GRPO-1.5B-Format-Old / all_results.json
LLucass's picture
Training in progress, step 250
df178ca verified
{
"total_flos": 0.0,
"train_loss": 0.00036543175232403515,
"train_runtime": 19061.8828,
"train_samples": 7000,
"train_samples_per_second": 1.259,
"train_steps_per_second": 0.013
}