Files
P2-split1_prob_Qwen3-8B-Bas…/train_results.json

9 lines
231 B
JSON
Raw Normal View History

{
"epoch": 3.0,
"total_flos": 616169609822208.0,
"train_loss": 0.4388826183339422,
"train_runtime": 59766.2651,
"train_samples": 24704,
"train_samples_per_second": 1.24,
"train_steps_per_second": 0.013
}