nash_dpo_merge_iter_6 / all_results.json
YYYYYYibo's picture
Model save
11ee063 verified
{
"epoch": 1.0,
"train_loss": 0.54004861034599,
"train_runtime": 9618.5725,
"train_samples": 26135,
"train_samples_per_second": 2.717,
"train_steps_per_second": 0.021
}