LLM-based YouTube Assistant for answering video-related questions. Users can ask questions about specific YouTube videos, and the assistant will provide relevant answers.
The code uses the CohereEmbeddings model ("embed-english-light-v3.0") for generating embeddings and Cohere LLM API for generating text based on the query. Get a Cohere API key and set it as an environment variable (COHERE_API_KEY).
Install the required packages:
pip install -r requirements.txt
Set Cohere API key in a .env file:
COHERE_API_KEY=your_api_key_here
Run the streamlit app:
streamlit run youtube_assistant.py