Skip to content

Latest commit

 

History

History
67 lines (50 loc) · 2.34 KB

README.md

File metadata and controls

67 lines (50 loc) · 2.34 KB

Jerry's AI Playground

This repo includes the tools based on OpenAI, LangChain and Streamlit to increase working efficiency.

  1. Auto Meeting Minutes is a tool to chat with video/audio and output the meeting minutes with build-in prompt.
  2. Casual Chat is a ChatGPT like ChatBot.
  3. FilesChat is a bot to chat with your own documents, supported format including PDF, PPTX, DOCX, TXT.
  4. AI translator is a chatbot to translate every words you input between Chinese and English.
  5. CSV Chatbot is a chatbot to analyse tabula data, still in developing... Sample Screenshots

1. How to use

  1. Install ffmpeg, ffmpeg official website.

  2. Install the required dependencies:

pip install -r requirements.txt
  1. Run the Streamlit application:
streamlit run .\home_page.py

2. Setup OpenAI and Azure OpenAI

To configure the application, you will need to create a key.txt file containing your Azure OpenAI API key and a config.json file with your desired settings.

key.txt

Create a file named key.txt and add your Azure OpenAI API key, BING_SUBSCRIPTION_KEY(optional) and AZURE_COGNITIVE_SEARCH_API_KEY(optional).

config.json (Only For Azure)

Create a config.json file with the following configuration:

{
    "CHATGPT_MODEL": "xxxxx",
    "OPENAI_API_BASE": "https://xxxxxx.openai.azure.com/",
    "OPENAI_API_VERSION": "xxxxx",
    "EMBEDDING_MODEL": "xxxxx",
    "EMBEDDING_MODEL_VERSION": "xxxxx",
    "BING_SEARCH_URL": "https://api.bing.microsoft.com/v7.0/search",
    "AZURE_COGNITIVE_SEARCH_SERVICE_NAME": "xxxxxxxxxxxx",
    "AZURE_COGNITIVE_SEARCH_INDEX_NAME": "xxxxxxxxxx"
}

3. How it works

3-1 Auto Meeting Minutes

  1. Ffmpeg is used to convert video to audio;
  2. Faster-Whisper is used to convert audio to subtitle;
  3. AgglomerativeClustering from sklearn is used to distinguish the speakers;
  4. OpenAI is used to do the meeting minute summary.

3-2 Casual Chatbot

  1. Azure OpenAI GPT3.5 as the server;
  2. LangChain is used to create the conversational Chatbot;
  3. Streamlit is used to create the web-ui

3-3 Files ChatBot

  1. Azure OpenAI as server and LangChain used to create a QA chatbot over documents.

3-4 CSV ChatBot

  1. Azure OpenAI as server and LangChain used to create a chatbot to create python code based on user's query, and give output.