Files
DeepSeek-R1-Distill-Qwen-1.5B/all_results.json

9 lines
231 B
JSON
Raw Normal View History

{
"epoch": 3.0,
"total_flos": 1.847238457911296e+16,
"train_loss": 0.4201909417197818,
"train_runtime": 1415.4735,
"train_samples": 1,
"train_samples_per_second": 2.119,
"train_steps_per_second": 0.134
}