| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 10560, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.2974101305007935, | |
| "learning_rate": 4.75e-05, | |
| "loss": 0.7682, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8662923581436948, | |
| "eval_f1": 0.18337078651685396, | |
| "eval_loss": 0.43944087624549866, | |
| "eval_precision": 0.40476190476190477, | |
| "eval_recall": 0.11853573503776874, | |
| "eval_runtime": 4.4831, | |
| "eval_samples_per_second": 208.562, | |
| "eval_steps_per_second": 3.346, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.8939604759216309, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.3466, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9326741528371261, | |
| "eval_f1": 0.6631853785900783, | |
| "eval_loss": 0.22170819342136383, | |
| "eval_precision": 0.602181128496918, | |
| "eval_recall": 0.7379430563625798, | |
| "eval_runtime": 4.5771, | |
| "eval_samples_per_second": 204.278, | |
| "eval_steps_per_second": 3.277, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.3257710933685303, | |
| "learning_rate": 4.25e-05, | |
| "loss": 0.2131, | |
| "step": 1584 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9428102951406141, | |
| "eval_f1": 0.7492870106300232, | |
| "eval_loss": 0.172767773270607, | |
| "eval_precision": 0.6764981273408239, | |
| "eval_recall": 0.8396281231841952, | |
| "eval_runtime": 4.55, | |
| "eval_samples_per_second": 205.495, | |
| "eval_steps_per_second": 3.297, | |
| "step": 1584 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.8121825456619263, | |
| "learning_rate": 4e-05, | |
| "loss": 0.1759, | |
| "step": 2112 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9515552022259763, | |
| "eval_f1": 0.7833023132145706, | |
| "eval_loss": 0.15085110068321228, | |
| "eval_precision": 0.7220588235294118, | |
| "eval_recall": 0.8558977338756537, | |
| "eval_runtime": 4.5248, | |
| "eval_samples_per_second": 206.638, | |
| "eval_steps_per_second": 3.315, | |
| "step": 2112 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.214239478111267, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.1563, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9533439332207095, | |
| "eval_f1": 0.7900773539610563, | |
| "eval_loss": 0.14215560257434845, | |
| "eval_precision": 0.730276134122288, | |
| "eval_recall": 0.8605461940732132, | |
| "eval_runtime": 4.5421, | |
| "eval_samples_per_second": 205.851, | |
| "eval_steps_per_second": 3.302, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 1.626373291015625, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.1464, | |
| "step": 3168 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9541389247739243, | |
| "eval_f1": 0.7889618922470434, | |
| "eval_loss": 0.14294518530368805, | |
| "eval_precision": 0.7202495201535508, | |
| "eval_recall": 0.8721673445671121, | |
| "eval_runtime": 4.5002, | |
| "eval_samples_per_second": 207.766, | |
| "eval_steps_per_second": 3.333, | |
| "step": 3168 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 3.158684015274048, | |
| "learning_rate": 3.2500000000000004e-05, | |
| "loss": 0.1394, | |
| "step": 3696 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9525489416674948, | |
| "eval_f1": 0.7869281045751634, | |
| "eval_loss": 0.14403149485588074, | |
| "eval_precision": 0.7153041825095057, | |
| "eval_recall": 0.8744915746658919, | |
| "eval_runtime": 4.5154, | |
| "eval_samples_per_second": 207.067, | |
| "eval_steps_per_second": 3.322, | |
| "step": 3696 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.7790516018867493, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1325, | |
| "step": 4224 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9553314121037464, | |
| "eval_f1": 0.796106287818995, | |
| "eval_loss": 0.1397974044084549, | |
| "eval_precision": 0.7274038461538461, | |
| "eval_recall": 0.8791400348634515, | |
| "eval_runtime": 4.5264, | |
| "eval_samples_per_second": 206.567, | |
| "eval_steps_per_second": 3.314, | |
| "step": 4224 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 1.278489589691162, | |
| "learning_rate": 2.7500000000000004e-05, | |
| "loss": 0.1269, | |
| "step": 4752 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9579151346516943, | |
| "eval_f1": 0.7998928475756765, | |
| "eval_loss": 0.13412921130657196, | |
| "eval_precision": 0.7420477137176938, | |
| "eval_recall": 0.8675188843695526, | |
| "eval_runtime": 4.5121, | |
| "eval_samples_per_second": 207.219, | |
| "eval_steps_per_second": 3.324, | |
| "step": 4752 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 1.4628835916519165, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.124, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.9564742124614926, | |
| "eval_f1": 0.8013807753584705, | |
| "eval_loss": 0.13306674361228943, | |
| "eval_precision": 0.7378973105134474, | |
| "eval_recall": 0.8768158047646717, | |
| "eval_runtime": 4.5165, | |
| "eval_samples_per_second": 207.02, | |
| "eval_steps_per_second": 3.321, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 2.4508931636810303, | |
| "learning_rate": 2.25e-05, | |
| "loss": 0.1194, | |
| "step": 5808 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.9568717082381, | |
| "eval_f1": 0.803921568627451, | |
| "eval_loss": 0.13287858664989471, | |
| "eval_precision": 0.7389186556259133, | |
| "eval_recall": 0.8814642649622313, | |
| "eval_runtime": 4.525, | |
| "eval_samples_per_second": 206.629, | |
| "eval_steps_per_second": 3.315, | |
| "step": 5808 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 1.7830716371536255, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1171, | |
| "step": 6336 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.9566729603497963, | |
| "eval_f1": 0.8026525198938992, | |
| "eval_loss": 0.13365061581134796, | |
| "eval_precision": 0.7384089799902391, | |
| "eval_recall": 0.8791400348634515, | |
| "eval_runtime": 4.5462, | |
| "eval_samples_per_second": 205.667, | |
| "eval_steps_per_second": 3.299, | |
| "step": 6336 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 2.1622517108917236, | |
| "learning_rate": 1.75e-05, | |
| "loss": 0.1153, | |
| "step": 6864 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.9586604392328332, | |
| "eval_f1": 0.8043826830571886, | |
| "eval_loss": 0.129431813955307, | |
| "eval_precision": 0.7446808510638298, | |
| "eval_recall": 0.8744915746658919, | |
| "eval_runtime": 4.5078, | |
| "eval_samples_per_second": 207.416, | |
| "eval_steps_per_second": 3.328, | |
| "step": 6864 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 1.1562857627868652, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.1119, | |
| "step": 7392 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.9572692040147074, | |
| "eval_f1": 0.807794981313401, | |
| "eval_loss": 0.13104024529457092, | |
| "eval_precision": 0.7471604938271605, | |
| "eval_recall": 0.8791400348634515, | |
| "eval_runtime": 4.5437, | |
| "eval_samples_per_second": 205.78, | |
| "eval_steps_per_second": 3.301, | |
| "step": 7392 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 1.6378480195999146, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.1109, | |
| "step": 7920 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.9578654476796185, | |
| "eval_f1": 0.803963577932512, | |
| "eval_loss": 0.1311807483434677, | |
| "eval_precision": 0.7456532538499752, | |
| "eval_recall": 0.8721673445671121, | |
| "eval_runtime": 4.5022, | |
| "eval_samples_per_second": 207.676, | |
| "eval_steps_per_second": 3.332, | |
| "step": 7920 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 2.3020808696746826, | |
| "learning_rate": 1e-05, | |
| "loss": 0.1102, | |
| "step": 8448 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.9580641955679221, | |
| "eval_f1": 0.8060735215769846, | |
| "eval_loss": 0.13092459738254547, | |
| "eval_precision": 0.7442203639940974, | |
| "eval_recall": 0.8791400348634515, | |
| "eval_runtime": 4.5251, | |
| "eval_samples_per_second": 206.624, | |
| "eval_steps_per_second": 3.315, | |
| "step": 8448 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 1.444940209388733, | |
| "learning_rate": 7.5e-06, | |
| "loss": 0.1095, | |
| "step": 8976 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.9586604392328332, | |
| "eval_f1": 0.8073443320915381, | |
| "eval_loss": 0.13140013813972473, | |
| "eval_precision": 0.7447226313205695, | |
| "eval_recall": 0.8814642649622313, | |
| "eval_runtime": 4.5218, | |
| "eval_samples_per_second": 206.775, | |
| "eval_steps_per_second": 3.317, | |
| "step": 8976 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 1.5871490240097046, | |
| "learning_rate": 5e-06, | |
| "loss": 0.1073, | |
| "step": 9504 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.9576666997913147, | |
| "eval_f1": 0.8018114011720832, | |
| "eval_loss": 0.1323314905166626, | |
| "eval_precision": 0.7402852926709297, | |
| "eval_recall": 0.8744915746658919, | |
| "eval_runtime": 4.5193, | |
| "eval_samples_per_second": 206.892, | |
| "eval_steps_per_second": 3.319, | |
| "step": 9504 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 0.673528254032135, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.107, | |
| "step": 10032 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.9580641955679221, | |
| "eval_f1": 0.800106894708712, | |
| "eval_loss": 0.1300082504749298, | |
| "eval_precision": 0.7407224146462148, | |
| "eval_recall": 0.8698431144683324, | |
| "eval_runtime": 4.5363, | |
| "eval_samples_per_second": 206.115, | |
| "eval_steps_per_second": 3.307, | |
| "step": 10032 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 2.0137035846710205, | |
| "learning_rate": 0.0, | |
| "loss": 0.1073, | |
| "step": 10560 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.9578654476796185, | |
| "eval_f1": 0.800106894708712, | |
| "eval_loss": 0.1304602026939392, | |
| "eval_precision": 0.7407224146462148, | |
| "eval_recall": 0.8698431144683324, | |
| "eval_runtime": 4.5054, | |
| "eval_samples_per_second": 207.53, | |
| "eval_steps_per_second": 3.329, | |
| "step": 10560 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 10560, | |
| "total_flos": 4545096690358590.0, | |
| "train_loss": 0.17226147796168473, | |
| "train_runtime": 1232.3096, | |
| "train_samples_per_second": 136.93, | |
| "train_steps_per_second": 8.569 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 10560, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 4545096690358590.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |