| { |
| "best_metric": 0.7850116201409046, |
| "best_model_checkpoint": "roberta-base/checkpoint-7073", |
| "epoch": 15.0, |
| "eval_steps": 500, |
| "global_step": 9645, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.78, |
| "grad_norm": 16.478424072265625, |
| "learning_rate": 4.742871954380509e-05, |
| "loss": 1.4921, |
| "step": 500 |
| }, |
| { |
| "epoch": 1.0, |
| "eval_accuracy": 0.6762199845081333, |
| "eval_f1_macro": 0.36762215914350266, |
| "eval_f1_micro": 0.6762199845081333, |
| "eval_f1_weighted": 0.6381553142782734, |
| "eval_loss": 1.0307424068450928, |
| "eval_macro_fpr": 0.03508857639652692, |
| "eval_macro_sensitivity": 0.4027183655220987, |
| "eval_macro_specificity": 0.9746655126680946, |
| "eval_precision": 0.6361951106730956, |
| "eval_precision_macro": 0.3745780279088578, |
| "eval_recall": 0.6762199845081333, |
| "eval_recall_macro": 0.4027183655220987, |
| "eval_runtime": 65.0146, |
| "eval_samples_per_second": 19.857, |
| "eval_steps_per_second": 2.492, |
| "eval_weighted_fpr": 0.03306962025316456, |
| "eval_weighted_sensitivity": 0.6762199845081333, |
| "eval_weighted_specificity": 0.9437627055132857, |
| "step": 643 |
| }, |
| { |
| "epoch": 1.56, |
| "grad_norm": 11.052178382873535, |
| "learning_rate": 4.483670295489891e-05, |
| "loss": 0.9019, |
| "step": 1000 |
| }, |
| { |
| "epoch": 2.0, |
| "eval_accuracy": 0.737412858249419, |
| "eval_f1_macro": 0.4767493392755858, |
| "eval_f1_micro": 0.737412858249419, |
| "eval_f1_weighted": 0.7197263095885438, |
| "eval_loss": 0.8135831952095032, |
| "eval_macro_fpr": 0.025093207469991694, |
| "eval_macro_sensitivity": 0.5148741493607558, |
| "eval_macro_specificity": 0.9804978566683986, |
| "eval_precision": 0.7158327481302543, |
| "eval_precision_macro": 0.4625056436667857, |
| "eval_recall": 0.737412858249419, |
| "eval_recall_macro": 0.5148741493607558, |
| "eval_runtime": 78.3833, |
| "eval_samples_per_second": 16.47, |
| "eval_steps_per_second": 2.067, |
| "eval_weighted_fpr": 0.024804273066510573, |
| "eval_weighted_sensitivity": 0.737412858249419, |
| "eval_weighted_specificity": 0.9700549917765612, |
| "step": 1286 |
| }, |
| { |
| "epoch": 2.33, |
| "grad_norm": 16.875791549682617, |
| "learning_rate": 4.224468636599274e-05, |
| "loss": 0.7714, |
| "step": 1500 |
| }, |
| { |
| "epoch": 3.0, |
| "eval_accuracy": 0.7621998450813323, |
| "eval_f1_macro": 0.5064301619528891, |
| "eval_f1_micro": 0.7621998450813323, |
| "eval_f1_weighted": 0.7440787580787277, |
| "eval_loss": 0.8493772149085999, |
| "eval_macro_fpr": 0.022439395825223803, |
| "eval_macro_sensitivity": 0.5353612181578932, |
| "eval_macro_specificity": 0.9822327546765126, |
| "eval_precision": 0.7385639759374297, |
| "eval_precision_macro": 0.505572707108006, |
| "eval_recall": 0.7621998450813323, |
| "eval_recall_macro": 0.5353612181578932, |
| "eval_runtime": 62.8149, |
| "eval_samples_per_second": 20.552, |
| "eval_steps_per_second": 2.579, |
| "eval_weighted_fpr": 0.02179933252858056, |
| "eval_weighted_sensitivity": 0.7621998450813323, |
| "eval_weighted_specificity": 0.9712914750663563, |
| "step": 1929 |
| }, |
| { |
| "epoch": 3.11, |
| "grad_norm": 59.48842239379883, |
| "learning_rate": 3.9652669777086575e-05, |
| "loss": 0.7018, |
| "step": 2000 |
| }, |
| { |
| "epoch": 3.89, |
| "grad_norm": 1.6855363845825195, |
| "learning_rate": 3.706583722135822e-05, |
| "loss": 0.5351, |
| "step": 2500 |
| }, |
| { |
| "epoch": 4.0, |
| "eval_accuracy": 0.7645236250968241, |
| "eval_f1_macro": 0.544445143585482, |
| "eval_f1_micro": 0.7645236250968241, |
| "eval_f1_weighted": 0.7545581478665218, |
| "eval_loss": 0.9904704093933105, |
| "eval_macro_fpr": 0.022368675533492623, |
| "eval_macro_sensitivity": 0.5653277909563001, |
| "eval_macro_specificity": 0.9821237739107975, |
| "eval_precision": 0.7656722240785093, |
| "eval_precision_macro": 0.5495765557734748, |
| "eval_recall": 0.7645236250968241, |
| "eval_recall_macro": 0.5653277909563001, |
| "eval_runtime": 80.7211, |
| "eval_samples_per_second": 15.993, |
| "eval_steps_per_second": 2.007, |
| "eval_weighted_fpr": 0.021526695935419914, |
| "eval_weighted_sensitivity": 0.7645236250968241, |
| "eval_weighted_specificity": 0.9673329835651372, |
| "step": 2572 |
| }, |
| { |
| "epoch": 4.67, |
| "grad_norm": 41.18514633178711, |
| "learning_rate": 3.447382063245205e-05, |
| "loss": 0.4521, |
| "step": 3000 |
| }, |
| { |
| "epoch": 5.0, |
| "eval_accuracy": 0.7900852052672347, |
| "eval_f1_macro": 0.6116525431724917, |
| "eval_f1_micro": 0.7900852052672347, |
| "eval_f1_weighted": 0.7794301250682801, |
| "eval_loss": 1.0259578227996826, |
| "eval_macro_fpr": 0.019043445988564375, |
| "eval_macro_sensitivity": 0.6315170743424566, |
| "eval_macro_specificity": 0.984305159354804, |
| "eval_precision": 0.7882093327742286, |
| "eval_precision_macro": 0.6420940275483907, |
| "eval_recall": 0.7900852052672347, |
| "eval_recall_macro": 0.6315170743424566, |
| "eval_runtime": 56.9964, |
| "eval_samples_per_second": 22.651, |
| "eval_steps_per_second": 2.842, |
| "eval_weighted_fpr": 0.018624149542986735, |
| "eval_weighted_sensitivity": 0.7900852052672347, |
| "eval_weighted_specificity": 0.9744921850548264, |
| "step": 3215 |
| }, |
| { |
| "epoch": 5.44, |
| "grad_norm": 92.02596282958984, |
| "learning_rate": 3.1881804043545884e-05, |
| "loss": 0.3466, |
| "step": 3500 |
| }, |
| { |
| "epoch": 6.0, |
| "eval_accuracy": 0.7869868319132456, |
| "eval_f1_macro": 0.6118575336738351, |
| "eval_f1_micro": 0.7869868319132455, |
| "eval_f1_weighted": 0.7889880056842663, |
| "eval_loss": 1.0384949445724487, |
| "eval_macro_fpr": 0.019220806323432427, |
| "eval_macro_sensitivity": 0.633518783056355, |
| "eval_macro_specificity": 0.984312476194111, |
| "eval_precision": 0.8134105657387156, |
| "eval_precision_macro": 0.6722188161248251, |
| "eval_recall": 0.7869868319132456, |
| "eval_recall_macro": 0.633518783056355, |
| "eval_runtime": 58.1615, |
| "eval_samples_per_second": 22.197, |
| "eval_steps_per_second": 2.785, |
| "eval_weighted_fpr": 0.018966825298296436, |
| "eval_weighted_sensitivity": 0.7869868319132456, |
| "eval_weighted_specificity": 0.9777003109984174, |
| "step": 3858 |
| }, |
| { |
| "epoch": 6.22, |
| "grad_norm": 80.3629379272461, |
| "learning_rate": 2.9294971487817523e-05, |
| "loss": 0.3321, |
| "step": 4000 |
| }, |
| { |
| "epoch": 7.0, |
| "grad_norm": 0.3377046585083008, |
| "learning_rate": 2.6702954898911353e-05, |
| "loss": 0.2333, |
| "step": 4500 |
| }, |
| { |
| "epoch": 7.0, |
| "eval_accuracy": 0.8117738187451587, |
| "eval_f1_macro": 0.665659747836216, |
| "eval_f1_micro": 0.8117738187451587, |
| "eval_f1_weighted": 0.8057624075524522, |
| "eval_loss": 1.1464879512786865, |
| "eval_macro_fpr": 0.01693769060060091, |
| "eval_macro_sensitivity": 0.6784424177464591, |
| "eval_macro_specificity": 0.9857572234742, |
| "eval_precision": 0.8064435087918365, |
| "eval_precision_macro": 0.6794791733053616, |
| "eval_recall": 0.8117738187451587, |
| "eval_recall_macro": 0.6784424177464591, |
| "eval_runtime": 54.6308, |
| "eval_samples_per_second": 23.631, |
| "eval_steps_per_second": 2.965, |
| "eval_weighted_fpr": 0.016292323164599398, |
| "eval_weighted_sensitivity": 0.8117738187451587, |
| "eval_weighted_specificity": 0.9745845333678407, |
| "step": 4501 |
| }, |
| { |
| "epoch": 7.78, |
| "grad_norm": 0.19802448153495789, |
| "learning_rate": 2.4110938310005186e-05, |
| "loss": 0.1658, |
| "step": 5000 |
| }, |
| { |
| "epoch": 8.0, |
| "eval_accuracy": 0.814872192099148, |
| "eval_f1_macro": 0.7328612504365998, |
| "eval_f1_micro": 0.814872192099148, |
| "eval_f1_weighted": 0.8120704640541792, |
| "eval_loss": 1.241939902305603, |
| "eval_macro_fpr": 0.016489901852411035, |
| "eval_macro_sensitivity": 0.7440371076362254, |
| "eval_macro_specificity": 0.9861286253461167, |
| "eval_precision": 0.8148683401292235, |
| "eval_precision_macro": 0.7318970148814324, |
| "eval_recall": 0.814872192099148, |
| "eval_recall_macro": 0.7440371076362254, |
| "eval_runtime": 57.719, |
| "eval_samples_per_second": 22.367, |
| "eval_steps_per_second": 2.807, |
| "eval_weighted_fpr": 0.015968463954032203, |
| "eval_weighted_sensitivity": 0.814872192099148, |
| "eval_weighted_specificity": 0.9770571880926026, |
| "step": 5144 |
| }, |
| { |
| "epoch": 8.55, |
| "grad_norm": 0.0870741754770279, |
| "learning_rate": 2.151892172109902e-05, |
| "loss": 0.1597, |
| "step": 5500 |
| }, |
| { |
| "epoch": 9.0, |
| "eval_accuracy": 0.8179705654531371, |
| "eval_f1_macro": 0.7804377723463752, |
| "eval_f1_micro": 0.817970565453137, |
| "eval_f1_weighted": 0.819003836369197, |
| "eval_loss": 1.3440543413162231, |
| "eval_macro_fpr": 0.015977205363795557, |
| "eval_macro_sensitivity": 0.770217713596195, |
| "eval_macro_specificity": 0.9863181364411756, |
| "eval_precision": 0.8259393911103454, |
| "eval_precision_macro": 0.8314287989687705, |
| "eval_recall": 0.8179705654531371, |
| "eval_recall_macro": 0.770217713596195, |
| "eval_runtime": 90.353, |
| "eval_samples_per_second": 14.288, |
| "eval_steps_per_second": 1.793, |
| "eval_weighted_fpr": 0.01564684732671949, |
| "eval_weighted_sensitivity": 0.8179705654531371, |
| "eval_weighted_specificity": 0.9768014811644973, |
| "step": 5787 |
| }, |
| { |
| "epoch": 9.33, |
| "grad_norm": 0.013821585103869438, |
| "learning_rate": 1.8926905132192844e-05, |
| "loss": 0.11, |
| "step": 6000 |
| }, |
| { |
| "epoch": 10.0, |
| "eval_accuracy": 0.8024786986831913, |
| "eval_f1_macro": 0.7500128517100584, |
| "eval_f1_micro": 0.8024786986831914, |
| "eval_f1_weighted": 0.8064576225913597, |
| "eval_loss": 1.3504691123962402, |
| "eval_macro_fpr": 0.01779575845463305, |
| "eval_macro_sensitivity": 0.765413440395755, |
| "eval_macro_specificity": 0.9852391037313044, |
| "eval_precision": 0.8152310151046661, |
| "eval_precision_macro": 0.7519799672719661, |
| "eval_recall": 0.8024786986831913, |
| "eval_recall_macro": 0.765413440395755, |
| "eval_runtime": 55.2658, |
| "eval_samples_per_second": 23.36, |
| "eval_steps_per_second": 2.931, |
| "eval_weighted_fpr": 0.017277593332881633, |
| "eval_weighted_sensitivity": 0.8024786986831913, |
| "eval_weighted_specificity": 0.9761078572863756, |
| "step": 6430 |
| }, |
| { |
| "epoch": 10.11, |
| "grad_norm": 0.009372069500386715, |
| "learning_rate": 1.6334888543286677e-05, |
| "loss": 0.0741, |
| "step": 6500 |
| }, |
| { |
| "epoch": 10.89, |
| "grad_norm": 0.017231620848178864, |
| "learning_rate": 1.374805598755832e-05, |
| "loss": 0.0747, |
| "step": 7000 |
| }, |
| { |
| "epoch": 11.0, |
| "eval_accuracy": 0.8280402788536019, |
| "eval_f1_macro": 0.7850116201409046, |
| "eval_f1_micro": 0.8280402788536019, |
| "eval_f1_weighted": 0.8252613057748646, |
| "eval_loss": 1.3682292699813843, |
| "eval_macro_fpr": 0.015248972295162136, |
| "eval_macro_sensitivity": 0.7819796586438489, |
| "eval_macro_specificity": 0.9869417363825405, |
| "eval_precision": 0.8277410447478016, |
| "eval_precision_macro": 0.8096252898700513, |
| "eval_recall": 0.8280402788536019, |
| "eval_recall_macro": 0.7819796586438489, |
| "eval_runtime": 58.5549, |
| "eval_samples_per_second": 22.048, |
| "eval_steps_per_second": 2.767, |
| "eval_weighted_fpr": 0.014616802739004477, |
| "eval_weighted_sensitivity": 0.8280402788536019, |
| "eval_weighted_specificity": 0.976085766884506, |
| "step": 7073 |
| }, |
| { |
| "epoch": 11.66, |
| "grad_norm": 0.039045702666044235, |
| "learning_rate": 1.1156039398652151e-05, |
| "loss": 0.0519, |
| "step": 7500 |
| }, |
| { |
| "epoch": 12.0, |
| "eval_accuracy": 0.8164213787761425, |
| "eval_f1_macro": 0.7698779986816631, |
| "eval_f1_micro": 0.8164213787761425, |
| "eval_f1_weighted": 0.8170206182613167, |
| "eval_loss": 1.4436842203140259, |
| "eval_macro_fpr": 0.016423979995821146, |
| "eval_macro_sensitivity": 0.7750707038634387, |
| "eval_macro_specificity": 0.9861294812620465, |
| "eval_precision": 0.8188455079677682, |
| "eval_precision_macro": 0.7730875104850559, |
| "eval_recall": 0.8164213787761425, |
| "eval_recall_macro": 0.7750707038634387, |
| "eval_runtime": 56.1974, |
| "eval_samples_per_second": 22.973, |
| "eval_steps_per_second": 2.883, |
| "eval_weighted_fpr": 0.01580737677582872, |
| "eval_weighted_sensitivity": 0.8164213787761425, |
| "eval_weighted_specificity": 0.9755208401545548, |
| "step": 7716 |
| }, |
| { |
| "epoch": 12.44, |
| "grad_norm": 220.98574829101562, |
| "learning_rate": 8.564022809745984e-06, |
| "loss": 0.0324, |
| "step": 8000 |
| }, |
| { |
| "epoch": 13.0, |
| "eval_accuracy": 0.8086754453911696, |
| "eval_f1_macro": 0.7637730945937946, |
| "eval_f1_micro": 0.8086754453911695, |
| "eval_f1_weighted": 0.8096616104204358, |
| "eval_loss": 1.4511244297027588, |
| "eval_macro_fpr": 0.017099071788154886, |
| "eval_macro_sensitivity": 0.7801835333077355, |
| "eval_macro_specificity": 0.9857069816183608, |
| "eval_precision": 0.8126700320980054, |
| "eval_precision_macro": 0.755164725157429, |
| "eval_recall": 0.8086754453911696, |
| "eval_recall_macro": 0.7801835333077355, |
| "eval_runtime": 62.3649, |
| "eval_samples_per_second": 20.701, |
| "eval_steps_per_second": 2.598, |
| "eval_weighted_fpr": 0.016618448496265894, |
| "eval_weighted_sensitivity": 0.8086754453911696, |
| "eval_weighted_specificity": 0.9769292788842445, |
| "step": 8359 |
| }, |
| { |
| "epoch": 13.22, |
| "grad_norm": 0.02690161019563675, |
| "learning_rate": 5.9720062208398135e-06, |
| "loss": 0.0349, |
| "step": 8500 |
| }, |
| { |
| "epoch": 14.0, |
| "grad_norm": 0.014679946005344391, |
| "learning_rate": 3.379989631933644e-06, |
| "loss": 0.0184, |
| "step": 9000 |
| }, |
| { |
| "epoch": 14.0, |
| "eval_accuracy": 0.8140975987606507, |
| "eval_f1_macro": 0.7681382733031227, |
| "eval_f1_micro": 0.8140975987606507, |
| "eval_f1_weighted": 0.8158709627059858, |
| "eval_loss": 1.5005292892456055, |
| "eval_macro_fpr": 0.016486730283890944, |
| "eval_macro_sensitivity": 0.7847943200142441, |
| "eval_macro_specificity": 0.9861151984488189, |
| "eval_precision": 0.8196264811835474, |
| "eval_precision_macro": 0.7613127212329543, |
| "eval_recall": 0.8140975987606507, |
| "eval_recall_macro": 0.7847943200142441, |
| "eval_runtime": 55.9241, |
| "eval_samples_per_second": 23.085, |
| "eval_steps_per_second": 2.897, |
| "eval_weighted_fpr": 0.016049217600641968, |
| "eval_weighted_sensitivity": 0.8140975987606507, |
| "eval_weighted_specificity": 0.9776303779716344, |
| "step": 9002 |
| }, |
| { |
| "epoch": 14.77, |
| "grad_norm": 0.007979592308402061, |
| "learning_rate": 7.931570762052877e-07, |
| "loss": 0.0137, |
| "step": 9500 |
| }, |
| { |
| "epoch": 15.0, |
| "eval_accuracy": 0.8164213787761425, |
| "eval_f1_macro": 0.7664691580433438, |
| "eval_f1_micro": 0.8164213787761425, |
| "eval_f1_weighted": 0.8173112279420088, |
| "eval_loss": 1.4994738101959229, |
| "eval_macro_fpr": 0.01630779929417336, |
| "eval_macro_sensitivity": 0.777963226880043, |
| "eval_macro_specificity": 0.9862332623102597, |
| "eval_precision": 0.8193184364806528, |
| "eval_precision_macro": 0.7619970590670772, |
| "eval_recall": 0.8164213787761425, |
| "eval_recall_macro": 0.777963226880043, |
| "eval_runtime": 61.5289, |
| "eval_samples_per_second": 20.982, |
| "eval_steps_per_second": 2.633, |
| "eval_weighted_fpr": 0.01580737677582872, |
| "eval_weighted_sensitivity": 0.8164213787761425, |
| "eval_weighted_specificity": 0.9770775558777535, |
| "step": 9645 |
| } |
| ], |
| "logging_steps": 500, |
| "max_steps": 9645, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 15, |
| "save_steps": 500, |
| "total_flos": 2.028033591671808e+16, |
| "train_batch_size": 8, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
