dpo-selective-redteaming / train_results.json
wxzhang's picture
Model save
b1209ae verified
raw
history blame contribute delete
192 Bytes
{
"epoch": 0.99,
"train_loss": 5218.799974524457,
"train_runtime": 254.7215,
"train_samples": 1487,
"train_samples_per_second": 5.838,
"train_steps_per_second": 0.181
}