qwen2.5-0.5b-expo-DPO-noES3-0.1 / train_results.json
hZzy's picture
Model save
7894361 verified
raw
history blame contribute delete
232 Bytes
{
"epoch": 1.995276334435522,
"total_flos": 0.0,
"train_loss": 0.3641128831289031,
"train_runtime": 23012.8835,
"train_samples": 50802,
"train_samples_per_second": 4.415,
"train_steps_per_second": 0.031
}