| { | |
| "best_metric": 0.5430528375733855, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-1/checkpoint-1532", | |
| "epoch": 8.0, | |
| "eval_steps": 500, | |
| "global_step": 6128, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.4840396046638489, | |
| "learning_rate": 3.344893789847612e-05, | |
| "loss": 0.1172, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5401174168297456, | |
| "eval_f1": 0.20068027210884357, | |
| "eval_loss": 0.10136038810014725, | |
| "eval_precision": 0.7662337662337663, | |
| "eval_recall": 0.11545988258317025, | |
| "eval_runtime": 133.3642, | |
| "eval_samples_per_second": 7.663, | |
| "eval_steps_per_second": 1.92, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.713012456893921, | |
| "learning_rate": 2.8670518198693823e-05, | |
| "loss": 0.1055, | |
| "step": 1532 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5430528375733855, | |
| "eval_f1": 0.21247892074198985, | |
| "eval_loss": 0.10160499066114426, | |
| "eval_precision": 0.7682926829268293, | |
| "eval_recall": 0.1232876712328767, | |
| "eval_runtime": 132.4691, | |
| "eval_samples_per_second": 7.715, | |
| "eval_steps_per_second": 1.933, | |
| "step": 1532 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.46175065636634827, | |
| "learning_rate": 2.389209849891152e-05, | |
| "loss": 0.1039, | |
| "step": 2298 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5322896281800391, | |
| "eval_f1": 0.17301038062283738, | |
| "eval_loss": 0.10080397874116898, | |
| "eval_precision": 0.746268656716418, | |
| "eval_recall": 0.09784735812133072, | |
| "eval_runtime": 133.9804, | |
| "eval_samples_per_second": 7.628, | |
| "eval_steps_per_second": 1.911, | |
| "step": 2298 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.8960716128349304, | |
| "learning_rate": 1.9113678799129215e-05, | |
| "loss": 0.1038, | |
| "step": 3064 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.5362035225048923, | |
| "eval_f1": 0.18835616438356165, | |
| "eval_loss": 0.10024181753396988, | |
| "eval_precision": 0.7534246575342466, | |
| "eval_recall": 0.10763209393346379, | |
| "eval_runtime": 131.9665, | |
| "eval_samples_per_second": 7.744, | |
| "eval_steps_per_second": 1.94, | |
| "step": 3064 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.2918031215667725, | |
| "learning_rate": 1.4335259099346911e-05, | |
| "loss": 0.103, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.5401174168297456, | |
| "eval_f1": 0.20068027210884357, | |
| "eval_loss": 0.10036557912826538, | |
| "eval_precision": 0.7662337662337663, | |
| "eval_recall": 0.11545988258317025, | |
| "eval_runtime": 134.3935, | |
| "eval_samples_per_second": 7.605, | |
| "eval_steps_per_second": 1.905, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 1.0010789632797241, | |
| "learning_rate": 9.556839399564608e-06, | |
| "loss": 0.1025, | |
| "step": 4596 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.5420743639921722, | |
| "eval_f1": 0.2067796610169491, | |
| "eval_loss": 0.10058682411909103, | |
| "eval_precision": 0.7721518987341772, | |
| "eval_recall": 0.11937377690802348, | |
| "eval_runtime": 132.5868, | |
| "eval_samples_per_second": 7.708, | |
| "eval_steps_per_second": 1.931, | |
| "step": 4596 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.8870960474014282, | |
| "learning_rate": 4.778419699782304e-06, | |
| "loss": 0.1024, | |
| "step": 5362 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.541095890410959, | |
| "eval_f1": 0.20373514431239387, | |
| "eval_loss": 0.10063320398330688, | |
| "eval_precision": 0.7692307692307693, | |
| "eval_recall": 0.11741682974559686, | |
| "eval_runtime": 132.2356, | |
| "eval_samples_per_second": 7.729, | |
| "eval_steps_per_second": 1.936, | |
| "step": 5362 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.9253156185150146, | |
| "learning_rate": 0.0, | |
| "loss": 0.1023, | |
| "step": 6128 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.541095890410959, | |
| "eval_f1": 0.20373514431239387, | |
| "eval_loss": 0.10009610652923584, | |
| "eval_precision": 0.7692307692307693, | |
| "eval_recall": 0.11741682974559686, | |
| "eval_runtime": 131.3578, | |
| "eval_samples_per_second": 7.78, | |
| "eval_steps_per_second": 1.949, | |
| "step": 6128 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 6128, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 8, | |
| "save_steps": 500, | |
| "total_flos": 1885561578240.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.1218013407365569, | |
| "learning_rate": 3.822735759825843e-05, | |
| "num_train_epochs": 8, | |
| "temperature": 6 | |
| } | |
| } | |