dpo-selective-alpaca / train_results.json
wxzhang's picture
Model save
fb2ade4 verified
{
"epoch": 1.0,
"train_loss": 4602.791315297503,
"train_runtime": 29489.6101,
"train_samples": 297394,
"train_samples_per_second": 10.085,
"train_steps_per_second": 0.158
}