qwen2.5-0.5b-expo-DPO-ES-TRY2 / train_results.json
hZzy's picture
Model save
7f977a6 verified
{
"epoch": 2.9976381672177608,
"total_flos": 0.0,
"train_loss": 0.023984534440167023,
"train_runtime": 3310.3799,
"train_samples": 50802,
"train_samples_per_second": 46.039,
"train_steps_per_second": 0.159
}