qwen2.5-0.5b-expo-DPO-ES-TRY / train_results.json
hZzy's picture
Model save
9ff305a verified
raw
history blame contribute delete
234 Bytes
{
"epoch": 0.8502598016060463,
"total_flos": 0.0,
"train_loss": 0.15444423039754232,
"train_runtime": 3628.8607,
"train_samples": 50802,
"train_samples_per_second": 41.998,
"train_steps_per_second": 0.583
}