| { | |
| "best_metric": 0.2672373652458191, | |
| "best_model_checkpoint": "bert_base_lda_20_v1_book_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": 3.422664165496826, | |
| "learning_rate": 4.9e-05, | |
| "loss": 0.3572, | |
| "step": 1422 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8733366312144447, | |
| "eval_combined_score": 0.8525603104828916, | |
| "eval_f1": 0.8317839897513386, | |
| "eval_loss": 0.2866745889186859, | |
| "eval_runtime": 26.2406, | |
| "eval_samples_per_second": 1540.741, | |
| "eval_steps_per_second": 6.021, | |
| "step": 1422 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 3.693434476852417, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.2449, | |
| "step": 2844 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8808805342567401, | |
| "eval_combined_score": 0.8654760681370246, | |
| "eval_f1": 0.850071602017309, | |
| "eval_loss": 0.2800416052341461, | |
| "eval_runtime": 26.839, | |
| "eval_samples_per_second": 1506.387, | |
| "eval_steps_per_second": 5.887, | |
| "step": 2844 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 3.081462860107422, | |
| "learning_rate": 4.7e-05, | |
| "loss": 0.1784, | |
| "step": 4266 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8944595597328716, | |
| "eval_combined_score": 0.8776961619020907, | |
| "eval_f1": 0.8609327640713098, | |
| "eval_loss": 0.2672373652458191, | |
| "eval_runtime": 26.5933, | |
| "eval_samples_per_second": 1520.307, | |
| "eval_steps_per_second": 5.941, | |
| "step": 4266 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 3.004323959350586, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.1265, | |
| "step": 5688 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8950779124412565, | |
| "eval_combined_score": 0.8781572094100296, | |
| "eval_f1": 0.8612365063788028, | |
| "eval_loss": 0.28892993927001953, | |
| "eval_runtime": 26.563, | |
| "eval_samples_per_second": 1522.04, | |
| "eval_steps_per_second": 5.948, | |
| "step": 5688 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 2.8927578926086426, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.0934, | |
| "step": 7110 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8965619589413801, | |
| "eval_combined_score": 0.8765829956550268, | |
| "eval_f1": 0.8566040323686737, | |
| "eval_loss": 0.3325226604938507, | |
| "eval_runtime": 26.1913, | |
| "eval_samples_per_second": 1543.645, | |
| "eval_steps_per_second": 6.033, | |
| "step": 7110 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 1.2701847553253174, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 0.0723, | |
| "step": 8532 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8954736581746228, | |
| "eval_combined_score": 0.8780008554899517, | |
| "eval_f1": 0.8605280528052807, | |
| "eval_loss": 0.3647216558456421, | |
| "eval_runtime": 26.5512, | |
| "eval_samples_per_second": 1522.718, | |
| "eval_steps_per_second": 5.951, | |
| "step": 8532 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 6.682582855224609, | |
| "learning_rate": 4.3e-05, | |
| "loss": 0.0576, | |
| "step": 9954 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8960178085580015, | |
| "eval_combined_score": 0.879708930273282, | |
| "eval_f1": 0.8634000519885625, | |
| "eval_loss": 0.4286547601222992, | |
| "eval_runtime": 26.3205, | |
| "eval_samples_per_second": 1536.066, | |
| "eval_steps_per_second": 6.003, | |
| "step": 9954 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 2.5855891704559326, | |
| "learning_rate": 4.2e-05, | |
| "loss": 0.0486, | |
| "step": 11376 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8966856294830572, | |
| "eval_combined_score": 0.8783599041762369, | |
| "eval_f1": 0.8600341788694166, | |
| "eval_loss": 0.39322641491889954, | |
| "eval_runtime": 26.61, | |
| "eval_samples_per_second": 1519.352, | |
| "eval_steps_per_second": 5.938, | |
| "step": 11376 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "step": 11376, | |
| "total_flos": 3.8292762019405824e+17, | |
| "train_loss": 0.147349518562671, | |
| "train_runtime": 4974.4295, | |
| "train_samples_per_second": 3657.163, | |
| "train_steps_per_second": 14.293 | |
| } | |
| ], | |
| "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 | |
| } | |