Teaching Assistant 2.0 using WizardLM

• Upload any video and chat with its content. Also gives you the source. (For ex: {video_name}-{min_5_to_10})
• This implementation only uses opensource models and tools.
• Microservices hosted on the AWS and GPU services (embedding generation and LLM inference) are hosted on Lambda labs.
• This project uses instructor-xl model for embedding db generation and WizardLM LLM for chat.

Architecture:

Screenshots:

• I uploaded one of our Cloud Computing lectures which was about common terminologies used in Cloud Computing.

• Chat interface:

Acknowledgement:

• Thank you Keshav for helping out with frontend and designing.:relaxed: