Qwen2.5-1.5B-Open-R1-GRPO / train_results.json
zyl2023's picture
Model save
22f038f verified
{
"total_flos": 417145578061824.0,
"train_loss": 0.47688108040986815,
"train_runtime": 15636.0619,
"train_samples": 93733,
"train_samples_per_second": 29.973,
"train_steps_per_second": 0.234
}