Files
Llama3.2-3B_Paper_Impact_me…/train_results.json

8 lines
209 B
JSON
Raw Normal View History

{
"epoch": 1.0,
"total_flos": 2.8024067250035098e+17,
"train_loss": 0.08477670932749667,
"train_runtime": 1255.2183,
"train_samples_per_second": 16.86,
"train_steps_per_second": 0.132
}