| { | |
| "best_metric": 6.0799336433410645, | |
| "best_model_checkpoint": "./output_dir/checkpoint-629", | |
| "epoch": 0.9996027016289233, | |
| "eval_steps": 50.0, | |
| "global_step": 629, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.9626450538635254, | |
| "learning_rate": 9.205087440381559e-05, | |
| "loss": 0.8035, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.889542281627655, | |
| "learning_rate": 8.410174880763116e-05, | |
| "loss": 0.3005, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.5076414942741394, | |
| "learning_rate": 7.615262321144675e-05, | |
| "loss": 0.2731, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 1.0295907258987427, | |
| "learning_rate": 6.820349761526232e-05, | |
| "loss": 0.2261, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.3473876416683197, | |
| "learning_rate": 6.0254372019077906e-05, | |
| "loss": 0.1866, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.5202713012695312, | |
| "learning_rate": 5.2305246422893485e-05, | |
| "loss": 0.1937, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.25799375772476196, | |
| "learning_rate": 4.4356120826709064e-05, | |
| "loss": 0.1769, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.2543068826198578, | |
| "learning_rate": 3.640699523052464e-05, | |
| "loss": 0.1693, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.5472003221511841, | |
| "learning_rate": 2.8457869634340222e-05, | |
| "loss": 0.1611, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.19463592767715454, | |
| "learning_rate": 2.05087440381558e-05, | |
| "loss": 0.1487, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.7376086711883545, | |
| "learning_rate": 1.2559618441971382e-05, | |
| "loss": 0.1111, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.5419193506240845, | |
| "learning_rate": 4.610492845786963e-06, | |
| "loss": 0.122, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_bleu-1": 78.7702, | |
| "eval_bleu-2": 75.9001, | |
| "eval_bleu-3": 75.2623, | |
| "eval_bleu-4": 74.9418, | |
| "eval_gen_len": 128.0, | |
| "eval_loss": 6.0799336433410645, | |
| "eval_runtime": 5.1216, | |
| "eval_samples_per_second": 2.343, | |
| "eval_steps_per_second": 0.586, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 629, | |
| "total_flos": 8.181451760192717e+16, | |
| "train_loss": 0.23391734473466494, | |
| "train_runtime": 1027.8255, | |
| "train_samples_per_second": 9.795, | |
| "train_steps_per_second": 0.612 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 629, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "total_flos": 8.181451760192717e+16, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |