qwen2.5-0.5b-expo-DPO-W0-noES5-1 / train_results.json
hZzy's picture
Model save
9068903 verified
{
"epoch": 2.992914501653283,
"total_flos": 0.0,
"train_loss": 275.78432337443036,
"train_runtime": 34690.4645,
"train_samples": 50802,
"train_samples_per_second": 4.393,
"train_steps_per_second": 0.03
}