Qwen-2.5-7B-Simple-GRPO / train_results.json
Mingsmilet's picture
Model save
10fa216 verified
raw
history blame contribute delete
200 Bytes
{
"total_flos": 0.0,
"train_loss": 0.004416576032875535,
"train_runtime": 7299.7333,
"train_samples": 7500,
"train_samples_per_second": 1.027,
"train_steps_per_second": 0.008
}