ymcui / Chinese-LLaMA-Alpaca

中文LLaMA&Alpaca大语言模型+本地CPU/GPU训练部署 (Chinese LLaMA & Alpaca LLMs)
https://github.com/ymcui/Chinese-LLaMA-Alpaca/wiki
Apache License 2.0
18.23k stars 1.86k forks source link

训练保存checkpoint之后如何完成推理、评估 #877

Closed char-con closed 8 months ago

char-con commented 9 months ago

提交前必须检查以下项目

问题类型

模型推理

基础模型

LLaMA-7B

操作系统

Linux

详细描述问题

[```] 在lora训练完成以后保存了checkpoint,如何进行加载推理,还需要再次和原来的lora还有basemodel合并吗? 0LEY~7)CU96~0B)@WCU46 8

又或者如同此图吗?


### 依赖情况(代码类问题务必提供)

请在此处粘贴依赖情况



### 运行日志或截图

_No response_
github-actions[bot] commented 8 months ago

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your consideration.

github-actions[bot] commented 8 months ago

Closing the issue, since no updates observed. Feel free to re-open if you need any further assistance.