Zhang Weixiao
Zhang Weixiao
做RAG时如果能同时做语义搜索,这样既能看到AI具体查找结果的回答,也能看到相关内容在哪些文档的哪里出现过,感觉会好很多。
### Installation Method | 安装方法与平台 OneKeyInstall (一键安装脚本-windows) ### Version | 版本 Latest | 最新版 ### OS | 操作系统 Windows ### Describe the bug | 简述 bridge_qwen_local.py: def llm_stream_generator(self, **kwargs): #...
 it just keeps deleting. How can I mannually clear all the database? And could anyone tell me where the chat history is?
data = { "inputs": convert_lm_input_to_basic_string(input), "parameters": { "temperature": 0.0, "max_tokens": self._max_output_tokens, }, } Recommend this part of code be modified in custom_llm.py so that temperature can be set in .env...
I asked the same query both in search and chat tab but the answers are different. It seems in the search tab it could correctly return the relevant document chunks...
### 🥰 Feature Description Recommend allowing users to customize Serper API base since there are other endpoints available. ### 🧐 Proposed Solution add an environment variable in .env for users...
Tried litellm for hours but didn't work. Not so familiar with langchain. It would be better if some more examples (together_ai, openrouters) are provided. Customizable openai endpoint would be a...
## Description I am using docker. Calling the API and generate an answer would cause the websocket to go offline. ## Info in Chrome Unchecked runtime.lastError: The message port closed...
我看百炼模型里面有对Cosyvoice的支持了,要用到Dashscope,可以加入到字幕转音频功能里吗?
Is it possible to upload files like csv to analyze? What might be the basic roadmap to realize it?