Files
qwen3-4b-beauty-mix/train_results.json

8 lines
210 B
JSON
Raw Normal View History

{
"epoch": 1.0,
"total_flos": 1.7136144572775137e+19,
"train_loss": 0.7275813773740559,
"train_runtime": 21314.9758,
"train_samples_per_second": 58.741,
"train_steps_per_second": 0.459
}