dpo-selective-buffer-safeipo / train_results.json
wxzhang's picture
Model save
a2514bf verified
{
"epoch": 1.0,
"train_loss": 5859.617769083399,
"train_runtime": 32772.3871,
"train_samples": 120613,
"train_samples_per_second": 3.68,
"train_steps_per_second": 0.057
}