dpo-selective-buffer-spo-shift / train_results.json
wxzhang's picture
Model save
5d84c8e verified
raw
history blame contribute delete
193 Bytes
{
"epoch": 1.0,
"train_loss": 67.04043597990268,
"train_runtime": 46860.0347,
"train_samples": 59478,
"train_samples_per_second": 1.269,
"train_steps_per_second": 0.04
}