| { | |
| "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": 1.7431385517120361, | |
| "learning_rate": 4.75e-05, | |
| "loss": 1.2665, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8448836627470603, | |
| "eval_f1": 0.0, | |
| "eval_loss": 0.7136898040771484, | |
| "eval_precision": 0.0, | |
| "eval_recall": 0.0, | |
| "eval_runtime": 0.9087, | |
| "eval_samples_per_second": 205.786, | |
| "eval_steps_per_second": 3.301, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.2968225479125977, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.713, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8451338503877909, | |
| "eval_f1": 0.0, | |
| "eval_loss": 0.6074723601341248, | |
| "eval_precision": 0.0, | |
| "eval_recall": 0.0, | |
| "eval_runtime": 0.9286, | |
| "eval_samples_per_second": 201.374, | |
| "eval_steps_per_second": 3.231, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.28258216381073, | |
| "learning_rate": 4.25e-05, | |
| "loss": 0.6346, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8493870402802102, | |
| "eval_f1": 0.022222222222222223, | |
| "eval_loss": 0.5231208801269531, | |
| "eval_precision": 0.19047619047619047, | |
| "eval_recall": 0.011799410029498525, | |
| "eval_runtime": 0.957, | |
| "eval_samples_per_second": 195.406, | |
| "eval_steps_per_second": 3.135, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.9985119104385376, | |
| "learning_rate": 4e-05, | |
| "loss": 0.5555, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8656492369276958, | |
| "eval_f1": 0.10501193317422433, | |
| "eval_loss": 0.44577664136886597, | |
| "eval_precision": 0.275, | |
| "eval_recall": 0.06489675516224189, | |
| "eval_runtime": 0.9294, | |
| "eval_samples_per_second": 201.195, | |
| "eval_steps_per_second": 3.228, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.3407347202301025, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.4696, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8949211908931699, | |
| "eval_f1": 0.35859519408502777, | |
| "eval_loss": 0.37153393030166626, | |
| "eval_precision": 0.4801980198019802, | |
| "eval_recall": 0.2861356932153392, | |
| "eval_runtime": 0.9105, | |
| "eval_samples_per_second": 205.39, | |
| "eval_steps_per_second": 3.295, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 1.2348569631576538, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.3932, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9194395796847635, | |
| "eval_f1": 0.5402124430955993, | |
| "eval_loss": 0.3133937120437622, | |
| "eval_precision": 0.55625, | |
| "eval_recall": 0.5250737463126843, | |
| "eval_runtime": 0.8947, | |
| "eval_samples_per_second": 209.001, | |
| "eval_steps_per_second": 3.353, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.8620750904083252, | |
| "learning_rate": 3.2500000000000004e-05, | |
| "loss": 0.3299, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9276957718288716, | |
| "eval_f1": 0.6284916201117318, | |
| "eval_loss": 0.2705799341201782, | |
| "eval_precision": 0.596816976127321, | |
| "eval_recall": 0.6637168141592921, | |
| "eval_runtime": 0.8955, | |
| "eval_samples_per_second": 208.832, | |
| "eval_steps_per_second": 3.35, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 1.5275648832321167, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2896, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9339504628471353, | |
| "eval_f1": 0.6711772665764547, | |
| "eval_loss": 0.24331486225128174, | |
| "eval_precision": 0.62, | |
| "eval_recall": 0.7315634218289085, | |
| "eval_runtime": 0.9165, | |
| "eval_samples_per_second": 204.036, | |
| "eval_steps_per_second": 3.273, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 1.4570423364639282, | |
| "learning_rate": 2.7500000000000004e-05, | |
| "loss": 0.2656, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9354515886915187, | |
| "eval_f1": 0.6923076923076923, | |
| "eval_loss": 0.22765140235424042, | |
| "eval_precision": 0.6289156626506024, | |
| "eval_recall": 0.7699115044247787, | |
| "eval_runtime": 0.9292, | |
| "eval_samples_per_second": 201.246, | |
| "eval_steps_per_second": 3.229, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 2.4647762775421143, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.2442, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.9387040280210157, | |
| "eval_f1": 0.7088948787061994, | |
| "eval_loss": 0.20824101567268372, | |
| "eval_precision": 0.652605459057072, | |
| "eval_recall": 0.775811209439528, | |
| "eval_runtime": 0.9229, | |
| "eval_samples_per_second": 202.614, | |
| "eval_steps_per_second": 3.25, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 1.644608974456787, | |
| "learning_rate": 2.25e-05, | |
| "loss": 0.23, | |
| "step": 1166 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.9382036527395546, | |
| "eval_f1": 0.7078947368421052, | |
| "eval_loss": 0.2019660323858261, | |
| "eval_precision": 0.6389548693586699, | |
| "eval_recall": 0.7935103244837758, | |
| "eval_runtime": 0.908, | |
| "eval_samples_per_second": 205.937, | |
| "eval_steps_per_second": 3.304, | |
| "step": 1166 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 2.0320627689361572, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2229, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.9384538403802852, | |
| "eval_f1": 0.7220026350461134, | |
| "eval_loss": 0.19768132269382477, | |
| "eval_precision": 0.6523809523809524, | |
| "eval_recall": 0.8082595870206489, | |
| "eval_runtime": 0.9098, | |
| "eval_samples_per_second": 205.544, | |
| "eval_steps_per_second": 3.297, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 1.1117591857910156, | |
| "learning_rate": 1.75e-05, | |
| "loss": 0.2132, | |
| "step": 1378 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.940205153865399, | |
| "eval_f1": 0.7267904509283818, | |
| "eval_loss": 0.18858253955841064, | |
| "eval_precision": 0.6602409638554216, | |
| "eval_recall": 0.8082595870206489, | |
| "eval_runtime": 0.9058, | |
| "eval_samples_per_second": 206.445, | |
| "eval_steps_per_second": 3.312, | |
| "step": 1378 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 1.4440829753875732, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.2055, | |
| "step": 1484 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.9414560920690518, | |
| "eval_f1": 0.7294751009421264, | |
| "eval_loss": 0.18096885085105896, | |
| "eval_precision": 0.6707920792079208, | |
| "eval_recall": 0.799410029498525, | |
| "eval_runtime": 0.9231, | |
| "eval_samples_per_second": 202.575, | |
| "eval_steps_per_second": 3.25, | |
| "step": 1484 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 0.746239423751831, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.2038, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.9404553415061296, | |
| "eval_f1": 0.7275132275132277, | |
| "eval_loss": 0.18217705190181732, | |
| "eval_precision": 0.6594724220623501, | |
| "eval_recall": 0.8112094395280236, | |
| "eval_runtime": 0.9121, | |
| "eval_samples_per_second": 205.015, | |
| "eval_steps_per_second": 3.289, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 1.0166144371032715, | |
| "learning_rate": 1e-05, | |
| "loss": 0.2004, | |
| "step": 1696 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.9429572179134351, | |
| "eval_f1": 0.7393617021276596, | |
| "eval_loss": 0.17875301837921143, | |
| "eval_precision": 0.6731234866828087, | |
| "eval_recall": 0.8200589970501475, | |
| "eval_runtime": 0.9135, | |
| "eval_samples_per_second": 204.697, | |
| "eval_steps_per_second": 3.284, | |
| "step": 1696 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 1.0570679903030396, | |
| "learning_rate": 7.5e-06, | |
| "loss": 0.1966, | |
| "step": 1802 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.9432074055541656, | |
| "eval_f1": 0.7417218543046358, | |
| "eval_loss": 0.1774715930223465, | |
| "eval_precision": 0.6730769230769231, | |
| "eval_recall": 0.8259587020648967, | |
| "eval_runtime": 0.9125, | |
| "eval_samples_per_second": 204.93, | |
| "eval_steps_per_second": 3.288, | |
| "step": 1802 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 1.0107321739196777, | |
| "learning_rate": 5e-06, | |
| "loss": 0.1931, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.9434575931948962, | |
| "eval_f1": 0.7387862796833773, | |
| "eval_loss": 0.17654407024383545, | |
| "eval_precision": 0.6682577565632458, | |
| "eval_recall": 0.8259587020648967, | |
| "eval_runtime": 0.9156, | |
| "eval_samples_per_second": 204.236, | |
| "eval_steps_per_second": 3.277, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 1.0605403184890747, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.1937, | |
| "step": 2014 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.9437077808356267, | |
| "eval_f1": 0.7427055702917772, | |
| "eval_loss": 0.17490608990192413, | |
| "eval_precision": 0.6746987951807228, | |
| "eval_recall": 0.8259587020648967, | |
| "eval_runtime": 0.9207, | |
| "eval_samples_per_second": 203.097, | |
| "eval_steps_per_second": 3.258, | |
| "step": 2014 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 1.278860092163086, | |
| "learning_rate": 0.0, | |
| "loss": 0.1888, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.9444583437578183, | |
| "eval_f1": 0.7483355525965381, | |
| "eval_loss": 0.17425791919231415, | |
| "eval_precision": 0.6820388349514563, | |
| "eval_recall": 0.8289085545722714, | |
| "eval_runtime": 0.9189, | |
| "eval_samples_per_second": 203.506, | |
| "eval_steps_per_second": 3.265, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 2120, | |
| "total_flos": 901149122771520.0, | |
| "train_loss": 0.36049160057643675, | |
| "train_runtime": 244.1252, | |
| "train_samples_per_second": 138.208, | |
| "train_steps_per_second": 8.684 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2120, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 901149122771520.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |