qwen_cpo_entropy_0_01 / train_results.json
yakazimir's picture
Model save
4309889 verified
raw
history blame contribute delete
232 Bytes
{
"epoch": 2.999297541394882,
"total_flos": 0.0,
"train_loss": 0.5516859670777903,
"train_runtime": 39088.5646,
"train_samples": 59790,
"train_samples_per_second": 4.589,
"train_steps_per_second": 0.143
}