Files
llama3-8b-full-pretrain-was…/all_results.json
ModelHub XC 259b0c5548 初始化项目,由ModelHub XC社区提供模型
Model: shuoxing/llama3-8b-full-pretrain-wash-c4-0-6m-bs4
Source: Original Platform
2026-06-12 17:14:16 +08:00

8 lines
201 B
JSON

{
"epoch": 3.0,
"total_flos": 2578979020800.0,
"train_loss": 1.6905082198137384,
"train_runtime": 1266.0882,
"train_samples_per_second": 2.945,
"train_steps_per_second": 0.737
}