| { | |
| "best_metric": 0.4823857545852661, | |
| "best_model_checkpoint": "ft-roberta-toxicity/checkpoint-200", | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.1541271209716797, | |
| "learning_rate": 5e-06, | |
| "loss": 0.7051, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 4.209589958190918, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6954, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 5.086088180541992, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.719, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 4.385767936706543, | |
| "learning_rate": 2e-05, | |
| "loss": 0.6785, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 3.3053433895111084, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.6874, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 2.8885927200317383, | |
| "learning_rate": 3e-05, | |
| "loss": 0.6988, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 2.523589611053467, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.7674, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 8.986541748046875, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6891, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 6.452502727508545, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.8348, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 6.318837642669678, | |
| "learning_rate": 5e-05, | |
| "loss": 0.7019, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 3.7034225463867188, | |
| "learning_rate": 4.9444444444444446e-05, | |
| "loss": 0.65, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 3.1009528636932373, | |
| "learning_rate": 4.888888888888889e-05, | |
| "loss": 0.6593, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 7.615400791168213, | |
| "learning_rate": 4.8333333333333334e-05, | |
| "loss": 0.6755, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 3.228475332260132, | |
| "learning_rate": 4.7777777777777784e-05, | |
| "loss": 0.6034, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 4.151826858520508, | |
| "learning_rate": 4.722222222222222e-05, | |
| "loss": 0.7218, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 11.06347370147705, | |
| "learning_rate": 4.666666666666667e-05, | |
| "loss": 0.7539, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 24.23040771484375, | |
| "learning_rate": 4.6111111111111115e-05, | |
| "loss": 0.6296, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 6.828371047973633, | |
| "learning_rate": 4.555555555555556e-05, | |
| "loss": 0.559, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 7.590404510498047, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.4996, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 51.719970703125, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 0.6665, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.715, | |
| "eval_auc": 0.795187969924812, | |
| "eval_f1": 0.7219512195121951, | |
| "eval_loss": 0.5786986351013184, | |
| "eval_precision": 0.74, | |
| "eval_recall": 0.7047619047619048, | |
| "eval_runtime": 32.3877, | |
| "eval_samples_per_second": 6.175, | |
| "eval_steps_per_second": 0.401, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 13.197641372680664, | |
| "learning_rate": 4.388888888888889e-05, | |
| "loss": 0.6974, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 4.645239353179932, | |
| "learning_rate": 4.3333333333333334e-05, | |
| "loss": 0.4852, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 4.36188268661499, | |
| "learning_rate": 4.277777777777778e-05, | |
| "loss": 0.7566, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 6.4677300453186035, | |
| "learning_rate": 4.222222222222222e-05, | |
| "loss": 0.4989, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 6.412233829498291, | |
| "learning_rate": 4.166666666666667e-05, | |
| "loss": 0.5496, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 12.167023658752441, | |
| "learning_rate": 4.111111111111111e-05, | |
| "loss": 0.5181, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 4.28169584274292, | |
| "learning_rate": 4.055555555555556e-05, | |
| "loss": 0.4636, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 20.04325294494629, | |
| "learning_rate": 4e-05, | |
| "loss": 0.4278, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 18.524545669555664, | |
| "learning_rate": 3.944444444444445e-05, | |
| "loss": 0.574, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 12.304378509521484, | |
| "learning_rate": 3.888888888888889e-05, | |
| "loss": 0.4926, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 16.889781951904297, | |
| "learning_rate": 3.8333333333333334e-05, | |
| "loss": 0.2429, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 18.878406524658203, | |
| "learning_rate": 3.777777777777778e-05, | |
| "loss": 0.4433, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 21.254526138305664, | |
| "learning_rate": 3.722222222222222e-05, | |
| "loss": 0.644, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 21.575471878051758, | |
| "learning_rate": 3.6666666666666666e-05, | |
| "loss": 0.6373, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 8.538455963134766, | |
| "learning_rate": 3.611111111111111e-05, | |
| "loss": 0.5708, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 7.5060577392578125, | |
| "learning_rate": 3.555555555555556e-05, | |
| "loss": 0.7443, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 7.51224946975708, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.4896, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 5.721360683441162, | |
| "learning_rate": 3.444444444444445e-05, | |
| "loss": 0.544, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 6.373270511627197, | |
| "learning_rate": 3.388888888888889e-05, | |
| "loss": 0.4813, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 5.219586372375488, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.3724, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.79, | |
| "eval_auc": 0.8741854636591478, | |
| "eval_f1": 0.7666666666666667, | |
| "eval_loss": 0.4823857545852661, | |
| "eval_precision": 0.92, | |
| "eval_recall": 0.6571428571428571, | |
| "eval_runtime": 32.4699, | |
| "eval_samples_per_second": 6.16, | |
| "eval_steps_per_second": 0.4, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.01 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 105244422144000.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |