Qwen2.5-7B-GRPO-NM-COT_2048 / all_results.json
Haitao999's picture
Model save
4a4c630 verified
raw
history blame
203 Bytes
{
"total_flos": 0.0,
"train_loss": 0.0007264473253309804,
"train_runtime": 13704.5305,
"train_samples": 85949,
"train_samples_per_second": 6.272,
"train_steps_per_second": 0.056
}