CiteChat is a helper for your academic journey, offering a unique and interactive approach to assisting with research. It turns PDF documents into knowledge resources and operates as a chatbot. Researchers can easily upload PDFs or search through Arxiv and SemanticScholar, then submit questions to CiteChat. By default, it uses an internet connection to Google Bard, but you also have the option of Llama-2 7B GGML (2 bits quantize) which can be run locally. With CiteChat, I hope your research becomes not only more accessible but also efficient.
If there's any error or bugs, feel free to open a discussion to report them.
Install all library using python venv
cd citechat
pip install -r requirements.txt
Put BardAPI tokens on config.yaml
, check how to get Bard tokens
# Bard Tokens
__Secure-1PSID : "Put your token here"
__Secure-1PSIDCC : "Put your token here"
__Secure-1PSIDTS : "Put your token here"
To run the app
python app.py
CTRL + C
in your terminal to close the app and the port
- Go to Google Bard
- Right click, click Inspect or
Ctrl + Shift + C
- Click
>>
symbol and selectApplication
- Search
__Secure-1PSID
and copy toconfig.yaml
all the tokens you need