Enlighten AI

Dataset

    playlist_tag,
    playlist_url,
    video_url,
    channel_id,
    channel_url,
    video_title,
    video_length,
    video_publish_date,
    video_rating,
    video_views,
    video_author,
    video_keywords,
    video_description,
    video_transcript

TODO

• Identify Data source ( videos from heath gamer YouTube Channel)
• Use pytube + OpenAI Whisper to transcribe the videos
• Save the Daa into a csv with meta data
• Conduct Analysis on the Dataset
• Build an report on the dataset
• Design the A RAG based Chatbot
• Implement the knwoledge base data store (Melvius, Faiss, ChromaDB, ...)+ embeddings using Open Source model.
• Use Hugging face transformers library + LlamaIndex ( Llama 3 8B, Phi3 mini 128k, Llama mini, mistral 7B)
• Technical tests
• Working Tests
• priliminary deployment on the web.
• gather opinions to improve on the web.