| { | |
| "best_metric": 0.9052917957305908, | |
| "best_model_checkpoint": "./tst-translation-output/checkpoint-12000", | |
| "epoch": 1.8653816259909841, | |
| "eval_steps": 3000, | |
| "global_step": 12000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 9.712509712509713e-06, | |
| "loss": 1.9995, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 1.9425019425019425e-05, | |
| "loss": 1.4173, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 2.913752913752914e-05, | |
| "loss": 1.3137, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 3.885003885003885e-05, | |
| "loss": 1.2381, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 4.856254856254856e-05, | |
| "loss": 1.2023, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 4.991638730343166e-05, | |
| "loss": 1.1751, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "eval_bleu": 7.6766, | |
| "eval_gen_len": 17.4644, | |
| "eval_loss": 1.1288801431655884, | |
| "eval_runtime": 851.3289, | |
| "eval_samples_per_second": 15.112, | |
| "eval_steps_per_second": 0.946, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 4.981825033562843e-05, | |
| "loss": 1.1267, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 4.972011336782521e-05, | |
| "loss": 1.1013, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 4.962197640002198e-05, | |
| "loss": 1.0821, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 4.952383943221876e-05, | |
| "loss": 1.0599, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 4.9425702464415536e-05, | |
| "loss": 1.0336, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 4.9327565496612316e-05, | |
| "loss": 1.0268, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "eval_bleu": 9.2277, | |
| "eval_gen_len": 17.7668, | |
| "eval_loss": 0.9894970655441284, | |
| "eval_runtime": 891.3112, | |
| "eval_samples_per_second": 14.434, | |
| "eval_steps_per_second": 0.903, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 4.922942852880909e-05, | |
| "loss": 0.9631, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 4.913129156100586e-05, | |
| "loss": 0.8021, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 4.903315459320265e-05, | |
| "loss": 0.8189, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "learning_rate": 4.893501762539942e-05, | |
| "loss": 0.8034, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "learning_rate": 4.8836880657596196e-05, | |
| "loss": 0.8156, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 4.873874368979297e-05, | |
| "loss": 0.8075, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "eval_bleu": 9.1197, | |
| "eval_gen_len": 17.6811, | |
| "eval_loss": 0.945686399936676, | |
| "eval_runtime": 839.7466, | |
| "eval_samples_per_second": 15.32, | |
| "eval_steps_per_second": 0.959, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "learning_rate": 4.864060672198975e-05, | |
| "loss": 0.81, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "learning_rate": 4.854246975418653e-05, | |
| "loss": 0.8101, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "learning_rate": 4.84443327863833e-05, | |
| "loss": 0.7964, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "learning_rate": 4.8346195818580075e-05, | |
| "loss": 0.806, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 4.8248058850776855e-05, | |
| "loss": 0.8042, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 4.814992188297363e-05, | |
| "loss": 0.8082, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "eval_bleu": 8.4837, | |
| "eval_gen_len": 17.4826, | |
| "eval_loss": 0.9052917957305908, | |
| "eval_runtime": 837.7614, | |
| "eval_samples_per_second": 15.356, | |
| "eval_steps_per_second": 0.961, | |
| "step": 12000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 257320, | |
| "num_train_epochs": 40, | |
| "save_steps": 3000, | |
| "total_flos": 4.16088915443712e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |