| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 2120, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 3.486544132232666, | |
| "learning_rate": 4.75e-05, | |
| "loss": 0.3438, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9492119089316988, | |
| "eval_f1": 0.7840440165061899, | |
| "eval_loss": 0.16529129445552826, | |
| "eval_precision": 0.7345360824742269, | |
| "eval_recall": 0.8407079646017699, | |
| "eval_runtime": 0.8917, | |
| "eval_samples_per_second": 209.705, | |
| "eval_steps_per_second": 3.364, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 2.105318307876587, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.1133, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9569677257943457, | |
| "eval_f1": 0.8272108843537416, | |
| "eval_loss": 0.13404884934425354, | |
| "eval_precision": 0.7676767676767676, | |
| "eval_recall": 0.8967551622418879, | |
| "eval_runtime": 0.9428, | |
| "eval_samples_per_second": 198.343, | |
| "eval_steps_per_second": 3.182, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 2.1835291385650635, | |
| "learning_rate": 4.25e-05, | |
| "loss": 0.0736, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9589692269201902, | |
| "eval_f1": 0.8222222222222222, | |
| "eval_loss": 0.1448056399822235, | |
| "eval_precision": 0.7769028871391076, | |
| "eval_recall": 0.8731563421828908, | |
| "eval_runtime": 0.893, | |
| "eval_samples_per_second": 209.417, | |
| "eval_steps_per_second": 3.36, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 4.041090965270996, | |
| "learning_rate": 4e-05, | |
| "loss": 0.0473, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.961220915686765, | |
| "eval_f1": 0.8274894810659187, | |
| "eval_loss": 0.158503457903862, | |
| "eval_precision": 0.7887700534759359, | |
| "eval_recall": 0.8702064896755162, | |
| "eval_runtime": 0.9078, | |
| "eval_samples_per_second": 206.001, | |
| "eval_steps_per_second": 3.305, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 2.1877193450927734, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.0311, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9604703527645734, | |
| "eval_f1": 0.8344923504867873, | |
| "eval_loss": 0.18446023762226105, | |
| "eval_precision": 0.7894736842105263, | |
| "eval_recall": 0.8849557522123894, | |
| "eval_runtime": 0.911, | |
| "eval_samples_per_second": 205.279, | |
| "eval_steps_per_second": 3.293, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 2.7992465496063232, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.0179, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9602201651238429, | |
| "eval_f1": 0.8263305322128851, | |
| "eval_loss": 0.21451354026794434, | |
| "eval_precision": 0.7866666666666666, | |
| "eval_recall": 0.8702064896755162, | |
| "eval_runtime": 0.9051, | |
| "eval_samples_per_second": 206.598, | |
| "eval_steps_per_second": 3.314, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.3002428412437439, | |
| "learning_rate": 3.2500000000000004e-05, | |
| "loss": 0.0126, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9567175381536153, | |
| "eval_f1": 0.82336578581363, | |
| "eval_loss": 0.22246094048023224, | |
| "eval_precision": 0.7789473684210526, | |
| "eval_recall": 0.8731563421828908, | |
| "eval_runtime": 0.9044, | |
| "eval_samples_per_second": 206.757, | |
| "eval_steps_per_second": 3.317, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.2764192223548889, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0091, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9582186639979985, | |
| "eval_f1": 0.8326417704011065, | |
| "eval_loss": 0.2556192874908447, | |
| "eval_precision": 0.7838541666666666, | |
| "eval_recall": 0.887905604719764, | |
| "eval_runtime": 0.9143, | |
| "eval_samples_per_second": 204.535, | |
| "eval_steps_per_second": 3.281, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.3533385396003723, | |
| "learning_rate": 2.7500000000000004e-05, | |
| "loss": 0.0041, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9609707280460346, | |
| "eval_f1": 0.8321579689703806, | |
| "eval_loss": 0.2573556900024414, | |
| "eval_precision": 0.7972972972972973, | |
| "eval_recall": 0.8702064896755162, | |
| "eval_runtime": 0.9137, | |
| "eval_samples_per_second": 204.666, | |
| "eval_steps_per_second": 3.283, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 3.1199791431427, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0036, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.9554665999499625, | |
| "eval_f1": 0.817174515235457, | |
| "eval_loss": 0.3124300241470337, | |
| "eval_precision": 0.7702349869451697, | |
| "eval_recall": 0.8702064896755162, | |
| "eval_runtime": 0.9012, | |
| "eval_samples_per_second": 207.51, | |
| "eval_steps_per_second": 3.329, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.014265856705605984, | |
| "learning_rate": 2.25e-05, | |
| "loss": 0.0038, | |
| "step": 1166 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.960720540405304, | |
| "eval_f1": 0.8324022346368715, | |
| "eval_loss": 0.2836814224720001, | |
| "eval_precision": 0.7904509283819628, | |
| "eval_recall": 0.8790560471976401, | |
| "eval_runtime": 0.916, | |
| "eval_samples_per_second": 204.138, | |
| "eval_steps_per_second": 3.275, | |
| "step": 1166 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 1.1654541492462158, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0017, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.9574681010758068, | |
| "eval_f1": 0.825, | |
| "eval_loss": 0.3034752905368805, | |
| "eval_precision": 0.7795275590551181, | |
| "eval_recall": 0.8761061946902655, | |
| "eval_runtime": 0.91, | |
| "eval_samples_per_second": 205.506, | |
| "eval_steps_per_second": 3.297, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 0.017983168363571167, | |
| "learning_rate": 1.75e-05, | |
| "loss": 0.0015, | |
| "step": 1378 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.9604703527645734, | |
| "eval_f1": 0.8333333333333335, | |
| "eval_loss": 0.3068053126335144, | |
| "eval_precision": 0.7874015748031497, | |
| "eval_recall": 0.8849557522123894, | |
| "eval_runtime": 0.902, | |
| "eval_samples_per_second": 207.324, | |
| "eval_steps_per_second": 3.326, | |
| "step": 1378 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 0.011942153796553612, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.0012, | |
| "step": 1484 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.9577182887165374, | |
| "eval_f1": 0.8326417704011065, | |
| "eval_loss": 0.32863807678222656, | |
| "eval_precision": 0.7838541666666666, | |
| "eval_recall": 0.887905604719764, | |
| "eval_runtime": 0.8986, | |
| "eval_samples_per_second": 208.111, | |
| "eval_steps_per_second": 3.339, | |
| "step": 1484 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 0.0062804995104670525, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.0006, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.960720540405304, | |
| "eval_f1": 0.840782122905028, | |
| "eval_loss": 0.3137037754058838, | |
| "eval_precision": 0.7984084880636605, | |
| "eval_recall": 0.887905604719764, | |
| "eval_runtime": 0.9082, | |
| "eval_samples_per_second": 205.909, | |
| "eval_steps_per_second": 3.303, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.2206662893295288, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0008, | |
| "step": 1696 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.9617212909682261, | |
| "eval_f1": 0.8382559774964837, | |
| "eval_loss": 0.3065112233161926, | |
| "eval_precision": 0.8010752688172043, | |
| "eval_recall": 0.8790560471976401, | |
| "eval_runtime": 0.9053, | |
| "eval_samples_per_second": 206.56, | |
| "eval_steps_per_second": 3.314, | |
| "step": 1696 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 0.00254653743468225, | |
| "learning_rate": 7.5e-06, | |
| "loss": 0.0014, | |
| "step": 1802 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.9589692269201902, | |
| "eval_f1": 0.8365650969529086, | |
| "eval_loss": 0.33052197098731995, | |
| "eval_precision": 0.7885117493472585, | |
| "eval_recall": 0.8908554572271387, | |
| "eval_runtime": 0.9027, | |
| "eval_samples_per_second": 207.161, | |
| "eval_steps_per_second": 3.323, | |
| "step": 1802 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 0.0013508679112419486, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0005, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.9597197898423818, | |
| "eval_f1": 0.8344923504867873, | |
| "eval_loss": 0.3244776129722595, | |
| "eval_precision": 0.7894736842105263, | |
| "eval_recall": 0.8849557522123894, | |
| "eval_runtime": 0.8996, | |
| "eval_samples_per_second": 207.879, | |
| "eval_steps_per_second": 3.335, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 0.0013845202047377825, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.0004, | |
| "step": 2014 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.9602201651238429, | |
| "eval_f1": 0.8372739916550764, | |
| "eval_loss": 0.32481154799461365, | |
| "eval_precision": 0.7921052631578948, | |
| "eval_recall": 0.887905604719764, | |
| "eval_runtime": 0.9114, | |
| "eval_samples_per_second": 205.168, | |
| "eval_steps_per_second": 3.291, | |
| "step": 2014 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 0.003232144983485341, | |
| "learning_rate": 0.0, | |
| "loss": 0.0003, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.9604703527645734, | |
| "eval_f1": 0.8384401114206128, | |
| "eval_loss": 0.3243328630924225, | |
| "eval_precision": 0.7941952506596306, | |
| "eval_recall": 0.887905604719764, | |
| "eval_runtime": 0.9046, | |
| "eval_samples_per_second": 206.727, | |
| "eval_steps_per_second": 3.316, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 2120, | |
| "total_flos": 898035701398080.0, | |
| "train_loss": 0.03343723254489168, | |
| "train_runtime": 509.8629, | |
| "train_samples_per_second": 66.175, | |
| "train_steps_per_second": 4.158 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2120, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 898035701398080.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |