| { | |
| "best_metric": 0.3611699044704437, | |
| "best_model_checkpoint": "bert_base_lda_20_v1_qqp/checkpoint-4266", | |
| "epoch": 8.0, | |
| "eval_steps": 500, | |
| "global_step": 11376, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.913527011871338, | |
| "learning_rate": 4.9e-05, | |
| "loss": 0.4706, | |
| "step": 1422 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7975760573831313, | |
| "eval_combined_score": 0.7412216058960324, | |
| "eval_f1": 0.6848671544089334, | |
| "eval_loss": 0.41439592838287354, | |
| "eval_runtime": 26.0568, | |
| "eval_samples_per_second": 1551.613, | |
| "eval_steps_per_second": 6.064, | |
| "step": 1422 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 2.9698476791381836, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.3635, | |
| "step": 2844 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8274301261439525, | |
| "eval_combined_score": 0.8024355964223224, | |
| "eval_f1": 0.7774410667006922, | |
| "eval_loss": 0.3808220326900482, | |
| "eval_runtime": 26.0976, | |
| "eval_samples_per_second": 1549.186, | |
| "eval_steps_per_second": 6.054, | |
| "step": 2844 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 3.8180556297302246, | |
| "learning_rate": 4.7e-05, | |
| "loss": 0.2981, | |
| "step": 4266 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.840563937670047, | |
| "eval_combined_score": 0.8157680574270204, | |
| "eval_f1": 0.7909721771839938, | |
| "eval_loss": 0.3611699044704437, | |
| "eval_runtime": 26.154, | |
| "eval_samples_per_second": 1545.844, | |
| "eval_steps_per_second": 6.041, | |
| "step": 4266 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 3.3878448009490967, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.2419, | |
| "step": 5688 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8491219391540935, | |
| "eval_combined_score": 0.8199943211952585, | |
| "eval_f1": 0.7908667032364235, | |
| "eval_loss": 0.4086599349975586, | |
| "eval_runtime": 25.573, | |
| "eval_samples_per_second": 1580.967, | |
| "eval_steps_per_second": 6.178, | |
| "step": 5688 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 3.262328863143921, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.1933, | |
| "step": 7110 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8506059856542172, | |
| "eval_combined_score": 0.8207251134725826, | |
| "eval_f1": 0.7908442412909481, | |
| "eval_loss": 0.44820770621299744, | |
| "eval_runtime": 26.1949, | |
| "eval_samples_per_second": 1543.43, | |
| "eval_steps_per_second": 6.032, | |
| "step": 7110 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 4.035948276519775, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 0.1514, | |
| "step": 8532 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8534504081127875, | |
| "eval_combined_score": 0.8276283651201168, | |
| "eval_f1": 0.8018063221274461, | |
| "eval_loss": 0.43122583627700806, | |
| "eval_runtime": 26.1244, | |
| "eval_samples_per_second": 1547.595, | |
| "eval_steps_per_second": 6.048, | |
| "step": 8532 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 4.258938312530518, | |
| "learning_rate": 4.3e-05, | |
| "loss": 0.1208, | |
| "step": 9954 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8498392282958199, | |
| "eval_combined_score": 0.8269528560391477, | |
| "eval_f1": 0.8040664837824754, | |
| "eval_loss": 0.5434415936470032, | |
| "eval_runtime": 25.8296, | |
| "eval_samples_per_second": 1565.258, | |
| "eval_steps_per_second": 6.117, | |
| "step": 9954 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 3.5687382221221924, | |
| "learning_rate": 4.2e-05, | |
| "loss": 0.097, | |
| "step": 11376 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.853227801137769, | |
| "eval_combined_score": 0.8277139005688845, | |
| "eval_f1": 0.8022000000000001, | |
| "eval_loss": 0.5604811310768127, | |
| "eval_runtime": 25.8751, | |
| "eval_samples_per_second": 1562.506, | |
| "eval_steps_per_second": 6.106, | |
| "step": 11376 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "step": 11376, | |
| "total_flos": 3.8292762019405824e+17, | |
| "train_loss": 0.2420723964728719, | |
| "train_runtime": 4923.4333, | |
| "train_samples_per_second": 3695.043, | |
| "train_steps_per_second": 14.441 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 71100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 5 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.8292762019405824e+17, | |
| "train_batch_size": 256, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |