Qwen3-8B-math220k-run3 / train_results.json
HectorHe's picture
Model save
bc44d9f verified
{
"total_flos": 5328655832612864.0,
"train_loss": 0.3040030678062444,
"train_runtime": 78243.4815,
"train_samples": 93733,
"train_samples_per_second": 3.594,
"train_steps_per_second": 0.112
}