| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 1587, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1890359168241966, | |
| "grad_norm": 7.269421577453613, | |
| "learning_rate": 0.0005, | |
| "loss": 1.9195, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3780718336483932, | |
| "grad_norm": 1.071650505065918, | |
| "learning_rate": 0.0005, | |
| "loss": 1.9396, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5671077504725898, | |
| "grad_norm": 0.5662499070167542, | |
| "learning_rate": 0.0005, | |
| "loss": 2.3641, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7561436672967864, | |
| "grad_norm": 0.47088223695755005, | |
| "learning_rate": 0.0005, | |
| "loss": 1.8969, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.945179584120983, | |
| "grad_norm": 0.3653932511806488, | |
| "learning_rate": 0.0005, | |
| "loss": 1.8531, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6214529147982063, | |
| "eval_loss": 1.5580899715423584, | |
| "eval_runtime": 5.56, | |
| "eval_samples_per_second": 89.928, | |
| "eval_steps_per_second": 11.331, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_exact_match": 14.4, | |
| "eval_f1": 20.983333333333345, | |
| "eval_qa_bleu": 12.514681390986086, | |
| "eval_qa_exact_match": 0.102, | |
| "eval_recite_bleu": 12.61307158352263, | |
| "eval_recite_exact_match": 0.0, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 1.1342155009451795, | |
| "grad_norm": 0.5318541526794434, | |
| "learning_rate": 0.0005, | |
| "loss": 1.6622, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.3232514177693762, | |
| "grad_norm": 235.6529998779297, | |
| "learning_rate": 0.0005, | |
| "loss": 1.5596, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.5122873345935728, | |
| "grad_norm": 0.5057249069213867, | |
| "learning_rate": 0.0005, | |
| "loss": 1.5296, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.7013232514177694, | |
| "grad_norm": 0.5339187383651733, | |
| "learning_rate": 0.0005, | |
| "loss": 1.5036, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.8903591682419658, | |
| "grad_norm": 0.5649825930595398, | |
| "learning_rate": 0.0005, | |
| "loss": 1.4884, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6643587443946188, | |
| "eval_loss": 1.2189537286758423, | |
| "eval_runtime": 6.192, | |
| "eval_samples_per_second": 80.749, | |
| "eval_steps_per_second": 10.174, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_exact_match": 11.2, | |
| "eval_f1": 18.31238095238096, | |
| "eval_qa_bleu": 10.766593815425745, | |
| "eval_qa_exact_match": 0.096, | |
| "eval_recite_bleu": 13.257993783037875, | |
| "eval_recite_exact_match": 0.002, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 2.0793950850661624, | |
| "grad_norm": 0.6042512655258179, | |
| "learning_rate": 0.0005, | |
| "loss": 1.2966, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.268431001890359, | |
| "grad_norm": 965.2459106445312, | |
| "learning_rate": 0.0005, | |
| "loss": 8.8153, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.4574669187145557, | |
| "grad_norm": 3.0128896236419678, | |
| "learning_rate": 0.0005, | |
| "loss": 9.1241, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.6465028355387523, | |
| "grad_norm": 108062179328.0, | |
| "learning_rate": 0.0005, | |
| "loss": 10.5947, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.835538752362949, | |
| "grad_norm": 5002999889920.0, | |
| "learning_rate": 0.0005, | |
| "loss": 12.5972, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.21054708520179372, | |
| "eval_loss": 19.724363327026367, | |
| "eval_runtime": 6.5285, | |
| "eval_samples_per_second": 76.587, | |
| "eval_steps_per_second": 9.65, | |
| "step": 1587 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_exact_match": 0.0, | |
| "eval_f1": 0.0, | |
| "eval_qa_bleu": 0.0, | |
| "eval_qa_exact_match": 0.0, | |
| "eval_recite_bleu": 0.0, | |
| "eval_recite_exact_match": 0.0, | |
| "step": 1587 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 26450, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "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": 1.446358715245527e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |