| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 8.0, | |
| "eval_steps": 500, | |
| "global_step": 7280, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.5494505494505495, | |
| "grad_norm": 7.208542823791504, | |
| "learning_rate": 1.862637362637363e-05, | |
| "loss": 1.4229, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_bleu": 45.1493, | |
| "eval_gen_len": 21.1512, | |
| "eval_loss": 0.9880260229110718, | |
| "eval_runtime": 159.8583, | |
| "eval_samples_per_second": 11.379, | |
| "eval_steps_per_second": 0.713, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.098901098901099, | |
| "grad_norm": 5.381701469421387, | |
| "learning_rate": 1.7252747252747256e-05, | |
| "loss": 1.1197, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.6483516483516483, | |
| "grad_norm": 4.710964679718018, | |
| "learning_rate": 1.587912087912088e-05, | |
| "loss": 0.9205, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_bleu": 48.9448, | |
| "eval_gen_len": 21.4244, | |
| "eval_loss": 0.8794811367988586, | |
| "eval_runtime": 164.115, | |
| "eval_samples_per_second": 11.084, | |
| "eval_steps_per_second": 0.695, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 2.197802197802198, | |
| "grad_norm": 3.468052864074707, | |
| "learning_rate": 1.4505494505494506e-05, | |
| "loss": 0.8318, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.7472527472527473, | |
| "grad_norm": 3.9397318363189697, | |
| "learning_rate": 1.3131868131868134e-05, | |
| "loss": 0.7509, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_bleu": 50.7507, | |
| "eval_gen_len": 21.4739, | |
| "eval_loss": 0.83347487449646, | |
| "eval_runtime": 167.1805, | |
| "eval_samples_per_second": 10.88, | |
| "eval_steps_per_second": 0.682, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 3.2967032967032965, | |
| "grad_norm": 4.654043197631836, | |
| "learning_rate": 1.1758241758241759e-05, | |
| "loss": 0.7075, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 3.244563341140747, | |
| "learning_rate": 1.0384615384615386e-05, | |
| "loss": 0.6522, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_bleu": 51.8451, | |
| "eval_gen_len": 21.4134, | |
| "eval_loss": 0.8046153783798218, | |
| "eval_runtime": 167.5891, | |
| "eval_samples_per_second": 10.854, | |
| "eval_steps_per_second": 0.68, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 4.395604395604396, | |
| "grad_norm": 4.202299118041992, | |
| "learning_rate": 9.010989010989011e-06, | |
| "loss": 0.6097, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 4.945054945054945, | |
| "grad_norm": 6.508543968200684, | |
| "learning_rate": 7.637362637362638e-06, | |
| "loss": 0.5877, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_bleu": 52.489, | |
| "eval_gen_len": 21.453, | |
| "eval_loss": 0.7915568947792053, | |
| "eval_runtime": 170.447, | |
| "eval_samples_per_second": 10.672, | |
| "eval_steps_per_second": 0.669, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 5.4945054945054945, | |
| "grad_norm": 4.693784236907959, | |
| "learning_rate": 6.2637362637362645e-06, | |
| "loss": 0.5454, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_bleu": 53.267, | |
| "eval_gen_len": 21.4063, | |
| "eval_loss": 0.7819094657897949, | |
| "eval_runtime": 166.0221, | |
| "eval_samples_per_second": 10.956, | |
| "eval_steps_per_second": 0.687, | |
| "step": 5460 | |
| }, | |
| { | |
| "epoch": 6.043956043956044, | |
| "grad_norm": 3.2858574390411377, | |
| "learning_rate": 4.890109890109891e-06, | |
| "loss": 0.5319, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 6.593406593406593, | |
| "grad_norm": 4.065381050109863, | |
| "learning_rate": 3.516483516483517e-06, | |
| "loss": 0.5039, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_bleu": 53.3718, | |
| "eval_gen_len": 21.5014, | |
| "eval_loss": 0.7780753970146179, | |
| "eval_runtime": 172.3653, | |
| "eval_samples_per_second": 10.553, | |
| "eval_steps_per_second": 0.661, | |
| "step": 6370 | |
| }, | |
| { | |
| "epoch": 7.142857142857143, | |
| "grad_norm": 3.438549280166626, | |
| "learning_rate": 2.1428571428571427e-06, | |
| "loss": 0.5018, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 7.6923076923076925, | |
| "grad_norm": 4.464223384857178, | |
| "learning_rate": 7.692307692307694e-07, | |
| "loss": 0.4847, | |
| "step": 7000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 7280, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 8, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2444491603574784.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |