| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 4460, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.9550127983093262, | |
| "learning_rate": 0.0008, | |
| "loss": 3.1762, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 2.2190332412719727, | |
| "eval_rouge1": 0.7537, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.7549, | |
| "eval_rougeLsum": 0.7532, | |
| "eval_runtime": 54.3281, | |
| "eval_samples_per_second": 13.805, | |
| "eval_steps_per_second": 0.442, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.9039629101753235, | |
| "learning_rate": 0.0006, | |
| "loss": 2.6396, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 1.8944951295852661, | |
| "eval_rouge1": 0.6963, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6949, | |
| "eval_rougeLsum": 0.6934, | |
| "eval_runtime": 51.6713, | |
| "eval_samples_per_second": 14.515, | |
| "eval_steps_per_second": 0.464, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.2377656698226929, | |
| "learning_rate": 0.0004, | |
| "loss": 2.3694, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 1.6951366662979126, | |
| "eval_rouge1": 0.6912, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6912, | |
| "eval_rougeLsum": 0.6889, | |
| "eval_runtime": 51.91, | |
| "eval_samples_per_second": 14.448, | |
| "eval_steps_per_second": 0.462, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.069547414779663, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1722, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 1.5595394372940063, | |
| "eval_rouge1": 0.6883, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6851, | |
| "eval_rougeLsum": 0.6883, | |
| "eval_runtime": 55.5167, | |
| "eval_samples_per_second": 13.509, | |
| "eval_steps_per_second": 0.432, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.9467520713806152, | |
| "learning_rate": 0.0, | |
| "loss": 2.0114, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 1.0, | |
| "eval_loss": 1.4508676528930664, | |
| "eval_rouge1": 0.6893, | |
| "eval_rouge2": 0.0, | |
| "eval_rougeL": 0.6872, | |
| "eval_rougeLsum": 0.6862, | |
| "eval_runtime": 50.2723, | |
| "eval_samples_per_second": 14.919, | |
| "eval_steps_per_second": 0.477, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 4460, | |
| "total_flos": 5.530296634048512e+16, | |
| "train_loss": 2.4737384607973656, | |
| "train_runtime": 5670.119, | |
| "train_samples_per_second": 12.576, | |
| "train_steps_per_second": 0.787 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4460, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 5.530296634048512e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |