Files
qwen2.5-3b-dpo-mini/train_results.json

9 lines
222 B
JSON
Raw Normal View History

{
"epoch": 1.9999360981532366,
"total_flos": 0.0,
"train_loss": 0.0,
"train_runtime": 42.011,
"train_samples": 1001516,
"train_samples_per_second": 23839.372,
"train_steps_per_second": 186.237
}