steps_per_second": 0.148,
      "step": 2695
    },
    {
      "epoch": 0.34903065022494095,
      "grad_norm": 0.1817868819971395,
      "learning_rate": 1.0284347873312804e-06,
      "loss": 0.6821,
      "step": 2696
    },
    {
      "epoch": 0.3491601126322944,
      "grad_norm": 0.11870942962323508,
      "learning_rate": 1.0282190006315584e-06,
      "loss": 0.6689,
      "step": 2697
    },
    {
      "epoch": 0.34928957503964786,
      "grad_norm": 0.19407363199818237,
      "learning_rate": 1.0280031315661828e-06,
      "loss": 0.858,
      "step": 2698
    },
    {
      "epoch": 0.3494190374470013,
      "grad_norm": 0.10666355644435606,
      "learning_rate": 1.0277871801792493e-06,
      "loss": 0.7075,
      "step": 2699
    },
    {
      "epoch": 0.34954849985435477,
      "grad_norm": 0.109787623204738,
      "learning_rate": 1.0275711465148699e-06,
      "loss": 0.762,
      "step": 2700
    },
    {
      "epoch": 0.34954849985435477,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6244673132896423,
      "eval_runtime": 20.7565,
      "eval_samples_per_second": 2.12,
      "eval_steps_per_second": 0.145,
      "step": 2700
    },
    {
      "epoch": 0.3496779622617083,
      "grad_norm": 0.16854566995053066,
      "learning_rate": 1.027355030617174e-06,
      "loss": 0.6534,
      "step": 2701
    },
    {
      "epoch": 0.34980742466906173,
      "grad_norm": 0.18780546546749913,
      "learning_rate": 1.027138832530308e-06,
      "loss": 0.6724,
      "step": 2702
    },
    {
      "epoch": 0.3499368870764152,
      "grad_norm": 0.14995509366403234,
      "learning_rate": 1.0269225522984342e-06,
      "loss": 0.9214,
      "step": 2703
    },
    {
      "epoch": 0.35006634948376864,
      "grad_norm": 0.12674617365309715,
      "learning_rate": 1.0267061899657327e-06,
      "loss": 0.738,
      "step": 2704
    },
    {
      "epoch": 0.3501958118911221,
      "grad_norm": 0.12061226051122682,
      "learning_rate": 1.0264897455763996e-06,
      "loss": 0.8054,
      "step": 2705
    },
    {
      "epoch": 0.3501958118911221,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6246448755264282,
      "eval_runtime": 20.7527,
      "eval_samples_per_second": 2.12,
      "eval_steps_per_second": 0.145,
      "step": 2705
    },
    {
      "epoch": 0.3503252742984756,
      "grad_norm": 0.12638993342007726,
      "learning_rate": 1.026273219174648e-06,
      "loss": 0.7728,
      "step": 2706
    },
    {
      "epoch": 0.35045473670582905,
      "grad_norm": 0.1318676914925145,
      "learning_rate": 1.0260566108047081e-06,
      "loss": 0.6862,
      "step": 2707
    },
    {
      "epoch": 0.3505841991131825,
      "grad_norm": 0.11870047900732393,
      "learning_rate": 1.0258399205108263e-06,
      "loss": 0.6953,
      "step": 2708
    },
    {
      "epoch": 0.35071366152053596,
      "grad_norm": 0.15554761835008327,
      "learning_rate": 1.0256231483372662e-06,
      "loss": 0.6822,
      "step": 2709
    },
    {
      "epoch": 0.35084312392788947,
      "grad_norm": 0.1283686676932774,
      "learning_rate": 1.0254062943283078e-06,
      "loss": 0.7934,
      "step": 2710
    },
    {
      "epoch": 0.35084312392788947,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.625,
      "eval_runtime": 20.9696,
      "eval_samples_per_second": 2.098,
      "eval_steps_per_second": 0.143,
      "step": 2710
    },
    {
      "epoch": 0.3509725863352429,
      "grad_norm": 0.1387498407129159,
      "learning_rate": 1.0251893585282477e-06,
      "loss": 0.7303,
      "step": 2711
    },
    {
      "epoch": 0.3511020487425964,
      "grad_norm": 0.14614957357735758,
      "learning_rate": 1.0249723409814e-06,
      "loss": 0.7035,
      "step": 2712
    },
    {
      "epoch": 0.3512315111499498,
      "grad_norm": 0.12361410582688825,
      "learning_rate": 1.0247552417320942e-06,
      "loss": 0.6467,
      "step": 2713
    },
    {
      "epoch": 0.3513609735573033,
      "grad_norm": 0.1366073718909575,
      "learning_rate": 1.024538060824678e-06,
      "loss": 0.7207,
      "step": 2714
    },
    {
      "epoch": 0.3514904359646568,
      "grad_norm": 0.12849696107517175,
      "learning_rate": 1.0243207983035143e-06,
      "loss": 0.7145,
      "step": 2715
    },
    {
      "epoch": 0.3514904359646568,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6260653138160706,
      "eval_runtime": 20.1479,
      "eval_samples_per_second": 2.184,
      "eval_steps_per_second": 0.149,
      "step": 2715
    },
    {
      "epoch": 0.35161989837201024,
      "grad_norm": 0.21569020090123464,
      "learning_rate": 1.0241034542129838e-06,
      "loss": 0.8474,
      "step": 2716
    },
    {
      "epoch": 0.3517493607793637,
      "grad_norm": 0.1985221261226252,
      "learning_rate": 1.0238860285974833e-06,
      "loss": 0.6548,
      "step": 2717
    },
    {
      "epoch": 0.35187882318671715,
      "grad_norm": 0.12307991479943844,
      "learning_rate": 1.0236685215014264e-06,
      "loss": 0.8794,
      "step": 2718
    },
    {
      "epoch": 0.3520082855940706,
      "grad_norm": 0.1299775778247633,
      "learning_rate": 1.0234509329692432e-06,
      "loss": 0.6559,
      "step": 2719
    },
    {
      "epoch": 0.3521377480014241,
      "grad_norm": 0.19516247011483848,
      "learning_rate": 1.0232332630453811e-06,
      "loss": 0.6024,
      "step": 2720
    },
    {
      "epoch": 0.3521377480014241,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6257102489471436,
      "eval_runtime": 20.0734,
      "eval_samples_per_second": 2.192,
      "eval_steps_per_second": 0.149,
      "step": 2720
    },
    {
      "epoch": 0.35226721040877756,
      "grad_norm": 0.1347558549548527,
      "learning_rate": 1.023015511774303e-06,
      "loss": 0.8176,
      "step": 2721
    },
    {
      "epoch": 0.352396672816131,
      "grad_norm": 0.13407150459822206,
      "learning_rate": 1.022797679200489e-06,
      "loss": 0.6418,
      "step": 2722
    },
    {
      "epoch": 0.35252613522348447,
      "grad_norm": 0.15077833826565806,
      "learning_rate": 1.0225797653684364e-06,
      "loss": 0.8223,
      "step": 2723
    },
    {
      "epoch": 0.3526555976308379,
      "grad_norm": 0.18608418341572158,
      "learning_rate": 1.022361770322658e-06,
      "loss": 0.8953,
      "step": 2724
    },
    {
      "epoch": 0.35278506003819143,
      "grad_norm": 0.11848103082757616,
      "learning_rate": 1.0221436941076835e-06,
      "loss": 0.8021,
      "step": 2725
    },
    {
      "epoch": 0.35278506003819143,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8415300546448088,
      "eval_PRM F1 AUC": 0.463855421686747,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.77,
      "eval_PRM Recall": 0.927710843373494,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6251775622367859,
      "eval_runtime": 20.1668,
      "eval_samples_per_second": 2.182,
      "eval_steps_per_second": 0.149,
      "step": 2725
    },
    {
      "epoch": 0.3529145224455449,
      "grad_norm": 0.2116902047789001,
      "learning_rate": 1.0219255367680603e-06,
      "loss": 0.897,
      "step": 2726
    },
    {
      "epoch": 0.35304398485289834,
      "grad_norm": 0.16139614503438354,
      "learning_rate": 1.0217072983483507e-06,
      "loss": 0.6772,
      "step": 2727
    },
    {
      "epoch": 0.3531734472602518,
      "grad_norm": 0.22100733821146543,
      "learning_rate": 1.0214889788931343e-06,
      "loss": 0.963,
      "step": 2728
    },
    {
      "epoch": 0.35330290966760525,
      "grad_norm": 0.12001734517579127,
      "learning_rate": 1.0212705784470077e-06,
      "loss": 0.7664,
      "step": 2729
    },
    {
      "epoch": 0.35343237207495876,
      "grad_norm": 0.14850484282638185,
      "learning_rate": 1.0210520970545835e-06,
      "loss": 0.7637,
      "step": 2730
    },
    {
      "epoch": 0.35343237207495876,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6253551244735718,
      "eval_runtime": 20.3843,
      "eval_samples_per_second": 2.159,
      "eval_steps_per_second": 0.147,
      "step": 2730
    },
    {
      "epoch": 0.3535618344823122,
      "grad_norm": 0.11724829907098497,
      "learning_rate": 1.0208335347604907e-06,
      "loss": 0.6985,
      "step": 2731
    },
    {
      "epoch": 0.35369129688966566,
      "grad_norm": 0.11813228026793304,
      "learning_rate": 1.0206148916093753e-06,
      "loss": 0.7053,
      "step": 2732
    },
    {
      "epoch": 0.3538207592970191,
      "grad_norm": 0.11701838798036542,
      "learning_rate": 1.0203961676458994e-06,
      "loss": 0.7621,
      "step": 2733
    },
    {
      "epoch": 0.35395022170437257,
      "grad_norm": 0.12572205898941735,
      "learning_rate": 1.020177362914742e-06,
      "loss": 0.8169,
      "step": 2734
    },
    {
      "epoch": 0.3540796841117261,
      "grad_norm": 0.16882291578642253,
      "learning_rate": 1.0199584774605985e-06,
      "loss": 0.7242,
      "step": 2735
    },
    {
      "epoch": 0.3540796841117261,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6232244372367859,
      "eval_runtime": 20.8221,
      "eval_samples_per_second": 2.113,
      "eval_steps_per_second": 0.144,
      "step": 2735
    },
    {
      "epoch": 0.35420914651907953,
      "grad_norm": 0.11653044732796512,
      "learning_rate": 1.0197395113281804e-06,
      "loss": 0.7321,
      "step": 2736
    },
    {
      "epoch": 0.354338608926433,
      "grad_norm": 0.11571717859652658,
      "learning_rate": 1.0195204645622163e-06,
      "loss": 0.7371,
      "step": 2737
    },
    {
      "epoch": 0.35446807133378644,
      "grad_norm": 0.11754856623706685,
      "learning_rate": 1.0193013372074507e-06,
      "loss": 0.7966,
      "step": 2738
    },
    {
      "epoch": 0.3545975337411399,
      "grad_norm": 0.13469327786209867,
      "learning_rate": 1.0190821293086448e-06,
      "loss": 0.7305,
      "step": 2739
    },
    {
      "epoch": 0.3547269961484934,
      "grad_norm": 0.12976069946690014,
      "learning_rate": 1.018862840910576e-06,
      "loss": 0.6887,
      "step": 2740
    },
    {
      "epoch": 0.3547269961484934,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6246448755264282,
      "eval_runtime": 19.6303,
      "eval_samples_per_second": 2.241,
      "eval_steps_per_second": 0.153,
      "step": 2740
    },
    {
      "epoch": 0.35485645855584685,
      "grad_norm": 0.12786321279416296,
      "learning_rate": 1.0186434720580391e-06,
      "loss": 0.7698,
      "step": 2741
    },
    {
      "epoch": 0.3549859209632003,
      "grad_norm": 0.10220789842983692,
      "learning_rate": 1.0184240227958442e-06,
      "loss": 0.7167,
      "step": 2742
    },
    {
      "epoch": 0.35511538337055376,
      "grad_norm": 0.21902866455938091,
      "learning_rate": 1.018204493168818e-06,
      "loss": 0.666,
      "step": 2743
    },
    {
      "epoch": 0.3552448457779072,
      "grad_norm": 0.16010945358101447,
      "learning_rate": 1.0179848832218044e-06,
      "loss": 0.6352,
      "step": 2744
    },
    {
      "epoch": 0.3553743081852607,
      "grad_norm": 0.11510753384554984,
      "learning_rate": 1.0177651929996632e-06,
      "loss": 0.8212,
      "step": 2745
    },
    {
      "epoch": 0.3553743081852607,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6258878111839294,
      "eval_runtime": 20.5043,
      "eval_samples_per_second": 2.146,
      "eval_steps_per_second": 0.146,
      "step": 2745
    },
    {
      "epoch": 0.3555037705926142,
      "grad_norm": 0.1734162631080907,
      "learning_rate": 1.01754542254727e-06,
      "loss": 0.6626,
      "step": 2746
    },
    {
      "epoch": 0.35563323299996763,
      "grad_norm": 0.10734341116006511,
      "learning_rate": 1.0173255719095178e-06,
      "loss": 0.7329,
      "step": 2747
    },
    {
      "epoch": 0.3557626954073211,
      "grad_norm": 0.1597415965179653,
      "learning_rate": 1.0171056411313155e-06,
      "loss": 0.8628,
      "step": 2748
    },
    {
      "epoch": 0.35589215781467454,
      "grad_norm": 0.1402622554858559,
      "learning_rate": 1.0168856302575884e-06,
      "loss": 0.7981,
      "step": 2749
    },
    {
      "epoch": 0.35602162022202805,
      "grad_norm": 0.17858580163135207,
      "learning_rate": 1.016665539333278e-06,
      "loss": 0.891,
      "step": 2750
    },
    {
      "epoch": 0.35602162022202805,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6242897510528564,
      "eval_runtime": 20.5857,
      "eval_samples_per_second": 2.137,
      "eval_steps_per_second": 0.146,
      "step": 2750
    },
    {
      "epoch": 0.3561510826293815,
      "grad_norm": 0.11870390244309614,
      "learning_rate": 1.0164453684033429e-06,
      "loss": 0.7383,
      "step": 2751
    },
    {
      "epoch": 0.35628054503673495,
      "grad_norm": 0.13655355148023363,
      "learning_rate": 1.0162251175127565e-06,
      "loss": 0.6854,
      "step": 2752
    },
    {
      "epoch": 0.3564100074440884,
      "grad_norm": 0.11107294337546103,
      "learning_rate": 1.0160047867065107e-06,
      "loss": 0.7694,
      "step": 2753
    },
    {
      "epoch": 0.3565394698514419,
      "grad_norm": 0.1132502104250624,
      "learning_rate": 1.0157843760296114e-06,
      "loss": 0.7491,
      "step": 2754
    },
    {
      "epoch": 0.35666893225879537,
      "grad_norm": 0.11393725242873398,
      "learning_rate": 1.0155638855270824e-06,
      "loss": 0.7587,
      "step": 2755
    },
    {
      "epoch": 0.35666893225879537,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6248224377632141,
      "eval_runtime": 20.6429,
      "eval_samples_per_second": 2.131,
      "eval_steps_per_second": 0.145,
      "step": 2755
    },
    {
      "epoch": 0.3567983946661488,
      "grad_norm": 0.131522192438505,
      "learning_rate": 1.0153433152439639e-06,
      "loss": 0.6835,
      "step": 2756
    },
    {
      "epoch": 0.3569278570735023,
      "grad_norm": 0.12180491173852623,
      "learning_rate": 1.015122665225311e-06,
      "loss": 0.7074,
      "step": 2757
    },
    {
      "epoch": 0.35705731948085573,
      "grad_norm": 0.20500319445625226,
      "learning_rate": 1.0149019355161964e-06,
      "loss": 0.6676,
      "step": 2758
    },
    {
      "epoch": 0.35718678188820924,
      "grad_norm": 0.12044705076411193,
      "learning_rate": 1.0146811261617085e-06,
      "loss": 0.7926,
      "step": 2759
    },
    {
      "epoch": 0.3573162442955627,
      "grad_norm": 0.10619269377339655,
      "learning_rate": 1.0144602372069522e-06,
      "loss": 0.7463,
      "step": 2760
    },
    {
      "epoch": 0.3573162442955627,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6260653138160706,
      "eval_runtime": 20.7202,
      "eval_samples_per_second": 2.124,
      "eval_steps_per_second": 0.145,
      "step": 2760
    },
    {
      "epoch": 0.35744570670291614,
      "grad_norm": 0.11891466555305893,
      "learning_rate": 1.0142392686970485e-06,
      "loss": 0.7551,
      "step": 2761
    },
    {
      "epoch": 0.3575751691102696,
      "grad_norm": 0.13609856749765495,
      "learning_rate": 1.0140182206771342e-06,
      "loss": 0.6885,
      "step": 2762
    },
    {
      "epoch": 0.35770463151762305,
      "grad_norm": 0.11425428381603756,
      "learning_rate": 1.0137970931923638e-06,
      "loss": 0.7473,
      "step": 2763
    },
    {
      "epoch": 0.35783409392497656,
      "grad_norm": 0.1582178144541826,
      "learning_rate": 1.0135758862879064e-06,
      "loss": 0.9011,
      "step": 2764
    },
    {
      "epoch": 0.35796355633233,
      "grad_norm": 0.1120917242428175,
      "learning_rate": 1.013354600008948e-06,
      "loss": 0.7391,
      "step": 2765
    },
    {
      "epoch": 0.35796355633233,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6246448755264282,
      "eval_runtime": 20.3676,
      "eval_samples_per_second": 2.16,
      "eval_steps_per_second": 0.147,
      "step": 2765
    },
    {
      "epoch": 0.35809301873968347,
      "grad_norm": 0.15703359705665684,
      "learning_rate": 1.0131332344006912e-06,
      "loss": 0.7189,
      "step": 2766
    },
    {
      "epoch": 0.3582224811470369,
      "grad_norm": 0.10909456100818825,
      "learning_rate": 1.012911789508354e-06,
      "loss": 0.7361,
      "step": 2767
    },
    {
      "epoch": 0.3583519435543904,
      "grad_norm": 0.11376200325607701,
      "learning_rate": 1.0126902653771713e-06,
      "loss": 0.6575,
      "step": 2768
    },
    {
      "epoch": 0.3584814059617439,
      "grad_norm": 0.1262968396491346,
      "learning_rate": 1.0124686620523935e-06,
      "loss": 0.8145,
      "step": 2769
    },
    {
      "epoch": 0.35861086836909734,
      "grad_norm": 0.13926513650013556,
      "learning_rate": 1.0122469795792882e-06,
      "loss": 0.8274,
      "step": 2770
    },
    {
      "epoch": 0.35861086836909734,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.625,
      "eval_runtime": 20.6053,
      "eval_samples_per_second": 2.135,
      "eval_steps_per_second": 0.146,
      "step": 2770
    },
    {
      "epoch": 0.3587403307764508,
      "grad_norm": 0.15627731270756506,
      "learning_rate": 1.012025218003138e-06,
      "loss": 0.6954,
      "step": 2771
    },
    {
      "epoch": 0.35886979318380424,
      "grad_norm": 0.12242087056096711,
      "learning_rate": 1.0118033773692423e-06,
      "loss": 0.7094,
      "step": 2772
    },
    {
      "epoch": 0.3589992555911577,
      "grad_norm": 0.1264656957752253,
      "learning_rate": 1.011581457722917e-06,
      "loss": 0.7654,
      "step": 2773
    },
    {
      "epoch": 0.3591287179985112,
      "grad_norm": 0.17999341009987485,
      "learning_rate": 1.011359459109493e-06,
      "loss": 0.5994,
      "step": 2774
    },
    {
      "epoch": 0.35925818040586466,
      "grad_norm": 0.10936517829904863,
      "learning_rate": 1.0111373815743181e-06,
      "loss": 0.7556,
      "step": 2775
    },
    {
      "epoch": 0.35925818040586466,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6253551244735718,
      "eval_runtime": 20.2195,
      "eval_samples_per_second": 2.176,
      "eval_steps_per_second": 0.148,
      "step": 2775
    },
    {
      "epoch": 0.3593876428132181,
      "grad_norm": 0.17755414915472456,
      "learning_rate": 1.010915225162757e-06,
      "loss": 0.6494,
      "step": 2776
    },
    {
      "epoch": 0.35951710522057156,
      "grad_norm": 0.1611005102659837,
      "learning_rate": 1.0106929899201884e-06,
      "loss": 0.7393,
      "step": 2777
    },
    {
      "epoch": 0.359646567627925,
      "grad_norm": 0.18229186403387432,
      "learning_rate": 1.0104706758920093e-06,
      "loss": 0.6657,
      "step": 2778
    },
    {
      "epoch": 0.3597760300352785,
      "grad_norm": 0.1896330364951413,
      "learning_rate": 1.0102482831236314e-06,
      "loss": 0.6361,
      "step": 2779
    },
    {
      "epoch": 0.359905492442632,
      "grad_norm": 0.12466767141565187,
      "learning_rate": 1.010025811660483e-06,
      "loss": 0.8573,
      "step": 2780
    },
    {
      "epoch": 0.359905492442632,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.625,
      "eval_runtime": 20.283,
      "eval_samples_per_second": 2.169,
      "eval_steps_per_second": 0.148,
      "step": 2780
    },
    {
      "epoch": 0.36003495484998543,
      "grad_norm": 0.14535080032320413,
      "learning_rate": 1.0098032615480086e-06,
      "loss": 0.8849,
      "step": 2781
    },
    {
      "epoch": 0.3601644172573389,
      "grad_norm": 0.1999147027160217,
      "learning_rate": 1.0095806328316683e-06,
      "loss": 0.6954,
      "step": 2782
    },
    {
      "epoch": 0.36029387966469234,
      "grad_norm": 0.15927990805136685,
      "learning_rate": 1.0093579255569387e-06,
      "loss": 0.6552,
      "step": 2783
    },
    {
      "epoch": 0.36042334207204585,
      "grad_norm": 0.15835118981907673,
      "learning_rate": 1.0091351397693123e-06,
      "loss": 0.6371,
      "step": 2784
    },
    {
      "epoch": 0.3605528044793993,
      "grad_norm": 0.10386258069867864,
      "learning_rate": 1.0089122755142974e-06,
      "loss": 0.7109,
      "step": 2785
    },
    {
      "epoch": 0.3605528044793993,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6257102489471436,
      "eval_runtime": 20.8722,
      "eval_samples_per_second": 2.108,
      "eval_steps_per_second": 0.144,
      "step": 2785
    },
    {
      "epoch": 0.36068226688675276,
      "grad_norm": 0.17030186734538136,
      "learning_rate": 1.0086893328374187e-06,
      "loss": 0.7106,
      "step": 2786
    },
    {
      "epoch": 0.3608117292941062,
      "grad_norm": 0.1550846234054508,
      "learning_rate": 1.0084663117842168e-06,
      "loss": 0.6976,
      "step": 2787
    },
    {
      "epoch": 0.36094119170145966,
      "grad_norm": 0.2264992995774288,
      "learning_rate": 1.0082432124002483e-06,
      "loss": 0.8905,
      "step": 2788
    },
    {
      "epoch": 0.36107065410881317,
      "grad_norm": 0.2007499073979965,
      "learning_rate": 1.0080200347310853e-06,
      "loss": 0.6783,
      "step": 2789
    },
    {
      "epoch": 0.3612001165161666,
      "grad_norm": 0.12898495169101912,
      "learning_rate": 1.0077967788223172e-06,
      "loss": 0.7843,
      "step": 2790
    },
    {
      "epoch": 0.3612001165161666,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6251775622367859,
      "eval_runtime": 20.7158,
      "eval_samples_per_second": 2.124,
      "eval_steps_per_second": 0.145,
      "step": 2790
    },
    {
      "epoch": 0.3613295789235201,
      "grad_norm": 0.1550602620277481,
      "learning_rate": 1.0075734447195478e-06,
      "loss": 0.6628,
      "step": 2791
    },
    {
      "epoch": 0.36145904133087353,
      "grad_norm": 0.12766404479282328,
      "learning_rate": 1.0073500324683983e-06,
      "loss": 0.6885,
      "step": 2792
    },
    {
      "epoch": 0.361588503738227,
      "grad_norm": 0.12716156577780283,
      "learning_rate": 1.0071265421145042e-06,
      "loss": 0.696,
      "step": 2793
    },
    {
      "epoch": 0.3617179661455805,
      "grad_norm": 0.10859967223212445,
      "learning_rate": 1.0069029737035188e-06,
      "loss": 0.7709,
      "step": 2794
    },
    {
      "epoch": 0.36184742855293395,
      "grad_norm": 0.14587594387496694,
      "learning_rate": 1.0066793272811104e-06,
      "loss": 0.6575,
      "step": 2795
    },
    {
      "epoch": 0.36184742855293395,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6225141882896423,
      "eval_runtime": 20.7019,
      "eval_samples_per_second": 2.125,
      "eval_steps_per_second": 0.145,
      "step": 2795
    },
    {
      "epoch": 0.3619768909602874,
      "grad_norm": 0.1745755996814566,
      "learning_rate": 1.0064556028929628e-06,
      "loss": 0.8647,
      "step": 2796
    },
    {
      "epoch": 0.36210635336764085,
      "grad_norm": 0.21949138484570782,
      "learning_rate": 1.0062318005847769e-06,
      "loss": 0.624,
      "step": 2797
    },
    {
      "epoch": 0.36223581577499436,
      "grad_norm": 0.14046554710327563,
      "learning_rate": 1.0060079204022681e-06,
      "loss": 0.6584,
      "step": 2798
    },
    {
      "epoch": 0.3623652781823478,
      "grad_norm": 0.2817723441517102,
      "learning_rate": 1.0057839623911692e-06,
      "loss": 0.925,
      "step": 2799
    },
    {
      "epoch": 0.36249474058970127,
      "grad_norm": 0.11721244501914892,
      "learning_rate": 1.0055599265972277e-06,
      "loss": 0.5902,
      "step": 2800
    },
    {
      "epoch": 0.36249474058970127,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6253551244735718,
      "eval_runtime": 20.7911,
      "eval_samples_per_second": 2.116,
      "eval_steps_per_second": 0.144,
      "step": 2800
    },
    {
      "epoch": 0.3626242029970547,
      "grad_norm": 0.22128517155194466,
      "learning_rate": 1.0053358130662076e-06,
      "loss": 0.6342,
      "step": 2801
    },
    {
      "epoch": 0.3627536654044082,
      "grad_norm": 0.17332877937342372,
      "learning_rate": 1.0051116218438886e-06,
      "loss": 0.5484,
      "step": 2802
    },
    {
      "epoch": 0.3628831278117617,
      "grad_norm": 0.18114344254945527,
      "learning_rate": 1.0048873529760663e-06,
      "loss": 0.6136,
      "step": 2803
    },
    {
      "epoch": 0.36301259021911514,
      "grad_norm": 0.13865993771006485,
      "learning_rate": 1.0046630065085524e-06,
      "loss": 0.795,
      "step": 2804
    },
    {
      "epoch": 0.3631420526264686,
      "grad_norm": 0.19357737146631498,
      "learning_rate": 1.0044385824871738e-06,
      "loss": 0.8451,
      "step": 2805
    },
    {
      "epoch": 0.3631420526264686,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6246448755264282,
      "eval_runtime": 20.4191,
      "eval_samples_per_second": 2.155,
      "eval_steps_per_second": 0.147,
      "step": 2805
    },
    {
      "epoch": 0.36327151503382205,
      "grad_norm": 0.11869806826167473,
      "learning_rate": 1.004214080957774e-06,
      "loss": 0.6781,
      "step": 2806
    },
    {
      "epoch": 0.3634009774411755,
      "grad_norm": 0.15712017771014644,
      "learning_rate": 1.0039895019662115e-06,
      "loss": 0.8055,
      "step": 2807
    },
    {
      "epoch": 0.363530439848529,
      "grad_norm": 0.1886070857501483,
      "learning_rate": 1.0037648455583619e-06,
      "loss": 0.7979,
      "step": 2808
    },
    {
      "epoch": 0.36365990225588246,
      "grad_norm": 0.09699037192668848,
      "learning_rate": 1.0035401117801152e-06,
      "loss": 0.6787,
      "step": 2809
    },
    {
      "epoch": 0.3637893646632359,
      "grad_norm": 0.1301472360259722,
      "learning_rate": 1.003315300677378e-06,
      "loss": 0.8563,
      "step": 2810
    },
    {
      "epoch": 0.3637893646632359,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6251775622367859,
      "eval_runtime": 20.73,
      "eval_samples_per_second": 2.123,
      "eval_steps_per_second": 0.145,
      "step": 2810
    },
    {
      "epoch": 0.36391882707058937,
      "grad_norm": 0.17629520010489655,
      "learning_rate": 1.0030904122960725e-06,
      "loss": 0.6196,
      "step": 2811
    },
    {
      "epoch": 0.3640482894779428,
      "grad_norm": 0.17110750514810258,
      "learning_rate": 1.002865446682137e-06,
      "loss": 0.8717,
      "step": 2812
    },
    {
      "epoch": 0.36417775188529633,
      "grad_norm": 0.11661881615758461,
      "learning_rate": 1.0026404038815246e-06,
      "loss": 0.6324,
      "step": 2813
    },
    {
      "epoch": 0.3643072142926498,
      "grad_norm": 0.17321263941495776,
      "learning_rate": 1.0024152839402055e-06,
      "loss": 0.6805,
      "step": 2814
    },
    {
      "epoch": 0.36443667670000324,
      "grad_norm": 0.18243577324490526,
      "learning_rate": 1.0021900869041648e-06,
      "loss": 0.9711,
      "step": 2815
    },
    {
      "epoch": 0.36443667670000324,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6253551244735718,
      "eval_runtime": 20.174,
      "eval_samples_per_second": 2.181,
      "eval_steps_per_second": 0.149,
      "step": 2815
    },
    {
      "epoch": 0.3645661391073567,
      "grad_norm": 0.35969707544601937,
      "learning_rate": 1.0019648128194035e-06,
      "loss": 1.0547,
      "step": 2816
    },
    {
      "epoch": 0.36469560151471014,
      "grad_norm": 0.15667076978530703,
      "learning_rate": 1.0017394617319387e-06,
      "loss": 0.8734,
      "step": 2817
    },
    {
      "epoch": 0.36482506392206365,
      "grad_norm": 0.1583859620711222,
      "learning_rate": 1.0015140336878024e-06,
      "loss": 0.6794,
      "step": 2818
    },
    {
      "epoch": 0.3649545263294171,
      "grad_norm": 0.13531924142054677,
      "learning_rate": 1.0012885287330432e-06,
      "loss": 0.8154,
      "step": 2819
    },
    {
      "epoch": 0.36508398873677056,
      "grad_norm": 0.23001703415401786,
      "learning_rate": 1.001062946913725e-06,
      "loss": 0.9592,
      "step": 2820
    },
    {
      "epoch": 0.36508398873677056,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6255326867103577,
      "eval_runtime": 20.3471,
      "eval_samples_per_second": 2.162,
      "eval_steps_per_second": 0.147,
      "step": 2820
    },
    {
      "epoch": 0.365213451144124,
      "grad_norm": 0.10967483737984751,
      "learning_rate": 1.0008372882759275e-06,
      "loss": 0.6429,
      "step": 2821
    },
    {
      "epoch": 0.36534291355147747,
      "grad_norm": 0.18289014759832528,
      "learning_rate": 1.000611552865746e-06,
      "loss": 0.7673,
      "step": 2822
    },
    {
      "epoch": 0.365472375958831,
      "grad_norm": 0.11244747425236269,
      "learning_rate": 1.0003857407292913e-06,
      "loss": 0.7706,
      "step": 2823
    },
    {
      "epoch": 0.36560183836618443,
      "grad_norm": 0.10748853614266381,
      "learning_rate": 1.0001598519126906e-06,
      "loss": 0.7705,
      "step": 2824
    },
    {
      "epoch": 0.3657313007735379,
      "grad_norm": 0.18569021687321619,
      "learning_rate": 9.999338864620857e-07,
      "loss": 0.6119,
      "step": 2825
    },
    {
      "epoch": 0.3657313007735379,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.623046875,
      "eval_runtime": 20.6616,
      "eval_samples_per_second": 2.13,
      "eval_steps_per_second": 0.145,
      "step": 2825
    },
    {
      "epoch": 0.36586076318089134,
      "grad_norm": 0.1404997605463703,
      "learning_rate": 9.997078444236353e-07,
      "loss": 0.8663,
      "step": 2826
    },
    {
      "epoch": 0.3659902255882448,
      "grad_norm": 0.12635455672583473,
      "learning_rate": 9.994817258435125e-07,
      "loss": 0.745,
      "step": 2827
    },
    {
      "epoch": 0.3661196879955983,
      "grad_norm": 0.19570866842944618,
      "learning_rate": 9.992555307679069e-07,
      "loss": 0.6388,
      "step": 2828
    },
    {
      "epoch": 0.36624915040295175,
      "grad_norm": 0.13974682735196295,
      "learning_rate": 9.990292592430233e-07,
      "loss": 0.7522,
      "step": 2829
    },
    {
      "epoch": 0.3663786128103052,
      "grad_norm": 0.12582437913045555,
      "learning_rate": 9.98802911315082e-07,
      "loss": 0.7726,
      "step": 2830
    },
    {
      "epoch": 0.3663786128103052,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6207386255264282,
      "eval_runtime": 19.9816,
      "eval_samples_per_second": 2.202,
      "eval_steps_per_second": 0.15,
      "step": 2830
    },
    {
      "epoch": 0.36650807521765866,
      "grad_norm": 0.12879098767911193,
      "learning_rate": 9.9857648703032e-07,
      "loss": 0.7976,
      "step": 2831
    },
    {
      "epoch": 0.3666375376250121,
      "grad_norm": 0.1113684299180072,
      "learning_rate": 9.98349986434988e-07,
      "loss": 0.6083,
      "step": 2832
    },
    {
      "epoch": 0.3667670000323656,
      "grad_norm": 0.20367847289459148,
      "learning_rate": 9.981234095753538e-07,
      "loss": 0.741,
      "step": 2833
    },
    {
      "epoch": 0.3668964624397191,
      "grad_norm": 0.16565516350602624,
      "learning_rate": 9.978967564977006e-07,
      "loss": 0.6956,
      "step": 2834
    },
    {
      "epoch": 0.3670259248470725,
      "grad_norm": 0.149223061260409,
      "learning_rate": 9.976700272483263e-07,
      "loss": 0.8771,
      "step": 2835
    },
    {
      "epoch": 0.3670259248470725,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6255326867103577,
      "eval_runtime": 19.825,
      "eval_samples_per_second": 2.219,
      "eval_steps_per_second": 0.151,
      "step": 2835
    },
    {
      "epoch": 0.367155387254426,
      "grad_norm": 0.1368947890017426,
      "learning_rate": 9.974432218735455e-07,
      "loss": 0.8422,
      "step": 2836
    },
    {
      "epoch": 0.36728484966177943,
      "grad_norm": 0.18992851341581415,
      "learning_rate": 9.972163404196876e-07,
      "loss": 0.8458,
      "step": 2837
    },
    {
      "epoch": 0.36741431206913294,
      "grad_norm": 0.1020817477929157,
      "learning_rate": 9.969893829330976e-07,
      "loss": 0.7202,
      "step": 2838
    },
    {
      "epoch": 0.3675437744764864,
      "grad_norm": 0.22575054944290548,
      "learning_rate": 9.96762349460136e-07,
      "loss": 0.693,
      "step": 2839
    },
    {
      "epoch": 0.36767323688383985,
      "grad_norm": 0.149959951158413,
      "learning_rate": 9.965352400471796e-07,
      "loss": 0.7114,
      "step": 2840
    },
    {
      "epoch": 0.36767323688383985,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6242897510528564,
      "eval_runtime": 20.3143,
      "eval_samples_per_second": 2.166,
      "eval_steps_per_second": 0.148,
      "step": 2840
    },
    {
      "epoch": 0.3678026992911933,
      "grad_norm": 0.2256009121617485,
      "learning_rate": 9.963080547406194e-07,
      "loss": 0.8301,
      "step": 2841
    },
    {
      "epoch": 0.3679321616985468,
      "grad_norm": 0.10723478803848362,
      "learning_rate": 9.960807935868634e-07,
      "loss": 0.7053,
      "step": 2842
    },
    {
      "epoch": 0.36806162410590026,
      "grad_norm": 0.130704032270869,
      "learning_rate": 9.958534566323336e-07,
      "loss": 0.5854,
      "step": 2843
    },
    {
      "epoch": 0.3681910865132537,
      "grad_norm": 0.13733285112940952,
      "learning_rate": 9.956260439234683e-07,
      "loss": 0.7458,
      "step": 2844
    },
    {
      "epoch": 0.36832054892060717,
      "grad_norm": 0.10798055070264732,
      "learning_rate": 9.953985555067216e-07,
      "loss": 0.6639,
      "step": 2845
    },
    {
      "epoch": 0.36832054892060717,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6218039989471436,
      "eval_runtime": 19.2582,
      "eval_samples_per_second": 2.285,
      "eval_steps_per_second": 0.156,
      "step": 2845
    },
    {
      "epoch": 0.3684500113279606,
      "grad_norm": 0.17381986305454794,
      "learning_rate": 9.95170991428562e-07,
      "loss": 0.6963,
      "step": 2846
    },
    {
      "epoch": 0.36857947373531413,
      "grad_norm": 0.11041443323013606,
      "learning_rate": 9.949433517354745e-07,
      "loss": 0.8016,
      "step": 2847
    },
    {
      "epoch": 0.3687089361426676,
      "grad_norm": 0.1342537235632201,
      "learning_rate": 9.947156364739591e-07,
      "loss": 0.7283,
      "step": 2848
    },
    {
      "epoch": 0.36883839855002104,
      "grad_norm": 0.11340855921890433,
      "learning_rate": 9.94487845690531e-07,
      "loss": 0.8123,
      "step": 2849
    },
    {
      "epoch": 0.3689678609573745,
      "grad_norm": 0.1148836825914116,
      "learning_rate": 9.942599794317212e-07,
      "loss": 0.7489,
      "step": 2850
    },
    {
      "epoch": 0.3689678609573745,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6242897510528564,
      "eval_runtime": 20.0357,
      "eval_samples_per_second": 2.196,
      "eval_steps_per_second": 0.15,
      "step": 2850
    },
    {
      "epoch": 0.36909732336472795,
      "grad_norm": 0.11598163033572025,
      "learning_rate": 9.940320377440763e-07,
      "loss": 0.7452,
      "step": 2851
    },
    {
      "epoch": 0.36922678577208146,
      "grad_norm": 0.18576733544059618,
      "learning_rate": 9.938040206741574e-07,
      "loss": 0.9213,
      "step": 2852
    },
    {
      "epoch": 0.3693562481794349,
      "grad_norm": 0.1541508092868443,
      "learning_rate": 9.93575928268542e-07,
      "loss": 0.7845,
      "step": 2853
    },
    {
      "epoch": 0.36948571058678836,
      "grad_norm": 0.13106887946643705,
      "learning_rate": 9.933477605738228e-07,
      "loss": 0.8201,
      "step": 2854
    },
    {
      "epoch": 0.3696151729941418,
      "grad_norm": 0.12708264906921904,
      "learning_rate": 9.931195176366073e-07,
      "loss": 0.7166,
      "step": 2855
    },
    {
      "epoch": 0.3696151729941418,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6214488744735718,
      "eval_runtime": 20.9235,
      "eval_samples_per_second": 2.103,
      "eval_steps_per_second": 0.143,
      "step": 2855
    },
    {
      "epoch": 0.36974463540149527,
      "grad_norm": 0.20759925584505856,
      "learning_rate": 9.92891199503519e-07,
      "loss": 0.9233,
      "step": 2856
    },
    {
      "epoch": 0.3698740978088488,
      "grad_norm": 0.1182791972346759,
      "learning_rate": 9.926628062211963e-07,
      "loss": 0.783,
      "step": 2857
    },
    {
      "epoch": 0.37000356021620223,
      "grad_norm": 0.19592505470800373,
      "learning_rate": 9.924343378362934e-07,
      "loss": 0.6206,
      "step": 2858
    },
    {
      "epoch": 0.3701330226235557,
      "grad_norm": 0.09279704123529552,
      "learning_rate": 9.92205794395479e-07,
      "loss": 0.676,
      "step": 2859
    },
    {
      "epoch": 0.37026248503090914,
      "grad_norm": 0.13837394202458592,
      "learning_rate": 9.919771759454386e-07,
      "loss": 0.7532,
      "step": 2860
    },
    {
      "epoch": 0.37026248503090914,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6228693127632141,
      "eval_runtime": 20.4147,
      "eval_samples_per_second": 2.155,
      "eval_steps_per_second": 0.147,
      "step": 2860
    },
    {
      "epoch": 0.3703919474382626,
      "grad_norm": 0.2749197806110179,
      "learning_rate": 9.917484825328717e-07,
      "loss": 0.9917,
      "step": 2861
    },
    {
      "epoch": 0.3705214098456161,
      "grad_norm": 0.1316747837115949,
      "learning_rate": 9.915197142044933e-07,
      "loss": 0.7505,
      "step": 2862
    },
    {
      "epoch": 0.37065087225296955,
      "grad_norm": 0.08914873041230541,
      "learning_rate": 9.912908710070344e-07,
      "loss": 0.6862,
      "step": 2863
    },
    {
      "epoch": 0.370780334660323,
      "grad_norm": 0.15896317627144219,
      "learning_rate": 9.910619529872408e-07,
      "loss": 0.8815,
      "step": 2864
    },
    {
      "epoch": 0.37090979706767646,
      "grad_norm": 0.13847414093318494,
      "learning_rate": 9.908329601918735e-07,
      "loss": 0.7295,
      "step": 2865
    },
    {
      "epoch": 0.37090979706767646,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6244673132896423,
      "eval_runtime": 20.6997,
      "eval_samples_per_second": 2.126,
      "eval_steps_per_second": 0.145,
      "step": 2865
    },
    {
      "epoch": 0.3710392594750299,
      "grad_norm": 0.133110081768871,
      "learning_rate": 9.90603892667709e-07,
      "loss": 0.7982,
      "step": 2866
    },
    {
      "epoch": 0.3711687218823834,
      "grad_norm": 0.1386340023273188,
      "learning_rate": 9.903747504615391e-07,
      "loss": 0.8813,
      "step": 2867
    },
    {
      "epoch": 0.3712981842897369,
      "grad_norm": 0.12937511884512193,
      "learning_rate": 9.901455336201704e-07,
      "loss": 0.7798,
      "step": 2868
    },
    {
      "epoch": 0.37142764669709033,
      "grad_norm": 0.10752945915865027,
      "learning_rate": 9.899162421904254e-07,
      "loss": 0.7404,
      "step": 2869
    },
    {
      "epoch": 0.3715571091044438,
      "grad_norm": 0.19242984857130088,
      "learning_rate": 9.896868762191415e-07,
      "loss": 0.6317,
      "step": 2870
    },
    {
      "epoch": 0.3715571091044438,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.619140625,
      "eval_runtime": 21.217,
      "eval_samples_per_second": 2.074,
      "eval_steps_per_second": 0.141,
      "step": 2870
    },
    {
      "epoch": 0.37168657151179724,
      "grad_norm": 0.16927344739673844,
      "learning_rate": 9.894574357531708e-07,
      "loss": 0.6477,
      "step": 2871
    },
    {
      "epoch": 0.37181603391915075,
      "grad_norm": 0.17839618971087978,
      "learning_rate": 9.892279208393822e-07,
      "loss": 0.8459,
      "step": 2872
    },
    {
      "epoch": 0.3719454963265042,
      "grad_norm": 0.16120500834577478,
      "learning_rate": 9.889983315246579e-07,
      "loss": 0.7805,
      "step": 2873
    },
    {
      "epoch": 0.37207495873385765,
      "grad_norm": 0.12044896850124229,
      "learning_rate": 9.887686678558966e-07,
      "loss": 0.7936,
      "step": 2874
    },
    {
      "epoch": 0.3722044211412111,
      "grad_norm": 0.196596106119202,
      "learning_rate": 9.885389298800116e-07,
      "loss": 0.8745,
      "step": 2875
    },
    {
      "epoch": 0.3722044211412111,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6209161877632141,
      "eval_runtime": 20.5102,
      "eval_samples_per_second": 2.145,
      "eval_steps_per_second": 0.146,
      "step": 2875
    },
    {
      "epoch": 0.37233388354856456,
      "grad_norm": 0.10856748403868338,
      "learning_rate": 9.883091176439317e-07,
      "loss": 0.754,
      "step": 2876
    },
    {
      "epoch": 0.37246334595591807,
      "grad_norm": 0.10659709455841847,
      "learning_rate": 9.880792311946002e-07,
      "loss": 0.7786,
      "step": 2877
    },
    {
      "epoch": 0.3725928083632715,
      "grad_norm": 0.10193075910208234,
      "learning_rate": 9.878492705789768e-07,
      "loss": 0.703,
      "step": 2878
    },
    {
      "epoch": 0.372722270770625,
      "grad_norm": 0.27018595012761365,
      "learning_rate": 9.876192358440352e-07,
      "loss": 0.8665,
      "step": 2879
    },
    {
      "epoch": 0.37285173317797843,
      "grad_norm": 0.2192297338242662,
      "learning_rate": 9.873891270367647e-07,
      "loss": 0.8121,
      "step": 2880
    },
    {
      "epoch": 0.37285173317797843,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.62109375,
      "eval_runtime": 19.8857,
      "eval_samples_per_second": 2.213,
      "eval_steps_per_second": 0.151,
      "step": 2880
    },
    {
      "epoch": 0.3729811955853319,
      "grad_norm": 0.10687442096021815,
      "learning_rate": 9.871589442041697e-07,
      "loss": 0.6902,
      "step": 2881
    },
    {
      "epoch": 0.3731106579926854,
      "grad_norm": 0.1190306321432561,
      "learning_rate": 9.869286873932696e-07,
      "loss": 0.8167,
      "step": 2882
    },
    {
      "epoch": 0.37324012040003884,
      "grad_norm": 0.11254228096620711,
      "learning_rate": 9.866983566510994e-07,
      "loss": 0.6722,
      "step": 2883
    },
    {
      "epoch": 0.3733695828073923,
      "grad_norm": 0.20083797984382643,
      "learning_rate": 9.864679520247084e-07,
      "loss": 0.8613,
      "step": 2884
    },
    {
      "epoch": 0.37349904521474575,
      "grad_norm": 0.13950813833990097,
      "learning_rate": 9.862374735611615e-07,
      "loss": 0.6549,
      "step": 2885
    },
    {
      "epoch": 0.37349904521474575,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6202059388160706,
      "eval_runtime": 20.0603,
      "eval_samples_per_second": 2.193,
      "eval_steps_per_second": 0.15,
      "step": 2885
    },
    {
      "epoch": 0.37362850762209926,
      "grad_norm": 0.10845310820567529,
      "learning_rate": 9.860069213075388e-07,
      "loss": 0.7228,
      "step": 2886
    },
    {
      "epoch": 0.3737579700294527,
      "grad_norm": 0.29280655580377535,
      "learning_rate": 9.85776295310935e-07,
      "loss": 1.0416,
      "step": 2887
    },
    {
      "epoch": 0.37388743243680617,
      "grad_norm": 0.1732519069824405,
      "learning_rate": 9.855455956184603e-07,
      "loss": 0.9031,
      "step": 2888
    },
    {
      "epoch": 0.3740168948441596,
      "grad_norm": 0.1442694528254996,
      "learning_rate": 9.853148222772401e-07,
      "loss": 0.6379,
      "step": 2889
    },
    {
      "epoch": 0.3741463572515131,
      "grad_norm": 0.210322397461802,
      "learning_rate": 9.85083975334414e-07,
      "loss": 0.5677,
      "step": 2890
    },
    {
      "epoch": 0.3741463572515131,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6193181872367859,
      "eval_runtime": 20.1002,
      "eval_samples_per_second": 2.189,
      "eval_steps_per_second": 0.149,
      "step": 2890
    },
    {
      "epoch": 0.3742758196588666,
      "grad_norm": 0.13601654045992892,
      "learning_rate": 9.848530548371373e-07,
      "loss": 0.6717,
      "step": 2891
    },
    {
      "epoch": 0.37440528206622004,
      "grad_norm": 0.161632015639682,
      "learning_rate": 9.846220608325803e-07,
      "loss": 0.8059,
      "step": 2892
    },
    {
      "epoch": 0.3745347444735735,
      "grad_norm": 0.14065969069575485,
      "learning_rate": 9.843909933679281e-07,
      "loss": 0.6122,
      "step": 2893
    },
    {
      "epoch": 0.37466420688092694,
      "grad_norm": 0.15436419168454768,
      "learning_rate": 9.84159852490381e-07,
      "loss": 0.7066,
      "step": 2894
    },
    {
      "epoch": 0.3747936692882804,
      "grad_norm": 0.10746396080468937,
      "learning_rate": 9.839286382471544e-07,
      "loss": 0.676,
      "step": 2895
    },
    {
      "epoch": 0.3747936692882804,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6221590638160706,
      "eval_runtime": 20.844,
      "eval_samples_per_second": 2.111,
      "eval_steps_per_second": 0.144,
      "step": 2895
    },
    {
      "epoch": 0.3749231316956339,
      "grad_norm": 0.1424428802203209,
      "learning_rate": 9.83697350685478e-07,
      "loss": 0.6363,
      "step": 2896
    },
    {
      "epoch": 0.37505259410298736,
      "grad_norm": 0.17816174444259716,
      "learning_rate": 9.834659898525973e-07,
      "loss": 0.7845,
      "step": 2897
    },
    {
      "epoch": 0.3751820565103408,
      "grad_norm": 0.14746849829737738,
      "learning_rate": 9.832345557957727e-07,
      "loss": 0.7157,
      "step": 2898
    },
    {
      "epoch": 0.37531151891769426,
      "grad_norm": 0.12054749979829008,
      "learning_rate": 9.830030485622787e-07,
      "loss": 0.6608,
      "step": 2899
    },
    {
      "epoch": 0.3754409813250477,
      "grad_norm": 0.12373977935547659,
      "learning_rate": 9.827714681994055e-07,
      "loss": 0.7675,
      "step": 2900
    },
    {
      "epoch": 0.3754409813250477,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6198508739471436,
      "eval_runtime": 20.7095,
      "eval_samples_per_second": 2.125,
      "eval_steps_per_second": 0.145,
      "step": 2900
    },
    {
      "epoch": 0.3755704437324012,
      "grad_norm": 0.1025071281951921,
      "learning_rate": 9.825398147544585e-07,
      "loss": 0.686,
      "step": 2901
    },
    {
      "epoch": 0.3756999061397547,
      "grad_norm": 0.13855938622203118,
      "learning_rate": 9.823080882747574e-07,
      "loss": 0.8522,
      "step": 2902
    },
    {
      "epoch": 0.37582936854710813,
      "grad_norm": 0.2007044350243728,
      "learning_rate": 9.820762888076369e-07,
      "loss": 0.8264,
      "step": 2903
    },
    {
      "epoch": 0.3759588309544616,
      "grad_norm": 0.14910523321869876,
      "learning_rate": 9.818444164004465e-07,
      "loss": 0.8308,
      "step": 2904
    },
    {
      "epoch": 0.37608829336181504,
      "grad_norm": 0.14577532600215753,
      "learning_rate": 9.816124711005514e-07,
      "loss": 0.7068,
      "step": 2905
    },
    {
      "epoch": 0.37608829336181504,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.619140625,
      "eval_runtime": 20.3865,
      "eval_samples_per_second": 2.158,
      "eval_steps_per_second": 0.147,
      "step": 2905
    },
    {
      "epoch": 0.37621775576916855,
      "grad_norm": 0.10747425253265232,
      "learning_rate": 9.81380452955331e-07,
      "loss": 0.7368,
      "step": 2906
    },
    {
      "epoch": 0.376347218176522,
      "grad_norm": 0.16717792766842757,
      "learning_rate": 9.811483620121796e-07,
      "loss": 0.6138,
      "step": 2907
    },
    {
      "epoch": 0.37647668058387546,
      "grad_norm": 0.1391080390369382,
      "learning_rate": 9.80916198318506e-07,
      "loss": 0.7,
      "step": 2908
    },
    {
      "epoch": 0.3766061429912289,
      "grad_norm": 0.17051142695814003,
      "learning_rate": 9.806839619217353e-07,
      "loss": 0.6503,
      "step": 2909
    },
    {
      "epoch": 0.37673560539858236,
      "grad_norm": 0.13019302956333167,
      "learning_rate": 9.804516528693056e-07,
      "loss": 0.8029,
      "step": 2910
    },
    {
      "epoch": 0.37673560539858236,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6207386255264282,
      "eval_runtime": 20.7218,
      "eval_samples_per_second": 2.123,
      "eval_steps_per_second": 0.145,
      "step": 2910
    },
    {
      "epoch": 0.37686506780593587,
      "grad_norm": 0.18289266796873432,
      "learning_rate": 9.802192712086713e-07,
      "loss": 0.8928,
      "step": 2911
    },
    {
      "epoch": 0.3769945302132893,
      "grad_norm": 0.16085694768119557,
      "learning_rate": 9.799868169873011e-07,
      "loss": 0.9163,
      "step": 2912
    },
    {
      "epoch": 0.3771239926206428,
      "grad_norm": 0.13196289845834508,
      "learning_rate": 9.797542902526779e-07,
      "loss": 0.6727,
      "step": 2913
    },
    {
      "epoch": 0.37725345502799623,
      "grad_norm": 0.18268665637029438,
      "learning_rate": 9.795216910523006e-07,
      "loss": 0.6888,
      "step": 2914
    },
    {
      "epoch": 0.3773829174353497,
      "grad_norm": 0.11628970806080995,
      "learning_rate": 9.79289019433682e-07,
      "loss": 0.7185,
      "step": 2915
    },
    {
      "epoch": 0.3773829174353497,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6207386255264282,
      "eval_runtime": 20.739,
      "eval_samples_per_second": 2.122,
      "eval_steps_per_second": 0.145,
      "step": 2915
    },
    {
      "epoch": 0.3775123798427032,
      "grad_norm": 0.14413704069575264,
      "learning_rate": 9.7905627544435e-07,
      "loss": 0.839,
      "step": 2916
    },
    {
      "epoch": 0.37764184225005665,
      "grad_norm": 0.1621074243863637,
      "learning_rate": 9.788234591318476e-07,
      "loss": 0.8204,
      "step": 2917
    },
    {
      "epoch": 0.3777713046574101,
      "grad_norm": 0.1433322417043379,
      "learning_rate": 9.785905705437318e-07,
      "loss": 0.8381,
      "step": 2918
    },
    {
      "epoch": 0.37790076706476355,
      "grad_norm": 0.16274607016418605,
      "learning_rate": 9.783576097275751e-07,
      "loss": 0.6924,
      "step": 2919
    },
    {
      "epoch": 0.378030229472117,
      "grad_norm": 0.12163896570491606,
      "learning_rate": 9.781245767309644e-07,
      "loss": 0.7842,
      "step": 2920
    },
    {
      "epoch": 0.378030229472117,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6219815611839294,
      "eval_runtime": 19.7696,
      "eval_samples_per_second": 2.226,
      "eval_steps_per_second": 0.152,
      "step": 2920
    },
    {
      "epoch": 0.3781596918794705,
      "grad_norm": 0.1711905994356605,
      "learning_rate": 9.778914716015012e-07,
      "loss": 0.6746,
      "step": 2921
    },
    {
      "epoch": 0.37828915428682397,
      "grad_norm": 0.16140728647998293,
      "learning_rate": 9.776582943868023e-07,
      "loss": 0.6947,
      "step": 2922
    },
    {
      "epoch": 0.3784186166941774,
      "grad_norm": 0.12306153651013194,
      "learning_rate": 9.774250451344987e-07,
      "loss": 0.7122,
      "step": 2923
    },
    {
      "epoch": 0.3785480791015309,
      "grad_norm": 0.1033864125430614,
      "learning_rate": 9.771917238922362e-07,
      "loss": 0.7742,
      "step": 2924
    },
    {
      "epoch": 0.37867754150888433,
      "grad_norm": 0.14437274575784365,
      "learning_rate": 9.769583307076757e-07,
      "loss": 0.8065,
      "step": 2925
    },
    {
      "epoch": 0.37867754150888433,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.619140625,
      "eval_runtime": 20.1743,
      "eval_samples_per_second": 2.181,
      "eval_steps_per_second": 0.149,
      "step": 2925
    },
    {
      "epoch": 0.37880700391623784,
      "grad_norm": 0.12768225459881963,
      "learning_rate": 9.767248656284923e-07,
      "loss": 0.6436,
      "step": 2926
    },
    {
      "epoch": 0.3789364663235913,
      "grad_norm": 0.19376344204495813,
      "learning_rate": 9.764913287023756e-07,
      "loss": 0.6767,
      "step": 2927
    },
    {
      "epoch": 0.37906592873094475,
      "grad_norm": 0.17283468523541107,
      "learning_rate": 9.762577199770308e-07,
      "loss": 0.8026,
      "step": 2928
    },
    {
      "epoch": 0.3791953911382982,
      "grad_norm": 0.12603899729750628,
      "learning_rate": 9.76024039500177e-07,
      "loss": 0.8136,
      "step": 2929
    },
    {
      "epoch": 0.3793248535456517,
      "grad_norm": 0.11609354010733448,
      "learning_rate": 9.757902873195483e-07,
      "loss": 0.6344,
      "step": 2930
    },
    {
      "epoch": 0.3793248535456517,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6194957494735718,
      "eval_runtime": 19.8324,
      "eval_samples_per_second": 2.219,
      "eval_steps_per_second": 0.151,
      "step": 2930
    },
    {
      "epoch": 0.37945431595300516,
      "grad_norm": 0.20049524647476305,
      "learning_rate": 9.755564634828932e-07,
      "loss": 0.6322,
      "step": 2931
    },
    {
      "epoch": 0.3795837783603586,
      "grad_norm": 0.12885743062912688,
      "learning_rate": 9.75322568037975e-07,
      "loss": 0.8063,
      "step": 2932
    },
    {
      "epoch": 0.37971324076771207,
      "grad_norm": 0.12177216476930756,
      "learning_rate": 9.750886010325714e-07,
      "loss": 0.6625,
      "step": 2933
    },
    {
      "epoch": 0.3798427031750655,
      "grad_norm": 0.3506098334993272,
      "learning_rate": 9.748545625144754e-07,
      "loss": 0.9276,
      "step": 2934
    },
    {
      "epoch": 0.37997216558241903,
      "grad_norm": 0.12137965121930647,
      "learning_rate": 9.746204525314933e-07,
      "loss": 0.7565,
      "step": 2935
    },
    {
      "epoch": 0.37997216558241903,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6189630627632141,
      "eval_runtime": 19.9211,
      "eval_samples_per_second": 2.209,
      "eval_steps_per_second": 0.151,
      "step": 2935
    },
    {
      "epoch": 0.3801016279897725,
      "grad_norm": 0.1699646772733835,
      "learning_rate": 9.743862711314474e-07,
      "loss": 0.8575,
      "step": 2936
    },
    {
      "epoch": 0.38023109039712594,
      "grad_norm": 0.12775046294973874,
      "learning_rate": 9.741520183621743e-07,
      "loss": 0.8253,
      "step": 2937
    },
    {
      "epoch": 0.3803605528044794,
      "grad_norm": 0.13359175217216981,
      "learning_rate": 9.739176942715242e-07,
      "loss": 0.6843,
      "step": 2938
    },
    {
      "epoch": 0.38049001521183284,
      "grad_norm": 0.1341214894303281,
      "learning_rate": 9.736832989073627e-07,
      "loss": 0.6556,
      "step": 2939
    },
    {
      "epoch": 0.38061947761918635,
      "grad_norm": 0.11736389042064316,
      "learning_rate": 9.7344883231757e-07,
      "loss": 0.791,
      "step": 2940
    },
    {
      "epoch": 0.38061947761918635,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6212713122367859,
      "eval_runtime": 20.0101,
      "eval_samples_per_second": 2.199,
      "eval_steps_per_second": 0.15,
      "step": 2940
    },
    {
      "epoch": 0.3807489400265398,
      "grad_norm": 0.11781251235836517,
      "learning_rate": 9.732142945500406e-07,
      "loss": 0.7035,
      "step": 2941
    },
    {
      "epoch": 0.38087840243389326,
      "grad_norm": 0.13578805227585336,
      "learning_rate": 9.729796856526835e-07,
      "loss": 0.7827,
      "step": 2942
    },
    {
      "epoch": 0.3810078648412467,
      "grad_norm": 0.12253230081529547,
      "learning_rate": 9.727450056734224e-07,
      "loss": 0.8142,
      "step": 2943
    },
    {
      "epoch": 0.38113732724860017,
      "grad_norm": 0.12978567074701472,
      "learning_rate": 9.725102546601954e-07,
      "loss": 0.6324,
      "step": 2944
    },
    {
      "epoch": 0.3812667896559537,
      "grad_norm": 0.15150022400599913,
      "learning_rate": 9.72275432660955e-07,
      "loss": 0.8481,
      "step": 2945
    },
    {
      "epoch": 0.3812667896559537,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6187855005264282,
      "eval_runtime": 20.1174,
      "eval_samples_per_second": 2.187,
      "eval_steps_per_second": 0.149,
      "step": 2945
    },
    {
      "epoch": 0.38139625206330713,
      "grad_norm": 0.12496904279896774,
      "learning_rate": 9.720405397236691e-07,
      "loss": 0.7777,
      "step": 2946
    },
    {
      "epoch": 0.3815257144706606,
      "grad_norm": 0.17337355995235101,
      "learning_rate": 9.718055758963181e-07,
      "loss": 0.8457,
      "step": 2947
    },
    {
      "epoch": 0.38165517687801404,
      "grad_norm": 0.29201564674071445,
      "learning_rate": 9.71570541226899e-07,
      "loss": 0.9115,
      "step": 2948
    },
    {
      "epoch": 0.3817846392853675,
      "grad_norm": 0.11403922625743408,
      "learning_rate": 9.713354357634226e-07,
      "loss": 0.7421,
      "step": 2949
    },
    {
      "epoch": 0.381914101692721,
      "grad_norm": 0.17551327863611849,
      "learning_rate": 9.71100259553913e-07,
      "loss": 0.719,
      "step": 2950
    },
    {
      "epoch": 0.381914101692721,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6175426244735718,
      "eval_runtime": 20.7676,
      "eval_samples_per_second": 2.119,
      "eval_steps_per_second": 0.144,
      "step": 2950
    },
    {
      "epoch": 0.38204356410007445,
      "grad_norm": 0.2683226372989131,
      "learning_rate": 9.708650126464107e-07,
      "loss": 0.8911,
      "step": 2951
    },
    {
      "epoch": 0.3821730265074279,
      "grad_norm": 0.12698481816769971,
      "learning_rate": 9.706296950889689e-07,
      "loss": 0.802,
      "step": 2952
    },
    {
      "epoch": 0.38230248891478136,
      "grad_norm": 0.14449187930380666,
      "learning_rate": 9.703943069296563e-07,
      "loss": 0.8191,
      "step": 2953
    },
    {
      "epoch": 0.3824319513221348,
      "grad_norm": 0.17288285607499515,
      "learning_rate": 9.701588482165559e-07,
      "loss": 0.9005,
      "step": 2954
    },
    {
      "epoch": 0.3825614137294883,
      "grad_norm": 0.11128933102164836,
      "learning_rate": 9.699233189977645e-07,
      "loss": 0.7004,
      "step": 2955
    },
    {
      "epoch": 0.3825614137294883,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6205610632896423,
      "eval_runtime": 20.4147,
      "eval_samples_per_second": 2.155,
      "eval_steps_per_second": 0.147,
      "step": 2955
    },
    {
      "epoch": 0.3826908761368418,
      "grad_norm": 0.14522726841511438,
      "learning_rate": 9.696877193213938e-07,
      "loss": 0.8127,
      "step": 2956
    },
    {
      "epoch": 0.3828203385441952,
      "grad_norm": 0.10892132776729214,
      "learning_rate": 9.694520492355702e-07,
      "loss": 0.7451,
      "step": 2957
    },
    {
      "epoch": 0.3829498009515487,
      "grad_norm": 0.2910502845523496,
      "learning_rate": 9.692163087884336e-07,
      "loss": 0.8813,
      "step": 2958
    },
    {
      "epoch": 0.38307926335890213,
      "grad_norm": 0.1327059346049371,
      "learning_rate": 9.689804980281392e-07,
      "loss": 0.6716,
      "step": 2959
    },
    {
      "epoch": 0.38320872576625564,
      "grad_norm": 0.112394389186232,
      "learning_rate": 9.687446170028559e-07,
      "loss": 0.7485,
      "step": 2960
    },
    {
      "epoch": 0.38320872576625564,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6184303760528564,
      "eval_runtime": 20.5626,
      "eval_samples_per_second": 2.14,
      "eval_steps_per_second": 0.146,
      "step": 2960
    },
    {
      "epoch": 0.3833381881736091,
      "grad_norm": 0.11608336240880529,
      "learning_rate": 9.685086657607668e-07,
      "loss": 0.7906,
      "step": 2961
    },
    {
      "epoch": 0.38346765058096255,
      "grad_norm": 0.1069424901883626,
      "learning_rate": 9.682726443500707e-07,
      "loss": 0.6702,
      "step": 2962
    },
    {
      "epoch": 0.383597112988316,
      "grad_norm": 0.12422148674825287,
      "learning_rate": 9.680365528189787e-07,
      "loss": 0.7229,
      "step": 2963
    },
    {
      "epoch": 0.38372657539566946,
      "grad_norm": 0.11725133943594489,
      "learning_rate": 9.678003912157178e-07,
      "loss": 0.6699,
      "step": 2964
    },
    {
      "epoch": 0.38385603780302296,
      "grad_norm": 0.16245212612765417,
      "learning_rate": 9.67564159588529e-07,
      "loss": 0.8082,
      "step": 2965
    },
    {
      "epoch": 0.38385603780302296,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6159446239471436,
      "eval_runtime": 20.4146,
      "eval_samples_per_second": 2.155,
      "eval_steps_per_second": 0.147,
      "step": 2965
    },
    {
      "epoch": 0.3839855002103764,
      "grad_norm": 0.2723255227454983,
      "learning_rate": 9.67327857985667e-07,
      "loss": 0.99,
      "step": 2966
    },
    {
      "epoch": 0.38411496261772987,
      "grad_norm": 0.11293641649025954,
      "learning_rate": 9.670914864554014e-07,
      "loss": 0.6996,
      "step": 2967
    },
    {
      "epoch": 0.3842444250250833,
      "grad_norm": 0.127342427222284,
      "learning_rate": 9.668550450460156e-07,
      "loss": 0.804,
      "step": 2968
    },
    {
      "epoch": 0.38437388743243683,
      "grad_norm": 0.12020174994253749,
      "learning_rate": 9.666185338058083e-07,
      "loss": 0.7474,
      "step": 2969
    },
    {
      "epoch": 0.3845033498397903,
      "grad_norm": 0.15999556900609463,
      "learning_rate": 9.663819527830906e-07,
      "loss": 0.6677,
      "step": 2970
    },
    {
      "epoch": 0.3845033498397903,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6166548132896423,
      "eval_runtime": 20.3236,
      "eval_samples_per_second": 2.165,
      "eval_steps_per_second": 0.148,
      "step": 2970
    },
    {
      "epoch": 0.38463281224714374,
      "grad_norm": 0.16644370180520565,
      "learning_rate": 9.661453020261898e-07,
      "loss": 0.7926,
      "step": 2971
    },
    {
      "epoch": 0.3847622746544972,
      "grad_norm": 0.13131534994832633,
      "learning_rate": 9.659085815834463e-07,
      "loss": 0.7896,
      "step": 2972
    },
    {
      "epoch": 0.38489173706185065,
      "grad_norm": 0.10275317655902884,
      "learning_rate": 9.656717915032153e-07,
      "loss": 0.6989,
      "step": 2973
    },
    {
      "epoch": 0.38502119946920416,
      "grad_norm": 0.1332665948345099,
      "learning_rate": 9.654349318338655e-07,
      "loss": 0.7637,
      "step": 2974
    },
    {
      "epoch": 0.3851506618765576,
      "grad_norm": 0.1414660732726309,
      "learning_rate": 9.651980026237805e-07,
      "loss": 0.7838,
      "step": 2975
    },
    {
      "epoch": 0.3851506618765576,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8540540540540541,
      "eval_PRM F1 AUC": 0.4759036144578313,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7745098039215687,
      "eval_PRM Recall": 0.9518072289156626,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6166548132896423,
      "eval_runtime": 19.8971,
      "eval_samples_per_second": 2.211,
      "eval_steps_per_second": 0.151,
      "step": 2975
    },
    {
      "epoch": 0.38528012428391106,
      "grad_norm": 0.10030760074649209,
      "learning_rate": 9.649610039213584e-07,
      "loss": 0.6753,
      "step": 2976
    },
    {
      "epoch": 0.3854095866912645,
      "grad_norm": 0.16798701373016434,
      "learning_rate": 9.647239357750101e-07,
      "loss": 0.9323,
      "step": 2977
    },
    {
      "epoch": 0.38553904909861797,
      "grad_norm": 0.13719937524078818,
      "learning_rate": 9.644867982331623e-07,
      "loss": 0.6757,
      "step": 2978
    },
    {
      "epoch": 0.3856685115059715,
      "grad_norm": 0.3258988939345767,
      "learning_rate": 9.642495913442548e-07,
      "loss": 1.0037,
      "step": 2979
    },
    {
      "epoch": 0.38579797391332493,
      "grad_norm": 0.2500398927720127,
      "learning_rate": 9.64012315156742e-07,
      "loss": 0.6485,
      "step": 2980
    },
    {
      "epoch": 0.38579797391332493,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6166548132896423,
      "eval_runtime": 20.3246,
      "eval_samples_per_second": 2.165,
      "eval_steps_per_second": 0.148,
      "step": 2980
    },
    {
      "epoch": 0.3859274363206784,
      "grad_norm": 0.14195969638249414,
      "learning_rate": 9.637749697190924e-07,
      "loss": 0.847,
      "step": 2981
    },
    {
      "epoch": 0.38605689872803184,
      "grad_norm": 0.19284062836922308,
      "learning_rate": 9.635375550797885e-07,
      "loss": 0.626,
      "step": 2982
    },
    {
      "epoch": 0.3861863611353853,
      "grad_norm": 0.14693518026601537,
      "learning_rate": 9.633000712873272e-07,
      "loss": 0.8036,
      "step": 2983
    },
    {
      "epoch": 0.3863158235427388,
      "grad_norm": 0.1276245523638328,
      "learning_rate": 9.630625183902193e-07,
      "loss": 0.8201,
      "step": 2984
    },
    {
      "epoch": 0.38644528595009225,
      "grad_norm": 0.12996353056823864,
      "learning_rate": 9.628248964369898e-07,
      "loss": 0.6917,
      "step": 2985
    },
    {
      "epoch": 0.38644528595009225,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6184303760528564,
      "eval_runtime": 19.8595,
      "eval_samples_per_second": 2.216,
      "eval_steps_per_second": 0.151,
      "step": 2985
    },
    {
      "epoch": 0.3865747483574457,
      "grad_norm": 0.14461082624521332,
      "learning_rate": 9.625872054761777e-07,
      "loss": 0.7872,
      "step": 2986
    },
    {
      "epoch": 0.38670421076479916,
      "grad_norm": 0.10981781767572543,
      "learning_rate": 9.623494455563365e-07,
      "loss": 0.7173,
      "step": 2987
    },
    {
      "epoch": 0.3868336731721526,
      "grad_norm": 0.18090521427014442,
      "learning_rate": 9.62111616726033e-07,
      "loss": 0.6464,
      "step": 2988
    },
    {
      "epoch": 0.3869631355795061,
      "grad_norm": 0.15986378981868013,
      "learning_rate": 9.61873719033849e-07,
      "loss": 0.8995,
      "step": 2989
    },
    {
      "epoch": 0.3870925979868596,
      "grad_norm": 0.11538584014845356,
      "learning_rate": 9.6163575252838e-07,
      "loss": 0.7437,
      "step": 2990
    },
    {
      "epoch": 0.3870925979868596,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6162996888160706,
      "eval_runtime": 20.4872,
      "eval_samples_per_second": 2.148,
      "eval_steps_per_second": 0.146,
      "step": 2990
    },
    {
      "epoch": 0.38722206039421303,
      "grad_norm": 0.1886340654569701,
      "learning_rate": 9.61397717258235e-07,
      "loss": 0.6417,
      "step": 2991
    },
    {
      "epoch": 0.3873515228015665,
      "grad_norm": 0.12209644550522541,
      "learning_rate": 9.611596132720377e-07,
      "loss": 0.7263,
      "step": 2992
    },
    {
      "epoch": 0.38748098520891994,
      "grad_norm": 0.13047979406009533,
      "learning_rate": 9.609214406184258e-07,
      "loss": 0.6543,
      "step": 2993
    },
    {
      "epoch": 0.38761044761627345,
      "grad_norm": 0.22390987067116666,
      "learning_rate": 9.60683199346051e-07,
      "loss": 0.9119,
      "step": 2994
    },
    {
      "epoch": 0.3877399100236269,
      "grad_norm": 0.20221461348124425,
      "learning_rate": 9.604448895035787e-07,
      "loss": 0.9195,
      "step": 2995
    },
    {
      "epoch": 0.3877399100236269,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6148792505264282,
      "eval_runtime": 20.7237,
      "eval_samples_per_second": 2.123,
      "eval_steps_per_second": 0.145,
      "step": 2995
    },
    {
      "epoch": 0.38786937243098035,
      "grad_norm": 0.18992182302509658,
      "learning_rate": 9.602065111396886e-07,
      "loss": 0.6474,
      "step": 2996
    },
    {
      "epoch": 0.3879988348383338,
      "grad_norm": 0.13798151412899176,
      "learning_rate": 9.59968064303074e-07,
      "loss": 0.7939,
      "step": 2997
    },
    {
      "epoch": 0.38812829724568726,
      "grad_norm": 0.14072398182924187,
      "learning_rate": 9.59729549042443e-07,
      "loss": 0.7848,
      "step": 2998
    },
    {
      "epoch": 0.38825775965304077,
      "grad_norm": 0.08599506231353685,
      "learning_rate": 9.594909654065168e-07,
      "loss": 0.6457,
      "step": 2999
    },
    {
      "epoch": 0.3883872220603942,
      "grad_norm": 0.11391958686440753,
      "learning_rate": 9.592523134440312e-07,
      "loss": 0.6804,
      "step": 3000
    },
    {
      "epoch": 0.3883872220603942,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8478260869565217,
      "eval_PRM F1 AUC": 0.46987951807228917,
      "eval_PRM F1 Neg": 0.0,
      "eval_PRM NPV": 0.0,
      "eval_PRM Precision": 0.7722772277227723,
      "eval_PRM Recall": 0.9397590361445783,
      "eval_PRM Specificty": 0.0,
      "eval_loss": 0.6141690611839294,
      "eval_runtime": 20.919,
      "eval_samples_per_second": 2.103,
      "eval_steps_per_second": 0.143,
      "step": 3000
    }
  ],
  "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": 2518198355329024.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}