OpenRS-GRPO-sft / all_results.json
chunli-peng's picture
Model save
14a99d2 verified
{
"total_flos": 6.028549985560166e+17,
"train_loss": 1.1274345985717245,
"train_runtime": 1714.4302,
"train_samples": 7000,
"train_samples_per_second": 40.83,
"train_steps_per_second": 0.21
}