{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.388462659026901, "eval_steps": 5, "global_step": 1500, "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 }, { "epoch": 0.12974652811498494, "grad_norm": 5.434972663555558, "learning_rate": 1.246681703653885e-06, "loss": 0.3905, "step": 501 }, { "epoch": 0.13000550322100288, "grad_norm": 4.58952122757678, "learning_rate": 1.2466232853035596e-06, "loss": 0.5206, "step": 502 }, { "epoch": 0.13026447832702082, "grad_norm": 7.372300675268678, "learning_rate": 1.246564358597927e-06, "loss": 0.4235, "step": 503 }, { "epoch": 0.13052345343303876, "grad_norm": 4.142484155230843, "learning_rate": 1.2465049235851768e-06, "loss": 0.4143, "step": 504 }, { "epoch": 0.1307824285390567, "grad_norm": 10.4043112649764, "learning_rate": 1.2464449803139142e-06, "loss": 0.4165, "step": 505 }, { "epoch": 0.1307824285390567, "eval_PRM Accuracy": 0.8018867924528302, "eval_PRM F1": 0.8609271523178808, "eval_PRM F1 AUC": 0.826348873755893, "eval_PRM F1 AUC (fixed)": 0.6356731272917758, "eval_PRM F1 Neg": 0.6557377049180327, "eval_PRM NPV": 0.5263157894736842, "eval_PRM Precision": 0.9558823529411765, "eval_PRM Recall": 0.7831325301204819, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.3677997589111328, "eval_runtime": 15.0664, "eval_samples_per_second": 2.92, "eval_steps_per_second": 0.133, "step": 505 }, { "epoch": 0.1310414036450746, "grad_norm": 9.922312215454863, "learning_rate": 1.2463845288331596e-06, "loss": 0.3574, "step": 506 }, { "epoch": 0.13130037875109254, "grad_norm": 6.670265269248358, "learning_rate": 1.2463235691923503e-06, "loss": 0.4276, "step": 507 }, { "epoch": 0.13155935385711048, "grad_norm": 7.307285016571106, "learning_rate": 1.2462621014413373e-06, "loss": 0.3718, "step": 508 }, { "epoch": 0.13181832896312842, "grad_norm": 7.0173057930362095, "learning_rate": 1.2462001256303888e-06, "loss": 0.3169, "step": 509 }, { "epoch": 0.13207730406914636, "grad_norm": 5.3637100732393135, "learning_rate": 1.2461376418101877e-06, "loss": 0.3718, "step": 510 }, { "epoch": 0.13207730406914636, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.89171974522293, "eval_PRM F1 AUC": 0.8347302252488213, "eval_PRM F1 AUC (fixed)": 0.644578313253012, "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": 0.323575884103775, "eval_runtime": 14.7572, "eval_samples_per_second": 2.982, "eval_steps_per_second": 0.136, "step": 510 }, { "epoch": 0.1323362791751643, "grad_norm": 3.949656761693001, "learning_rate": 1.246074650031832e-06, "loss": 0.3234, "step": 511 }, { "epoch": 0.1325952542811822, "grad_norm": 4.248192544379391, "learning_rate": 1.2460111503468363e-06, "loss": 0.4262, "step": 512 }, { "epoch": 0.13285422938720015, "grad_norm": 3.891048084630212, "learning_rate": 1.245947142807129e-06, "loss": 0.3222, "step": 513 }, { "epoch": 0.1331132044932181, "grad_norm": 4.148693831751682, "learning_rate": 1.2458826274650552e-06, "loss": 0.3477, "step": 514 }, { "epoch": 0.13337217959923603, "grad_norm": 6.540096957578557, "learning_rate": 1.2458176043733744e-06, "loss": 0.3139, "step": 515 }, { "epoch": 0.13337217959923603, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8930817610062893, "eval_PRM F1 AUC": 0.819015191199581, "eval_PRM F1 AUC (fixed)": 0.6547930853850183, "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": 0.3057812452316284, "eval_runtime": 14.6202, "eval_samples_per_second": 3.01, "eval_steps_per_second": 0.137, "step": 515 }, { "epoch": 0.13363115470525397, "grad_norm": 4.703165811412571, "learning_rate": 1.2457520735852618e-06, "loss": 0.4059, "step": 516 }, { "epoch": 0.13389012981127188, "grad_norm": 6.028053885312962, "learning_rate": 1.2456860351543074e-06, "loss": 0.4418, "step": 517 }, { "epoch": 0.13414910491728982, "grad_norm": 3.816982454189665, "learning_rate": 1.2456194891345167e-06, "loss": 0.3853, "step": 518 }, { "epoch": 0.13440808002330776, "grad_norm": 4.1845349649740395, "learning_rate": 1.2455524355803102e-06, "loss": 0.3931, "step": 519 }, { "epoch": 0.1346670551293257, "grad_norm": 5.069175129972723, "learning_rate": 1.2454848745465234e-06, "loss": 0.3706, "step": 520 }, { "epoch": 0.1346670551293257, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.8974358974358975, "eval_PRM F1 AUC": 0.8564693556836039, "eval_PRM F1 AUC (fixed)": 0.6519119958093242, "eval_PRM F1 Neg": 0.7142857142857143, "eval_PRM NPV": 0.6060606060606061, "eval_PRM Precision": 0.958904109589041, "eval_PRM Recall": 0.8433734939759037, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.31878462433815, "eval_runtime": 14.9989, "eval_samples_per_second": 2.934, "eval_steps_per_second": 0.133, "step": 520 }, { "epoch": 0.13492603023534364, "grad_norm": 5.554389828528387, "learning_rate": 1.2454168060884065e-06, "loss": 0.3495, "step": 521 }, { "epoch": 0.13518500534136157, "grad_norm": 4.269474285741495, "learning_rate": 1.2453482302616254e-06, "loss": 0.351, "step": 522 }, { "epoch": 0.1354439804473795, "grad_norm": 4.285143074157683, "learning_rate": 1.2452791471222605e-06, "loss": 0.4821, "step": 523 }, { "epoch": 0.13570295555339743, "grad_norm": 4.800867083848419, "learning_rate": 1.2452095567268069e-06, "loss": 0.3048, "step": 524 }, { "epoch": 0.13596193065941536, "grad_norm": 5.37470151118888, "learning_rate": 1.2451394591321747e-06, "loss": 0.4761, "step": 525 }, { "epoch": 0.13596193065941536, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8930817610062893, "eval_PRM F1 AUC": 0.819015191199581, "eval_PRM F1 AUC (fixed)": 0.6521739130434783, "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": 0.3053739368915558, "eval_runtime": 14.8675, "eval_samples_per_second": 2.959, "eval_steps_per_second": 0.135, "step": 525 }, { "epoch": 0.1362209057654333, "grad_norm": 4.76701924346624, "learning_rate": 1.2450688543956887e-06, "loss": 0.416, "step": 526 }, { "epoch": 0.13647988087145124, "grad_norm": 4.626643528344459, "learning_rate": 1.2449977425750886e-06, "loss": 0.3598, "step": 527 }, { "epoch": 0.13673885597746915, "grad_norm": 8.703126246641308, "learning_rate": 1.244926123728529e-06, "loss": 0.3911, "step": 528 }, { "epoch": 0.1369978310834871, "grad_norm": 15.454895749771413, "learning_rate": 1.2448539979145786e-06, "loss": 0.4538, "step": 529 }, { "epoch": 0.13725680618950503, "grad_norm": 3.4674946395586135, "learning_rate": 1.2447813651922208e-06, "loss": 0.34, "step": 530 }, { "epoch": 0.13725680618950503, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8944099378881988, "eval_PRM F1 AUC": 0.8033001571503404, "eval_PRM F1 AUC (fixed)": 0.6511262441068622, "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": 0.3079970180988312, "eval_runtime": 14.6555, "eval_samples_per_second": 3.002, "eval_steps_per_second": 0.136, "step": 530 }, { "epoch": 0.13751578129552297, "grad_norm": 9.901461765314394, "learning_rate": 1.2447082256208538e-06, "loss": 0.4309, "step": 531 }, { "epoch": 0.1377747564015409, "grad_norm": 8.819595922470516, "learning_rate": 1.2446345792602901e-06, "loss": 0.3939, "step": 532 }, { "epoch": 0.13803373150755885, "grad_norm": 7.240407887637497, "learning_rate": 1.244560426170757e-06, "loss": 0.426, "step": 533 }, { "epoch": 0.13829270661357676, "grad_norm": 3.304557932787079, "learning_rate": 1.2444857664128957e-06, "loss": 0.3499, "step": 534 }, { "epoch": 0.1385516817195947, "grad_norm": 4.779209710817938, "learning_rate": 1.2444106000477619e-06, "loss": 0.3743, "step": 535 }, { "epoch": 0.1385516817195947, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.89171974522293, "eval_PRM F1 AUC": 0.8347302252488213, "eval_PRM F1 AUC (fixed)": 0.6550550026191724, "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": 0.331030011177063, "eval_runtime": 14.3834, "eval_samples_per_second": 3.059, "eval_steps_per_second": 0.139, "step": 535 }, { "epoch": 0.13881065682561264, "grad_norm": 5.984315015375788, "learning_rate": 1.2443349271368257e-06, "loss": 0.2657, "step": 536 }, { "epoch": 0.13906963193163058, "grad_norm": 8.492618114157318, "learning_rate": 1.2442587477419717e-06, "loss": 0.3658, "step": 537 }, { "epoch": 0.13932860703764852, "grad_norm": 9.882539638982529, "learning_rate": 1.244182061925498e-06, "loss": 0.4839, "step": 538 }, { "epoch": 0.13958758214366643, "grad_norm": 4.006257788106991, "learning_rate": 1.2441048697501174e-06, "loss": 0.4156, "step": 539 }, { "epoch": 0.13984655724968437, "grad_norm": 6.065415435767625, "learning_rate": 1.2440271712789568e-06, "loss": 0.4337, "step": 540 }, { "epoch": 0.13984655724968437, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.8987341772151899, "eval_PRM F1 AUC": 0.8407543216343635, "eval_PRM F1 AUC (fixed)": 0.6665793609219487, "eval_PRM F1 Neg": 0.7037037037037037, "eval_PRM NPV": 0.6129032258064516, "eval_PRM Precision": 0.9466666666666667, "eval_PRM Recall": 0.8554216867469879, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.31409063935279846, "eval_runtime": 14.4095, "eval_samples_per_second": 3.054, "eval_steps_per_second": 0.139, "step": 540 }, { "epoch": 0.1401055323557023, "grad_norm": 8.055749515946466, "learning_rate": 1.2439489665755566e-06, "loss": 0.4103, "step": 541 }, { "epoch": 0.14036450746172024, "grad_norm": 3.159244262365573, "learning_rate": 1.243870255703872e-06, "loss": 0.2871, "step": 542 }, { "epoch": 0.14062348256773818, "grad_norm": 3.3484268018384755, "learning_rate": 1.2437910387282715e-06, "loss": 0.3617, "step": 543 }, { "epoch": 0.14088245767375612, "grad_norm": 8.017493622586152, "learning_rate": 1.2437113157135375e-06, "loss": 0.344, "step": 544 }, { "epoch": 0.14114143277977403, "grad_norm": 8.661634985456647, "learning_rate": 1.2436310867248668e-06, "loss": 0.4113, "step": 545 }, { "epoch": 0.14114143277977403, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.8987341772151899, "eval_PRM F1 AUC": 0.8407543216343635, "eval_PRM F1 AUC (fixed)": 0.6678889470927187, "eval_PRM F1 Neg": 0.7037037037037037, "eval_PRM NPV": 0.6129032258064516, "eval_PRM Precision": 0.9466666666666667, "eval_PRM Recall": 0.8554216867469879, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.31337836384773254, "eval_runtime": 14.6079, "eval_samples_per_second": 3.012, "eval_steps_per_second": 0.137, "step": 545 }, { "epoch": 0.14140040788579197, "grad_norm": 4.4569376746391445, "learning_rate": 1.2435503518278695e-06, "loss": 0.3847, "step": 546 }, { "epoch": 0.1416593829918099, "grad_norm": 3.5855721585468143, "learning_rate": 1.2434691110885695e-06, "loss": 0.3407, "step": 547 }, { "epoch": 0.14191835809782785, "grad_norm": 6.692845918255269, "learning_rate": 1.2433873645734042e-06, "loss": 0.5017, "step": 548 }, { "epoch": 0.1421773332038458, "grad_norm": 4.714823671241315, "learning_rate": 1.243305112349225e-06, "loss": 0.377, "step": 549 }, { "epoch": 0.1424363083098637, "grad_norm": 10.08925546130851, "learning_rate": 1.2432223544832969e-06, "loss": 0.4127, "step": 550 }, { "epoch": 0.1424363083098637, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.9, "eval_PRM F1 AUC": 0.8250392875851232, "eval_PRM F1 AUC (fixed)": 0.6571503404924044, "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": 0.30864760279655457, "eval_runtime": 14.6728, "eval_samples_per_second": 2.999, "eval_steps_per_second": 0.136, "step": 550 }, { "epoch": 0.14269528341588164, "grad_norm": 9.255856783180022, "learning_rate": 1.2431390910432978e-06, "loss": 0.4187, "step": 551 }, { "epoch": 0.14295425852189958, "grad_norm": 3.203425763387509, "learning_rate": 1.2430553220973196e-06, "loss": 0.344, "step": 552 }, { "epoch": 0.14321323362791752, "grad_norm": 7.163800749072193, "learning_rate": 1.2429710477138676e-06, "loss": 0.3816, "step": 553 }, { "epoch": 0.14347220873393546, "grad_norm": 3.4812956449664845, "learning_rate": 1.24288626796186e-06, "loss": 0.3907, "step": 554 }, { "epoch": 0.1437311838399534, "grad_norm": 7.383612072825209, "learning_rate": 1.2428009829106288e-06, "loss": 0.438, "step": 555 }, { "epoch": 0.1437311838399534, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.8974358974358975, "eval_PRM F1 AUC": 0.8564693556836039, "eval_PRM F1 AUC (fixed)": 0.6568884232582504, "eval_PRM F1 Neg": 0.7142857142857143, "eval_PRM NPV": 0.6060606060606061, "eval_PRM Precision": 0.958904109589041, "eval_PRM Recall": 0.8433734939759037, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.3244212865829468, "eval_runtime": 14.2781, "eval_samples_per_second": 3.082, "eval_steps_per_second": 0.14, "step": 555 }, { "epoch": 0.1439901589459713, "grad_norm": 5.631807558276998, "learning_rate": 1.2427151926299187e-06, "loss": 0.3592, "step": 556 }, { "epoch": 0.14424913405198925, "grad_norm": 5.0220728120088145, "learning_rate": 1.2426288971898881e-06, "loss": 0.4078, "step": 557 }, { "epoch": 0.1445081091580072, "grad_norm": 9.058383349897746, "learning_rate": 1.2425420966611081e-06, "loss": 0.3206, "step": 558 }, { "epoch": 0.14476708426402513, "grad_norm": 7.986430368212883, "learning_rate": 1.2424547911145635e-06, "loss": 0.3563, "step": 559 }, { "epoch": 0.14502605937004306, "grad_norm": 3.5609507470899473, "learning_rate": 1.2423669806216509e-06, "loss": 0.4017, "step": 560 }, { "epoch": 0.14502605937004306, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.89171974522293, "eval_PRM F1 AUC": 0.8347302252488213, "eval_PRM F1 AUC (fixed)": 0.6705081194342588, "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": 0.31759682297706604, "eval_runtime": 14.5424, "eval_samples_per_second": 3.026, "eval_steps_per_second": 0.138, "step": 560 }, { "epoch": 0.14528503447606098, "grad_norm": 9.331894617265462, "learning_rate": 1.242278665254181e-06, "loss": 0.3495, "step": 561 }, { "epoch": 0.14554400958207891, "grad_norm": 4.2029122364941855, "learning_rate": 1.2421898450843768e-06, "loss": 0.4254, "step": 562 }, { "epoch": 0.14580298468809685, "grad_norm": 3.7627145658128334, "learning_rate": 1.2421005201848744e-06, "loss": 0.3512, "step": 563 }, { "epoch": 0.1460619597941148, "grad_norm": 7.780938416484376, "learning_rate": 1.2420106906287224e-06, "loss": 0.3139, "step": 564 }, { "epoch": 0.14632093490013273, "grad_norm": 5.945589615268719, "learning_rate": 1.2419203564893822e-06, "loss": 0.4031, "step": 565 }, { "epoch": 0.14632093490013273, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.9079754601226994, "eval_PRM F1 AUC": 0.8153483499214248, "eval_PRM F1 AUC (fixed)": 0.6846516500785752, "eval_PRM F1 Neg": 0.6938775510204082, "eval_PRM NPV": 0.6538461538461539, "eval_PRM Precision": 0.925, "eval_PRM Recall": 0.891566265060241, "eval_PRM Specificty": 0.7391304347826086, "eval_loss": 0.300191730260849, "eval_runtime": 14.3339, "eval_samples_per_second": 3.07, "eval_steps_per_second": 0.14, "step": 565 }, { "epoch": 0.14657991000615067, "grad_norm": 6.2263899172585955, "learning_rate": 1.241829517840728e-06, "loss": 0.3507, "step": 566 }, { "epoch": 0.14683888511216858, "grad_norm": 6.387650973352577, "learning_rate": 1.2417381747570464e-06, "loss": 0.3567, "step": 567 }, { "epoch": 0.14709786021818652, "grad_norm": 6.671960513858668, "learning_rate": 1.2416463273130364e-06, "loss": 0.3629, "step": 568 }, { "epoch": 0.14735683532420446, "grad_norm": 7.090385235466389, "learning_rate": 1.24155397558381e-06, "loss": 0.4472, "step": 569 }, { "epoch": 0.1476158104302224, "grad_norm": 5.756477153036491, "learning_rate": 1.2414611196448906e-06, "loss": 0.4065, "step": 570 }, { "epoch": 0.1476158104302224, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.89171974522293, "eval_PRM F1 AUC": 0.8347302252488213, "eval_PRM F1 AUC (fixed)": 0.680460974332111, "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": 0.3101547360420227, "eval_runtime": 14.3796, "eval_samples_per_second": 3.06, "eval_steps_per_second": 0.139, "step": 570 }, { "epoch": 0.14787478553624034, "grad_norm": 9.003905398536311, "learning_rate": 1.2413677595722154e-06, "loss": 0.3899, "step": 571 }, { "epoch": 0.14813376064225825, "grad_norm": 4.523238842150813, "learning_rate": 1.241273895442132e-06, "loss": 0.3687, "step": 572 }, { "epoch": 0.1483927357482762, "grad_norm": 4.1866996609752904, "learning_rate": 1.2411795273314025e-06, "loss": 0.33, "step": 573 }, { "epoch": 0.14865171085429413, "grad_norm": 8.369351825162653, "learning_rate": 1.2410846553171988e-06, "loss": 0.4078, "step": 574 }, { "epoch": 0.14891068596031207, "grad_norm": 9.110580389178104, "learning_rate": 1.2409892794771066e-06, "loss": 0.4616, "step": 575 }, { "epoch": 0.14891068596031207, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8846153846153846, "eval_PRM F1 AUC": 0.8287061288632793, "eval_PRM F1 AUC (fixed)": 0.6781037192247249, "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": 0.3210476040840149, "eval_runtime": 14.8749, "eval_samples_per_second": 2.958, "eval_steps_per_second": 0.134, "step": 575 }, { "epoch": 0.14916966106633, "grad_norm": 10.564572190452354, "learning_rate": 1.2408933998891226e-06, "loss": 0.3938, "step": 576 }, { "epoch": 0.14942863617234795, "grad_norm": 15.194471093181686, "learning_rate": 1.2407970166316562e-06, "loss": 0.4729, "step": 577 }, { "epoch": 0.14968761127836586, "grad_norm": 9.4388656054538, "learning_rate": 1.2407001297835281e-06, "loss": 0.4167, "step": 578 }, { "epoch": 0.1499465863843838, "grad_norm": 3.577291285694216, "learning_rate": 1.2406027394239712e-06, "loss": 0.3604, "step": 579 }, { "epoch": 0.15020556149040173, "grad_norm": 6.76903188686745, "learning_rate": 1.2405048456326302e-06, "loss": 0.3794, "step": 580 }, { "epoch": 0.15020556149040173, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.9012345679012346, "eval_PRM F1 AUC": 0.8093242535358826, "eval_PRM F1 AUC (fixed)": 0.6901519119958093, "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": 0.30129262804985046, "eval_runtime": 14.3739, "eval_samples_per_second": 3.061, "eval_steps_per_second": 0.139, "step": 580 }, { "epoch": 0.15046453659641967, "grad_norm": 7.344346715445107, "learning_rate": 1.240406448489561e-06, "loss": 0.4737, "step": 581 }, { "epoch": 0.1507235117024376, "grad_norm": 7.796548314532856, "learning_rate": 1.2403075480752318e-06, "loss": 0.3963, "step": 582 }, { "epoch": 0.15098248680845552, "grad_norm": 4.3140902460127695, "learning_rate": 1.2402081444705218e-06, "loss": 0.4505, "step": 583 }, { "epoch": 0.15124146191447346, "grad_norm": 10.426328309205195, "learning_rate": 1.2401082377567222e-06, "loss": 0.4682, "step": 584 }, { "epoch": 0.1515004370204914, "grad_norm": 3.1885828757437666, "learning_rate": 1.2400078280155352e-06, "loss": 0.2957, "step": 585 }, { "epoch": 0.1515004370204914, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8930817610062893, "eval_PRM F1 AUC": 0.819015191199581, "eval_PRM F1 AUC (fixed)": 0.6809848088004191, "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": 0.31476715207099915, "eval_runtime": 14.6253, "eval_samples_per_second": 3.008, "eval_steps_per_second": 0.137, "step": 585 }, { "epoch": 0.15175941212650934, "grad_norm": 3.149213053009891, "learning_rate": 1.2399069153290748e-06, "loss": 0.2859, "step": 586 }, { "epoch": 0.15201838723252728, "grad_norm": 4.484121045797467, "learning_rate": 1.2398054997798657e-06, "loss": 0.3807, "step": 587 }, { "epoch": 0.15227736233854522, "grad_norm": 4.509486231542979, "learning_rate": 1.2397035814508447e-06, "loss": 0.4663, "step": 588 }, { "epoch": 0.15253633744456313, "grad_norm": 3.8777329311615483, "learning_rate": 1.2396011604253585e-06, "loss": 0.3775, "step": 589 }, { "epoch": 0.15279531255058107, "grad_norm": 7.543389414522049, "learning_rate": 1.2394982367871663e-06, "loss": 0.2858, "step": 590 }, { "epoch": 0.15279531255058107, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8903225806451613, "eval_PRM F1 AUC": 0.8504452592980618, "eval_PRM F1 AUC (fixed)": 0.6733892090099528, "eval_PRM F1 Neg": 0.7017543859649122, "eval_PRM NPV": 0.5882352941176471, "eval_PRM Precision": 0.9583333333333334, "eval_PRM Recall": 0.8313253012048193, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.33046090602874756, "eval_runtime": 14.6589, "eval_samples_per_second": 3.002, "eval_steps_per_second": 0.136, "step": 590 }, { "epoch": 0.153054287656599, "grad_norm": 5.42693700134366, "learning_rate": 1.2393948106204376e-06, "loss": 0.3109, "step": 591 }, { "epoch": 0.15331326276261695, "grad_norm": 7.0826540601798404, "learning_rate": 1.2392908820097526e-06, "loss": 0.434, "step": 592 }, { "epoch": 0.1535722378686349, "grad_norm": 4.604182378707424, "learning_rate": 1.2391864510401034e-06, "loss": 0.3272, "step": 593 }, { "epoch": 0.1538312129746528, "grad_norm": 5.012793342377323, "learning_rate": 1.2390815177968915e-06, "loss": 0.3389, "step": 594 }, { "epoch": 0.15409018808067074, "grad_norm": 4.471973602793316, "learning_rate": 1.2389760823659305e-06, "loss": 0.3548, "step": 595 }, { "epoch": 0.15409018808067074, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8930817610062893, "eval_PRM F1 AUC": 0.819015191199581, "eval_PRM F1 AUC (fixed)": 0.6838658983761131, "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": 0.31398123502731323, "eval_runtime": 14.9235, "eval_samples_per_second": 2.948, "eval_steps_per_second": 0.134, "step": 595 }, { "epoch": 0.15434916318668868, "grad_norm": 5.292384586964645, "learning_rate": 1.2388701448334438e-06, "loss": 0.3323, "step": 596 }, { "epoch": 0.15460813829270662, "grad_norm": 7.288360954102471, "learning_rate": 1.2387637052860657e-06, "loss": 0.4074, "step": 597 }, { "epoch": 0.15486711339872455, "grad_norm": 4.833374417379775, "learning_rate": 1.2386567638108412e-06, "loss": 0.3843, "step": 598 }, { "epoch": 0.1551260885047425, "grad_norm": 10.124157768027903, "learning_rate": 1.238549320495226e-06, "loss": 0.4332, "step": 599 }, { "epoch": 0.1553850636107604, "grad_norm": 8.37696636780662, "learning_rate": 1.238441375427085e-06, "loss": 0.38, "step": 600 }, { "epoch": 0.1553850636107604, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.9012345679012346, "eval_PRM F1 AUC": 0.8093242535358826, "eval_PRM F1 AUC (fixed)": 0.6849135673127291, "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": 0.3145059645175934, "eval_runtime": 14.6733, "eval_samples_per_second": 2.999, "eval_steps_per_second": 0.136, "step": 600 }, { "epoch": 0.15564403871677834, "grad_norm": 10.573652158959245, "learning_rate": 1.2383329286946946e-06, "loss": 0.3895, "step": 601 }, { "epoch": 0.15590301382279628, "grad_norm": 6.215297625313593, "learning_rate": 1.2382239803867413e-06, "loss": 0.3188, "step": 602 }, { "epoch": 0.15616198892881422, "grad_norm": 3.8865032915789315, "learning_rate": 1.2381145305923215e-06, "loss": 0.3603, "step": 603 }, { "epoch": 0.15642096403483216, "grad_norm": 4.305704623668269, "learning_rate": 1.2380045794009416e-06, "loss": 0.415, "step": 604 }, { "epoch": 0.1566799391408501, "grad_norm": 7.777263912530598, "learning_rate": 1.2378941269025184e-06, "loss": 0.3643, "step": 605 }, { "epoch": 0.1566799391408501, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8903225806451613, "eval_PRM F1 AUC": 0.8504452592980618, "eval_PRM F1 AUC (fixed)": 0.6843897328444212, "eval_PRM F1 Neg": 0.7017543859649122, "eval_PRM NPV": 0.5882352941176471, "eval_PRM Precision": 0.9583333333333334, "eval_PRM Recall": 0.8313253012048193, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.33305254578590393, "eval_runtime": 14.4625, "eval_samples_per_second": 3.042, "eval_steps_per_second": 0.138, "step": 605 }, { "epoch": 0.156938914246868, "grad_norm": 13.549319660584498, "learning_rate": 1.2377831731873786e-06, "loss": 0.3611, "step": 606 }, { "epoch": 0.15719788935288595, "grad_norm": 5.115617377808031, "learning_rate": 1.2376717183462583e-06, "loss": 0.4097, "step": 607 }, { "epoch": 0.1574568644589039, "grad_norm": 8.064002200115613, "learning_rate": 1.2375597624703042e-06, "loss": 0.3544, "step": 608 }, { "epoch": 0.15771583956492183, "grad_norm": 8.743740116901071, "learning_rate": 1.2374473056510723e-06, "loss": 0.3979, "step": 609 }, { "epoch": 0.15797481467093977, "grad_norm": 9.05779305197767, "learning_rate": 1.237334347980528e-06, "loss": 0.3519, "step": 610 }, { "epoch": 0.15797481467093977, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.89171974522293, "eval_PRM F1 AUC": 0.8347302252488213, "eval_PRM F1 AUC (fixed)": 0.6898899947616554, "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": 0.30802619457244873, "eval_runtime": 15.0217, "eval_samples_per_second": 2.929, "eval_steps_per_second": 0.133, "step": 610 }, { "epoch": 0.15823378977695768, "grad_norm": 5.861247610988427, "learning_rate": 1.2372208895510469e-06, "loss": 0.3434, "step": 611 }, { "epoch": 0.15849276488297562, "grad_norm": 3.670151363461034, "learning_rate": 1.2371069304554133e-06, "loss": 0.3702, "step": 612 }, { "epoch": 0.15875173998899356, "grad_norm": 8.084637417212509, "learning_rate": 1.2369924707868222e-06, "loss": 0.4298, "step": 613 }, { "epoch": 0.1590107150950115, "grad_norm": 5.687659548763739, "learning_rate": 1.236877510638877e-06, "loss": 0.3255, "step": 614 }, { "epoch": 0.15926969020102943, "grad_norm": 7.056523326388455, "learning_rate": 1.23676205010559e-06, "loss": 0.3883, "step": 615 }, { "epoch": 0.15926969020102943, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8944099378881988, "eval_PRM F1 AUC": 0.8033001571503404, "eval_PRM F1 AUC (fixed)": 0.6964379256155055, "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": 0.2987489402294159, "eval_runtime": 14.7965, "eval_samples_per_second": 2.974, "eval_steps_per_second": 0.135, "step": 615 }, { "epoch": 0.15952866530704737, "grad_norm": 8.052724316864127, "learning_rate": 1.2366460892813839e-06, "loss": 0.4153, "step": 616 }, { "epoch": 0.15978764041306528, "grad_norm": 11.44407409138463, "learning_rate": 1.2365296282610897e-06, "loss": 0.4042, "step": 617 }, { "epoch": 0.16004661551908322, "grad_norm": 14.598720324528122, "learning_rate": 1.2364126671399477e-06, "loss": 0.4369, "step": 618 }, { "epoch": 0.16030559062510116, "grad_norm": 5.482057017022813, "learning_rate": 1.236295206013607e-06, "loss": 0.3319, "step": 619 }, { "epoch": 0.1605645657311191, "grad_norm": 3.3221121701486727, "learning_rate": 1.2361772449781263e-06, "loss": 0.3924, "step": 620 }, { "epoch": 0.1605645657311191, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8846153846153846, "eval_PRM F1 AUC": 0.8287061288632793, "eval_PRM F1 AUC (fixed)": 0.6783656364588789, "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": 0.3130224049091339, "eval_runtime": 14.3415, "eval_samples_per_second": 3.068, "eval_steps_per_second": 0.139, "step": 620 }, { "epoch": 0.16082354083713704, "grad_norm": 3.3351070027555, "learning_rate": 1.236058784129972e-06, "loss": 0.3507, "step": 621 }, { "epoch": 0.16108251594315495, "grad_norm": 5.057842166618868, "learning_rate": 1.23593982356602e-06, "loss": 0.344, "step": 622 }, { "epoch": 0.1613414910491729, "grad_norm": 9.181148751364352, "learning_rate": 1.2358203633835549e-06, "loss": 0.4648, "step": 623 }, { "epoch": 0.16160046615519083, "grad_norm": 6.083979457492019, "learning_rate": 1.2357004036802695e-06, "loss": 0.2599, "step": 624 }, { "epoch": 0.16185944126120877, "grad_norm": 4.382520565652978, "learning_rate": 1.2355799445542649e-06, "loss": 0.3058, "step": 625 }, { "epoch": 0.16185944126120877, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8903225806451613, "eval_PRM F1 AUC": 0.8504452592980618, "eval_PRM F1 AUC (fixed)": 0.6846516500785752, "eval_PRM F1 Neg": 0.7017543859649122, "eval_PRM NPV": 0.5882352941176471, "eval_PRM Precision": 0.9583333333333334, "eval_PRM Recall": 0.8313253012048193, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.3220345973968506, "eval_runtime": 14.5696, "eval_samples_per_second": 3.02, "eval_steps_per_second": 0.137, "step": 625 }, { "epoch": 0.1621184163672267, "grad_norm": 4.1437604506815955, "learning_rate": 1.2354589861040514e-06, "loss": 0.3002, "step": 626 }, { "epoch": 0.16237739147324465, "grad_norm": 4.374843106284846, "learning_rate": 1.2353375284285475e-06, "loss": 0.3704, "step": 627 }, { "epoch": 0.16263636657926256, "grad_norm": 5.608222695272279, "learning_rate": 1.235215571627079e-06, "loss": 0.344, "step": 628 }, { "epoch": 0.1628953416852805, "grad_norm": 4.8582080854128895, "learning_rate": 1.2350931157993807e-06, "loss": 0.3336, "step": 629 }, { "epoch": 0.16315431679129844, "grad_norm": 12.482439714954587, "learning_rate": 1.2349701610455959e-06, "loss": 0.5153, "step": 630 }, { "epoch": 0.16315431679129844, "eval_PRM Accuracy": 0.8113207547169812, "eval_PRM F1": 0.8734177215189873, "eval_PRM F1 AUC": 0.785227867993714, "eval_PRM F1 AUC (fixed)": 0.6856993190151912, "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": 0.31282979249954224, "eval_runtime": 14.476, "eval_samples_per_second": 3.04, "eval_steps_per_second": 0.138, "step": 630 }, { "epoch": 0.16341329189731638, "grad_norm": 2.8269173986313767, "learning_rate": 1.2348467074662746e-06, "loss": 0.3002, "step": 631 }, { "epoch": 0.16367226700333432, "grad_norm": 9.367758085039979, "learning_rate": 1.234722755162376e-06, "loss": 0.4783, "step": 632 }, { "epoch": 0.16393124210935223, "grad_norm": 4.360443477858389, "learning_rate": 1.2345983042352668e-06, "loss": 0.3304, "step": 633 }, { "epoch": 0.16419021721537017, "grad_norm": 4.814617876325908, "learning_rate": 1.2344733547867211e-06, "loss": 0.3975, "step": 634 }, { "epoch": 0.1644491923213881, "grad_norm": 8.736415057185651, "learning_rate": 1.2343479069189208e-06, "loss": 0.3967, "step": 635 }, { "epoch": 0.1644491923213881, "eval_PRM Accuracy": 0.8207547169811321, "eval_PRM F1": 0.8789808917197452, "eval_PRM F1 AUC": 0.8069669984284967, "eval_PRM F1 AUC (fixed)": 0.6841278156102671, "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": 0.32189351320266724, "eval_runtime": 14.5716, "eval_samples_per_second": 3.02, "eval_steps_per_second": 0.137, "step": 635 }, { "epoch": 0.16470816742740604, "grad_norm": 8.16922264657535, "learning_rate": 1.2342219607344553e-06, "loss": 0.3905, "step": 636 }, { "epoch": 0.16496714253342398, "grad_norm": 4.102239363110395, "learning_rate": 1.2340955163363223e-06, "loss": 0.3864, "step": 637 }, { "epoch": 0.16522611763944192, "grad_norm": 6.666667743598508, "learning_rate": 1.2339685738279262e-06, "loss": 0.3439, "step": 638 }, { "epoch": 0.16548509274545983, "grad_norm": 6.359302938917453, "learning_rate": 1.2338411333130786e-06, "loss": 0.3799, "step": 639 }, { "epoch": 0.16574406785147777, "grad_norm": 3.4027973170905943, "learning_rate": 1.2337131948959988e-06, "loss": 0.3117, "step": 640 }, { "epoch": 0.16574406785147777, "eval_PRM Accuracy": 0.8113207547169812, "eval_PRM F1": 0.8734177215189873, "eval_PRM F1 AUC": 0.785227867993714, "eval_PRM F1 AUC (fixed)": 0.6901519119958093, "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": 0.3099837303161621, "eval_runtime": 14.876, "eval_samples_per_second": 2.958, "eval_steps_per_second": 0.134, "step": 640 }, { "epoch": 0.1660030429574957, "grad_norm": 4.065907555813363, "learning_rate": 1.2335847586813134e-06, "loss": 0.359, "step": 641 }, { "epoch": 0.16626201806351365, "grad_norm": 9.506205075827122, "learning_rate": 1.2334558247740555e-06, "loss": 0.3711, "step": 642 }, { "epoch": 0.1665209931695316, "grad_norm": 5.585007233460523, "learning_rate": 1.2333263932796658e-06, "loss": 0.3794, "step": 643 }, { "epoch": 0.1667799682755495, "grad_norm": 6.2824853591678895, "learning_rate": 1.2331964643039914e-06, "loss": 0.3819, "step": 644 }, { "epoch": 0.16703894338156744, "grad_norm": 3.0945268049557955, "learning_rate": 1.233066037953287e-06, "loss": 0.2563, "step": 645 }, { "epoch": 0.16703894338156744, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8860759493670886, "eval_PRM F1 AUC": 0.8129910948140389, "eval_PRM F1 AUC (fixed)": 0.7008905185961236, "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": 0.2902960777282715, "eval_runtime": 14.3898, "eval_samples_per_second": 3.058, "eval_steps_per_second": 0.139, "step": 645 }, { "epoch": 0.16729791848758538, "grad_norm": 8.26070229351453, "learning_rate": 1.2329351143342127e-06, "loss": 0.4367, "step": 646 }, { "epoch": 0.16755689359360332, "grad_norm": 3.5692756270657853, "learning_rate": 1.232803693553837e-06, "loss": 0.3088, "step": 647 }, { "epoch": 0.16781586869962126, "grad_norm": 6.322229970930503, "learning_rate": 1.2326717757196335e-06, "loss": 0.3815, "step": 648 }, { "epoch": 0.1680748438056392, "grad_norm": 5.4301781501602955, "learning_rate": 1.232539360939483e-06, "loss": 0.4233, "step": 649 }, { "epoch": 0.1683338189116571, "grad_norm": 4.037554188342987, "learning_rate": 1.2324064493216729e-06, "loss": 0.4186, "step": 650 }, { "epoch": 0.1683338189116571, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8846153846153846, "eval_PRM F1 AUC": 0.8287061288632793, "eval_PRM F1 AUC (fixed)": 0.7037716081718176, "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": 0.3051375150680542, "eval_runtime": 14.6736, "eval_samples_per_second": 2.999, "eval_steps_per_second": 0.136, "step": 650 }, { "epoch": 0.16859279401767505, "grad_norm": 5.903651242662739, "learning_rate": 1.2322730409748958e-06, "loss": 0.4349, "step": 651 }, { "epoch": 0.16885176912369299, "grad_norm": 4.79677556638012, "learning_rate": 1.2321391360082521e-06, "loss": 0.3621, "step": 652 }, { "epoch": 0.16911074422971092, "grad_norm": 5.0682489678833536, "learning_rate": 1.2320047345312468e-06, "loss": 0.4286, "step": 653 }, { "epoch": 0.16936971933572886, "grad_norm": 10.204257727871076, "learning_rate": 1.2318698366537924e-06, "loss": 0.418, "step": 654 }, { "epoch": 0.16962869444174677, "grad_norm": 5.9807619771124845, "learning_rate": 1.2317344424862062e-06, "loss": 0.4272, "step": 655 }, { "epoch": 0.16962869444174677, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8846153846153846, "eval_PRM F1 AUC": 0.8287061288632793, "eval_PRM F1 AUC (fixed)": 0.7003666841278156, "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": 0.3076668977737427, "eval_runtime": 14.5657, "eval_samples_per_second": 3.021, "eval_steps_per_second": 0.137, "step": 655 }, { "epoch": 0.1698876695477647, "grad_norm": 8.183477991339371, "learning_rate": 1.2315985521392116e-06, "loss": 0.3848, "step": 656 }, { "epoch": 0.17014664465378265, "grad_norm": 4.350350964575722, "learning_rate": 1.2314621657239382e-06, "loss": 0.4005, "step": 657 }, { "epoch": 0.1704056197598006, "grad_norm": 3.8484281205534305, "learning_rate": 1.2313252833519207e-06, "loss": 0.3491, "step": 658 }, { "epoch": 0.17066459486581853, "grad_norm": 3.8215249184870355, "learning_rate": 1.2311879051351e-06, "loss": 0.365, "step": 659 }, { "epoch": 0.17092356997183647, "grad_norm": 3.8839060530983347, "learning_rate": 1.2310500311858224e-06, "loss": 0.3976, "step": 660 }, { "epoch": 0.17092356997183647, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.89171974522293, "eval_PRM F1 AUC": 0.8347302252488213, "eval_PRM F1 AUC (fixed)": 0.6909376636982713, "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": 0.2925107777118683, "eval_runtime": 14.7018, "eval_samples_per_second": 2.993, "eval_steps_per_second": 0.136, "step": 660 }, { "epoch": 0.17118254507785438, "grad_norm": 3.2291152137659855, "learning_rate": 1.2309116616168388e-06, "loss": 0.3187, "step": 661 }, { "epoch": 0.17144152018387232, "grad_norm": 3.6725860645126867, "learning_rate": 1.2307727965413063e-06, "loss": 0.3757, "step": 662 }, { "epoch": 0.17170049528989026, "grad_norm": 3.1002808823226986, "learning_rate": 1.230633436072787e-06, "loss": 0.3074, "step": 663 }, { "epoch": 0.1719594703959082, "grad_norm": 8.746795861527746, "learning_rate": 1.2304935803252479e-06, "loss": 0.4095, "step": 664 }, { "epoch": 0.17221844550192614, "grad_norm": 10.82989204467529, "learning_rate": 1.2303532294130613e-06, "loss": 0.4814, "step": 665 }, { "epoch": 0.17221844550192614, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8930817610062893, "eval_PRM F1 AUC": 0.819015191199581, "eval_PRM F1 AUC (fixed)": 0.6875327396542692, "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": 0.30050474405288696, "eval_runtime": 14.9203, "eval_samples_per_second": 2.949, "eval_steps_per_second": 0.134, "step": 665 }, { "epoch": 0.17247742060794405, "grad_norm": 8.264683855258928, "learning_rate": 1.2302123834510042e-06, "loss": 0.4312, "step": 666 }, { "epoch": 0.172736395713962, "grad_norm": 4.929731323927467, "learning_rate": 1.2300710425542586e-06, "loss": 0.3498, "step": 667 }, { "epoch": 0.17299537081997993, "grad_norm": 4.665278640952237, "learning_rate": 1.2299292068384114e-06, "loss": 0.3288, "step": 668 }, { "epoch": 0.17325434592599787, "grad_norm": 4.455799796574909, "learning_rate": 1.2297868764194538e-06, "loss": 0.3569, "step": 669 }, { "epoch": 0.1735133210320158, "grad_norm": 3.97792111313062, "learning_rate": 1.2296440514137816e-06, "loss": 0.3906, "step": 670 }, { "epoch": 0.1735133210320158, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8903225806451613, "eval_PRM F1 AUC": 0.8504452592980618, "eval_PRM F1 AUC (fixed)": 0.6836039811419592, "eval_PRM F1 Neg": 0.7017543859649122, "eval_PRM NPV": 0.5882352941176471, "eval_PRM Precision": 0.9583333333333334, "eval_PRM Recall": 0.8313253012048193, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.32599514722824097, "eval_runtime": 14.7092, "eval_samples_per_second": 2.991, "eval_steps_per_second": 0.136, "step": 670 }, { "epoch": 0.17377229613803374, "grad_norm": 8.57169595641781, "learning_rate": 1.2295007319381956e-06, "loss": 0.3743, "step": 671 }, { "epoch": 0.17403127124405166, "grad_norm": 9.193160162768763, "learning_rate": 1.2293569181099002e-06, "loss": 0.5468, "step": 672 }, { "epoch": 0.1742902463500696, "grad_norm": 7.97975844919332, "learning_rate": 1.2292126100465047e-06, "loss": 0.4063, "step": 673 }, { "epoch": 0.17454922145608753, "grad_norm": 3.8452977252610747, "learning_rate": 1.229067807866022e-06, "loss": 0.3421, "step": 674 }, { "epoch": 0.17480819656210547, "grad_norm": 10.931204111910318, "learning_rate": 1.2289225116868696e-06, "loss": 0.3536, "step": 675 }, { "epoch": 0.17480819656210547, "eval_PRM Accuracy": 0.8113207547169812, "eval_PRM F1": 0.8734177215189873, "eval_PRM F1 AUC": 0.785227867993714, "eval_PRM F1 AUC (fixed)": 0.6864850707176532, "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": 0.32675302028656006, "eval_runtime": 14.7963, "eval_samples_per_second": 2.974, "eval_steps_per_second": 0.135, "step": 675 }, { "epoch": 0.1750671716681234, "grad_norm": 5.2549991101724425, "learning_rate": 1.228776721627869e-06, "loss": 0.4517, "step": 676 }, { "epoch": 0.17532614677414132, "grad_norm": 5.623571928637202, "learning_rate": 1.2286304378082454e-06, "loss": 0.403, "step": 677 }, { "epoch": 0.17558512188015926, "grad_norm": 7.871866515710695, "learning_rate": 1.2284836603476272e-06, "loss": 0.3853, "step": 678 }, { "epoch": 0.1758440969861772, "grad_norm": 3.150820890154312, "learning_rate": 1.2283363893660478e-06, "loss": 0.3173, "step": 679 }, { "epoch": 0.17610307209219514, "grad_norm": 4.865805406292621, "learning_rate": 1.2281886249839429e-06, "loss": 0.3076, "step": 680 }, { "epoch": 0.17610307209219514, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8860759493670886, "eval_PRM F1 AUC": 0.8129910948140389, "eval_PRM F1 AUC (fixed)": 0.6946045049764276, "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": 0.30507519841194153, "eval_runtime": 15.0032, "eval_samples_per_second": 2.933, "eval_steps_per_second": 0.133, "step": 680 }, { "epoch": 0.17636204719821308, "grad_norm": 3.6089302563745327, "learning_rate": 1.2280403673221522e-06, "loss": 0.3024, "step": 681 }, { "epoch": 0.17662102230423102, "grad_norm": 6.5551214656445485, "learning_rate": 1.2278916165019195e-06, "loss": 0.395, "step": 682 }, { "epoch": 0.17687999741024893, "grad_norm": 3.729361087732829, "learning_rate": 1.2277423726448905e-06, "loss": 0.3875, "step": 683 }, { "epoch": 0.17713897251626687, "grad_norm": 5.2636108819488365, "learning_rate": 1.2275926358731156e-06, "loss": 0.3809, "step": 684 }, { "epoch": 0.1773979476222848, "grad_norm": 3.124690927806883, "learning_rate": 1.227442406309047e-06, "loss": 0.3772, "step": 685 }, { "epoch": 0.1773979476222848, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8860759493670886, "eval_PRM F1 AUC": 0.8129910948140389, "eval_PRM F1 AUC (fixed)": 0.7019381875327396, "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": 0.29255253076553345, "eval_runtime": 14.5029, "eval_samples_per_second": 3.034, "eval_steps_per_second": 0.138, "step": 685 }, { "epoch": 0.17765692272830275, "grad_norm": 6.064065924164856, "learning_rate": 1.227291684075541e-06, "loss": 0.4751, "step": 686 }, { "epoch": 0.17791589783432069, "grad_norm": 3.8170721320489993, "learning_rate": 1.2271404692958553e-06, "loss": 0.327, "step": 687 }, { "epoch": 0.1781748729403386, "grad_norm": 6.515137825706018, "learning_rate": 1.2269887620936524e-06, "loss": 0.3708, "step": 688 }, { "epoch": 0.17843384804635654, "grad_norm": 2.9103376080730987, "learning_rate": 1.2268365625929956e-06, "loss": 0.3284, "step": 689 }, { "epoch": 0.17869282315237447, "grad_norm": 4.771365939250701, "learning_rate": 1.2266838709183523e-06, "loss": 0.4002, "step": 690 }, { "epoch": 0.17869282315237447, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8860759493670886, "eval_PRM F1 AUC": 0.8129910948140389, "eval_PRM F1 AUC (fixed)": 0.6938187532739654, "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": 0.29877424240112305, "eval_runtime": 14.538, "eval_samples_per_second": 3.027, "eval_steps_per_second": 0.138, "step": 690 }, { "epoch": 0.1789517982583924, "grad_norm": 6.18545393540289, "learning_rate": 1.2265306871945913e-06, "loss": 0.2963, "step": 691 }, { "epoch": 0.17921077336441035, "grad_norm": 3.6651890147345014, "learning_rate": 1.2263770115469844e-06, "loss": 0.3844, "step": 692 }, { "epoch": 0.1794697484704283, "grad_norm": 3.0288719572638274, "learning_rate": 1.2262228441012055e-06, "loss": 0.3152, "step": 693 }, { "epoch": 0.1797287235764462, "grad_norm": 4.805350206780974, "learning_rate": 1.2260681849833307e-06, "loss": 0.4042, "step": 694 }, { "epoch": 0.17998769868246414, "grad_norm": 4.11172461983298, "learning_rate": 1.2259130343198382e-06, "loss": 0.3348, "step": 695 }, { "epoch": 0.17998769868246414, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8930817610062893, "eval_PRM F1 AUC": 0.819015191199581, "eval_PRM F1 AUC (fixed)": 0.6953902566788894, "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": 0.298240065574646, "eval_runtime": 14.723, "eval_samples_per_second": 2.989, "eval_steps_per_second": 0.136, "step": 695 }, { "epoch": 0.18024667378848208, "grad_norm": 4.577272985951415, "learning_rate": 1.2257573922376082e-06, "loss": 0.3048, "step": 696 }, { "epoch": 0.18050564889450002, "grad_norm": 6.9027920443285655, "learning_rate": 1.2256012588639224e-06, "loss": 0.4526, "step": 697 }, { "epoch": 0.18076462400051796, "grad_norm": 6.007320601254431, "learning_rate": 1.2254446343264652e-06, "loss": 0.4382, "step": 698 }, { "epoch": 0.18102359910653587, "grad_norm": 6.4059971503781705, "learning_rate": 1.2252875187533216e-06, "loss": 0.2843, "step": 699 }, { "epoch": 0.1812825742125538, "grad_norm": 4.816960316434854, "learning_rate": 1.2251299122729788e-06, "loss": 0.3413, "step": 700 }, { "epoch": 0.1812825742125538, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.89171974522293, "eval_PRM F1 AUC": 0.8347302252488213, "eval_PRM F1 AUC (fixed)": 0.6927710843373494, "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": 0.309047669172287, "eval_runtime": 14.3751, "eval_samples_per_second": 3.061, "eval_steps_per_second": 0.139, "step": 700 }, { "epoch": 0.18154154931857175, "grad_norm": 5.101486987466127, "learning_rate": 1.2249718150143253e-06, "loss": 0.3951, "step": 701 }, { "epoch": 0.1818005244245897, "grad_norm": 8.502311209246374, "learning_rate": 1.224813227106651e-06, "loss": 0.4991, "step": 702 }, { "epoch": 0.18205949953060763, "grad_norm": 5.728862100407766, "learning_rate": 1.2246541486796468e-06, "loss": 0.3105, "step": 703 }, { "epoch": 0.18231847463662557, "grad_norm": 10.721025982358185, "learning_rate": 1.2244945798634053e-06, "loss": 0.5076, "step": 704 }, { "epoch": 0.18257744974264348, "grad_norm": 7.180199889514728, "learning_rate": 1.2243345207884193e-06, "loss": 0.3735, "step": 705 }, { "epoch": 0.18257744974264348, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8944099378881988, "eval_PRM F1 AUC": 0.8033001571503404, "eval_PRM F1 AUC (fixed)": 0.6888423258250393, "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": 0.30524346232414246, "eval_runtime": 14.8186, "eval_samples_per_second": 2.969, "eval_steps_per_second": 0.135, "step": 705 }, { "epoch": 0.18283642484866142, "grad_norm": 4.033301201721461, "learning_rate": 1.2241739715855833e-06, "loss": 0.3323, "step": 706 }, { "epoch": 0.18309539995467936, "grad_norm": 7.540320350893967, "learning_rate": 1.2240129323861923e-06, "loss": 0.4018, "step": 707 }, { "epoch": 0.1833543750606973, "grad_norm": 4.023342552193731, "learning_rate": 1.223851403321942e-06, "loss": 0.4608, "step": 708 }, { "epoch": 0.18361335016671523, "grad_norm": 3.4906951118001435, "learning_rate": 1.2236893845249286e-06, "loss": 0.3207, "step": 709 }, { "epoch": 0.18387232527273314, "grad_norm": 3.516295428942307, "learning_rate": 1.2235268761276492e-06, "loss": 0.4116, "step": 710 }, { "epoch": 0.18387232527273314, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.9, "eval_PRM F1 AUC": 0.8250392875851232, "eval_PRM F1 AUC (fixed)": 0.6888423258250392, "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": 0.2983972430229187, "eval_runtime": 14.7691, "eval_samples_per_second": 2.979, "eval_steps_per_second": 0.135, "step": 710 }, { "epoch": 0.18413130037875108, "grad_norm": 6.3107471370478345, "learning_rate": 1.2233638782630008e-06, "loss": 0.3446, "step": 711 }, { "epoch": 0.18439027548476902, "grad_norm": 4.766463349421268, "learning_rate": 1.223200391064281e-06, "loss": 0.4559, "step": 712 }, { "epoch": 0.18464925059078696, "grad_norm": 4.079425145872921, "learning_rate": 1.2230364146651874e-06, "loss": 0.3296, "step": 713 }, { "epoch": 0.1849082256968049, "grad_norm": 3.558588858327177, "learning_rate": 1.2228719491998178e-06, "loss": 0.3722, "step": 714 }, { "epoch": 0.18516720080282284, "grad_norm": 3.6740128427409187, "learning_rate": 1.22270699480267e-06, "loss": 0.3337, "step": 715 }, { "epoch": 0.18516720080282284, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8831168831168831, "eval_PRM F1 AUC": 0.8444211629125196, "eval_PRM F1 AUC (fixed)": 0.6893661602933473, "eval_PRM F1 Neg": 0.6896551724137931, "eval_PRM NPV": 0.5714285714285714, "eval_PRM Precision": 0.9577464788732394, "eval_PRM Recall": 0.8192771084337349, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.3114849328994751, "eval_runtime": 14.6726, "eval_samples_per_second": 2.999, "eval_steps_per_second": 0.136, "step": 715 }, { "epoch": 0.18542617590884075, "grad_norm": 6.842617200338519, "learning_rate": 1.222541551608641e-06, "loss": 0.4139, "step": 716 }, { "epoch": 0.1856851510148587, "grad_norm": 8.077453143122918, "learning_rate": 1.2223756197530287e-06, "loss": 0.4138, "step": 717 }, { "epoch": 0.18594412612087663, "grad_norm": 10.71942617315954, "learning_rate": 1.2222091993715296e-06, "loss": 0.3837, "step": 718 }, { "epoch": 0.18620310122689457, "grad_norm": 5.7851074969807526, "learning_rate": 1.2220422906002402e-06, "loss": 0.4341, "step": 719 }, { "epoch": 0.1864620763329125, "grad_norm": 4.993275457162052, "learning_rate": 1.2218748935756563e-06, "loss": 0.4591, "step": 720 }, { "epoch": 0.1864620763329125, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.89171974522293, "eval_PRM F1 AUC": 0.8347302252488213, "eval_PRM F1 AUC (fixed)": 0.6901519119958094, "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": 0.30387333035469055, "eval_runtime": 15.0727, "eval_samples_per_second": 2.919, "eval_steps_per_second": 0.133, "step": 720 }, { "epoch": 0.18672105143893045, "grad_norm": 9.287087083955917, "learning_rate": 1.221707008434673e-06, "loss": 0.3178, "step": 721 }, { "epoch": 0.18698002654494836, "grad_norm": 8.33504855523225, "learning_rate": 1.2215386353145848e-06, "loss": 0.3372, "step": 722 }, { "epoch": 0.1872390016509663, "grad_norm": 3.0068926378723506, "learning_rate": 1.2213697743530841e-06, "loss": 0.3144, "step": 723 }, { "epoch": 0.18749797675698424, "grad_norm": 4.173934848386845, "learning_rate": 1.2212004256882643e-06, "loss": 0.4498, "step": 724 }, { "epoch": 0.18775695186300217, "grad_norm": 3.919799887702488, "learning_rate": 1.2210305894586158e-06, "loss": 0.3239, "step": 725 }, { "epoch": 0.18775695186300217, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8860759493670886, "eval_PRM F1 AUC": 0.8129910948140389, "eval_PRM F1 AUC (fixed)": 0.6930330015715034, "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": 0.3035014569759369, "eval_runtime": 14.3599, "eval_samples_per_second": 3.064, "eval_steps_per_second": 0.139, "step": 725 }, { "epoch": 0.18801592696902011, "grad_norm": 3.465868954032309, "learning_rate": 1.2208602658030287e-06, "loss": 0.3427, "step": 726 }, { "epoch": 0.18827490207503803, "grad_norm": 3.9040806750182657, "learning_rate": 1.2206894548607914e-06, "loss": 0.3492, "step": 727 }, { "epoch": 0.18853387718105596, "grad_norm": 4.379059220952106, "learning_rate": 1.2205181567715905e-06, "loss": 0.4384, "step": 728 }, { "epoch": 0.1887928522870739, "grad_norm": 4.134944823065672, "learning_rate": 1.2203463716755117e-06, "loss": 0.2962, "step": 729 }, { "epoch": 0.18905182739309184, "grad_norm": 4.76895358133757, "learning_rate": 1.2201740997130388e-06, "loss": 0.4152, "step": 730 }, { "epoch": 0.18905182739309184, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8930817610062893, "eval_PRM F1 AUC": 0.819015191199581, "eval_PRM F1 AUC (fixed)": 0.6951283394447355, "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": 0.29401424527168274, "eval_runtime": 14.755, "eval_samples_per_second": 2.982, "eval_steps_per_second": 0.136, "step": 730 }, { "epoch": 0.18931080249910978, "grad_norm": 3.5289293203995435, "learning_rate": 1.2200013410250528e-06, "loss": 0.3676, "step": 731 }, { "epoch": 0.18956977760512772, "grad_norm": 3.544637575803375, "learning_rate": 1.219828095752834e-06, "loss": 0.3552, "step": 732 }, { "epoch": 0.18982875271114563, "grad_norm": 8.342332576696858, "learning_rate": 1.2196543640380601e-06, "loss": 0.4627, "step": 733 }, { "epoch": 0.19008772781716357, "grad_norm": 6.969226338650388, "learning_rate": 1.2194801460228063e-06, "loss": 0.4385, "step": 734 }, { "epoch": 0.1903467029231815, "grad_norm": 4.161712948947366, "learning_rate": 1.219305441849546e-06, "loss": 0.3898, "step": 735 }, { "epoch": 0.1903467029231815, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8944099378881988, "eval_PRM F1 AUC": 0.8033001571503404, "eval_PRM F1 AUC (fixed)": 0.6938187532739655, "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": 0.2999347448348999, "eval_runtime": 15.0306, "eval_samples_per_second": 2.927, "eval_steps_per_second": 0.133, "step": 735 }, { "epoch": 0.19060567802919945, "grad_norm": 6.84476769969144, "learning_rate": 1.21913025166115e-06, "loss": 0.4164, "step": 736 }, { "epoch": 0.1908646531352174, "grad_norm": 5.136634266840099, "learning_rate": 1.2189545756008864e-06, "loss": 0.3763, "step": 737 }, { "epoch": 0.1911236282412353, "grad_norm": 7.2580426628791255, "learning_rate": 1.2187784138124207e-06, "loss": 0.3112, "step": 738 }, { "epoch": 0.19138260334725324, "grad_norm": 5.718167380444347, "learning_rate": 1.2186017664398158e-06, "loss": 0.3891, "step": 739 }, { "epoch": 0.19164157845327118, "grad_norm": 6.763967879398377, "learning_rate": 1.218424633627532e-06, "loss": 0.342, "step": 740 }, { "epoch": 0.19164157845327118, "eval_PRM Accuracy": 0.8679245283018868, "eval_PRM F1": 0.9135802469135802, "eval_PRM F1 AUC": 0.8370874803562075, "eval_PRM F1 AUC (fixed)": 0.6940806705081194, "eval_PRM F1 Neg": 0.72, "eval_PRM NPV": 0.6666666666666666, "eval_PRM Precision": 0.9367088607594937, "eval_PRM Recall": 0.891566265060241, "eval_PRM Specificty": 0.782608695652174, "eval_loss": 0.2982929050922394, "eval_runtime": 14.3814, "eval_samples_per_second": 3.06, "eval_steps_per_second": 0.139, "step": 740 }, { "epoch": 0.19190055355928912, "grad_norm": 4.5433183235959005, "learning_rate": 1.2182470155204253e-06, "loss": 0.3741, "step": 741 }, { "epoch": 0.19215952866530706, "grad_norm": 5.452456164324737, "learning_rate": 1.2180689122637499e-06, "loss": 0.397, "step": 742 }, { "epoch": 0.192418503771325, "grad_norm": 3.4223013167649468, "learning_rate": 1.2178903240031563e-06, "loss": 0.2803, "step": 743 }, { "epoch": 0.1926774788773429, "grad_norm": 5.385205369812769, "learning_rate": 1.217711250884692e-06, "loss": 0.3934, "step": 744 }, { "epoch": 0.19293645398336084, "grad_norm": 6.183148792048128, "learning_rate": 1.2175316930547997e-06, "loss": 0.4244, "step": 745 }, { "epoch": 0.19293645398336084, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.906832298136646, "eval_PRM F1 AUC": 0.8310633839706653, "eval_PRM F1 AUC (fixed)": 0.6911995809324253, "eval_PRM F1 Neg": 0.7058823529411765, "eval_PRM NPV": 0.6428571428571429, "eval_PRM Precision": 0.9358974358974359, "eval_PRM Recall": 0.8795180722891566, "eval_PRM Specificty": 0.782608695652174, "eval_loss": 0.29950782656669617, "eval_runtime": 14.3784, "eval_samples_per_second": 3.06, "eval_steps_per_second": 0.139, "step": 745 }, { "epoch": 0.19319542908937878, "grad_norm": 3.935237830232094, "learning_rate": 1.2173516506603203e-06, "loss": 0.3224, "step": 746 }, { "epoch": 0.19345440419539672, "grad_norm": 4.276810366583379, "learning_rate": 1.2171711238484896e-06, "loss": 0.3685, "step": 747 }, { "epoch": 0.19371337930141466, "grad_norm": 8.221097875723666, "learning_rate": 1.2169901127669401e-06, "loss": 0.4421, "step": 748 }, { "epoch": 0.19397235440743257, "grad_norm": 3.1276241732515877, "learning_rate": 1.2168086175637006e-06, "loss": 0.3708, "step": 749 }, { "epoch": 0.1942313295134505, "grad_norm": 5.711025740988542, "learning_rate": 1.216626638387195e-06, "loss": 0.3505, "step": 750 }, { "epoch": 0.1942313295134505, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.89171974522293, "eval_PRM F1 AUC": 0.8347302252488213, "eval_PRM F1 AUC (fixed)": 0.6893661602933472, "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": 0.3052200376987457, "eval_runtime": 14.9201, "eval_samples_per_second": 2.949, "eval_steps_per_second": 0.134, "step": 750 }, { "epoch": 0.19449030461946845, "grad_norm": 4.218636733039021, "learning_rate": 1.2164441753862438e-06, "loss": 0.3969, "step": 751 }, { "epoch": 0.1947492797254864, "grad_norm": 6.769566136370222, "learning_rate": 1.2162612287100627e-06, "loss": 0.3348, "step": 752 }, { "epoch": 0.19500825483150433, "grad_norm": 3.4857801672629707, "learning_rate": 1.216077798508263e-06, "loss": 0.3256, "step": 753 }, { "epoch": 0.19526722993752227, "grad_norm": 9.174244423244135, "learning_rate": 1.2158938849308519e-06, "loss": 0.368, "step": 754 }, { "epoch": 0.19552620504354018, "grad_norm": 9.747075045547675, "learning_rate": 1.2157094881282308e-06, "loss": 0.336, "step": 755 }, { "epoch": 0.19552620504354018, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.89171974522293, "eval_PRM F1 AUC": 0.8347302252488213, "eval_PRM F1 AUC (fixed)": 0.6990570979570455, "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": 0.2928756773471832, "eval_runtime": 14.2527, "eval_samples_per_second": 3.087, "eval_steps_per_second": 0.14, "step": 755 }, { "epoch": 0.19578518014955812, "grad_norm": 3.251806890809459, "learning_rate": 1.2155246082511975e-06, "loss": 0.2617, "step": 756 }, { "epoch": 0.19604415525557606, "grad_norm": 4.98149048932274, "learning_rate": 1.2153392454509442e-06, "loss": 0.3699, "step": 757 }, { "epoch": 0.196303130361594, "grad_norm": 3.9074566654750433, "learning_rate": 1.2151533998790583e-06, "loss": 0.4173, "step": 758 }, { "epoch": 0.19656210546761194, "grad_norm": 3.752364680965662, "learning_rate": 1.2149670716875215e-06, "loss": 0.3821, "step": 759 }, { "epoch": 0.19682108057362985, "grad_norm": 5.348304560974054, "learning_rate": 1.2147802610287107e-06, "loss": 0.3873, "step": 760 }, { "epoch": 0.19682108057362985, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8860759493670886, "eval_PRM F1 AUC": 0.8129910948140389, "eval_PRM F1 AUC (fixed)": 0.711629125196438, "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": 0.28274041414260864, "eval_runtime": 14.5446, "eval_samples_per_second": 3.025, "eval_steps_per_second": 0.138, "step": 760 }, { "epoch": 0.1970800556796478, "grad_norm": 7.55785178982506, "learning_rate": 1.2145929680553973e-06, "loss": 0.324, "step": 761 }, { "epoch": 0.19733903078566573, "grad_norm": 3.0868830616823852, "learning_rate": 1.2144051929207466e-06, "loss": 0.2994, "step": 762 }, { "epoch": 0.19759800589168366, "grad_norm": 3.5417146803588118, "learning_rate": 1.2142169357783194e-06, "loss": 0.3483, "step": 763 }, { "epoch": 0.1978569809977016, "grad_norm": 4.0208469768721224, "learning_rate": 1.214028196782069e-06, "loss": 0.4194, "step": 764 }, { "epoch": 0.19811595610371954, "grad_norm": 3.8431699488974886, "learning_rate": 1.2138389760863443e-06, "loss": 0.3649, "step": 765 }, { "epoch": 0.19811595610371954, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8930817610062893, "eval_PRM F1 AUC": 0.819015191199581, "eval_PRM F1 AUC (fixed)": 0.7257726558407543, "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": 0.274412602186203, "eval_runtime": 14.9209, "eval_samples_per_second": 2.949, "eval_steps_per_second": 0.134, "step": 765 }, { "epoch": 0.19837493120973745, "grad_norm": 4.5792155524540386, "learning_rate": 1.2136492738458874e-06, "loss": 0.3576, "step": 766 }, { "epoch": 0.1986339063157554, "grad_norm": 8.636004621584396, "learning_rate": 1.2134590902158339e-06, "loss": 0.4316, "step": 767 }, { "epoch": 0.19889288142177333, "grad_norm": 4.140044364503692, "learning_rate": 1.213268425351714e-06, "loss": 0.3677, "step": 768 }, { "epoch": 0.19915185652779127, "grad_norm": 5.324756102689371, "learning_rate": 1.2130772794094504e-06, "loss": 0.3403, "step": 769 }, { "epoch": 0.1994108316338092, "grad_norm": 8.4411250018076, "learning_rate": 1.21288565254536e-06, "loss": 0.391, "step": 770 }, { "epoch": 0.1994108316338092, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8930817610062893, "eval_PRM F1 AUC": 0.819015191199581, "eval_PRM F1 AUC (fixed)": 0.7247249869041383, "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": 0.2825393080711365, "eval_runtime": 15.1448, "eval_samples_per_second": 2.905, "eval_steps_per_second": 0.132, "step": 770 }, { "epoch": 0.19966980673982712, "grad_norm": 7.501946899653349, "learning_rate": 1.2126935449161529e-06, "loss": 0.3256, "step": 771 }, { "epoch": 0.19992878184584506, "grad_norm": 5.065470925957987, "learning_rate": 1.2125009566789318e-06, "loss": 0.3201, "step": 772 }, { "epoch": 0.200187756951863, "grad_norm": 3.8048849885969136, "learning_rate": 1.2123078879911928e-06, "loss": 0.3268, "step": 773 }, { "epoch": 0.20044673205788094, "grad_norm": 7.635096076804379, "learning_rate": 1.2121143390108253e-06, "loss": 0.4506, "step": 774 }, { "epoch": 0.20070570716389888, "grad_norm": 3.4186284893273022, "learning_rate": 1.2119203098961107e-06, "loss": 0.3222, "step": 775 }, { "epoch": 0.20070570716389888, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.8974358974358975, "eval_PRM F1 AUC": 0.8564693556836039, "eval_PRM F1 AUC (fixed)": 0.714772132006286, "eval_PRM F1 Neg": 0.7142857142857143, "eval_PRM NPV": 0.6060606060606061, "eval_PRM Precision": 0.958904109589041, "eval_PRM Recall": 0.8433734939759037, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.2990555763244629, "eval_runtime": 14.5907, "eval_samples_per_second": 3.016, "eval_steps_per_second": 0.137, "step": 775 }, { "epoch": 0.20096468226991682, "grad_norm": 5.857241588906598, "learning_rate": 1.211725800805724e-06, "loss": 0.412, "step": 776 }, { "epoch": 0.20122365737593473, "grad_norm": 8.636257256129669, "learning_rate": 1.2115308118987314e-06, "loss": 0.3583, "step": 777 }, { "epoch": 0.20148263248195267, "grad_norm": 4.001799400777121, "learning_rate": 1.2113353433345928e-06, "loss": 0.4281, "step": 778 }, { "epoch": 0.2017416075879706, "grad_norm": 5.457582326684946, "learning_rate": 1.2111393952731599e-06, "loss": 0.426, "step": 779 }, { "epoch": 0.20200058269398855, "grad_norm": 5.88931960974729, "learning_rate": 1.2109429678746757e-06, "loss": 0.3624, "step": 780 }, { "epoch": 0.20200058269398855, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.8987341772151899, "eval_PRM F1 AUC": 0.8407543216343635, "eval_PRM F1 AUC (fixed)": 0.7197485594552121, "eval_PRM F1 Neg": 0.7037037037037037, "eval_PRM NPV": 0.6129032258064516, "eval_PRM Precision": 0.9466666666666667, "eval_PRM Recall": 0.8554216867469879, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.2832641005516052, "eval_runtime": 15.2286, "eval_samples_per_second": 2.889, "eval_steps_per_second": 0.131, "step": 780 }, { "epoch": 0.20225955780000648, "grad_norm": 4.5606319970193905, "learning_rate": 1.2107460612997768e-06, "loss": 0.4612, "step": 781 }, { "epoch": 0.2025185329060244, "grad_norm": 10.640629778364074, "learning_rate": 1.2105486757094902e-06, "loss": 0.3714, "step": 782 }, { "epoch": 0.20277750801204233, "grad_norm": 10.81610998670503, "learning_rate": 1.2103508112652357e-06, "loss": 0.4135, "step": 783 }, { "epoch": 0.20303648311806027, "grad_norm": 5.278795898437136, "learning_rate": 1.2101524681288238e-06, "loss": 0.4048, "step": 784 }, { "epoch": 0.2032954582240782, "grad_norm": 9.07962614012763, "learning_rate": 1.209953646462457e-06, "loss": 0.3948, "step": 785 }, { "epoch": 0.2032954582240782, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.9056603773584906, "eval_PRM F1 AUC": 0.8467784180199058, "eval_PRM F1 AUC (fixed)": 0.7100576217915139, "eval_PRM F1 Neg": 0.7169811320754716, "eval_PRM NPV": 0.6333333333333333, "eval_PRM Precision": 0.9473684210526315, "eval_PRM Recall": 0.8674698795180723, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.28009822964668274, "eval_runtime": 14.868, "eval_samples_per_second": 2.959, "eval_steps_per_second": 0.135, "step": 785 }, { "epoch": 0.20355443333009615, "grad_norm": 4.945794786482562, "learning_rate": 1.2097543464287291e-06, "loss": 0.3626, "step": 786 }, { "epoch": 0.2038134084361141, "grad_norm": 4.210535900275439, "learning_rate": 1.2095545681906252e-06, "loss": 0.4574, "step": 787 }, { "epoch": 0.204072383542132, "grad_norm": 7.8934875649755885, "learning_rate": 1.2093543119115211e-06, "loss": 0.4204, "step": 788 }, { "epoch": 0.20433135864814994, "grad_norm": 5.9298063027561705, "learning_rate": 1.2091535777551837e-06, "loss": 0.3544, "step": 789 }, { "epoch": 0.20459033375416788, "grad_norm": 4.61640368794494, "learning_rate": 1.2089523658857708e-06, "loss": 0.3672, "step": 790 }, { "epoch": 0.20459033375416788, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8831168831168831, "eval_PRM F1 AUC": 0.8444211629125196, "eval_PRM F1 AUC (fixed)": 0.6946045049764275, "eval_PRM F1 Neg": 0.6896551724137931, "eval_PRM NPV": 0.5714285714285714, "eval_PRM Precision": 0.9577464788732394, "eval_PRM Recall": 0.8192771084337349, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.30553391575813293, "eval_runtime": 14.5411, "eval_samples_per_second": 3.026, "eval_steps_per_second": 0.138, "step": 790 }, { "epoch": 0.20484930886018582, "grad_norm": 3.8906893870399473, "learning_rate": 1.2087506764678309e-06, "loss": 0.3811, "step": 791 }, { "epoch": 0.20510828396620376, "grad_norm": 5.048510996715175, "learning_rate": 1.2085485096663026e-06, "loss": 0.5011, "step": 792 }, { "epoch": 0.20536725907222167, "grad_norm": 11.837946428436359, "learning_rate": 1.2083458656465159e-06, "loss": 0.3957, "step": 793 }, { "epoch": 0.2056262341782396, "grad_norm": 11.305888044956315, "learning_rate": 1.2081427445741896e-06, "loss": 0.3496, "step": 794 }, { "epoch": 0.20588520928425755, "grad_norm": 9.260364149994357, "learning_rate": 1.207939146615434e-06, "loss": 0.3768, "step": 795 }, { "epoch": 0.20588520928425755, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8846153846153846, "eval_PRM F1 AUC": 0.8287061288632793, "eval_PRM F1 AUC (fixed)": 0.6946045049764275, "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": 0.3177688419818878, "eval_runtime": 14.5323, "eval_samples_per_second": 3.028, "eval_steps_per_second": 0.138, "step": 795 }, { "epoch": 0.2061441843902755, "grad_norm": 6.166600089935319, "learning_rate": 1.207735071936749e-06, "loss": 0.3425, "step": 796 }, { "epoch": 0.20640315949629343, "grad_norm": 5.291436952965745, "learning_rate": 1.2075305207050233e-06, "loss": 0.3083, "step": 797 }, { "epoch": 0.20666213460231136, "grad_norm": 4.946759609608186, "learning_rate": 1.2073254930875372e-06, "loss": 0.346, "step": 798 }, { "epoch": 0.20692110970832928, "grad_norm": 3.2952179929570984, "learning_rate": 1.2071199892519587e-06, "loss": 0.3274, "step": 799 }, { "epoch": 0.20718008481434722, "grad_norm": 3.9838080756980765, "learning_rate": 1.2069140093663468e-06, "loss": 0.4127, "step": 800 }, { "epoch": 0.20718008481434722, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.9079754601226994, "eval_PRM F1 AUC": 0.8153483499214248, "eval_PRM F1 AUC (fixed)": 0.710843373493976, "eval_PRM F1 Neg": 0.6938775510204082, "eval_PRM NPV": 0.6538461538461539, "eval_PRM Precision": 0.925, "eval_PRM Recall": 0.891566265060241, "eval_PRM Specificty": 0.7391304347826086, "eval_loss": 0.3012973964214325, "eval_runtime": 14.7148, "eval_samples_per_second": 2.99, "eval_steps_per_second": 0.136, "step": 800 }, { "epoch": 0.20743905992036515, "grad_norm": 5.84440802349982, "learning_rate": 1.2067075535991485e-06, "loss": 0.3121, "step": 801 }, { "epoch": 0.2076980350263831, "grad_norm": 18.634255107630327, "learning_rate": 1.206500622119201e-06, "loss": 0.6074, "step": 802 }, { "epoch": 0.20795701013240103, "grad_norm": 3.4684820156592164, "learning_rate": 1.2062932150957302e-06, "loss": 0.2838, "step": 803 }, { "epoch": 0.20821598523841894, "grad_norm": 6.985530687578916, "learning_rate": 1.2060853326983506e-06, "loss": 0.3661, "step": 804 }, { "epoch": 0.20847496034443688, "grad_norm": 15.523633503922179, "learning_rate": 1.205876975097066e-06, "loss": 0.5244, "step": 805 }, { "epoch": 0.20847496034443688, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.9079754601226994, "eval_PRM F1 AUC": 0.8153483499214248, "eval_PRM F1 AUC (fixed)": 0.717391304347826, "eval_PRM F1 Neg": 0.6938775510204082, "eval_PRM NPV": 0.6538461538461539, "eval_PRM Precision": 0.925, "eval_PRM Recall": 0.891566265060241, "eval_PRM Specificty": 0.7391304347826086, "eval_loss": 0.3000315725803375, "eval_runtime": 14.3543, "eval_samples_per_second": 3.065, "eval_steps_per_second": 0.139, "step": 805 }, { "epoch": 0.20873393545045482, "grad_norm": 3.215330481636461, "learning_rate": 1.2056681424622682e-06, "loss": 0.3055, "step": 806 }, { "epoch": 0.20899291055647276, "grad_norm": 3.3904213572865647, "learning_rate": 1.2054588349647376e-06, "loss": 0.3612, "step": 807 }, { "epoch": 0.2092518856624907, "grad_norm": 4.089578401872159, "learning_rate": 1.2052490527756435e-06, "loss": 0.3688, "step": 808 }, { "epoch": 0.20951086076850864, "grad_norm": 5.149244379565398, "learning_rate": 1.205038796066543e-06, "loss": 0.3794, "step": 809 }, { "epoch": 0.20976983587452655, "grad_norm": 7.982346861326755, "learning_rate": 1.2048280650093808e-06, "loss": 0.4222, "step": 810 }, { "epoch": 0.20976983587452655, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.89171974522293, "eval_PRM F1 AUC": 0.8347302252488213, "eval_PRM F1 AUC (fixed)": 0.7192247249869042, "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": 0.3014485538005829, "eval_runtime": 14.5923, "eval_samples_per_second": 3.015, "eval_steps_per_second": 0.137, "step": 810 }, { "epoch": 0.2100288109805445, "grad_norm": 3.3870290321601293, "learning_rate": 1.2046168597764905e-06, "loss": 0.3392, "step": 811 }, { "epoch": 0.21028778608656243, "grad_norm": 5.422644541842962, "learning_rate": 1.2044051805405923e-06, "loss": 0.3788, "step": 812 }, { "epoch": 0.21054676119258037, "grad_norm": 12.103695395274976, "learning_rate": 1.2041930274747955e-06, "loss": 0.394, "step": 813 }, { "epoch": 0.2108057362985983, "grad_norm": 5.780235540532406, "learning_rate": 1.2039804007525956e-06, "loss": 0.4104, "step": 814 }, { "epoch": 0.21106471140461622, "grad_norm": 8.248186220584325, "learning_rate": 1.203767300547876e-06, "loss": 0.3479, "step": 815 }, { "epoch": 0.21106471140461622, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8903225806451613, "eval_PRM F1 AUC": 0.8504452592980618, "eval_PRM F1 AUC (fixed)": 0.7168674698795181, "eval_PRM F1 Neg": 0.7017543859649122, "eval_PRM NPV": 0.5882352941176471, "eval_PRM Precision": 0.9583333333333334, "eval_PRM Recall": 0.8313253012048193, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.3021652400493622, "eval_runtime": 14.4687, "eval_samples_per_second": 3.041, "eval_steps_per_second": 0.138, "step": 815 }, { "epoch": 0.21132368651063416, "grad_norm": 10.434436452029244, "learning_rate": 1.203553727034907e-06, "loss": 0.3728, "step": 816 }, { "epoch": 0.2115826616166521, "grad_norm": 4.209869566782344, "learning_rate": 1.2033396803883467e-06, "loss": 0.3499, "step": 817 }, { "epoch": 0.21184163672267003, "grad_norm": 5.604499786309078, "learning_rate": 1.2031251607832393e-06, "loss": 0.4211, "step": 818 }, { "epoch": 0.21210061182868797, "grad_norm": 3.4492243819967734, "learning_rate": 1.2029101683950161e-06, "loss": 0.4104, "step": 819 }, { "epoch": 0.2123595869347059, "grad_norm": 5.627721634673, "learning_rate": 1.202694703399495e-06, "loss": 0.4138, "step": 820 }, { "epoch": 0.2123595869347059, "eval_PRM Accuracy": 0.8679245283018868, "eval_PRM F1": 0.9125, "eval_PRM F1 AUC": 0.8528025144054479, "eval_PRM F1 AUC (fixed)": 0.7187008905185961, "eval_PRM F1 Neg": 0.7307692307692307, "eval_PRM NPV": 0.6551724137931034, "eval_PRM Precision": 0.948051948051948, "eval_PRM Recall": 0.8795180722891566, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.272720068693161, "eval_runtime": 14.4143, "eval_samples_per_second": 3.053, "eval_steps_per_second": 0.139, "step": 820 }, { "epoch": 0.21261856204072382, "grad_norm": 4.1223242524960275, "learning_rate": 1.2024787659728805e-06, "loss": 0.3767, "step": 821 }, { "epoch": 0.21287753714674176, "grad_norm": 4.041731071615488, "learning_rate": 1.2022623562917634e-06, "loss": 0.3681, "step": 822 }, { "epoch": 0.2131365122527597, "grad_norm": 12.213894926599707, "learning_rate": 1.2020454745331207e-06, "loss": 0.5263, "step": 823 }, { "epoch": 0.21339548735877764, "grad_norm": 4.048567978079006, "learning_rate": 1.2018281208743155e-06, "loss": 0.2961, "step": 824 }, { "epoch": 0.21365446246479558, "grad_norm": 3.4978630432290467, "learning_rate": 1.2016102954930965e-06, "loss": 0.2978, "step": 825 }, { "epoch": 0.21365446246479558, "eval_PRM Accuracy": 0.8679245283018868, "eval_PRM F1": 0.9125, "eval_PRM F1 AUC": 0.8528025144054479, "eval_PRM F1 AUC (fixed)": 0.7160817181770561, "eval_PRM F1 Neg": 0.7307692307692307, "eval_PRM NPV": 0.6551724137931034, "eval_PRM Precision": 0.948051948051948, "eval_PRM Recall": 0.8795180722891566, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.27165505290031433, "eval_runtime": 14.2242, "eval_samples_per_second": 3.093, "eval_steps_per_second": 0.141, "step": 825 }, { "epoch": 0.21391343757081352, "grad_norm": 6.136696971864226, "learning_rate": 1.2013919985675985e-06, "loss": 0.4115, "step": 826 }, { "epoch": 0.21417241267683143, "grad_norm": 7.112543853206797, "learning_rate": 1.2011732302763419e-06, "loss": 0.3849, "step": 827 }, { "epoch": 0.21443138778284937, "grad_norm": 3.7784105346607646, "learning_rate": 1.2009539907982324e-06, "loss": 0.3712, "step": 828 }, { "epoch": 0.2146903628888673, "grad_norm": 4.433101135916309, "learning_rate": 1.2007342803125616e-06, "loss": 0.3525, "step": 829 }, { "epoch": 0.21494933799488525, "grad_norm": 3.219684081039623, "learning_rate": 1.2005140989990051e-06, "loss": 0.325, "step": 830 }, { "epoch": 0.21494933799488525, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.9056603773584906, "eval_PRM F1 AUC": 0.8467784180199058, "eval_PRM F1 AUC (fixed)": 0.7145102147721321, "eval_PRM F1 Neg": 0.7169811320754716, "eval_PRM NPV": 0.6333333333333333, "eval_PRM Precision": 0.9473684210526315, "eval_PRM Recall": 0.8674698795180723, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.2757134437561035, "eval_runtime": 14.4793, "eval_samples_per_second": 3.039, "eval_steps_per_second": 0.138, "step": 830 }, { "epoch": 0.2152083131009032, "grad_norm": 10.491754275926994, "learning_rate": 1.2002934470376248e-06, "loss": 0.4024, "step": 831 }, { "epoch": 0.2154672882069211, "grad_norm": 4.998222363731994, "learning_rate": 1.200072324608867e-06, "loss": 0.3625, "step": 832 }, { "epoch": 0.21572626331293904, "grad_norm": 4.470005863558178, "learning_rate": 1.1998507318935624e-06, "loss": 0.3332, "step": 833 }, { "epoch": 0.21598523841895698, "grad_norm": 7.993509681116096, "learning_rate": 1.1996286690729266e-06, "loss": 0.369, "step": 834 }, { "epoch": 0.21624421352497492, "grad_norm": 4.708532369425207, "learning_rate": 1.1994061363285598e-06, "loss": 0.3882, "step": 835 }, { "epoch": 0.21624421352497492, "eval_PRM Accuracy": 0.8679245283018868, "eval_PRM F1": 0.9113924050632911, "eval_PRM F1 AUC": 0.8685175484546883, "eval_PRM F1 AUC (fixed)": 0.7145102147721321, "eval_PRM F1 Neg": 0.7407407407407407, "eval_PRM NPV": 0.6451612903225806, "eval_PRM Precision": 0.96, "eval_PRM Recall": 0.8674698795180723, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.27347585558891296, "eval_runtime": 15.0518, "eval_samples_per_second": 2.923, "eval_steps_per_second": 0.133, "step": 835 }, { "epoch": 0.21650318863099285, "grad_norm": 3.187417433180603, "learning_rate": 1.1991831338424465e-06, "loss": 0.3189, "step": 836 }, { "epoch": 0.2167621637370108, "grad_norm": 5.108923083012557, "learning_rate": 1.1989596617969546e-06, "loss": 0.468, "step": 837 }, { "epoch": 0.2170211388430287, "grad_norm": 3.5092540984088227, "learning_rate": 1.198735720374837e-06, "loss": 0.2836, "step": 838 }, { "epoch": 0.21728011394904664, "grad_norm": 6.365958781576424, "learning_rate": 1.1985113097592302e-06, "loss": 0.3103, "step": 839 }, { "epoch": 0.21753908905506458, "grad_norm": 7.018918766044205, "learning_rate": 1.1982864301336541e-06, "loss": 0.3434, "step": 840 }, { "epoch": 0.21753908905506458, "eval_PRM Accuracy": 0.8679245283018868, "eval_PRM F1": 0.9125, "eval_PRM F1 AUC": 0.8528025144054479, "eval_PRM F1 AUC (fixed)": 0.7176532215819801, "eval_PRM F1 Neg": 0.7307692307692307, "eval_PRM NPV": 0.6551724137931034, "eval_PRM Precision": 0.948051948051948, "eval_PRM Recall": 0.8795180722891566, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.26224979758262634, "eval_runtime": 14.8565, "eval_samples_per_second": 2.962, "eval_steps_per_second": 0.135, "step": 840 }, { "epoch": 0.21779806416108252, "grad_norm": 3.179988198620702, "learning_rate": 1.198061081682012e-06, "loss": 0.292, "step": 841 }, { "epoch": 0.21805703926710046, "grad_norm": 5.252994006042356, "learning_rate": 1.1978352645885913e-06, "loss": 0.469, "step": 842 }, { "epoch": 0.21831601437311837, "grad_norm": 5.41230067605829, "learning_rate": 1.1976089790380619e-06, "loss": 0.3918, "step": 843 }, { "epoch": 0.2185749894791363, "grad_norm": 6.237162379573823, "learning_rate": 1.197382225215477e-06, "loss": 0.4238, "step": 844 }, { "epoch": 0.21883396458515425, "grad_norm": 5.260596050836425, "learning_rate": 1.1971550033062737e-06, "loss": 0.437, "step": 845 }, { "epoch": 0.21883396458515425, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8903225806451613, "eval_PRM F1 AUC": 0.8504452592980618, "eval_PRM F1 AUC (fixed)": 0.7082242011524359, "eval_PRM F1 Neg": 0.7017543859649122, "eval_PRM NPV": 0.5882352941176471, "eval_PRM Precision": 0.9583333333333334, "eval_PRM Recall": 0.8313253012048193, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.2731127440929413, "eval_runtime": 15.0799, "eval_samples_per_second": 2.918, "eval_steps_per_second": 0.133, "step": 845 }, { "epoch": 0.2190929396911722, "grad_norm": 3.8809428532933423, "learning_rate": 1.19692731349627e-06, "loss": 0.4406, "step": 846 }, { "epoch": 0.21935191479719013, "grad_norm": 10.17440956775982, "learning_rate": 1.1966991559716685e-06, "loss": 0.3781, "step": 847 }, { "epoch": 0.21961088990320807, "grad_norm": 8.224360422397567, "learning_rate": 1.1964705309190531e-06, "loss": 0.3955, "step": 848 }, { "epoch": 0.21986986500922598, "grad_norm": 6.917395590081431, "learning_rate": 1.1962414385253898e-06, "loss": 0.3605, "step": 849 }, { "epoch": 0.22012884011524392, "grad_norm": 5.291250727527, "learning_rate": 1.1960118789780282e-06, "loss": 0.3373, "step": 850 }, { "epoch": 0.22012884011524392, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8903225806451613, "eval_PRM F1 AUC": 0.8504452592980618, "eval_PRM F1 AUC (fixed)": 0.7048192771084338, "eval_PRM F1 Neg": 0.7017543859649122, "eval_PRM NPV": 0.5882352941176471, "eval_PRM Precision": 0.9583333333333334, "eval_PRM Recall": 0.8313253012048193, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.2853865921497345, "eval_runtime": 14.6096, "eval_samples_per_second": 3.012, "eval_steps_per_second": 0.137, "step": 850 }, { "epoch": 0.22038781522126186, "grad_norm": 4.169023666456844, "learning_rate": 1.1957818524646987e-06, "loss": 0.4194, "step": 851 }, { "epoch": 0.2206467903272798, "grad_norm": 3.8927309085937742, "learning_rate": 1.1955513591735137e-06, "loss": 0.3856, "step": 852 }, { "epoch": 0.22090576543329773, "grad_norm": 4.616525144202818, "learning_rate": 1.195320399292968e-06, "loss": 0.3722, "step": 853 }, { "epoch": 0.22116474053931565, "grad_norm": 4.846970920598354, "learning_rate": 1.1950889730119374e-06, "loss": 0.4251, "step": 854 }, { "epoch": 0.22142371564533359, "grad_norm": 3.389416043470305, "learning_rate": 1.194857080519679e-06, "loss": 0.3592, "step": 855 }, { "epoch": 0.22142371564533359, "eval_PRM Accuracy": 0.8867924528301887, "eval_PRM F1": 0.925, "eval_PRM F1 AUC": 0.8805657412257727, "eval_PRM F1 AUC (fixed)": 0.7032477737035098, "eval_PRM F1 Neg": 0.7692307692307693, "eval_PRM NPV": 0.6896551724137931, "eval_PRM Precision": 0.961038961038961, "eval_PRM Recall": 0.891566265060241, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.26908186078071594, "eval_runtime": 14.562, "eval_samples_per_second": 3.022, "eval_steps_per_second": 0.137, "step": 855 }, { "epoch": 0.22168269075135152, "grad_norm": 5.107553289534113, "learning_rate": 1.1946247220058315e-06, "loss": 0.3464, "step": 856 }, { "epoch": 0.22194166585736946, "grad_norm": 4.546089365240697, "learning_rate": 1.194391897660415e-06, "loss": 0.4066, "step": 857 }, { "epoch": 0.2222006409633874, "grad_norm": 3.0429013571459382, "learning_rate": 1.1941586076738298e-06, "loss": 0.3494, "step": 858 }, { "epoch": 0.22245961606940534, "grad_norm": 4.515604834735779, "learning_rate": 1.1939248522368574e-06, "loss": 0.3247, "step": 859 }, { "epoch": 0.22271859117542325, "grad_norm": 6.328796625019218, "learning_rate": 1.1936906315406602e-06, "loss": 0.4285, "step": 860 }, { "epoch": 0.22271859117542325, "eval_PRM Accuracy": 0.8773584905660378, "eval_PRM F1": 0.9202453987730062, "eval_PRM F1 AUC": 0.8431115767417497, "eval_PRM F1 AUC (fixed)": 0.7066526977475118, "eval_PRM F1 Neg": 0.7346938775510204, "eval_PRM NPV": 0.6923076923076923, "eval_PRM Precision": 0.9375, "eval_PRM Recall": 0.9036144578313253, "eval_PRM Specificty": 0.782608695652174, "eval_loss": 0.2668800354003906, "eval_runtime": 14.4032, "eval_samples_per_second": 3.055, "eval_steps_per_second": 0.139, "step": 860 }, { "epoch": 0.2229775662814412, "grad_norm": 4.003627406455415, "learning_rate": 1.1934559457767803e-06, "loss": 0.3303, "step": 861 }, { "epoch": 0.22323654138745913, "grad_norm": 2.9568343764262015, "learning_rate": 1.193220795137141e-06, "loss": 0.3514, "step": 862 }, { "epoch": 0.22349551649347707, "grad_norm": 8.797213234245561, "learning_rate": 1.1929851798140452e-06, "loss": 0.3842, "step": 863 }, { "epoch": 0.223754491599495, "grad_norm": 2.5474802896901565, "learning_rate": 1.1927491000001762e-06, "loss": 0.2886, "step": 864 }, { "epoch": 0.22401346670551292, "grad_norm": 3.3874524772324275, "learning_rate": 1.1925125558885966e-06, "loss": 0.3237, "step": 865 }, { "epoch": 0.22401346670551292, "eval_PRM Accuracy": 0.8867924528301887, "eval_PRM F1": 0.9259259259259259, "eval_PRM F1 AUC": 0.8648507071765322, "eval_PRM F1 AUC (fixed)": 0.7056050288108958, "eval_PRM F1 Neg": 0.76, "eval_PRM NPV": 0.7037037037037037, "eval_PRM Precision": 0.9493670886075949, "eval_PRM Recall": 0.9036144578313253, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.26945021748542786, "eval_runtime": 14.3503, "eval_samples_per_second": 3.066, "eval_steps_per_second": 0.139, "step": 865 }, { "epoch": 0.22427244181153086, "grad_norm": 6.980144378433547, "learning_rate": 1.1922755476727493e-06, "loss": 0.4359, "step": 866 }, { "epoch": 0.2245314169175488, "grad_norm": 3.675259629117122, "learning_rate": 1.1920380755464567e-06, "loss": 0.4484, "step": 867 }, { "epoch": 0.22479039202356674, "grad_norm": 3.970800178023882, "learning_rate": 1.19180013970392e-06, "loss": 0.4649, "step": 868 }, { "epoch": 0.22504936712958468, "grad_norm": 5.593595197945199, "learning_rate": 1.1915617403397198e-06, "loss": 0.406, "step": 869 }, { "epoch": 0.22530834223560262, "grad_norm": 3.4055613682905475, "learning_rate": 1.1913228776488164e-06, "loss": 0.3813, "step": 870 }, { "epoch": 0.22530834223560262, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.89171974522293, "eval_PRM F1 AUC": 0.8347302252488213, "eval_PRM F1 AUC (fixed)": 0.6998428496595076, "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": 0.2825016975402832, "eval_runtime": 14.7056, "eval_samples_per_second": 2.992, "eval_steps_per_second": 0.136, "step": 870 }, { "epoch": 0.22556731734162053, "grad_norm": 4.766583139189498, "learning_rate": 1.1910835518265488e-06, "loss": 0.406, "step": 871 }, { "epoch": 0.22582629244763847, "grad_norm": 5.360701882443408, "learning_rate": 1.1908437630686337e-06, "loss": 0.2926, "step": 872 }, { "epoch": 0.2260852675536564, "grad_norm": 8.516057938457399, "learning_rate": 1.1906035115711675e-06, "loss": 0.3621, "step": 873 }, { "epoch": 0.22634424265967434, "grad_norm": 6.787723362329659, "learning_rate": 1.190362797530625e-06, "loss": 0.327, "step": 874 }, { "epoch": 0.22660321776569228, "grad_norm": 3.657201137384946, "learning_rate": 1.1901216211438586e-06, "loss": 0.3264, "step": 875 }, { "epoch": 0.22660321776569228, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.8974358974358975, "eval_PRM F1 AUC": 0.8564693556836039, "eval_PRM F1 AUC (fixed)": 0.6948664222105815, "eval_PRM F1 Neg": 0.7142857142857143, "eval_PRM NPV": 0.6060606060606061, "eval_PRM Precision": 0.958904109589041, "eval_PRM Recall": 0.8433734939759037, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.29068121314048767, "eval_runtime": 14.7585, "eval_samples_per_second": 2.981, "eval_steps_per_second": 0.136, "step": 875 }, { "epoch": 0.2268621928717102, "grad_norm": 4.078588375063669, "learning_rate": 1.1898799826080994e-06, "loss": 0.4119, "step": 876 }, { "epoch": 0.22712116797772813, "grad_norm": 3.2953125555811984, "learning_rate": 1.189637882120956e-06, "loss": 0.3044, "step": 877 }, { "epoch": 0.22738014308374607, "grad_norm": 7.182771434003031, "learning_rate": 1.1893953198804152e-06, "loss": 0.3758, "step": 878 }, { "epoch": 0.227639118189764, "grad_norm": 4.247871537462031, "learning_rate": 1.1891522960848409e-06, "loss": 0.3703, "step": 879 }, { "epoch": 0.22789809329578195, "grad_norm": 3.4826649057243224, "learning_rate": 1.1889088109329752e-06, "loss": 0.3483, "step": 880 }, { "epoch": 0.22789809329578195, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.9056603773584906, "eval_PRM F1 AUC": 0.8467784180199058, "eval_PRM F1 AUC (fixed)": 0.7103195390256679, "eval_PRM F1 Neg": 0.7169811320754716, "eval_PRM NPV": 0.6333333333333333, "eval_PRM Precision": 0.9473684210526315, "eval_PRM Recall": 0.8674698795180723, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.2728223502635956, "eval_runtime": 15.0548, "eval_samples_per_second": 2.923, "eval_steps_per_second": 0.133, "step": 880 }, { "epoch": 0.2281570684017999, "grad_norm": 5.741320223413004, "learning_rate": 1.1886648646239365e-06, "loss": 0.3574, "step": 881 }, { "epoch": 0.2284160435078178, "grad_norm": 4.062167022584372, "learning_rate": 1.188420457357221e-06, "loss": 0.4174, "step": 882 }, { "epoch": 0.22867501861383574, "grad_norm": 2.526453290531781, "learning_rate": 1.1881755893327022e-06, "loss": 0.317, "step": 883 }, { "epoch": 0.22893399371985368, "grad_norm": 3.0807541085440304, "learning_rate": 1.1879302607506295e-06, "loss": 0.3288, "step": 884 }, { "epoch": 0.22919296882587162, "grad_norm": 3.8308465141169763, "learning_rate": 1.1876844718116293e-06, "loss": 0.4456, "step": 885 }, { "epoch": 0.22919296882587162, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.9056603773584906, "eval_PRM F1 AUC": 0.8467784180199058, "eval_PRM F1 AUC (fixed)": 0.71634363541121, "eval_PRM F1 Neg": 0.7169811320754716, "eval_PRM NPV": 0.6333333333333333, "eval_PRM Precision": 0.9473684210526315, "eval_PRM Recall": 0.8674698795180723, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.26763397455215454, "eval_runtime": 14.9954, "eval_samples_per_second": 2.934, "eval_steps_per_second": 0.133, "step": 885 }, { "epoch": 0.22945194393188956, "grad_norm": 3.024120524976241, "learning_rate": 1.1874382227167049e-06, "loss": 0.3371, "step": 886 }, { "epoch": 0.22971091903790747, "grad_norm": 3.82411760033903, "learning_rate": 1.1871915136672352e-06, "loss": 0.4065, "step": 887 }, { "epoch": 0.2299698941439254, "grad_norm": 3.8886412588863, "learning_rate": 1.1869443448649758e-06, "loss": 0.269, "step": 888 }, { "epoch": 0.23022886924994335, "grad_norm": 4.071708730287361, "learning_rate": 1.1866967165120583e-06, "loss": 0.2922, "step": 889 }, { "epoch": 0.23048784435596129, "grad_norm": 5.072664995850144, "learning_rate": 1.1864486288109894e-06, "loss": 0.4196, "step": 890 }, { "epoch": 0.23048784435596129, "eval_PRM Accuracy": 0.8679245283018868, "eval_PRM F1": 0.9113924050632911, "eval_PRM F1 AUC": 0.8685175484546883, "eval_PRM F1 AUC (fixed)": 0.7239392352016762, "eval_PRM F1 Neg": 0.7407407407407407, "eval_PRM NPV": 0.6451612903225806, "eval_PRM Precision": 0.96, "eval_PRM Recall": 0.8674698795180723, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.2615525424480438, "eval_runtime": 14.7332, "eval_samples_per_second": 2.986, "eval_steps_per_second": 0.136, "step": 890 }, { "epoch": 0.23074681946197922, "grad_norm": 3.0759707900811657, "learning_rate": 1.186200081964652e-06, "loss": 0.3235, "step": 891 }, { "epoch": 0.23100579456799716, "grad_norm": 4.164881209108914, "learning_rate": 1.185951076176305e-06, "loss": 0.3374, "step": 892 }, { "epoch": 0.23126476967401507, "grad_norm": 4.746338390873242, "learning_rate": 1.1857016116495815e-06, "loss": 0.3589, "step": 893 }, { "epoch": 0.23152374478003301, "grad_norm": 4.619158986837625, "learning_rate": 1.1854516885884903e-06, "loss": 0.2584, "step": 894 }, { "epoch": 0.23178271988605095, "grad_norm": 3.02965756184151, "learning_rate": 1.1852013071974155e-06, "loss": 0.3908, "step": 895 }, { "epoch": 0.23178271988605095, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.9044585987261147, "eval_PRM F1 AUC": 0.8624934520691462, "eval_PRM F1 AUC (fixed)": 0.7202723939235202, "eval_PRM F1 Neg": 0.7272727272727273, "eval_PRM NPV": 0.625, "eval_PRM Precision": 0.9594594594594594, "eval_PRM Recall": 0.8554216867469879, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.2613525986671448, "eval_runtime": 14.5307, "eval_samples_per_second": 3.028, "eval_steps_per_second": 0.138, "step": 895 }, { "epoch": 0.2320416949920689, "grad_norm": 6.098313346871557, "learning_rate": 1.1849504676811156e-06, "loss": 0.3305, "step": 896 }, { "epoch": 0.23230067009808683, "grad_norm": 6.268537137431423, "learning_rate": 1.1846991702447233e-06, "loss": 0.3817, "step": 897 }, { "epoch": 0.23255964520410474, "grad_norm": 4.935579305034822, "learning_rate": 1.184447415093747e-06, "loss": 0.3596, "step": 898 }, { "epoch": 0.23281862031012268, "grad_norm": 4.467145665647315, "learning_rate": 1.1841952024340688e-06, "loss": 0.416, "step": 899 }, { "epoch": 0.23307759541614062, "grad_norm": 4.170804728572236, "learning_rate": 1.1839425324719442e-06, "loss": 0.3559, "step": 900 }, { "epoch": 0.23307759541614062, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.8987341772151899, "eval_PRM F1 AUC": 0.8407543216343635, "eval_PRM F1 AUC (fixed)": 0.71503404924044, "eval_PRM F1 Neg": 0.7037037037037037, "eval_PRM NPV": 0.6129032258064516, "eval_PRM Precision": 0.9466666666666667, "eval_PRM Recall": 0.8554216867469879, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.2619946002960205, "eval_runtime": 14.6743, "eval_samples_per_second": 2.998, "eval_steps_per_second": 0.136, "step": 900 }, { "epoch": 0.23333657052215856, "grad_norm": 3.561851713362729, "learning_rate": 1.183689405414004e-06, "loss": 0.37, "step": 901 }, { "epoch": 0.2335955456281765, "grad_norm": 5.391340666602356, "learning_rate": 1.183435821467252e-06, "loss": 0.2998, "step": 902 }, { "epoch": 0.23385452073419444, "grad_norm": 4.479966992834465, "learning_rate": 1.1831817808390655e-06, "loss": 0.4816, "step": 903 }, { "epoch": 0.23411349584021235, "grad_norm": 3.788280129003505, "learning_rate": 1.1829272837371958e-06, "loss": 0.4445, "step": 904 }, { "epoch": 0.2343724709462303, "grad_norm": 6.620416351536968, "learning_rate": 1.1826723303697673e-06, "loss": 0.4148, "step": 905 }, { "epoch": 0.2343724709462303, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.9056603773584906, "eval_PRM F1 AUC": 0.8467784180199058, "eval_PRM F1 AUC (fixed)": 0.7116291251964378, "eval_PRM F1 Neg": 0.7169811320754716, "eval_PRM NPV": 0.6333333333333333, "eval_PRM Precision": 0.9473684210526315, "eval_PRM Recall": 0.8674698795180723, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.2720625102519989, "eval_runtime": 15.08, "eval_samples_per_second": 2.918, "eval_steps_per_second": 0.133, "step": 905 }, { "epoch": 0.23463144605224823, "grad_norm": 3.4257473666984324, "learning_rate": 1.1824169209452774e-06, "loss": 0.3041, "step": 906 }, { "epoch": 0.23489042115826617, "grad_norm": 4.939739069415202, "learning_rate": 1.1821610556725966e-06, "loss": 0.435, "step": 907 }, { "epoch": 0.2351493962642841, "grad_norm": 5.255878093381102, "learning_rate": 1.181904734760968e-06, "loss": 0.4064, "step": 908 }, { "epoch": 0.23540837137030202, "grad_norm": 8.85032004547169, "learning_rate": 1.1816479584200078e-06, "loss": 0.3155, "step": 909 }, { "epoch": 0.23566734647631996, "grad_norm": 3.6772122279451294, "learning_rate": 1.1813907268597038e-06, "loss": 0.3787, "step": 910 }, { "epoch": 0.23566734647631996, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.9056603773584906, "eval_PRM F1 AUC": 0.8467784180199058, "eval_PRM F1 AUC (fixed)": 0.712676794133054, "eval_PRM F1 Neg": 0.7169811320754716, "eval_PRM NPV": 0.6333333333333333, "eval_PRM Precision": 0.9473684210526315, "eval_PRM Recall": 0.8674698795180723, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.280823677778244, "eval_runtime": 13.5068, "eval_samples_per_second": 3.258, "eval_steps_per_second": 0.148, "step": 910 }, { "epoch": 0.2359263215823379, "grad_norm": 4.360653413098666, "learning_rate": 1.1811330402904171e-06, "loss": 0.3677, "step": 911 }, { "epoch": 0.23618529668835583, "grad_norm": 3.384814109696639, "learning_rate": 1.18087489892288e-06, "loss": 0.313, "step": 912 }, { "epoch": 0.23644427179437377, "grad_norm": 3.0761177527594104, "learning_rate": 1.1806163029681972e-06, "loss": 0.2976, "step": 913 }, { "epoch": 0.2367032469003917, "grad_norm": 3.0945972922107345, "learning_rate": 1.1803572526378453e-06, "loss": 0.3056, "step": 914 }, { "epoch": 0.23696222200640962, "grad_norm": 7.625548103055213, "learning_rate": 1.1800977481436718e-06, "loss": 0.3014, "step": 915 }, { "epoch": 0.23696222200640962, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.9056603773584906, "eval_PRM F1 AUC": 0.8467784180199058, "eval_PRM F1 AUC (fixed)": 0.7210581456259822, "eval_PRM F1 Neg": 0.7169811320754716, "eval_PRM NPV": 0.6333333333333333, "eval_PRM Precision": 0.9473684210526315, "eval_PRM Recall": 0.8674698795180723, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.2804136276245117, "eval_runtime": 14.4877, "eval_samples_per_second": 3.037, "eval_steps_per_second": 0.138, "step": 915 }, { "epoch": 0.23722119711242756, "grad_norm": 4.106895071256707, "learning_rate": 1.1798377896978964e-06, "loss": 0.3447, "step": 916 }, { "epoch": 0.2374801722184455, "grad_norm": 3.4879165088232758, "learning_rate": 1.1795773775131094e-06, "loss": 0.3609, "step": 917 }, { "epoch": 0.23773914732446344, "grad_norm": 4.94351615997525, "learning_rate": 1.1793165118022725e-06, "loss": 0.464, "step": 918 }, { "epoch": 0.23799812243048138, "grad_norm": 9.395892112887031, "learning_rate": 1.1790551927787186e-06, "loss": 0.3857, "step": 919 }, { "epoch": 0.2382570975364993, "grad_norm": 4.4953539881083575, "learning_rate": 1.1787934206561506e-06, "loss": 0.3387, "step": 920 }, { "epoch": 0.2382570975364993, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.8987341772151899, "eval_PRM F1 AUC": 0.8407543216343635, "eval_PRM F1 AUC (fixed)": 0.7482975379779989, "eval_PRM F1 Neg": 0.7037037037037037, "eval_PRM NPV": 0.6129032258064516, "eval_PRM Precision": 0.9466666666666667, "eval_PRM Recall": 0.8554216867469879, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.2711288332939148, "eval_runtime": 14.5297, "eval_samples_per_second": 3.028, "eval_steps_per_second": 0.138, "step": 920 }, { "epoch": 0.23851607264251723, "grad_norm": 4.288113954542933, "learning_rate": 1.178531195648642e-06, "loss": 0.3936, "step": 921 }, { "epoch": 0.23877504774853517, "grad_norm": 4.977864839923052, "learning_rate": 1.1782685179706375e-06, "loss": 0.4172, "step": 922 }, { "epoch": 0.2390340228545531, "grad_norm": 6.67201876896054, "learning_rate": 1.1780053878369513e-06, "loss": 0.388, "step": 923 }, { "epoch": 0.23929299796057105, "grad_norm": 6.880680332320413, "learning_rate": 1.1777418054627677e-06, "loss": 0.3382, "step": 924 }, { "epoch": 0.23955197306658899, "grad_norm": 3.9508208402930105, "learning_rate": 1.1774777710636406e-06, "loss": 0.3966, "step": 925 }, { "epoch": 0.23955197306658899, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8888888888888888, "eval_PRM F1 AUC": 0.8661602933473023, "eval_PRM F1 AUC (fixed)": 0.7482975379779989, "eval_PRM F1 Neg": 0.711864406779661, "eval_PRM NPV": 0.5833333333333334, "eval_PRM Precision": 0.9714285714285714, "eval_PRM Recall": 0.8192771084337349, "eval_PRM Specificty": 0.9130434782608695, "eval_loss": 0.2894640564918518, "eval_runtime": 14.5853, "eval_samples_per_second": 3.017, "eval_steps_per_second": 0.137, "step": 925 }, { "epoch": 0.2398109481726069, "grad_norm": 10.535536833190756, "learning_rate": 1.1772132848554942e-06, "loss": 0.3728, "step": 926 }, { "epoch": 0.24006992327862484, "grad_norm": 6.648969924884001, "learning_rate": 1.1769483470546216e-06, "loss": 0.3397, "step": 927 }, { "epoch": 0.24032889838464278, "grad_norm": 3.638658941745989, "learning_rate": 1.1766829578776858e-06, "loss": 0.3116, "step": 928 }, { "epoch": 0.24058787349066071, "grad_norm": 6.1916710161098525, "learning_rate": 1.176417117541718e-06, "loss": 0.3743, "step": 929 }, { "epoch": 0.24084684859667865, "grad_norm": 6.466301555599048, "learning_rate": 1.1761508262641195e-06, "loss": 0.432, "step": 930 }, { "epoch": 0.24084684859667865, "eval_PRM Accuracy": 0.8679245283018868, "eval_PRM F1": 0.9102564102564102, "eval_PRM F1 AUC": 0.8842325825039288, "eval_PRM F1 AUC (fixed)": 0.7551073860660031, "eval_PRM F1 Neg": 0.75, "eval_PRM NPV": 0.6363636363636364, "eval_PRM Precision": 0.9726027397260274, "eval_PRM Recall": 0.8554216867469879, "eval_PRM Specificty": 0.9130434782608695, "eval_loss": 0.26407501101493835, "eval_runtime": 14.429, "eval_samples_per_second": 3.049, "eval_steps_per_second": 0.139, "step": 930 }, { "epoch": 0.24110582370269656, "grad_norm": 3.5072923022044473, "learning_rate": 1.1758840842626593e-06, "loss": 0.3742, "step": 931 }, { "epoch": 0.2413647988087145, "grad_norm": 4.137152539221647, "learning_rate": 1.1756168917554759e-06, "loss": 0.3656, "step": 932 }, { "epoch": 0.24162377391473244, "grad_norm": 3.7639799069758637, "learning_rate": 1.1753492489610755e-06, "loss": 0.3303, "step": 933 }, { "epoch": 0.24188274902075038, "grad_norm": 2.896848508314599, "learning_rate": 1.175081156098333e-06, "loss": 0.3401, "step": 934 }, { "epoch": 0.24214172412676832, "grad_norm": 9.300379803883216, "learning_rate": 1.1748126133864908e-06, "loss": 0.4256, "step": 935 }, { "epoch": 0.24214172412676832, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.8987341772151899, "eval_PRM F1 AUC": 0.8407543216343635, "eval_PRM F1 AUC (fixed)": 0.7503928758512309, "eval_PRM F1 Neg": 0.7037037037037037, "eval_PRM NPV": 0.6129032258064516, "eval_PRM Precision": 0.9466666666666667, "eval_PRM Recall": 0.8554216867469879, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.25802791118621826, "eval_runtime": 14.7761, "eval_samples_per_second": 2.978, "eval_steps_per_second": 0.135, "step": 935 }, { "epoch": 0.24240069923278626, "grad_norm": 6.945735757870045, "learning_rate": 1.1745436210451603e-06, "loss": 0.403, "step": 936 }, { "epoch": 0.24265967433880417, "grad_norm": 5.168673525936932, "learning_rate": 1.1742741792943193e-06, "loss": 0.3643, "step": 937 }, { "epoch": 0.2429186494448221, "grad_norm": 3.9301017982836597, "learning_rate": 1.1740042883543139e-06, "loss": 0.2894, "step": 938 }, { "epoch": 0.24317762455084005, "grad_norm": 3.8086762466056783, "learning_rate": 1.1737339484458574e-06, "loss": 0.3478, "step": 939 }, { "epoch": 0.243436599656858, "grad_norm": 10.598569677321471, "learning_rate": 1.1734631597900301e-06, "loss": 0.4391, "step": 940 }, { "epoch": 0.243436599656858, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.881578947368421, "eval_PRM F1 AUC": 0.8601361969617601, "eval_PRM F1 AUC (fixed)": 0.7454164484023048, "eval_PRM F1 Neg": 0.7, "eval_PRM NPV": 0.5675675675675675, "eval_PRM Precision": 0.9710144927536232, "eval_PRM Recall": 0.8072289156626506, "eval_PRM Specificty": 0.9130434782608695, "eval_loss": 0.26957109570503235, "eval_runtime": 14.7665, "eval_samples_per_second": 2.98, "eval_steps_per_second": 0.135, "step": 940 }, { "epoch": 0.24369557476287593, "grad_norm": 5.764724413638612, "learning_rate": 1.1731919226082793e-06, "loss": 0.365, "step": 941 }, { "epoch": 0.24395454986889387, "grad_norm": 5.9502788053858895, "learning_rate": 1.1729202371224194e-06, "loss": 0.3258, "step": 942 }, { "epoch": 0.24421352497491178, "grad_norm": 6.031412246354583, "learning_rate": 1.1726481035546311e-06, "loss": 0.3288, "step": 943 }, { "epoch": 0.24447250008092972, "grad_norm": 3.945656622656256, "learning_rate": 1.1723755221274619e-06, "loss": 0.28, "step": 944 }, { "epoch": 0.24473147518694766, "grad_norm": 3.3620655142691787, "learning_rate": 1.1721024930638248e-06, "loss": 0.34, "step": 945 }, { "epoch": 0.24473147518694766, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.8974358974358975, "eval_PRM F1 AUC": 0.8564693556836039, "eval_PRM F1 AUC (fixed)": 0.7407019381875327, "eval_PRM F1 Neg": 0.7142857142857143, "eval_PRM NPV": 0.6060606060606061, "eval_PRM Precision": 0.958904109589041, "eval_PRM Recall": 0.8433734939759037, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.2563490867614746, "eval_runtime": 14.652, "eval_samples_per_second": 3.003, "eval_steps_per_second": 0.136, "step": 945 }, { "epoch": 0.2449904502929656, "grad_norm": 3.5293565846531973, "learning_rate": 1.1718290165869995e-06, "loss": 0.3143, "step": 946 }, { "epoch": 0.24524942539898353, "grad_norm": 3.4890087445954756, "learning_rate": 1.1715550929206317e-06, "loss": 0.3351, "step": 947 }, { "epoch": 0.24550840050500145, "grad_norm": 2.986678147022999, "learning_rate": 1.1712807222887324e-06, "loss": 0.3392, "step": 948 }, { "epoch": 0.24576737561101938, "grad_norm": 11.359804178521635, "learning_rate": 1.1710059049156783e-06, "loss": 0.4761, "step": 949 }, { "epoch": 0.24602635071703732, "grad_norm": 10.1120871833462, "learning_rate": 1.1707306410262112e-06, "loss": 0.4559, "step": 950 }, { "epoch": 0.24602635071703732, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8930817610062893, "eval_PRM F1 AUC": 0.819015191199581, "eval_PRM F1 AUC (fixed)": 0.7393923520167627, "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": 0.2529611885547638, "eval_runtime": 14.6632, "eval_samples_per_second": 3.001, "eval_steps_per_second": 0.136, "step": 950 }, { "epoch": 0.24628532582305526, "grad_norm": 6.21559932436371, "learning_rate": 1.1704549308454385e-06, "loss": 0.3531, "step": 951 }, { "epoch": 0.2465443009290732, "grad_norm": 4.139390120568164, "learning_rate": 1.1701787745988325e-06, "loss": 0.4025, "step": 952 }, { "epoch": 0.24680327603509114, "grad_norm": 4.066293812063238, "learning_rate": 1.1699021725122297e-06, "loss": 0.3229, "step": 953 }, { "epoch": 0.24706225114110905, "grad_norm": 2.899098952795315, "learning_rate": 1.169625124811832e-06, "loss": 0.3156, "step": 954 }, { "epoch": 0.247321226247127, "grad_norm": 2.8517635816301694, "learning_rate": 1.169347631724205e-06, "loss": 0.3273, "step": 955 }, { "epoch": 0.247321226247127, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.8974358974358975, "eval_PRM F1 AUC": 0.8564693556836039, "eval_PRM F1 AUC (fixed)": 0.7333682556312205, "eval_PRM F1 Neg": 0.7142857142857143, "eval_PRM NPV": 0.6060606060606061, "eval_PRM Precision": 0.958904109589041, "eval_PRM Recall": 0.8433734939759037, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.2633054554462433, "eval_runtime": 14.3335, "eval_samples_per_second": 3.07, "eval_steps_per_second": 0.14, "step": 955 }, { "epoch": 0.24758020135314493, "grad_norm": 3.3564577121473618, "learning_rate": 1.1690696934762794e-06, "loss": 0.3227, "step": 956 }, { "epoch": 0.24783917645916287, "grad_norm": 3.8662849481785635, "learning_rate": 1.168791310295349e-06, "loss": 0.357, "step": 957 }, { "epoch": 0.2480981515651808, "grad_norm": 3.5997682609517634, "learning_rate": 1.1685124824090719e-06, "loss": 0.417, "step": 958 }, { "epoch": 0.24835712667119872, "grad_norm": 2.7034258306479293, "learning_rate": 1.1682332100454699e-06, "loss": 0.3167, "step": 959 }, { "epoch": 0.24861610177721666, "grad_norm": 4.988676707709703, "learning_rate": 1.1679534934329283e-06, "loss": 0.2986, "step": 960 }, { "epoch": 0.24861610177721666, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.9, "eval_PRM F1 AUC": 0.8250392875851232, "eval_PRM F1 AUC (fixed)": 0.7354635935044526, "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": 0.26185938715934753, "eval_runtime": 14.5293, "eval_samples_per_second": 3.028, "eval_steps_per_second": 0.138, "step": 960 }, { "epoch": 0.2488750768832346, "grad_norm": 5.29107206320586, "learning_rate": 1.167673332800196e-06, "loss": 0.4215, "step": 961 }, { "epoch": 0.24913405198925254, "grad_norm": 5.437935849441125, "learning_rate": 1.167392728376384e-06, "loss": 0.4476, "step": 962 }, { "epoch": 0.24939302709527048, "grad_norm": 3.8839556426836372, "learning_rate": 1.1671116803909674e-06, "loss": 0.3246, "step": 963 }, { "epoch": 0.24965200220128841, "grad_norm": 10.37270496386567, "learning_rate": 1.1668301890737833e-06, "loss": 0.4899, "step": 964 }, { "epoch": 0.24991097730730633, "grad_norm": 5.908063708001343, "learning_rate": 1.1665482546550314e-06, "loss": 0.3955, "step": 965 }, { "epoch": 0.24991097730730633, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.89171974522293, "eval_PRM F1 AUC": 0.8347302252488213, "eval_PRM F1 AUC (fixed)": 0.7260345730749084, "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": 0.2727225124835968, "eval_runtime": 14.3383, "eval_samples_per_second": 3.069, "eval_steps_per_second": 0.139, "step": 965 }, { "epoch": 0.2501699524133243, "grad_norm": 5.6447939854256735, "learning_rate": 1.1662658773652744e-06, "loss": 0.4093, "step": 966 }, { "epoch": 0.2504289275193422, "grad_norm": 2.850052957391009, "learning_rate": 1.1659830574354363e-06, "loss": 0.3873, "step": 967 }, { "epoch": 0.2506879026253601, "grad_norm": 4.787640638975102, "learning_rate": 1.1656997950968037e-06, "loss": 0.4238, "step": 968 }, { "epoch": 0.2509468777313781, "grad_norm": 5.680659064340013, "learning_rate": 1.1654160905810246e-06, "loss": 0.4614, "step": 969 }, { "epoch": 0.251205852837396, "grad_norm": 3.3362762459579023, "learning_rate": 1.1651319441201094e-06, "loss": 0.344, "step": 970 }, { "epoch": 0.251205852837396, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.89171974522293, "eval_PRM F1 AUC": 0.8347302252488213, "eval_PRM F1 AUC (fixed)": 0.716081718177056, "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": 0.28485599160194397, "eval_runtime": 14.3151, "eval_samples_per_second": 3.074, "eval_steps_per_second": 0.14, "step": 970 }, { "epoch": 0.25146482794341396, "grad_norm": 5.758747112832516, "learning_rate": 1.1648473559464287e-06, "loss": 0.3374, "step": 971 }, { "epoch": 0.25172380304943187, "grad_norm": 8.58153415473367, "learning_rate": 1.164562326292715e-06, "loss": 0.538, "step": 972 }, { "epoch": 0.2519827781554498, "grad_norm": 5.950271313129463, "learning_rate": 1.1642768553920621e-06, "loss": 0.3176, "step": 973 }, { "epoch": 0.25224175326146775, "grad_norm": 3.1460089924178947, "learning_rate": 1.163990943477924e-06, "loss": 0.3241, "step": 974 }, { "epoch": 0.25250072836748566, "grad_norm": 6.985297468419992, "learning_rate": 1.163704590784116e-06, "loss": 0.3705, "step": 975 }, { "epoch": 0.25250072836748566, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.89171974522293, "eval_PRM F1 AUC": 0.8347302252488213, "eval_PRM F1 AUC (fixed)": 0.7087480356207437, "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": 0.27933061122894287, "eval_runtime": 14.5344, "eval_samples_per_second": 3.027, "eval_steps_per_second": 0.138, "step": 975 }, { "epoch": 0.2527597034735036, "grad_norm": 8.012879551894242, "learning_rate": 1.1634177975448133e-06, "loss": 0.3826, "step": 976 }, { "epoch": 0.25301867857952154, "grad_norm": 7.96887475220341, "learning_rate": 1.1631305639945517e-06, "loss": 0.3364, "step": 977 }, { "epoch": 0.25327765368553945, "grad_norm": 3.612692739178533, "learning_rate": 1.162842890368227e-06, "loss": 0.4073, "step": 978 }, { "epoch": 0.2535366287915574, "grad_norm": 7.720990027200896, "learning_rate": 1.162554776901095e-06, "loss": 0.3994, "step": 979 }, { "epoch": 0.25379560389757533, "grad_norm": 2.783639066391506, "learning_rate": 1.1622662238287708e-06, "loss": 0.3031, "step": 980 }, { "epoch": 0.25379560389757533, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.9056603773584906, "eval_PRM F1 AUC": 0.8467784180199058, "eval_PRM F1 AUC (fixed)": 0.7145102147721321, "eval_PRM F1 Neg": 0.7169811320754716, "eval_PRM NPV": 0.6333333333333333, "eval_PRM Precision": 0.9473684210526315, "eval_PRM Recall": 0.8674698795180723, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.27657219767570496, "eval_runtime": 14.9306, "eval_samples_per_second": 2.947, "eval_steps_per_second": 0.134, "step": 980 }, { "epoch": 0.2540545790035933, "grad_norm": 3.4357231189414783, "learning_rate": 1.1619772313872295e-06, "loss": 0.3553, "step": 981 }, { "epoch": 0.2543135541096112, "grad_norm": 4.935806590676655, "learning_rate": 1.1616877998128053e-06, "loss": 0.3536, "step": 982 }, { "epoch": 0.2545725292156292, "grad_norm": 3.193095180565718, "learning_rate": 1.1613979293421918e-06, "loss": 0.2857, "step": 983 }, { "epoch": 0.2548315043216471, "grad_norm": 5.53188970528004, "learning_rate": 1.1611076202124404e-06, "loss": 0.3127, "step": 984 }, { "epoch": 0.255090479427665, "grad_norm": 3.229345173101998, "learning_rate": 1.160816872660963e-06, "loss": 0.4002, "step": 985 }, { "epoch": 0.255090479427665, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.89171974522293, "eval_PRM F1 AUC": 0.8347302252488213, "eval_PRM F1 AUC (fixed)": 0.7160817181770561, "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": 0.27515214681625366, "eval_runtime": 14.4985, "eval_samples_per_second": 3.035, "eval_steps_per_second": 0.138, "step": 985 }, { "epoch": 0.25534945453368296, "grad_norm": 4.293361296415176, "learning_rate": 1.1605256869255285e-06, "loss": 0.3489, "step": 986 }, { "epoch": 0.2556084296397009, "grad_norm": 4.129626468223985, "learning_rate": 1.1602340632442652e-06, "loss": 0.3799, "step": 987 }, { "epoch": 0.25586740474571884, "grad_norm": 9.096540369766284, "learning_rate": 1.1599420018556586e-06, "loss": 0.4538, "step": 988 }, { "epoch": 0.25612637985173675, "grad_norm": 5.247960343985199, "learning_rate": 1.159649502998553e-06, "loss": 0.3674, "step": 989 }, { "epoch": 0.25638535495775466, "grad_norm": 7.994078985517792, "learning_rate": 1.1593565669121497e-06, "loss": 0.3333, "step": 990 }, { "epoch": 0.25638535495775466, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8903225806451613, "eval_PRM F1 AUC": 0.8504452592980618, "eval_PRM F1 AUC (fixed)": 0.7239392352016761, "eval_PRM F1 Neg": 0.7017543859649122, "eval_PRM NPV": 0.5882352941176471, "eval_PRM Precision": 0.9583333333333334, "eval_PRM Recall": 0.8313253012048193, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.287078857421875, "eval_runtime": 14.6202, "eval_samples_per_second": 3.01, "eval_steps_per_second": 0.137, "step": 990 }, { "epoch": 0.25664433006377263, "grad_norm": 8.594303185516962, "learning_rate": 1.1590631938360086e-06, "loss": 0.3525, "step": 991 }, { "epoch": 0.25690330516979054, "grad_norm": 7.048469269863221, "learning_rate": 1.158769384010046e-06, "loss": 0.3893, "step": 992 }, { "epoch": 0.2571622802758085, "grad_norm": 13.317768952534582, "learning_rate": 1.1584751376745355e-06, "loss": 0.3186, "step": 993 }, { "epoch": 0.2574212553818264, "grad_norm": 5.726905993306342, "learning_rate": 1.1581804550701078e-06, "loss": 0.3263, "step": 994 }, { "epoch": 0.25768023048784433, "grad_norm": 4.893574502139262, "learning_rate": 1.157885336437751e-06, "loss": 0.4972, "step": 995 }, { "epoch": 0.25768023048784433, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.89171974522293, "eval_PRM F1 AUC": 0.8347302252488213, "eval_PRM F1 AUC (fixed)": 0.7409638554216867, "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": 0.2642107307910919, "eval_runtime": 14.5194, "eval_samples_per_second": 3.03, "eval_steps_per_second": 0.138, "step": 995 }, { "epoch": 0.2579392055938623, "grad_norm": 3.5760726549084616, "learning_rate": 1.1575897820188087e-06, "loss": 0.3395, "step": 996 }, { "epoch": 0.2581981806998802, "grad_norm": 4.246352890964505, "learning_rate": 1.1572937920549816e-06, "loss": 0.4059, "step": 997 }, { "epoch": 0.2584571558058982, "grad_norm": 5.18629173452879, "learning_rate": 1.1569973667883264e-06, "loss": 0.3453, "step": 998 }, { "epoch": 0.2587161309119161, "grad_norm": 3.3078206651369015, "learning_rate": 1.1567005064612558e-06, "loss": 0.3284, "step": 999 }, { "epoch": 0.25897510601793405, "grad_norm": 3.0615848926933, "learning_rate": 1.156403211316538e-06, "loss": 0.2652, "step": 1000 }, { "epoch": 0.25897510601793405, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.9056603773584906, "eval_PRM F1 AUC": 0.8467784180199058, "eval_PRM F1 AUC (fixed)": 0.7443687794656888, "eval_PRM F1 Neg": 0.7169811320754716, "eval_PRM NPV": 0.6333333333333333, "eval_PRM Precision": 0.9473684210526315, "eval_PRM Recall": 0.8674698795180723, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.25816041231155396, "eval_runtime": 14.3987, "eval_samples_per_second": 3.056, "eval_steps_per_second": 0.139, "step": 1000 }, { "epoch": 0.25923408112395196, "grad_norm": 8.62964960291262, "learning_rate": 1.1561054815972973e-06, "loss": 0.4371, "step": 1001 }, { "epoch": 0.2594930562299699, "grad_norm": 3.752764520211669, "learning_rate": 1.155807317547013e-06, "loss": 0.3212, "step": 1002 }, { "epoch": 0.25975203133598784, "grad_norm": 11.589612656246816, "learning_rate": 1.1555087194095199e-06, "loss": 0.3769, "step": 1003 }, { "epoch": 0.26001100644200575, "grad_norm": 3.8830914204755986, "learning_rate": 1.1552096874290074e-06, "loss": 0.3566, "step": 1004 }, { "epoch": 0.2602699815480237, "grad_norm": 4.374775346676012, "learning_rate": 1.1549102218500199e-06, "loss": 0.3932, "step": 1005 }, { "epoch": 0.2602699815480237, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.9056603773584906, "eval_PRM F1 AUC": 0.8467784180199058, "eval_PRM F1 AUC (fixed)": 0.7412257726558408, "eval_PRM F1 Neg": 0.7169811320754716, "eval_PRM NPV": 0.6333333333333333, "eval_PRM Precision": 0.9473684210526315, "eval_PRM Recall": 0.8674698795180723, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.2624988853931427, "eval_runtime": 14.6088, "eval_samples_per_second": 3.012, "eval_steps_per_second": 0.137, "step": 1005 }, { "epoch": 0.26052895665404163, "grad_norm": 6.288991137216826, "learning_rate": 1.1546103229174566e-06, "loss": 0.3921, "step": 1006 }, { "epoch": 0.26078793176005954, "grad_norm": 3.4011240081976823, "learning_rate": 1.1543099908765706e-06, "loss": 0.3222, "step": 1007 }, { "epoch": 0.2610469068660775, "grad_norm": 8.545602163381849, "learning_rate": 1.1540092259729698e-06, "loss": 0.3048, "step": 1008 }, { "epoch": 0.2613058819720954, "grad_norm": 3.5424382509880132, "learning_rate": 1.1537080284526156e-06, "loss": 0.3191, "step": 1009 }, { "epoch": 0.2615648570781134, "grad_norm": 7.447874599354588, "learning_rate": 1.1534063985618231e-06, "loss": 0.3002, "step": 1010 }, { "epoch": 0.2615648570781134, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.8987341772151899, "eval_PRM F1 AUC": 0.8407543216343635, "eval_PRM F1 AUC (fixed)": 0.7391304347826088, "eval_PRM F1 Neg": 0.7037037037037037, "eval_PRM NPV": 0.6129032258064516, "eval_PRM Precision": 0.9466666666666667, "eval_PRM Recall": 0.8554216867469879, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.2732199430465698, "eval_runtime": 14.6566, "eval_samples_per_second": 3.002, "eval_steps_per_second": 0.136, "step": 1010 }, { "epoch": 0.2618238321841313, "grad_norm": 4.934117068568929, "learning_rate": 1.153104336547262e-06, "loss": 0.4552, "step": 1011 }, { "epoch": 0.2620828072901492, "grad_norm": 5.94232166921095, "learning_rate": 1.1528018426559534e-06, "loss": 0.3295, "step": 1012 }, { "epoch": 0.2623417823961672, "grad_norm": 4.68230926071834, "learning_rate": 1.152498917135274e-06, "loss": 0.3659, "step": 1013 }, { "epoch": 0.2626007575021851, "grad_norm": 3.9028364151194155, "learning_rate": 1.1521955602329516e-06, "loss": 0.375, "step": 1014 }, { "epoch": 0.26285973260820306, "grad_norm": 3.820941550762621, "learning_rate": 1.1518917721970679e-06, "loss": 0.387, "step": 1015 }, { "epoch": 0.26285973260820306, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.89171974522293, "eval_PRM F1 AUC": 0.8347302252488213, "eval_PRM F1 AUC (fixed)": 0.7252488213724463, "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": 0.28247568011283875, "eval_runtime": 15.3011, "eval_samples_per_second": 2.876, "eval_steps_per_second": 0.131, "step": 1015 }, { "epoch": 0.26311870771422097, "grad_norm": 6.272094131495291, "learning_rate": 1.1515875532760563e-06, "loss": 0.4004, "step": 1016 }, { "epoch": 0.2633776828202389, "grad_norm": 9.61148233057241, "learning_rate": 1.1512829037187032e-06, "loss": 0.3623, "step": 1017 }, { "epoch": 0.26363665792625685, "grad_norm": 3.5927848620945237, "learning_rate": 1.150977823774147e-06, "loss": 0.3681, "step": 1018 }, { "epoch": 0.26389563303227476, "grad_norm": 4.669022408853148, "learning_rate": 1.150672313691878e-06, "loss": 0.329, "step": 1019 }, { "epoch": 0.2641546081382927, "grad_norm": 4.151941292781019, "learning_rate": 1.1503663737217385e-06, "loss": 0.3425, "step": 1020 }, { "epoch": 0.2641546081382927, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.89171974522293, "eval_PRM F1 AUC": 0.8347302252488213, "eval_PRM F1 AUC (fixed)": 0.7155578837087481, "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": 0.29420801997184753, "eval_runtime": 14.6768, "eval_samples_per_second": 2.998, "eval_steps_per_second": 0.136, "step": 1020 }, { "epoch": 0.26441358324431063, "grad_norm": 3.1002475083020897, "learning_rate": 1.150060004113922e-06, "loss": 0.3625, "step": 1021 }, { "epoch": 0.2646725583503286, "grad_norm": 7.00755462604261, "learning_rate": 1.1497532051189734e-06, "loss": 0.3134, "step": 1022 }, { "epoch": 0.2649315334563465, "grad_norm": 3.205399859551509, "learning_rate": 1.1494459769877892e-06, "loss": 0.4001, "step": 1023 }, { "epoch": 0.2651905085623644, "grad_norm": 4.163906104739272, "learning_rate": 1.1491383199716162e-06, "loss": 0.3586, "step": 1024 }, { "epoch": 0.2654494836683824, "grad_norm": 3.3459054714717604, "learning_rate": 1.1488302343220522e-06, "loss": 0.3757, "step": 1025 }, { "epoch": 0.2654494836683824, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.9056603773584906, "eval_PRM F1 AUC": 0.8467784180199058, "eval_PRM F1 AUC (fixed)": 0.7097957045573599, "eval_PRM F1 Neg": 0.7169811320754716, "eval_PRM NPV": 0.6333333333333333, "eval_PRM Precision": 0.9473684210526315, "eval_PRM Recall": 0.8674698795180723, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.2874738276004791, "eval_runtime": 14.7665, "eval_samples_per_second": 2.98, "eval_steps_per_second": 0.135, "step": 1025 }, { "epoch": 0.2657084587744003, "grad_norm": 3.062475684970944, "learning_rate": 1.148521720291046e-06, "loss": 0.3169, "step": 1026 }, { "epoch": 0.26596743388041827, "grad_norm": 5.559405808823225, "learning_rate": 1.1482127781308959e-06, "loss": 0.3497, "step": 1027 }, { "epoch": 0.2662264089864362, "grad_norm": 2.8845109010736802, "learning_rate": 1.1479034080942506e-06, "loss": 0.2926, "step": 1028 }, { "epoch": 0.2664853840924541, "grad_norm": 2.7290208271682483, "learning_rate": 1.1475936104341092e-06, "loss": 0.2948, "step": 1029 }, { "epoch": 0.26674435919847206, "grad_norm": 4.275627121965084, "learning_rate": 1.1472833854038202e-06, "loss": 0.3048, "step": 1030 }, { "epoch": 0.26674435919847206, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.9, "eval_PRM F1 AUC": 0.8250392875851232, "eval_PRM F1 AUC (fixed)": 0.7176532215819801, "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": 0.2746083736419678, "eval_runtime": 14.8637, "eval_samples_per_second": 2.96, "eval_steps_per_second": 0.135, "step": 1030 }, { "epoch": 0.26700333430448997, "grad_norm": 5.096607444133815, "learning_rate": 1.146972733257081e-06, "loss": 0.3847, "step": 1031 }, { "epoch": 0.26726230941050794, "grad_norm": 5.906775619461052, "learning_rate": 1.1466616542479392e-06, "loss": 0.4176, "step": 1032 }, { "epoch": 0.26752128451652585, "grad_norm": 3.3864533892662276, "learning_rate": 1.1463501486307913e-06, "loss": 0.4073, "step": 1033 }, { "epoch": 0.26778025962254376, "grad_norm": 7.674881770284153, "learning_rate": 1.146038216660382e-06, "loss": 0.3887, "step": 1034 }, { "epoch": 0.2680392347285617, "grad_norm": 3.9737886137856604, "learning_rate": 1.1457258585918055e-06, "loss": 0.3936, "step": 1035 }, { "epoch": 0.2680392347285617, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.8974358974358975, "eval_PRM F1 AUC": 0.8564693556836039, "eval_PRM F1 AUC (fixed)": 0.7317967522262965, "eval_PRM F1 Neg": 0.7142857142857143, "eval_PRM NPV": 0.6060606060606061, "eval_PRM Precision": 0.958904109589041, "eval_PRM Recall": 0.8433734939759037, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.2778846025466919, "eval_runtime": 14.5607, "eval_samples_per_second": 3.022, "eval_steps_per_second": 0.137, "step": 1035 }, { "epoch": 0.26829820983457964, "grad_norm": 3.299595057990445, "learning_rate": 1.1454130746805036e-06, "loss": 0.3571, "step": 1036 }, { "epoch": 0.2685571849405976, "grad_norm": 3.1859027589559115, "learning_rate": 1.1450998651822676e-06, "loss": 0.2808, "step": 1037 }, { "epoch": 0.2688161600466155, "grad_norm": 5.266318626904086, "learning_rate": 1.1447862303532357e-06, "loss": 0.3535, "step": 1038 }, { "epoch": 0.2690751351526334, "grad_norm": 6.0832686097080835, "learning_rate": 1.144472170449894e-06, "loss": 0.2934, "step": 1039 }, { "epoch": 0.2693341102586514, "grad_norm": 5.449410893831182, "learning_rate": 1.144157685729077e-06, "loss": 0.3511, "step": 1040 }, { "epoch": 0.2693341102586514, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8831168831168831, "eval_PRM F1 AUC": 0.8444211629125196, "eval_PRM F1 AUC (fixed)": 0.7393923520167627, "eval_PRM F1 Neg": 0.6896551724137931, "eval_PRM NPV": 0.5714285714285714, "eval_PRM Precision": 0.9577464788732394, "eval_PRM Recall": 0.8192771084337349, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.2817172706127167, "eval_runtime": 14.4198, "eval_samples_per_second": 3.051, "eval_steps_per_second": 0.139, "step": 1040 }, { "epoch": 0.2695930853646693, "grad_norm": 10.491775913753377, "learning_rate": 1.143842776447966e-06, "loss": 0.4059, "step": 1041 }, { "epoch": 0.26985206047068727, "grad_norm": 4.163129039105927, "learning_rate": 1.1435274428640896e-06, "loss": 0.3379, "step": 1042 }, { "epoch": 0.2701110355767052, "grad_norm": 4.035370510840349, "learning_rate": 1.1432116852353236e-06, "loss": 0.3394, "step": 1043 }, { "epoch": 0.27037001068272315, "grad_norm": 5.014839844833506, "learning_rate": 1.1428955038198902e-06, "loss": 0.2666, "step": 1044 }, { "epoch": 0.27062898578874106, "grad_norm": 3.3777543126356533, "learning_rate": 1.1425788988763587e-06, "loss": 0.3968, "step": 1045 }, { "epoch": 0.27062898578874106, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.9044585987261147, "eval_PRM F1 AUC": 0.8624934520691462, "eval_PRM F1 AUC (fixed)": 0.7545835515976952, "eval_PRM F1 Neg": 0.7272727272727273, "eval_PRM NPV": 0.625, "eval_PRM Precision": 0.9594594594594594, "eval_PRM Recall": 0.8554216867469879, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.25391244888305664, "eval_runtime": 14.6404, "eval_samples_per_second": 3.005, "eval_steps_per_second": 0.137, "step": 1045 }, { "epoch": 0.270887960894759, "grad_norm": 4.769686269210307, "learning_rate": 1.142261870663644e-06, "loss": 0.3934, "step": 1046 }, { "epoch": 0.27114693600077694, "grad_norm": 4.340672008315693, "learning_rate": 1.1419444194410078e-06, "loss": 0.3723, "step": 1047 }, { "epoch": 0.27140591110679485, "grad_norm": 7.286389338199718, "learning_rate": 1.1416265454680578e-06, "loss": 0.3354, "step": 1048 }, { "epoch": 0.2716648862128128, "grad_norm": 6.061977229152866, "learning_rate": 1.1413082490047465e-06, "loss": 0.2896, "step": 1049 }, { "epoch": 0.27192386131883073, "grad_norm": 6.780070674559419, "learning_rate": 1.140989530311373e-06, "loss": 0.3464, "step": 1050 }, { "epoch": 0.27192386131883073, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.9056603773584906, "eval_PRM F1 AUC": 0.8467784180199058, "eval_PRM F1 AUC (fixed)": 0.7582503928758513, "eval_PRM F1 Neg": 0.7169811320754716, "eval_PRM NPV": 0.6333333333333333, "eval_PRM Precision": 0.9473684210526315, "eval_PRM Recall": 0.8674698795180723, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.250600129365921, "eval_runtime": 14.7351, "eval_samples_per_second": 2.986, "eval_steps_per_second": 0.136, "step": 1050 }, { "epoch": 0.27218283642484864, "grad_norm": 3.0057781248611124, "learning_rate": 1.1406703896485814e-06, "loss": 0.3507, "step": 1051 }, { "epoch": 0.2724418115308666, "grad_norm": 2.3656334187644323, "learning_rate": 1.14035082727736e-06, "loss": 0.2873, "step": 1052 }, { "epoch": 0.2727007866368845, "grad_norm": 4.327035964713289, "learning_rate": 1.1400308434590434e-06, "loss": 0.3582, "step": 1053 }, { "epoch": 0.2729597617429025, "grad_norm": 4.409821289635134, "learning_rate": 1.1397104384553099e-06, "loss": 0.3955, "step": 1054 }, { "epoch": 0.2732187368489204, "grad_norm": 6.070528106749317, "learning_rate": 1.1393896125281823e-06, "loss": 0.3542, "step": 1055 }, { "epoch": 0.2732187368489204, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8903225806451613, "eval_PRM F1 AUC": 0.8504452592980618, "eval_PRM F1 AUC (fixed)": 0.7572027239392352, "eval_PRM F1 Neg": 0.7017543859649122, "eval_PRM NPV": 0.5882352941176471, "eval_PRM Precision": 0.9583333333333334, "eval_PRM Recall": 0.8313253012048193, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.256532222032547, "eval_runtime": 15.0852, "eval_samples_per_second": 2.917, "eval_steps_per_second": 0.133, "step": 1055 }, { "epoch": 0.2734777119549383, "grad_norm": 3.5562342168376575, "learning_rate": 1.139068365940028e-06, "loss": 0.3447, "step": 1056 }, { "epoch": 0.2737366870609563, "grad_norm": 5.912361465122927, "learning_rate": 1.1387466989535581e-06, "loss": 0.3846, "step": 1057 }, { "epoch": 0.2739956621669742, "grad_norm": 5.0568559263393595, "learning_rate": 1.1384246118318279e-06, "loss": 0.3529, "step": 1058 }, { "epoch": 0.27425463727299215, "grad_norm": 3.890988872505904, "learning_rate": 1.1381021048382354e-06, "loss": 0.3636, "step": 1059 }, { "epoch": 0.27451361237901006, "grad_norm": 5.104942614408514, "learning_rate": 1.1377791782365234e-06, "loss": 0.4364, "step": 1060 }, { "epoch": 0.27451361237901006, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.8961038961038961, "eval_PRM F1 AUC": 0.8721843897328445, "eval_PRM F1 AUC (fixed)": 0.7527501309586171, "eval_PRM F1 Neg": 0.7241379310344828, "eval_PRM NPV": 0.6, "eval_PRM Precision": 0.971830985915493, "eval_PRM Recall": 0.8313253012048193, "eval_PRM Specificty": 0.9130434782608695, "eval_loss": 0.26171520352363586, "eval_runtime": 14.5773, "eval_samples_per_second": 3.018, "eval_steps_per_second": 0.137, "step": 1060 }, { "epoch": 0.274772587485028, "grad_norm": 7.7662853415479525, "learning_rate": 1.137455832290776e-06, "loss": 0.3633, "step": 1061 }, { "epoch": 0.27503156259104594, "grad_norm": 5.938423981059565, "learning_rate": 1.137132067265422e-06, "loss": 0.3852, "step": 1062 }, { "epoch": 0.27529053769706385, "grad_norm": 5.107382832753897, "learning_rate": 1.136807883425232e-06, "loss": 0.2787, "step": 1063 }, { "epoch": 0.2755495128030818, "grad_norm": 3.6572572981324556, "learning_rate": 1.136483281035319e-06, "loss": 0.3835, "step": 1064 }, { "epoch": 0.27580848790909973, "grad_norm": 4.0782743725531505, "learning_rate": 1.136158260361139e-06, "loss": 0.3717, "step": 1065 }, { "epoch": 0.27580848790909973, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.8974358974358975, "eval_PRM F1 AUC": 0.8564693556836039, "eval_PRM F1 AUC (fixed)": 0.7456783656364588, "eval_PRM F1 Neg": 0.7142857142857143, "eval_PRM NPV": 0.6060606060606061, "eval_PRM Precision": 0.958904109589041, "eval_PRM Recall": 0.8433734939759037, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.26301947236061096, "eval_runtime": 14.8292, "eval_samples_per_second": 2.967, "eval_steps_per_second": 0.135, "step": 1065 }, { "epoch": 0.2760674630151177, "grad_norm": 5.013671476758816, "learning_rate": 1.1358328216684891e-06, "loss": 0.2519, "step": 1066 }, { "epoch": 0.2763264381211356, "grad_norm": 3.4108222416181158, "learning_rate": 1.135506965223509e-06, "loss": 0.2606, "step": 1067 }, { "epoch": 0.2765854132271535, "grad_norm": 6.796678122355816, "learning_rate": 1.13518069129268e-06, "loss": 0.4095, "step": 1068 }, { "epoch": 0.2768443883331715, "grad_norm": 4.326362848524545, "learning_rate": 1.1348540001428243e-06, "loss": 0.3535, "step": 1069 }, { "epoch": 0.2771033634391894, "grad_norm": 4.655760102750851, "learning_rate": 1.134526892041106e-06, "loss": 0.3831, "step": 1070 }, { "epoch": 0.2771033634391894, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.9, "eval_PRM F1 AUC": 0.8250392875851232, "eval_PRM F1 AUC (fixed)": 0.7569408067050812, "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": 0.26515859365463257, "eval_runtime": 14.8422, "eval_samples_per_second": 2.965, "eval_steps_per_second": 0.135, "step": 1070 }, { "epoch": 0.27736233854520737, "grad_norm": 8.05051403901847, "learning_rate": 1.134199367255029e-06, "loss": 0.3539, "step": 1071 }, { "epoch": 0.2776213136512253, "grad_norm": 3.378675019359202, "learning_rate": 1.1338714260524393e-06, "loss": 0.3111, "step": 1072 }, { "epoch": 0.2778802887572432, "grad_norm": 3.7446683598641615, "learning_rate": 1.133543068701523e-06, "loss": 0.4096, "step": 1073 }, { "epoch": 0.27813926386326115, "grad_norm": 6.913150764414947, "learning_rate": 1.1332142954708064e-06, "loss": 0.3498, "step": 1074 }, { "epoch": 0.27839823896927907, "grad_norm": 3.781401402320106, "learning_rate": 1.1328851066291555e-06, "loss": 0.3295, "step": 1075 }, { "epoch": 0.27839823896927907, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.8987341772151899, "eval_PRM F1 AUC": 0.8407543216343635, "eval_PRM F1 AUC (fixed)": 0.7537977998952331, "eval_PRM F1 Neg": 0.7037037037037037, "eval_PRM NPV": 0.6129032258064516, "eval_PRM Precision": 0.9466666666666667, "eval_PRM Recall": 0.8554216867469879, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.2676940858364105, "eval_runtime": 14.9396, "eval_samples_per_second": 2.945, "eval_steps_per_second": 0.134, "step": 1075 }, { "epoch": 0.27865721407529703, "grad_norm": 5.819523266576147, "learning_rate": 1.1325555024457768e-06, "loss": 0.4179, "step": 1076 }, { "epoch": 0.27891618918131494, "grad_norm": 4.099575759852333, "learning_rate": 1.1322254831902162e-06, "loss": 0.4283, "step": 1077 }, { "epoch": 0.27917516428733286, "grad_norm": 2.388694788644085, "learning_rate": 1.1318950491323593e-06, "loss": 0.3189, "step": 1078 }, { "epoch": 0.2794341393933508, "grad_norm": 3.8430182333509477, "learning_rate": 1.1315642005424304e-06, "loss": 0.3609, "step": 1079 }, { "epoch": 0.27969311449936873, "grad_norm": 6.0484009009962865, "learning_rate": 1.1312329376909934e-06, "loss": 0.4167, "step": 1080 }, { "epoch": 0.27969311449936873, "eval_PRM Accuracy": 0.8207547169811321, "eval_PRM F1": 0.8758169934640523, "eval_PRM F1 AUC": 0.8383970665269774, "eval_PRM F1 AUC (fixed)": 0.749083289680461, "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": 0.28986474871635437, "eval_runtime": 14.6064, "eval_samples_per_second": 3.012, "eval_steps_per_second": 0.137, "step": 1080 }, { "epoch": 0.2799520896053867, "grad_norm": 7.64904237886481, "learning_rate": 1.1309012608489504e-06, "loss": 0.3803, "step": 1081 }, { "epoch": 0.2802110647114046, "grad_norm": 9.823767001803189, "learning_rate": 1.1305691702875423e-06, "loss": 0.2893, "step": 1082 }, { "epoch": 0.2804700398174225, "grad_norm": 5.004452607333411, "learning_rate": 1.1302366662783488e-06, "loss": 0.4203, "step": 1083 }, { "epoch": 0.2807290149234405, "grad_norm": 8.144551217692342, "learning_rate": 1.129903749093287e-06, "loss": 0.3629, "step": 1084 }, { "epoch": 0.2809879900294584, "grad_norm": 7.145958462099795, "learning_rate": 1.129570419004612e-06, "loss": 0.289, "step": 1085 }, { "epoch": 0.2809879900294584, "eval_PRM Accuracy": 0.8113207547169812, "eval_PRM F1": 0.868421052631579, "eval_PRM F1 AUC": 0.8323729701414354, "eval_PRM F1 AUC (fixed)": 0.748821372446307, "eval_PRM F1 Neg": 0.6666666666666666, "eval_PRM NPV": 0.5405405405405406, "eval_PRM Precision": 0.9565217391304348, "eval_PRM Recall": 0.7951807228915663, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.294299453496933, "eval_runtime": 14.5707, "eval_samples_per_second": 3.02, "eval_steps_per_second": 0.137, "step": 1085 }, { "epoch": 0.28124696513547637, "grad_norm": 4.960838845669801, "learning_rate": 1.129236676284917e-06, "loss": 0.3983, "step": 1086 }, { "epoch": 0.2815059402414943, "grad_norm": 4.9063190727136, "learning_rate": 1.1289025212071324e-06, "loss": 0.3552, "step": 1087 }, { "epoch": 0.28176491534751225, "grad_norm": 6.111329302463371, "learning_rate": 1.1285679540445259e-06, "loss": 0.4539, "step": 1088 }, { "epoch": 0.28202389045353016, "grad_norm": 3.1683233714486176, "learning_rate": 1.1282329750707018e-06, "loss": 0.3701, "step": 1089 }, { "epoch": 0.28228286555954807, "grad_norm": 5.4822402949350115, "learning_rate": 1.1278975845596016e-06, "loss": 0.3055, "step": 1090 }, { "epoch": 0.28228286555954807, "eval_PRM Accuracy": 0.8207547169811321, "eval_PRM F1": 0.8789808917197452, "eval_PRM F1 AUC": 0.8069669984284967, "eval_PRM F1 AUC (fixed)": 0.753273965426925, "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": 0.2796691656112671, "eval_runtime": 14.4984, "eval_samples_per_second": 3.035, "eval_steps_per_second": 0.138, "step": 1090 }, { "epoch": 0.28254184066556604, "grad_norm": 3.480490868188625, "learning_rate": 1.127561782785503e-06, "loss": 0.3824, "step": 1091 }, { "epoch": 0.28280081577158395, "grad_norm": 3.805686299465861, "learning_rate": 1.1272255700230209e-06, "loss": 0.3474, "step": 1092 }, { "epoch": 0.2830597908776019, "grad_norm": 4.48044589020009, "learning_rate": 1.126888946547105e-06, "loss": 0.3661, "step": 1093 }, { "epoch": 0.2833187659836198, "grad_norm": 2.877380173271708, "learning_rate": 1.1265519126330417e-06, "loss": 0.3087, "step": 1094 }, { "epoch": 0.28357774108963774, "grad_norm": 2.542242455848216, "learning_rate": 1.126214468556453e-06, "loss": 0.315, "step": 1095 }, { "epoch": 0.28357774108963774, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8875, "eval_PRM F1 AUC": 0.7972760607647982, "eval_PRM F1 AUC (fixed)": 0.7637506547930855, "eval_PRM F1 Neg": 0.6538461538461539, "eval_PRM NPV": 0.5862068965517241, "eval_PRM Precision": 0.922077922077922, "eval_PRM Recall": 0.8554216867469879, "eval_PRM Specificty": 0.7391304347826086, "eval_loss": 0.2715385854244232, "eval_runtime": 14.3759, "eval_samples_per_second": 3.061, "eval_steps_per_second": 0.139, "step": 1095 }, { "epoch": 0.2838367161956557, "grad_norm": 5.416422720177891, "learning_rate": 1.1258766145932964e-06, "loss": 0.5243, "step": 1096 }, { "epoch": 0.2840956913016736, "grad_norm": 7.765499713875308, "learning_rate": 1.125538351019864e-06, "loss": 0.4824, "step": 1097 }, { "epoch": 0.2843546664076916, "grad_norm": 4.784714891415138, "learning_rate": 1.1251996781127832e-06, "loss": 0.3516, "step": 1098 }, { "epoch": 0.2846136415137095, "grad_norm": 4.1611214148825315, "learning_rate": 1.1248605961490168e-06, "loss": 0.4139, "step": 1099 }, { "epoch": 0.2848726166197274, "grad_norm": 9.588537077132573, "learning_rate": 1.1245211054058611e-06, "loss": 0.3736, "step": 1100 }, { "epoch": 0.2848726166197274, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.89171974522293, "eval_PRM F1 AUC": 0.8347302252488213, "eval_PRM F1 AUC (fixed)": 0.7532739654269252, "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": 0.27907219529151917, "eval_runtime": 14.5308, "eval_samples_per_second": 3.028, "eval_steps_per_second": 0.138, "step": 1100 }, { "epoch": 0.28513159172574537, "grad_norm": 8.28027340948901, "learning_rate": 1.1241812061609475e-06, "loss": 0.3066, "step": 1101 }, { "epoch": 0.2853905668317633, "grad_norm": 5.2275923345878565, "learning_rate": 1.1238408986922405e-06, "loss": 0.2485, "step": 1102 }, { "epoch": 0.28564954193778125, "grad_norm": 3.9191319661849118, "learning_rate": 1.1235001832780397e-06, "loss": 0.2648, "step": 1103 }, { "epoch": 0.28590851704379916, "grad_norm": 3.803575863891714, "learning_rate": 1.1231590601969773e-06, "loss": 0.3509, "step": 1104 }, { "epoch": 0.2861674921498171, "grad_norm": 5.062747812604048, "learning_rate": 1.1228175297280193e-06, "loss": 0.3306, "step": 1105 }, { "epoch": 0.2861674921498171, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.8987341772151899, "eval_PRM F1 AUC": 0.8407543216343635, "eval_PRM F1 AUC (fixed)": 0.7564169722367732, "eval_PRM F1 Neg": 0.7037037037037037, "eval_PRM NPV": 0.6129032258064516, "eval_PRM Precision": 0.9466666666666667, "eval_PRM Recall": 0.8554216867469879, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.2660982012748718, "eval_runtime": 14.627, "eval_samples_per_second": 3.008, "eval_steps_per_second": 0.137, "step": 1105 }, { "epoch": 0.28642646725583504, "grad_norm": 4.898710258003656, "learning_rate": 1.1224755921504648e-06, "loss": 0.4031, "step": 1106 }, { "epoch": 0.28668544236185295, "grad_norm": 3.561715323993499, "learning_rate": 1.1221332477439459e-06, "loss": 0.3674, "step": 1107 }, { "epoch": 0.2869444174678709, "grad_norm": 3.7429726684658853, "learning_rate": 1.1217904967884274e-06, "loss": 0.3967, "step": 1108 }, { "epoch": 0.2872033925738888, "grad_norm": 3.178179402931704, "learning_rate": 1.1214473395642064e-06, "loss": 0.3632, "step": 1109 }, { "epoch": 0.2874623676799068, "grad_norm": 6.2815507059756, "learning_rate": 1.1211037763519122e-06, "loss": 0.3744, "step": 1110 }, { "epoch": 0.2874623676799068, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.89171974522293, "eval_PRM F1 AUC": 0.8347302252488213, "eval_PRM F1 AUC (fixed)": 0.7608695652173912, "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": 0.26640403270721436, "eval_runtime": 14.4791, "eval_samples_per_second": 3.039, "eval_steps_per_second": 0.138, "step": 1110 }, { "epoch": 0.2877213427859247, "grad_norm": 4.2386671161357405, "learning_rate": 1.1207598074325066e-06, "loss": 0.3827, "step": 1111 }, { "epoch": 0.2879803178919426, "grad_norm": 3.3655970950015295, "learning_rate": 1.1204154330872824e-06, "loss": 0.3525, "step": 1112 }, { "epoch": 0.2882392929979606, "grad_norm": 7.982714386079896, "learning_rate": 1.1200706535978647e-06, "loss": 0.3293, "step": 1113 }, { "epoch": 0.2884982681039785, "grad_norm": 9.602171181971372, "learning_rate": 1.1197254692462096e-06, "loss": 0.3855, "step": 1114 }, { "epoch": 0.28875724320999646, "grad_norm": 3.4844020908031554, "learning_rate": 1.1193798803146042e-06, "loss": 0.3128, "step": 1115 }, { "epoch": 0.28875724320999646, "eval_PRM Accuracy": 0.8207547169811321, "eval_PRM F1": 0.8774193548387097, "eval_PRM F1 AUC": 0.8226820324777371, "eval_PRM F1 AUC (fixed)": 0.7627029858564695, "eval_PRM F1 Neg": 0.6666666666666666, "eval_PRM NPV": 0.5588235294117647, "eval_PRM Precision": 0.9444444444444444, "eval_PRM Recall": 0.8192771084337349, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.26669546961784363, "eval_runtime": 14.4819, "eval_samples_per_second": 3.038, "eval_steps_per_second": 0.138, "step": 1115 }, { "epoch": 0.2890162183160144, "grad_norm": 6.597406785592498, "learning_rate": 1.1190338870856666e-06, "loss": 0.2791, "step": 1116 }, { "epoch": 0.2892751934220323, "grad_norm": 5.9280718737016045, "learning_rate": 1.1186874898423455e-06, "loss": 0.3807, "step": 1117 }, { "epoch": 0.28953416852805025, "grad_norm": 2.8334208382831605, "learning_rate": 1.1183406888679199e-06, "loss": 0.3188, "step": 1118 }, { "epoch": 0.28979314363406816, "grad_norm": 7.612503860515019, "learning_rate": 1.117993484445999e-06, "loss": 0.3501, "step": 1119 }, { "epoch": 0.29005211874008613, "grad_norm": 8.11990763445414, "learning_rate": 1.1176458768605225e-06, "loss": 0.4403, "step": 1120 }, { "epoch": 0.29005211874008613, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.9, "eval_PRM F1 AUC": 0.8250392875851232, "eval_PRM F1 AUC (fixed)": 0.7747511786275538, "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": 0.2532932758331299, "eval_runtime": 14.6892, "eval_samples_per_second": 2.995, "eval_steps_per_second": 0.136, "step": 1120 }, { "epoch": 0.29031109384610404, "grad_norm": 4.263431566665755, "learning_rate": 1.1172978663957587e-06, "loss": 0.3335, "step": 1121 }, { "epoch": 0.29057006895212195, "grad_norm": 5.260238174654083, "learning_rate": 1.1169494533363062e-06, "loss": 0.34, "step": 1122 }, { "epoch": 0.2908290440581399, "grad_norm": 4.355002531212862, "learning_rate": 1.1166006379670924e-06, "loss": 0.3562, "step": 1123 }, { "epoch": 0.29108801916415783, "grad_norm": 4.538866881853854, "learning_rate": 1.1162514205733744e-06, "loss": 0.343, "step": 1124 }, { "epoch": 0.2913469942701758, "grad_norm": 4.793637250566336, "learning_rate": 1.115901801440737e-06, "loss": 0.3588, "step": 1125 }, { "epoch": 0.2913469942701758, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.906832298136646, "eval_PRM F1 AUC": 0.8310633839706653, "eval_PRM F1 AUC (fixed)": 0.7755369303300158, "eval_PRM F1 Neg": 0.7058823529411765, "eval_PRM NPV": 0.6428571428571429, "eval_PRM Precision": 0.9358974358974359, "eval_PRM Recall": 0.8795180722891566, "eval_PRM Specificty": 0.782608695652174, "eval_loss": 0.2523413598537445, "eval_runtime": 14.579, "eval_samples_per_second": 3.018, "eval_steps_per_second": 0.137, "step": 1125 }, { "epoch": 0.2916059693761937, "grad_norm": 3.7268330544818213, "learning_rate": 1.1155517808550941e-06, "loss": 0.4181, "step": 1126 }, { "epoch": 0.2918649444822117, "grad_norm": 5.165577278078134, "learning_rate": 1.1152013591026884e-06, "loss": 0.3887, "step": 1127 }, { "epoch": 0.2921239195882296, "grad_norm": 5.040742010231609, "learning_rate": 1.11485053647009e-06, "loss": 0.3919, "step": 1128 }, { "epoch": 0.2923828946942475, "grad_norm": 4.725959357988391, "learning_rate": 1.1144993132441967e-06, "loss": 0.3668, "step": 1129 }, { "epoch": 0.29264186980026546, "grad_norm": 4.032660789825553, "learning_rate": 1.1141476897122348e-06, "loss": 0.3183, "step": 1130 }, { "epoch": 0.29264186980026546, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.89171974522293, "eval_PRM F1 AUC": 0.8347302252488213, "eval_PRM F1 AUC (fixed)": 0.7661079099004714, "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": 0.26497408747673035, "eval_runtime": 14.3768, "eval_samples_per_second": 3.06, "eval_steps_per_second": 0.139, "step": 1130 }, { "epoch": 0.2929008449062834, "grad_norm": 7.673167100360908, "learning_rate": 1.1137956661617566e-06, "loss": 0.3832, "step": 1131 }, { "epoch": 0.29315982001230134, "grad_norm": 7.881068518051551, "learning_rate": 1.113443242880643e-06, "loss": 0.3027, "step": 1132 }, { "epoch": 0.29341879511831925, "grad_norm": 5.782281218579323, "learning_rate": 1.1130904201571006e-06, "loss": 0.3171, "step": 1133 }, { "epoch": 0.29367777022433716, "grad_norm": 3.762358561390526, "learning_rate": 1.1127371982796636e-06, "loss": 0.3372, "step": 1134 }, { "epoch": 0.29393674533035513, "grad_norm": 3.7955893572075787, "learning_rate": 1.1123835775371919e-06, "loss": 0.3413, "step": 1135 }, { "epoch": 0.29393674533035513, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.906832298136646, "eval_PRM F1 AUC": 0.8310633839706653, "eval_PRM F1 AUC (fixed)": 0.7619172341540074, "eval_PRM F1 Neg": 0.7058823529411765, "eval_PRM NPV": 0.6428571428571429, "eval_PRM Precision": 0.9358974358974359, "eval_PRM Recall": 0.8795180722891566, "eval_PRM Specificty": 0.782608695652174, "eval_loss": 0.25865188241004944, "eval_runtime": 14.9603, "eval_samples_per_second": 2.941, "eval_steps_per_second": 0.134, "step": 1135 }, { "epoch": 0.29419572043637304, "grad_norm": 2.959177369623632, "learning_rate": 1.1120295582188718e-06, "loss": 0.3441, "step": 1136 }, { "epoch": 0.294454695542391, "grad_norm": 3.3679792765512704, "learning_rate": 1.111675140614216e-06, "loss": 0.3377, "step": 1137 }, { "epoch": 0.2947136706484089, "grad_norm": 3.597086189327973, "learning_rate": 1.1113203250130623e-06, "loss": 0.3644, "step": 1138 }, { "epoch": 0.29497264575442683, "grad_norm": 2.9662259450420168, "learning_rate": 1.110965111705574e-06, "loss": 0.3793, "step": 1139 }, { "epoch": 0.2952316208604448, "grad_norm": 10.258782090900965, "learning_rate": 1.1106095009822401e-06, "loss": 0.4473, "step": 1140 }, { "epoch": 0.2952316208604448, "eval_PRM Accuracy": 0.8773584905660378, "eval_PRM F1": 0.9202453987730062, "eval_PRM F1 AUC": 0.8431115767417497, "eval_PRM F1 AUC (fixed)": 0.7621791513881614, "eval_PRM F1 Neg": 0.7346938775510204, "eval_PRM NPV": 0.6923076923076923, "eval_PRM Precision": 0.9375, "eval_PRM Recall": 0.9036144578313253, "eval_PRM Specificty": 0.782608695652174, "eval_loss": 0.2613238990306854, "eval_runtime": 14.3424, "eval_samples_per_second": 3.068, "eval_steps_per_second": 0.139, "step": 1140 }, { "epoch": 0.2954905959664627, "grad_norm": 9.086209755968815, "learning_rate": 1.1102534931338743e-06, "loss": 0.357, "step": 1141 }, { "epoch": 0.2957495710724807, "grad_norm": 4.4360234766203845, "learning_rate": 1.1098970884516153e-06, "loss": 0.3027, "step": 1142 }, { "epoch": 0.2960085461784986, "grad_norm": 3.6234523247900623, "learning_rate": 1.1095402872269257e-06, "loss": 0.3617, "step": 1143 }, { "epoch": 0.2962675212845165, "grad_norm": 3.214590685530097, "learning_rate": 1.1091830897515932e-06, "loss": 0.361, "step": 1144 }, { "epoch": 0.29652649639053447, "grad_norm": 4.836324193876224, "learning_rate": 1.108825496317729e-06, "loss": 0.3596, "step": 1145 }, { "epoch": 0.29652649639053447, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8860759493670886, "eval_PRM F1 AUC": 0.8129910948140389, "eval_PRM F1 AUC (fixed)": 0.7619172341540074, "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": 0.2637127935886383, "eval_runtime": 14.6897, "eval_samples_per_second": 2.995, "eval_steps_per_second": 0.136, "step": 1145 }, { "epoch": 0.2967854714965524, "grad_norm": 4.40229983630881, "learning_rate": 1.1084675072177684e-06, "loss": 0.2748, "step": 1146 }, { "epoch": 0.29704444660257034, "grad_norm": 6.446443974590935, "learning_rate": 1.10810912274447e-06, "loss": 0.3959, "step": 1147 }, { "epoch": 0.29730342170858826, "grad_norm": 4.162136530137705, "learning_rate": 1.1077503431909161e-06, "loss": 0.3092, "step": 1148 }, { "epoch": 0.2975623968146062, "grad_norm": 3.379942752653197, "learning_rate": 1.1073911688505118e-06, "loss": 0.3569, "step": 1149 }, { "epoch": 0.29782137192062413, "grad_norm": 4.790526956890661, "learning_rate": 1.1070316000169847e-06, "loss": 0.3305, "step": 1150 }, { "epoch": 0.29782137192062413, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.906832298136646, "eval_PRM F1 AUC": 0.8310633839706653, "eval_PRM F1 AUC (fixed)": 0.7668936616029335, "eval_PRM F1 Neg": 0.7058823529411765, "eval_PRM NPV": 0.6428571428571429, "eval_PRM Precision": 0.9358974358974359, "eval_PRM Recall": 0.8795180722891566, "eval_PRM Specificty": 0.782608695652174, "eval_loss": 0.256858229637146, "eval_runtime": 14.6627, "eval_samples_per_second": 3.001, "eval_steps_per_second": 0.136, "step": 1150 }, { "epoch": 0.29808034702664205, "grad_norm": 4.646326337168363, "learning_rate": 1.1066716369843863e-06, "loss": 0.4038, "step": 1151 }, { "epoch": 0.29833932213266, "grad_norm": 6.96599879604997, "learning_rate": 1.1063112800470888e-06, "loss": 0.3765, "step": 1152 }, { "epoch": 0.2985982972386779, "grad_norm": 3.1915204817949783, "learning_rate": 1.105950529499788e-06, "loss": 0.2686, "step": 1153 }, { "epoch": 0.2988572723446959, "grad_norm": 3.798299543621722, "learning_rate": 1.1055893856375003e-06, "loss": 0.3423, "step": 1154 }, { "epoch": 0.2991162474507138, "grad_norm": 5.672302650751939, "learning_rate": 1.1052278487555649e-06, "loss": 0.383, "step": 1155 }, { "epoch": 0.2991162474507138, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.906832298136646, "eval_PRM F1 AUC": 0.8310633839706653, "eval_PRM F1 AUC (fixed)": 0.7750130958617076, "eval_PRM F1 Neg": 0.7058823529411765, "eval_PRM NPV": 0.6428571428571429, "eval_PRM Precision": 0.9358974358974359, "eval_PRM Recall": 0.8795180722891566, "eval_PRM Specificty": 0.782608695652174, "eval_loss": 0.25225791335105896, "eval_runtime": 14.7283, "eval_samples_per_second": 2.987, "eval_steps_per_second": 0.136, "step": 1155 }, { "epoch": 0.2993752225567317, "grad_norm": 5.845418089666656, "learning_rate": 1.1048659191496416e-06, "loss": 0.417, "step": 1156 }, { "epoch": 0.2996341976627497, "grad_norm": 5.937672783763917, "learning_rate": 1.1045035971157117e-06, "loss": 0.4291, "step": 1157 }, { "epoch": 0.2998931727687676, "grad_norm": 3.4584606230076447, "learning_rate": 1.1041408829500773e-06, "loss": 0.3194, "step": 1158 }, { "epoch": 0.30015214787478556, "grad_norm": 4.451876308852337, "learning_rate": 1.1037777769493613e-06, "loss": 0.3549, "step": 1159 }, { "epoch": 0.30041112298080347, "grad_norm": 3.602995497692067, "learning_rate": 1.103414279410507e-06, "loss": 0.3692, "step": 1160 }, { "epoch": 0.30041112298080347, "eval_PRM Accuracy": 0.8207547169811321, "eval_PRM F1": 0.8789808917197452, "eval_PRM F1 AUC": 0.8069669984284967, "eval_PRM F1 AUC (fixed)": 0.7773703509690938, "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": 0.26475226879119873, "eval_runtime": 14.7743, "eval_samples_per_second": 2.978, "eval_steps_per_second": 0.135, "step": 1160 }, { "epoch": 0.3006700980868214, "grad_norm": 5.62440934358399, "learning_rate": 1.1030503906307781e-06, "loss": 0.2798, "step": 1161 }, { "epoch": 0.30092907319283935, "grad_norm": 4.339810549098005, "learning_rate": 1.1026861109077579e-06, "loss": 0.3868, "step": 1162 }, { "epoch": 0.30118804829885726, "grad_norm": 12.165816417405841, "learning_rate": 1.1023214405393492e-06, "loss": 0.3984, "step": 1163 }, { "epoch": 0.3014470234048752, "grad_norm": 5.7935098656181365, "learning_rate": 1.1019563798237752e-06, "loss": 0.5339, "step": 1164 }, { "epoch": 0.30170599851089314, "grad_norm": 3.5268608216207467, "learning_rate": 1.1015909290595771e-06, "loss": 0.4078, "step": 1165 }, { "epoch": 0.30170599851089314, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8860759493670886, "eval_PRM F1 AUC": 0.8129910948140389, "eval_PRM F1 AUC (fixed)": 0.7729177579884756, "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": 0.26175162196159363, "eval_runtime": 14.2847, "eval_samples_per_second": 3.08, "eval_steps_per_second": 0.14, "step": 1165 }, { "epoch": 0.30196497361691105, "grad_norm": 4.127649198536379, "learning_rate": 1.1012250885456164e-06, "loss": 0.4231, "step": 1166 }, { "epoch": 0.302223948722929, "grad_norm": 3.4331514655877773, "learning_rate": 1.100858858581072e-06, "loss": 0.4242, "step": 1167 }, { "epoch": 0.3024829238289469, "grad_norm": 6.40016116126962, "learning_rate": 1.1004922394654423e-06, "loss": 0.4285, "step": 1168 }, { "epoch": 0.3027418989349649, "grad_norm": 3.656133214972611, "learning_rate": 1.1001252314985433e-06, "loss": 0.3593, "step": 1169 }, { "epoch": 0.3030008740409828, "grad_norm": 7.158557853449496, "learning_rate": 1.0997578349805096e-06, "loss": 0.4875, "step": 1170 }, { "epoch": 0.3030008740409828, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8930817610062893, "eval_PRM F1 AUC": 0.819015191199581, "eval_PRM F1 AUC (fixed)": 0.7590361445783133, "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": 0.2685607969760895, "eval_runtime": 14.5565, "eval_samples_per_second": 3.023, "eval_steps_per_second": 0.137, "step": 1170 }, { "epoch": 0.30325984914700077, "grad_norm": 3.0556367903726644, "learning_rate": 1.0993900502117929e-06, "loss": 0.3849, "step": 1171 }, { "epoch": 0.3035188242530187, "grad_norm": 2.802202697649935, "learning_rate": 1.099021877493163e-06, "loss": 0.3063, "step": 1172 }, { "epoch": 0.3037777993590366, "grad_norm": 5.612109421532672, "learning_rate": 1.0986533171257064e-06, "loss": 0.3155, "step": 1173 }, { "epoch": 0.30403677446505456, "grad_norm": 7.349516478413747, "learning_rate": 1.0982843694108272e-06, "loss": 0.3551, "step": 1174 }, { "epoch": 0.30429574957107247, "grad_norm": 5.372861942029928, "learning_rate": 1.0979150346502456e-06, "loss": 0.3109, "step": 1175 }, { "epoch": 0.30429574957107247, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.9, "eval_PRM F1 AUC": 0.8250392875851232, "eval_PRM F1 AUC (fixed)": 0.7627029858564693, "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": 0.2636399567127228, "eval_runtime": 14.6856, "eval_samples_per_second": 2.996, "eval_steps_per_second": 0.136, "step": 1175 }, { "epoch": 0.30455472467709044, "grad_norm": 2.9856816961371804, "learning_rate": 1.097545313145999e-06, "loss": 0.3356, "step": 1176 }, { "epoch": 0.30481369978310835, "grad_norm": 4.151685285025255, "learning_rate": 1.0971752052004406e-06, "loss": 0.3121, "step": 1177 }, { "epoch": 0.30507267488912626, "grad_norm": 3.2390232780479025, "learning_rate": 1.0968047111162399e-06, "loss": 0.3094, "step": 1178 }, { "epoch": 0.3053316499951442, "grad_norm": 4.673769546912608, "learning_rate": 1.096433831196382e-06, "loss": 0.3841, "step": 1179 }, { "epoch": 0.30559062510116214, "grad_norm": 3.818883587684225, "learning_rate": 1.0960625657441678e-06, "loss": 0.3289, "step": 1180 }, { "epoch": 0.30559062510116214, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.9, "eval_PRM F1 AUC": 0.8250392875851232, "eval_PRM F1 AUC (fixed)": 0.7797276060764797, "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": 0.25430724024772644, "eval_runtime": 14.6663, "eval_samples_per_second": 3.0, "eval_steps_per_second": 0.136, "step": 1180 }, { "epoch": 0.3058496002071801, "grad_norm": 3.871524702893708, "learning_rate": 1.0956909150632127e-06, "loss": 0.3914, "step": 1181 }, { "epoch": 0.306108575313198, "grad_norm": 4.465570175748092, "learning_rate": 1.0953188794574485e-06, "loss": 0.4142, "step": 1182 }, { "epoch": 0.30636755041921593, "grad_norm": 6.871748586463232, "learning_rate": 1.0949464592311203e-06, "loss": 0.4566, "step": 1183 }, { "epoch": 0.3066265255252339, "grad_norm": 5.5128943255174985, "learning_rate": 1.0945736546887888e-06, "loss": 0.3573, "step": 1184 }, { "epoch": 0.3068855006312518, "grad_norm": 6.356115342327138, "learning_rate": 1.0942004661353286e-06, "loss": 0.4997, "step": 1185 }, { "epoch": 0.3068855006312518, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8930817610062893, "eval_PRM F1 AUC": 0.819015191199581, "eval_PRM F1 AUC (fixed)": 0.7844421162912519, "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": 0.25315046310424805, "eval_runtime": 14.4533, "eval_samples_per_second": 3.044, "eval_steps_per_second": 0.138, "step": 1185 }, { "epoch": 0.3071444757372698, "grad_norm": 3.5379039136569537, "learning_rate": 1.0938268938759283e-06, "loss": 0.379, "step": 1186 }, { "epoch": 0.3074034508432877, "grad_norm": 4.202834808405025, "learning_rate": 1.0934529382160903e-06, "loss": 0.3091, "step": 1187 }, { "epoch": 0.3076624259493056, "grad_norm": 3.8281927280708548, "learning_rate": 1.0930785994616308e-06, "loss": 0.4309, "step": 1188 }, { "epoch": 0.30792140105532356, "grad_norm": 6.333845590798601, "learning_rate": 1.092703877918679e-06, "loss": 0.4023, "step": 1189 }, { "epoch": 0.3081803761613415, "grad_norm": 4.1471122361908845, "learning_rate": 1.0923287738936771e-06, "loss": 0.4087, "step": 1190 }, { "epoch": 0.3081803761613415, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8903225806451613, "eval_PRM F1 AUC": 0.8504452592980618, "eval_PRM F1 AUC (fixed)": 0.7799895233106339, "eval_PRM F1 Neg": 0.7017543859649122, "eval_PRM NPV": 0.5882352941176471, "eval_PRM Precision": 0.9583333333333334, "eval_PRM Recall": 0.8313253012048193, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.26387837529182434, "eval_runtime": 15.0772, "eval_samples_per_second": 2.918, "eval_steps_per_second": 0.133, "step": 1190 }, { "epoch": 0.30843935126735944, "grad_norm": 3.490983419168074, "learning_rate": 1.0919532876933803e-06, "loss": 0.4005, "step": 1191 }, { "epoch": 0.30869832637337735, "grad_norm": 4.300796741862126, "learning_rate": 1.0915774196248564e-06, "loss": 0.4054, "step": 1192 }, { "epoch": 0.3089573014793953, "grad_norm": 5.991455393032304, "learning_rate": 1.091201169995485e-06, "loss": 0.3517, "step": 1193 }, { "epoch": 0.30921627658541323, "grad_norm": 4.34016764245388, "learning_rate": 1.0908245391129584e-06, "loss": 0.3691, "step": 1194 }, { "epoch": 0.30947525169143114, "grad_norm": 3.006782408187908, "learning_rate": 1.0904475272852802e-06, "loss": 0.3554, "step": 1195 }, { "epoch": 0.30947525169143114, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8860759493670886, "eval_PRM F1 AUC": 0.8129910948140389, "eval_PRM F1 AUC (fixed)": 0.788108957569408, "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": 0.25778552889823914, "eval_runtime": 14.6532, "eval_samples_per_second": 3.003, "eval_steps_per_second": 0.136, "step": 1195 }, { "epoch": 0.3097342267974491, "grad_norm": 7.834331309453101, "learning_rate": 1.0900701348207656e-06, "loss": 0.3341, "step": 1196 }, { "epoch": 0.309993201903467, "grad_norm": 4.837884152961991, "learning_rate": 1.0896923620280416e-06, "loss": 0.3391, "step": 1197 }, { "epoch": 0.310252177009485, "grad_norm": 3.458578607508161, "learning_rate": 1.0893142092160451e-06, "loss": 0.3735, "step": 1198 }, { "epoch": 0.3105111521155029, "grad_norm": 3.243852160400903, "learning_rate": 1.0889356766940252e-06, "loss": 0.3834, "step": 1199 }, { "epoch": 0.3107701272215208, "grad_norm": 4.370429860764285, "learning_rate": 1.0885567647715404e-06, "loss": 0.2463, "step": 1200 }, { "epoch": 0.3107701272215208, "eval_PRM Accuracy": 0.8679245283018868, "eval_PRM F1": 0.9135802469135802, "eval_PRM F1 AUC": 0.8370874803562075, "eval_PRM F1 AUC (fixed)": 0.7998952331063384, "eval_PRM F1 Neg": 0.72, "eval_PRM NPV": 0.6666666666666666, "eval_PRM Precision": 0.9367088607594937, "eval_PRM Recall": 0.891566265060241, "eval_PRM Specificty": 0.782608695652174, "eval_loss": 0.25024715065956116, "eval_runtime": 14.8849, "eval_samples_per_second": 2.956, "eval_steps_per_second": 0.134, "step": 1200 }, { "epoch": 0.3110291023275388, "grad_norm": 3.9693767544882834, "learning_rate": 1.0881774737584597e-06, "loss": 0.2953, "step": 1201 }, { "epoch": 0.3112880774335567, "grad_norm": 3.45629528989537, "learning_rate": 1.0877978039649628e-06, "loss": 0.3799, "step": 1202 }, { "epoch": 0.31154705253957465, "grad_norm": 3.465792543374586, "learning_rate": 1.087417755701538e-06, "loss": 0.2799, "step": 1203 }, { "epoch": 0.31180602764559257, "grad_norm": 4.392253680638649, "learning_rate": 1.0870373292789844e-06, "loss": 0.4396, "step": 1204 }, { "epoch": 0.3120650027516105, "grad_norm": 7.80351799300313, "learning_rate": 1.0866565250084092e-06, "loss": 0.3509, "step": 1205 }, { "epoch": 0.3120650027516105, "eval_PRM Accuracy": 0.8773584905660378, "eval_PRM F1": 0.9212121212121213, "eval_PRM F1 AUC": 0.8273965426925091, "eval_PRM F1 AUC (fixed)": 0.8140387637506548, "eval_PRM F1 Neg": 0.723404255319149, "eval_PRM NPV": 0.7083333333333334, "eval_PRM Precision": 0.926829268292683, "eval_PRM Recall": 0.9156626506024096, "eval_PRM Specificty": 0.7391304347826086, "eval_loss": 0.2483535259962082, "eval_runtime": 14.8603, "eval_samples_per_second": 2.961, "eval_steps_per_second": 0.135, "step": 1205 }, { "epoch": 0.31232397785762844, "grad_norm": 5.156928269505667, "learning_rate": 1.0862753432012293e-06, "loss": 0.2728, "step": 1206 }, { "epoch": 0.31258295296364635, "grad_norm": 12.132211890537517, "learning_rate": 1.08589378416917e-06, "loss": 0.4319, "step": 1207 }, { "epoch": 0.3128419280696643, "grad_norm": 6.915697049487834, "learning_rate": 1.0855118482242655e-06, "loss": 0.4666, "step": 1208 }, { "epoch": 0.31310090317568223, "grad_norm": 6.968230933752234, "learning_rate": 1.085129535678858e-06, "loss": 0.2983, "step": 1209 }, { "epoch": 0.3133598782817002, "grad_norm": 3.594238368899922, "learning_rate": 1.084746846845597e-06, "loss": 0.4571, "step": 1210 }, { "epoch": 0.3133598782817002, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.8961038961038961, "eval_PRM F1 AUC": 0.8721843897328445, "eval_PRM F1 AUC (fixed)": 0.8085385018334206, "eval_PRM F1 Neg": 0.7241379310344828, "eval_PRM NPV": 0.6, "eval_PRM Precision": 0.971830985915493, "eval_PRM Recall": 0.8313253012048193, "eval_PRM Specificty": 0.9130434782608695, "eval_loss": 0.26208773255348206, "eval_runtime": 14.5572, "eval_samples_per_second": 3.023, "eval_steps_per_second": 0.137, "step": 1210 }, { "epoch": 0.3136188533877181, "grad_norm": 10.986999998465846, "learning_rate": 1.0843637820374415e-06, "loss": 0.348, "step": 1211 }, { "epoch": 0.313877828493736, "grad_norm": 6.007952874878346, "learning_rate": 1.0839803415676555e-06, "loss": 0.3719, "step": 1212 }, { "epoch": 0.314136803599754, "grad_norm": 6.548551815265104, "learning_rate": 1.0835965257498127e-06, "loss": 0.3069, "step": 1213 }, { "epoch": 0.3143957787057719, "grad_norm": 6.907837361489586, "learning_rate": 1.083212334897792e-06, "loss": 0.4288, "step": 1214 }, { "epoch": 0.31465475381178987, "grad_norm": 4.619638739193467, "learning_rate": 1.0828277693257797e-06, "loss": 0.3227, "step": 1215 }, { "epoch": 0.31465475381178987, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.8961038961038961, "eval_PRM F1 AUC": 0.8721843897328445, "eval_PRM F1 AUC (fixed)": 0.8129910948140388, "eval_PRM F1 Neg": 0.7241379310344828, "eval_PRM NPV": 0.6, "eval_PRM Precision": 0.971830985915493, "eval_PRM Recall": 0.8313253012048193, "eval_PRM Specificty": 0.9130434782608695, "eval_loss": 0.2673983871936798, "eval_runtime": 14.6708, "eval_samples_per_second": 2.999, "eval_steps_per_second": 0.136, "step": 1215 }, { "epoch": 0.3149137289178078, "grad_norm": 6.725501666695132, "learning_rate": 1.0824428293482682e-06, "loss": 0.3644, "step": 1216 }, { "epoch": 0.3151727040238257, "grad_norm": 4.1162153720484715, "learning_rate": 1.082057515280057e-06, "loss": 0.371, "step": 1217 }, { "epoch": 0.31543167912984366, "grad_norm": 6.834652646331434, "learning_rate": 1.0816718274362505e-06, "loss": 0.2873, "step": 1218 }, { "epoch": 0.31569065423586157, "grad_norm": 3.0933559887087174, "learning_rate": 1.0812857661322589e-06, "loss": 0.3626, "step": 1219 }, { "epoch": 0.31594962934187953, "grad_norm": 3.994407339934042, "learning_rate": 1.0808993316837982e-06, "loss": 0.418, "step": 1220 }, { "epoch": 0.31594962934187953, "eval_PRM Accuracy": 0.8867924528301887, "eval_PRM F1": 0.926829268292683, "eval_PRM F1 AUC": 0.8491356731272918, "eval_PRM F1 AUC (fixed)": 0.8122053431115768, "eval_PRM F1 Neg": 0.75, "eval_PRM NPV": 0.72, "eval_PRM Precision": 0.9382716049382716, "eval_PRM Recall": 0.9156626506024096, "eval_PRM Specificty": 0.782608695652174, "eval_loss": 0.24850182235240936, "eval_runtime": 15.0241, "eval_samples_per_second": 2.929, "eval_steps_per_second": 0.133, "step": 1220 }, { "epoch": 0.31620860444789745, "grad_norm": 3.5620307606032946, "learning_rate": 1.0805125244068892e-06, "loss": 0.3168, "step": 1221 }, { "epoch": 0.31646757955391536, "grad_norm": 7.342654037434144, "learning_rate": 1.0801253446178583e-06, "loss": 0.4416, "step": 1222 }, { "epoch": 0.3167265546599333, "grad_norm": 9.728744933510296, "learning_rate": 1.0797377926333355e-06, "loss": 0.4929, "step": 1223 }, { "epoch": 0.31698552976595123, "grad_norm": 6.491475501077239, "learning_rate": 1.079349868770256e-06, "loss": 0.4135, "step": 1224 }, { "epoch": 0.3172445048719692, "grad_norm": 6.106952798989202, "learning_rate": 1.0789615733458586e-06, "loss": 0.4119, "step": 1225 }, { "epoch": 0.3172445048719692, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.906832298136646, "eval_PRM F1 AUC": 0.8310633839706653, "eval_PRM F1 AUC (fixed)": 0.7930853850183341, "eval_PRM F1 Neg": 0.7058823529411765, "eval_PRM NPV": 0.6428571428571429, "eval_PRM Precision": 0.9358974358974359, "eval_PRM Recall": 0.8795180722891566, "eval_PRM Specificty": 0.782608695652174, "eval_loss": 0.2541673481464386, "eval_runtime": 14.5856, "eval_samples_per_second": 3.017, "eval_steps_per_second": 0.137, "step": 1225 }, { "epoch": 0.3175034799779871, "grad_norm": 3.0338791452656144, "learning_rate": 1.0785729066776864e-06, "loss": 0.3631, "step": 1226 }, { "epoch": 0.317762455084005, "grad_norm": 3.6359197190099266, "learning_rate": 1.0781838690835856e-06, "loss": 0.3624, "step": 1227 }, { "epoch": 0.318021430190023, "grad_norm": 4.79808651922732, "learning_rate": 1.0777944608817061e-06, "loss": 0.353, "step": 1228 }, { "epoch": 0.3182804052960409, "grad_norm": 3.2920462609639523, "learning_rate": 1.0774046823905007e-06, "loss": 0.3106, "step": 1229 }, { "epoch": 0.31853938040205887, "grad_norm": 5.105740527499564, "learning_rate": 1.0770145339287254e-06, "loss": 0.3778, "step": 1230 }, { "epoch": 0.31853938040205887, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8860759493670886, "eval_PRM F1 AUC": 0.8129910948140389, "eval_PRM F1 AUC (fixed)": 0.7794656888423258, "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": 0.264245867729187, "eval_runtime": 14.9669, "eval_samples_per_second": 2.94, "eval_steps_per_second": 0.134, "step": 1230 }, { "epoch": 0.3187983555080768, "grad_norm": 2.5851477388861754, "learning_rate": 1.0766240158154382e-06, "loss": 0.3024, "step": 1231 }, { "epoch": 0.31905733061409475, "grad_norm": 5.233677920244123, "learning_rate": 1.0762331283699997e-06, "loss": 0.3715, "step": 1232 }, { "epoch": 0.31931630572011266, "grad_norm": 2.695238376578191, "learning_rate": 1.0758418719120723e-06, "loss": 0.3182, "step": 1233 }, { "epoch": 0.31957528082613057, "grad_norm": 5.177819112800959, "learning_rate": 1.0754502467616207e-06, "loss": 0.3243, "step": 1234 }, { "epoch": 0.31983425593214854, "grad_norm": 4.881959830665304, "learning_rate": 1.0750582532389107e-06, "loss": 0.3412, "step": 1235 }, { "epoch": 0.31983425593214854, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8860759493670886, "eval_PRM F1 AUC": 0.8129910948140389, "eval_PRM F1 AUC (fixed)": 0.7802514405447879, "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": 0.26019617915153503, "eval_runtime": 14.7283, "eval_samples_per_second": 2.987, "eval_steps_per_second": 0.136, "step": 1235 }, { "epoch": 0.32009323103816645, "grad_norm": 3.170497689713328, "learning_rate": 1.074665891664509e-06, "loss": 0.3291, "step": 1236 }, { "epoch": 0.3203522061441844, "grad_norm": 4.371669431913452, "learning_rate": 1.0742731623592846e-06, "loss": 0.3604, "step": 1237 }, { "epoch": 0.3206111812502023, "grad_norm": 6.077137989216843, "learning_rate": 1.0738800656444053e-06, "loss": 0.3834, "step": 1238 }, { "epoch": 0.32087015635622024, "grad_norm": 4.292347676028572, "learning_rate": 1.0734866018413413e-06, "loss": 0.3677, "step": 1239 }, { "epoch": 0.3211291314622382, "grad_norm": 6.327860085742873, "learning_rate": 1.0730927712718614e-06, "loss": 0.3754, "step": 1240 }, { "epoch": 0.3211291314622382, "eval_PRM Accuracy": 0.8773584905660378, "eval_PRM F1": 0.9202453987730062, "eval_PRM F1 AUC": 0.8431115767417497, "eval_PRM F1 AUC (fixed)": 0.7854897852278679, "eval_PRM F1 Neg": 0.7346938775510204, "eval_PRM NPV": 0.6923076923076923, "eval_PRM Precision": 0.9375, "eval_PRM Recall": 0.9036144578313253, "eval_PRM Specificty": 0.782608695652174, "eval_loss": 0.2534755766391754, "eval_runtime": 14.3876, "eval_samples_per_second": 3.058, "eval_steps_per_second": 0.139, "step": 1240 }, { "epoch": 0.3213881065682561, "grad_norm": 5.573037635059935, "learning_rate": 1.0726985742580355e-06, "loss": 0.399, "step": 1241 }, { "epoch": 0.3216470816742741, "grad_norm": 5.900603463933119, "learning_rate": 1.0723040111222325e-06, "loss": 0.3801, "step": 1242 }, { "epoch": 0.321906056780292, "grad_norm": 7.0207752592900485, "learning_rate": 1.0719090821871208e-06, "loss": 0.3531, "step": 1243 }, { "epoch": 0.3221650318863099, "grad_norm": 4.396846712017389, "learning_rate": 1.0715137877756683e-06, "loss": 0.3045, "step": 1244 }, { "epoch": 0.32242400699232787, "grad_norm": 3.10027769254378, "learning_rate": 1.0711181282111415e-06, "loss": 0.3093, "step": 1245 }, { "epoch": 0.32242400699232787, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8860759493670886, "eval_PRM F1 AUC": 0.8129910948140389, "eval_PRM F1 AUC (fixed)": 0.7805133577789417, "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": 0.25553369522094727, "eval_runtime": 14.6501, "eval_samples_per_second": 3.003, "eval_steps_per_second": 0.137, "step": 1245 }, { "epoch": 0.3226829820983458, "grad_norm": 3.5643964755719777, "learning_rate": 1.0707221038171056e-06, "loss": 0.362, "step": 1246 }, { "epoch": 0.32294195720436375, "grad_norm": 4.684952259372974, "learning_rate": 1.0703257149174238e-06, "loss": 0.2909, "step": 1247 }, { "epoch": 0.32320093231038166, "grad_norm": 5.515834741111073, "learning_rate": 1.0699289618362579e-06, "loss": 0.3027, "step": 1248 }, { "epoch": 0.3234599074163996, "grad_norm": 3.841355440414757, "learning_rate": 1.0695318448980677e-06, "loss": 0.3314, "step": 1249 }, { "epoch": 0.32371888252241754, "grad_norm": 3.259683425761671, "learning_rate": 1.0691343644276095e-06, "loss": 0.3673, "step": 1250 }, { "epoch": 0.32371888252241754, "eval_PRM Accuracy": 0.8207547169811321, "eval_PRM F1": 0.8789808917197452, "eval_PRM F1 AUC": 0.8069669984284967, "eval_PRM F1 AUC (fixed)": 0.7815610267155578, "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": 0.2531985342502594, "eval_runtime": 14.525, "eval_samples_per_second": 3.029, "eval_steps_per_second": 0.138, "step": 1250 }, { "epoch": 0.32397785762843545, "grad_norm": 3.9427534379874714, "learning_rate": 1.0687365207499381e-06, "loss": 0.3362, "step": 1251 }, { "epoch": 0.3242368327344534, "grad_norm": 4.561687981910411, "learning_rate": 1.0683383141904046e-06, "loss": 0.2931, "step": 1252 }, { "epoch": 0.32449580784047133, "grad_norm": 4.234865118537673, "learning_rate": 1.0679397450746573e-06, "loss": 0.3682, "step": 1253 }, { "epoch": 0.3247547829464893, "grad_norm": 5.463852961847639, "learning_rate": 1.0675408137286403e-06, "loss": 0.3418, "step": 1254 }, { "epoch": 0.3250137580525072, "grad_norm": 3.8455546613574847, "learning_rate": 1.0671415204785948e-06, "loss": 0.3141, "step": 1255 }, { "epoch": 0.3250137580525072, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.9, "eval_PRM F1 AUC": 0.8250392875851232, "eval_PRM F1 AUC (fixed)": 0.7820848611838659, "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": 0.2530316114425659, "eval_runtime": 14.6154, "eval_samples_per_second": 3.011, "eval_steps_per_second": 0.137, "step": 1255 }, { "epoch": 0.3252727331585251, "grad_norm": 3.095249401099387, "learning_rate": 1.066741865651057e-06, "loss": 0.3307, "step": 1256 }, { "epoch": 0.3255317082645431, "grad_norm": 4.028023928603981, "learning_rate": 1.06634184957286e-06, "loss": 0.4218, "step": 1257 }, { "epoch": 0.325790683370561, "grad_norm": 3.2652804916367, "learning_rate": 1.0659414725711314e-06, "loss": 0.3492, "step": 1258 }, { "epoch": 0.32604965847657896, "grad_norm": 6.934006280547388, "learning_rate": 1.0655407349732942e-06, "loss": 0.3953, "step": 1259 }, { "epoch": 0.3263086335825969, "grad_norm": 5.3012226263917, "learning_rate": 1.065139637107066e-06, "loss": 0.3463, "step": 1260 }, { "epoch": 0.3263086335825969, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8930817610062893, "eval_PRM F1 AUC": 0.819015191199581, "eval_PRM F1 AUC (fixed)": 0.782870612886328, "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": 0.25366419553756714, "eval_runtime": 14.8022, "eval_samples_per_second": 2.973, "eval_steps_per_second": 0.135, "step": 1260 }, { "epoch": 0.3265676086886148, "grad_norm": 5.814350884881637, "learning_rate": 1.06473817930046e-06, "loss": 0.2951, "step": 1261 }, { "epoch": 0.32682658379463275, "grad_norm": 5.444085139661884, "learning_rate": 1.0643363618817825e-06, "loss": 0.4134, "step": 1262 }, { "epoch": 0.32708555890065066, "grad_norm": 9.9526531408028, "learning_rate": 1.0639341851796348e-06, "loss": 0.5132, "step": 1263 }, { "epoch": 0.32734453400666863, "grad_norm": 4.3954758322259835, "learning_rate": 1.0635316495229113e-06, "loss": 0.2878, "step": 1264 }, { "epoch": 0.32760350911268654, "grad_norm": 4.885386307335036, "learning_rate": 1.063128755240801e-06, "loss": 0.3412, "step": 1265 }, { "epoch": 0.32760350911268654, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8831168831168831, "eval_PRM F1 AUC": 0.8444211629125196, "eval_PRM F1 AUC (fixed)": 0.7799895233106339, "eval_PRM F1 Neg": 0.6896551724137931, "eval_PRM NPV": 0.5714285714285714, "eval_PRM Precision": 0.9577464788732394, "eval_PRM Recall": 0.8192771084337349, "eval_PRM Specificty": 0.8695652173913043, "eval_loss": 0.2644989490509033, "eval_runtime": 14.5446, "eval_samples_per_second": 3.025, "eval_steps_per_second": 0.138, "step": 1265 }, { "epoch": 0.32786248421870445, "grad_norm": 2.883792563856442, "learning_rate": 1.062725502662785e-06, "loss": 0.3397, "step": 1266 }, { "epoch": 0.3281214593247224, "grad_norm": 2.666039333847741, "learning_rate": 1.062321892118638e-06, "loss": 0.2813, "step": 1267 }, { "epoch": 0.32838043443074033, "grad_norm": 3.2249721487819643, "learning_rate": 1.0619179239384274e-06, "loss": 0.2993, "step": 1268 }, { "epoch": 0.3286394095367583, "grad_norm": 4.124141804714215, "learning_rate": 1.0615135984525133e-06, "loss": 0.3636, "step": 1269 }, { "epoch": 0.3288983846427762, "grad_norm": 4.252191767294118, "learning_rate": 1.0611089159915473e-06, "loss": 0.2837, "step": 1270 }, { "epoch": 0.3288983846427762, "eval_PRM Accuracy": 0.8207547169811321, "eval_PRM F1": 0.8789808917197452, "eval_PRM F1 AUC": 0.8069669984284967, "eval_PRM F1 AUC (fixed)": 0.7852278679937139, "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": 0.2600552439689636, "eval_runtime": 14.5574, "eval_samples_per_second": 3.023, "eval_steps_per_second": 0.137, "step": 1270 }, { "epoch": 0.3291573597487941, "grad_norm": 3.4884179116143366, "learning_rate": 1.0607038768864737e-06, "loss": 0.2719, "step": 1271 }, { "epoch": 0.3294163348548121, "grad_norm": 7.575658883276662, "learning_rate": 1.060298481468528e-06, "loss": 0.4286, "step": 1272 }, { "epoch": 0.32967530996083, "grad_norm": 2.7108707510626338, "learning_rate": 1.0598927300692372e-06, "loss": 0.2576, "step": 1273 }, { "epoch": 0.32993428506684797, "grad_norm": 4.385897164351068, "learning_rate": 1.0594866230204195e-06, "loss": 0.2862, "step": 1274 }, { "epoch": 0.3301932601728659, "grad_norm": 3.3979497479914396, "learning_rate": 1.0590801606541839e-06, "loss": 0.4011, "step": 1275 }, { "epoch": 0.3301932601728659, "eval_PRM Accuracy": 0.8679245283018868, "eval_PRM F1": 0.9135802469135802, "eval_PRM F1 AUC": 0.8370874803562075, "eval_PRM F1 AUC (fixed)": 0.7878470403352541, "eval_PRM F1 Neg": 0.72, "eval_PRM NPV": 0.6666666666666666, "eval_PRM Precision": 0.9367088607594937, "eval_PRM Recall": 0.891566265060241, "eval_PRM Specificty": 0.782608695652174, "eval_loss": 0.25437629222869873, "eval_runtime": 14.7355, "eval_samples_per_second": 2.986, "eval_steps_per_second": 0.136, "step": 1275 }, { "epoch": 0.33045223527888384, "grad_norm": 3.0241464378789225, "learning_rate": 1.0586733433029298e-06, "loss": 0.3214, "step": 1276 }, { "epoch": 0.33071121038490175, "grad_norm": 4.362546620704584, "learning_rate": 1.058266171299347e-06, "loss": 0.3757, "step": 1277 }, { "epoch": 0.33097018549091967, "grad_norm": 3.935891246180922, "learning_rate": 1.0578586449764157e-06, "loss": 0.3495, "step": 1278 }, { "epoch": 0.33122916059693763, "grad_norm": 3.5114047359210065, "learning_rate": 1.0574507646674054e-06, "loss": 0.3644, "step": 1279 }, { "epoch": 0.33148813570295554, "grad_norm": 6.0952351579317625, "learning_rate": 1.0570425307058752e-06, "loss": 0.4252, "step": 1280 }, { "epoch": 0.33148813570295554, "eval_PRM Accuracy": 0.8867924528301887, "eval_PRM F1": 0.926829268292683, "eval_PRM F1 AUC": 0.8491356731272918, "eval_PRM F1 AUC (fixed)": 0.7930853850183343, "eval_PRM F1 Neg": 0.75, "eval_PRM NPV": 0.72, "eval_PRM Precision": 0.9382716049382716, "eval_PRM Recall": 0.9156626506024096, "eval_PRM Specificty": 0.782608695652174, "eval_loss": 0.2512930631637573, "eval_runtime": 14.9129, "eval_samples_per_second": 2.95, "eval_steps_per_second": 0.134, "step": 1280 }, { "epoch": 0.3317471108089735, "grad_norm": 10.255531897227266, "learning_rate": 1.0566339434256734e-06, "loss": 0.3373, "step": 1281 }, { "epoch": 0.3320060859149914, "grad_norm": 2.6011141556068256, "learning_rate": 1.0562250031609372e-06, "loss": 0.201, "step": 1282 }, { "epoch": 0.33226506102100933, "grad_norm": 5.756435201271477, "learning_rate": 1.0558157102460929e-06, "loss": 0.3607, "step": 1283 }, { "epoch": 0.3325240361270273, "grad_norm": 3.256313178124926, "learning_rate": 1.0554060650158543e-06, "loss": 0.293, "step": 1284 }, { "epoch": 0.3327830112330452, "grad_norm": 4.038686227297466, "learning_rate": 1.0549960678052243e-06, "loss": 0.3206, "step": 1285 }, { "epoch": 0.3327830112330452, "eval_PRM Accuracy": 0.839622641509434, "eval_PRM F1": 0.8930817610062893, "eval_PRM F1 AUC": 0.819015191199581, "eval_PRM F1 AUC (fixed)": 0.7886327920377161, "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": 0.251687616109848, "eval_runtime": 14.924, "eval_samples_per_second": 2.948, "eval_steps_per_second": 0.134, "step": 1285 }, { "epoch": 0.3330419863390632, "grad_norm": 5.903103517069564, "learning_rate": 1.054585718949493e-06, "loss": 0.4008, "step": 1286 }, { "epoch": 0.3333009614450811, "grad_norm": 3.6280182528048797, "learning_rate": 1.0541750187842384e-06, "loss": 0.4035, "step": 1287 }, { "epoch": 0.333559936551099, "grad_norm": 7.746890673662093, "learning_rate": 1.0537639676453254e-06, "loss": 0.2507, "step": 1288 }, { "epoch": 0.33381891165711697, "grad_norm": 4.728549191922307, "learning_rate": 1.0533525658689064e-06, "loss": 0.3587, "step": 1289 }, { "epoch": 0.3340778867631349, "grad_norm": 8.67075750836026, "learning_rate": 1.0529408137914204e-06, "loss": 0.2709, "step": 1290 }, { "epoch": 0.3340778867631349, "eval_PRM Accuracy": 0.8113207547169812, "eval_PRM F1": 0.8717948717948718, "eval_PRM F1 AUC": 0.8009429020429545, "eval_PRM F1 AUC (fixed)": 0.7936092194866423, "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": 0.2578117549419403, "eval_runtime": 14.7797, "eval_samples_per_second": 2.977, "eval_steps_per_second": 0.135, "step": 1290 }, { "epoch": 0.33433686186915285, "grad_norm": 4.534289687153858, "learning_rate": 1.0525287117495928e-06, "loss": 0.3352, "step": 1291 }, { "epoch": 0.33459583697517076, "grad_norm": 2.4350765866908084, "learning_rate": 1.0521162600804348e-06, "loss": 0.2627, "step": 1292 }, { "epoch": 0.33485481208118867, "grad_norm": 2.993341013394517, "learning_rate": 1.0517034591212447e-06, "loss": 0.3248, "step": 1293 }, { "epoch": 0.33511378718720664, "grad_norm": 6.045274041684217, "learning_rate": 1.0512903092096048e-06, "loss": 0.4829, "step": 1294 }, { "epoch": 0.33537276229322455, "grad_norm": 4.920969311687047, "learning_rate": 1.0508768106833844e-06, "loss": 0.4081, "step": 1295 }, { "epoch": 0.33537276229322455, "eval_PRM Accuracy": 0.8207547169811321, "eval_PRM F1": 0.8789808917197452, "eval_PRM F1 AUC": 0.8069669984284967, "eval_PRM F1 AUC (fixed)": 0.8067050811943426, "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": 0.24843551218509674, "eval_runtime": 14.645, "eval_samples_per_second": 3.004, "eval_steps_per_second": 0.137, "step": 1295 }, { "epoch": 0.3356317373992425, "grad_norm": 3.8198314302235783, "learning_rate": 1.0504629638807365e-06, "loss": 0.3589, "step": 1296 }, { "epoch": 0.3358907125052604, "grad_norm": 3.2373637606236456, "learning_rate": 1.0500487691400998e-06, "loss": 0.3175, "step": 1297 }, { "epoch": 0.3361496876112784, "grad_norm": 4.29295357505977, "learning_rate": 1.0496342268001975e-06, "loss": 0.3697, "step": 1298 }, { "epoch": 0.3364086627172963, "grad_norm": 5.516686019813932, "learning_rate": 1.0492193372000364e-06, "loss": 0.324, "step": 1299 }, { "epoch": 0.3366676378233142, "grad_norm": 3.812344311395989, "learning_rate": 1.0488041006789084e-06, "loss": 0.4356, "step": 1300 }, { "epoch": 0.3366676378233142, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8860759493670886, "eval_PRM F1 AUC": 0.8129910948140389, "eval_PRM F1 AUC (fixed)": 0.8040859088528025, "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": 0.24917888641357422, "eval_runtime": 14.8531, "eval_samples_per_second": 2.962, "eval_steps_per_second": 0.135, "step": 1300 }, { "epoch": 0.3369266129293322, "grad_norm": 3.196056765448634, "learning_rate": 1.0483885175763879e-06, "loss": 0.3073, "step": 1301 }, { "epoch": 0.3371855880353501, "grad_norm": 4.95895879113035, "learning_rate": 1.047972588232333e-06, "loss": 0.4303, "step": 1302 }, { "epoch": 0.33744456314136806, "grad_norm": 2.4644183637611103, "learning_rate": 1.047556312986886e-06, "loss": 0.2904, "step": 1303 }, { "epoch": 0.33770353824738597, "grad_norm": 5.294206024896391, "learning_rate": 1.0471396921804708e-06, "loss": 0.3325, "step": 1304 }, { "epoch": 0.3379625133534039, "grad_norm": 3.3704941228453094, "learning_rate": 1.0467227261537944e-06, "loss": 0.2584, "step": 1305 }, { "epoch": 0.3379625133534039, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.906832298136646, "eval_PRM F1 AUC": 0.8310633839706653, "eval_PRM F1 AUC (fixed)": 0.7954426401257203, "eval_PRM F1 Neg": 0.7058823529411765, "eval_PRM NPV": 0.6428571428571429, "eval_PRM Precision": 0.9358974358974359, "eval_PRM Recall": 0.8795180722891566, "eval_PRM Specificty": 0.782608695652174, "eval_loss": 0.25340020656585693, "eval_runtime": 14.6219, "eval_samples_per_second": 3.009, "eval_steps_per_second": 0.137, "step": 1305 }, { "epoch": 0.33822148845942185, "grad_norm": 4.249922053779406, "learning_rate": 1.0463054152478464e-06, "loss": 0.3716, "step": 1306 }, { "epoch": 0.33848046356543976, "grad_norm": 4.501080318218483, "learning_rate": 1.0458877598038978e-06, "loss": 0.2519, "step": 1307 }, { "epoch": 0.3387394386714577, "grad_norm": 2.8975562640932733, "learning_rate": 1.045469760163502e-06, "loss": 0.3074, "step": 1308 }, { "epoch": 0.33899841377747564, "grad_norm": 8.021087356933908, "learning_rate": 1.0450514166684933e-06, "loss": 0.3931, "step": 1309 }, { "epoch": 0.33925738888349355, "grad_norm": 7.004507103777712, "learning_rate": 1.0446327296609876e-06, "loss": 0.4529, "step": 1310 }, { "epoch": 0.33925738888349355, "eval_PRM Accuracy": 0.8113207547169812, "eval_PRM F1": 0.8717948717948718, "eval_PRM F1 AUC": 0.8009429020429545, "eval_PRM F1 AUC (fixed)": 0.7833944473546359, "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": 0.2645565867424011, "eval_runtime": 15.0492, "eval_samples_per_second": 2.924, "eval_steps_per_second": 0.133, "step": 1310 }, { "epoch": 0.3395163639895115, "grad_norm": 5.034753487824166, "learning_rate": 1.0442136994833815e-06, "loss": 0.3002, "step": 1311 }, { "epoch": 0.3397753390955294, "grad_norm": 4.159325525124912, "learning_rate": 1.0437943264783528e-06, "loss": 0.4791, "step": 1312 }, { "epoch": 0.3400343142015474, "grad_norm": 4.685979244764695, "learning_rate": 1.0433746109888585e-06, "loss": 0.3015, "step": 1313 }, { "epoch": 0.3402932893075653, "grad_norm": 3.849565505765079, "learning_rate": 1.042954553358137e-06, "loss": 0.333, "step": 1314 }, { "epoch": 0.3405522644135832, "grad_norm": 3.645168751223873, "learning_rate": 1.0425341539297054e-06, "loss": 0.3582, "step": 1315 }, { "epoch": 0.3405522644135832, "eval_PRM Accuracy": 0.8018867924528302, "eval_PRM F1": 0.864516129032258, "eval_PRM F1 AUC": 0.7949188056574124, "eval_PRM F1 AUC (fixed)": 0.7778941854374017, "eval_PRM F1 Neg": 0.631578947368421, "eval_PRM NPV": 0.5294117647058824, "eval_PRM Precision": 0.9305555555555556, "eval_PRM Recall": 0.8072289156626506, "eval_PRM Specificty": 0.782608695652174, "eval_loss": 0.27547842264175415, "eval_runtime": 14.6792, "eval_samples_per_second": 2.997, "eval_steps_per_second": 0.136, "step": 1315 }, { "epoch": 0.3408112395196012, "grad_norm": 2.9138626397617093, "learning_rate": 1.0421134130473612e-06, "loss": 0.3854, "step": 1316 }, { "epoch": 0.3410702146256191, "grad_norm": 2.857443915356315, "learning_rate": 1.0416923310551805e-06, "loss": 0.3204, "step": 1317 }, { "epoch": 0.34132918973163706, "grad_norm": 6.003891257495076, "learning_rate": 1.0412709082975184e-06, "loss": 0.2993, "step": 1318 }, { "epoch": 0.341588164837655, "grad_norm": 4.153935299144768, "learning_rate": 1.0408491451190093e-06, "loss": 0.389, "step": 1319 }, { "epoch": 0.34184713994367294, "grad_norm": 4.6586000639572305, "learning_rate": 1.0404270418645654e-06, "loss": 0.3316, "step": 1320 }, { "epoch": 0.34184713994367294, "eval_PRM Accuracy": 0.8207547169811321, "eval_PRM F1": 0.8789808917197452, "eval_PRM F1 AUC": 0.8069669984284967, "eval_PRM F1 AUC (fixed)": 0.7768465165007857, "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": 0.27451756596565247, "eval_runtime": 15.085, "eval_samples_per_second": 2.917, "eval_steps_per_second": 0.133, "step": 1320 }, { "epoch": 0.34210611504969085, "grad_norm": 4.479214897680437, "learning_rate": 1.0400045988793768e-06, "loss": 0.3974, "step": 1321 }, { "epoch": 0.34236509015570876, "grad_norm": 3.175295099148635, "learning_rate": 1.0395818165089125e-06, "loss": 0.3704, "step": 1322 }, { "epoch": 0.34262406526172673, "grad_norm": 4.047456516055003, "learning_rate": 1.039158695098918e-06, "loss": 0.3562, "step": 1323 }, { "epoch": 0.34288304036774464, "grad_norm": 3.496307850241482, "learning_rate": 1.0387352349954158e-06, "loss": 0.3671, "step": 1324 }, { "epoch": 0.3431420154737626, "grad_norm": 6.735416072169709, "learning_rate": 1.038311436544707e-06, "loss": 0.4127, "step": 1325 }, { "epoch": 0.3431420154737626, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8860759493670886, "eval_PRM F1 AUC": 0.8129910948140389, "eval_PRM F1 AUC (fixed)": 0.783918281822944, "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": 0.2647566795349121, "eval_runtime": 14.6853, "eval_samples_per_second": 2.996, "eval_steps_per_second": 0.136, "step": 1325 }, { "epoch": 0.3434009905797805, "grad_norm": 3.7720284176982295, "learning_rate": 1.0378873000933675e-06, "loss": 0.2992, "step": 1326 }, { "epoch": 0.34365996568579843, "grad_norm": 10.021450252138205, "learning_rate": 1.037462825988251e-06, "loss": 0.4193, "step": 1327 }, { "epoch": 0.3439189407918164, "grad_norm": 3.72757558238376, "learning_rate": 1.0370380145764868e-06, "loss": 0.2907, "step": 1328 }, { "epoch": 0.3441779158978343, "grad_norm": 5.12031159635696, "learning_rate": 1.03661286620548e-06, "loss": 0.4486, "step": 1329 }, { "epoch": 0.3444368910038523, "grad_norm": 2.935652764654381, "learning_rate": 1.0361873812229111e-06, "loss": 0.226, "step": 1330 }, { "epoch": 0.3444368910038523, "eval_PRM Accuracy": 0.8113207547169812, "eval_PRM F1": 0.8717948717948718, "eval_PRM F1 AUC": 0.8009429020429545, "eval_PRM F1 AUC (fixed)": 0.7771084337349398, "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": 0.2631690204143524, "eval_runtime": 14.4887, "eval_samples_per_second": 3.037, "eval_steps_per_second": 0.138, "step": 1330 }, { "epoch": 0.3446958661098702, "grad_norm": 3.485190804963136, "learning_rate": 1.0357615599767362e-06, "loss": 0.3748, "step": 1331 }, { "epoch": 0.3449548412158881, "grad_norm": 4.250749573394851, "learning_rate": 1.0353354028151864e-06, "loss": 0.3145, "step": 1332 }, { "epoch": 0.34521381632190606, "grad_norm": 7.923349942246661, "learning_rate": 1.0349089100867673e-06, "loss": 0.2599, "step": 1333 }, { "epoch": 0.345472791427924, "grad_norm": 6.3912608889410905, "learning_rate": 1.034482082140259e-06, "loss": 0.3285, "step": 1334 }, { "epoch": 0.34573176653394194, "grad_norm": 4.472668850805793, "learning_rate": 1.034054919324716e-06, "loss": 0.3863, "step": 1335 }, { "epoch": 0.34573176653394194, "eval_PRM Accuracy": 0.8113207547169812, "eval_PRM F1": 0.8701298701298701, "eval_PRM F1 AUC": 0.8166579360921948, "eval_PRM F1 AUC (fixed)": 0.784442116291252, "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": 0.25815823674201965, "eval_runtime": 14.7537, "eval_samples_per_second": 2.982, "eval_steps_per_second": 0.136, "step": 1335 }, { "epoch": 0.34599074163995985, "grad_norm": 4.239365211227992, "learning_rate": 1.033627421989466e-06, "loss": 0.3723, "step": 1336 }, { "epoch": 0.3462497167459778, "grad_norm": 3.6891728911365607, "learning_rate": 1.033199590484111e-06, "loss": 0.3842, "step": 1337 }, { "epoch": 0.34650869185199573, "grad_norm": 4.1821840352227255, "learning_rate": 1.0327714251585257e-06, "loss": 0.4589, "step": 1338 }, { "epoch": 0.34676766695801364, "grad_norm": 3.2232499807001544, "learning_rate": 1.0323429263628582e-06, "loss": 0.299, "step": 1339 }, { "epoch": 0.3470266420640316, "grad_norm": 5.441098902376691, "learning_rate": 1.0319140944475291e-06, "loss": 0.4802, "step": 1340 }, { "epoch": 0.3470266420640316, "eval_PRM Accuracy": 0.8207547169811321, "eval_PRM F1": 0.8774193548387097, "eval_PRM F1 AUC": 0.8226820324777371, "eval_PRM F1 AUC (fixed)": 0.786013619696176, "eval_PRM F1 Neg": 0.6666666666666666, "eval_PRM NPV": 0.5588235294117647, "eval_PRM Precision": 0.9444444444444444, "eval_PRM Recall": 0.8192771084337349, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.2553994953632355, "eval_runtime": 14.633, "eval_samples_per_second": 3.007, "eval_steps_per_second": 0.137, "step": 1340 }, { "epoch": 0.3472856171700495, "grad_norm": 4.884035995601269, "learning_rate": 1.0314849297632318e-06, "loss": 0.3386, "step": 1341 }, { "epoch": 0.3475445922760675, "grad_norm": 3.2695838689315853, "learning_rate": 1.031055432660931e-06, "loss": 0.3565, "step": 1342 }, { "epoch": 0.3478035673820854, "grad_norm": 4.304615322479972, "learning_rate": 1.0306256034918637e-06, "loss": 0.3045, "step": 1343 }, { "epoch": 0.3480625424881033, "grad_norm": 3.3317439490288208, "learning_rate": 1.0301954426075391e-06, "loss": 0.343, "step": 1344 }, { "epoch": 0.3483215175941213, "grad_norm": 3.656639575005203, "learning_rate": 1.0297649503597366e-06, "loss": 0.3504, "step": 1345 }, { "epoch": 0.3483215175941213, "eval_PRM Accuracy": 0.8207547169811321, "eval_PRM F1": 0.8774193548387097, "eval_PRM F1 AUC": 0.8226820324777371, "eval_PRM F1 AUC (fixed)": 0.784704033525406, "eval_PRM F1 Neg": 0.6666666666666666, "eval_PRM NPV": 0.5588235294117647, "eval_PRM Precision": 0.9444444444444444, "eval_PRM Recall": 0.8192771084337349, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.25615188479423523, "eval_runtime": 14.6776, "eval_samples_per_second": 2.998, "eval_steps_per_second": 0.136, "step": 1345 }, { "epoch": 0.3485804927001392, "grad_norm": 3.766623424533264, "learning_rate": 1.0293341271005073e-06, "loss": 0.284, "step": 1346 }, { "epoch": 0.34883946780615716, "grad_norm": 4.488867287571511, "learning_rate": 1.0289029731821725e-06, "loss": 0.3686, "step": 1347 }, { "epoch": 0.34909844291217507, "grad_norm": 3.1747978172581894, "learning_rate": 1.0284714889573248e-06, "loss": 0.3598, "step": 1348 }, { "epoch": 0.349357418018193, "grad_norm": 3.3137262712445894, "learning_rate": 1.0280396747788255e-06, "loss": 0.3896, "step": 1349 }, { "epoch": 0.34961639312421094, "grad_norm": 4.6326692615664316, "learning_rate": 1.027607530999807e-06, "loss": 0.3655, "step": 1350 }, { "epoch": 0.34961639312421094, "eval_PRM Accuracy": 0.8207547169811321, "eval_PRM F1": 0.8789808917197452, "eval_PRM F1 AUC": 0.8069669984284967, "eval_PRM F1 AUC (fixed)": 0.7915138816134102, "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": 0.2515624165534973, "eval_runtime": 14.6399, "eval_samples_per_second": 3.005, "eval_steps_per_second": 0.137, "step": 1350 }, { "epoch": 0.34987536823022886, "grad_norm": 5.029034145916732, "learning_rate": 1.0271750579736707e-06, "loss": 0.2834, "step": 1351 }, { "epoch": 0.3501343433362468, "grad_norm": 3.2262110030593965, "learning_rate": 1.026742256054087e-06, "loss": 0.3099, "step": 1352 }, { "epoch": 0.35039331844226473, "grad_norm": 4.277241669093875, "learning_rate": 1.0263091255949962e-06, "loss": 0.3033, "step": 1353 }, { "epoch": 0.35065229354828265, "grad_norm": 4.219679308248584, "learning_rate": 1.0258756669506065e-06, "loss": 0.3592, "step": 1354 }, { "epoch": 0.3509112686543006, "grad_norm": 3.873414393356266, "learning_rate": 1.0254418804753941e-06, "loss": 0.3591, "step": 1355 }, { "epoch": 0.3509112686543006, "eval_PRM Accuracy": 0.8113207547169812, "eval_PRM F1": 0.8717948717948718, "eval_PRM F1 AUC": 0.8009429020429545, "eval_PRM F1 AUC (fixed)": 0.7941330539549503, "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": 0.25376003980636597, "eval_runtime": 14.6093, "eval_samples_per_second": 3.012, "eval_steps_per_second": 0.137, "step": 1355 }, { "epoch": 0.3511702437603185, "grad_norm": 3.596581202851472, "learning_rate": 1.0250077665241045e-06, "loss": 0.3553, "step": 1356 }, { "epoch": 0.3514292188663365, "grad_norm": 3.0834953882859586, "learning_rate": 1.02457332545175e-06, "loss": 0.2994, "step": 1357 }, { "epoch": 0.3516881939723544, "grad_norm": 3.5752729654852904, "learning_rate": 1.0241385576136109e-06, "loss": 0.3762, "step": 1358 }, { "epoch": 0.35194716907837237, "grad_norm": 6.601488366950558, "learning_rate": 1.0237034633652345e-06, "loss": 0.3852, "step": 1359 }, { "epoch": 0.3522061441843903, "grad_norm": 3.8142506524141284, "learning_rate": 1.0232680430624352e-06, "loss": 0.3078, "step": 1360 }, { "epoch": 0.3522061441843903, "eval_PRM Accuracy": 0.8207547169811321, "eval_PRM F1": 0.8789808917197452, "eval_PRM F1 AUC": 0.8069669984284967, "eval_PRM F1 AUC (fixed)": 0.7977998952331063, "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": 0.24866238236427307, "eval_runtime": 14.679, "eval_samples_per_second": 2.997, "eval_steps_per_second": 0.136, "step": 1360 }, { "epoch": 0.3524651192904082, "grad_norm": 7.2589367790025685, "learning_rate": 1.022832297061294e-06, "loss": 0.3643, "step": 1361 }, { "epoch": 0.35272409439642616, "grad_norm": 6.2858784402296495, "learning_rate": 1.022396225718158e-06, "loss": 0.4068, "step": 1362 }, { "epoch": 0.35298306950244407, "grad_norm": 7.788782355650248, "learning_rate": 1.021959829389641e-06, "loss": 0.3245, "step": 1363 }, { "epoch": 0.35324204460846204, "grad_norm": 3.599017538260975, "learning_rate": 1.021523108432622e-06, "loss": 0.3178, "step": 1364 }, { "epoch": 0.35350101971447995, "grad_norm": 6.426781605853537, "learning_rate": 1.0210860632042453e-06, "loss": 0.4676, "step": 1365 }, { "epoch": 0.35350101971447995, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8860759493670886, "eval_PRM F1 AUC": 0.8129910948140389, "eval_PRM F1 AUC (fixed)": 0.8038239916186485, "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": 0.24092192947864532, "eval_runtime": 14.7974, "eval_samples_per_second": 2.974, "eval_steps_per_second": 0.135, "step": 1365 }, { "epoch": 0.35375999482049786, "grad_norm": 4.851902543247538, "learning_rate": 1.0206486940619207e-06, "loss": 0.3503, "step": 1366 }, { "epoch": 0.3540189699265158, "grad_norm": 3.260733464307621, "learning_rate": 1.0202110013633235e-06, "loss": 0.2659, "step": 1367 }, { "epoch": 0.35427794503253374, "grad_norm": 4.77562481552069, "learning_rate": 1.0197729854663925e-06, "loss": 0.3689, "step": 1368 }, { "epoch": 0.3545369201385517, "grad_norm": 6.9020702863177705, "learning_rate": 1.0193346467293314e-06, "loss": 0.3497, "step": 1369 }, { "epoch": 0.3547958952445696, "grad_norm": 2.8641402943880014, "learning_rate": 1.0188959855106082e-06, "loss": 0.2729, "step": 1370 }, { "epoch": 0.3547958952445696, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8846153846153846, "eval_PRM F1 AUC": 0.8287061288632793, "eval_PRM F1 AUC (fixed)": 0.7909900471451021, "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": 0.2476367950439453, "eval_runtime": 14.9235, "eval_samples_per_second": 2.948, "eval_steps_per_second": 0.134, "step": 1370 }, { "epoch": 0.3550548703505875, "grad_norm": 4.912052749316654, "learning_rate": 1.0184570021689536e-06, "loss": 0.3502, "step": 1371 }, { "epoch": 0.3553138454566055, "grad_norm": 5.308943310093734, "learning_rate": 1.018017697063363e-06, "loss": 0.3365, "step": 1372 }, { "epoch": 0.3555728205626234, "grad_norm": 3.630549995233943, "learning_rate": 1.0175780705530935e-06, "loss": 0.4088, "step": 1373 }, { "epoch": 0.35583179566864137, "grad_norm": 7.2185639503008545, "learning_rate": 1.0171381229976667e-06, "loss": 0.3197, "step": 1374 }, { "epoch": 0.3560907707746593, "grad_norm": 8.163397684150947, "learning_rate": 1.0166978547568656e-06, "loss": 0.2647, "step": 1375 }, { "epoch": 0.3560907707746593, "eval_PRM Accuracy": 0.8207547169811321, "eval_PRM F1": 0.8789808917197452, "eval_PRM F1 AUC": 0.8069669984284967, "eval_PRM F1 AUC (fixed)": 0.788108957569408, "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": 0.24819716811180115, "eval_runtime": 14.7698, "eval_samples_per_second": 2.979, "eval_steps_per_second": 0.135, "step": 1375 }, { "epoch": 0.3563497458806772, "grad_norm": 6.313737507680141, "learning_rate": 1.0162572661907358e-06, "loss": 0.3255, "step": 1376 }, { "epoch": 0.35660872098669516, "grad_norm": 4.288938839171069, "learning_rate": 1.0158163576595846e-06, "loss": 0.3833, "step": 1377 }, { "epoch": 0.35686769609271307, "grad_norm": 4.158728708310663, "learning_rate": 1.0153751295239813e-06, "loss": 0.464, "step": 1378 }, { "epoch": 0.35712667119873104, "grad_norm": 6.984820199164077, "learning_rate": 1.0149335821447565e-06, "loss": 0.3083, "step": 1379 }, { "epoch": 0.35738564630474895, "grad_norm": 7.100930397132101, "learning_rate": 1.0144917158830015e-06, "loss": 0.3468, "step": 1380 }, { "epoch": 0.35738564630474895, "eval_PRM Accuracy": 0.8207547169811321, "eval_PRM F1": 0.8789808917197452, "eval_PRM F1 AUC": 0.8069669984284967, "eval_PRM F1 AUC (fixed)": 0.7870612886327921, "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": 0.2420676201581955, "eval_runtime": 14.5031, "eval_samples_per_second": 3.034, "eval_steps_per_second": 0.138, "step": 1380 }, { "epoch": 0.3576446214107669, "grad_norm": 3.402060831951842, "learning_rate": 1.014049531100069e-06, "loss": 0.3431, "step": 1381 }, { "epoch": 0.3579035965167848, "grad_norm": 4.1631232404476215, "learning_rate": 1.0136070281575718e-06, "loss": 0.4072, "step": 1382 }, { "epoch": 0.35816257162280274, "grad_norm": 2.96097909341508, "learning_rate": 1.013164207417383e-06, "loss": 0.2762, "step": 1383 }, { "epoch": 0.3584215467288207, "grad_norm": 3.589280367075051, "learning_rate": 1.0127210692416356e-06, "loss": 0.3656, "step": 1384 }, { "epoch": 0.3586805218348386, "grad_norm": 10.198150056204957, "learning_rate": 1.012277613992722e-06, "loss": 0.3689, "step": 1385 }, { "epoch": 0.3586805218348386, "eval_PRM Accuracy": 0.8867924528301887, "eval_PRM F1": 0.927710843373494, "eval_PRM F1 AUC": 0.8334206390780513, "eval_PRM F1 AUC (fixed)": 0.7888947092718701, "eval_PRM F1 Neg": 0.7391304347826086, "eval_PRM NPV": 0.7391304347826086, "eval_PRM Precision": 0.927710843373494, "eval_PRM Recall": 0.927710843373494, "eval_PRM Specificty": 0.7391304347826086, "eval_loss": 0.2418714016675949, "eval_runtime": 14.7389, "eval_samples_per_second": 2.985, "eval_steps_per_second": 0.136, "step": 1385 }, { "epoch": 0.3589394969408566, "grad_norm": 9.174095092421487, "learning_rate": 1.011833842033294e-06, "loss": 0.385, "step": 1386 }, { "epoch": 0.3591984720468745, "grad_norm": 5.068293049956948, "learning_rate": 1.0113897537262627e-06, "loss": 0.314, "step": 1387 }, { "epoch": 0.3594574471528924, "grad_norm": 6.1186335582351745, "learning_rate": 1.0109453494347973e-06, "loss": 0.3665, "step": 1388 }, { "epoch": 0.3597164222589104, "grad_norm": 3.0993215013454383, "learning_rate": 1.010500629522326e-06, "loss": 0.3367, "step": 1389 }, { "epoch": 0.3599753973649283, "grad_norm": 7.933153861235534, "learning_rate": 1.0100555943525348e-06, "loss": 0.3925, "step": 1390 }, { "epoch": 0.3599753973649283, "eval_PRM Accuracy": 0.8207547169811321, "eval_PRM F1": 0.8789808917197452, "eval_PRM F1 AUC": 0.8069669984284967, "eval_PRM F1 AUC (fixed)": 0.7891566265060241, "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": 0.24478855729103088, "eval_runtime": 14.5608, "eval_samples_per_second": 3.022, "eval_steps_per_second": 0.137, "step": 1390 }, { "epoch": 0.36023437247094625, "grad_norm": 3.69707566141284, "learning_rate": 1.0096102442893676e-06, "loss": 0.4158, "step": 1391 }, { "epoch": 0.36049334757696416, "grad_norm": 5.0661368363238175, "learning_rate": 1.0091645796970257e-06, "loss": 0.3552, "step": 1392 }, { "epoch": 0.3607523226829821, "grad_norm": 6.248551686579097, "learning_rate": 1.008718600939968e-06, "loss": 0.342, "step": 1393 }, { "epoch": 0.36101129778900004, "grad_norm": 6.792166057295491, "learning_rate": 1.0082723083829096e-06, "loss": 0.323, "step": 1394 }, { "epoch": 0.36127027289501795, "grad_norm": 3.779026986393951, "learning_rate": 1.007825702390823e-06, "loss": 0.3358, "step": 1395 }, { "epoch": 0.36127027289501795, "eval_PRM Accuracy": 0.8207547169811321, "eval_PRM F1": 0.8789808917197452, "eval_PRM F1 AUC": 0.8069669984284967, "eval_PRM F1 AUC (fixed)": 0.784180199057098, "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": 0.25908514857292175, "eval_runtime": 14.8271, "eval_samples_per_second": 2.968, "eval_steps_per_second": 0.135, "step": 1395 }, { "epoch": 0.3615292480010359, "grad_norm": 7.177709456902315, "learning_rate": 1.0073787833289364e-06, "loss": 0.339, "step": 1396 }, { "epoch": 0.36178822310705383, "grad_norm": 4.534515548735088, "learning_rate": 1.0069315515627346e-06, "loss": 0.3624, "step": 1397 }, { "epoch": 0.36204719821307174, "grad_norm": 3.516634251586425, "learning_rate": 1.0064840074579578e-06, "loss": 0.3755, "step": 1398 }, { "epoch": 0.3623061733190897, "grad_norm": 2.9834443537357753, "learning_rate": 1.0060361513806013e-06, "loss": 0.2756, "step": 1399 }, { "epoch": 0.3625651484251076, "grad_norm": 6.302837815954074, "learning_rate": 1.0055879836969164e-06, "loss": 0.3814, "step": 1400 }, { "epoch": 0.3625651484251076, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8860759493670886, "eval_PRM F1 AUC": 0.8129910948140389, "eval_PRM F1 AUC (fixed)": 0.7894185437401781, "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": 0.2521251142024994, "eval_runtime": 14.7735, "eval_samples_per_second": 2.978, "eval_steps_per_second": 0.135, "step": 1400 }, { "epoch": 0.3628241235311256, "grad_norm": 5.898802333208595, "learning_rate": 1.0051395047734083e-06, "loss": 0.3526, "step": 1401 }, { "epoch": 0.3630830986371435, "grad_norm": 3.827032648322295, "learning_rate": 1.004690714976837e-06, "loss": 0.3202, "step": 1402 }, { "epoch": 0.36334207374316146, "grad_norm": 3.767391984694187, "learning_rate": 1.0042416146742175e-06, "loss": 0.3372, "step": 1403 }, { "epoch": 0.3636010488491794, "grad_norm": 4.634860029918238, "learning_rate": 1.0037922042328177e-06, "loss": 0.4127, "step": 1404 }, { "epoch": 0.3638600239551973, "grad_norm": 3.1248109063152127, "learning_rate": 1.0033424840201595e-06, "loss": 0.3281, "step": 1405 }, { "epoch": 0.3638600239551973, "eval_PRM Accuracy": 0.8207547169811321, "eval_PRM F1": 0.8789808917197452, "eval_PRM F1 AUC": 0.8069669984284967, "eval_PRM F1 AUC (fixed)": 0.788108957569408, "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": 0.24834133684635162, "eval_runtime": 14.6414, "eval_samples_per_second": 3.005, "eval_steps_per_second": 0.137, "step": 1405 }, { "epoch": 0.36411899906121525, "grad_norm": 5.545453446821229, "learning_rate": 1.0028924544040179e-06, "loss": 0.384, "step": 1406 }, { "epoch": 0.36437797416723317, "grad_norm": 4.211034137415862, "learning_rate": 1.0024421157524216e-06, "loss": 0.3684, "step": 1407 }, { "epoch": 0.36463694927325113, "grad_norm": 4.057108424391074, "learning_rate": 1.0019914684336517e-06, "loss": 0.4328, "step": 1408 }, { "epoch": 0.36489592437926904, "grad_norm": 3.3320674863925537, "learning_rate": 1.0015405128162412e-06, "loss": 0.3834, "step": 1409 }, { "epoch": 0.36515489948528695, "grad_norm": 8.681780340750507, "learning_rate": 1.0010892492689758e-06, "loss": 0.351, "step": 1410 }, { "epoch": 0.36515489948528695, "eval_PRM Accuracy": 0.8301886792452831, "eval_PRM F1": 0.8860759493670886, "eval_PRM F1 AUC": 0.8129910948140389, "eval_PRM F1 AUC (fixed)": 0.788108957569408, "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": 0.24854445457458496, "eval_runtime": 14.7704, "eval_samples_per_second": 2.979, "eval_steps_per_second": 0.135, "step": 1410 }, { "epoch": 0.3654138745913049, "grad_norm": 3.47683496398654, "learning_rate": 1.000637678160893e-06, "loss": 0.3208, "step": 1411 }, { "epoch": 0.36567284969732283, "grad_norm": 3.9465011870396873, "learning_rate": 1.0001857998612815e-06, "loss": 0.4328, "step": 1412 }, { "epoch": 0.3659318248033408, "grad_norm": 2.7448030317458754, "learning_rate": 9.997336147396816e-07, "loss": 0.3005, "step": 1413 }, { "epoch": 0.3661907999093587, "grad_norm": 4.164297634945185, "learning_rate": 9.99281123165884e-07, "loss": 0.2506, "step": 1414 }, { "epoch": 0.3664497750153766, "grad_norm": 3.4588831820316877, "learning_rate": 9.988283255099306e-07, "loss": 0.3232, "step": 1415 }, { "epoch": 0.3664497750153766, "eval_PRM Accuracy": 0.8773584905660378, "eval_PRM F1": 0.9202453987730062, "eval_PRM F1 AUC": 0.8431115767417497, "eval_PRM F1 AUC (fixed)": 0.7909900471451021, "eval_PRM F1 Neg": 0.7346938775510204, "eval_PRM NPV": 0.6923076923076923, "eval_PRM Precision": 0.9375, "eval_PRM Recall": 0.9036144578313253, "eval_PRM Specificty": 0.782608695652174, "eval_loss": 0.2446984201669693, "eval_runtime": 14.2678, "eval_samples_per_second": 3.084, "eval_steps_per_second": 0.14, "step": 1415 }, { "epoch": 0.3667087501213946, "grad_norm": 5.711828100772439, "learning_rate": 9.983752221421134e-07, "loss": 0.3872, "step": 1416 }, { "epoch": 0.3669677252274125, "grad_norm": 3.83260643758672, "learning_rate": 9.979218134329739e-07, "loss": 0.2705, "step": 1417 }, { "epoch": 0.36722670033343047, "grad_norm": 7.755924384672744, "learning_rate": 9.974680997533043e-07, "loss": 0.4048, "step": 1418 }, { "epoch": 0.3674856754394484, "grad_norm": 3.4136788910922937, "learning_rate": 9.970140814741452e-07, "loss": 0.2815, "step": 1419 }, { "epoch": 0.3677446505454663, "grad_norm": 3.3743340350957682, "learning_rate": 9.965597589667871e-07, "loss": 0.3282, "step": 1420 }, { "epoch": 0.3677446505454663, "eval_PRM Accuracy": 0.8867924528301887, "eval_PRM F1": 0.926829268292683, "eval_PRM F1 AUC": 0.8491356731272918, "eval_PRM F1 AUC (fixed)": 0.8014667365112624, "eval_PRM F1 Neg": 0.75, "eval_PRM NPV": 0.72, "eval_PRM Precision": 0.9382716049382716, "eval_PRM Recall": 0.9156626506024096, "eval_PRM Specificty": 0.782608695652174, "eval_loss": 0.23952960968017578, "eval_runtime": 14.9505, "eval_samples_per_second": 2.943, "eval_steps_per_second": 0.134, "step": 1420 }, { "epoch": 0.36800362565148426, "grad_norm": 3.4721221499465535, "learning_rate": 9.961051326027688e-07, "loss": 0.2982, "step": 1421 }, { "epoch": 0.36826260075750217, "grad_norm": 6.65201998896593, "learning_rate": 9.956502027538776e-07, "loss": 0.3241, "step": 1422 }, { "epoch": 0.36852157586352013, "grad_norm": 3.462624885557569, "learning_rate": 9.951949697921493e-07, "loss": 0.2788, "step": 1423 }, { "epoch": 0.36878055096953805, "grad_norm": 3.3189609048980264, "learning_rate": 9.947394340898675e-07, "loss": 0.3762, "step": 1424 }, { "epoch": 0.369039526075556, "grad_norm": 2.9790395939751733, "learning_rate": 9.94283596019563e-07, "loss": 0.2558, "step": 1425 }, { "epoch": 0.369039526075556, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.9, "eval_PRM F1 AUC": 0.8250392875851232, "eval_PRM F1 AUC (fixed)": 0.8043478260869564, "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": 0.2378481775522232, "eval_runtime": 14.6257, "eval_samples_per_second": 3.008, "eval_steps_per_second": 0.137, "step": 1425 }, { "epoch": 0.3692985011815739, "grad_norm": 3.4004104085051905, "learning_rate": 9.938274559540144e-07, "loss": 0.3402, "step": 1426 }, { "epoch": 0.36955747628759184, "grad_norm": 4.1750959101791745, "learning_rate": 9.933710142662468e-07, "loss": 0.2943, "step": 1427 }, { "epoch": 0.3698164513936098, "grad_norm": 3.2247794918939894, "learning_rate": 9.92914271329532e-07, "loss": 0.35, "step": 1428 }, { "epoch": 0.3700754264996277, "grad_norm": 5.112086961413494, "learning_rate": 9.924572275173889e-07, "loss": 0.2642, "step": 1429 }, { "epoch": 0.3703344016056457, "grad_norm": 6.439724484325886, "learning_rate": 9.919998832035815e-07, "loss": 0.3536, "step": 1430 }, { "epoch": 0.3703344016056457, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.9056603773584906, "eval_PRM F1 AUC": 0.8467784180199058, "eval_PRM F1 AUC (fixed)": 0.8035620743844945, "eval_PRM F1 Neg": 0.7169811320754716, "eval_PRM NPV": 0.6333333333333333, "eval_PRM Precision": 0.9473684210526315, "eval_PRM Recall": 0.8674698795180723, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.23704619705677032, "eval_runtime": 14.4407, "eval_samples_per_second": 3.047, "eval_steps_per_second": 0.138, "step": 1430 }, { "epoch": 0.3705933767116636, "grad_norm": 3.443131454125624, "learning_rate": 9.9154223876212e-07, "loss": 0.3985, "step": 1431 }, { "epoch": 0.3708523518176815, "grad_norm": 5.567638713314668, "learning_rate": 9.910842945672602e-07, "loss": 0.3216, "step": 1432 }, { "epoch": 0.37111132692369947, "grad_norm": 2.9108306082811293, "learning_rate": 9.906260509935027e-07, "loss": 0.3545, "step": 1433 }, { "epoch": 0.3713703020297174, "grad_norm": 3.992555721910898, "learning_rate": 9.901675084155924e-07, "loss": 0.3514, "step": 1434 }, { "epoch": 0.37162927713573535, "grad_norm": 3.288250771978138, "learning_rate": 9.897086672085205e-07, "loss": 0.3328, "step": 1435 }, { "epoch": 0.37162927713573535, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.8987341772151899, "eval_PRM F1 AUC": 0.8407543216343635, "eval_PRM F1 AUC (fixed)": 0.8004190675746463, "eval_PRM F1 Neg": 0.7037037037037037, "eval_PRM NPV": 0.6129032258064516, "eval_PRM Precision": 0.9466666666666667, "eval_PRM Recall": 0.8554216867469879, "eval_PRM Specificty": 0.8260869565217391, "eval_loss": 0.23846612870693207, "eval_runtime": 14.7197, "eval_samples_per_second": 2.989, "eval_steps_per_second": 0.136, "step": 1435 }, { "epoch": 0.37188825224175326, "grad_norm": 3.3613173523286695, "learning_rate": 9.892495277475203e-07, "loss": 0.2902, "step": 1436 }, { "epoch": 0.37214722734777117, "grad_norm": 4.1674637022810845, "learning_rate": 9.887900904080704e-07, "loss": 0.2901, "step": 1437 }, { "epoch": 0.37240620245378914, "grad_norm": 4.785496501344485, "learning_rate": 9.883303555658928e-07, "loss": 0.3439, "step": 1438 }, { "epoch": 0.37266517755980705, "grad_norm": 4.6478897795229335, "learning_rate": 9.878703235969523e-07, "loss": 0.3821, "step": 1439 }, { "epoch": 0.372924152665825, "grad_norm": 4.519922141366181, "learning_rate": 9.874099948774568e-07, "loss": 0.4066, "step": 1440 }, { "epoch": 0.372924152665825, "eval_PRM Accuracy": 0.8584905660377359, "eval_PRM F1": 0.906832298136646, "eval_PRM F1 AUC": 0.8310633839706653, "eval_PRM F1 AUC (fixed)": 0.8064431639601886, "eval_PRM F1 Neg": 0.7058823529411765, "eval_PRM NPV": 0.6428571428571429, "eval_PRM Precision": 0.9358974358974359, "eval_PRM Recall": 0.8795180722891566, "eval_PRM Specificty": 0.782608695652174, "eval_loss": 0.23579718172550201, "eval_runtime": 14.5869, "eval_samples_per_second": 3.016, "eval_steps_per_second": 0.137, "step": 1440 }, { "epoch": 0.3731831277718429, "grad_norm": 8.082390323683258, "learning_rate": 9.869493697838573e-07, "loss": 0.3807, "step": 1441 }, { "epoch": 0.3734421028778609, "grad_norm": 3.9871931447775344, "learning_rate": 9.86488448692847e-07, "loss": 0.4278, "step": 1442 }, { "epoch": 0.3737010779838788, "grad_norm": 5.366527806558145, "learning_rate": 9.860272319813607e-07, "loss": 0.3911, "step": 1443 }, { "epoch": 0.3739600530898967, "grad_norm": 4.962218269420269, "learning_rate": 9.855657200265755e-07, "loss": 0.3959, "step": 1444 }, { "epoch": 0.3742190281959147, "grad_norm": 3.682545253283956, "learning_rate": 9.851039132059103e-07, "loss": 0.4014, "step": 1445 }, { "epoch": 0.3742190281959147, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.9, "eval_PRM F1 AUC": 0.8250392875851232, "eval_PRM F1 AUC (fixed)": 0.7991094814038764, "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": 0.24163256585597992, "eval_runtime": 14.9003, "eval_samples_per_second": 2.953, "eval_steps_per_second": 0.134, "step": 1445 }, { "epoch": 0.3744780033019326, "grad_norm": 4.012447848004918, "learning_rate": 9.846418118970238e-07, "loss": 0.3358, "step": 1446 }, { "epoch": 0.37473697840795056, "grad_norm": 3.6285946249500456, "learning_rate": 9.841794164778167e-07, "loss": 0.3036, "step": 1447 }, { "epoch": 0.37499595351396847, "grad_norm": 7.966353046705746, "learning_rate": 9.8371672732643e-07, "loss": 0.2241, "step": 1448 }, { "epoch": 0.3752549286199864, "grad_norm": 3.212966323254794, "learning_rate": 9.832537448212445e-07, "loss": 0.3485, "step": 1449 }, { "epoch": 0.37551390372600435, "grad_norm": 5.815003987423003, "learning_rate": 9.827904693408814e-07, "loss": 0.34, "step": 1450 }, { "epoch": 0.37551390372600435, "eval_PRM Accuracy": 0.8773584905660378, "eval_PRM F1": 0.9202453987730062, "eval_PRM F1 AUC": 0.8431115767417497, "eval_PRM F1 AUC (fixed)": 0.8004190675746464, "eval_PRM F1 Neg": 0.7346938775510204, "eval_PRM NPV": 0.6923076923076923, "eval_PRM Precision": 0.9375, "eval_PRM Recall": 0.9036144578313253, "eval_PRM Specificty": 0.782608695652174, "eval_loss": 0.2403801530599594, "eval_runtime": 14.4933, "eval_samples_per_second": 3.036, "eval_steps_per_second": 0.138, "step": 1450 }, { "epoch": 0.37577287883202226, "grad_norm": 4.760451223227752, "learning_rate": 9.823269012642012e-07, "loss": 0.3352, "step": 1451 }, { "epoch": 0.37603185393804023, "grad_norm": 4.0186328881753335, "learning_rate": 9.818630409703036e-07, "loss": 0.2568, "step": 1452 }, { "epoch": 0.37629082904405814, "grad_norm": 4.164063275627754, "learning_rate": 9.813988888385275e-07, "loss": 0.3776, "step": 1453 }, { "epoch": 0.37654980415007605, "grad_norm": 8.55257640064162, "learning_rate": 9.809344452484508e-07, "loss": 0.3895, "step": 1454 }, { "epoch": 0.376808779256094, "grad_norm": 4.579909515061625, "learning_rate": 9.804697105798885e-07, "loss": 0.2611, "step": 1455 }, { "epoch": 0.376808779256094, "eval_PRM Accuracy": 0.8679245283018868, "eval_PRM F1": 0.9135802469135802, "eval_PRM F1 AUC": 0.8370874803562075, "eval_PRM F1 AUC (fixed)": 0.8067050811943426, "eval_PRM F1 Neg": 0.72, "eval_PRM NPV": 0.6666666666666666, "eval_PRM Precision": 0.9367088607594937, "eval_PRM Recall": 0.891566265060241, "eval_PRM Specificty": 0.782608695652174, "eval_loss": 0.23761196434497833, "eval_runtime": 14.4681, "eval_samples_per_second": 3.041, "eval_steps_per_second": 0.138, "step": 1455 }, { "epoch": 0.37706775436211193, "grad_norm": 3.5982328749856105, "learning_rate": 9.80004685212895e-07, "loss": 0.3832, "step": 1456 }, { "epoch": 0.3773267294681299, "grad_norm": 5.440234201873794, "learning_rate": 9.79539369527762e-07, "loss": 0.3383, "step": 1457 }, { "epoch": 0.3775857045741478, "grad_norm": 5.14995747106743, "learning_rate": 9.790737639050182e-07, "loss": 0.4605, "step": 1458 }, { "epoch": 0.3778446796801657, "grad_norm": 3.3322124466293124, "learning_rate": 9.786078687254297e-07, "loss": 0.3834, "step": 1459 }, { "epoch": 0.3781036547861837, "grad_norm": 6.726995927479046, "learning_rate": 9.781416843699999e-07, "loss": 0.305, "step": 1460 }, { "epoch": 0.3781036547861837, "eval_PRM Accuracy": 0.8679245283018868, "eval_PRM F1": 0.9135802469135802, "eval_PRM F1 AUC": 0.8370874803562075, "eval_PRM F1 AUC (fixed)": 0.8114195914091147, "eval_PRM F1 Neg": 0.72, "eval_PRM NPV": 0.6666666666666666, "eval_PRM Precision": 0.9367088607594937, "eval_PRM Recall": 0.891566265060241, "eval_PRM Specificty": 0.782608695652174, "eval_loss": 0.23749984800815582, "eval_runtime": 14.8026, "eval_samples_per_second": 2.972, "eval_steps_per_second": 0.135, "step": 1460 }, { "epoch": 0.3783626298922016, "grad_norm": 3.9906455982204188, "learning_rate": 9.776752112199674e-07, "loss": 0.3707, "step": 1461 }, { "epoch": 0.37862160499821956, "grad_norm": 5.0238473308596925, "learning_rate": 9.772084496568088e-07, "loss": 0.3443, "step": 1462 }, { "epoch": 0.3788805801042375, "grad_norm": 6.058458226117506, "learning_rate": 9.767414000622346e-07, "loss": 0.3752, "step": 1463 }, { "epoch": 0.37913955521025544, "grad_norm": 6.055971624389849, "learning_rate": 9.762740628181923e-07, "loss": 0.3526, "step": 1464 }, { "epoch": 0.37939853031627335, "grad_norm": 4.379766953792154, "learning_rate": 9.758064383068643e-07, "loss": 0.3024, "step": 1465 }, { "epoch": 0.37939853031627335, "eval_PRM Accuracy": 0.8867924528301887, "eval_PRM F1": 0.926829268292683, "eval_PRM F1 AUC": 0.8491356731272918, "eval_PRM F1 AUC (fixed)": 0.8171817705605029, "eval_PRM F1 Neg": 0.75, "eval_PRM NPV": 0.72, "eval_PRM Precision": 0.9382716049382716, "eval_PRM Recall": 0.9156626506024096, "eval_PRM Specificty": 0.782608695652174, "eval_loss": 0.2357960194349289, "eval_runtime": 14.6708, "eval_samples_per_second": 2.999, "eval_steps_per_second": 0.136, "step": 1465 }, { "epoch": 0.37965750542229126, "grad_norm": 3.721510725794544, "learning_rate": 9.753385269106671e-07, "loss": 0.3405, "step": 1466 }, { "epoch": 0.37991648052830923, "grad_norm": 4.6197481619434795, "learning_rate": 9.74870329012253e-07, "loss": 0.3346, "step": 1467 }, { "epoch": 0.38017545563432714, "grad_norm": 3.318853840301337, "learning_rate": 9.744018449945078e-07, "loss": 0.342, "step": 1468 }, { "epoch": 0.3804344307403451, "grad_norm": 6.607534071714149, "learning_rate": 9.739330752405516e-07, "loss": 0.3281, "step": 1469 }, { "epoch": 0.380693405846363, "grad_norm": 3.3108665759335176, "learning_rate": 9.73464020133738e-07, "loss": 0.3388, "step": 1470 }, { "epoch": 0.380693405846363, "eval_PRM Accuracy": 0.8773584905660378, "eval_PRM F1": 0.9212121212121213, "eval_PRM F1 AUC": 0.8273965426925091, "eval_PRM F1 AUC (fixed)": 0.8192771084337349, "eval_PRM F1 Neg": 0.723404255319149, "eval_PRM NPV": 0.7083333333333334, "eval_PRM Precision": 0.926829268292683, "eval_PRM Recall": 0.9156626506024096, "eval_PRM Specificty": 0.7391304347826086, "eval_loss": 0.23642498254776, "eval_runtime": 14.734, "eval_samples_per_second": 2.986, "eval_steps_per_second": 0.136, "step": 1470 }, { "epoch": 0.38095238095238093, "grad_norm": 4.88496391866845, "learning_rate": 9.729946800576542e-07, "loss": 0.4142, "step": 1471 }, { "epoch": 0.3812113560583989, "grad_norm": 5.27916049939447, "learning_rate": 9.725250553961202e-07, "loss": 0.3391, "step": 1472 }, { "epoch": 0.3814703311644168, "grad_norm": 8.900314491547281, "learning_rate": 9.72055146533189e-07, "loss": 0.377, "step": 1473 }, { "epoch": 0.3817293062704348, "grad_norm": 3.0011588762933292, "learning_rate": 9.71584953853146e-07, "loss": 0.2884, "step": 1474 }, { "epoch": 0.3819882813764527, "grad_norm": 4.099551127706329, "learning_rate": 9.71114477740508e-07, "loss": 0.3726, "step": 1475 }, { "epoch": 0.3819882813764527, "eval_PRM Accuracy": 0.8773584905660378, "eval_PRM F1": 0.9212121212121213, "eval_PRM F1 AUC": 0.8273965426925091, "eval_PRM F1 AUC (fixed)": 0.8258250392875852, "eval_PRM F1 Neg": 0.723404255319149, "eval_PRM NPV": 0.7083333333333334, "eval_PRM Precision": 0.926829268292683, "eval_PRM Recall": 0.9156626506024096, "eval_PRM Specificty": 0.7391304347826086, "eval_loss": 0.23649762570858002, "eval_runtime": 14.8796, "eval_samples_per_second": 2.957, "eval_steps_per_second": 0.134, "step": 1475 }, { "epoch": 0.3822472564824706, "grad_norm": 3.6773421349404165, "learning_rate": 9.706437185800247e-07, "loss": 0.3779, "step": 1476 }, { "epoch": 0.38250623158848857, "grad_norm": 4.032342903491971, "learning_rate": 9.701726767566764e-07, "loss": 0.4382, "step": 1477 }, { "epoch": 0.3827652066945065, "grad_norm": 3.5491298285114325, "learning_rate": 9.697013526556751e-07, "loss": 0.3316, "step": 1478 }, { "epoch": 0.38302418180052444, "grad_norm": 4.341491391745633, "learning_rate": 9.692297466624633e-07, "loss": 0.3436, "step": 1479 }, { "epoch": 0.38328315690654235, "grad_norm": 2.8643056611618674, "learning_rate": 9.687578591627138e-07, "loss": 0.2758, "step": 1480 }, { "epoch": 0.38328315690654235, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.9, "eval_PRM F1 AUC": 0.8250392875851232, "eval_PRM F1 AUC (fixed)": 0.8187532739654269, "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": 0.24352265894412994, "eval_runtime": 14.8616, "eval_samples_per_second": 2.961, "eval_steps_per_second": 0.135, "step": 1480 }, { "epoch": 0.38354213201256027, "grad_norm": 3.7035412141544444, "learning_rate": 9.682856905423307e-07, "loss": 0.2516, "step": 1481 }, { "epoch": 0.38380110711857823, "grad_norm": 6.508781566083587, "learning_rate": 9.678132411874468e-07, "loss": 0.4246, "step": 1482 }, { "epoch": 0.38406008222459614, "grad_norm": 3.258716176187973, "learning_rate": 9.67340511484425e-07, "loss": 0.3599, "step": 1483 }, { "epoch": 0.3843190573306141, "grad_norm": 3.4622217784743956, "learning_rate": 9.668675018198576e-07, "loss": 0.3031, "step": 1484 }, { "epoch": 0.384578032436632, "grad_norm": 5.0275928854286045, "learning_rate": 9.66394212580565e-07, "loss": 0.3924, "step": 1485 }, { "epoch": 0.384578032436632, "eval_PRM Accuracy": 0.8490566037735849, "eval_PRM F1": 0.9, "eval_PRM F1 AUC": 0.8250392875851232, "eval_PRM F1 AUC (fixed)": 0.8171817705605029, "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": 0.24826860427856445, "eval_runtime": 14.271, "eval_samples_per_second": 3.083, "eval_steps_per_second": 0.14, "step": 1485 }, { "epoch": 0.38483700754265, "grad_norm": 6.109340461968093, "learning_rate": 9.659206441535976e-07, "loss": 0.3058, "step": 1486 }, { "epoch": 0.3850959826486679, "grad_norm": 5.412763780216841, "learning_rate": 9.654467969262333e-07, "loss": 0.3662, "step": 1487 }, { "epoch": 0.3853549577546858, "grad_norm": 4.1727011697393195, "learning_rate": 9.649726712859777e-07, "loss": 0.3745, "step": 1488 }, { "epoch": 0.3856139328607038, "grad_norm": 7.060653823640019, "learning_rate": 9.64498267620565e-07, "loss": 0.326, "step": 1489 }, { "epoch": 0.3858729079667217, "grad_norm": 3.036608182108932, "learning_rate": 9.640235863179559e-07, "loss": 0.2627, "step": 1490 }, { "epoch": 0.3858729079667217, "eval_PRM Accuracy": 0.8773584905660378, "eval_PRM F1": 0.9202453987730062, "eval_PRM F1 AUC": 0.8431115767417497, "eval_PRM F1 AUC (fixed)": 0.8268727082242011, "eval_PRM F1 Neg": 0.7346938775510204, "eval_PRM NPV": 0.6923076923076923, "eval_PRM Precision": 0.9375, "eval_PRM Recall": 0.9036144578313253, "eval_PRM Specificty": 0.782608695652174, "eval_loss": 0.23949016630649567, "eval_runtime": 14.7217, "eval_samples_per_second": 2.989, "eval_steps_per_second": 0.136, "step": 1490 }, { "epoch": 0.38613188307273966, "grad_norm": 3.9043233046211663, "learning_rate": 9.635486277663383e-07, "loss": 0.3002, "step": 1491 }, { "epoch": 0.38639085817875757, "grad_norm": 3.180806958564625, "learning_rate": 9.630733923541277e-07, "loss": 0.2761, "step": 1492 }, { "epoch": 0.3866498332847755, "grad_norm": 3.4766085563957825, "learning_rate": 9.625978804699646e-07, "loss": 0.2817, "step": 1493 }, { "epoch": 0.38690880839079345, "grad_norm": 12.109662314263637, "learning_rate": 9.621220925027168e-07, "loss": 0.4113, "step": 1494 }, { "epoch": 0.38716778349681136, "grad_norm": 4.082401544792815, "learning_rate": 9.616460288414776e-07, "loss": 0.3458, "step": 1495 }, { "epoch": 0.38716778349681136, "eval_PRM Accuracy": 0.8867924528301887, "eval_PRM F1": 0.9285714285714286, "eval_PRM F1 AUC": 0.817705605028811, "eval_PRM F1 AUC (fixed)": 0.8300157150340493, "eval_PRM F1 Neg": 0.7272727272727273, "eval_PRM NPV": 0.7619047619047619, "eval_PRM Precision": 0.9176470588235294, "eval_PRM Recall": 0.9397590361445783, "eval_PRM Specificty": 0.6956521739130435, "eval_loss": 0.23893995583057404, "eval_runtime": 14.7386, "eval_samples_per_second": 2.985, "eval_steps_per_second": 0.136, "step": 1495 }, { "epoch": 0.3874267586028293, "grad_norm": 3.5917626862055987, "learning_rate": 9.611696898755649e-07, "loss": 0.2712, "step": 1496 }, { "epoch": 0.38768573370884724, "grad_norm": 9.289904299091432, "learning_rate": 9.606930759945227e-07, "loss": 0.4151, "step": 1497 }, { "epoch": 0.38794470881486515, "grad_norm": 6.707271661263045, "learning_rate": 9.602161875881196e-07, "loss": 0.3717, "step": 1498 }, { "epoch": 0.3882036839208831, "grad_norm": 2.4515236018126743, "learning_rate": 9.59739025046349e-07, "loss": 0.2463, "step": 1499 }, { "epoch": 0.388462659026901, "grad_norm": 3.005546252622098, "learning_rate": 9.592615887594276e-07, "loss": 0.3076, "step": 1500 }, { "epoch": 0.388462659026901, "eval_PRM Accuracy": 0.9056603773584906, "eval_PRM F1": 0.9397590361445783, "eval_PRM F1 AUC": 0.8611838658983763, "eval_PRM F1 AUC (fixed)": 0.8328968046097434, "eval_PRM F1 Neg": 0.782608695652174, "eval_PRM NPV": 0.782608695652174, "eval_PRM Precision": 0.9397590361445783, "eval_PRM Recall": 0.9397590361445783, "eval_PRM Specificty": 0.782608695652174, "eval_loss": 0.229969784617424, "eval_runtime": 14.7198, "eval_samples_per_second": 2.989, "eval_steps_per_second": 0.136, "step": 1500 } ], "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": 126724469153792.0, "train_batch_size": 2, "trial_name": null, "trial_params": null }