Qwen-2.5-3B-grpo-code / all_results.json
mlxha's picture
Model save
d055e09 verified
{
"total_flos": 0.0,
"train_loss": 0.001293145966497858,
"train_runtime": 17459.8588,
"train_samples": 35735,
"train_samples_per_second": 0.512,
"train_steps_per_second": 0.032
}