{ "best_metric": 0.9235158618894997, "best_model_checkpoint": "./saved_models/cliroberta_prompt_sbdh_gpt4_v2_0/checkpoint-408", "epoch": 17.0, "eval_steps": 500, "global_step": 408, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 1.0, "grad_norm": 6.645392417907715, "learning_rate": 2.5e-06, "loss": 1.8965, "step": 24 }, { "epoch": 1.0, "eval_acc_macro": 0.07351644449088429, "eval_acc_micro": 0.16844170403586972, "eval_auc_macro": 0.5519856670284093, "eval_auc_micro": 0.6260212922777759, "eval_f1_at_5": 0.25157489791625554, "eval_f1_at_8": 0.2244759753473909, "eval_f1_macro": 0.12396975795306268, "eval_f1_micro": 0.2883185416166808, "eval_loss": 0.3488869071006775, "eval_prec_at_5": 0.16232876712328767, "eval_prec_at_8": 0.13227739726027396, "eval_prec_macro": 0.1633962475788553, "eval_prec_micro": 0.20283496456293612, "eval_rec_at_5": 0.5587899543378996, "eval_rec_at_8": 0.7408675799086758, "eval_rec_macro": 0.3086744639374022, "eval_rec_micro": 0.4983416252072555, "eval_runtime": 4.7343, "eval_samples_per_second": 185.032, "eval_steps_per_second": 23.235, "step": 24 }, { "epoch": 2.0, "grad_norm": 6.26666784286499, "learning_rate": 5e-06, "loss": 0.3531, "step": 48 }, { "epoch": 2.0, "eval_acc_macro": 0.11095425738710896, "eval_acc_micro": 0.23921319796952797, "eval_auc_macro": 0.8260589461583061, "eval_auc_micro": 0.7608045184921441, "eval_f1_at_5": 0.3274833326039068, "eval_f1_at_8": 0.26273329679353813, "eval_f1_macro": 0.16550844622182223, "eval_f1_micro": 0.3860727086533143, "eval_loss": 0.3172319531440735, "eval_prec_at_5": 0.2100456621004566, "eval_prec_at_8": 0.1545376712328767, "eval_prec_macro": 0.36732877721657925, "eval_prec_micro": 0.5046854082997986, "eval_rec_at_5": 0.7427701674277017, "eval_rec_at_8": 0.8761415525114156, "eval_rec_macro": 0.1808019430800684, "eval_rec_micro": 0.312603648424518, "eval_runtime": 6.9454, "eval_samples_per_second": 126.127, "eval_steps_per_second": 15.838, "step": 48 }, { "epoch": 3.0, "grad_norm": 1.4328076839447021, "learning_rate": 7.500000000000001e-06, "loss": 0.2409, "step": 72 }, { "epoch": 3.0, "eval_acc_macro": 0.6542871174273012, "eval_acc_micro": 0.6627906976743733, "eval_auc_macro": 0.9683337499122936, "eval_auc_micro": 0.9677785939562165, "eval_f1_at_5": 0.4172108144155159, "eval_f1_at_8": 0.29078564566802617, "eval_f1_macro": 0.7752458106791931, "eval_f1_micro": 0.7972027972027316, "eval_loss": 0.1295301616191864, "eval_prec_at_5": 0.2671232876712329, "eval_prec_at_8": 0.1709474885844749, "eval_prec_macro": 0.7878146103091771, "eval_prec_micro": 0.7910204081632007, "eval_rec_at_5": 0.9522450532724505, "eval_rec_at_8": 0.9726027397260274, "eval_rec_macro": 0.7778465051663405, "eval_rec_micro": 0.80348258706461, "eval_runtime": 7.0065, "eval_samples_per_second": 125.028, "eval_steps_per_second": 15.7, "step": 72 }, { "epoch": 4.0, "grad_norm": 2.6871750354766846, "learning_rate": 1e-05, "loss": 0.1262, "step": 96 }, { "epoch": 4.0, "eval_acc_macro": 0.7556596335774451, "eval_acc_micro": 0.7631578947367863, "eval_auc_macro": 0.9859339445130657, "eval_auc_micro": 0.9878248630762655, "eval_f1_at_5": 0.4281991387591176, "eval_f1_at_8": 0.2925113305987028, "eval_f1_macro": 0.8563120619008121, "eval_f1_micro": 0.865671641790973, "eval_loss": 0.086879663169384, "eval_prec_at_5": 0.2742009132420092, "eval_prec_at_8": 0.17194634703196346, "eval_prec_macro": 0.8563984905449645, "eval_prec_micro": 0.865671641790973, "eval_rec_at_5": 0.9767884322678843, "eval_rec_at_8": 0.9788812785388128, "eval_rec_macro": 0.8591516601823509, "eval_rec_micro": 0.865671641790973, "eval_runtime": 7.0436, "eval_samples_per_second": 124.369, "eval_steps_per_second": 15.617, "step": 96 }, { "epoch": 5.0, "grad_norm": 1.409122347831726, "learning_rate": 9.722222222222223e-06, "loss": 0.0916, "step": 120 }, { "epoch": 5.0, "eval_acc_macro": 0.8023972093976516, "eval_acc_micro": 0.8076079005120111, "eval_auc_macro": 0.9910191597807554, "eval_auc_micro": 0.99171908563779, "eval_f1_at_5": 0.4281442700601774, "eval_f1_at_8": 0.29274329072668936, "eval_f1_macro": 0.8879651934832161, "eval_f1_micro": 0.893565358154521, "eval_loss": 0.06939995288848877, "eval_prec_at_5": 0.2742009132420092, "eval_prec_at_8": 0.1720890410958904, "eval_prec_macro": 0.8664154958281926, "eval_prec_micro": 0.8727272727272037, "eval_rec_at_5": 0.9762176560121765, "eval_rec_at_8": 0.9794520547945206, "eval_rec_macro": 0.9116566511084603, "eval_rec_micro": 0.9154228855720634, "eval_runtime": 7.006, "eval_samples_per_second": 125.035, "eval_steps_per_second": 15.701, "step": 120 }, { "epoch": 6.0, "grad_norm": 1.1020153760910034, "learning_rate": 9.444444444444445e-06, "loss": 0.0785, "step": 144 }, { "epoch": 6.0, "eval_acc_macro": 0.8155311593684006, "eval_acc_micro": 0.8230421686746368, "eval_auc_macro": 0.9922239037332781, "eval_auc_micro": 0.9932765964575004, "eval_f1_at_5": 0.4288655829720509, "eval_f1_at_8": 0.2925113305987028, "eval_f1_macro": 0.8957459298097094, "eval_f1_micro": 0.9029326724493264, "eval_loss": 0.06519585102796555, "eval_prec_at_5": 0.27465753424657535, "eval_prec_at_8": 0.17194634703196346, "eval_prec_macro": 0.8920600176459569, "eval_prec_micro": 0.8995884773661811, "eval_rec_at_5": 0.9779299847792998, "eval_rec_at_8": 0.9788812785388128, "eval_rec_macro": 0.900216189836664, "eval_rec_micro": 0.9063018242121967, "eval_runtime": 7.0038, "eval_samples_per_second": 125.075, "eval_steps_per_second": 15.706, "step": 144 }, { "epoch": 7.0, "grad_norm": 1.087604284286499, "learning_rate": 9.166666666666666e-06, "loss": 0.0692, "step": 168 }, { "epoch": 7.0, "eval_acc_macro": 0.8256823767025346, "eval_acc_micro": 0.8327024981074312, "eval_auc_macro": 0.9930553828317482, "eval_auc_micro": 0.9940669351858088, "eval_f1_at_5": 0.4291987915672543, "eval_f1_at_8": 0.2925113305987028, "eval_f1_macro": 0.9023354543659394, "eval_f1_micro": 0.9087154068565958, "eval_loss": 0.06074792891740799, "eval_prec_at_5": 0.27488584474885847, "eval_prec_at_8": 0.17194634703196346, "eval_prec_macro": 0.8963122585533565, "eval_prec_micro": 0.9053497942386086, "eval_rec_at_5": 0.9785007610350076, "eval_rec_at_8": 0.9788812785388128, "eval_rec_macro": 0.9088224265936838, "eval_rec_micro": 0.9121061359866574, "eval_runtime": 6.9556, "eval_samples_per_second": 125.942, "eval_steps_per_second": 15.815, "step": 168 }, { "epoch": 8.0, "grad_norm": 1.107313632965088, "learning_rate": 8.888888888888888e-06, "loss": 0.0624, "step": 192 }, { "epoch": 8.0, "eval_acc_macro": 0.8278930665915176, "eval_acc_micro": 0.832705350414406, "eval_auc_macro": 0.9935760753995465, "eval_auc_micro": 0.9946024409081464, "eval_f1_at_5": 0.4291987915672543, "eval_f1_at_8": 0.2925113305987028, "eval_f1_macro": 0.9037268724978929, "eval_f1_micro": 0.9087171052630831, "eval_loss": 0.060165293514728546, "eval_prec_at_5": 0.27488584474885847, "eval_prec_at_8": 0.17194634703196346, "eval_prec_macro": 0.8946308202984045, "eval_prec_micro": 0.9013050570961744, "eval_rec_at_5": 0.9785007610350076, "eval_rec_at_8": 0.9788812785388128, "eval_rec_macro": 0.9137053638007582, "eval_rec_micro": 0.9162520729684148, "eval_runtime": 6.9888, "eval_samples_per_second": 125.344, "eval_steps_per_second": 15.74, "step": 192 }, { "epoch": 9.0, "grad_norm": 0.9251410961151123, "learning_rate": 8.611111111111112e-06, "loss": 0.058, "step": 216 }, { "epoch": 9.0, "eval_acc_macro": 0.8361146010604618, "eval_acc_micro": 0.8392307692307046, "eval_auc_macro": 0.9940701262860925, "eval_auc_micro": 0.9949116264083977, "eval_f1_at_5": 0.4295136788854873, "eval_f1_at_8": 0.2925113305987028, "eval_f1_macro": 0.9089752463509541, "eval_f1_micro": 0.9125888749476442, "eval_loss": 0.05600544437766075, "eval_prec_at_5": 0.2751141552511416, "eval_prec_at_8": 0.17194634703196346, "eval_prec_macro": 0.9171011361277855, "eval_prec_micro": 0.9206751054851543, "eval_rec_at_5": 0.9788812785388128, "eval_rec_at_8": 0.9788812785388128, "eval_rec_macro": 0.90156843038978, "eval_rec_micro": 0.9046434494194938, "eval_runtime": 7.0672, "eval_samples_per_second": 123.953, "eval_steps_per_second": 15.565, "step": 216 }, { "epoch": 10.0, "grad_norm": 1.4386526346206665, "learning_rate": 8.333333333333334e-06, "loss": 0.0556, "step": 240 }, { "epoch": 10.0, "eval_acc_macro": 0.8346145293076771, "eval_acc_micro": 0.8394103956554818, "eval_auc_macro": 0.9940972007833064, "eval_auc_micro": 0.9950393993211819, "eval_f1_at_5": 0.42918048633589306, "eval_f1_at_8": 0.2925113305987028, "eval_f1_macro": 0.9076071562375437, "eval_f1_micro": 0.9126950653731831, "eval_loss": 0.05595849081873894, "eval_prec_at_5": 0.27488584474885847, "eval_prec_at_8": 0.17194634703196346, "eval_prec_macro": 0.9241533819936333, "eval_prec_micro": 0.9287553648067872, "eval_rec_at_5": 0.978310502283105, "eval_rec_at_8": 0.9788812785388128, "eval_rec_macro": 0.8925287390946574, "eval_rec_micro": 0.8971807628523302, "eval_runtime": 6.9532, "eval_samples_per_second": 125.984, "eval_steps_per_second": 15.82, "step": 240 }, { "epoch": 11.0, "grad_norm": 1.3220967054367065, "learning_rate": 8.055555555555557e-06, "loss": 0.0519, "step": 264 }, { "epoch": 11.0, "eval_acc_macro": 0.8442462130549724, "eval_acc_micro": 0.8494957331263887, "eval_auc_macro": 0.9943053813562361, "eval_auc_micro": 0.995346713669991, "eval_f1_at_5": 0.4288655829720509, "eval_f1_at_8": 0.2925113305987028, "eval_f1_macro": 0.9133293255310709, "eval_f1_micro": 0.9186241610737484, "eval_loss": 0.05388868972659111, "eval_prec_at_5": 0.27465753424657535, "eval_prec_at_8": 0.17194634703196346, "eval_prec_macro": 0.9247426584222537, "eval_prec_micro": 0.9295415959252181, "eval_rec_at_5": 0.9779299847792998, "eval_rec_at_8": 0.9788812785388128, "eval_rec_macro": 0.9033312145477272, "eval_rec_micro": 0.9079601990048998, "eval_runtime": 6.9787, "eval_samples_per_second": 125.524, "eval_steps_per_second": 15.762, "step": 264 }, { "epoch": 12.0, "grad_norm": 0.9175763130187988, "learning_rate": 7.77777777777778e-06, "loss": 0.0492, "step": 288 }, { "epoch": 12.0, "eval_acc_macro": 0.8470976768477376, "eval_acc_micro": 0.8500376789750678, "eval_auc_macro": 0.994169602401962, "eval_auc_micro": 0.9953780777322643, "eval_f1_at_5": 0.42853236537508876, "eval_f1_at_8": 0.2925113305987028, "eval_f1_macro": 0.9152662917583787, "eval_f1_micro": 0.9189409368634688, "eval_loss": 0.056095000356435776, "eval_prec_at_5": 0.2744292237442923, "eval_prec_at_8": 0.17194634703196346, "eval_prec_macro": 0.9004415240992042, "eval_prec_micro": 0.9031224979983263, "eval_rec_at_5": 0.977359208523592, "eval_rec_at_8": 0.9788812785388128, "eval_rec_macro": 0.9315482026514185, "eval_rec_micro": 0.9353233830844995, "eval_runtime": 6.9839, "eval_samples_per_second": 125.432, "eval_steps_per_second": 15.751, "step": 288 }, { "epoch": 13.0, "grad_norm": 1.00278639793396, "learning_rate": 7.500000000000001e-06, "loss": 0.048, "step": 312 }, { "epoch": 13.0, "eval_acc_macro": 0.8486995058416006, "eval_acc_micro": 0.8536776212831882, "eval_auc_macro": 0.9946152044988873, "eval_auc_micro": 0.9955295518966525, "eval_f1_at_5": 0.42918048633589306, "eval_f1_at_8": 0.2925113305987028, "eval_f1_macro": 0.9163503224179071, "eval_f1_micro": 0.921063739974595, "eval_loss": 0.05099855735898018, "eval_prec_at_5": 0.27488584474885847, "eval_prec_at_8": 0.17194634703196346, "eval_prec_macro": 0.9313959829408701, "eval_prec_micro": 0.9380911435940723, "eval_rec_at_5": 0.978310502283105, "eval_rec_at_8": 0.9788812785388128, "eval_rec_macro": 0.9025756160114312, "eval_rec_micro": 0.9046434494194938, "eval_runtime": 6.9796, "eval_samples_per_second": 125.509, "eval_steps_per_second": 15.76, "step": 312 }, { "epoch": 14.0, "grad_norm": 0.7232435941696167, "learning_rate": 7.222222222222223e-06, "loss": 0.0457, "step": 336 }, { "epoch": 14.0, "eval_acc_macro": 0.8525145546996621, "eval_acc_micro": 0.8559322033897645, "eval_auc_macro": 0.9948837193480031, "eval_auc_micro": 0.9959032475704436, "eval_f1_at_5": 0.4291987915672543, "eval_f1_at_8": 0.2925113305987028, "eval_f1_macro": 0.9187720987706429, "eval_f1_micro": 0.9223744292236677, "eval_loss": 0.04985074698925018, "eval_prec_at_5": 0.27488584474885847, "eval_prec_at_8": 0.17194634703196346, "eval_prec_macro": 0.9187488547711858, "eval_prec_micro": 0.9235245220281859, "eval_rec_at_5": 0.9785007610350076, "eval_rec_at_8": 0.9788812785388128, "eval_rec_macro": 0.9190659938481747, "eval_rec_micro": 0.9212271973465239, "eval_runtime": 6.9629, "eval_samples_per_second": 125.809, "eval_steps_per_second": 15.798, "step": 336 }, { "epoch": 15.0, "grad_norm": 0.7620428204536438, "learning_rate": 6.944444444444445e-06, "loss": 0.0444, "step": 360 }, { "epoch": 15.0, "eval_acc_macro": 0.8563973908810851, "eval_acc_micro": 0.8611532625189028, "eval_auc_macro": 0.9949728097109144, "eval_auc_micro": 0.9959143853766486, "eval_f1_at_5": 0.42918048633589306, "eval_f1_at_8": 0.2925113305987028, "eval_f1_macro": 0.9209429321639129, "eval_f1_micro": 0.9253974724825988, "eval_loss": 0.051042765378952026, "eval_prec_at_5": 0.27488584474885847, "eval_prec_at_8": 0.17194634703196346, "eval_prec_macro": 0.9035271919581529, "eval_prec_micro": 0.9101844426623167, "eval_rec_at_5": 0.978310502283105, "eval_rec_at_8": 0.9788812785388128, "eval_rec_macro": 0.9396785113976804, "eval_rec_micro": 0.9411276948589601, "eval_runtime": 7.0373, "eval_samples_per_second": 124.48, "eval_steps_per_second": 15.631, "step": 360 }, { "epoch": 16.0, "grad_norm": 1.0121941566467285, "learning_rate": 6.666666666666667e-06, "loss": 0.042, "step": 384 }, { "epoch": 16.0, "eval_acc_macro": 0.8578953265912642, "eval_acc_micro": 0.8632148377124526, "eval_auc_macro": 0.9950318833304016, "eval_auc_micro": 0.9960295057415836, "eval_f1_at_5": 0.42918048633589306, "eval_f1_at_8": 0.2925113305987028, "eval_f1_macro": 0.9219762526743717, "eval_f1_micro": 0.9265864786394916, "eval_loss": 0.04983897507190704, "eval_prec_at_5": 0.27488584474885847, "eval_prec_at_8": 0.17194634703196346, "eval_prec_macro": 0.9224728362696459, "eval_prec_micro": 0.9269709543567695, "eval_rec_at_5": 0.978310502283105, "eval_rec_at_8": 0.9788812785388128, "eval_rec_macro": 0.9228006946898867, "eval_rec_micro": 0.9262023217246329, "eval_runtime": 6.9735, "eval_samples_per_second": 125.619, "eval_steps_per_second": 15.774, "step": 384 }, { "epoch": 17.0, "grad_norm": 0.6501286029815674, "learning_rate": 6.3888888888888885e-06, "loss": 0.0401, "step": 408 }, { "epoch": 17.0, "eval_acc_macro": 0.8609431757906035, "eval_acc_micro": 0.8669760247485795, "eval_auc_macro": 0.9951994943771586, "eval_auc_micro": 0.996281264713042, "eval_f1_at_5": 0.42918048633589306, "eval_f1_at_8": 0.2925113305987028, "eval_f1_macro": 0.9235158618894997, "eval_f1_micro": 0.9287489643744051, "eval_loss": 0.04911518469452858, "eval_prec_at_5": 0.27488584474885847, "eval_prec_at_8": 0.17194634703196346, "eval_prec_macro": 0.9229370933414865, "eval_prec_micro": 0.9279801324502542, "eval_rec_at_5": 0.978310502283105, "eval_rec_at_8": 0.9788812785388128, "eval_rec_macro": 0.9243968142875807, "eval_rec_micro": 0.929519071310039, "eval_runtime": 6.9496, "eval_samples_per_second": 126.05, "eval_steps_per_second": 15.828, "step": 408 } ], "logging_steps": 500, "max_steps": 960, "num_input_tokens_seen": 0, "num_train_epochs": 40, "save_steps": 500, "total_flos": 1.3730832326295552e+16, "train_batch_size": 32, "trial_name": null, "trial_params": null }