| { | |
| "best_metric": 0.8307240704500979, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-38/checkpoint-672", | |
| "epoch": 8.0, | |
| "eval_steps": 500, | |
| "global_step": 768, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.8633686304092407, | |
| "learning_rate": 0.00011164310281576651, | |
| "loss": 0.584, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7690802348336595, | |
| "eval_f1": 0.7993197278911565, | |
| "eval_loss": 0.48832422494888306, | |
| "eval_precision": 0.706766917293233, | |
| "eval_recall": 0.9197651663405088, | |
| "eval_runtime": 28.3969, | |
| "eval_samples_per_second": 35.99, | |
| "eval_steps_per_second": 1.127, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.9557557106018066, | |
| "learning_rate": 9.76877149637957e-05, | |
| "loss": 0.4827, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8003913894324853, | |
| "eval_f1": 0.8222996515679443, | |
| "eval_loss": 0.44451940059661865, | |
| "eval_precision": 0.7409733124018838, | |
| "eval_recall": 0.923679060665362, | |
| "eval_runtime": 28.5294, | |
| "eval_samples_per_second": 35.823, | |
| "eval_steps_per_second": 1.122, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 3.0629470348358154, | |
| "learning_rate": 8.373232711182488e-05, | |
| "loss": 0.448, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.824853228962818, | |
| "eval_f1": 0.8353265869365226, | |
| "eval_loss": 0.4216720759868622, | |
| "eval_precision": 0.7881944444444444, | |
| "eval_recall": 0.8884540117416829, | |
| "eval_runtime": 28.7944, | |
| "eval_samples_per_second": 35.493, | |
| "eval_steps_per_second": 1.111, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 5.688210964202881, | |
| "learning_rate": 6.977693925985407e-05, | |
| "loss": 0.4341, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8209393346379648, | |
| "eval_f1": 0.825214899713467, | |
| "eval_loss": 0.42298367619514465, | |
| "eval_precision": 0.8059701492537313, | |
| "eval_recall": 0.8454011741682974, | |
| "eval_runtime": 28.2399, | |
| "eval_samples_per_second": 36.19, | |
| "eval_steps_per_second": 1.133, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 2.334526300430298, | |
| "learning_rate": 5.5821551407883254e-05, | |
| "loss": 0.4202, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8287671232876712, | |
| "eval_f1": 0.8436103663985702, | |
| "eval_loss": 0.4076910614967346, | |
| "eval_precision": 0.7763157894736842, | |
| "eval_recall": 0.923679060665362, | |
| "eval_runtime": 28.534, | |
| "eval_samples_per_second": 35.817, | |
| "eval_steps_per_second": 1.121, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 4.195997714996338, | |
| "learning_rate": 4.186616355591244e-05, | |
| "loss": 0.4109, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8258317025440313, | |
| "eval_f1": 0.842756183745583, | |
| "eval_loss": 0.4115942418575287, | |
| "eval_precision": 0.7681159420289855, | |
| "eval_recall": 0.9334637964774951, | |
| "eval_runtime": 28.3073, | |
| "eval_samples_per_second": 36.104, | |
| "eval_steps_per_second": 1.13, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 2.727288007736206, | |
| "learning_rate": 2.7910775703941627e-05, | |
| "loss": 0.4017, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8307240704500979, | |
| "eval_f1": 0.8467670504871568, | |
| "eval_loss": 0.41019657254219055, | |
| "eval_precision": 0.7734627831715211, | |
| "eval_recall": 0.9354207436399217, | |
| "eval_runtime": 28.5087, | |
| "eval_samples_per_second": 35.849, | |
| "eval_steps_per_second": 1.122, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 3.721353530883789, | |
| "learning_rate": 1.3955387851970814e-05, | |
| "loss": 0.4014, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8258317025440313, | |
| "eval_f1": 0.8405017921146953, | |
| "eval_loss": 0.40405774116516113, | |
| "eval_precision": 0.775206611570248, | |
| "eval_recall": 0.9178082191780822, | |
| "eval_runtime": 28.5925, | |
| "eval_samples_per_second": 35.744, | |
| "eval_steps_per_second": 1.119, | |
| "step": 768 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 864, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9, | |
| "save_steps": 500, | |
| "total_flos": 1885561578240.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.8898885181069172, | |
| "learning_rate": 0.00012559849066773733, | |
| "num_train_epochs": 9, | |
| "temperature": 3 | |
| } | |
| } | |