| { | |
| "best_metric": 72.5364, | |
| "best_model_checkpoint": "bin/indosum-base-0/checkpoint-4460", | |
| "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": 1.1983163356781006, | |
| "learning_rate": 0.0008, | |
| "loss": 1.2132, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 94.092, | |
| "eval_loss": 0.7741625905036926, | |
| "eval_rouge1": 67.4414, | |
| "eval_rouge2": 59.7409, | |
| "eval_rougeL": 64.517, | |
| "eval_rougeLsum": 66.4918, | |
| "eval_runtime": 191.8213, | |
| "eval_samples_per_second": 3.91, | |
| "eval_steps_per_second": 0.125, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.1038604974746704, | |
| "learning_rate": 0.0006, | |
| "loss": 0.686, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 100.29333333333334, | |
| "eval_loss": 0.6673194766044617, | |
| "eval_rouge1": 70.2138, | |
| "eval_rouge2": 62.8202, | |
| "eval_rougeL": 67.1553, | |
| "eval_rougeLsum": 69.3063, | |
| "eval_runtime": 191.8169, | |
| "eval_samples_per_second": 3.91, | |
| "eval_steps_per_second": 0.125, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.5924516916275024, | |
| "learning_rate": 0.0004, | |
| "loss": 0.491, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 100.944, | |
| "eval_loss": 0.6274383664131165, | |
| "eval_rouge1": 71.2142, | |
| "eval_rouge2": 63.9943, | |
| "eval_rougeL": 68.2722, | |
| "eval_rougeLsum": 70.2971, | |
| "eval_runtime": 193.0753, | |
| "eval_samples_per_second": 3.884, | |
| "eval_steps_per_second": 0.124, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.4989166855812073, | |
| "learning_rate": 0.0002, | |
| "loss": 0.343, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 98.82266666666666, | |
| "eval_loss": 0.6469114422798157, | |
| "eval_rouge1": 71.7114, | |
| "eval_rouge2": 64.489, | |
| "eval_rougeL": 68.7214, | |
| "eval_rougeLsum": 70.7949, | |
| "eval_runtime": 192.2303, | |
| "eval_samples_per_second": 3.902, | |
| "eval_steps_per_second": 0.125, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.6214805245399475, | |
| "learning_rate": 0.0, | |
| "loss": 0.2059, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 98.90533333333333, | |
| "eval_loss": 0.7169591784477234, | |
| "eval_rouge1": 72.5364, | |
| "eval_rouge2": 65.2519, | |
| "eval_rougeL": 69.5637, | |
| "eval_rougeLsum": 71.6884, | |
| "eval_runtime": 195.8892, | |
| "eval_samples_per_second": 3.829, | |
| "eval_steps_per_second": 0.123, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 4460, | |
| "total_flos": 4.883004680306688e+16, | |
| "train_loss": 0.5878115769458994, | |
| "train_runtime": 5026.0592, | |
| "train_samples_per_second": 14.188, | |
| "train_steps_per_second": 0.887 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4460, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 4.883004680306688e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |