Qwen-2.5-7B-GRPO-Base-1Action_991 / train_results.json
luckeciano's picture
Model save
291b166 verified
{
"total_flos": 0.0,
"train_loss": -0.5921875196695328,
"train_runtime": 17144.8933,
"train_samples": 7500,
"train_samples_per_second": 0.56,
"train_steps_per_second": 0.006
}