Qwen3-8B-math220k-run4 / train_results.json
HectorHe's picture
Model save
faa0fd8 verified
{
"total_flos": 5359601230217216.0,
"train_loss": 0.3001025055014911,
"train_runtime": 79773.54,
"train_samples": 93733,
"train_samples_per_second": 3.525,
"train_steps_per_second": 0.11
}