Qwen2.5-1.5B-R1-GRPO-debug / train_results.json
laolaorkk's picture
Model save
ec36b72 verified
raw
history blame
199 Bytes
{
"total_flos": 0.0,
"train_loss": 4.6566128730773926e-09,
"train_runtime": 1048.9444,
"train_samples": 10,
"train_samples_per_second": 0.01,
"train_steps_per_second": 0.001
}