初始化项目,由ModelHub XC社区提供模型

Model: W-61/llama-3-8b-base-ultrachat-sft-4xh100
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-04-27 21:03:30 +08:00
commit 5aa224cf18
26 changed files with 414547 additions and 0 deletions

9
train_results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"epoch": 0.9994655264564404,
"total_flos": 2.759237889794507e+18,
"train_loss": 1.1061673424460672,
"train_runtime": 11435.0508,
"train_samples": 207864,
"train_samples_per_second": 10.471,
"train_steps_per_second": 0.082
}