Qwen2.5-0.5B-Open-R1-GRPOmemory_combine-20-2000-0/train_results.json

{
    "total_flos": 0.0,
    "train_loss": 0.006824205948613517,
    "train_runtime": 18399.2985,
    "train_samples": 2000,
    "train_samples_per_second": 0.435,
    "train_steps_per_second": 0.013
}