| { | |
| "best_metric": 0.4735751152038574, | |
| "best_model_checkpoint": "./vit-lr-cosine-warmup/checkpoint-963", | |
| "epoch": 13.0, | |
| "eval_steps": 500, | |
| "global_step": 4173, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 7.814958572387695, | |
| "learning_rate": 4.1428571428571437e-05, | |
| "loss": 0.86, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8141470180305131, | |
| "eval_f1": 0.8010598763076213, | |
| "eval_loss": 0.5250416994094849, | |
| "eval_precision": 0.8100096575743447, | |
| "eval_recall": 0.8141470180305131, | |
| "eval_runtime": 36.0914, | |
| "eval_samples_per_second": 79.908, | |
| "eval_steps_per_second": 10.002, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 6.845722198486328, | |
| "learning_rate": 8.311688311688312e-05, | |
| "loss": 0.4517, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8221220527045769, | |
| "eval_f1": 0.8099830282273331, | |
| "eval_loss": 0.5117006897926331, | |
| "eval_precision": 0.8347375649374938, | |
| "eval_recall": 0.8221220527045769, | |
| "eval_runtime": 37.2073, | |
| "eval_samples_per_second": 77.512, | |
| "eval_steps_per_second": 9.702, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 2.5204238891601562, | |
| "learning_rate": 9.985334621908699e-05, | |
| "loss": 0.3512, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8394590846047156, | |
| "eval_f1": 0.830758624078281, | |
| "eval_loss": 0.4735751152038574, | |
| "eval_precision": 0.8318317467279469, | |
| "eval_recall": 0.8394590846047156, | |
| "eval_runtime": 35.8524, | |
| "eval_samples_per_second": 80.441, | |
| "eval_steps_per_second": 10.069, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 5.594597816467285, | |
| "learning_rate": 9.894936461151184e-05, | |
| "loss": 0.2184, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8567961165048543, | |
| "eval_f1": 0.85051019948, | |
| "eval_loss": 0.4796653389930725, | |
| "eval_precision": 0.8536361493542505, | |
| "eval_recall": 0.8567961165048543, | |
| "eval_runtime": 35.3028, | |
| "eval_samples_per_second": 81.693, | |
| "eval_steps_per_second": 10.226, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 9.165299415588379, | |
| "learning_rate": 9.723506398349735e-05, | |
| "loss": 0.1264, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8547156726768377, | |
| "eval_f1": 0.8530254035056796, | |
| "eval_loss": 0.6211732029914856, | |
| "eval_precision": 0.8551837556766221, | |
| "eval_recall": 0.8547156726768377, | |
| "eval_runtime": 36.6331, | |
| "eval_samples_per_second": 78.727, | |
| "eval_steps_per_second": 9.854, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.9397739768028259, | |
| "learning_rate": 9.473882326123909e-05, | |
| "loss": 0.0687, | |
| "step": 1926 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8463938973647711, | |
| "eval_f1": 0.840249522586874, | |
| "eval_loss": 0.7659199237823486, | |
| "eval_precision": 0.8475689441425499, | |
| "eval_recall": 0.8463938973647711, | |
| "eval_runtime": 35.8316, | |
| "eval_samples_per_second": 80.488, | |
| "eval_steps_per_second": 10.075, | |
| "step": 1926 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 11.412993431091309, | |
| "learning_rate": 9.15019657867844e-05, | |
| "loss": 0.0463, | |
| "step": 2247 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8519417475728155, | |
| "eval_f1": 0.84690540461018, | |
| "eval_loss": 0.8237490057945251, | |
| "eval_precision": 0.8546320390871954, | |
| "eval_recall": 0.8519417475728155, | |
| "eval_runtime": 36.3237, | |
| "eval_samples_per_second": 79.397, | |
| "eval_steps_per_second": 9.938, | |
| "step": 2247 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.15119314193725586, | |
| "learning_rate": 8.759130166350091e-05, | |
| "loss": 0.0373, | |
| "step": 2568 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8377253814147018, | |
| "eval_f1": 0.8414971604167042, | |
| "eval_loss": 0.871150553226471, | |
| "eval_precision": 0.8492780112281874, | |
| "eval_recall": 0.8377253814147018, | |
| "eval_runtime": 37.0971, | |
| "eval_samples_per_second": 77.742, | |
| "eval_steps_per_second": 9.731, | |
| "step": 2568 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.9838098883628845, | |
| "learning_rate": 8.304716115113689e-05, | |
| "loss": 0.0347, | |
| "step": 2889 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.8567961165048543, | |
| "eval_f1": 0.8533749018674412, | |
| "eval_loss": 0.8180708885192871, | |
| "eval_precision": 0.8549859977362129, | |
| "eval_recall": 0.8567961165048543, | |
| "eval_runtime": 36.43, | |
| "eval_samples_per_second": 79.166, | |
| "eval_steps_per_second": 9.909, | |
| "step": 2889 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 2.4911880493164062, | |
| "learning_rate": 7.795595034552552e-05, | |
| "loss": 0.0263, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.8446601941747572, | |
| "eval_f1": 0.8308350673322552, | |
| "eval_loss": 1.0705382823944092, | |
| "eval_precision": 0.8388632159592988, | |
| "eval_recall": 0.8446601941747572, | |
| "eval_runtime": 36.2711, | |
| "eval_samples_per_second": 79.512, | |
| "eval_steps_per_second": 9.953, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.003689270233735442, | |
| "learning_rate": 7.240195031927308e-05, | |
| "loss": 0.0289, | |
| "step": 3531 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.858876560332871, | |
| "eval_f1": 0.855018220816544, | |
| "eval_loss": 0.9376017451286316, | |
| "eval_precision": 0.8605983316828895, | |
| "eval_recall": 0.858876560332871, | |
| "eval_runtime": 36.7783, | |
| "eval_samples_per_second": 78.416, | |
| "eval_steps_per_second": 9.816, | |
| "step": 3531 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.08966358751058578, | |
| "learning_rate": 6.647710326399964e-05, | |
| "loss": 0.0164, | |
| "step": 3852 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.863384188626907, | |
| "eval_f1": 0.8610588511525862, | |
| "eval_loss": 0.9714025259017944, | |
| "eval_precision": 0.8611342448885915, | |
| "eval_recall": 0.863384188626907, | |
| "eval_runtime": 36.4831, | |
| "eval_samples_per_second": 79.05, | |
| "eval_steps_per_second": 9.895, | |
| "step": 3852 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 0.05049363151192665, | |
| "learning_rate": 6.027949045818934e-05, | |
| "loss": 0.0077, | |
| "step": 4173 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.8398058252427184, | |
| "eval_f1": 0.8242916893123671, | |
| "eval_loss": 1.2992373704910278, | |
| "eval_precision": 0.8395816522197255, | |
| "eval_recall": 0.8398058252427184, | |
| "eval_runtime": 37.0798, | |
| "eval_samples_per_second": 77.778, | |
| "eval_steps_per_second": 9.736, | |
| "step": 4173 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "step": 4173, | |
| "total_flos": 5.166157498470679e+18, | |
| "train_loss": 0.1749291451406399, | |
| "train_runtime": 1863.5469, | |
| "train_samples_per_second": 275.174, | |
| "train_steps_per_second": 17.225 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 32100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 100, | |
| "save_steps": 500, | |
| "total_flos": 5.166157498470679e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |