ℹ️: This App has been currently disabled as I am currently out of the credits for LLM model vendors. Also I am working out on a better way to extract insights from research papers and Scientific Q&A, will restart this in some time. # Research Buddy A research buddy which helps you asks questions on certain research papers, get insights on top research papers. * Team:- Bhavish Pahwa :surfing_man: ([GitHub](https://github.com/bp-high)) ([LinkedIn](https://www.linkedin.com/in/bhavishpahwa/)) * Video:- https://www.youtube.com/watch?v=c3Ae7DBuE-U * Streamlit Community Cloud:- https://gpt-research-buddy.streamlit.app/ * Most Updated versions can be found on HuggingFace Spaces:- https://huggingface.co/spaces/bpHigh/AI-Research-Buddy # System Design ![Streamlit](https://github.com/bp-high/research_buddy/assets/53102161/2dd2e37b-4711-4dfd-bde2-e80360fe5513) # Tech Stack:- - Platforms:- ClarifAI, Modal - Frameworks:- Streamlit, LLamaIndex, Langchain(for integration with ClarifAI), - Models:- GPT-3.5-turbo, OpenAI Embeddings, Nougat Transformer(https://facebookresearch.github.io/nougat/) and ClarifAI multilingual moderation model # Usage Videos * How to get started with the App:- https://github.com/bp-high/research_buddy/assets/53102161/c0661ee2-2de4-4932-b1ab-37447fb1a448 * How to use Chatbot to get answers to questions https://github.com/bp-high/research_buddy/assets/53102161/8d5e51fb-91c7-46e1-86d6-d7a36377285a * How the app identifies obscene, toxic, hateful questions and intervenes whenever required https://github.com/bp-high/research_buddy/assets/53102161/da42b75f-e94c-4516-bddc-81b786b75b0f