Files
capsd-Qwen3-1.7B-Base-math_…/train_results.json

8 lines
200 B
JSON
Raw Permalink Normal View History

{
"epoch": 1.0,
"total_flos": 1655861018624.0,
"train_loss": 0.4739113450050354,
"train_runtime": 261.6357,
"train_samples_per_second": 7.644,
"train_steps_per_second": 0.122
}