qwen2.5-0.5b-expo-L1EXPO-noES-0.1 / train_results.json
hZzy's picture
Model save
fc395d0 verified
raw
history blame
233 Bytes
{
"epoch": 2.992914501653283,
"total_flos": 0.0,
"train_loss": 0.08480868444806247,
"train_runtime": 47353.1169,
"train_samples": 50802,
"train_samples_per_second": 3.218,
"train_steps_per_second": 0.022
}