Qwen2.5-1.5B-GRPO / all_results.json
CjangCjengh's picture
Model save
8e6d464 verified
raw
history blame contribute delete
200 Bytes
{
"total_flos": 0.0,
"train_loss": 0.13987601281700246,
"train_runtime": 35127.6818,
"train_samples": 6565,
"train_samples_per_second": 0.934,
"train_steps_per_second": 0.117
}