Files
qwen3-14b-fft-if/train_results.json

8 lines
209 B
JSON
Raw Normal View History

{
"epoch": 2.0,
"total_flos": 1.3931784786122113e+19,
"train_loss": 0.5872682382094778,
"train_runtime": 12619.0484,
"train_samples_per_second": 8.865,
"train_steps_per_second": 0.069
}