qwen2.5-0.5b-expo-DPO-ES-TRY0 / all_results.json
hZzy's picture
Model save
a1a29fe verified
raw
history blame
233 Bytes
{
"epoch": 0.8162494095418045,
"total_flos": 0.0,
"train_loss": 0.6411769655015733,
"train_runtime": 5149.1432,
"train_samples": 50802,
"train_samples_per_second": 19.732,
"train_steps_per_second": 0.068
}