stratified_10m_curriculum_roberta_perplexity_increasing / checkpoints /checkpoint-1566 /trainer_state.json
| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9988041138483617, | |
| "eval_steps": 500, | |
| "global_step": 1566, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.09567089213106912, | |
| "grad_norm": 117.45254516601562, | |
| "learning_rate": 2.5e-06, | |
| "loss": 31905.095, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.19134178426213824, | |
| "grad_norm": 71.91663360595703, | |
| "learning_rate": 5e-06, | |
| "loss": 2260638924.8, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2870126763932074, | |
| "grad_norm": 56.18342590332031, | |
| "learning_rate": 7.5e-06, | |
| "loss": 63.7996, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3826835685242765, | |
| "grad_norm": 47.14400100708008, | |
| "learning_rate": 1e-05, | |
| "loss": 12.4659, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4783544606553456, | |
| "grad_norm": 57.82217025756836, | |
| "learning_rate": 1.25e-05, | |
| "loss": 9.1207, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5740253527864148, | |
| "grad_norm": 36.61322021484375, | |
| "learning_rate": 1.5e-05, | |
| "loss": 8.5922, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6696962449174838, | |
| "grad_norm": 30.304691314697266, | |
| "learning_rate": 1.7500000000000002e-05, | |
| "loss": 7.5559, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.765367137048553, | |
| "grad_norm": 24.35045051574707, | |
| "learning_rate": 2e-05, | |
| "loss": 7.2955, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8610380291796221, | |
| "grad_norm": 24.114551544189453, | |
| "learning_rate": 2.2499999999999998e-05, | |
| "loss": 7.2298, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9567089213106912, | |
| "grad_norm": 22.68138313293457, | |
| "learning_rate": 2.5e-05, | |
| "loss": 7.5636, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.9988041138483617, | |
| "eval_accuracy": 0.0, | |
| "eval_loss": 10.570171356201172, | |
| "eval_normalizer": 94012.0, | |
| "eval_runtime": 76.7018, | |
| "eval_samples_per_second": 696.945, | |
| "eval_steps_per_second": 1.369, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 1.0535756995933987, | |
| "grad_norm": 136.1419219970703, | |
| "learning_rate": 2.75e-05, | |
| "loss": 244134.34, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.1492465917244679, | |
| "grad_norm": 83.12769317626953, | |
| "learning_rate": 3e-05, | |
| "loss": 422854923386.88, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.244917483855537, | |
| "grad_norm": 75.21563720703125, | |
| "learning_rate": 3.2500000000000004e-05, | |
| "loss": 122855.21, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.3405883759866062, | |
| "grad_norm": 53.24630355834961, | |
| "learning_rate": 3.5000000000000004e-05, | |
| "loss": 8.0066, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.4362592681176751, | |
| "grad_norm": 78.36145782470703, | |
| "learning_rate": 3.75e-05, | |
| "loss": 5.7671, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.5319301602487443, | |
| "grad_norm": 42.857173919677734, | |
| "learning_rate": 4e-05, | |
| "loss": 5.6876, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.6276010523798135, | |
| "grad_norm": 68.13167572021484, | |
| "learning_rate": 4.25e-05, | |
| "loss": 5.532, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.7232719445108826, | |
| "grad_norm": 84.04206085205078, | |
| "learning_rate": 4.4999999999999996e-05, | |
| "loss": 5.8722, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.8189428366419516, | |
| "grad_norm": 172.8482666015625, | |
| "learning_rate": 4.75e-05, | |
| "loss": 6.2774, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.9146137287730207, | |
| "grad_norm": 71.3455810546875, | |
| "learning_rate": 5e-05, | |
| "loss": 6.8607, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.9988041138483617, | |
| "eval_accuracy": 0.0, | |
| "eval_loss": 12.504220962524414, | |
| "eval_normalizer": 93558.0, | |
| "eval_runtime": 76.0703, | |
| "eval_samples_per_second": 702.731, | |
| "eval_steps_per_second": 1.38, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 2.0114805070557282, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.25e-05, | |
| "loss": 49.308, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.1071513991867974, | |
| "grad_norm": 95.2462158203125, | |
| "learning_rate": 5.5e-05, | |
| "loss": 15104857.6, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.2028222913178666, | |
| "grad_norm": 47.490631103515625, | |
| "learning_rate": 5.75e-05, | |
| "loss": 164198594.56, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.2984931834489357, | |
| "grad_norm": 36.559452056884766, | |
| "learning_rate": 6e-05, | |
| "loss": 19.3007, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.394164075580005, | |
| "grad_norm": 71.66349029541016, | |
| "learning_rate": 6.25e-05, | |
| "loss": 4.3654, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.489834967711074, | |
| "grad_norm": 62.64774703979492, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 4.5691, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.5855058598421428, | |
| "grad_norm": 82.2427749633789, | |
| "learning_rate": 6.75e-05, | |
| "loss": 4.6713, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.6811767519732124, | |
| "grad_norm": 28.598220825195312, | |
| "learning_rate": 7.000000000000001e-05, | |
| "loss": 4.9054, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.776847644104281, | |
| "grad_norm": 55.590911865234375, | |
| "learning_rate": 7.25e-05, | |
| "loss": 5.3171, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.8725185362353503, | |
| "grad_norm": 52.226806640625, | |
| "learning_rate": 7.5e-05, | |
| "loss": 5.8207, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.9681894283664194, | |
| "grad_norm": 41.47724151611328, | |
| "learning_rate": 7.75e-05, | |
| "loss": 6.6854, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.9988041138483617, | |
| "eval_accuracy": 0.0, | |
| "eval_loss": 13.740432739257812, | |
| "eval_normalizer": 94216.0, | |
| "eval_runtime": 75.872, | |
| "eval_samples_per_second": 704.568, | |
| "eval_steps_per_second": 1.384, | |
| "step": 1566 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 5220, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.2563861448012595e+17, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |