{ "best_metric": 0.7408195079003096, "best_model_checkpoint": "bart-large/checkpoint-64225", "epoch": 30.0, "eval_steps": 500, "global_step": 77070, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.19, "learning_rate": 4.967886337096146e-05, "loss": 2.0273, "step": 500 }, { "epoch": 0.39, "learning_rate": 4.9354482937589206e-05, "loss": 1.4983, "step": 1000 }, { "epoch": 0.58, "learning_rate": 4.903140002595043e-05, "loss": 1.4727, "step": 1500 }, { "epoch": 0.78, "learning_rate": 4.870701959257818e-05, "loss": 1.3945, "step": 2000 }, { "epoch": 0.97, "learning_rate": 4.8383287920072665e-05, "loss": 1.1685, "step": 2500 }, { "epoch": 1.0, "eval_accuracy": 0.6847405112316034, "eval_f1_macro": 0.4296023125768376, "eval_f1_micro": 0.6847405112316034, "eval_f1_weighted": 0.6471198955620663, "eval_loss": 1.258675456047058, "eval_macro_fpr": 0.03310536835500512, "eval_macro_sensitivity": 0.47197476920519743, "eval_macro_specificity": 0.9759825618150952, "eval_precision": 0.6359796796355663, "eval_precision_macro": 0.4175869918662546, "eval_recall": 0.6847405112316034, "eval_recall_macro": 0.47197476920519743, "eval_runtime": 166.2079, "eval_samples_per_second": 7.767, "eval_steps_per_second": 3.887, "eval_weighted_fpr": 0.03183916138621607, "eval_weighted_sensitivity": 0.6847405112316034, "eval_weighted_specificity": 0.9549979159948256, "step": 2569 }, { "epoch": 1.17, "learning_rate": 4.80589074867004e-05, "loss": 1.2814, "step": 3000 }, { "epoch": 1.36, "learning_rate": 4.7734527053328146e-05, "loss": 1.2579, "step": 3500 }, { "epoch": 1.56, "learning_rate": 4.741079538082263e-05, "loss": 1.246, "step": 4000 }, { "epoch": 1.75, "learning_rate": 4.708641494745037e-05, "loss": 1.0685, "step": 4500 }, { "epoch": 1.95, "learning_rate": 4.676203451407811e-05, "loss": 1.1965, "step": 5000 }, { "epoch": 2.0, "eval_accuracy": 0.6638264910921766, "eval_f1_macro": 0.3954707529203812, "eval_f1_micro": 0.6638264910921766, "eval_f1_weighted": 0.6467821848395333, "eval_loss": 1.1623183488845825, "eval_macro_fpr": 0.03422706134984445, "eval_macro_sensitivity": 0.4261017849773373, "eval_macro_specificity": 0.9752839802406176, "eval_precision": 0.6943370263824146, "eval_precision_macro": 0.45636498568634104, "eval_recall": 0.6638264910921766, "eval_recall_macro": 0.4261017849773373, "eval_runtime": 320.2013, "eval_samples_per_second": 4.032, "eval_steps_per_second": 2.017, "eval_weighted_fpr": 0.03490990990990991, "eval_weighted_sensitivity": 0.6638264910921766, "eval_weighted_specificity": 0.9654332125170877, "step": 5138 }, { "epoch": 2.14, "learning_rate": 4.643765408070585e-05, "loss": 1.0697, "step": 5500 }, { "epoch": 2.34, "learning_rate": 4.611327364733359e-05, "loss": 1.1042, "step": 6000 }, { "epoch": 2.53, "learning_rate": 4.578954197482808e-05, "loss": 1.124, "step": 6500 }, { "epoch": 2.72, "learning_rate": 4.546516154145582e-05, "loss": 1.1449, "step": 7000 }, { "epoch": 2.92, "learning_rate": 4.514078110808356e-05, "loss": 1.189, "step": 7500 }, { "epoch": 3.0, "eval_accuracy": 0.7234701781564679, "eval_f1_macro": 0.5196282925085354, "eval_f1_micro": 0.7234701781564679, "eval_f1_weighted": 0.7031290239201134, "eval_loss": 1.357356309890747, "eval_macro_fpr": 0.027073020005110214, "eval_macro_sensitivity": 0.5527845991176076, "eval_macro_specificity": 0.9790845865453512, "eval_precision": 0.7219921821470505, "eval_precision_macro": 0.5413306235947684, "eval_recall": 0.7234701781564679, "eval_recall_macro": 0.5527845991176076, "eval_runtime": 290.3268, "eval_samples_per_second": 4.447, "eval_steps_per_second": 2.225, "eval_weighted_fpr": 0.026576341844710787, "eval_weighted_sensitivity": 0.7234701781564679, "eval_weighted_specificity": 0.9627986200238022, "step": 7707 }, { "epoch": 3.11, "learning_rate": 4.481704943557805e-05, "loss": 0.9829, "step": 8000 }, { "epoch": 3.31, "learning_rate": 4.4492669002205786e-05, "loss": 0.9172, "step": 8500 }, { "epoch": 3.5, "learning_rate": 4.416828856883353e-05, "loss": 0.9911, "step": 9000 }, { "epoch": 3.7, "learning_rate": 4.384390813546127e-05, "loss": 0.9934, "step": 9500 }, { "epoch": 3.89, "learning_rate": 4.351952770208901e-05, "loss": 1.0127, "step": 10000 }, { "epoch": 4.0, "eval_accuracy": 0.7668474051123161, "eval_f1_macro": 0.6233324855492023, "eval_f1_micro": 0.7668474051123161, "eval_f1_weighted": 0.7569127961671379, "eval_loss": 1.4684507846832275, "eval_macro_fpr": 0.022371017412255348, "eval_macro_sensitivity": 0.6201844239757082, "eval_macro_specificity": 0.9821444214871504, "eval_precision": 0.7584494799967734, "eval_precision_macro": 0.6670503035550461, "eval_recall": 0.7668474051123161, "eval_recall_macro": 0.6201844239757082, "eval_runtime": 296.0277, "eval_samples_per_second": 4.361, "eval_steps_per_second": 2.182, "eval_weighted_fpr": 0.021255561047948592, "eval_weighted_sensitivity": 0.7668474051123161, "eval_weighted_specificity": 0.9653189171949409, "step": 10276 }, { "epoch": 4.09, "learning_rate": 4.3196444790450244e-05, "loss": 0.9926, "step": 10500 }, { "epoch": 4.28, "learning_rate": 4.287206435707798e-05, "loss": 1.1046, "step": 11000 }, { "epoch": 4.48, "learning_rate": 4.2547683923705725e-05, "loss": 0.8572, "step": 11500 }, { "epoch": 4.67, "learning_rate": 4.222330349033346e-05, "loss": 0.8623, "step": 12000 }, { "epoch": 4.87, "learning_rate": 4.1898923056961206e-05, "loss": 1.0205, "step": 12500 }, { "epoch": 5.0, "eval_accuracy": 0.7668474051123161, "eval_f1_macro": 0.6732447131282889, "eval_f1_micro": 0.7668474051123161, "eval_f1_weighted": 0.7643097206437405, "eval_loss": 1.4232454299926758, "eval_macro_fpr": 0.021536303183803004, "eval_macro_sensitivity": 0.6871530786461169, "eval_macro_specificity": 0.9827017397959152, "eval_precision": 0.7710604143740172, "eval_precision_macro": 0.6764905863857386, "eval_recall": 0.7668474051123161, "eval_recall_macro": 0.6871530786461169, "eval_runtime": 288.8657, "eval_samples_per_second": 4.469, "eval_steps_per_second": 2.236, "eval_weighted_fpr": 0.021255561047948592, "eval_weighted_sensitivity": 0.7668474051123161, "eval_weighted_specificity": 0.9736786918264111, "step": 12845 }, { "epoch": 5.06, "learning_rate": 4.1574542623588944e-05, "loss": 0.888, "step": 13000 }, { "epoch": 5.25, "learning_rate": 4.125016219021669e-05, "loss": 0.7739, "step": 13500 }, { "epoch": 5.45, "learning_rate": 4.0925781756844425e-05, "loss": 0.9572, "step": 14000 }, { "epoch": 5.64, "learning_rate": 4.060140132347217e-05, "loss": 0.8296, "step": 14500 }, { "epoch": 5.84, "learning_rate": 4.0277020890099906e-05, "loss": 0.7927, "step": 15000 }, { "epoch": 6.0, "eval_accuracy": 0.7428350116189001, "eval_f1_macro": 0.6107557579512543, "eval_f1_micro": 0.7428350116189001, "eval_f1_weighted": 0.729163687728552, "eval_loss": 1.5677722692489624, "eval_macro_fpr": 0.024790721920250693, "eval_macro_sensitivity": 0.6332883060366477, "eval_macro_specificity": 0.9807865890062467, "eval_precision": 0.7451349692354797, "eval_precision_macro": 0.6488747447950202, "eval_recall": 0.7428350116189001, "eval_recall_macro": 0.6332883060366477, "eval_runtime": 291.455, "eval_samples_per_second": 4.429, "eval_steps_per_second": 2.216, "eval_weighted_fpr": 0.024131414449774676, "eval_weighted_sensitivity": 0.7428350116189001, "eval_weighted_specificity": 0.9689638234747989, "step": 15414 }, { "epoch": 6.03, "learning_rate": 3.995264045672765e-05, "loss": 0.8196, "step": 15500 }, { "epoch": 6.23, "learning_rate": 3.9628260023355394e-05, "loss": 0.6989, "step": 16000 }, { "epoch": 6.42, "learning_rate": 3.930387958998313e-05, "loss": 0.7149, "step": 16500 }, { "epoch": 6.62, "learning_rate": 3.8979499156610875e-05, "loss": 0.7459, "step": 17000 }, { "epoch": 6.81, "learning_rate": 3.865511872323861e-05, "loss": 0.7701, "step": 17500 }, { "epoch": 7.0, "eval_accuracy": 0.7467079783113865, "eval_f1_macro": 0.6583574261846535, "eval_f1_micro": 0.7467079783113864, "eval_f1_weighted": 0.7399270546375634, "eval_loss": 1.7336597442626953, "eval_macro_fpr": 0.023989153033760927, "eval_macro_sensitivity": 0.6536163621805403, "eval_macro_specificity": 0.9809809544706846, "eval_precision": 0.759972067375925, "eval_precision_macro": 0.6862877418910309, "eval_recall": 0.7467079783113865, "eval_recall_macro": 0.6536163621805403, "eval_runtime": 289.659, "eval_samples_per_second": 4.457, "eval_steps_per_second": 2.23, "eval_weighted_fpr": 0.023656225132026334, "eval_weighted_sensitivity": 0.7467079783113865, "eval_weighted_specificity": 0.9680063387488825, "step": 17983 }, { "epoch": 7.01, "learning_rate": 3.83313870507331e-05, "loss": 0.826, "step": 18000 }, { "epoch": 7.2, "learning_rate": 3.800765537822758e-05, "loss": 0.6803, "step": 18500 }, { "epoch": 7.4, "learning_rate": 3.7683274944855327e-05, "loss": 0.6779, "step": 19000 }, { "epoch": 7.59, "learning_rate": 3.7358894511483064e-05, "loss": 0.616, "step": 19500 }, { "epoch": 7.79, "learning_rate": 3.703451407811081e-05, "loss": 0.608, "step": 20000 }, { "epoch": 7.98, "learning_rate": 3.671013364473855e-05, "loss": 0.584, "step": 20500 }, { "epoch": 8.0, "eval_accuracy": 0.7691711851278079, "eval_f1_macro": 0.6979809418287721, "eval_f1_micro": 0.7691711851278079, "eval_f1_weighted": 0.7682883483181749, "eval_loss": 1.6188185214996338, "eval_macro_fpr": 0.021432185448306138, "eval_macro_sensitivity": 0.7065204224528159, "eval_macro_specificity": 0.9826542042505794, "eval_precision": 0.7766202006674988, "eval_precision_macro": 0.6979437669181305, "eval_recall": 0.7691711851278079, "eval_recall_macro": 0.7065204224528159, "eval_runtime": 303.6709, "eval_samples_per_second": 4.251, "eval_steps_per_second": 2.127, "eval_weighted_fpr": 0.020985915492957745, "eval_weighted_sensitivity": 0.7691711851278079, "eval_weighted_specificity": 0.9706418786308822, "step": 20552 }, { "epoch": 8.17, "learning_rate": 3.638575321136629e-05, "loss": 0.5266, "step": 21000 }, { "epoch": 8.37, "learning_rate": 3.606137277799403e-05, "loss": 0.6717, "step": 21500 }, { "epoch": 8.56, "learning_rate": 3.573699234462177e-05, "loss": 0.5474, "step": 22000 }, { "epoch": 8.76, "learning_rate": 3.5412611911249514e-05, "loss": 0.638, "step": 22500 }, { "epoch": 8.95, "learning_rate": 3.5088880238743996e-05, "loss": 0.5659, "step": 23000 }, { "epoch": 9.0, "eval_accuracy": 0.7598760650658405, "eval_f1_macro": 0.6728187845754294, "eval_f1_micro": 0.7598760650658405, "eval_f1_weighted": 0.7542471407266271, "eval_loss": 1.6982707977294922, "eval_macro_fpr": 0.02271263520390573, "eval_macro_sensitivity": 0.680403785625184, "eval_macro_specificity": 0.9819588600237938, "eval_precision": 0.766529422163196, "eval_precision_macro": 0.7000337979719602, "eval_recall": 0.7598760650658405, "eval_recall_macro": 0.680403785625184, "eval_runtime": 286.113, "eval_samples_per_second": 4.512, "eval_steps_per_second": 2.258, "eval_weighted_fpr": 0.02207348333808032, "eval_weighted_sensitivity": 0.7598760650658405, "eval_weighted_specificity": 0.9695068352910698, "step": 23121 }, { "epoch": 9.15, "learning_rate": 3.476449980537174e-05, "loss": 0.5062, "step": 23500 }, { "epoch": 9.34, "learning_rate": 3.4440119371999484e-05, "loss": 0.4485, "step": 24000 }, { "epoch": 9.54, "learning_rate": 3.411573893862722e-05, "loss": 0.6239, "step": 24500 }, { "epoch": 9.73, "learning_rate": 3.379200726612171e-05, "loss": 0.6764, "step": 25000 }, { "epoch": 9.93, "learning_rate": 3.3467626832749447e-05, "loss": 0.7021, "step": 25500 }, { "epoch": 10.0, "eval_accuracy": 0.7699457784663052, "eval_f1_macro": 0.6953722525071192, "eval_f1_micro": 0.7699457784663051, "eval_f1_weighted": 0.7634449085212252, "eval_loss": 1.6445409059524536, "eval_macro_fpr": 0.022344139928734137, "eval_macro_sensitivity": 0.6856644071349105, "eval_macro_specificity": 0.9820521636968861, "eval_precision": 0.7656296240571903, "eval_precision_macro": 0.7143810783711486, "eval_recall": 0.7699457784663052, "eval_recall_macro": 0.6856644071349105, "eval_runtime": 288.6164, "eval_samples_per_second": 4.473, "eval_steps_per_second": 2.238, "eval_weighted_fpr": 0.0208963624850489, "eval_weighted_sensitivity": 0.7699457784663052, "eval_weighted_specificity": 0.9608366769869864, "step": 25690 }, { "epoch": 10.12, "learning_rate": 3.314324639937719e-05, "loss": 0.6243, "step": 26000 }, { "epoch": 10.32, "learning_rate": 3.281886596600493e-05, "loss": 0.6952, "step": 26500 }, { "epoch": 10.51, "learning_rate": 3.249448553263267e-05, "loss": 0.7705, "step": 27000 }, { "epoch": 10.7, "learning_rate": 3.217010509926041e-05, "loss": 0.6582, "step": 27500 }, { "epoch": 10.9, "learning_rate": 3.184572466588815e-05, "loss": 0.6216, "step": 28000 }, { "epoch": 11.0, "eval_accuracy": 0.7676219984508134, "eval_f1_macro": 0.6786279912338002, "eval_f1_micro": 0.7676219984508135, "eval_f1_weighted": 0.7623696014823479, "eval_loss": 1.656176209449768, "eval_macro_fpr": 0.022330481497328643, "eval_macro_sensitivity": 0.6775994679116955, "eval_macro_specificity": 0.9821081561015023, "eval_precision": 0.7633744243747118, "eval_precision_macro": 0.6856198086678005, "eval_recall": 0.7676219984508134, "eval_recall_macro": 0.6775994679116955, "eval_runtime": 286.2474, "eval_samples_per_second": 4.51, "eval_steps_per_second": 2.257, "eval_weighted_fpr": 0.021165514321998025, "eval_weighted_sensitivity": 0.7676219984508134, "eval_weighted_specificity": 0.9640003430717192, "step": 28259 }, { "epoch": 11.09, "learning_rate": 3.15213442325159e-05, "loss": 0.716, "step": 28500 }, { "epoch": 11.29, "learning_rate": 3.119826132087712e-05, "loss": 0.5161, "step": 29000 }, { "epoch": 11.48, "learning_rate": 3.087388088750487e-05, "loss": 0.5121, "step": 29500 }, { "epoch": 11.68, "learning_rate": 3.0549500454132604e-05, "loss": 0.4694, "step": 30000 }, { "epoch": 11.87, "learning_rate": 3.0225120020760352e-05, "loss": 0.6408, "step": 30500 }, { "epoch": 12.0, "eval_accuracy": 0.7668474051123161, "eval_f1_macro": 0.6665576721498584, "eval_f1_micro": 0.7668474051123161, "eval_f1_weighted": 0.760771317042359, "eval_loss": 1.6681983470916748, "eval_macro_fpr": 0.022265012904393446, "eval_macro_sensitivity": 0.6719396221017415, "eval_macro_specificity": 0.9822275271243499, "eval_precision": 0.7628875217070834, "eval_precision_macro": 0.6705561723407458, "eval_recall": 0.7668474051123161, "eval_recall_macro": 0.6719396221017415, "eval_runtime": 285.4858, "eval_samples_per_second": 4.522, "eval_steps_per_second": 2.263, "eval_weighted_fpr": 0.021255561047948592, "eval_weighted_sensitivity": 0.7668474051123161, "eval_weighted_specificity": 0.9665655017529341, "step": 30828 }, { "epoch": 12.07, "learning_rate": 2.990073958738809e-05, "loss": 0.518, "step": 31000 }, { "epoch": 12.26, "learning_rate": 2.9577007914882577e-05, "loss": 0.3736, "step": 31500 }, { "epoch": 12.46, "learning_rate": 2.9252627481510315e-05, "loss": 0.4537, "step": 32000 }, { "epoch": 12.65, "learning_rate": 2.892824704813806e-05, "loss": 0.5017, "step": 32500 }, { "epoch": 12.85, "learning_rate": 2.86038666147658e-05, "loss": 0.523, "step": 33000 }, { "epoch": 13.0, "eval_accuracy": 0.7652982184353214, "eval_f1_macro": 0.7065905031891068, "eval_f1_micro": 0.7652982184353214, "eval_f1_weighted": 0.7533906957445078, "eval_loss": 1.7727322578430176, "eval_macro_fpr": 0.02260077041026825, "eval_macro_sensitivity": 0.6934075054851363, "eval_macro_specificity": 0.9820770689956008, "eval_precision": 0.7674159154971928, "eval_precision_macro": 0.8238391848974315, "eval_recall": 0.7652982184353214, "eval_recall_macro": 0.6934075054851363, "eval_runtime": 282.4043, "eval_samples_per_second": 4.571, "eval_steps_per_second": 2.288, "eval_weighted_fpr": 0.02143615139724089, "eval_weighted_sensitivity": 0.7652982184353214, "eval_weighted_specificity": 0.9658578164986904, "step": 33397 }, { "epoch": 13.04, "learning_rate": 2.8280134942260284e-05, "loss": 0.4438, "step": 33500 }, { "epoch": 13.23, "learning_rate": 2.7955754508888028e-05, "loss": 0.3444, "step": 34000 }, { "epoch": 13.43, "learning_rate": 2.7631374075515765e-05, "loss": 0.3751, "step": 34500 }, { "epoch": 13.62, "learning_rate": 2.730699364214351e-05, "loss": 0.3835, "step": 35000 }, { "epoch": 13.82, "learning_rate": 2.6982613208771247e-05, "loss": 0.3688, "step": 35500 }, { "epoch": 14.0, "eval_accuracy": 0.7792408985282726, "eval_f1_macro": 0.6960443476037717, "eval_f1_micro": 0.7792408985282726, "eval_f1_weighted": 0.7731313238375468, "eval_loss": 1.840445876121521, "eval_macro_fpr": 0.020905039822376084, "eval_macro_sensitivity": 0.6920728491541597, "eval_macro_specificity": 0.9831130726980227, "eval_precision": 0.7787988846597094, "eval_precision_macro": 0.7228983794850382, "eval_recall": 0.7792408985282726, "eval_recall_macro": 0.6920728491541597, "eval_runtime": 289.1315, "eval_samples_per_second": 4.465, "eval_steps_per_second": 2.234, "eval_weighted_fpr": 0.019834365648270583, "eval_weighted_sensitivity": 0.7792408985282726, "eval_weighted_specificity": 0.9674551919420677, "step": 35966 }, { "epoch": 14.01, "learning_rate": 2.665823277539899e-05, "loss": 0.3816, "step": 36000 }, { "epoch": 14.21, "learning_rate": 2.633385234202673e-05, "loss": 0.3767, "step": 36500 }, { "epoch": 14.4, "learning_rate": 2.6009471908654472e-05, "loss": 0.3859, "step": 37000 }, { "epoch": 14.6, "learning_rate": 2.5685091475282212e-05, "loss": 0.3066, "step": 37500 }, { "epoch": 14.79, "learning_rate": 2.5360711041909953e-05, "loss": 0.402, "step": 38000 }, { "epoch": 14.99, "learning_rate": 2.5036330608537694e-05, "loss": 0.2394, "step": 38500 }, { "epoch": 15.0, "eval_accuracy": 0.7815646785437645, "eval_f1_macro": 0.7230355619154473, "eval_f1_micro": 0.7815646785437644, "eval_f1_weighted": 0.7765270107358122, "eval_loss": 1.7884501218795776, "eval_macro_fpr": 0.020978052492475165, "eval_macro_sensitivity": 0.711481022624799, "eval_macro_specificity": 0.9829555212885144, "eval_precision": 0.7809140931231899, "eval_precision_macro": 0.7440727141565046, "eval_recall": 0.7815646785437645, "eval_recall_macro": 0.711481022624799, "eval_runtime": 286.1698, "eval_samples_per_second": 4.511, "eval_steps_per_second": 2.257, "eval_weighted_fpr": 0.01957245974458634, "eval_weighted_sensitivity": 0.7815646785437645, "eval_weighted_specificity": 0.9627681407839513, "step": 38535 }, { "epoch": 15.18, "learning_rate": 2.4711950175165434e-05, "loss": 0.4596, "step": 39000 }, { "epoch": 15.38, "learning_rate": 2.4387569741793175e-05, "loss": 0.5262, "step": 39500 }, { "epoch": 15.57, "learning_rate": 2.4063189308420915e-05, "loss": 0.4665, "step": 40000 }, { "epoch": 15.76, "learning_rate": 2.3738808875048656e-05, "loss": 0.3048, "step": 40500 }, { "epoch": 15.96, "learning_rate": 2.341572596340989e-05, "loss": 0.2734, "step": 41000 }, { "epoch": 16.0, "eval_accuracy": 0.7776917118512781, "eval_f1_macro": 0.7295133712588342, "eval_f1_micro": 0.7776917118512781, "eval_f1_weighted": 0.7776503309668394, "eval_loss": 1.8943842649459839, "eval_macro_fpr": 0.0202894340480138, "eval_macro_sensitivity": 0.7265129858866516, "eval_macro_specificity": 0.9833416395469147, "eval_precision": 0.787044263189245, "eval_precision_macro": 0.7538684462791384, "eval_recall": 0.7776917118512781, "eval_recall_macro": 0.7265129858866516, "eval_runtime": 289.6948, "eval_samples_per_second": 4.456, "eval_steps_per_second": 2.23, "eval_weighted_fpr": 0.02000976085895559, "eval_weighted_sensitivity": 0.7776917118512781, "eval_weighted_specificity": 0.9724328813524437, "step": 41104 }, { "epoch": 16.15, "learning_rate": 2.309134553003763e-05, "loss": 0.2378, "step": 41500 }, { "epoch": 16.35, "learning_rate": 2.276696509666537e-05, "loss": 0.2691, "step": 42000 }, { "epoch": 16.54, "learning_rate": 2.244258466329311e-05, "loss": 0.3306, "step": 42500 }, { "epoch": 16.74, "learning_rate": 2.211820422992085e-05, "loss": 0.2808, "step": 43000 }, { "epoch": 16.93, "learning_rate": 2.1794472557415336e-05, "loss": 0.4319, "step": 43500 }, { "epoch": 17.0, "eval_accuracy": 0.7885360185902401, "eval_f1_macro": 0.7269032248062757, "eval_f1_micro": 0.7885360185902401, "eval_f1_weighted": 0.7855493274616836, "eval_loss": 1.7744460105895996, "eval_macro_fpr": 0.01947959283678764, "eval_macro_sensitivity": 0.7319521995692674, "eval_macro_specificity": 0.9840228815937803, "eval_precision": 0.7847388778381785, "eval_precision_macro": 0.7247172757956307, "eval_recall": 0.7885360185902401, "eval_recall_macro": 0.7319521995692674, "eval_runtime": 288.2495, "eval_samples_per_second": 4.479, "eval_steps_per_second": 2.241, "eval_weighted_fpr": 0.018795180722891568, "eval_weighted_sensitivity": 0.7885360185902401, "eval_weighted_specificity": 0.9718072053164638, "step": 43673 }, { "epoch": 17.13, "learning_rate": 2.1470092124043077e-05, "loss": 0.1443, "step": 44000 }, { "epoch": 17.32, "learning_rate": 2.1145711690670818e-05, "loss": 0.3099, "step": 44500 }, { "epoch": 17.52, "learning_rate": 2.0821331257298558e-05, "loss": 0.2612, "step": 45000 }, { "epoch": 17.71, "learning_rate": 2.0496950823926302e-05, "loss": 0.1737, "step": 45500 }, { "epoch": 17.91, "learning_rate": 2.0173219151420787e-05, "loss": 0.2347, "step": 46000 }, { "epoch": 18.0, "eval_accuracy": 0.7412858249419055, "eval_f1_macro": 0.6825268580987619, "eval_f1_micro": 0.7412858249419055, "eval_f1_weighted": 0.7354314259956967, "eval_loss": 2.0036449432373047, "eval_macro_fpr": 0.025493876443533448, "eval_macro_sensitivity": 0.6799085681343341, "eval_macro_specificity": 0.980067466354203, "eval_precision": 0.7352104687796146, "eval_precision_macro": 0.6933956510787752, "eval_recall": 0.7412858249419055, "eval_recall_macro": 0.6799085681343341, "eval_runtime": 284.7586, "eval_samples_per_second": 4.534, "eval_steps_per_second": 2.269, "eval_weighted_fpr": 0.024322749781532187, "eval_weighted_sensitivity": 0.7412858249419055, "eval_weighted_specificity": 0.9597261703711405, "step": 46242 }, { "epoch": 18.1, "learning_rate": 1.9848838718048528e-05, "loss": 0.2375, "step": 46500 }, { "epoch": 18.3, "learning_rate": 1.952445828467627e-05, "loss": 0.1648, "step": 47000 }, { "epoch": 18.49, "learning_rate": 1.9200726612170754e-05, "loss": 0.1856, "step": 47500 }, { "epoch": 18.68, "learning_rate": 1.8876346178798494e-05, "loss": 0.3288, "step": 48000 }, { "epoch": 18.88, "learning_rate": 1.8551965745426235e-05, "loss": 0.1882, "step": 48500 }, { "epoch": 19.0, "eval_accuracy": 0.7815646785437645, "eval_f1_macro": 0.7224569664119563, "eval_f1_micro": 0.7815646785437644, "eval_f1_weighted": 0.7792243003941057, "eval_loss": 1.9297939538955688, "eval_macro_fpr": 0.020184156724894552, "eval_macro_sensitivity": 0.7261931252462459, "eval_macro_specificity": 0.9834902475280387, "eval_precision": 0.7803663621729094, "eval_precision_macro": 0.7242785000954916, "eval_recall": 0.7815646785437645, "eval_recall_macro": 0.7261931252462459, "eval_runtime": 290.7761, "eval_samples_per_second": 4.44, "eval_steps_per_second": 2.222, "eval_weighted_fpr": 0.01957245974458634, "eval_weighted_sensitivity": 0.7815646785437645, "eval_weighted_specificity": 0.9707890343768164, "step": 48811 }, { "epoch": 19.07, "learning_rate": 1.8227585312053975e-05, "loss": 0.2023, "step": 49000 }, { "epoch": 19.27, "learning_rate": 1.7903204878681716e-05, "loss": 0.1941, "step": 49500 }, { "epoch": 19.46, "learning_rate": 1.757882444530946e-05, "loss": 0.1347, "step": 50000 }, { "epoch": 19.66, "learning_rate": 1.72544440119372e-05, "loss": 0.1518, "step": 50500 }, { "epoch": 19.85, "learning_rate": 1.693006357856494e-05, "loss": 0.1799, "step": 51000 }, { "epoch": 20.0, "eval_accuracy": 0.7792408985282726, "eval_f1_macro": 0.7242164682622322, "eval_f1_micro": 0.7792408985282726, "eval_f1_weighted": 0.77792735433031, "eval_loss": 1.9687796831130981, "eval_macro_fpr": 0.020494577008212903, "eval_macro_sensitivity": 0.7343346620007255, "eval_macro_specificity": 0.9833745449065493, "eval_precision": 0.7891902482075702, "eval_precision_macro": 0.7312255238513404, "eval_recall": 0.7792408985282726, "eval_recall_macro": 0.7343346620007255, "eval_runtime": 285.0268, "eval_samples_per_second": 4.529, "eval_steps_per_second": 2.266, "eval_weighted_fpr": 0.019834365648270583, "eval_weighted_sensitivity": 0.7792408985282726, "eval_weighted_specificity": 0.9713772750699678, "step": 51380 }, { "epoch": 20.05, "learning_rate": 1.660568314519268e-05, "loss": 0.1389, "step": 51500 }, { "epoch": 20.24, "learning_rate": 1.6281302711820422e-05, "loss": 0.113, "step": 52000 }, { "epoch": 20.44, "learning_rate": 1.5956922278448163e-05, "loss": 0.1118, "step": 52500 }, { "epoch": 20.63, "learning_rate": 1.5632541845075907e-05, "loss": 0.1845, "step": 53000 }, { "epoch": 20.83, "learning_rate": 1.5308161411703647e-05, "loss": 0.1366, "step": 53500 }, { "epoch": 21.0, "eval_accuracy": 0.7846630518977536, "eval_f1_macro": 0.7265114805047838, "eval_f1_micro": 0.7846630518977536, "eval_f1_weighted": 0.7832740678095101, "eval_loss": 1.9910492897033691, "eval_macro_fpr": 0.019831051752647592, "eval_macro_sensitivity": 0.745505358404397, "eval_macro_specificity": 0.9838452330123495, "eval_precision": 0.784616464055549, "eval_precision_macro": 0.7147611761715634, "eval_recall": 0.7846630518977536, "eval_recall_macro": 0.745505358404397, "eval_runtime": 290.3707, "eval_samples_per_second": 4.446, "eval_steps_per_second": 2.225, "eval_weighted_fpr": 0.01922544951590595, "eval_weighted_sensitivity": 0.7846630518977536, "eval_weighted_specificity": 0.9730154432874895, "step": 53949 }, { "epoch": 21.02, "learning_rate": 1.4984429739198132e-05, "loss": 0.2057, "step": 54000 }, { "epoch": 21.21, "learning_rate": 1.4660049305825873e-05, "loss": 0.1708, "step": 54500 }, { "epoch": 21.41, "learning_rate": 1.4335668872453614e-05, "loss": 0.084, "step": 55000 }, { "epoch": 21.6, "learning_rate": 1.4011288439081354e-05, "loss": 0.1225, "step": 55500 }, { "epoch": 21.8, "learning_rate": 1.3686908005709095e-05, "loss": 0.1008, "step": 56000 }, { "epoch": 21.99, "learning_rate": 1.3362527572336835e-05, "loss": 0.1793, "step": 56500 }, { "epoch": 22.0, "eval_accuracy": 0.7629744384198296, "eval_f1_macro": 0.714971434368363, "eval_f1_micro": 0.7629744384198296, "eval_f1_weighted": 0.7582223914248779, "eval_loss": 2.2548303604125977, "eval_macro_fpr": 0.02295397517276209, "eval_macro_sensitivity": 0.7272583059978417, "eval_macro_specificity": 0.9817539896299595, "eval_precision": 0.7648005911254243, "eval_precision_macro": 0.7150489472675985, "eval_recall": 0.7629744384198296, "eval_recall_macro": 0.7272583059978417, "eval_runtime": 352.8043, "eval_samples_per_second": 3.659, "eval_steps_per_second": 1.831, "eval_weighted_fpr": 0.021708286038592508, "eval_weighted_sensitivity": 0.7629744384198296, "eval_weighted_specificity": 0.9633354060295618, "step": 56518 }, { "epoch": 22.19, "learning_rate": 1.303814713896458e-05, "loss": 0.1293, "step": 57000 }, { "epoch": 22.38, "learning_rate": 1.271376670559232e-05, "loss": 0.1532, "step": 57500 }, { "epoch": 22.58, "learning_rate": 1.2390035033086805e-05, "loss": 0.2066, "step": 58000 }, { "epoch": 22.77, "learning_rate": 1.2065654599714546e-05, "loss": 0.1326, "step": 58500 }, { "epoch": 22.97, "learning_rate": 1.174192292720903e-05, "loss": 0.1749, "step": 59000 }, { "epoch": 23.0, "eval_accuracy": 0.7815646785437645, "eval_f1_macro": 0.7289459925102588, "eval_f1_micro": 0.7815646785437644, "eval_f1_weighted": 0.7773605214390553, "eval_loss": 2.1108603477478027, "eval_macro_fpr": 0.02054851528566883, "eval_macro_sensitivity": 0.7230322921449998, "eval_macro_specificity": 0.9833739271591507, "eval_precision": 0.7767807321927734, "eval_precision_macro": 0.7466305539706889, "eval_recall": 0.7815646785437645, "eval_recall_macro": 0.7230322921449998, "eval_runtime": 297.526, "eval_samples_per_second": 4.339, "eval_steps_per_second": 2.171, "eval_weighted_fpr": 0.01957245974458634, "eval_weighted_sensitivity": 0.7815646785437645, "eval_weighted_specificity": 0.9690442288434956, "step": 59087 }, { "epoch": 23.16, "learning_rate": 1.1417542493836771e-05, "loss": 0.1763, "step": 59500 }, { "epoch": 23.36, "learning_rate": 1.1093162060464514e-05, "loss": 0.1039, "step": 60000 }, { "epoch": 23.55, "learning_rate": 1.0768781627092254e-05, "loss": 0.1895, "step": 60500 }, { "epoch": 23.74, "learning_rate": 1.0444401193719995e-05, "loss": 0.1544, "step": 61000 }, { "epoch": 23.94, "learning_rate": 1.0120020760347735e-05, "loss": 0.1154, "step": 61500 }, { "epoch": 24.0, "eval_accuracy": 0.7877614252517429, "eval_f1_macro": 0.7331374609108605, "eval_f1_micro": 0.7877614252517429, "eval_f1_weighted": 0.7827775199528373, "eval_loss": 2.0637075901031494, "eval_macro_fpr": 0.019584306852913787, "eval_macro_sensitivity": 0.7268862975255689, "eval_macro_specificity": 0.9839707825927626, "eval_precision": 0.7836989509025826, "eval_precision_macro": 0.7590439050861235, "eval_recall": 0.7877614252517429, "eval_recall_macro": 0.7268862975255689, "eval_runtime": 293.3276, "eval_samples_per_second": 4.401, "eval_steps_per_second": 2.202, "eval_weighted_fpr": 0.01888092613009923, "eval_weighted_sensitivity": 0.7877614252517429, "eval_weighted_specificity": 0.9718003136396952, "step": 61656 }, { "epoch": 24.13, "learning_rate": 9.795640326975478e-06, "loss": 0.1484, "step": 62000 }, { "epoch": 24.33, "learning_rate": 9.471259893603218e-06, "loss": 0.1147, "step": 62500 }, { "epoch": 24.52, "learning_rate": 9.146879460230959e-06, "loss": 0.1212, "step": 63000 }, { "epoch": 24.72, "learning_rate": 8.822499026858701e-06, "loss": 0.0781, "step": 63500 }, { "epoch": 24.91, "learning_rate": 8.498767354353186e-06, "loss": 0.1447, "step": 64000 }, { "epoch": 25.0, "eval_accuracy": 0.7916343919442292, "eval_f1_macro": 0.7408195079003096, "eval_f1_micro": 0.7916343919442292, "eval_f1_weighted": 0.7861355334433522, "eval_loss": 2.0027315616607666, "eval_macro_fpr": 0.019407099074489707, "eval_macro_sensitivity": 0.7299369476165393, "eval_macro_specificity": 0.984089840852184, "eval_precision": 0.7857715570004769, "eval_precision_macro": 0.7750096561329829, "eval_recall": 0.7916343919442292, "eval_recall_macro": 0.7299369476165393, "eval_runtime": 300.7258, "eval_samples_per_second": 4.293, "eval_steps_per_second": 2.148, "eval_weighted_fpr": 0.01845372847636688, "eval_weighted_sensitivity": 0.7916343919442292, "eval_weighted_specificity": 0.9697132208385308, "step": 64225 }, { "epoch": 25.11, "learning_rate": 8.174386920980927e-06, "loss": 0.1036, "step": 64500 }, { "epoch": 25.3, "learning_rate": 7.850006487608667e-06, "loss": 0.1165, "step": 65000 }, { "epoch": 25.5, "learning_rate": 7.52562605423641e-06, "loss": 0.0632, "step": 65500 }, { "epoch": 25.69, "learning_rate": 7.20124562086415e-06, "loss": 0.1215, "step": 66000 }, { "epoch": 25.89, "learning_rate": 6.876865187491891e-06, "loss": 0.0806, "step": 66500 }, { "epoch": 26.0, "eval_accuracy": 0.7885360185902401, "eval_f1_macro": 0.7118193374018945, "eval_f1_micro": 0.7885360185902401, "eval_f1_weighted": 0.7839622559886584, "eval_loss": 2.077707529067993, "eval_macro_fpr": 0.019568063989489137, "eval_macro_sensitivity": 0.713410599172433, "eval_macro_specificity": 0.9840056220595941, "eval_precision": 0.783060971098355, "eval_precision_macro": 0.7162166073441082, "eval_recall": 0.7885360185902401, "eval_recall_macro": 0.713410599172433, "eval_runtime": 293.3091, "eval_samples_per_second": 4.401, "eval_steps_per_second": 2.202, "eval_weighted_fpr": 0.018795180722891568, "eval_weighted_sensitivity": 0.7885360185902401, "eval_weighted_specificity": 0.9715483123036702, "step": 66794 }, { "epoch": 26.08, "learning_rate": 6.552484754119632e-06, "loss": 0.0969, "step": 67000 }, { "epoch": 26.27, "learning_rate": 6.228104320747373e-06, "loss": 0.1048, "step": 67500 }, { "epoch": 26.47, "learning_rate": 5.904372648241859e-06, "loss": 0.1223, "step": 68000 }, { "epoch": 26.66, "learning_rate": 5.5799922148695995e-06, "loss": 0.079, "step": 68500 }, { "epoch": 26.86, "learning_rate": 5.25561178149734e-06, "loss": 0.0407, "step": 69000 }, { "epoch": 27.0, "eval_accuracy": 0.7885360185902401, "eval_f1_macro": 0.710517336505339, "eval_f1_micro": 0.7885360185902401, "eval_f1_weighted": 0.7866391697664283, "eval_loss": 2.175379991531372, "eval_macro_fpr": 0.01935082910687802, "eval_macro_sensitivity": 0.7079594722658412, "eval_macro_specificity": 0.9840709457196527, "eval_precision": 0.7862704372568858, "eval_precision_macro": 0.7192207626055576, "eval_recall": 0.7885360185902401, "eval_recall_macro": 0.7079594722658412, "eval_runtime": 323.9439, "eval_samples_per_second": 3.985, "eval_steps_per_second": 1.994, "eval_weighted_fpr": 0.018795180722891568, "eval_weighted_sensitivity": 0.7885360185902401, "eval_weighted_specificity": 0.9725281672045502, "step": 69363 }, { "epoch": 27.05, "learning_rate": 4.931880108991826e-06, "loss": 0.144, "step": 69500 }, { "epoch": 27.25, "learning_rate": 4.6074996756195675e-06, "loss": 0.1027, "step": 70000 }, { "epoch": 27.44, "learning_rate": 4.283119242247308e-06, "loss": 0.0783, "step": 70500 }, { "epoch": 27.64, "learning_rate": 3.958738808875049e-06, "loss": 0.1075, "step": 71000 }, { "epoch": 27.83, "learning_rate": 3.63435837550279e-06, "loss": 0.0701, "step": 71500 }, { "epoch": 28.0, "eval_accuracy": 0.7823392718822618, "eval_f1_macro": 0.7065654727253492, "eval_f1_micro": 0.7823392718822618, "eval_f1_weighted": 0.7810038357002493, "eval_loss": 2.1577563285827637, "eval_macro_fpr": 0.020110114145651264, "eval_macro_sensitivity": 0.7096633019016597, "eval_macro_specificity": 0.9835831643616377, "eval_precision": 0.7816909903991323, "eval_precision_macro": 0.7129613601879862, "eval_recall": 0.7823392718822618, "eval_recall_macro": 0.7096633019016597, "eval_runtime": 309.7699, "eval_samples_per_second": 4.168, "eval_steps_per_second": 2.085, "eval_weighted_fpr": 0.019485472574717427, "eval_weighted_sensitivity": 0.7823392718822618, "eval_weighted_specificity": 0.9714081935423033, "step": 71932 }, { "epoch": 28.03, "learning_rate": 3.3099779421305306e-06, "loss": 0.0839, "step": 72000 }, { "epoch": 28.22, "learning_rate": 2.986246269625016e-06, "loss": 0.0878, "step": 72500 }, { "epoch": 28.42, "learning_rate": 2.6618658362527576e-06, "loss": 0.0574, "step": 73000 }, { "epoch": 28.61, "learning_rate": 2.3374854028804986e-06, "loss": 0.0807, "step": 73500 }, { "epoch": 28.8, "learning_rate": 2.0131049695082392e-06, "loss": 0.0909, "step": 74000 }, { "epoch": 29.0, "learning_rate": 1.6887245361359802e-06, "loss": 0.1034, "step": 74500 }, { "epoch": 29.0, "eval_accuracy": 0.78001549186677, "eval_f1_macro": 0.7063839754791109, "eval_f1_micro": 0.78001549186677, "eval_f1_weighted": 0.778480865627508, "eval_loss": 2.213203191757202, "eval_macro_fpr": 0.020340341198432747, "eval_macro_sensitivity": 0.7043594983806052, "eval_macro_specificity": 0.9834222624437919, "eval_precision": 0.7788510250837947, "eval_precision_macro": 0.7162913991417985, "eval_recall": 0.78001549186677, "eval_recall_macro": 0.7043594983806052, "eval_runtime": 356.7589, "eval_samples_per_second": 3.619, "eval_steps_per_second": 1.811, "eval_weighted_fpr": 0.0197469058545404, "eval_weighted_sensitivity": 0.78001549186677, "eval_weighted_specificity": 0.971318444790108, "step": 74501 }, { "epoch": 29.19, "learning_rate": 1.3643441027637214e-06, "loss": 0.0292, "step": 75000 }, { "epoch": 29.39, "learning_rate": 1.0399636693914622e-06, "loss": 0.0937, "step": 75500 }, { "epoch": 29.58, "learning_rate": 7.155832360192033e-07, "loss": 0.1159, "step": 76000 }, { "epoch": 29.78, "learning_rate": 3.9185156351368886e-07, "loss": 0.1325, "step": 76500 }, { "epoch": 29.97, "learning_rate": 6.747113014142988e-08, "loss": 0.0388, "step": 77000 }, { "epoch": 30.0, "eval_accuracy": 0.7823392718822618, "eval_f1_macro": 0.708030615398878, "eval_f1_micro": 0.7823392718822618, "eval_f1_weighted": 0.780149605946002, "eval_loss": 2.183318853378296, "eval_macro_fpr": 0.020101371906798372, "eval_macro_sensitivity": 0.7056009725554542, "eval_macro_specificity": 0.9835845291455382, "eval_precision": 0.7806128914771521, "eval_precision_macro": 0.7201007317849041, "eval_recall": 0.7823392718822618, "eval_recall_macro": 0.7056009725554542, "eval_runtime": 331.848, "eval_samples_per_second": 3.89, "eval_steps_per_second": 1.947, "eval_weighted_fpr": 0.019485472574717427, "eval_weighted_sensitivity": 0.7823392718822618, "eval_weighted_specificity": 0.971428665300812, "step": 77070 } ], "logging_steps": 500, "max_steps": 77070, "num_train_epochs": 30, "save_steps": 500, "total_flos": 1.6752303660201984e+17, "trial_name": null, "trial_params": null }