| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 50.0, | |
| "eval_steps": 1000, | |
| "global_step": 4850, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 5.15, | |
| "learning_rate": 4.88e-05, | |
| "loss": 9.8461, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 10.31, | |
| "learning_rate": 4.4390804597701155e-05, | |
| "loss": 2.7876, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 10.31, | |
| "eval_bleu": 15.7275, | |
| "eval_gen_len": 15.5938, | |
| "eval_loss": 4.260565757751465, | |
| "eval_runtime": 13.4099, | |
| "eval_samples_per_second": 14.318, | |
| "eval_steps_per_second": 0.895, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 15.46, | |
| "learning_rate": 3.864367816091955e-05, | |
| "loss": 0.5407, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 20.62, | |
| "learning_rate": 3.289655172413793e-05, | |
| "loss": 0.1404, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 20.62, | |
| "eval_bleu": 16.6706, | |
| "eval_gen_len": 17.4375, | |
| "eval_loss": 4.24957799911499, | |
| "eval_runtime": 13.1685, | |
| "eval_samples_per_second": 14.58, | |
| "eval_steps_per_second": 0.911, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 25.77, | |
| "learning_rate": 2.7149425287356323e-05, | |
| "loss": 0.072, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 30.93, | |
| "learning_rate": 2.1402298850574715e-05, | |
| "loss": 0.0398, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 30.93, | |
| "eval_bleu": 19.2786, | |
| "eval_gen_len": 17.8385, | |
| "eval_loss": 4.348609447479248, | |
| "eval_runtime": 13.5832, | |
| "eval_samples_per_second": 14.135, | |
| "eval_steps_per_second": 0.883, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 36.08, | |
| "learning_rate": 1.5655172413793103e-05, | |
| "loss": 0.0203, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 41.24, | |
| "learning_rate": 9.908045977011495e-06, | |
| "loss": 0.0107, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 41.24, | |
| "eval_bleu": 21.5085, | |
| "eval_gen_len": 17.2917, | |
| "eval_loss": 4.341139316558838, | |
| "eval_runtime": 13.34, | |
| "eval_samples_per_second": 14.393, | |
| "eval_steps_per_second": 0.9, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 46.39, | |
| "learning_rate": 4.160919540229885e-06, | |
| "loss": 0.0049, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "step": 4850, | |
| "total_flos": 1.681692694549627e+17, | |
| "train_loss": 1.3880607915416205, | |
| "train_runtime": 3276.7904, | |
| "train_samples_per_second": 23.544, | |
| "train_steps_per_second": 1.48 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4850, | |
| "num_train_epochs": 50, | |
| "save_steps": 1000, | |
| "total_flos": 1.681692694549627e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |