OpenRS-GRPO3 / train_results.json
xiwenc1's picture
Model save
270a755 verified
{
"total_flos": 0.0,
"train_loss": 0.007592763372914685,
"train_runtime": 38120.128,
"train_samples": 7000,
"train_samples_per_second": 0.63,
"train_steps_per_second": 0.013
}