"sgl-kernel/python/vscode:/vscode.git/clone" did not exist on "f200af0d8cde04ad746c37f51a40f7e218b6b581"
Commit d9254821 authored by chenych's avatar chenych
Browse files

update train loss

parent 037608eb
......@@ -32,6 +32,7 @@ GLM-4.1V-Thinking能够将图像和视频以其原始的分辨率和宽高比进
## 环境配置
### 硬件需求
DCU型号:K100_AI,节点数量:1台,卡数:1张。
`-v 路径``docker_name``imageID`根据实际情况修改
### Docker(方法一)
......@@ -104,7 +105,7 @@ pip install transformers==4.53.2
## 训练
### Llama Factory 微调方法(推荐)
因为transformers版本与[LLaMA-Factory](https://developer.sourcefind.cn/codes/OpenDAS/llama-factory)版本不一致, 启动训练前需要先增加环境变量来跳过版本检查,环境变量如下:
因为transformers版本与[LLaMA-Factory](https://developer.sourcefind.cn/codes/OpenDAS/llama-factory)版本不一致,启动训练前需要先增加环境变量来跳过版本检查,环境变量如下:
```
export DISABLE_VERSION_CHECK=1
......@@ -147,7 +148,7 @@ python inference/trans_infer_cli.py
| device | iters | train_loss |
| :------: | :------: | :------: |
| A800 | 375 | |
| A800 | 375 | 0.5245 |
| K100_AI | 375 | 0.5264 |
## 应用场景
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment