Files
HT-ht-analysis-Qwen-instruc…/all_results.json

8 lines
203 B
JSON
Raw Normal View History

{
"epoch": 3.0,
"total_flos": 154685539491840.0,
"train_loss": 0.3971013587030393,
"train_runtime": 54304.2861,
"train_samples_per_second": 4.076,
"train_steps_per_second": 0.17
}