| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 7.377049180327869, | |
| "global_step": 9000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 1.918032786885246e-05, | |
| "loss": 1.0017, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 1.836065573770492e-05, | |
| "loss": 0.0813, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_bleu": 0.0, | |
| "eval_gen_len": 9.8363, | |
| "eval_loss": 0.11947569251060486, | |
| "eval_runtime": 48.7643, | |
| "eval_samples_per_second": 41.957, | |
| "eval_steps_per_second": 0.656, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "learning_rate": 1.7540983606557377e-05, | |
| "loss": 0.0534, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "learning_rate": 1.6721311475409837e-05, | |
| "loss": 0.038, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_bleu": 0.0, | |
| "eval_gen_len": 9.8495, | |
| "eval_loss": 0.126708522439003, | |
| "eval_runtime": 51.3358, | |
| "eval_samples_per_second": 39.855, | |
| "eval_steps_per_second": 0.623, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "learning_rate": 1.5901639344262295e-05, | |
| "loss": 0.0347, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "learning_rate": 1.5081967213114754e-05, | |
| "loss": 0.0187, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "learning_rate": 1.4262295081967214e-05, | |
| "loss": 0.0198, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_bleu": 0.0, | |
| "eval_gen_len": 9.8822, | |
| "eval_loss": 0.13768209517002106, | |
| "eval_runtime": 49.8051, | |
| "eval_samples_per_second": 41.08, | |
| "eval_steps_per_second": 0.643, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 3.28, | |
| "learning_rate": 1.3442622950819673e-05, | |
| "loss": 0.0134, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 3.69, | |
| "learning_rate": 1.2622950819672132e-05, | |
| "loss": 0.0108, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_bleu": 0.0, | |
| "eval_gen_len": 9.9233, | |
| "eval_loss": 0.15018606185913086, | |
| "eval_runtime": 50.7382, | |
| "eval_samples_per_second": 40.325, | |
| "eval_steps_per_second": 0.631, | |
| "step": 4880 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "learning_rate": 1.1803278688524591e-05, | |
| "loss": 0.0099, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 4.51, | |
| "learning_rate": 1.0983606557377052e-05, | |
| "loss": 0.0065, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 4.92, | |
| "learning_rate": 1.0163934426229509e-05, | |
| "loss": 0.0067, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_bleu": 0.0, | |
| "eval_gen_len": 9.8421, | |
| "eval_loss": 0.1598789393901825, | |
| "eval_runtime": 51.4429, | |
| "eval_samples_per_second": 39.772, | |
| "eval_steps_per_second": 0.622, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 5.33, | |
| "learning_rate": 9.344262295081968e-06, | |
| "loss": 0.0051, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 5.74, | |
| "learning_rate": 8.524590163934427e-06, | |
| "loss": 0.0045, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_bleu": 0.0, | |
| "eval_gen_len": 9.8827, | |
| "eval_loss": 0.16899947822093964, | |
| "eval_runtime": 50.0655, | |
| "eval_samples_per_second": 40.866, | |
| "eval_steps_per_second": 0.639, | |
| "step": 7320 | |
| }, | |
| { | |
| "epoch": 6.15, | |
| "learning_rate": 7.704918032786886e-06, | |
| "loss": 0.0043, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 6.56, | |
| "learning_rate": 6.885245901639345e-06, | |
| "loss": 0.0035, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 6.97, | |
| "learning_rate": 6.065573770491804e-06, | |
| "loss": 0.0036, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_bleu": 0.0, | |
| "eval_gen_len": 9.8832, | |
| "eval_loss": 0.17745506763458252, | |
| "eval_runtime": 50.7877, | |
| "eval_samples_per_second": 40.285, | |
| "eval_steps_per_second": 0.63, | |
| "step": 8540 | |
| }, | |
| { | |
| "epoch": 7.38, | |
| "learning_rate": 5.245901639344263e-06, | |
| "loss": 0.0028, | |
| "step": 9000 | |
| } | |
| ], | |
| "max_steps": 12200, | |
| "num_train_epochs": 10, | |
| "total_flos": 5.951675547814134e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |