CAR:A chat robot with any fields of given knowledge
pip+virtualenv
-
Clone the project repository:
git clone https://github.com/Bili-Sakura/Chat-Anything-Robot.git cd Chat-Anything-Robot
-
Create and activate virtual environment:
Recommended python version is 3.8.x
python -m venv .venv/car # Create a virtual environment named car, located in .venv source .venv/car/bin/activate # For Unix-like systems .\.venv\car\Scripts\activate # For Windows
-
Install dependencies:
pip install -r requirements.txt
-
Deactivate the virtual environment:
source .venv/car/bin/deactivate # For Unix-like systems .\.venv\car\Scripts\deactivate # For Windows
conda
-
Create and activate conda environment:
conda create --name car python=3.8 conda activate car
-
Install dependencies:
conda install --file requirements.txt
-
Deactivate the conda environment:
conda deactivate
Follow the steps below to prepare your OpenAI API key:
- Sign up for an account on OpenAI's website.
- Generate an API key in your account settings.
- Copy the API key to use in your application.
It is avaliable to use api key transfered from supported website. Set base url parameter whenever you use.
cd Chat-Anything-Robot
run.bat # which equals to [cmd].venv\car\Scripts\activate + [cmd]streamlit run app.py
Congratulations! You've start your own program!
Put files (*.docx, *.txt, et al.) under the data
folder.
Re-init the program, the corpus will be added into vectorstore automatically (may cost time in the first time).
- Add button for vercterization
- Add cost log info locally
- Add knowledge info in sidebar
- Asynchronously chat
- Web Application Project Structure