nash_dpo_rank4_iter_plus_3 / train_results.json
YYYYYYibo's picture
Model save
d6acae0 verified
raw
history blame contribute delete
183 Bytes
{
"epoch": 1.0,
"train_loss": 0.0,
"train_runtime": 0.0096,
"train_samples": 5000,
"train_samples_per_second": 521173.985,
"train_steps_per_second": 4065.157
}