ymcui / Chinese-LLaMA-Alpaca-2

中文LLaMA-2 & Alpaca-2大模型二期项目 + 64K超长上下文模型 (Chinese LLaMA-2 & Alpaca-2 LLMs with 64K long context models)
Apache License 2.0
7.01k stars 571 forks source link

Model 在 RAG推理中不断重覆生成相同的tokens #476

Closed williamyu20 closed 6 months ago

williamyu20 commented 7 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

Chinese-Alpaca-2-16K (7B/13B)

操作系统

Linux

详细描述问题

用Langchain 搭建了一个RAG,并且给model一些從Chatgpt 3.5 生成假的银行开帐户的资料,用llama.cpp作推理。看了Langchain retrieval的output 是没有任何问题的,能output相对应的资料,但model的回应却一直出现重复回覆相同tokens的情况,即使调节了repetition_penalty也没用。这是这个model原本的缺陷吗? model用的是ggml-model-q8_0.gguf,以下是retrieval 得到的output和model的回应。

image

image

依赖情况(代码类问题务必提供)

No response

运行日志或截图

No response

ymcui commented 7 months ago

直接用llama.cpp做推理是否有这个问题?以此来判断是哪个环节出了问题。

github-actions[bot] commented 6 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 6 months ago

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