| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.0, | |
| "global_step": 11644, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 1.9312950875987636e-05, | |
| "loss": 0.5996, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 1.8625901751975268e-05, | |
| "loss": 0.6564, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 1.7938852627962902e-05, | |
| "loss": 0.6654, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.7251803503950534e-05, | |
| "loss": 0.6449, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 1.6564754379938165e-05, | |
| "loss": 0.6363, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_bleu": 11.080936212068192, | |
| "eval_loss": 0.7257852554321289, | |
| "eval_runtime": 19.0528, | |
| "eval_samples_per_second": 67.917, | |
| "eval_steps_per_second": 4.251, | |
| "step": 2911 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "learning_rate": 1.58777052559258e-05, | |
| "loss": 0.6256, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 1.5192030230161459e-05, | |
| "loss": 0.513, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "learning_rate": 1.4504981106149092e-05, | |
| "loss": 0.5154, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "learning_rate": 1.3817931982136723e-05, | |
| "loss": 0.5233, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "learning_rate": 1.3130882858124357e-05, | |
| "loss": 0.5053, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 1.244383373411199e-05, | |
| "loss": 0.5053, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_bleu": 11.760028545810009, | |
| "eval_loss": 0.7024416923522949, | |
| "eval_runtime": 19.0637, | |
| "eval_samples_per_second": 67.878, | |
| "eval_steps_per_second": 4.249, | |
| "step": 5822 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "learning_rate": 1.1756784610099623e-05, | |
| "loss": 0.4781, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "learning_rate": 1.1069735486087256e-05, | |
| "loss": 0.4301, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 1.038268636207489e-05, | |
| "loss": 0.4228, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "learning_rate": 9.697011336310546e-06, | |
| "loss": 0.4341, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "learning_rate": 9.009962212298181e-06, | |
| "loss": 0.4195, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "learning_rate": 8.322913088285814e-06, | |
| "loss": 0.4226, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_bleu": 16.148157890934268, | |
| "eval_loss": 0.6924545168876648, | |
| "eval_runtime": 18.804, | |
| "eval_samples_per_second": 68.815, | |
| "eval_steps_per_second": 4.308, | |
| "step": 8733 | |
| }, | |
| { | |
| "epoch": 3.09, | |
| "learning_rate": 7.63723806252147e-06, | |
| "loss": 0.3849, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.26, | |
| "learning_rate": 6.9501889385091035e-06, | |
| "loss": 0.3686, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "learning_rate": 6.263139814496737e-06, | |
| "loss": 0.3713, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "learning_rate": 5.5760906904843705e-06, | |
| "loss": 0.3681, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 3.78, | |
| "learning_rate": 4.8890415664720035e-06, | |
| "loss": 0.3708, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 3.95, | |
| "learning_rate": 4.2019924424596366e-06, | |
| "loss": 0.3665, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_bleu": 15.230107565625683, | |
| "eval_loss": 0.6867288947105408, | |
| "eval_runtime": 18.8096, | |
| "eval_samples_per_second": 68.795, | |
| "eval_steps_per_second": 4.306, | |
| "step": 11644 | |
| } | |
| ], | |
| "max_steps": 14555, | |
| "num_train_epochs": 5, | |
| "total_flos": 284658341511168.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |