lin-jun-xiang/docgpt-langchain

🔐Free GPT-3.5 chat with your docs (PDF, WORD, CSV, TXT)

docgpt-langchain's Issues

ش

試了很多provider 只有chatbase能用

New Feature

Issue you'd like to raise.

Hi there! I wanted to make changes in it with which the uploaded documents stay in the memory and we get an option to later on add more delete from previous documents. How can that be done?
thanks.

Suggestion:

No response

RetrievalQA costs long time to get the answer

Problem

The langchain retrievalQA will costs long time to get the answer.

Solution

1. Use streaming to solve. (Langchain Issue)
2. Use streamlit_chat to response streamling. But this feature has not been implemented. (Streamlit Issue)

Resource limit: Oh No, Error running app.

Problem

1. The free model will use HuggingFaceEmbeddings which cause lots of memory.
2. The streamlit free server just have 1GB RAM.
3. When there have multiple users using free model in the same time, our app will "over the resource limits".

Reference

• Streamlit app resource limit
• Same problem in streamlit forum

Suggestion

1. improve streamlit app performance
2. RAM in streamlit
3. Select a lower embedding model of HuggingFace

Run script without streamlit

Issue you'd like to raise.

Hey. I was wondering if there was a way to run the app.py script without need for streamlit. I want to do that so I can run it on my own program/website later.

Suggestion:

No response

crashed

docker deployment error

Issue you'd like to raise.

After deploying with docker, it prompts: /bin/sh: 1: [streamlit,: not found

Suggestion:

No response