| { | |
| "best_metric": 0.8302752293577982, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-0/checkpoint-4216", | |
| "epoch": 8.0, | |
| "eval_steps": 500, | |
| "global_step": 4216, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 16.55626678466797, | |
| "learning_rate": 5.1383294230414005e-05, | |
| "loss": 1.5907, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8061926605504587, | |
| "eval_loss": 1.2464169263839722, | |
| "eval_runtime": 2.4007, | |
| "eval_samples_per_second": 363.232, | |
| "eval_steps_per_second": 2.916, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.405675241279466e-05, | |
| "loss": 0.9038, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8107798165137615, | |
| "eval_loss": 1.1235342025756836, | |
| "eval_runtime": 2.3533, | |
| "eval_samples_per_second": 370.54, | |
| "eval_steps_per_second": 2.975, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 48.55740737915039, | |
| "learning_rate": 3.67162818084498e-05, | |
| "loss": 0.6946, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8176605504587156, | |
| "eval_loss": 1.1027159690856934, | |
| "eval_runtime": 2.3719, | |
| "eval_samples_per_second": 367.637, | |
| "eval_steps_per_second": 2.951, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 16.267213821411133, | |
| "learning_rate": 2.9375811204104943e-05, | |
| "loss": 0.595, | |
| "step": 2108 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8256880733944955, | |
| "eval_loss": 1.0700539350509644, | |
| "eval_runtime": 2.3638, | |
| "eval_samples_per_second": 368.893, | |
| "eval_steps_per_second": 2.961, | |
| "step": 2108 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 10.482604026794434, | |
| "learning_rate": 2.2035340599760084e-05, | |
| "loss": 0.5308, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8291284403669725, | |
| "eval_loss": 1.1128482818603516, | |
| "eval_runtime": 2.3733, | |
| "eval_samples_per_second": 367.419, | |
| "eval_steps_per_second": 2.949, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 24.14480209350586, | |
| "learning_rate": 1.4694869995415228e-05, | |
| "loss": 0.4895, | |
| "step": 3162 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8291284403669725, | |
| "eval_loss": 1.1355042457580566, | |
| "eval_runtime": 2.3485, | |
| "eval_samples_per_second": 371.294, | |
| "eval_steps_per_second": 2.981, | |
| "step": 3162 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 25.623193740844727, | |
| "learning_rate": 7.35439939107037e-06, | |
| "loss": 0.4606, | |
| "step": 3689 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8268348623853211, | |
| "eval_loss": 1.1558915376663208, | |
| "eval_runtime": 2.3654, | |
| "eval_samples_per_second": 368.653, | |
| "eval_steps_per_second": 2.959, | |
| "step": 3689 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 11.750224113464355, | |
| "learning_rate": 1.3928786725512063e-08, | |
| "loss": 0.449, | |
| "step": 4216 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8302752293577982, | |
| "eval_loss": 1.155808687210083, | |
| "eval_runtime": 2.3688, | |
| "eval_samples_per_second": 368.125, | |
| "eval_steps_per_second": 2.955, | |
| "step": 4216 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4216, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 8, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 64687167744780.0, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.5167874928728581, | |
| "learning_rate": 5.872376483475886e-05, | |
| "num_train_epochs": 8, | |
| "temperature": 5 | |
| } | |
| } | |