Tutor++ is a powerful AI-driven Q&A framework that seamlessly queries local repositories, pinpointing answers within documents and videos with unmatched precision.
Tutor++ harnesses the prowess of the state-of-the-art CPM-Bee model with 100 billion parameters and integrates the Langchain large model framework, offering exceptional performance in both English and Chinese. Deployed on Amazon AWS, this advanced AI Q&A indexing system utilizes the FAISS vector database for swift searches, processing real-time queries by delving deep into local, user-defined repositories spanning various media formats, from PDFs to videos. Its unparalleled ability to pinpoint exact locations within documents and videos for answers is groundbreaking. Coupled with its user-centric design, expansive functionality, and superior adaptability, Tutor++ epitomizes the zenith of automated knowledge extraction and retrieval.
- python 3.8 +
- langchain
- boto3
- streamlit
- fastapi
git clone https://github.com/mrn3088/tutor-plus-plus.git
pip install -r requirements.txt
streamlit run app.py
@Xiu Yuan
@Ruinan Ma
@Bowen Yao
@Zeming Luo
@Yiqing Dai
Inspiration, code snippets, etc.