qwen2.5-0.5b-expo-DPO-ES-0.01 / train_results.json
hZzy's picture
Model save
9f7a057 verified
raw
history blame
232 Bytes
{
"epoch": 0.8502598016060463,
"total_flos": 0.0,
"train_loss": 0.6054576412836711,
"train_runtime": 8287.5068,
"train_samples": 50802,
"train_samples_per_second": 30.65,
"train_steps_per_second": 0.212
}