Qwen3-0.6B-Distill_001 / train_results.json
jeehwon's picture
Model save
b398752 verified
{
"total_flos": 241025679360.0,
"train_loss": 1.7981508374214172,
"train_runtime": 27.6272,
"train_samples": 1000,
"train_samples_per_second": 36.196,
"train_steps_per_second": 0.29
}