Files
qwen3-8b-sft/all_results.json

12 lines
346 B
JSON
Raw Normal View History

{
"epoch": 1.0,
"eval_loss": 0.06474341452121735,
"eval_runtime": 487.1248,
"eval_samples_per_second": 4.2,
"eval_steps_per_second": 0.302,
"total_flos": 3995069728161792.0,
"train_loss": 0.10921624170816473,
"train_runtime": 69121.8578,
"train_samples_per_second": 0.955,
"train_steps_per_second": 0.068
}