{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.12948755300896703,
  "eval_steps": 5,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0,
      "eval_PRM Accuracy": 0.3867924528301887,
      "eval_PRM F1": 0.4144144144144144,
      "eval_PRM F1 AUC": 0.5298585646935569,
      "eval_PRM F1 AUC (fixed)": 0.33813514929282346,
      "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": 0.8678227066993713,
      "eval_runtime": 16.2347,
      "eval_samples_per_second": 2.71,
      "eval_steps_per_second": 0.123,
      "step": 0
    },
    {
      "epoch": 0.000258975106017934,
      "grad_norm": 63.037703867152466,
      "learning_rate": 3.2299741602067187e-09,
      "loss": 0.9478,
      "step": 1
    },
    {
      "epoch": 0.000517950212035868,
      "grad_norm": 39.472536090782,
      "learning_rate": 6.4599483204134375e-09,
      "loss": 0.9176,
      "step": 2
    },
    {
      "epoch": 0.0007769253180538021,
      "grad_norm": 40.61452474226669,
      "learning_rate": 9.689922480620155e-09,
      "loss": 0.994,
      "step": 3
    },
    {
      "epoch": 0.001035900424071736,
      "grad_norm": 36.95892027782838,
      "learning_rate": 1.2919896640826875e-08,
      "loss": 0.9114,
      "step": 4
    },
    {
      "epoch": 0.00129487553008967,
      "grad_norm": 26.036252649177914,
      "learning_rate": 1.6149870801033593e-08,
      "loss": 0.9668,
      "step": 5
    },
    {
      "epoch": 0.00129487553008967,
      "eval_PRM Accuracy": 0.4056603773584906,
      "eval_PRM F1": 0.4424778761061947,
      "eval_PRM F1 AUC": 0.5419067574646412,
      "eval_PRM F1 AUC (fixed)": 0.3386589837611315,
      "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": 0.8697418570518494,
      "eval_runtime": 14.7356,
      "eval_samples_per_second": 2.986,
      "eval_steps_per_second": 0.136,
      "step": 5
    },
    {
      "epoch": 0.0015538506361076042,
      "grad_norm": 40.337292444544296,
      "learning_rate": 1.937984496124031e-08,
      "loss": 1.0354,
      "step": 6
    },
    {
      "epoch": 0.0018128257421255382,
      "grad_norm": 51.91063484757722,
      "learning_rate": 2.260981912144703e-08,
      "loss": 0.9882,
      "step": 7
    },
    {
      "epoch": 0.002071800848143472,
      "grad_norm": 51.663899635887994,
      "learning_rate": 2.583979328165375e-08,
      "loss": 0.9176,
      "step": 8
    },
    {
      "epoch": 0.0023307759541614063,
      "grad_norm": 21.81186522663788,
      "learning_rate": 2.9069767441860468e-08,
      "loss": 0.8714,
      "step": 9
    },
    {
      "epoch": 0.00258975106017934,
      "grad_norm": 25.886467614209714,
      "learning_rate": 3.2299741602067186e-08,
      "loss": 0.7792,
      "step": 10
    },
    {
      "epoch": 0.00258975106017934,
      "eval_PRM Accuracy": 0.4056603773584906,
      "eval_PRM F1": 0.43243243243243246,
      "eval_PRM F1 AUC": 0.5576217915138816,
      "eval_PRM F1 AUC (fixed)": 0.3397066526977475,
      "eval_PRM F1 Neg": 0.37623762376237624,
      "eval_PRM NPV": 0.24358974358974358,
      "eval_PRM Precision": 0.8571428571428571,
      "eval_PRM Recall": 0.2891566265060241,
      "eval_PRM Specificty": 0.8260869565217391,
      "eval_loss": 0.8677586913108826,
      "eval_runtime": 14.9411,
      "eval_samples_per_second": 2.945,
      "eval_steps_per_second": 0.134,
      "step": 10
    },
    {
      "epoch": 0.0028487261661972745,
      "grad_norm": 44.33097404486639,
      "learning_rate": 3.552971576227391e-08,
      "loss": 0.9529,
      "step": 11
    },
    {
      "epoch": 0.0031077012722152083,
      "grad_norm": 39.28825527565396,
      "learning_rate": 3.875968992248062e-08,
      "loss": 0.9612,
      "step": 12
    },
    {
      "epoch": 0.003366676378233142,
      "grad_norm": 26.936193809280923,
      "learning_rate": 4.198966408268734e-08,
      "loss": 0.9125,
      "step": 13
    },
    {
      "epoch": 0.0036256514842510764,
      "grad_norm": 63.91618838754899,
      "learning_rate": 4.521963824289406e-08,
      "loss": 1.0074,
      "step": 14
    },
    {
      "epoch": 0.0038846265902690103,
      "grad_norm": 46.647418475437426,
      "learning_rate": 4.844961240310078e-08,
      "loss": 0.9009,
      "step": 15
    },
    {
      "epoch": 0.0038846265902690103,
      "eval_PRM Accuracy": 0.4056603773584906,
      "eval_PRM F1": 0.4424778761061947,
      "eval_PRM F1 AUC": 0.5419067574646412,
      "eval_PRM F1 AUC (fixed)": 0.3407543216343636,
      "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": 0.8698608875274658,
      "eval_runtime": 10.4028,
      "eval_samples_per_second": 4.23,
      "eval_steps_per_second": 0.192,
      "step": 15
    },
    {
      "epoch": 0.004143601696286944,
      "grad_norm": 57.58762296760476,
      "learning_rate": 5.16795865633075e-08,
      "loss": 0.9692,
      "step": 16
    },
    {
      "epoch": 0.004402576802304879,
      "grad_norm": 30.661267401226237,
      "learning_rate": 5.4909560723514214e-08,
      "loss": 0.8378,
      "step": 17
    },
    {
      "epoch": 0.004661551908322813,
      "grad_norm": 41.39269160524219,
      "learning_rate": 5.8139534883720935e-08,
      "loss": 0.9658,
      "step": 18
    },
    {
      "epoch": 0.0049205270143407465,
      "grad_norm": 42.52726716449715,
      "learning_rate": 6.136950904392765e-08,
      "loss": 0.9987,
      "step": 19
    },
    {
      "epoch": 0.00517950212035868,
      "grad_norm": 49.82620333190475,
      "learning_rate": 6.459948320413437e-08,
      "loss": 0.9629,
      "step": 20
    },
    {
      "epoch": 0.00517950212035868,
      "eval_PRM Accuracy": 0.4056603773584906,
      "eval_PRM F1": 0.4424778761061947,
      "eval_PRM F1 AUC": 0.5419067574646412,
      "eval_PRM F1 AUC (fixed)": 0.33944473546359355,
      "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": 0.8663817048072815,
      "eval_runtime": 10.5034,
      "eval_samples_per_second": 4.189,
      "eval_steps_per_second": 0.19,
      "step": 20
    },
    {
      "epoch": 0.005438477226376614,
      "grad_norm": 40.625187370837494,
      "learning_rate": 6.782945736434109e-08,
      "loss": 0.9758,
      "step": 21
    },
    {
      "epoch": 0.005697452332394549,
      "grad_norm": 29.05974126113703,
      "learning_rate": 7.105943152454781e-08,
      "loss": 0.871,
      "step": 22
    },
    {
      "epoch": 0.005956427438412483,
      "grad_norm": 30.951805478184244,
      "learning_rate": 7.428940568475452e-08,
      "loss": 0.8993,
      "step": 23
    },
    {
      "epoch": 0.006215402544430417,
      "grad_norm": 40.05628218268796,
      "learning_rate": 7.751937984496124e-08,
      "loss": 0.939,
      "step": 24
    },
    {
      "epoch": 0.0064743776504483505,
      "grad_norm": 64.28427259734805,
      "learning_rate": 8.074935400516796e-08,
      "loss": 1.0051,
      "step": 25
    },
    {
      "epoch": 0.0064743776504483505,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 AUC (fixed)": 0.3399685699319015,
      "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": 0.8646280169487,
      "eval_runtime": 14.5121,
      "eval_samples_per_second": 3.032,
      "eval_steps_per_second": 0.138,
      "step": 25
    },
    {
      "epoch": 0.006733352756466284,
      "grad_norm": 56.89452099484511,
      "learning_rate": 8.397932816537469e-08,
      "loss": 0.986,
      "step": 26
    },
    {
      "epoch": 0.006992327862484219,
      "grad_norm": 49.624626549694966,
      "learning_rate": 8.72093023255814e-08,
      "loss": 0.9552,
      "step": 27
    },
    {
      "epoch": 0.007251302968502153,
      "grad_norm": 40.268884202852206,
      "learning_rate": 9.043927648578811e-08,
      "loss": 0.8838,
      "step": 28
    },
    {
      "epoch": 0.007510278074520087,
      "grad_norm": 35.58022077420101,
      "learning_rate": 9.366925064599485e-08,
      "loss": 0.8722,
      "step": 29
    },
    {
      "epoch": 0.007769253180538021,
      "grad_norm": 42.379835980673334,
      "learning_rate": 9.689922480620156e-08,
      "loss": 0.9022,
      "step": 30
    },
    {
      "epoch": 0.007769253180538021,
      "eval_PRM Accuracy": 0.4056603773584906,
      "eval_PRM F1": 0.4424778761061947,
      "eval_PRM F1 AUC": 0.5419067574646412,
      "eval_PRM F1 AUC (fixed)": 0.33970665269774747,
      "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": 0.8620710372924805,
      "eval_runtime": 14.5493,
      "eval_samples_per_second": 3.024,
      "eval_steps_per_second": 0.137,
      "step": 30
    },
    {
      "epoch": 0.008028228286555954,
      "grad_norm": 34.42647540348086,
      "learning_rate": 1.0012919896640826e-07,
      "loss": 0.9226,
      "step": 31
    },
    {
      "epoch": 0.008287203392573888,
      "grad_norm": 16.754621860181707,
      "learning_rate": 1.03359173126615e-07,
      "loss": 1.0476,
      "step": 32
    },
    {
      "epoch": 0.008546178498591822,
      "grad_norm": 41.44570425860884,
      "learning_rate": 1.0658914728682171e-07,
      "loss": 0.906,
      "step": 33
    },
    {
      "epoch": 0.008805153604609758,
      "grad_norm": 38.57423040321388,
      "learning_rate": 1.0981912144702843e-07,
      "loss": 0.9152,
      "step": 34
    },
    {
      "epoch": 0.009064128710627692,
      "grad_norm": 53.23182535449974,
      "learning_rate": 1.1304909560723515e-07,
      "loss": 1.0426,
      "step": 35
    },
    {
      "epoch": 0.009064128710627692,
      "eval_PRM Accuracy": 0.41509433962264153,
      "eval_PRM F1": 0.45614035087719296,
      "eval_PRM F1 AUC": 0.5479308538501834,
      "eval_PRM F1 AUC (fixed)": 0.3397066526977475,
      "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": 0.8567813634872437,
      "eval_runtime": 15.0601,
      "eval_samples_per_second": 2.922,
      "eval_steps_per_second": 0.133,
      "step": 35
    },
    {
      "epoch": 0.009323103816645625,
      "grad_norm": 50.17753754565222,
      "learning_rate": 1.1627906976744187e-07,
      "loss": 0.9266,
      "step": 36
    },
    {
      "epoch": 0.00958207892266356,
      "grad_norm": 43.70938983864694,
      "learning_rate": 1.1950904392764858e-07,
      "loss": 0.8187,
      "step": 37
    },
    {
      "epoch": 0.009841054028681493,
      "grad_norm": 44.16135701565119,
      "learning_rate": 1.227390180878553e-07,
      "loss": 0.9802,
      "step": 38
    },
    {
      "epoch": 0.010100029134699427,
      "grad_norm": 45.38369838408391,
      "learning_rate": 1.2596899224806202e-07,
      "loss": 0.8363,
      "step": 39
    },
    {
      "epoch": 0.01035900424071736,
      "grad_norm": 52.82723187942695,
      "learning_rate": 1.2919896640826874e-07,
      "loss": 1.0043,
      "step": 40
    },
    {
      "epoch": 0.01035900424071736,
      "eval_PRM Accuracy": 0.4339622641509434,
      "eval_PRM F1": 0.4827586206896552,
      "eval_PRM F1 AUC": 0.5599790466212677,
      "eval_PRM F1 AUC (fixed)": 0.33708748035620745,
      "eval_PRM F1 Neg": 0.375,
      "eval_PRM NPV": 0.2465753424657534,
      "eval_PRM Precision": 0.8484848484848485,
      "eval_PRM Recall": 0.3373493975903614,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8464098572731018,
      "eval_runtime": 14.8041,
      "eval_samples_per_second": 2.972,
      "eval_steps_per_second": 0.135,
      "step": 40
    },
    {
      "epoch": 0.010617979346735295,
      "grad_norm": 35.02370077516764,
      "learning_rate": 1.3242894056847546e-07,
      "loss": 0.8568,
      "step": 41
    },
    {
      "epoch": 0.010876954452753228,
      "grad_norm": 38.713476056558825,
      "learning_rate": 1.3565891472868218e-07,
      "loss": 0.9738,
      "step": 42
    },
    {
      "epoch": 0.011135929558771162,
      "grad_norm": 48.99054930472666,
      "learning_rate": 1.3888888888888888e-07,
      "loss": 0.8334,
      "step": 43
    },
    {
      "epoch": 0.011394904664789098,
      "grad_norm": 35.69913234062826,
      "learning_rate": 1.4211886304909563e-07,
      "loss": 0.842,
      "step": 44
    },
    {
      "epoch": 0.011653879770807032,
      "grad_norm": 47.80559671444046,
      "learning_rate": 1.4534883720930235e-07,
      "loss": 0.9162,
      "step": 45
    },
    {
      "epoch": 0.011653879770807032,
      "eval_PRM Accuracy": 0.4339622641509434,
      "eval_PRM F1": 0.4827586206896552,
      "eval_PRM F1 AUC": 0.5599790466212677,
      "eval_PRM F1 AUC (fixed)": 0.3428496595075956,
      "eval_PRM F1 Neg": 0.375,
      "eval_PRM NPV": 0.2465753424657534,
      "eval_PRM Precision": 0.8484848484848485,
      "eval_PRM Recall": 0.3373493975903614,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.8389459252357483,
      "eval_runtime": 14.6941,
      "eval_samples_per_second": 2.994,
      "eval_steps_per_second": 0.136,
      "step": 45
    },
    {
      "epoch": 0.011912854876824966,
      "grad_norm": 38.95935999442876,
      "learning_rate": 1.4857881136950904e-07,
      "loss": 1.0287,
      "step": 46
    },
    {
      "epoch": 0.0121718299828429,
      "grad_norm": 45.84894726759758,
      "learning_rate": 1.518087855297158e-07,
      "loss": 0.9409,
      "step": 47
    },
    {
      "epoch": 0.012430805088860833,
      "grad_norm": 49.66077506241443,
      "learning_rate": 1.5503875968992249e-07,
      "loss": 0.882,
      "step": 48
    },
    {
      "epoch": 0.012689780194878767,
      "grad_norm": 26.814130429582953,
      "learning_rate": 1.582687338501292e-07,
      "loss": 0.7798,
      "step": 49
    },
    {
      "epoch": 0.012948755300896701,
      "grad_norm": 32.7672953300368,
      "learning_rate": 1.6149870801033593e-07,
      "loss": 0.8485,
      "step": 50
    },
    {
      "epoch": 0.012948755300896701,
      "eval_PRM Accuracy": 0.4339622641509434,
      "eval_PRM F1": 0.4915254237288136,
      "eval_PRM F1 AUC": 0.5442640125720272,
      "eval_PRM F1 AUC (fixed)": 0.34154007333682557,
      "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": 0.8221114873886108,
      "eval_runtime": 15.0333,
      "eval_samples_per_second": 2.927,
      "eval_steps_per_second": 0.133,
      "step": 50
    },
    {
      "epoch": 0.013207730406914635,
      "grad_norm": 40.046840263198526,
      "learning_rate": 1.6472868217054268e-07,
      "loss": 0.9715,
      "step": 51
    },
    {
      "epoch": 0.013466705512932569,
      "grad_norm": 27.3304257780402,
      "learning_rate": 1.6795865633074937e-07,
      "loss": 0.7938,
      "step": 52
    },
    {
      "epoch": 0.013725680618950504,
      "grad_norm": 52.78649060661054,
      "learning_rate": 1.711886304909561e-07,
      "loss": 0.9807,
      "step": 53
    },
    {
      "epoch": 0.013984655724968438,
      "grad_norm": 31.60652199020792,
      "learning_rate": 1.744186046511628e-07,
      "loss": 0.8377,
      "step": 54
    },
    {
      "epoch": 0.014243630830986372,
      "grad_norm": 36.740966951868984,
      "learning_rate": 1.776485788113695e-07,
      "loss": 0.9034,
      "step": 55
    },
    {
      "epoch": 0.014243630830986372,
      "eval_PRM Accuracy": 0.4528301886792453,
      "eval_PRM F1": 0.5396825396825397,
      "eval_PRM F1 AUC": 0.5091671031953903,
      "eval_PRM F1 AUC (fixed)": 0.3428496595075956,
      "eval_PRM F1 Neg": 0.32558139534883723,
      "eval_PRM NPV": 0.2222222222222222,
      "eval_PRM Precision": 0.7906976744186046,
      "eval_PRM Recall": 0.40963855421686746,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.7895786762237549,
      "eval_runtime": 14.5999,
      "eval_samples_per_second": 3.014,
      "eval_steps_per_second": 0.137,
      "step": 55
    },
    {
      "epoch": 0.014502605937004306,
      "grad_norm": 13.797714830369891,
      "learning_rate": 1.8087855297157623e-07,
      "loss": 0.7442,
      "step": 56
    },
    {
      "epoch": 0.01476158104302224,
      "grad_norm": 35.405763235862786,
      "learning_rate": 1.8410852713178298e-07,
      "loss": 0.7909,
      "step": 57
    },
    {
      "epoch": 0.015020556149040173,
      "grad_norm": 40.58250255476464,
      "learning_rate": 1.873385012919897e-07,
      "loss": 0.8553,
      "step": 58
    },
    {
      "epoch": 0.015279531255058107,
      "grad_norm": 42.238417195704244,
      "learning_rate": 1.905684754521964e-07,
      "loss": 0.8204,
      "step": 59
    },
    {
      "epoch": 0.015538506361076041,
      "grad_norm": 42.32301478324044,
      "learning_rate": 1.9379844961240311e-07,
      "loss": 0.9246,
      "step": 60
    },
    {
      "epoch": 0.015538506361076041,
      "eval_PRM Accuracy": 0.4528301886792453,
      "eval_PRM F1": 0.5396825396825397,
      "eval_PRM F1 AUC": 0.5091671031953903,
      "eval_PRM F1 AUC (fixed)": 0.3391828182294395,
      "eval_PRM F1 Neg": 0.32558139534883723,
      "eval_PRM NPV": 0.2222222222222222,
      "eval_PRM Precision": 0.7906976744186046,
      "eval_PRM Recall": 0.40963855421686746,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.7745277285575867,
      "eval_runtime": 14.6553,
      "eval_samples_per_second": 3.002,
      "eval_steps_per_second": 0.136,
      "step": 60
    },
    {
      "epoch": 0.015797481467093977,
      "grad_norm": 16.34880758084372,
      "learning_rate": 1.9702842377260983e-07,
      "loss": 0.8876,
      "step": 61
    },
    {
      "epoch": 0.01605645657311191,
      "grad_norm": 33.018050747698034,
      "learning_rate": 2.0025839793281653e-07,
      "loss": 0.7665,
      "step": 62
    },
    {
      "epoch": 0.016315431679129844,
      "grad_norm": 35.941705491024585,
      "learning_rate": 2.0348837209302328e-07,
      "loss": 0.9996,
      "step": 63
    },
    {
      "epoch": 0.016574406785147777,
      "grad_norm": 21.384319021829366,
      "learning_rate": 2.0671834625323e-07,
      "loss": 0.7215,
      "step": 64
    },
    {
      "epoch": 0.016833381891165712,
      "grad_norm": 14.176127306038236,
      "learning_rate": 2.0994832041343672e-07,
      "loss": 0.7689,
      "step": 65
    },
    {
      "epoch": 0.016833381891165712,
      "eval_PRM Accuracy": 0.46226415094339623,
      "eval_PRM F1": 0.5511811023622047,
      "eval_PRM F1 AUC": 0.5151911995809324,
      "eval_PRM F1 AUC (fixed)": 0.3459926663174437,
      "eval_PRM F1 Neg": 0.32941176470588235,
      "eval_PRM NPV": 0.22580645161290322,
      "eval_PRM Precision": 0.7954545454545454,
      "eval_PRM Recall": 0.42168674698795183,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.7576757073402405,
      "eval_runtime": 14.7429,
      "eval_samples_per_second": 2.984,
      "eval_steps_per_second": 0.136,
      "step": 65
    },
    {
      "epoch": 0.017092356997183644,
      "grad_norm": 27.825752091392918,
      "learning_rate": 2.1317829457364341e-07,
      "loss": 0.8875,
      "step": 66
    },
    {
      "epoch": 0.01735133210320158,
      "grad_norm": 14.650186494847837,
      "learning_rate": 2.1640826873385014e-07,
      "loss": 0.8708,
      "step": 67
    },
    {
      "epoch": 0.017610307209219515,
      "grad_norm": 24.186659078983645,
      "learning_rate": 2.1963824289405686e-07,
      "loss": 0.8956,
      "step": 68
    },
    {
      "epoch": 0.017869282315237447,
      "grad_norm": 13.805001585257308,
      "learning_rate": 2.228682170542636e-07,
      "loss": 0.8644,
      "step": 69
    },
    {
      "epoch": 0.018128257421255383,
      "grad_norm": 15.468143300619719,
      "learning_rate": 2.260981912144703e-07,
      "loss": 0.9425,
      "step": 70
    },
    {
      "epoch": 0.018128257421255383,
      "eval_PRM Accuracy": 0.5094339622641509,
      "eval_PRM F1": 0.6060606060606061,
      "eval_PRM F1 AUC": 0.5453116815086433,
      "eval_PRM F1 AUC (fixed)": 0.34363541121005764,
      "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": 0.730787456035614,
      "eval_runtime": 14.2896,
      "eval_samples_per_second": 3.079,
      "eval_steps_per_second": 0.14,
      "step": 70
    },
    {
      "epoch": 0.018387232527273315,
      "grad_norm": 29.660136915528717,
      "learning_rate": 2.2932816537467702e-07,
      "loss": 0.8083,
      "step": 71
    },
    {
      "epoch": 0.01864620763329125,
      "grad_norm": 22.16011134465685,
      "learning_rate": 2.3255813953488374e-07,
      "loss": 0.8421,
      "step": 72
    },
    {
      "epoch": 0.018905182739309183,
      "grad_norm": 14.487608897102827,
      "learning_rate": 2.3578811369509044e-07,
      "loss": 0.8739,
      "step": 73
    },
    {
      "epoch": 0.01916415784532712,
      "grad_norm": 18.961269564998528,
      "learning_rate": 2.3901808785529716e-07,
      "loss": 0.7423,
      "step": 74
    },
    {
      "epoch": 0.01942313295134505,
      "grad_norm": 19.796017096354078,
      "learning_rate": 2.4224806201550393e-07,
      "loss": 0.8343,
      "step": 75
    },
    {
      "epoch": 0.01942313295134505,
      "eval_PRM Accuracy": 0.5754716981132075,
      "eval_PRM F1": 0.6808510638297872,
      "eval_PRM F1 AUC": 0.5717653221581981,
      "eval_PRM F1 AUC (fixed)": 0.3530644316396019,
      "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": 0.6725817322731018,
      "eval_runtime": 14.6895,
      "eval_samples_per_second": 2.995,
      "eval_steps_per_second": 0.136,
      "step": 75
    },
    {
      "epoch": 0.019682108057362986,
      "grad_norm": 18.187683674164898,
      "learning_rate": 2.454780361757106e-07,
      "loss": 0.9895,
      "step": 76
    },
    {
      "epoch": 0.01994108316338092,
      "grad_norm": 13.1508358242096,
      "learning_rate": 2.487080103359173e-07,
      "loss": 0.8959,
      "step": 77
    },
    {
      "epoch": 0.020200058269398854,
      "grad_norm": 12.8284219993238,
      "learning_rate": 2.5193798449612404e-07,
      "loss": 0.9145,
      "step": 78
    },
    {
      "epoch": 0.02045903337541679,
      "grad_norm": 12.116461823137387,
      "learning_rate": 2.5516795865633076e-07,
      "loss": 0.7406,
      "step": 79
    },
    {
      "epoch": 0.02071800848143472,
      "grad_norm": 19.994908230899348,
      "learning_rate": 2.583979328165375e-07,
      "loss": 0.6423,
      "step": 80
    },
    {
      "epoch": 0.02071800848143472,
      "eval_PRM Accuracy": 0.6037735849056604,
      "eval_PRM F1": 0.7123287671232876,
      "eval_PRM F1 AUC": 0.574122577265584,
      "eval_PRM F1 AUC (fixed)": 0.35620743844945,
      "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": 0.6553806066513062,
      "eval_runtime": 14.8024,
      "eval_samples_per_second": 2.972,
      "eval_steps_per_second": 0.135,
      "step": 80
    },
    {
      "epoch": 0.020976983587452657,
      "grad_norm": 21.174111131813078,
      "learning_rate": 2.616279069767442e-07,
      "loss": 0.8049,
      "step": 81
    },
    {
      "epoch": 0.02123595869347059,
      "grad_norm": 24.28580125692702,
      "learning_rate": 2.6485788113695093e-07,
      "loss": 0.6907,
      "step": 82
    },
    {
      "epoch": 0.021494933799488525,
      "grad_norm": 12.136668635564263,
      "learning_rate": 2.6808785529715765e-07,
      "loss": 0.7757,
      "step": 83
    },
    {
      "epoch": 0.021753908905506457,
      "grad_norm": 24.677938731940024,
      "learning_rate": 2.7131782945736437e-07,
      "loss": 0.6658,
      "step": 84
    },
    {
      "epoch": 0.022012884011524392,
      "grad_norm": 14.33727589348448,
      "learning_rate": 2.745478036175711e-07,
      "loss": 0.8024,
      "step": 85
    },
    {
      "epoch": 0.022012884011524392,
      "eval_PRM Accuracy": 0.6226415094339622,
      "eval_PRM F1": 0.726027397260274,
      "eval_PRM F1 AUC": 0.6018858040859089,
      "eval_PRM F1 AUC (fixed)": 0.35908852802514407,
      "eval_PRM F1 Neg": 0.3939393939393939,
      "eval_PRM NPV": 0.3023255813953488,
      "eval_PRM Precision": 0.8412698412698413,
      "eval_PRM Recall": 0.6385542168674698,
      "eval_PRM Specificty": 0.5652173913043478,
      "eval_loss": 0.6360267400741577,
      "eval_runtime": 14.5509,
      "eval_samples_per_second": 3.024,
      "eval_steps_per_second": 0.137,
      "step": 85
    },
    {
      "epoch": 0.022271859117542325,
      "grad_norm": 11.527980390816484,
      "learning_rate": 2.7777777777777776e-07,
      "loss": 0.6023,
      "step": 86
    },
    {
      "epoch": 0.02253083422356026,
      "grad_norm": 16.674272355584908,
      "learning_rate": 2.8100775193798453e-07,
      "loss": 0.6889,
      "step": 87
    },
    {
      "epoch": 0.022789809329578196,
      "grad_norm": 10.66583950521891,
      "learning_rate": 2.8423772609819125e-07,
      "loss": 0.7801,
      "step": 88
    },
    {
      "epoch": 0.023048784435596128,
      "grad_norm": 9.835671755671285,
      "learning_rate": 2.87467700258398e-07,
      "loss": 0.7201,
      "step": 89
    },
    {
      "epoch": 0.023307759541614063,
      "grad_norm": 26.661182986870347,
      "learning_rate": 2.906976744186047e-07,
      "loss": 0.7963,
      "step": 90
    },
    {
      "epoch": 0.023307759541614063,
      "eval_PRM Accuracy": 0.6320754716981132,
      "eval_PRM F1": 0.738255033557047,
      "eval_PRM F1 AUC": 0.5921948664222106,
      "eval_PRM F1 AUC (fixed)": 0.36223153483499215,
      "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": 0.620624303817749,
      "eval_runtime": 15.2162,
      "eval_samples_per_second": 2.892,
      "eval_steps_per_second": 0.131,
      "step": 90
    },
    {
      "epoch": 0.023566734647631996,
      "grad_norm": 8.061268722268196,
      "learning_rate": 2.9392764857881137e-07,
      "loss": 0.608,
      "step": 91
    },
    {
      "epoch": 0.02382570975364993,
      "grad_norm": 13.284851373348685,
      "learning_rate": 2.971576227390181e-07,
      "loss": 0.8424,
      "step": 92
    },
    {
      "epoch": 0.024084684859667863,
      "grad_norm": 10.094119910567679,
      "learning_rate": 3.0038759689922486e-07,
      "loss": 0.7432,
      "step": 93
    },
    {
      "epoch": 0.0243436599656858,
      "grad_norm": 14.804320906675862,
      "learning_rate": 3.036175710594316e-07,
      "loss": 0.7599,
      "step": 94
    },
    {
      "epoch": 0.02460263507170373,
      "grad_norm": 10.414202643458607,
      "learning_rate": 3.0684754521963825e-07,
      "loss": 0.5895,
      "step": 95
    },
    {
      "epoch": 0.02460263507170373,
      "eval_PRM Accuracy": 0.6415094339622641,
      "eval_PRM F1": 0.7466666666666667,
      "eval_PRM F1 AUC": 0.5982189628077528,
      "eval_PRM F1 AUC (fixed)": 0.36170770036668415,
      "eval_PRM F1 Neg": 0.3870967741935484,
      "eval_PRM NPV": 0.3076923076923077,
      "eval_PRM Precision": 0.835820895522388,
      "eval_PRM Recall": 0.6746987951807228,
      "eval_PRM Specificty": 0.5217391304347826,
      "eval_loss": 0.6061849594116211,
      "eval_runtime": 15.0049,
      "eval_samples_per_second": 2.932,
      "eval_steps_per_second": 0.133,
      "step": 95
    },
    {
      "epoch": 0.024861610177721667,
      "grad_norm": 10.672625979890972,
      "learning_rate": 3.1007751937984497e-07,
      "loss": 0.7713,
      "step": 96
    },
    {
      "epoch": 0.025120585283739602,
      "grad_norm": 12.251684326095432,
      "learning_rate": 3.1330749354005175e-07,
      "loss": 0.781,
      "step": 97
    },
    {
      "epoch": 0.025379560389757534,
      "grad_norm": 13.91913851389255,
      "learning_rate": 3.165374677002584e-07,
      "loss": 0.7431,
      "step": 98
    },
    {
      "epoch": 0.02563853549577547,
      "grad_norm": 9.507497299775913,
      "learning_rate": 3.1976744186046514e-07,
      "loss": 0.6107,
      "step": 99
    },
    {
      "epoch": 0.025897510601793402,
      "grad_norm": 34.88561213639425,
      "learning_rate": 3.2299741602067186e-07,
      "loss": 0.8904,
      "step": 100
    },
    {
      "epoch": 0.025897510601793402,
      "eval_PRM Accuracy": 0.6415094339622641,
      "eval_PRM F1": 0.7466666666666667,
      "eval_PRM F1 AUC": 0.5982189628077528,
      "eval_PRM F1 AUC (fixed)": 0.36406495547407025,
      "eval_PRM F1 Neg": 0.3870967741935484,
      "eval_PRM NPV": 0.3076923076923077,
      "eval_PRM Precision": 0.835820895522388,
      "eval_PRM Recall": 0.6746987951807228,
      "eval_PRM Specificty": 0.5217391304347826,
      "eval_loss": 0.5902833938598633,
      "eval_runtime": 14.8013,
      "eval_samples_per_second": 2.973,
      "eval_steps_per_second": 0.135,
      "step": 100
    },
    {
      "epoch": 0.026156485707811337,
      "grad_norm": 12.625263414672842,
      "learning_rate": 3.262273901808786e-07,
      "loss": 0.7125,
      "step": 101
    },
    {
      "epoch": 0.02641546081382927,
      "grad_norm": 21.834409355265553,
      "learning_rate": 3.2945736434108535e-07,
      "loss": 0.8078,
      "step": 102
    },
    {
      "epoch": 0.026674435919847205,
      "grad_norm": 16.68863184867011,
      "learning_rate": 3.32687338501292e-07,
      "loss": 0.6124,
      "step": 103
    },
    {
      "epoch": 0.026933411025865137,
      "grad_norm": 6.89532276240644,
      "learning_rate": 3.3591731266149874e-07,
      "loss": 0.5681,
      "step": 104
    },
    {
      "epoch": 0.027192386131883073,
      "grad_norm": 9.05599285768429,
      "learning_rate": 3.391472868217054e-07,
      "loss": 0.5845,
      "step": 105
    },
    {
      "epoch": 0.027192386131883073,
      "eval_PRM Accuracy": 0.660377358490566,
      "eval_PRM F1": 0.7631578947368421,
      "eval_PRM F1 AUC": 0.610267155578837,
      "eval_PRM F1 AUC (fixed)": 0.3672079622839183,
      "eval_PRM F1 Neg": 0.4,
      "eval_PRM NPV": 0.32432432432432434,
      "eval_PRM Precision": 0.8405797101449275,
      "eval_PRM Recall": 0.6987951807228916,
      "eval_PRM Specificty": 0.5217391304347826,
      "eval_loss": 0.5670130252838135,
      "eval_runtime": 14.5479,
      "eval_samples_per_second": 3.024,
      "eval_steps_per_second": 0.137,
      "step": 105
    },
    {
      "epoch": 0.02745136123790101,
      "grad_norm": 29.200812937204702,
      "learning_rate": 3.423772609819122e-07,
      "loss": 0.8313,
      "step": 106
    },
    {
      "epoch": 0.02771033634391894,
      "grad_norm": 7.770870043903371,
      "learning_rate": 3.4560723514211885e-07,
      "loss": 0.573,
      "step": 107
    },
    {
      "epoch": 0.027969311449936876,
      "grad_norm": 21.63547347590896,
      "learning_rate": 3.488372093023256e-07,
      "loss": 0.7248,
      "step": 108
    },
    {
      "epoch": 0.028228286555954808,
      "grad_norm": 13.313977325056683,
      "learning_rate": 3.5206718346253235e-07,
      "loss": 0.6245,
      "step": 109
    },
    {
      "epoch": 0.028487261661972744,
      "grad_norm": 10.424747237877718,
      "learning_rate": 3.55297157622739e-07,
      "loss": 0.6744,
      "step": 110
    },
    {
      "epoch": 0.028487261661972744,
      "eval_PRM Accuracy": 0.6698113207547169,
      "eval_PRM F1": 0.7712418300653595,
      "eval_PRM F1 AUC": 0.6162912519643793,
      "eval_PRM F1 AUC (fixed)": 0.3716605552645364,
      "eval_PRM F1 Neg": 0.4067796610169492,
      "eval_PRM NPV": 0.3333333333333333,
      "eval_PRM Precision": 0.8428571428571429,
      "eval_PRM Recall": 0.7108433734939759,
      "eval_PRM Specificty": 0.5217391304347826,
      "eval_loss": 0.55446857213974,
      "eval_runtime": 15.0221,
      "eval_samples_per_second": 2.929,
      "eval_steps_per_second": 0.133,
      "step": 110
    },
    {
      "epoch": 0.028746236767990676,
      "grad_norm": 19.990118643252117,
      "learning_rate": 3.585271317829458e-07,
      "loss": 0.7346,
      "step": 111
    },
    {
      "epoch": 0.02900521187400861,
      "grad_norm": 11.664641699650158,
      "learning_rate": 3.6175710594315246e-07,
      "loss": 0.6889,
      "step": 112
    },
    {
      "epoch": 0.029264186980026544,
      "grad_norm": 26.966194495147256,
      "learning_rate": 3.649870801033592e-07,
      "loss": 0.678,
      "step": 113
    },
    {
      "epoch": 0.02952316208604448,
      "grad_norm": 17.964853935178834,
      "learning_rate": 3.6821705426356595e-07,
      "loss": 0.7198,
      "step": 114
    },
    {
      "epoch": 0.02978213719206241,
      "grad_norm": 14.85706395717968,
      "learning_rate": 3.714470284237726e-07,
      "loss": 0.6366,
      "step": 115
    },
    {
      "epoch": 0.02978213719206241,
      "eval_PRM Accuracy": 0.6792452830188679,
      "eval_PRM F1": 0.7733333333333333,
      "eval_PRM F1 AUC": 0.6537454164484023,
      "eval_PRM F1 AUC (fixed)": 0.3740178103719225,
      "eval_PRM F1 Neg": 0.45161290322580644,
      "eval_PRM NPV": 0.358974358974359,
      "eval_PRM Precision": 0.8656716417910447,
      "eval_PRM Recall": 0.6987951807228916,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.5416440963745117,
      "eval_runtime": 14.3887,
      "eval_samples_per_second": 3.058,
      "eval_steps_per_second": 0.139,
      "step": 115
    },
    {
      "epoch": 0.030041112298080347,
      "grad_norm": 16.61418454996819,
      "learning_rate": 3.746770025839794e-07,
      "loss": 0.725,
      "step": 116
    },
    {
      "epoch": 0.030300087404098282,
      "grad_norm": 9.534591885083701,
      "learning_rate": 3.7790697674418606e-07,
      "loss": 0.6432,
      "step": 117
    },
    {
      "epoch": 0.030559062510116215,
      "grad_norm": 7.169368696427665,
      "learning_rate": 3.811369509043928e-07,
      "loss": 0.6227,
      "step": 118
    },
    {
      "epoch": 0.03081803761613415,
      "grad_norm": 10.08722788125257,
      "learning_rate": 3.843669250645995e-07,
      "loss": 0.7123,
      "step": 119
    },
    {
      "epoch": 0.031077012722152082,
      "grad_norm": 8.13895725897613,
      "learning_rate": 3.8759689922480623e-07,
      "loss": 0.6191,
      "step": 120
    },
    {
      "epoch": 0.031077012722152082,
      "eval_PRM Accuracy": 0.6698113207547169,
      "eval_PRM F1": 0.7586206896551724,
      "eval_PRM F1 AUC": 0.679151388161341,
      "eval_PRM F1 AUC (fixed)": 0.3795180722891566,
      "eval_PRM F1 Neg": 0.47761194029850745,
      "eval_PRM NPV": 0.36363636363636365,
      "eval_PRM Precision": 0.8870967741935484,
      "eval_PRM Recall": 0.6626506024096386,
      "eval_PRM Specificty": 0.6956521739130435,
      "eval_loss": 0.5411401391029358,
      "eval_runtime": 14.5257,
      "eval_samples_per_second": 3.029,
      "eval_steps_per_second": 0.138,
      "step": 120
    },
    {
      "epoch": 0.031335987828170014,
      "grad_norm": 6.215018883416692,
      "learning_rate": 3.90826873385013e-07,
      "loss": 0.5748,
      "step": 121
    },
    {
      "epoch": 0.03159496293418795,
      "grad_norm": 8.35797483869244,
      "learning_rate": 3.9405684754521967e-07,
      "loss": 0.5726,
      "step": 122
    },
    {
      "epoch": 0.031853938040205886,
      "grad_norm": 8.05244834758129,
      "learning_rate": 3.972868217054264e-07,
      "loss": 0.6634,
      "step": 123
    },
    {
      "epoch": 0.03211291314622382,
      "grad_norm": 6.625684614301109,
      "learning_rate": 4.0051679586563306e-07,
      "loss": 0.6036,
      "step": 124
    },
    {
      "epoch": 0.03237188825224176,
      "grad_norm": 9.791133812266091,
      "learning_rate": 4.0374677002583983e-07,
      "loss": 0.5487,
      "step": 125
    },
    {
      "epoch": 0.03237188825224176,
      "eval_PRM Accuracy": 0.6415094339622641,
      "eval_PRM F1": 0.7285714285714285,
      "eval_PRM F1 AUC": 0.676794133053955,
      "eval_PRM F1 AUC (fixed)": 0.38344683080146674,
      "eval_PRM F1 Neg": 0.4722222222222222,
      "eval_PRM NPV": 0.3469387755102041,
      "eval_PRM Precision": 0.8947368421052632,
      "eval_PRM Recall": 0.6144578313253012,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.5412638187408447,
      "eval_runtime": 14.6931,
      "eval_samples_per_second": 2.995,
      "eval_steps_per_second": 0.136,
      "step": 125
    },
    {
      "epoch": 0.03263086335825969,
      "grad_norm": 6.757491658349913,
      "learning_rate": 4.0697674418604655e-07,
      "loss": 0.5949,
      "step": 126
    },
    {
      "epoch": 0.03288983846427762,
      "grad_norm": 26.94630756049828,
      "learning_rate": 4.102067183462533e-07,
      "loss": 0.7258,
      "step": 127
    },
    {
      "epoch": 0.03314881357029555,
      "grad_norm": 14.68796120813943,
      "learning_rate": 4.1343669250646e-07,
      "loss": 0.6737,
      "step": 128
    },
    {
      "epoch": 0.03340778867631349,
      "grad_norm": 13.84775550814014,
      "learning_rate": 4.1666666666666667e-07,
      "loss": 0.5792,
      "step": 129
    },
    {
      "epoch": 0.033666763782331424,
      "grad_norm": 15.319435558173073,
      "learning_rate": 4.1989664082687344e-07,
      "loss": 0.7407,
      "step": 130
    },
    {
      "epoch": 0.033666763782331424,
      "eval_PRM Accuracy": 0.660377358490566,
      "eval_PRM F1": 0.7428571428571429,
      "eval_PRM F1 AUC": 0.7045573598742797,
      "eval_PRM F1 AUC (fixed)": 0.3805657412257727,
      "eval_PRM F1 Neg": 0.5,
      "eval_PRM NPV": 0.3673469387755102,
      "eval_PRM Precision": 0.9122807017543859,
      "eval_PRM Recall": 0.6265060240963856,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.5402398705482483,
      "eval_runtime": 14.4767,
      "eval_samples_per_second": 3.039,
      "eval_steps_per_second": 0.138,
      "step": 130
    },
    {
      "epoch": 0.033925738888349356,
      "grad_norm": 11.219710300102047,
      "learning_rate": 4.2312661498708016e-07,
      "loss": 0.6967,
      "step": 131
    },
    {
      "epoch": 0.03418471399436729,
      "grad_norm": 9.028415286023842,
      "learning_rate": 4.2635658914728683e-07,
      "loss": 0.5366,
      "step": 132
    },
    {
      "epoch": 0.03444368910038523,
      "grad_norm": 12.524874702176467,
      "learning_rate": 4.295865633074936e-07,
      "loss": 0.6446,
      "step": 133
    },
    {
      "epoch": 0.03470266420640316,
      "grad_norm": 13.610218061273281,
      "learning_rate": 4.3281653746770027e-07,
      "loss": 0.6463,
      "step": 134
    },
    {
      "epoch": 0.03496163931242109,
      "grad_norm": 8.571461600608345,
      "learning_rate": 4.3604651162790705e-07,
      "loss": 0.6441,
      "step": 135
    },
    {
      "epoch": 0.03496163931242109,
      "eval_PRM Accuracy": 0.6698113207547169,
      "eval_PRM F1": 0.75177304964539,
      "eval_PRM F1 AUC": 0.7105814562598219,
      "eval_PRM F1 AUC (fixed)": 0.38580408590885285,
      "eval_PRM F1 Neg": 0.5070422535211268,
      "eval_PRM NPV": 0.375,
      "eval_PRM Precision": 0.9137931034482759,
      "eval_PRM Recall": 0.6385542168674698,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.5414556860923767,
      "eval_runtime": 14.4384,
      "eval_samples_per_second": 3.047,
      "eval_steps_per_second": 0.139,
      "step": 135
    },
    {
      "epoch": 0.03522061441843903,
      "grad_norm": 7.16674151720117,
      "learning_rate": 4.392764857881137e-07,
      "loss": 0.6491,
      "step": 136
    },
    {
      "epoch": 0.03547958952445696,
      "grad_norm": 7.1083764687848605,
      "learning_rate": 4.4250645994832044e-07,
      "loss": 0.5751,
      "step": 137
    },
    {
      "epoch": 0.035738564630474895,
      "grad_norm": 9.698448189769538,
      "learning_rate": 4.457364341085272e-07,
      "loss": 0.5885,
      "step": 138
    },
    {
      "epoch": 0.03599753973649283,
      "grad_norm": 6.798621458927422,
      "learning_rate": 4.489664082687339e-07,
      "loss": 0.6095,
      "step": 139
    },
    {
      "epoch": 0.036256514842510766,
      "grad_norm": 7.661301814457179,
      "learning_rate": 4.521963824289406e-07,
      "loss": 0.6317,
      "step": 140
    },
    {
      "epoch": 0.036256514842510766,
      "eval_PRM Accuracy": 0.660377358490566,
      "eval_PRM F1": 0.7428571428571429,
      "eval_PRM F1 AUC": 0.7045573598742797,
      "eval_PRM F1 AUC (fixed)": 0.38789942378208486,
      "eval_PRM F1 Neg": 0.5,
      "eval_PRM NPV": 0.3673469387755102,
      "eval_PRM Precision": 0.9122807017543859,
      "eval_PRM Recall": 0.6265060240963856,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.5477471351623535,
      "eval_runtime": 14.1224,
      "eval_samples_per_second": 3.116,
      "eval_steps_per_second": 0.142,
      "step": 140
    },
    {
      "epoch": 0.0365154899485287,
      "grad_norm": 13.365479316439192,
      "learning_rate": 4.554263565891473e-07,
      "loss": 0.6813,
      "step": 141
    },
    {
      "epoch": 0.03677446505454663,
      "grad_norm": 10.854235911252928,
      "learning_rate": 4.5865633074935404e-07,
      "loss": 0.6343,
      "step": 142
    },
    {
      "epoch": 0.03703344016056456,
      "grad_norm": 6.478102222374132,
      "learning_rate": 4.618863049095608e-07,
      "loss": 0.5942,
      "step": 143
    },
    {
      "epoch": 0.0372924152665825,
      "grad_norm": 7.359112000214807,
      "learning_rate": 4.651162790697675e-07,
      "loss": 0.6848,
      "step": 144
    },
    {
      "epoch": 0.037551390372600434,
      "grad_norm": 13.133076829254982,
      "learning_rate": 4.683462532299742e-07,
      "loss": 0.6556,
      "step": 145
    },
    {
      "epoch": 0.037551390372600434,
      "eval_PRM Accuracy": 0.6698113207547169,
      "eval_PRM F1": 0.75177304964539,
      "eval_PRM F1 AUC": 0.7105814562598219,
      "eval_PRM F1 AUC (fixed)": 0.3899947616553169,
      "eval_PRM F1 Neg": 0.5070422535211268,
      "eval_PRM NPV": 0.375,
      "eval_PRM Precision": 0.9137931034482759,
      "eval_PRM Recall": 0.6385542168674698,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.5437543392181396,
      "eval_runtime": 14.4066,
      "eval_samples_per_second": 3.054,
      "eval_steps_per_second": 0.139,
      "step": 145
    },
    {
      "epoch": 0.037810365478618366,
      "grad_norm": 7.676893672467976,
      "learning_rate": 4.7157622739018087e-07,
      "loss": 0.6555,
      "step": 146
    },
    {
      "epoch": 0.038069340584636305,
      "grad_norm": 8.025843991999064,
      "learning_rate": 4.7480620155038765e-07,
      "loss": 0.5676,
      "step": 147
    },
    {
      "epoch": 0.03832831569065424,
      "grad_norm": 9.894553920917076,
      "learning_rate": 4.780361757105943e-07,
      "loss": 0.563,
      "step": 148
    },
    {
      "epoch": 0.03858729079667217,
      "grad_norm": 6.883678153936593,
      "learning_rate": 4.812661498708011e-07,
      "loss": 0.5804,
      "step": 149
    },
    {
      "epoch": 0.0388462659026901,
      "grad_norm": 8.261843621096295,
      "learning_rate": 4.844961240310079e-07,
      "loss": 0.6092,
      "step": 150
    },
    {
      "epoch": 0.0388462659026901,
      "eval_PRM Accuracy": 0.6792452830188679,
      "eval_PRM F1": 0.7605633802816901,
      "eval_PRM F1 AUC": 0.7166055526453641,
      "eval_PRM F1 AUC (fixed)": 0.38999476165531693,
      "eval_PRM F1 Neg": 0.5142857142857142,
      "eval_PRM NPV": 0.3829787234042553,
      "eval_PRM Precision": 0.9152542372881356,
      "eval_PRM Recall": 0.6506024096385542,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.540428102016449,
      "eval_runtime": 14.7992,
      "eval_samples_per_second": 2.973,
      "eval_steps_per_second": 0.135,
      "step": 150
    },
    {
      "epoch": 0.03910524100870804,
      "grad_norm": 7.354522673332907,
      "learning_rate": 4.877260981912145e-07,
      "loss": 0.5619,
      "step": 151
    },
    {
      "epoch": 0.03936421611472597,
      "grad_norm": 10.294306829596843,
      "learning_rate": 4.909560723514212e-07,
      "loss": 0.6696,
      "step": 152
    },
    {
      "epoch": 0.039623191220743904,
      "grad_norm": 8.641933608314154,
      "learning_rate": 4.94186046511628e-07,
      "loss": 0.5843,
      "step": 153
    },
    {
      "epoch": 0.03988216632676184,
      "grad_norm": 9.039792206894193,
      "learning_rate": 4.974160206718346e-07,
      "loss": 0.6636,
      "step": 154
    },
    {
      "epoch": 0.040141141432779776,
      "grad_norm": 7.8634665219535975,
      "learning_rate": 5.006459948320414e-07,
      "loss": 0.5328,
      "step": 155
    },
    {
      "epoch": 0.040141141432779776,
      "eval_PRM Accuracy": 0.6981132075471698,
      "eval_PRM F1": 0.7746478873239436,
      "eval_PRM F1 AUC": 0.7443687794656888,
      "eval_PRM F1 AUC (fixed)": 0.3988999476165532,
      "eval_PRM F1 Neg": 0.5428571428571428,
      "eval_PRM NPV": 0.40425531914893614,
      "eval_PRM Precision": 0.9322033898305084,
      "eval_PRM Recall": 0.6626506024096386,
      "eval_PRM Specificty": 0.8260869565217391,
      "eval_loss": 0.534451425075531,
      "eval_runtime": 14.5832,
      "eval_samples_per_second": 3.017,
      "eval_steps_per_second": 0.137,
      "step": 155
    },
    {
      "epoch": 0.04040011653879771,
      "grad_norm": 9.34303127109423,
      "learning_rate": 5.038759689922481e-07,
      "loss": 0.6202,
      "step": 156
    },
    {
      "epoch": 0.04065909164481564,
      "grad_norm": 6.195118698281523,
      "learning_rate": 5.071059431524549e-07,
      "loss": 0.572,
      "step": 157
    },
    {
      "epoch": 0.04091806675083358,
      "grad_norm": 8.108721137426325,
      "learning_rate": 5.103359173126615e-07,
      "loss": 0.7191,
      "step": 158
    },
    {
      "epoch": 0.04117704185685151,
      "grad_norm": 12.409323930943694,
      "learning_rate": 5.135658914728683e-07,
      "loss": 0.4859,
      "step": 159
    },
    {
      "epoch": 0.04143601696286944,
      "grad_norm": 13.610909760223025,
      "learning_rate": 5.16795865633075e-07,
      "loss": 0.5121,
      "step": 160
    },
    {
      "epoch": 0.04143601696286944,
      "eval_PRM Accuracy": 0.7075471698113207,
      "eval_PRM F1": 0.7891156462585034,
      "eval_PRM F1 AUC": 0.7189628077527501,
      "eval_PRM F1 AUC (fixed)": 0.39392352016762705,
      "eval_PRM F1 Neg": 0.5230769230769231,
      "eval_PRM NPV": 0.40476190476190477,
      "eval_PRM Precision": 0.90625,
      "eval_PRM Recall": 0.6987951807228916,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.5245890617370605,
      "eval_runtime": 14.5943,
      "eval_samples_per_second": 3.015,
      "eval_steps_per_second": 0.137,
      "step": 160
    },
    {
      "epoch": 0.041694992068887375,
      "grad_norm": 6.347252875639677,
      "learning_rate": 5.200258397932817e-07,
      "loss": 0.5312,
      "step": 161
    },
    {
      "epoch": 0.041953967174905314,
      "grad_norm": 7.171273813798448,
      "learning_rate": 5.232558139534884e-07,
      "loss": 0.5634,
      "step": 162
    },
    {
      "epoch": 0.042212942280923246,
      "grad_norm": 14.883956010412117,
      "learning_rate": 5.264857881136951e-07,
      "loss": 0.5834,
      "step": 163
    },
    {
      "epoch": 0.04247191738694118,
      "grad_norm": 5.869889554466555,
      "learning_rate": 5.297157622739019e-07,
      "loss": 0.5194,
      "step": 164
    },
    {
      "epoch": 0.04273089249295912,
      "grad_norm": 11.870697609022255,
      "learning_rate": 5.329457364341085e-07,
      "loss": 0.5417,
      "step": 165
    },
    {
      "epoch": 0.04273089249295912,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8053691275167785,
      "eval_PRM F1 AUC": 0.7310110005238345,
      "eval_PRM F1 AUC (fixed)": 0.3994237820848612,
      "eval_PRM F1 Neg": 0.5396825396825397,
      "eval_PRM NPV": 0.425,
      "eval_PRM Precision": 0.9090909090909091,
      "eval_PRM Recall": 0.7228915662650602,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.514342188835144,
      "eval_runtime": 14.3157,
      "eval_samples_per_second": 3.074,
      "eval_steps_per_second": 0.14,
      "step": 165
    },
    {
      "epoch": 0.04298986759897705,
      "grad_norm": 5.691938211646347,
      "learning_rate": 5.361757105943153e-07,
      "loss": 0.6143,
      "step": 166
    },
    {
      "epoch": 0.04324884270499498,
      "grad_norm": 9.493665131614394,
      "learning_rate": 5.394056847545221e-07,
      "loss": 0.5261,
      "step": 167
    },
    {
      "epoch": 0.043507817811012914,
      "grad_norm": 6.1221521139428,
      "learning_rate": 5.426356589147287e-07,
      "loss": 0.5844,
      "step": 168
    },
    {
      "epoch": 0.04376679291703085,
      "grad_norm": 5.228074975287801,
      "learning_rate": 5.458656330749355e-07,
      "loss": 0.571,
      "step": 169
    },
    {
      "epoch": 0.044025768023048785,
      "grad_norm": 14.423948863883327,
      "learning_rate": 5.490956072351422e-07,
      "loss": 0.6379,
      "step": 170
    },
    {
      "epoch": 0.044025768023048785,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8104575163398693,
      "eval_PRM F1 AUC": 0.6995809324253536,
      "eval_PRM F1 AUC (fixed)": 0.39837611314824517,
      "eval_PRM F1 Neg": 0.5084745762711864,
      "eval_PRM NPV": 0.4166666666666667,
      "eval_PRM Precision": 0.8857142857142857,
      "eval_PRM Recall": 0.7469879518072289,
      "eval_PRM Specificty": 0.6521739130434783,
      "eval_loss": 0.5093567967414856,
      "eval_runtime": 14.4938,
      "eval_samples_per_second": 3.036,
      "eval_steps_per_second": 0.138,
      "step": 170
    },
    {
      "epoch": 0.04428474312906672,
      "grad_norm": 10.344568168628063,
      "learning_rate": 5.523255813953489e-07,
      "loss": 0.5654,
      "step": 171
    },
    {
      "epoch": 0.04454371823508465,
      "grad_norm": 8.542836484798716,
      "learning_rate": 5.555555555555555e-07,
      "loss": 0.5971,
      "step": 172
    },
    {
      "epoch": 0.04480269334110259,
      "grad_norm": 8.19809400190519,
      "learning_rate": 5.587855297157623e-07,
      "loss": 0.5235,
      "step": 173
    },
    {
      "epoch": 0.04506166844712052,
      "grad_norm": 6.077323198910218,
      "learning_rate": 5.620155038759691e-07,
      "loss": 0.5564,
      "step": 174
    },
    {
      "epoch": 0.04532064355313845,
      "grad_norm": 5.288812927319927,
      "learning_rate": 5.652454780361757e-07,
      "loss": 0.5437,
      "step": 175
    },
    {
      "epoch": 0.04532064355313845,
      "eval_PRM Accuracy": 0.7075471698113207,
      "eval_PRM F1": 0.7947019867549668,
      "eval_PRM F1 AUC": 0.6875327396542692,
      "eval_PRM F1 AUC (fixed)": 0.4044002095337874,
      "eval_PRM F1 Neg": 0.4918032786885246,
      "eval_PRM NPV": 0.39473684210526316,
      "eval_PRM Precision": 0.8823529411764706,
      "eval_PRM Recall": 0.7228915662650602,
      "eval_PRM Specificty": 0.6521739130434783,
      "eval_loss": 0.5160964727401733,
      "eval_runtime": 14.5801,
      "eval_samples_per_second": 3.018,
      "eval_steps_per_second": 0.137,
      "step": 175
    },
    {
      "epoch": 0.04557961865915639,
      "grad_norm": 5.935722932915326,
      "learning_rate": 5.684754521963825e-07,
      "loss": 0.4989,
      "step": 176
    },
    {
      "epoch": 0.045838593765174324,
      "grad_norm": 10.165376980754218,
      "learning_rate": 5.717054263565892e-07,
      "loss": 0.4974,
      "step": 177
    },
    {
      "epoch": 0.046097568871192256,
      "grad_norm": 8.844537517788227,
      "learning_rate": 5.74935400516796e-07,
      "loss": 0.5733,
      "step": 178
    },
    {
      "epoch": 0.04635654397721019,
      "grad_norm": 6.136922569828234,
      "learning_rate": 5.781653746770026e-07,
      "loss": 0.5724,
      "step": 179
    },
    {
      "epoch": 0.04661551908322813,
      "grad_norm": 6.093958466635018,
      "learning_rate": 5.813953488372094e-07,
      "loss": 0.543,
      "step": 180
    },
    {
      "epoch": 0.04661551908322813,
      "eval_PRM Accuracy": 0.7075471698113207,
      "eval_PRM F1": 0.7973856209150327,
      "eval_PRM F1 AUC": 0.671817705605029,
      "eval_PRM F1 AUC (fixed)": 0.40518596123624934,
      "eval_PRM F1 Neg": 0.4745762711864407,
      "eval_PRM NPV": 0.3888888888888889,
      "eval_PRM Precision": 0.8714285714285714,
      "eval_PRM Recall": 0.7349397590361446,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.5174723863601685,
      "eval_runtime": 14.5846,
      "eval_samples_per_second": 3.017,
      "eval_steps_per_second": 0.137,
      "step": 180
    },
    {
      "epoch": 0.04687449418924606,
      "grad_norm": 10.776808527067512,
      "learning_rate": 5.846253229974161e-07,
      "loss": 0.5772,
      "step": 181
    },
    {
      "epoch": 0.04713346929526399,
      "grad_norm": 10.033903169098872,
      "learning_rate": 5.878552971576227e-07,
      "loss": 0.4632,
      "step": 182
    },
    {
      "epoch": 0.04739244440128193,
      "grad_norm": 6.481504308087625,
      "learning_rate": 5.910852713178295e-07,
      "loss": 0.516,
      "step": 183
    },
    {
      "epoch": 0.04765141950729986,
      "grad_norm": 6.015599645256489,
      "learning_rate": 5.943152454780362e-07,
      "loss": 0.5046,
      "step": 184
    },
    {
      "epoch": 0.047910394613317794,
      "grad_norm": 13.76113181680478,
      "learning_rate": 5.97545219638243e-07,
      "loss": 0.5823,
      "step": 185
    },
    {
      "epoch": 0.047910394613317794,
      "eval_PRM Accuracy": 0.7075471698113207,
      "eval_PRM F1": 0.7973856209150327,
      "eval_PRM F1 AUC": 0.671817705605029,
      "eval_PRM F1 AUC (fixed)": 0.4062336301728654,
      "eval_PRM F1 Neg": 0.4745762711864407,
      "eval_PRM NPV": 0.3888888888888889,
      "eval_PRM Precision": 0.8714285714285714,
      "eval_PRM Recall": 0.7349397590361446,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.5205953121185303,
      "eval_runtime": 14.829,
      "eval_samples_per_second": 2.967,
      "eval_steps_per_second": 0.135,
      "step": 185
    },
    {
      "epoch": 0.048169369719335726,
      "grad_norm": 6.725385680199521,
      "learning_rate": 6.007751937984497e-07,
      "loss": 0.4939,
      "step": 186
    },
    {
      "epoch": 0.048428344825353666,
      "grad_norm": 10.561746795952903,
      "learning_rate": 6.040051679586564e-07,
      "loss": 0.5041,
      "step": 187
    },
    {
      "epoch": 0.0486873199313716,
      "grad_norm": 9.334572719128575,
      "learning_rate": 6.072351421188632e-07,
      "loss": 0.5234,
      "step": 188
    },
    {
      "epoch": 0.04894629503738953,
      "grad_norm": 5.443979412272866,
      "learning_rate": 6.104651162790698e-07,
      "loss": 0.4918,
      "step": 189
    },
    {
      "epoch": 0.04920527014340746,
      "grad_norm": 5.453809365018693,
      "learning_rate": 6.136950904392765e-07,
      "loss": 0.5384,
      "step": 190
    },
    {
      "epoch": 0.04920527014340746,
      "eval_PRM Accuracy": 0.7358490566037735,
      "eval_PRM F1": 0.8205128205128205,
      "eval_PRM F1 AUC": 0.6898899947616554,
      "eval_PRM F1 AUC (fixed)": 0.40859088528025145,
      "eval_PRM F1 Neg": 0.5,
      "eval_PRM NPV": 0.42424242424242425,
      "eval_PRM Precision": 0.8767123287671232,
      "eval_PRM Recall": 0.7710843373493976,
      "eval_PRM Specificty": 0.6086956521739131,
      "eval_loss": 0.5035573840141296,
      "eval_runtime": 14.772,
      "eval_samples_per_second": 2.979,
      "eval_steps_per_second": 0.135,
      "step": 190
    },
    {
      "epoch": 0.0494642452494254,
      "grad_norm": 16.244963846443802,
      "learning_rate": 6.169250645994833e-07,
      "loss": 0.6508,
      "step": 191
    },
    {
      "epoch": 0.04972322035544333,
      "grad_norm": 8.932185819469234,
      "learning_rate": 6.201550387596899e-07,
      "loss": 0.6221,
      "step": 192
    },
    {
      "epoch": 0.049982195461461265,
      "grad_norm": 4.408328610073295,
      "learning_rate": 6.233850129198967e-07,
      "loss": 0.4826,
      "step": 193
    },
    {
      "epoch": 0.050241170567479204,
      "grad_norm": 13.406641686428832,
      "learning_rate": 6.266149870801035e-07,
      "loss": 0.6162,
      "step": 194
    },
    {
      "epoch": 0.050500145673497136,
      "grad_norm": 4.892120858093281,
      "learning_rate": 6.298449612403102e-07,
      "loss": 0.4512,
      "step": 195
    },
    {
      "epoch": 0.050500145673497136,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8235294117647058,
      "eval_PRM F1 AUC": 0.7273441592456783,
      "eval_PRM F1 AUC (fixed)": 0.41382922996333155,
      "eval_PRM F1 Neg": 0.5423728813559322,
      "eval_PRM NPV": 0.4444444444444444,
      "eval_PRM Precision": 0.9,
      "eval_PRM Recall": 0.7590361445783133,
      "eval_PRM Specificty": 0.6956521739130435,
      "eval_loss": 0.5021147131919861,
      "eval_runtime": 14.9083,
      "eval_samples_per_second": 2.951,
      "eval_steps_per_second": 0.134,
      "step": 195
    },
    {
      "epoch": 0.05075912077951507,
      "grad_norm": 5.6387526600892315,
      "learning_rate": 6.330749354005168e-07,
      "loss": 0.4776,
      "step": 196
    },
    {
      "epoch": 0.051018095885533,
      "grad_norm": 7.286720230552998,
      "learning_rate": 6.363049095607236e-07,
      "loss": 0.5219,
      "step": 197
    },
    {
      "epoch": 0.05127707099155094,
      "grad_norm": 4.997205287307194,
      "learning_rate": 6.395348837209303e-07,
      "loss": 0.5234,
      "step": 198
    },
    {
      "epoch": 0.05153604609756887,
      "grad_norm": 6.166767867082422,
      "learning_rate": 6.427648578811369e-07,
      "loss": 0.4966,
      "step": 199
    },
    {
      "epoch": 0.051795021203586804,
      "grad_norm": 5.2138937231508695,
      "learning_rate": 6.459948320413437e-07,
      "loss": 0.5545,
      "step": 200
    },
    {
      "epoch": 0.051795021203586804,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8235294117647058,
      "eval_PRM F1 AUC": 0.7273441592456783,
      "eval_PRM F1 AUC (fixed)": 0.4232582503928759,
      "eval_PRM F1 Neg": 0.5423728813559322,
      "eval_PRM NPV": 0.4444444444444444,
      "eval_PRM Precision": 0.9,
      "eval_PRM Recall": 0.7590361445783133,
      "eval_PRM Specificty": 0.6956521739130435,
      "eval_loss": 0.49177101254463196,
      "eval_runtime": 15.1242,
      "eval_samples_per_second": 2.909,
      "eval_steps_per_second": 0.132,
      "step": 200
    },
    {
      "epoch": 0.052053996309604736,
      "grad_norm": 13.586487332304372,
      "learning_rate": 6.492248062015505e-07,
      "loss": 0.5246,
      "step": 201
    },
    {
      "epoch": 0.052312971415622675,
      "grad_norm": 6.063157983759811,
      "learning_rate": 6.524547803617572e-07,
      "loss": 0.4992,
      "step": 202
    },
    {
      "epoch": 0.05257194652164061,
      "grad_norm": 6.388364255784616,
      "learning_rate": 6.556847545219638e-07,
      "loss": 0.4724,
      "step": 203
    },
    {
      "epoch": 0.05283092162765854,
      "grad_norm": 11.43907528969819,
      "learning_rate": 6.589147286821707e-07,
      "loss": 0.5736,
      "step": 204
    },
    {
      "epoch": 0.05308989673367648,
      "grad_norm": 16.818113355638936,
      "learning_rate": 6.621447028423774e-07,
      "loss": 0.6059,
      "step": 205
    },
    {
      "epoch": 0.05308989673367648,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8235294117647058,
      "eval_PRM F1 AUC": 0.7273441592456783,
      "eval_PRM F1 AUC (fixed)": 0.42744892613933994,
      "eval_PRM F1 Neg": 0.5423728813559322,
      "eval_PRM NPV": 0.4444444444444444,
      "eval_PRM Precision": 0.9,
      "eval_PRM Recall": 0.7590361445783133,
      "eval_PRM Specificty": 0.6956521739130435,
      "eval_loss": 0.4973340928554535,
      "eval_runtime": 14.7257,
      "eval_samples_per_second": 2.988,
      "eval_steps_per_second": 0.136,
      "step": 205
    },
    {
      "epoch": 0.05334887183969441,
      "grad_norm": 9.323152018153321,
      "learning_rate": 6.65374677002584e-07,
      "loss": 0.4315,
      "step": 206
    },
    {
      "epoch": 0.05360784694571234,
      "grad_norm": 6.5863056309687416,
      "learning_rate": 6.686046511627907e-07,
      "loss": 0.5517,
      "step": 207
    },
    {
      "epoch": 0.053866822051730275,
      "grad_norm": 5.274715923239226,
      "learning_rate": 6.718346253229975e-07,
      "loss": 0.4108,
      "step": 208
    },
    {
      "epoch": 0.054125797157748214,
      "grad_norm": 11.927133654509566,
      "learning_rate": 6.750645994832042e-07,
      "loss": 0.5993,
      "step": 209
    },
    {
      "epoch": 0.054384772263766146,
      "grad_norm": 6.0307980409502155,
      "learning_rate": 6.782945736434108e-07,
      "loss": 0.5751,
      "step": 210
    },
    {
      "epoch": 0.054384772263766146,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8289473684210527,
      "eval_PRM F1 AUC": 0.749083289680461,
      "eval_PRM F1 AUC (fixed)": 0.43766369827134627,
      "eval_PRM F1 Neg": 0.5666666666666667,
      "eval_PRM NPV": 0.4594594594594595,
      "eval_PRM Precision": 0.9130434782608695,
      "eval_PRM Recall": 0.7590361445783133,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.49228909611701965,
      "eval_runtime": 14.7251,
      "eval_samples_per_second": 2.988,
      "eval_steps_per_second": 0.136,
      "step": 210
    },
    {
      "epoch": 0.05464374736978408,
      "grad_norm": 12.144312992076157,
      "learning_rate": 6.815245478036177e-07,
      "loss": 0.4427,
      "step": 211
    },
    {
      "epoch": 0.05490272247580202,
      "grad_norm": 5.780920751157059,
      "learning_rate": 6.847545219638244e-07,
      "loss": 0.6016,
      "step": 212
    },
    {
      "epoch": 0.05516169758181995,
      "grad_norm": 5.163882391479198,
      "learning_rate": 6.87984496124031e-07,
      "loss": 0.5456,
      "step": 213
    },
    {
      "epoch": 0.05542067268783788,
      "grad_norm": 12.425837231569478,
      "learning_rate": 6.912144702842377e-07,
      "loss": 0.4962,
      "step": 214
    },
    {
      "epoch": 0.05567964779385581,
      "grad_norm": 4.927140435616092,
      "learning_rate": 6.944444444444446e-07,
      "loss": 0.4614,
      "step": 215
    },
    {
      "epoch": 0.05567964779385581,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8289473684210527,
      "eval_PRM F1 AUC": 0.749083289680461,
      "eval_PRM F1 AUC (fixed)": 0.43871136720796233,
      "eval_PRM F1 Neg": 0.5666666666666667,
      "eval_PRM NPV": 0.4594594594594595,
      "eval_PRM Precision": 0.9130434782608695,
      "eval_PRM Recall": 0.7590361445783133,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.48628750443458557,
      "eval_runtime": 14.8785,
      "eval_samples_per_second": 2.957,
      "eval_steps_per_second": 0.134,
      "step": 215
    },
    {
      "epoch": 0.05593862289987375,
      "grad_norm": 6.655940409281741,
      "learning_rate": 6.976744186046513e-07,
      "loss": 0.4614,
      "step": 216
    },
    {
      "epoch": 0.056197598005891684,
      "grad_norm": 8.717195609708812,
      "learning_rate": 7.009043927648579e-07,
      "loss": 0.6091,
      "step": 217
    },
    {
      "epoch": 0.056456573111909616,
      "grad_norm": 9.066839424841193,
      "learning_rate": 7.041343669250647e-07,
      "loss": 0.5783,
      "step": 218
    },
    {
      "epoch": 0.05671554821792755,
      "grad_norm": 6.088716178459927,
      "learning_rate": 7.073643410852714e-07,
      "loss": 0.4702,
      "step": 219
    },
    {
      "epoch": 0.05697452332394549,
      "grad_norm": 6.681353857769403,
      "learning_rate": 7.10594315245478e-07,
      "loss": 0.483,
      "step": 220
    },
    {
      "epoch": 0.05697452332394549,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8235294117647058,
      "eval_PRM F1 AUC": 0.7273441592456783,
      "eval_PRM F1 AUC (fixed)": 0.44473546359350447,
      "eval_PRM F1 Neg": 0.5423728813559322,
      "eval_PRM NPV": 0.4444444444444444,
      "eval_PRM Precision": 0.9,
      "eval_PRM Recall": 0.7590361445783133,
      "eval_PRM Specificty": 0.6956521739130435,
      "eval_loss": 0.4895632565021515,
      "eval_runtime": 14.7295,
      "eval_samples_per_second": 2.987,
      "eval_steps_per_second": 0.136,
      "step": 220
    },
    {
      "epoch": 0.05723349842996342,
      "grad_norm": 5.913098681014986,
      "learning_rate": 7.138242894056849e-07,
      "loss": 0.5357,
      "step": 221
    },
    {
      "epoch": 0.05749247353598135,
      "grad_norm": 5.158407140201309,
      "learning_rate": 7.170542635658916e-07,
      "loss": 0.4635,
      "step": 222
    },
    {
      "epoch": 0.05775144864199929,
      "grad_norm": 6.47016455998915,
      "learning_rate": 7.202842377260982e-07,
      "loss": 0.4952,
      "step": 223
    },
    {
      "epoch": 0.05801042374801722,
      "grad_norm": 10.23651185109097,
      "learning_rate": 7.235142118863049e-07,
      "loss": 0.5544,
      "step": 224
    },
    {
      "epoch": 0.058269398854035155,
      "grad_norm": 4.578430497378247,
      "learning_rate": 7.267441860465117e-07,
      "loss": 0.5148,
      "step": 225
    },
    {
      "epoch": 0.058269398854035155,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8289473684210527,
      "eval_PRM F1 AUC": 0.749083289680461,
      "eval_PRM F1 AUC (fixed)": 0.4499738082765845,
      "eval_PRM F1 Neg": 0.5666666666666667,
      "eval_PRM NPV": 0.4594594594594595,
      "eval_PRM Precision": 0.9130434782608695,
      "eval_PRM Recall": 0.7590361445783133,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.4968224763870239,
      "eval_runtime": 15.0589,
      "eval_samples_per_second": 2.922,
      "eval_steps_per_second": 0.133,
      "step": 225
    },
    {
      "epoch": 0.05852837396005309,
      "grad_norm": 12.668163398816281,
      "learning_rate": 7.299741602067184e-07,
      "loss": 0.4264,
      "step": 226
    },
    {
      "epoch": 0.058787349066071026,
      "grad_norm": 4.73940430219435,
      "learning_rate": 7.332041343669251e-07,
      "loss": 0.5648,
      "step": 227
    },
    {
      "epoch": 0.05904632417208896,
      "grad_norm": 5.39766175412595,
      "learning_rate": 7.364341085271319e-07,
      "loss": 0.4793,
      "step": 228
    },
    {
      "epoch": 0.05930529927810689,
      "grad_norm": 11.661839654745418,
      "learning_rate": 7.396640826873386e-07,
      "loss": 0.5498,
      "step": 229
    },
    {
      "epoch": 0.05956427438412482,
      "grad_norm": 6.043419075157401,
      "learning_rate": 7.428940568475452e-07,
      "loss": 0.4511,
      "step": 230
    },
    {
      "epoch": 0.05956427438412482,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8266666666666667,
      "eval_PRM F1 AUC": 0.7647983237297015,
      "eval_PRM F1 AUC (fixed)": 0.4562598218962808,
      "eval_PRM F1 Neg": 0.5806451612903226,
      "eval_PRM NPV": 0.46153846153846156,
      "eval_PRM Precision": 0.9253731343283582,
      "eval_PRM Recall": 0.7469879518072289,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.49978208541870117,
      "eval_runtime": 14.4473,
      "eval_samples_per_second": 3.046,
      "eval_steps_per_second": 0.138,
      "step": 230
    },
    {
      "epoch": 0.05982324949014276,
      "grad_norm": 7.968824617613928,
      "learning_rate": 7.461240310077519e-07,
      "loss": 0.5601,
      "step": 231
    },
    {
      "epoch": 0.060082224596160694,
      "grad_norm": 4.1475482933198595,
      "learning_rate": 7.493540051679588e-07,
      "loss": 0.508,
      "step": 232
    },
    {
      "epoch": 0.060341199702178626,
      "grad_norm": 10.162275739039046,
      "learning_rate": 7.525839793281655e-07,
      "loss": 0.4295,
      "step": 233
    },
    {
      "epoch": 0.060600174808196565,
      "grad_norm": 17.411414443373726,
      "learning_rate": 7.558139534883721e-07,
      "loss": 0.5595,
      "step": 234
    },
    {
      "epoch": 0.0608591499142145,
      "grad_norm": 6.93839153430805,
      "learning_rate": 7.590439276485789e-07,
      "loss": 0.5302,
      "step": 235
    },
    {
      "epoch": 0.0608591499142145,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8027210884353742,
      "eval_PRM F1 AUC": 0.746726034573075,
      "eval_PRM F1 AUC (fixed)": 0.45468831849135677,
      "eval_PRM F1 Neg": 0.5538461538461539,
      "eval_PRM NPV": 0.42857142857142855,
      "eval_PRM Precision": 0.921875,
      "eval_PRM Recall": 0.7108433734939759,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.5159228444099426,
      "eval_runtime": 14.8959,
      "eval_samples_per_second": 2.954,
      "eval_steps_per_second": 0.134,
      "step": 235
    },
    {
      "epoch": 0.06111812502023243,
      "grad_norm": 14.524608563458866,
      "learning_rate": 7.622739018087856e-07,
      "loss": 0.5027,
      "step": 236
    },
    {
      "epoch": 0.06137710012625036,
      "grad_norm": 12.416401593553305,
      "learning_rate": 7.655038759689922e-07,
      "loss": 0.5593,
      "step": 237
    },
    {
      "epoch": 0.0616360752322683,
      "grad_norm": 14.295373057361312,
      "learning_rate": 7.68733850129199e-07,
      "loss": 0.6659,
      "step": 238
    },
    {
      "epoch": 0.06189505033828623,
      "grad_norm": 5.522965374953522,
      "learning_rate": 7.719638242894058e-07,
      "loss": 0.5088,
      "step": 239
    },
    {
      "epoch": 0.062154025444304165,
      "grad_norm": 5.111014366876256,
      "learning_rate": 7.751937984496125e-07,
      "loss": 0.4997,
      "step": 240
    },
    {
      "epoch": 0.062154025444304165,
      "eval_PRM Accuracy": 0.7264150943396226,
      "eval_PRM F1": 0.8053691275167785,
      "eval_PRM F1 AUC": 0.7310110005238345,
      "eval_PRM F1 AUC (fixed)": 0.46071241487689885,
      "eval_PRM F1 Neg": 0.5396825396825397,
      "eval_PRM NPV": 0.425,
      "eval_PRM Precision": 0.9090909090909091,
      "eval_PRM Recall": 0.7228915662650602,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.5017386078834534,
      "eval_runtime": 14.7115,
      "eval_samples_per_second": 2.991,
      "eval_steps_per_second": 0.136,
      "step": 240
    },
    {
      "epoch": 0.062413000550322104,
      "grad_norm": 4.557931809240471,
      "learning_rate": 7.784237726098191e-07,
      "loss": 0.4011,
      "step": 241
    },
    {
      "epoch": 0.06267197565634003,
      "grad_norm": 14.003216936602616,
      "learning_rate": 7.81653746770026e-07,
      "loss": 0.3932,
      "step": 242
    },
    {
      "epoch": 0.06293095076235797,
      "grad_norm": 7.843679582137715,
      "learning_rate": 7.848837209302327e-07,
      "loss": 0.4951,
      "step": 243
    },
    {
      "epoch": 0.0631899258683759,
      "grad_norm": 7.845569085491128,
      "learning_rate": 7.881136950904393e-07,
      "loss": 0.4397,
      "step": 244
    },
    {
      "epoch": 0.06344890097439383,
      "grad_norm": 5.61831702884945,
      "learning_rate": 7.913436692506461e-07,
      "loss": 0.4852,
      "step": 245
    },
    {
      "epoch": 0.06344890097439383,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8311688311688312,
      "eval_PRM F1 AUC": 0.7333682556312205,
      "eval_PRM F1 AUC (fixed)": 0.46176008381351497,
      "eval_PRM F1 Neg": 0.5517241379310345,
      "eval_PRM NPV": 0.45714285714285713,
      "eval_PRM Precision": 0.9014084507042254,
      "eval_PRM Recall": 0.7710843373493976,
      "eval_PRM Specificty": 0.6956521739130435,
      "eval_loss": 0.47543448209762573,
      "eval_runtime": 14.8812,
      "eval_samples_per_second": 2.957,
      "eval_steps_per_second": 0.134,
      "step": 245
    },
    {
      "epoch": 0.06370787608041177,
      "grad_norm": 4.346504560592329,
      "learning_rate": 7.945736434108528e-07,
      "loss": 0.454,
      "step": 246
    },
    {
      "epoch": 0.06396685118642971,
      "grad_norm": 5.69092130458769,
      "learning_rate": 7.978036175710595e-07,
      "loss": 0.4296,
      "step": 247
    },
    {
      "epoch": 0.06422582629244764,
      "grad_norm": 5.621082028360281,
      "learning_rate": 8.010335917312661e-07,
      "loss": 0.4341,
      "step": 248
    },
    {
      "epoch": 0.06448480139846557,
      "grad_norm": 19.347757597432366,
      "learning_rate": 8.04263565891473e-07,
      "loss": 0.4821,
      "step": 249
    },
    {
      "epoch": 0.06474377650448351,
      "grad_norm": 17.86909915880335,
      "learning_rate": 8.074935400516797e-07,
      "loss": 0.6175,
      "step": 250
    },
    {
      "epoch": 0.06474377650448351,
      "eval_PRM Accuracy": 0.7735849056603774,
      "eval_PRM F1": 0.8461538461538461,
      "eval_PRM F1 AUC": 0.7454164484023049,
      "eval_PRM F1 AUC (fixed)": 0.4855945521215297,
      "eval_PRM F1 Neg": 0.5714285714285714,
      "eval_PRM NPV": 0.48484848484848486,
      "eval_PRM Precision": 0.9041095890410958,
      "eval_PRM Recall": 0.7951807228915663,
      "eval_PRM Specificty": 0.6956521739130435,
      "eval_loss": 0.44122856855392456,
      "eval_runtime": 14.5299,
      "eval_samples_per_second": 3.028,
      "eval_steps_per_second": 0.138,
      "step": 250
    },
    {
      "epoch": 0.06500275161050144,
      "grad_norm": 10.984232298038023,
      "learning_rate": 8.107235142118863e-07,
      "loss": 0.4992,
      "step": 251
    },
    {
      "epoch": 0.06526172671651938,
      "grad_norm": 16.24839802704603,
      "learning_rate": 8.139534883720931e-07,
      "loss": 0.542,
      "step": 252
    },
    {
      "epoch": 0.0655207018225373,
      "grad_norm": 4.4357811033412045,
      "learning_rate": 8.171834625322998e-07,
      "loss": 0.3877,
      "step": 253
    },
    {
      "epoch": 0.06577967692855524,
      "grad_norm": 4.016940858651593,
      "learning_rate": 8.204134366925066e-07,
      "loss": 0.4171,
      "step": 254
    },
    {
      "epoch": 0.06603865203457318,
      "grad_norm": 3.563876436826327,
      "learning_rate": 8.236434108527132e-07,
      "loss": 0.3769,
      "step": 255
    },
    {
      "epoch": 0.06603865203457318,
      "eval_PRM Accuracy": 0.7830188679245284,
      "eval_PRM F1": 0.8496732026143791,
      "eval_PRM F1 AUC": 0.782870612886328,
      "eval_PRM F1 AUC (fixed)": 0.4963331587218439,
      "eval_PRM F1 Neg": 0.6101694915254238,
      "eval_PRM NPV": 0.5,
      "eval_PRM Precision": 0.9285714285714286,
      "eval_PRM Recall": 0.7831325301204819,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.44621098041534424,
      "eval_runtime": 14.8363,
      "eval_samples_per_second": 2.966,
      "eval_steps_per_second": 0.135,
      "step": 255
    },
    {
      "epoch": 0.0662976271405911,
      "grad_norm": 10.242691540094073,
      "learning_rate": 8.2687338501292e-07,
      "loss": 0.4985,
      "step": 256
    },
    {
      "epoch": 0.06655660224660905,
      "grad_norm": 6.350176841521547,
      "learning_rate": 8.301033591731267e-07,
      "loss": 0.4529,
      "step": 257
    },
    {
      "epoch": 0.06681557735262698,
      "grad_norm": 5.5896844097563845,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.4475,
      "step": 258
    },
    {
      "epoch": 0.06707455245864491,
      "grad_norm": 8.47310401816812,
      "learning_rate": 8.365633074935402e-07,
      "loss": 0.5157,
      "step": 259
    },
    {
      "epoch": 0.06733352756466285,
      "grad_norm": 4.916322297452303,
      "learning_rate": 8.397932816537469e-07,
      "loss": 0.4844,
      "step": 260
    },
    {
      "epoch": 0.06733352756466285,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8187919463087249,
      "eval_PRM F1 AUC": 0.7587742273441593,
      "eval_PRM F1 AUC (fixed)": 0.5141435306443164,
      "eval_PRM F1 Neg": 0.5714285714285714,
      "eval_PRM NPV": 0.45,
      "eval_PRM Precision": 0.9242424242424242,
      "eval_PRM Recall": 0.7349397590361446,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.4647907018661499,
      "eval_runtime": 14.8252,
      "eval_samples_per_second": 2.968,
      "eval_steps_per_second": 0.135,
      "step": 260
    },
    {
      "epoch": 0.06759250267068079,
      "grad_norm": 5.819653797435914,
      "learning_rate": 8.430232558139535e-07,
      "loss": 0.5159,
      "step": 261
    },
    {
      "epoch": 0.06785147777669871,
      "grad_norm": 4.869616721137864,
      "learning_rate": 8.462532299741603e-07,
      "loss": 0.5286,
      "step": 262
    },
    {
      "epoch": 0.06811045288271665,
      "grad_norm": 6.158491506026584,
      "learning_rate": 8.49483204134367e-07,
      "loss": 0.4261,
      "step": 263
    },
    {
      "epoch": 0.06836942798873458,
      "grad_norm": 7.8691608657818355,
      "learning_rate": 8.527131782945737e-07,
      "loss": 0.4181,
      "step": 264
    },
    {
      "epoch": 0.06862840309475252,
      "grad_norm": 4.668507407839531,
      "learning_rate": 8.559431524547804e-07,
      "loss": 0.4744,
      "step": 265
    },
    {
      "epoch": 0.06862840309475252,
      "eval_PRM Accuracy": 0.7641509433962265,
      "eval_PRM F1": 0.8322147651006712,
      "eval_PRM F1 AUC": 0.7865374541644841,
      "eval_PRM F1 AUC (fixed)": 0.5146673651126245,
      "eval_PRM F1 Neg": 0.6031746031746031,
      "eval_PRM NPV": 0.475,
      "eval_PRM Precision": 0.9393939393939394,
      "eval_PRM Recall": 0.7469879518072289,
      "eval_PRM Specificty": 0.8260869565217391,
      "eval_loss": 0.4597613513469696,
      "eval_runtime": 14.7087,
      "eval_samples_per_second": 2.991,
      "eval_steps_per_second": 0.136,
      "step": 265
    },
    {
      "epoch": 0.06888737820077045,
      "grad_norm": 5.975907751002809,
      "learning_rate": 8.591731266149872e-07,
      "loss": 0.4263,
      "step": 266
    },
    {
      "epoch": 0.06914635330678838,
      "grad_norm": 9.117203887331963,
      "learning_rate": 8.624031007751939e-07,
      "loss": 0.4882,
      "step": 267
    },
    {
      "epoch": 0.06940532841280632,
      "grad_norm": 9.706182790356385,
      "learning_rate": 8.656330749354005e-07,
      "loss": 0.5266,
      "step": 268
    },
    {
      "epoch": 0.06966430351882426,
      "grad_norm": 5.136582860683114,
      "learning_rate": 8.688630490956073e-07,
      "loss": 0.407,
      "step": 269
    },
    {
      "epoch": 0.06992327862484218,
      "grad_norm": 8.3622633548138,
      "learning_rate": 8.720930232558141e-07,
      "loss": 0.436,
      "step": 270
    },
    {
      "epoch": 0.06992327862484218,
      "eval_PRM Accuracy": 0.7735849056603774,
      "eval_PRM F1": 0.8441558441558441,
      "eval_PRM F1 AUC": 0.7611314824515453,
      "eval_PRM F1 AUC (fixed)": 0.5154531168150864,
      "eval_PRM F1 Neg": 0.5862068965517241,
      "eval_PRM NPV": 0.4857142857142857,
      "eval_PRM Precision": 0.9154929577464789,
      "eval_PRM Recall": 0.7831325301204819,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.4233591854572296,
      "eval_runtime": 14.4694,
      "eval_samples_per_second": 3.041,
      "eval_steps_per_second": 0.138,
      "step": 270
    },
    {
      "epoch": 0.07018225373086012,
      "grad_norm": 10.168784870346583,
      "learning_rate": 8.753229974160208e-07,
      "loss": 0.513,
      "step": 271
    },
    {
      "epoch": 0.07044122883687806,
      "grad_norm": 4.326897308284535,
      "learning_rate": 8.785529715762274e-07,
      "loss": 0.4433,
      "step": 272
    },
    {
      "epoch": 0.07070020394289599,
      "grad_norm": 5.672475397941136,
      "learning_rate": 8.817829457364342e-07,
      "loss": 0.4609,
      "step": 273
    },
    {
      "epoch": 0.07095917904891393,
      "grad_norm": 17.141990085978755,
      "learning_rate": 8.850129198966409e-07,
      "loss": 0.6045,
      "step": 274
    },
    {
      "epoch": 0.07121815415493185,
      "grad_norm": 5.220531040231862,
      "learning_rate": 8.882428940568475e-07,
      "loss": 0.4902,
      "step": 275
    },
    {
      "epoch": 0.07121815415493185,
      "eval_PRM Accuracy": 0.7830188679245284,
      "eval_PRM F1": 0.8496732026143791,
      "eval_PRM F1 AUC": 0.782870612886328,
      "eval_PRM F1 AUC (fixed)": 0.5125720272393923,
      "eval_PRM F1 Neg": 0.6101694915254238,
      "eval_PRM NPV": 0.5,
      "eval_PRM Precision": 0.9285714285714286,
      "eval_PRM Recall": 0.7831325301204819,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.43242713809013367,
      "eval_runtime": 14.5138,
      "eval_samples_per_second": 3.032,
      "eval_steps_per_second": 0.138,
      "step": 275
    },
    {
      "epoch": 0.07147712926094979,
      "grad_norm": 3.5766465490704893,
      "learning_rate": 8.914728682170544e-07,
      "loss": 0.416,
      "step": 276
    },
    {
      "epoch": 0.07173610436696773,
      "grad_norm": 10.224622754440063,
      "learning_rate": 8.947028423772611e-07,
      "loss": 0.4704,
      "step": 277
    },
    {
      "epoch": 0.07199507947298565,
      "grad_norm": 7.746233043103798,
      "learning_rate": 8.979328165374678e-07,
      "loss": 0.4716,
      "step": 278
    },
    {
      "epoch": 0.0722540545790036,
      "grad_norm": 5.999186571327644,
      "learning_rate": 9.011627906976744e-07,
      "loss": 0.46,
      "step": 279
    },
    {
      "epoch": 0.07251302968502153,
      "grad_norm": 4.481831549526554,
      "learning_rate": 9.043927648578812e-07,
      "loss": 0.4955,
      "step": 280
    },
    {
      "epoch": 0.07251302968502153,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8266666666666667,
      "eval_PRM F1 AUC": 0.7647983237297015,
      "eval_PRM F1 AUC (fixed)": 0.504190675746464,
      "eval_PRM F1 Neg": 0.5806451612903226,
      "eval_PRM NPV": 0.46153846153846156,
      "eval_PRM Precision": 0.9253731343283582,
      "eval_PRM Recall": 0.7469879518072289,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.46966707706451416,
      "eval_runtime": 14.5021,
      "eval_samples_per_second": 3.034,
      "eval_steps_per_second": 0.138,
      "step": 280
    },
    {
      "epoch": 0.07277200479103946,
      "grad_norm": 15.916954907020221,
      "learning_rate": 9.07622739018088e-07,
      "loss": 0.4277,
      "step": 281
    },
    {
      "epoch": 0.0730309798970574,
      "grad_norm": 10.711836972773476,
      "learning_rate": 9.108527131782946e-07,
      "loss": 0.4199,
      "step": 282
    },
    {
      "epoch": 0.07328995500307534,
      "grad_norm": 17.339949092453057,
      "learning_rate": 9.140826873385014e-07,
      "loss": 0.5185,
      "step": 283
    },
    {
      "epoch": 0.07354893010909326,
      "grad_norm": 6.559593493304679,
      "learning_rate": 9.173126614987081e-07,
      "loss": 0.4751,
      "step": 284
    },
    {
      "epoch": 0.0738079052151112,
      "grad_norm": 12.211437994570204,
      "learning_rate": 9.205426356589148e-07,
      "loss": 0.3782,
      "step": 285
    },
    {
      "epoch": 0.0738079052151112,
      "eval_PRM Accuracy": 0.7641509433962265,
      "eval_PRM F1": 0.8344370860927153,
      "eval_PRM F1 AUC": 0.7708224201152437,
      "eval_PRM F1 AUC (fixed)": 0.5107386066003143,
      "eval_PRM F1 Neg": 0.5901639344262295,
      "eval_PRM NPV": 0.47368421052631576,
      "eval_PRM Precision": 0.9264705882352942,
      "eval_PRM Recall": 0.7590361445783133,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.4466630816459656,
      "eval_runtime": 14.8677,
      "eval_samples_per_second": 2.959,
      "eval_steps_per_second": 0.135,
      "step": 285
    },
    {
      "epoch": 0.07406688032112912,
      "grad_norm": 9.505648050872507,
      "learning_rate": 9.237726098191216e-07,
      "loss": 0.3501,
      "step": 286
    },
    {
      "epoch": 0.07432585542714706,
      "grad_norm": 20.876440461993184,
      "learning_rate": 9.270025839793283e-07,
      "loss": 0.5186,
      "step": 287
    },
    {
      "epoch": 0.074584830533165,
      "grad_norm": 4.231394956008767,
      "learning_rate": 9.30232558139535e-07,
      "loss": 0.3812,
      "step": 288
    },
    {
      "epoch": 0.07484380563918293,
      "grad_norm": 7.798047063861872,
      "learning_rate": 9.334625322997416e-07,
      "loss": 0.4313,
      "step": 289
    },
    {
      "epoch": 0.07510278074520087,
      "grad_norm": 10.845228628781953,
      "learning_rate": 9.366925064599484e-07,
      "loss": 0.4542,
      "step": 290
    },
    {
      "epoch": 0.07510278074520087,
      "eval_PRM Accuracy": 0.7924528301886793,
      "eval_PRM F1": 0.8607594936708861,
      "eval_PRM F1 AUC": 0.7574646411733892,
      "eval_PRM F1 AUC (fixed)": 0.5178103719224725,
      "eval_PRM F1 Neg": 0.5925925925925926,
      "eval_PRM NPV": 0.5161290322580645,
      "eval_PRM Precision": 0.9066666666666666,
      "eval_PRM Recall": 0.8192771084337349,
      "eval_PRM Specificty": 0.6956521739130435,
      "eval_loss": 0.4175150692462921,
      "eval_runtime": 14.2443,
      "eval_samples_per_second": 3.089,
      "eval_steps_per_second": 0.14,
      "step": 290
    },
    {
      "epoch": 0.0753617558512188,
      "grad_norm": 12.878911968596602,
      "learning_rate": 9.399224806201551e-07,
      "loss": 0.5125,
      "step": 291
    },
    {
      "epoch": 0.07562073095723673,
      "grad_norm": 10.632764116022619,
      "learning_rate": 9.431524547803617e-07,
      "loss": 0.413,
      "step": 292
    },
    {
      "epoch": 0.07587970606325467,
      "grad_norm": 8.314195563054465,
      "learning_rate": 9.463824289405686e-07,
      "loss": 0.4092,
      "step": 293
    },
    {
      "epoch": 0.07613868116927261,
      "grad_norm": 4.533822120001036,
      "learning_rate": 9.496124031007753e-07,
      "loss": 0.45,
      "step": 294
    },
    {
      "epoch": 0.07639765627529053,
      "grad_norm": 4.147278873212908,
      "learning_rate": 9.52842377260982e-07,
      "loss": 0.4271,
      "step": 295
    },
    {
      "epoch": 0.07639765627529053,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8266666666666667,
      "eval_PRM F1 AUC": 0.7647983237297015,
      "eval_PRM F1 AUC (fixed)": 0.5240963855421688,
      "eval_PRM F1 Neg": 0.5806451612903226,
      "eval_PRM NPV": 0.46153846153846156,
      "eval_PRM Precision": 0.9253731343283582,
      "eval_PRM Recall": 0.7469879518072289,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.4515579342842102,
      "eval_runtime": 14.6113,
      "eval_samples_per_second": 3.011,
      "eval_steps_per_second": 0.137,
      "step": 295
    },
    {
      "epoch": 0.07665663138130847,
      "grad_norm": 12.839286179768832,
      "learning_rate": 9.560723514211886e-07,
      "loss": 0.4118,
      "step": 296
    },
    {
      "epoch": 0.0769156064873264,
      "grad_norm": 14.42556307757587,
      "learning_rate": 9.593023255813955e-07,
      "loss": 0.5255,
      "step": 297
    },
    {
      "epoch": 0.07717458159334434,
      "grad_norm": 3.6706140600620834,
      "learning_rate": 9.625322997416022e-07,
      "loss": 0.4228,
      "step": 298
    },
    {
      "epoch": 0.07743355669936228,
      "grad_norm": 7.9469588293703595,
      "learning_rate": 9.657622739018088e-07,
      "loss": 0.4946,
      "step": 299
    },
    {
      "epoch": 0.0776925318053802,
      "grad_norm": 8.228295223889006,
      "learning_rate": 9.689922480620157e-07,
      "loss": 0.4881,
      "step": 300
    },
    {
      "epoch": 0.0776925318053802,
      "eval_PRM Accuracy": 0.7075471698113207,
      "eval_PRM F1": 0.7832167832167832,
      "eval_PRM F1 AUC": 0.7503928758512309,
      "eval_PRM F1 AUC (fixed)": 0.5301204819277108,
      "eval_PRM F1 Neg": 0.5507246376811594,
      "eval_PRM NPV": 0.41304347826086957,
      "eval_PRM Precision": 0.9333333333333333,
      "eval_PRM Recall": 0.6746987951807228,
      "eval_PRM Specificty": 0.8260869565217391,
      "eval_loss": 0.4778639078140259,
      "eval_runtime": 14.3702,
      "eval_samples_per_second": 3.062,
      "eval_steps_per_second": 0.139,
      "step": 300
    },
    {
      "epoch": 0.07795150691139814,
      "grad_norm": 7.801989344153817,
      "learning_rate": 9.722222222222224e-07,
      "loss": 0.4925,
      "step": 301
    },
    {
      "epoch": 0.07821048201741608,
      "grad_norm": 6.194112700855093,
      "learning_rate": 9.75452196382429e-07,
      "loss": 0.5078,
      "step": 302
    },
    {
      "epoch": 0.078469457123434,
      "grad_norm": 4.334517501586195,
      "learning_rate": 9.786821705426357e-07,
      "loss": 0.4482,
      "step": 303
    },
    {
      "epoch": 0.07872843222945194,
      "grad_norm": 7.02828276789377,
      "learning_rate": 9.819121447028424e-07,
      "loss": 0.4523,
      "step": 304
    },
    {
      "epoch": 0.07898740733546988,
      "grad_norm": 4.699628027384838,
      "learning_rate": 9.85142118863049e-07,
      "loss": 0.3978,
      "step": 305
    },
    {
      "epoch": 0.07898740733546988,
      "eval_PRM Accuracy": 0.7641509433962265,
      "eval_PRM F1": 0.8344370860927153,
      "eval_PRM F1 AUC": 0.7708224201152437,
      "eval_PRM F1 AUC (fixed)": 0.5319539025667889,
      "eval_PRM F1 Neg": 0.5901639344262295,
      "eval_PRM NPV": 0.47368421052631576,
      "eval_PRM Precision": 0.9264705882352942,
      "eval_PRM Recall": 0.7590361445783133,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.4269060790538788,
      "eval_runtime": 14.6875,
      "eval_samples_per_second": 2.996,
      "eval_steps_per_second": 0.136,
      "step": 305
    },
    {
      "epoch": 0.07924638244148781,
      "grad_norm": 4.731440582096809,
      "learning_rate": 9.88372093023256e-07,
      "loss": 0.3662,
      "step": 306
    },
    {
      "epoch": 0.07950535754750575,
      "grad_norm": 5.599732171367987,
      "learning_rate": 9.916020671834626e-07,
      "loss": 0.456,
      "step": 307
    },
    {
      "epoch": 0.07976433265352369,
      "grad_norm": 5.52753719513111,
      "learning_rate": 9.948320413436693e-07,
      "loss": 0.3806,
      "step": 308
    },
    {
      "epoch": 0.08002330775954161,
      "grad_norm": 11.708582980152471,
      "learning_rate": 9.98062015503876e-07,
      "loss": 0.4499,
      "step": 309
    },
    {
      "epoch": 0.08028228286555955,
      "grad_norm": 4.372614035219098,
      "learning_rate": 1.0012919896640828e-06,
      "loss": 0.4207,
      "step": 310
    },
    {
      "epoch": 0.08028228286555955,
      "eval_PRM Accuracy": 0.8113207547169812,
      "eval_PRM F1": 0.8734177215189873,
      "eval_PRM F1 AUC": 0.785227867993714,
      "eval_PRM F1 AUC (fixed)": 0.5395495023572552,
      "eval_PRM F1 Neg": 0.6296296296296297,
      "eval_PRM NPV": 0.5483870967741935,
      "eval_PRM Precision": 0.92,
      "eval_PRM Recall": 0.8313253012048193,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.3785098195075989,
      "eval_runtime": 14.9751,
      "eval_samples_per_second": 2.938,
      "eval_steps_per_second": 0.134,
      "step": 310
    },
    {
      "epoch": 0.08054125797157748,
      "grad_norm": 14.412298324521696,
      "learning_rate": 1.0045219638242895e-06,
      "loss": 0.527,
      "step": 311
    },
    {
      "epoch": 0.08080023307759542,
      "grad_norm": 10.519064993089941,
      "learning_rate": 1.0077519379844962e-06,
      "loss": 0.425,
      "step": 312
    },
    {
      "epoch": 0.08105920818361335,
      "grad_norm": 21.1615871003155,
      "learning_rate": 1.0109819121447028e-06,
      "loss": 0.5047,
      "step": 313
    },
    {
      "epoch": 0.08131818328963128,
      "grad_norm": 12.442814820549918,
      "learning_rate": 1.0142118863049097e-06,
      "loss": 0.5891,
      "step": 314
    },
    {
      "epoch": 0.08157715839564922,
      "grad_norm": 17.06252344116241,
      "learning_rate": 1.0174418604651164e-06,
      "loss": 0.597,
      "step": 315
    },
    {
      "epoch": 0.08157715839564922,
      "eval_PRM Accuracy": 0.7924528301886793,
      "eval_PRM F1": 0.8552631578947368,
      "eval_PRM F1 AUC": 0.8046097433211105,
      "eval_PRM F1 AUC (fixed)": 0.5468831849135674,
      "eval_PRM F1 Neg": 0.6333333333333333,
      "eval_PRM NPV": 0.5135135135135135,
      "eval_PRM Precision": 0.9420289855072463,
      "eval_PRM Recall": 0.7831325301204819,
      "eval_PRM Specificty": 0.8260869565217391,
      "eval_loss": 0.4063412845134735,
      "eval_runtime": 14.5989,
      "eval_samples_per_second": 3.014,
      "eval_steps_per_second": 0.137,
      "step": 315
    },
    {
      "epoch": 0.08183613350166716,
      "grad_norm": 6.2636774461754285,
      "learning_rate": 1.020671834625323e-06,
      "loss": 0.4252,
      "step": 316
    },
    {
      "epoch": 0.08209510860768508,
      "grad_norm": 6.054523107112382,
      "learning_rate": 1.02390180878553e-06,
      "loss": 0.5367,
      "step": 317
    },
    {
      "epoch": 0.08235408371370302,
      "grad_norm": 9.718566201674209,
      "learning_rate": 1.0271317829457366e-06,
      "loss": 0.3912,
      "step": 318
    },
    {
      "epoch": 0.08261305881972096,
      "grad_norm": 4.987727785104741,
      "learning_rate": 1.0303617571059433e-06,
      "loss": 0.4392,
      "step": 319
    },
    {
      "epoch": 0.08287203392573889,
      "grad_norm": 6.884006523120188,
      "learning_rate": 1.03359173126615e-06,
      "loss": 0.3608,
      "step": 320
    },
    {
      "epoch": 0.08287203392573889,
      "eval_PRM Accuracy": 0.7735849056603774,
      "eval_PRM F1": 0.84,
      "eval_PRM F1 AUC": 0.7925615505500263,
      "eval_PRM F1 AUC (fixed)": 0.5471451021477213,
      "eval_PRM F1 Neg": 0.6129032258064516,
      "eval_PRM NPV": 0.48717948717948717,
      "eval_PRM Precision": 0.9402985074626866,
      "eval_PRM Recall": 0.7590361445783133,
      "eval_PRM Specificty": 0.8260869565217391,
      "eval_loss": 0.4275074303150177,
      "eval_runtime": 14.8425,
      "eval_samples_per_second": 2.964,
      "eval_steps_per_second": 0.135,
      "step": 320
    },
    {
      "epoch": 0.08313100903175683,
      "grad_norm": 5.318854637746786,
      "learning_rate": 1.0368217054263566e-06,
      "loss": 0.4384,
      "step": 321
    },
    {
      "epoch": 0.08338998413777475,
      "grad_norm": 11.435651480653691,
      "learning_rate": 1.0400516795865635e-06,
      "loss": 0.3841,
      "step": 322
    },
    {
      "epoch": 0.08364895924379269,
      "grad_norm": 4.849662752105375,
      "learning_rate": 1.0432816537467702e-06,
      "loss": 0.433,
      "step": 323
    },
    {
      "epoch": 0.08390793434981063,
      "grad_norm": 4.105703022458649,
      "learning_rate": 1.0465116279069768e-06,
      "loss": 0.427,
      "step": 324
    },
    {
      "epoch": 0.08416690945582855,
      "grad_norm": 7.043811548213904,
      "learning_rate": 1.0497416020671835e-06,
      "loss": 0.4356,
      "step": 325
    },
    {
      "epoch": 0.08416690945582855,
      "eval_PRM Accuracy": 0.8018867924528302,
      "eval_PRM F1": 0.8662420382165605,
      "eval_PRM F1 AUC": 0.7792037716081718,
      "eval_PRM F1 AUC (fixed)": 0.5586694604504977,
      "eval_PRM F1 Neg": 0.6181818181818182,
      "eval_PRM NPV": 0.53125,
      "eval_PRM Precision": 0.918918918918919,
      "eval_PRM Recall": 0.8192771084337349,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.3831585943698883,
      "eval_runtime": 14.7099,
      "eval_samples_per_second": 2.991,
      "eval_steps_per_second": 0.136,
      "step": 325
    },
    {
      "epoch": 0.08442588456184649,
      "grad_norm": 5.753783230308081,
      "learning_rate": 1.0529715762273902e-06,
      "loss": 0.4569,
      "step": 326
    },
    {
      "epoch": 0.08468485966786443,
      "grad_norm": 7.813829276934702,
      "learning_rate": 1.0562015503875968e-06,
      "loss": 0.4649,
      "step": 327
    },
    {
      "epoch": 0.08494383477388236,
      "grad_norm": 16.4994165028651,
      "learning_rate": 1.0594315245478037e-06,
      "loss": 0.5808,
      "step": 328
    },
    {
      "epoch": 0.0852028098799003,
      "grad_norm": 15.370926504978964,
      "learning_rate": 1.0626614987080104e-06,
      "loss": 0.5716,
      "step": 329
    },
    {
      "epoch": 0.08546178498591823,
      "grad_norm": 7.9551510457294805,
      "learning_rate": 1.065891472868217e-06,
      "loss": 0.4672,
      "step": 330
    },
    {
      "epoch": 0.08546178498591823,
      "eval_PRM Accuracy": 0.8018867924528302,
      "eval_PRM F1": 0.8662420382165605,
      "eval_PRM F1 AUC": 0.7792037716081718,
      "eval_PRM F1 AUC (fixed)": 0.5576217915138817,
      "eval_PRM F1 Neg": 0.6181818181818182,
      "eval_PRM NPV": 0.53125,
      "eval_PRM Precision": 0.918918918918919,
      "eval_PRM Recall": 0.8192771084337349,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.3691127598285675,
      "eval_runtime": 14.6243,
      "eval_samples_per_second": 3.009,
      "eval_steps_per_second": 0.137,
      "step": 330
    },
    {
      "epoch": 0.08572076009193616,
      "grad_norm": 4.515299251404578,
      "learning_rate": 1.069121447028424e-06,
      "loss": 0.4701,
      "step": 331
    },
    {
      "epoch": 0.0859797351979541,
      "grad_norm": 5.9679777493199095,
      "learning_rate": 1.0723514211886306e-06,
      "loss": 0.466,
      "step": 332
    },
    {
      "epoch": 0.08623871030397202,
      "grad_norm": 4.152376826263897,
      "learning_rate": 1.0755813953488373e-06,
      "loss": 0.4128,
      "step": 333
    },
    {
      "epoch": 0.08649768540998996,
      "grad_norm": 9.31379252266598,
      "learning_rate": 1.0788113695090441e-06,
      "loss": 0.401,
      "step": 334
    },
    {
      "epoch": 0.0867566605160079,
      "grad_norm": 13.247200202082919,
      "learning_rate": 1.0820413436692508e-06,
      "loss": 0.5408,
      "step": 335
    },
    {
      "epoch": 0.0867566605160079,
      "eval_PRM Accuracy": 0.8113207547169812,
      "eval_PRM F1": 0.8717948717948718,
      "eval_PRM F1 AUC": 0.8009429020429545,
      "eval_PRM F1 AUC (fixed)": 0.5500261917234154,
      "eval_PRM F1 Neg": 0.6428571428571429,
      "eval_PRM NPV": 0.5454545454545454,
      "eval_PRM Precision": 0.9315068493150684,
      "eval_PRM Recall": 0.8192771084337349,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.39321768283843994,
      "eval_runtime": 14.484,
      "eval_samples_per_second": 3.038,
      "eval_steps_per_second": 0.138,
      "step": 335
    },
    {
      "epoch": 0.08701563562202583,
      "grad_norm": 5.385907638555454,
      "learning_rate": 1.0852713178294575e-06,
      "loss": 0.3754,
      "step": 336
    },
    {
      "epoch": 0.08727461072804377,
      "grad_norm": 7.009224169412602,
      "learning_rate": 1.0885012919896641e-06,
      "loss": 0.4683,
      "step": 337
    },
    {
      "epoch": 0.0875335858340617,
      "grad_norm": 7.664329467972336,
      "learning_rate": 1.091731266149871e-06,
      "loss": 0.4783,
      "step": 338
    },
    {
      "epoch": 0.08779256094007963,
      "grad_norm": 4.197296615684727,
      "learning_rate": 1.0949612403100777e-06,
      "loss": 0.4905,
      "step": 339
    },
    {
      "epoch": 0.08805153604609757,
      "grad_norm": 10.035421137511468,
      "learning_rate": 1.0981912144702844e-06,
      "loss": 0.3913,
      "step": 340
    },
    {
      "epoch": 0.08805153604609757,
      "eval_PRM Accuracy": 0.7547169811320755,
      "eval_PRM F1": 0.8243243243243243,
      "eval_PRM F1 AUC": 0.7805133577789418,
      "eval_PRM F1 AUC (fixed)": 0.5413829229963332,
      "eval_PRM F1 Neg": 0.59375,
      "eval_PRM NPV": 0.4634146341463415,
      "eval_PRM Precision": 0.9384615384615385,
      "eval_PRM Recall": 0.7349397590361446,
      "eval_PRM Specificty": 0.8260869565217391,
      "eval_loss": 0.4457910656929016,
      "eval_runtime": 14.5536,
      "eval_samples_per_second": 3.023,
      "eval_steps_per_second": 0.137,
      "step": 340
    },
    {
      "epoch": 0.08831051115211551,
      "grad_norm": 7.566916747027618,
      "learning_rate": 1.101421188630491e-06,
      "loss": 0.4404,
      "step": 341
    },
    {
      "epoch": 0.08856948625813343,
      "grad_norm": 7.718429079088188,
      "learning_rate": 1.1046511627906977e-06,
      "loss": 0.4188,
      "step": 342
    },
    {
      "epoch": 0.08882846136415137,
      "grad_norm": 13.14516943643952,
      "learning_rate": 1.1078811369509044e-06,
      "loss": 0.3584,
      "step": 343
    },
    {
      "epoch": 0.0890874364701693,
      "grad_norm": 5.774045581523776,
      "learning_rate": 1.111111111111111e-06,
      "loss": 0.406,
      "step": 344
    },
    {
      "epoch": 0.08934641157618724,
      "grad_norm": 3.599053746712549,
      "learning_rate": 1.114341085271318e-06,
      "loss": 0.3606,
      "step": 345
    },
    {
      "epoch": 0.08934641157618724,
      "eval_PRM Accuracy": 0.8113207547169812,
      "eval_PRM F1": 0.8717948717948718,
      "eval_PRM F1 AUC": 0.8009429020429545,
      "eval_PRM F1 AUC (fixed)": 0.5474070193818753,
      "eval_PRM F1 Neg": 0.6428571428571429,
      "eval_PRM NPV": 0.5454545454545454,
      "eval_PRM Precision": 0.9315068493150684,
      "eval_PRM Recall": 0.8192771084337349,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.40695610642433167,
      "eval_runtime": 14.8039,
      "eval_samples_per_second": 2.972,
      "eval_steps_per_second": 0.135,
      "step": 345
    },
    {
      "epoch": 0.08960538668220518,
      "grad_norm": 6.351538227956327,
      "learning_rate": 1.1175710594315246e-06,
      "loss": 0.4081,
      "step": 346
    },
    {
      "epoch": 0.0898643617882231,
      "grad_norm": 4.552702510203715,
      "learning_rate": 1.1208010335917313e-06,
      "loss": 0.4791,
      "step": 347
    },
    {
      "epoch": 0.09012333689424104,
      "grad_norm": 6.400764438436327,
      "learning_rate": 1.1240310077519381e-06,
      "loss": 0.4699,
      "step": 348
    },
    {
      "epoch": 0.09038231200025898,
      "grad_norm": 6.6839400360275265,
      "learning_rate": 1.1272609819121448e-06,
      "loss": 0.5084,
      "step": 349
    },
    {
      "epoch": 0.0906412871062769,
      "grad_norm": 6.068084345164869,
      "learning_rate": 1.1304909560723515e-06,
      "loss": 0.4268,
      "step": 350
    },
    {
      "epoch": 0.0906412871062769,
      "eval_PRM Accuracy": 0.8113207547169812,
      "eval_PRM F1": 0.8717948717948718,
      "eval_PRM F1 AUC": 0.8009429020429545,
      "eval_PRM F1 AUC (fixed)": 0.569408067050812,
      "eval_PRM F1 Neg": 0.6428571428571429,
      "eval_PRM NPV": 0.5454545454545454,
      "eval_PRM Precision": 0.9315068493150684,
      "eval_PRM Recall": 0.8192771084337349,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.38017958402633667,
      "eval_runtime": 14.8134,
      "eval_samples_per_second": 2.97,
      "eval_steps_per_second": 0.135,
      "step": 350
    },
    {
      "epoch": 0.09090026221229484,
      "grad_norm": 6.000182731927381,
      "learning_rate": 1.1337209302325581e-06,
      "loss": 0.4375,
      "step": 351
    },
    {
      "epoch": 0.09115923731831278,
      "grad_norm": 6.664169382629427,
      "learning_rate": 1.136950904392765e-06,
      "loss": 0.4991,
      "step": 352
    },
    {
      "epoch": 0.09141821242433071,
      "grad_norm": 7.306217251000588,
      "learning_rate": 1.1401808785529717e-06,
      "loss": 0.4768,
      "step": 353
    },
    {
      "epoch": 0.09167718753034865,
      "grad_norm": 7.9308828901201345,
      "learning_rate": 1.1434108527131784e-06,
      "loss": 0.4159,
      "step": 354
    },
    {
      "epoch": 0.09193616263636657,
      "grad_norm": 5.17121110105702,
      "learning_rate": 1.1466408268733852e-06,
      "loss": 0.416,
      "step": 355
    },
    {
      "epoch": 0.09193616263636657,
      "eval_PRM Accuracy": 0.8018867924528302,
      "eval_PRM F1": 0.8627450980392157,
      "eval_PRM F1 AUC": 0.8106338397066527,
      "eval_PRM F1 AUC (fixed)": 0.5573598742797277,
      "eval_PRM F1 Neg": 0.6440677966101694,
      "eval_PRM NPV": 0.5277777777777778,
      "eval_PRM Precision": 0.9428571428571428,
      "eval_PRM Recall": 0.7951807228915663,
      "eval_PRM Specificty": 0.8260869565217391,
      "eval_loss": 0.3932633399963379,
      "eval_runtime": 14.5079,
      "eval_samples_per_second": 3.033,
      "eval_steps_per_second": 0.138,
      "step": 355
    },
    {
      "epoch": 0.09219513774238451,
      "grad_norm": 8.013449320152551,
      "learning_rate": 1.149870801033592e-06,
      "loss": 0.3701,
      "step": 356
    },
    {
      "epoch": 0.09245411284840245,
      "grad_norm": 5.882074020186261,
      "learning_rate": 1.1531007751937986e-06,
      "loss": 0.4908,
      "step": 357
    },
    {
      "epoch": 0.09271308795442038,
      "grad_norm": 3.9901380175520953,
      "learning_rate": 1.1563307493540052e-06,
      "loss": 0.4386,
      "step": 358
    },
    {
      "epoch": 0.09297206306043831,
      "grad_norm": 6.734838796869919,
      "learning_rate": 1.159560723514212e-06,
      "loss": 0.4259,
      "step": 359
    },
    {
      "epoch": 0.09323103816645625,
      "grad_norm": 4.774912365406276,
      "learning_rate": 1.1627906976744188e-06,
      "loss": 0.4545,
      "step": 360
    },
    {
      "epoch": 0.09323103816645625,
      "eval_PRM Accuracy": 0.8301886792452831,
      "eval_PRM F1": 0.8860759493670886,
      "eval_PRM F1 AUC": 0.8129910948140389,
      "eval_PRM F1 AUC (fixed)": 0.5518596123624935,
      "eval_PRM F1 Neg": 0.6666666666666666,
      "eval_PRM NPV": 0.5806451612903226,
      "eval_PRM Precision": 0.9333333333333333,
      "eval_PRM Recall": 0.8433734939759037,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.37586334347724915,
      "eval_runtime": 14.6961,
      "eval_samples_per_second": 2.994,
      "eval_steps_per_second": 0.136,
      "step": 360
    },
    {
      "epoch": 0.09349001327247418,
      "grad_norm": 4.447339694833538,
      "learning_rate": 1.1660206718346255e-06,
      "loss": 0.419,
      "step": 361
    },
    {
      "epoch": 0.09374898837849212,
      "grad_norm": 7.095783943084598,
      "learning_rate": 1.1692506459948321e-06,
      "loss": 0.427,
      "step": 362
    },
    {
      "epoch": 0.09400796348451006,
      "grad_norm": 4.085236885353116,
      "learning_rate": 1.1724806201550388e-06,
      "loss": 0.3795,
      "step": 363
    },
    {
      "epoch": 0.09426693859052798,
      "grad_norm": 8.894911794969753,
      "learning_rate": 1.1757105943152455e-06,
      "loss": 0.3978,
      "step": 364
    },
    {
      "epoch": 0.09452591369654592,
      "grad_norm": 13.919588842134319,
      "learning_rate": 1.1789405684754523e-06,
      "loss": 0.4439,
      "step": 365
    },
    {
      "epoch": 0.09452591369654592,
      "eval_PRM Accuracy": 0.8018867924528302,
      "eval_PRM F1": 0.8695652173913043,
      "eval_PRM F1 AUC": 0.747773703509691,
      "eval_PRM F1 AUC (fixed)": 0.5531691985332635,
      "eval_PRM F1 Neg": 0.5882352941176471,
      "eval_PRM NPV": 0.5357142857142857,
      "eval_PRM Precision": 0.8974358974358975,
      "eval_PRM Recall": 0.8433734939759037,
      "eval_PRM Specificty": 0.6521739130434783,
      "eval_loss": 0.3621164560317993,
      "eval_runtime": 14.1653,
      "eval_samples_per_second": 3.106,
      "eval_steps_per_second": 0.141,
      "step": 365
    },
    {
      "epoch": 0.09478488880256386,
      "grad_norm": 7.367527828756034,
      "learning_rate": 1.182170542635659e-06,
      "loss": 0.3933,
      "step": 366
    },
    {
      "epoch": 0.09504386390858179,
      "grad_norm": 4.401857963264339,
      "learning_rate": 1.1854005167958657e-06,
      "loss": 0.372,
      "step": 367
    },
    {
      "epoch": 0.09530283901459972,
      "grad_norm": 4.066198280850839,
      "learning_rate": 1.1886304909560723e-06,
      "loss": 0.4406,
      "step": 368
    },
    {
      "epoch": 0.09556181412061765,
      "grad_norm": 3.046639350111974,
      "learning_rate": 1.1918604651162792e-06,
      "loss": 0.3952,
      "step": 369
    },
    {
      "epoch": 0.09582078922663559,
      "grad_norm": 3.6302489391357318,
      "learning_rate": 1.195090439276486e-06,
      "loss": 0.3457,
      "step": 370
    },
    {
      "epoch": 0.09582078922663559,
      "eval_PRM Accuracy": 0.8207547169811321,
      "eval_PRM F1": 0.8789808917197452,
      "eval_PRM F1 AUC": 0.8069669984284967,
      "eval_PRM F1 AUC (fixed)": 0.5586694604504976,
      "eval_PRM F1 Neg": 0.6545454545454545,
      "eval_PRM NPV": 0.5625,
      "eval_PRM Precision": 0.9324324324324325,
      "eval_PRM Recall": 0.8313253012048193,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.3780423104763031,
      "eval_runtime": 14.4931,
      "eval_samples_per_second": 3.036,
      "eval_steps_per_second": 0.138,
      "step": 370
    },
    {
      "epoch": 0.09607976433265353,
      "grad_norm": 6.25623119390206,
      "learning_rate": 1.1983204134366926e-06,
      "loss": 0.4254,
      "step": 371
    },
    {
      "epoch": 0.09633873943867145,
      "grad_norm": 6.9752169729123,
      "learning_rate": 1.2015503875968994e-06,
      "loss": 0.5261,
      "step": 372
    },
    {
      "epoch": 0.09659771454468939,
      "grad_norm": 5.294474073160396,
      "learning_rate": 1.2047803617571061e-06,
      "loss": 0.374,
      "step": 373
    },
    {
      "epoch": 0.09685668965070733,
      "grad_norm": 4.659531419947484,
      "learning_rate": 1.2080103359173128e-06,
      "loss": 0.3569,
      "step": 374
    },
    {
      "epoch": 0.09711566475672526,
      "grad_norm": 4.039346702256107,
      "learning_rate": 1.2112403100775194e-06,
      "loss": 0.4512,
      "step": 375
    },
    {
      "epoch": 0.09711566475672526,
      "eval_PRM Accuracy": 0.8301886792452831,
      "eval_PRM F1": 0.8846153846153846,
      "eval_PRM F1 AUC": 0.8287061288632793,
      "eval_PRM F1 AUC (fixed)": 0.5746464117338921,
      "eval_PRM F1 Neg": 0.6785714285714286,
      "eval_PRM NPV": 0.5757575757575758,
      "eval_PRM Precision": 0.9452054794520548,
      "eval_PRM Recall": 0.8313253012048193,
      "eval_PRM Specificty": 0.8260869565217391,
      "eval_loss": 0.37546131014823914,
      "eval_runtime": 14.8182,
      "eval_samples_per_second": 2.969,
      "eval_steps_per_second": 0.135,
      "step": 375
    },
    {
      "epoch": 0.0973746398627432,
      "grad_norm": 6.1017493311919635,
      "learning_rate": 1.2144702842377263e-06,
      "loss": 0.4177,
      "step": 376
    },
    {
      "epoch": 0.09763361496876113,
      "grad_norm": 10.032807377159754,
      "learning_rate": 1.217700258397933e-06,
      "loss": 0.4064,
      "step": 377
    },
    {
      "epoch": 0.09789259007477906,
      "grad_norm": 9.509320958414998,
      "learning_rate": 1.2209302325581397e-06,
      "loss": 0.4294,
      "step": 378
    },
    {
      "epoch": 0.098151565180797,
      "grad_norm": 6.893477527872258,
      "learning_rate": 1.2241602067183463e-06,
      "loss": 0.3717,
      "step": 379
    },
    {
      "epoch": 0.09841054028681492,
      "grad_norm": 4.619892022039742,
      "learning_rate": 1.227390180878553e-06,
      "loss": 0.4551,
      "step": 380
    },
    {
      "epoch": 0.09841054028681492,
      "eval_PRM Accuracy": 0.839622641509434,
      "eval_PRM F1": 0.8930817610062893,
      "eval_PRM F1 AUC": 0.819015191199581,
      "eval_PRM F1 AUC (fixed)": 0.5914091147197484,
      "eval_PRM F1 Neg": 0.6792452830188679,
      "eval_PRM NPV": 0.6,
      "eval_PRM Precision": 0.9342105263157895,
      "eval_PRM Recall": 0.8554216867469879,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.35004815459251404,
      "eval_runtime": 14.5841,
      "eval_samples_per_second": 3.017,
      "eval_steps_per_second": 0.137,
      "step": 380
    },
    {
      "epoch": 0.09866951539283286,
      "grad_norm": 8.222153879654302,
      "learning_rate": 1.2306201550387597e-06,
      "loss": 0.4807,
      "step": 381
    },
    {
      "epoch": 0.0989284904988508,
      "grad_norm": 6.9014811801050895,
      "learning_rate": 1.2338501291989665e-06,
      "loss": 0.4988,
      "step": 382
    },
    {
      "epoch": 0.09918746560486873,
      "grad_norm": 6.180778661225844,
      "learning_rate": 1.2370801033591732e-06,
      "loss": 0.3045,
      "step": 383
    },
    {
      "epoch": 0.09944644071088667,
      "grad_norm": 9.700681223535515,
      "learning_rate": 1.2403100775193799e-06,
      "loss": 0.458,
      "step": 384
    },
    {
      "epoch": 0.0997054158169046,
      "grad_norm": 4.728499860184501,
      "learning_rate": 1.2435400516795866e-06,
      "loss": 0.4022,
      "step": 385
    },
    {
      "epoch": 0.0997054158169046,
      "eval_PRM Accuracy": 0.8301886792452831,
      "eval_PRM F1": 0.8860759493670886,
      "eval_PRM F1 AUC": 0.8129910948140389,
      "eval_PRM F1 AUC (fixed)": 0.606862231534835,
      "eval_PRM F1 Neg": 0.6666666666666666,
      "eval_PRM NPV": 0.5806451612903226,
      "eval_PRM Precision": 0.9333333333333333,
      "eval_PRM Recall": 0.8433734939759037,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.34537962079048157,
      "eval_runtime": 14.7382,
      "eval_samples_per_second": 2.985,
      "eval_steps_per_second": 0.136,
      "step": 385
    },
    {
      "epoch": 0.09996439092292253,
      "grad_norm": 8.353583602250334,
      "learning_rate": 1.2467700258397934e-06,
      "loss": 0.4462,
      "step": 386
    },
    {
      "epoch": 0.10022336602894047,
      "grad_norm": 4.259457118742542,
      "learning_rate": 1.25e-06,
      "loss": 0.4462,
      "step": 387
    },
    {
      "epoch": 0.10048234113495841,
      "grad_norm": 6.045747029072608,
      "learning_rate": 1.2499997444416303e-06,
      "loss": 0.316,
      "step": 388
    },
    {
      "epoch": 0.10074131624097633,
      "grad_norm": 6.549109829669429,
      "learning_rate": 1.2499989777667297e-06,
      "loss": 0.3509,
      "step": 389
    },
    {
      "epoch": 0.10100029134699427,
      "grad_norm": 5.862650772896159,
      "learning_rate": 1.2499976999759252e-06,
      "loss": 0.4487,
      "step": 390
    },
    {
      "epoch": 0.10100029134699427,
      "eval_PRM Accuracy": 0.839622641509434,
      "eval_PRM F1": 0.89171974522293,
      "eval_PRM F1 AUC": 0.8347302252488213,
      "eval_PRM F1 AUC (fixed)": 0.6191723415400734,
      "eval_PRM F1 Neg": 0.6909090909090909,
      "eval_PRM NPV": 0.59375,
      "eval_PRM Precision": 0.9459459459459459,
      "eval_PRM Recall": 0.8433734939759037,
      "eval_PRM Specificty": 0.8260869565217391,
      "eval_loss": 0.3483627140522003,
      "eval_runtime": 14.609,
      "eval_samples_per_second": 3.012,
      "eval_steps_per_second": 0.137,
      "step": 390
    },
    {
      "epoch": 0.1012592664530122,
      "grad_norm": 4.633090184138419,
      "learning_rate": 1.249995911070262e-06,
      "loss": 0.4869,
      "step": 391
    },
    {
      "epoch": 0.10151824155903014,
      "grad_norm": 14.179051341269778,
      "learning_rate": 1.2499936110512028e-06,
      "loss": 0.4242,
      "step": 392
    },
    {
      "epoch": 0.10177721666504808,
      "grad_norm": 3.6200503143835547,
      "learning_rate": 1.2499907999206288e-06,
      "loss": 0.3389,
      "step": 393
    },
    {
      "epoch": 0.102036191771066,
      "grad_norm": 8.94604690308416,
      "learning_rate": 1.2499874776808384e-06,
      "loss": 0.4031,
      "step": 394
    },
    {
      "epoch": 0.10229516687708394,
      "grad_norm": 6.1031012803457365,
      "learning_rate": 1.2499836443345493e-06,
      "loss": 0.4491,
      "step": 395
    },
    {
      "epoch": 0.10229516687708394,
      "eval_PRM Accuracy": 0.8113207547169812,
      "eval_PRM F1": 0.8701298701298701,
      "eval_PRM F1 AUC": 0.8166579360921948,
      "eval_PRM F1 AUC (fixed)": 0.6191723415400733,
      "eval_PRM F1 Neg": 0.6551724137931034,
      "eval_PRM NPV": 0.5428571428571428,
      "eval_PRM Precision": 0.9436619718309859,
      "eval_PRM Recall": 0.8072289156626506,
      "eval_PRM Specificty": 0.8260869565217391,
      "eval_loss": 0.3896908760070801,
      "eval_runtime": 14.657,
      "eval_samples_per_second": 3.002,
      "eval_steps_per_second": 0.136,
      "step": 395
    },
    {
      "epoch": 0.10255414198310188,
      "grad_norm": 11.484707576135863,
      "learning_rate": 1.2499792998848956e-06,
      "loss": 0.3512,
      "step": 396
    },
    {
      "epoch": 0.1028131170891198,
      "grad_norm": 8.423723647150908,
      "learning_rate": 1.2499744443354305e-06,
      "loss": 0.5061,
      "step": 397
    },
    {
      "epoch": 0.10307209219513774,
      "grad_norm": 3.9133153276637427,
      "learning_rate": 1.2499690776901246e-06,
      "loss": 0.3537,
      "step": 398
    },
    {
      "epoch": 0.10333106730115568,
      "grad_norm": 5.0512139528828355,
      "learning_rate": 1.2499631999533668e-06,
      "loss": 0.5046,
      "step": 399
    },
    {
      "epoch": 0.10359004240717361,
      "grad_norm": 5.887848469989917,
      "learning_rate": 1.2499568111299641e-06,
      "loss": 0.5188,
      "step": 400
    },
    {
      "epoch": 0.10359004240717361,
      "eval_PRM Accuracy": 0.8301886792452831,
      "eval_PRM F1": 0.8860759493670886,
      "eval_PRM F1 AUC": 0.8129910948140389,
      "eval_PRM F1 AUC (fixed)": 0.6097433211105291,
      "eval_PRM F1 Neg": 0.6666666666666666,
      "eval_PRM NPV": 0.5806451612903226,
      "eval_PRM Precision": 0.9333333333333333,
      "eval_PRM Recall": 0.8433734939759037,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.37192729115486145,
      "eval_runtime": 14.852,
      "eval_samples_per_second": 2.963,
      "eval_steps_per_second": 0.135,
      "step": 400
    },
    {
      "epoch": 0.10384901751319155,
      "grad_norm": 6.188275540711092,
      "learning_rate": 1.2499499112251407e-06,
      "loss": 0.4381,
      "step": 401
    },
    {
      "epoch": 0.10410799261920947,
      "grad_norm": 4.422397712181315,
      "learning_rate": 1.2499425002445396e-06,
      "loss": 0.3981,
      "step": 402
    },
    {
      "epoch": 0.10436696772522741,
      "grad_norm": 3.848781758597374,
      "learning_rate": 1.2499345781942212e-06,
      "loss": 0.4092,
      "step": 403
    },
    {
      "epoch": 0.10462594283124535,
      "grad_norm": 11.553444118243934,
      "learning_rate": 1.2499261450806644e-06,
      "loss": 0.424,
      "step": 404
    },
    {
      "epoch": 0.10488491793726328,
      "grad_norm": 5.996576278633823,
      "learning_rate": 1.249917200910765e-06,
      "loss": 0.4393,
      "step": 405
    },
    {
      "epoch": 0.10488491793726328,
      "eval_PRM Accuracy": 0.8301886792452831,
      "eval_PRM F1": 0.8846153846153846,
      "eval_PRM F1 AUC": 0.8287061288632793,
      "eval_PRM F1 AUC (fixed)": 0.5955997904662127,
      "eval_PRM F1 Neg": 0.6785714285714286,
      "eval_PRM NPV": 0.5757575757575758,
      "eval_PRM Precision": 0.9452054794520548,
      "eval_PRM Recall": 0.8313253012048193,
      "eval_PRM Specificty": 0.8260869565217391,
      "eval_loss": 0.37841737270355225,
      "eval_runtime": 14.4713,
      "eval_samples_per_second": 3.041,
      "eval_steps_per_second": 0.138,
      "step": 405
    },
    {
      "epoch": 0.10514389304328121,
      "grad_norm": 9.190741668299102,
      "learning_rate": 1.249907745691838e-06,
      "loss": 0.4367,
      "step": 406
    },
    {
      "epoch": 0.10540286814929915,
      "grad_norm": 5.8752181526111515,
      "learning_rate": 1.2498977794316158e-06,
      "loss": 0.5232,
      "step": 407
    },
    {
      "epoch": 0.10566184325531708,
      "grad_norm": 4.1488194676706005,
      "learning_rate": 1.2498873021382483e-06,
      "loss": 0.4059,
      "step": 408
    },
    {
      "epoch": 0.10592081836133502,
      "grad_norm": 5.550763774298918,
      "learning_rate": 1.2498763138203038e-06,
      "loss": 0.4206,
      "step": 409
    },
    {
      "epoch": 0.10617979346735296,
      "grad_norm": 5.264681618753274,
      "learning_rate": 1.2498648144867686e-06,
      "loss": 0.3559,
      "step": 410
    },
    {
      "epoch": 0.10617979346735296,
      "eval_PRM Accuracy": 0.839622641509434,
      "eval_PRM F1": 0.8930817610062893,
      "eval_PRM F1 AUC": 0.819015191199581,
      "eval_PRM F1 AUC (fixed)": 0.5835515976951283,
      "eval_PRM F1 Neg": 0.6792452830188679,
      "eval_PRM NPV": 0.6,
      "eval_PRM Precision": 0.9342105263157895,
      "eval_PRM Recall": 0.8554216867469879,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.36815035343170166,
      "eval_runtime": 14.7219,
      "eval_samples_per_second": 2.989,
      "eval_steps_per_second": 0.136,
      "step": 410
    },
    {
      "epoch": 0.10643876857337088,
      "grad_norm": 13.688204458169853,
      "learning_rate": 1.2498528041470464e-06,
      "loss": 0.4144,
      "step": 411
    },
    {
      "epoch": 0.10669774367938882,
      "grad_norm": 8.91627110732376,
      "learning_rate": 1.2498402828109594e-06,
      "loss": 0.3403,
      "step": 412
    },
    {
      "epoch": 0.10695671878540676,
      "grad_norm": 8.443031410437765,
      "learning_rate": 1.2498272504887472e-06,
      "loss": 0.414,
      "step": 413
    },
    {
      "epoch": 0.10721569389142468,
      "grad_norm": 12.324495442513667,
      "learning_rate": 1.2498137071910677e-06,
      "loss": 0.4573,
      "step": 414
    },
    {
      "epoch": 0.10747466899744262,
      "grad_norm": 4.4998378067906994,
      "learning_rate": 1.249799652928996e-06,
      "loss": 0.3958,
      "step": 415
    },
    {
      "epoch": 0.10747466899744262,
      "eval_PRM Accuracy": 0.839622641509434,
      "eval_PRM F1": 0.89171974522293,
      "eval_PRM F1 AUC": 0.8347302252488213,
      "eval_PRM F1 AUC (fixed)": 0.5845992666317443,
      "eval_PRM F1 Neg": 0.6909090909090909,
      "eval_PRM NPV": 0.59375,
      "eval_PRM Precision": 0.9459459459459459,
      "eval_PRM Recall": 0.8433734939759037,
      "eval_PRM Specificty": 0.8260869565217391,
      "eval_loss": 0.36931851506233215,
      "eval_runtime": 14.5493,
      "eval_samples_per_second": 3.024,
      "eval_steps_per_second": 0.137,
      "step": 415
    },
    {
      "epoch": 0.10773364410346055,
      "grad_norm": 5.59070218598532,
      "learning_rate": 1.2497850877140258e-06,
      "loss": 0.4822,
      "step": 416
    },
    {
      "epoch": 0.10799261920947849,
      "grad_norm": 8.968937575600412,
      "learning_rate": 1.2497700115580683e-06,
      "loss": 0.483,
      "step": 417
    },
    {
      "epoch": 0.10825159431549643,
      "grad_norm": 5.078418619404218,
      "learning_rate": 1.2497544244734524e-06,
      "loss": 0.4509,
      "step": 418
    },
    {
      "epoch": 0.10851056942151435,
      "grad_norm": 12.05259697828282,
      "learning_rate": 1.2497383264729254e-06,
      "loss": 0.49,
      "step": 419
    },
    {
      "epoch": 0.10876954452753229,
      "grad_norm": 5.62697187532568,
      "learning_rate": 1.2497217175696517e-06,
      "loss": 0.5459,
      "step": 420
    },
    {
      "epoch": 0.10876954452753229,
      "eval_PRM Accuracy": 0.7735849056603774,
      "eval_PRM F1": 0.8378378378378378,
      "eval_PRM F1 AUC": 0.8082765845992665,
      "eval_PRM F1 AUC (fixed)": 0.5971712938711368,
      "eval_PRM F1 Neg": 0.625,
      "eval_PRM NPV": 0.4878048780487805,
      "eval_PRM Precision": 0.9538461538461539,
      "eval_PRM Recall": 0.7469879518072289,
      "eval_PRM Specificty": 0.8695652173913043,
      "eval_loss": 0.38043850660324097,
      "eval_runtime": 14.9479,
      "eval_samples_per_second": 2.944,
      "eval_steps_per_second": 0.134,
      "step": 420
    },
    {
      "epoch": 0.10902851963355023,
      "grad_norm": 9.10491260940364,
      "learning_rate": 1.2497045977772138e-06,
      "loss": 0.4919,
      "step": 421
    },
    {
      "epoch": 0.10928749473956816,
      "grad_norm": 9.647920546907573,
      "learning_rate": 1.2496869671096123e-06,
      "loss": 0.4094,
      "step": 422
    },
    {
      "epoch": 0.1095464698455861,
      "grad_norm": 10.094138514190103,
      "learning_rate": 1.2496688255812653e-06,
      "loss": 0.35,
      "step": 423
    },
    {
      "epoch": 0.10980544495160403,
      "grad_norm": 5.931036453890864,
      "learning_rate": 1.2496501732070086e-06,
      "loss": 0.3879,
      "step": 424
    },
    {
      "epoch": 0.11006442005762196,
      "grad_norm": 3.8019022867593213,
      "learning_rate": 1.2496310100020957e-06,
      "loss": 0.3735,
      "step": 425
    },
    {
      "epoch": 0.11006442005762196,
      "eval_PRM Accuracy": 0.8490566037735849,
      "eval_PRM F1": 0.9,
      "eval_PRM F1 AUC": 0.8250392875851232,
      "eval_PRM F1 AUC (fixed)": 0.6160293347302253,
      "eval_PRM F1 Neg": 0.6923076923076923,
      "eval_PRM NPV": 0.6206896551724138,
      "eval_PRM Precision": 0.935064935064935,
      "eval_PRM Recall": 0.8674698795180723,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.328229695558548,
      "eval_runtime": 14.6799,
      "eval_samples_per_second": 2.997,
      "eval_steps_per_second": 0.136,
      "step": 425
    },
    {
      "epoch": 0.1103233951636399,
      "grad_norm": 5.07041457856754,
      "learning_rate": 1.2496113359821982e-06,
      "loss": 0.3717,
      "step": 426
    },
    {
      "epoch": 0.11058237026965782,
      "grad_norm": 3.9324565423627926,
      "learning_rate": 1.2495911511634052e-06,
      "loss": 0.3909,
      "step": 427
    },
    {
      "epoch": 0.11084134537567576,
      "grad_norm": 4.646597382905398,
      "learning_rate": 1.2495704555622238e-06,
      "loss": 0.3355,
      "step": 428
    },
    {
      "epoch": 0.1111003204816937,
      "grad_norm": 9.385735386008209,
      "learning_rate": 1.2495492491955783e-06,
      "loss": 0.4172,
      "step": 429
    },
    {
      "epoch": 0.11135929558771163,
      "grad_norm": 4.973918058742094,
      "learning_rate": 1.2495275320808112e-06,
      "loss": 0.4354,
      "step": 430
    },
    {
      "epoch": 0.11135929558771163,
      "eval_PRM Accuracy": 0.839622641509434,
      "eval_PRM F1": 0.8957055214723927,
      "eval_PRM F1 AUC": 0.7875851231011,
      "eval_PRM F1 AUC (fixed)": 0.6296490309062337,
      "eval_PRM F1 Neg": 0.6530612244897959,
      "eval_PRM NPV": 0.6153846153846154,
      "eval_PRM Precision": 0.9125,
      "eval_PRM Recall": 0.8795180722891566,
      "eval_PRM Specificty": 0.6956521739130435,
      "eval_loss": 0.31378573179244995,
      "eval_runtime": 14.6402,
      "eval_samples_per_second": 3.005,
      "eval_steps_per_second": 0.137,
      "step": 430
    },
    {
      "epoch": 0.11161827069372957,
      "grad_norm": 17.05390807031748,
      "learning_rate": 1.2495053042356823e-06,
      "loss": 0.4862,
      "step": 431
    },
    {
      "epoch": 0.1118772457997475,
      "grad_norm": 7.012072282963652,
      "learning_rate": 1.2494825656783692e-06,
      "loss": 0.4364,
      "step": 432
    },
    {
      "epoch": 0.11213622090576543,
      "grad_norm": 3.7158958642428055,
      "learning_rate": 1.2494593164274673e-06,
      "loss": 0.3721,
      "step": 433
    },
    {
      "epoch": 0.11239519601178337,
      "grad_norm": 11.095370577712787,
      "learning_rate": 1.2494355565019897e-06,
      "loss": 0.4619,
      "step": 434
    },
    {
      "epoch": 0.11265417111780131,
      "grad_norm": 8.423672434963388,
      "learning_rate": 1.2494112859213667e-06,
      "loss": 0.4653,
      "step": 435
    },
    {
      "epoch": 0.11265417111780131,
      "eval_PRM Accuracy": 0.8207547169811321,
      "eval_PRM F1": 0.8789808917197452,
      "eval_PRM F1 AUC": 0.8069669984284967,
      "eval_PRM F1 AUC (fixed)": 0.6272917757988475,
      "eval_PRM F1 Neg": 0.6545454545454545,
      "eval_PRM NPV": 0.5625,
      "eval_PRM Precision": 0.9324324324324325,
      "eval_PRM Recall": 0.8313253012048193,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.33313944935798645,
      "eval_runtime": 14.7779,
      "eval_samples_per_second": 2.977,
      "eval_steps_per_second": 0.135,
      "step": 435
    },
    {
      "epoch": 0.11291314622381923,
      "grad_norm": 6.314062135089114,
      "learning_rate": 1.2493865047054465e-06,
      "loss": 0.3681,
      "step": 436
    },
    {
      "epoch": 0.11317212132983717,
      "grad_norm": 5.757644590025226,
      "learning_rate": 1.2493612128744948e-06,
      "loss": 0.5514,
      "step": 437
    },
    {
      "epoch": 0.1134310964358551,
      "grad_norm": 5.523908619118926,
      "learning_rate": 1.249335410449195e-06,
      "loss": 0.4419,
      "step": 438
    },
    {
      "epoch": 0.11369007154187304,
      "grad_norm": 5.943773174013052,
      "learning_rate": 1.2493090974506483e-06,
      "loss": 0.4776,
      "step": 439
    },
    {
      "epoch": 0.11394904664789098,
      "grad_norm": 20.29662028796124,
      "learning_rate": 1.2492822739003727e-06,
      "loss": 0.5249,
      "step": 440
    },
    {
      "epoch": 0.11394904664789098,
      "eval_PRM Accuracy": 0.7452830188679245,
      "eval_PRM F1": 0.8163265306122449,
      "eval_PRM F1 AUC": 0.7744892613933997,
      "eval_PRM F1 AUC (fixed)": 0.6115767417496071,
      "eval_PRM F1 Neg": 0.5846153846153846,
      "eval_PRM NPV": 0.4523809523809524,
      "eval_PRM Precision": 0.9375,
      "eval_PRM Recall": 0.7228915662650602,
      "eval_PRM Specificty": 0.8260869565217391,
      "eval_loss": 0.3871985375881195,
      "eval_runtime": 14.7587,
      "eval_samples_per_second": 2.981,
      "eval_steps_per_second": 0.136,
      "step": 440
    },
    {
      "epoch": 0.1142080217539089,
      "grad_norm": 8.993042445023589,
      "learning_rate": 1.2492549398203041e-06,
      "loss": 0.4704,
      "step": 441
    },
    {
      "epoch": 0.11446699685992684,
      "grad_norm": 12.199717818133639,
      "learning_rate": 1.2492270952327962e-06,
      "loss": 0.3264,
      "step": 442
    },
    {
      "epoch": 0.11472597196594478,
      "grad_norm": 8.978520690223112,
      "learning_rate": 1.2491987401606198e-06,
      "loss": 0.4283,
      "step": 443
    },
    {
      "epoch": 0.1149849470719627,
      "grad_norm": 3.809395479659997,
      "learning_rate": 1.2491698746269634e-06,
      "loss": 0.4611,
      "step": 444
    },
    {
      "epoch": 0.11524392217798064,
      "grad_norm": 3.3687122852485687,
      "learning_rate": 1.2491404986554328e-06,
      "loss": 0.385,
      "step": 445
    },
    {
      "epoch": 0.11524392217798064,
      "eval_PRM Accuracy": 0.839622641509434,
      "eval_PRM F1": 0.89171974522293,
      "eval_PRM F1 AUC": 0.8347302252488213,
      "eval_PRM F1 AUC (fixed)": 0.6162912519643792,
      "eval_PRM F1 Neg": 0.6909090909090909,
      "eval_PRM NPV": 0.59375,
      "eval_PRM Precision": 0.9459459459459459,
      "eval_PRM Recall": 0.8433734939759037,
      "eval_PRM Specificty": 0.8260869565217391,
      "eval_loss": 0.3389197587966919,
      "eval_runtime": 14.7519,
      "eval_samples_per_second": 2.983,
      "eval_steps_per_second": 0.136,
      "step": 445
    },
    {
      "epoch": 0.11550289728399858,
      "grad_norm": 5.513267646704666,
      "learning_rate": 1.2491106122700512e-06,
      "loss": 0.373,
      "step": 446
    },
    {
      "epoch": 0.11576187239001651,
      "grad_norm": 4.586155749253553,
      "learning_rate": 1.2490802154952595e-06,
      "loss": 0.3982,
      "step": 447
    },
    {
      "epoch": 0.11602084749603445,
      "grad_norm": 4.643300680941585,
      "learning_rate": 1.2490493083559156e-06,
      "loss": 0.3947,
      "step": 448
    },
    {
      "epoch": 0.11627982260205237,
      "grad_norm": 12.093335727107991,
      "learning_rate": 1.249017890877295e-06,
      "loss": 0.475,
      "step": 449
    },
    {
      "epoch": 0.11653879770807031,
      "grad_norm": 8.246043522729842,
      "learning_rate": 1.2489859630850904e-06,
      "loss": 0.4301,
      "step": 450
    },
    {
      "epoch": 0.11653879770807031,
      "eval_PRM Accuracy": 0.8490566037735849,
      "eval_PRM F1": 0.9012345679012346,
      "eval_PRM F1 AUC": 0.8093242535358826,
      "eval_PRM F1 AUC (fixed)": 0.6278156102671556,
      "eval_PRM F1 Neg": 0.68,
      "eval_PRM NPV": 0.6296296296296297,
      "eval_PRM Precision": 0.9240506329113924,
      "eval_PRM Recall": 0.8795180722891566,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.32794228196144104,
      "eval_runtime": 14.7823,
      "eval_samples_per_second": 2.977,
      "eval_steps_per_second": 0.135,
      "step": 450
    },
    {
      "epoch": 0.11679777281408825,
      "grad_norm": 6.077070805952264,
      "learning_rate": 1.2489535250054122e-06,
      "loss": 0.4336,
      "step": 451
    },
    {
      "epoch": 0.11705674792010617,
      "grad_norm": 4.370192483038455,
      "learning_rate": 1.2489205766647878e-06,
      "loss": 0.379,
      "step": 452
    },
    {
      "epoch": 0.11731572302612411,
      "grad_norm": 3.0466331122024153,
      "learning_rate": 1.2488871180901615e-06,
      "loss": 0.3485,
      "step": 453
    },
    {
      "epoch": 0.11757469813214205,
      "grad_norm": 3.505329612931744,
      "learning_rate": 1.2488531493088958e-06,
      "loss": 0.3886,
      "step": 454
    },
    {
      "epoch": 0.11783367323815998,
      "grad_norm": 4.757786960622849,
      "learning_rate": 1.2488186703487692e-06,
      "loss": 0.3205,
      "step": 455
    },
    {
      "epoch": 0.11783367323815998,
      "eval_PRM Accuracy": 0.8301886792452831,
      "eval_PRM F1": 0.8860759493670886,
      "eval_PRM F1 AUC": 0.8129910948140389,
      "eval_PRM F1 AUC (fixed)": 0.6220534311157674,
      "eval_PRM F1 Neg": 0.6666666666666666,
      "eval_PRM NPV": 0.5806451612903226,
      "eval_PRM Precision": 0.9333333333333333,
      "eval_PRM Recall": 0.8433734939759037,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.33780306577682495,
      "eval_runtime": 14.8292,
      "eval_samples_per_second": 2.967,
      "eval_steps_per_second": 0.135,
      "step": 455
    },
    {
      "epoch": 0.11809264834417792,
      "grad_norm": 4.590493377069403,
      "learning_rate": 1.248783681237979e-06,
      "loss": 0.3612,
      "step": 456
    },
    {
      "epoch": 0.11835162345019586,
      "grad_norm": 3.476272900504724,
      "learning_rate": 1.2487481820051384e-06,
      "loss": 0.382,
      "step": 457
    },
    {
      "epoch": 0.11861059855621378,
      "grad_norm": 3.1106107478564233,
      "learning_rate": 1.2487121726792784e-06,
      "loss": 0.3405,
      "step": 458
    },
    {
      "epoch": 0.11886957366223172,
      "grad_norm": 4.674172376087834,
      "learning_rate": 1.2486756532898465e-06,
      "loss": 0.3694,
      "step": 459
    },
    {
      "epoch": 0.11912854876824965,
      "grad_norm": 4.708991145064097,
      "learning_rate": 1.2486386238667083e-06,
      "loss": 0.3287,
      "step": 460
    },
    {
      "epoch": 0.11912854876824965,
      "eval_PRM Accuracy": 0.839622641509434,
      "eval_PRM F1": 0.8930817610062893,
      "eval_PRM F1 AUC": 0.819015191199581,
      "eval_PRM F1 AUC (fixed)": 0.63986380303824,
      "eval_PRM F1 Neg": 0.6792452830188679,
      "eval_PRM NPV": 0.6,
      "eval_PRM Precision": 0.9342105263157895,
      "eval_PRM Recall": 0.8554216867469879,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.32568544149398804,
      "eval_runtime": 14.6822,
      "eval_samples_per_second": 2.997,
      "eval_steps_per_second": 0.136,
      "step": 460
    },
    {
      "epoch": 0.11938752387426758,
      "grad_norm": 3.2665701190298364,
      "learning_rate": 1.2486010844401458e-06,
      "loss": 0.309,
      "step": 461
    },
    {
      "epoch": 0.11964649898028552,
      "grad_norm": 14.608377136123803,
      "learning_rate": 1.2485630350408577e-06,
      "loss": 0.4372,
      "step": 462
    },
    {
      "epoch": 0.11990547408630345,
      "grad_norm": 4.692121241159457,
      "learning_rate": 1.2485244756999615e-06,
      "loss": 0.3702,
      "step": 463
    },
    {
      "epoch": 0.12016444919232139,
      "grad_norm": 3.5397062939111357,
      "learning_rate": 1.2484854064489895e-06,
      "loss": 0.3609,
      "step": 464
    },
    {
      "epoch": 0.12042342429833933,
      "grad_norm": 3.4262687284743394,
      "learning_rate": 1.2484458273198923e-06,
      "loss": 0.3629,
      "step": 465
    },
    {
      "epoch": 0.12042342429833933,
      "eval_PRM Accuracy": 0.8207547169811321,
      "eval_PRM F1": 0.8789808917197452,
      "eval_PRM F1 AUC": 0.8069669984284967,
      "eval_PRM F1 AUC (fixed)": 0.6490309062336301,
      "eval_PRM F1 Neg": 0.6545454545454545,
      "eval_PRM NPV": 0.5625,
      "eval_PRM Precision": 0.9324324324324325,
      "eval_PRM Recall": 0.8313253012048193,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.3342059850692749,
      "eval_runtime": 14.3984,
      "eval_samples_per_second": 3.056,
      "eval_steps_per_second": 0.139,
      "step": 465
    },
    {
      "epoch": 0.12068239940435725,
      "grad_norm": 6.765235304929425,
      "learning_rate": 1.2484057383450372e-06,
      "loss": 0.5116,
      "step": 466
    },
    {
      "epoch": 0.12094137451037519,
      "grad_norm": 4.3624619071051765,
      "learning_rate": 1.2483651395572086e-06,
      "loss": 0.4132,
      "step": 467
    },
    {
      "epoch": 0.12120034961639313,
      "grad_norm": 8.833253754878113,
      "learning_rate": 1.2483240309896073e-06,
      "loss": 0.4045,
      "step": 468
    },
    {
      "epoch": 0.12145932472241106,
      "grad_norm": 4.98960724246876,
      "learning_rate": 1.2482824126758517e-06,
      "loss": 0.3841,
      "step": 469
    },
    {
      "epoch": 0.121718299828429,
      "grad_norm": 4.3471531348502275,
      "learning_rate": 1.2482402846499764e-06,
      "loss": 0.3716,
      "step": 470
    },
    {
      "epoch": 0.121718299828429,
      "eval_PRM Accuracy": 0.8301886792452831,
      "eval_PRM F1": 0.8860759493670886,
      "eval_PRM F1 AUC": 0.8129910948140389,
      "eval_PRM F1 AUC (fixed)": 0.6485070717653222,
      "eval_PRM F1 Neg": 0.6666666666666666,
      "eval_PRM NPV": 0.5806451612903226,
      "eval_PRM Precision": 0.9333333333333333,
      "eval_PRM Recall": 0.8433734939759037,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.32934725284576416,
      "eval_runtime": 14.705,
      "eval_samples_per_second": 2.992,
      "eval_steps_per_second": 0.136,
      "step": 470
    },
    {
      "epoch": 0.12197727493444693,
      "grad_norm": 4.861294081639487,
      "learning_rate": 1.2481976469464334e-06,
      "loss": 0.4188,
      "step": 471
    },
    {
      "epoch": 0.12223625004046486,
      "grad_norm": 7.2929021973775345,
      "learning_rate": 1.2481544996000913e-06,
      "loss": 0.3866,
      "step": 472
    },
    {
      "epoch": 0.1224952251464828,
      "grad_norm": 5.221068095367759,
      "learning_rate": 1.2481108426462352e-06,
      "loss": 0.3554,
      "step": 473
    },
    {
      "epoch": 0.12275420025250072,
      "grad_norm": 3.392837388495582,
      "learning_rate": 1.248066676120567e-06,
      "loss": 0.3725,
      "step": 474
    },
    {
      "epoch": 0.12301317535851866,
      "grad_norm": 3.7543275834934877,
      "learning_rate": 1.248022000059206e-06,
      "loss": 0.3458,
      "step": 475
    },
    {
      "epoch": 0.12301317535851866,
      "eval_PRM Accuracy": 0.839622641509434,
      "eval_PRM F1": 0.8944099378881988,
      "eval_PRM F1 AUC": 0.8033001571503404,
      "eval_PRM F1 AUC (fixed)": 0.6437925615505501,
      "eval_PRM F1 Neg": 0.6666666666666666,
      "eval_PRM NPV": 0.6071428571428571,
      "eval_PRM Precision": 0.9230769230769231,
      "eval_PRM Recall": 0.8674698795180723,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.31160077452659607,
      "eval_runtime": 14.6465,
      "eval_samples_per_second": 3.004,
      "eval_steps_per_second": 0.137,
      "step": 475
    },
    {
      "epoch": 0.1232721504645366,
      "grad_norm": 21.593384607104042,
      "learning_rate": 1.2479768144986874e-06,
      "loss": 0.564,
      "step": 476
    },
    {
      "epoch": 0.12353112557055453,
      "grad_norm": 6.930229379845048,
      "learning_rate": 1.2479311194759636e-06,
      "loss": 0.4011,
      "step": 477
    },
    {
      "epoch": 0.12379010067657246,
      "grad_norm": 5.781015235162338,
      "learning_rate": 1.2478849150284029e-06,
      "loss": 0.2612,
      "step": 478
    },
    {
      "epoch": 0.1240490757825904,
      "grad_norm": 7.084988924381968,
      "learning_rate": 1.247838201193791e-06,
      "loss": 0.435,
      "step": 479
    },
    {
      "epoch": 0.12430805088860833,
      "grad_norm": 8.18854984265065,
      "learning_rate": 1.24779097801033e-06,
      "loss": 0.4731,
      "step": 480
    },
    {
      "epoch": 0.12430805088860833,
      "eval_PRM Accuracy": 0.839622641509434,
      "eval_PRM F1": 0.8930817610062893,
      "eval_PRM F1 AUC": 0.819015191199581,
      "eval_PRM F1 AUC (fixed)": 0.6348873755893137,
      "eval_PRM F1 Neg": 0.6792452830188679,
      "eval_PRM NPV": 0.6,
      "eval_PRM Precision": 0.9342105263157895,
      "eval_PRM Recall": 0.8554216867469879,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.3187015652656555,
      "eval_runtime": 14.4311,
      "eval_samples_per_second": 3.049,
      "eval_steps_per_second": 0.139,
      "step": 480
    },
    {
      "epoch": 0.12456702599462627,
      "grad_norm": 4.355038554690252,
      "learning_rate": 1.2477432455166382e-06,
      "loss": 0.3797,
      "step": 481
    },
    {
      "epoch": 0.12482600110064421,
      "grad_norm": 5.838501143643666,
      "learning_rate": 1.2476950037517505e-06,
      "loss": 0.4502,
      "step": 482
    },
    {
      "epoch": 0.12508497620666215,
      "grad_norm": 4.824847173062183,
      "learning_rate": 1.2476462527551186e-06,
      "loss": 0.4074,
      "step": 483
    },
    {
      "epoch": 0.12534395131268006,
      "grad_norm": 10.703781791565072,
      "learning_rate": 1.2475969925666104e-06,
      "loss": 0.3866,
      "step": 484
    },
    {
      "epoch": 0.125602926418698,
      "grad_norm": 11.792915909373495,
      "learning_rate": 1.2475472232265102e-06,
      "loss": 0.4068,
      "step": 485
    },
    {
      "epoch": 0.125602926418698,
      "eval_PRM Accuracy": 0.8207547169811321,
      "eval_PRM F1": 0.8758169934640523,
      "eval_PRM F1 AUC": 0.8383970665269774,
      "eval_PRM F1 AUC (fixed)": 0.6272917757988475,
      "eval_PRM F1 Neg": 0.6779661016949152,
      "eval_PRM NPV": 0.5555555555555556,
      "eval_PRM Precision": 0.9571428571428572,
      "eval_PRM Recall": 0.8072289156626506,
      "eval_PRM Specificty": 0.8695652173913043,
      "eval_loss": 0.34957724809646606,
      "eval_runtime": 14.9385,
      "eval_samples_per_second": 2.945,
      "eval_steps_per_second": 0.134,
      "step": 485
    },
    {
      "epoch": 0.12586190152471594,
      "grad_norm": 12.342288855982682,
      "learning_rate": 1.2474969447755184e-06,
      "loss": 0.4546,
      "step": 486
    },
    {
      "epoch": 0.12612087663073387,
      "grad_norm": 7.4421966783654865,
      "learning_rate": 1.2474461572547525e-06,
      "loss": 0.4749,
      "step": 487
    },
    {
      "epoch": 0.1263798517367518,
      "grad_norm": 3.933100369825377,
      "learning_rate": 1.2473948607057456e-06,
      "loss": 0.3958,
      "step": 488
    },
    {
      "epoch": 0.12663882684276973,
      "grad_norm": 7.290452928627164,
      "learning_rate": 1.2473430551704477e-06,
      "loss": 0.4805,
      "step": 489
    },
    {
      "epoch": 0.12689780194878766,
      "grad_norm": 3.4175667460178616,
      "learning_rate": 1.2472907406912242e-06,
      "loss": 0.3287,
      "step": 490
    },
    {
      "epoch": 0.12689780194878766,
      "eval_PRM Accuracy": 0.839622641509434,
      "eval_PRM F1": 0.8930817610062893,
      "eval_PRM F1 AUC": 0.819015191199581,
      "eval_PRM F1 AUC (fixed)": 0.6346254583551597,
      "eval_PRM F1 Neg": 0.6792452830188679,
      "eval_PRM NPV": 0.6,
      "eval_PRM Precision": 0.9342105263157895,
      "eval_PRM Recall": 0.8554216867469879,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.3145117461681366,
      "eval_runtime": 14.5492,
      "eval_samples_per_second": 3.024,
      "eval_steps_per_second": 0.137,
      "step": 490
    },
    {
      "epoch": 0.1271567770548056,
      "grad_norm": 4.335883611097396,
      "learning_rate": 1.2472379173108574e-06,
      "loss": 0.2983,
      "step": 491
    },
    {
      "epoch": 0.12741575216082354,
      "grad_norm": 6.626921553086188,
      "learning_rate": 1.2471845850725456e-06,
      "loss": 0.4223,
      "step": 492
    },
    {
      "epoch": 0.12767472726684148,
      "grad_norm": 3.180075529152741,
      "learning_rate": 1.247130744019903e-06,
      "loss": 0.3306,
      "step": 493
    },
    {
      "epoch": 0.12793370237285942,
      "grad_norm": 8.566289040757376,
      "learning_rate": 1.2470763941969604e-06,
      "loss": 0.4057,
      "step": 494
    },
    {
      "epoch": 0.12819267747887733,
      "grad_norm": 19.349772589892186,
      "learning_rate": 1.2470215356481644e-06,
      "loss": 0.553,
      "step": 495
    },
    {
      "epoch": 0.12819267747887733,
      "eval_PRM Accuracy": 0.8490566037735849,
      "eval_PRM F1": 0.9012345679012346,
      "eval_PRM F1 AUC": 0.8093242535358826,
      "eval_PRM F1 AUC (fixed)": 0.6369827134625459,
      "eval_PRM F1 Neg": 0.68,
      "eval_PRM NPV": 0.6296296296296297,
      "eval_PRM Precision": 0.9240506329113924,
      "eval_PRM Recall": 0.8795180722891566,
      "eval_PRM Specificty": 0.7391304347826086,
      "eval_loss": 0.3201056122779846,
      "eval_runtime": 14.5466,
      "eval_samples_per_second": 3.025,
      "eval_steps_per_second": 0.137,
      "step": 495
    },
    {
      "epoch": 0.12845165258489527,
      "grad_norm": 3.4846284567808343,
      "learning_rate": 1.246966168418377e-06,
      "loss": 0.3323,
      "step": 496
    },
    {
      "epoch": 0.1287106276909132,
      "grad_norm": 3.0603555272843113,
      "learning_rate": 1.2469102925528776e-06,
      "loss": 0.3422,
      "step": 497
    },
    {
      "epoch": 0.12896960279693115,
      "grad_norm": 4.009331855682547,
      "learning_rate": 1.2468539080973603e-06,
      "loss": 0.3721,
      "step": 498
    },
    {
      "epoch": 0.1292285779029491,
      "grad_norm": 6.272480918620614,
      "learning_rate": 1.2467970150979356e-06,
      "loss": 0.427,
      "step": 499
    },
    {
      "epoch": 0.12948755300896703,
      "grad_norm": 13.427629672952166,
      "learning_rate": 1.2467396136011298e-06,
      "loss": 0.4693,
      "step": 500
    },
    {
      "epoch": 0.12948755300896703,
      "eval_PRM Accuracy": 0.8301886792452831,
      "eval_PRM F1": 0.8860759493670886,
      "eval_PRM F1 AUC": 0.8129910948140389,
      "eval_PRM F1 AUC (fixed)": 0.6327920377160817,
      "eval_PRM F1 Neg": 0.6666666666666666,
      "eval_PRM NPV": 0.5806451612903226,
      "eval_PRM Precision": 0.9333333333333333,
      "eval_PRM Recall": 0.8433734939759037,
      "eval_PRM Specificty": 0.782608695652174,
      "eval_loss": 0.33357107639312744,
      "eval_runtime": 14.4212,
      "eval_samples_per_second": 3.051,
      "eval_steps_per_second": 0.139,
      "step": 500
    }
  ],
  "logging_steps": 1,
  "max_steps": 3861,
  "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": 42279677911040.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}