Files
llama3-8b-full-pretrain-was…/train_results.json

8 lines
201 B
JSON
Raw Normal View History

{
"epoch": 3.0,
"total_flos": 5124506050560.0,
"train_loss": 1.6660761659738874,
"train_runtime": 2303.3437,
"train_samples_per_second": 3.274,
"train_steps_per_second": 0.819
}