Skip to content

Create a chatbot for a YouTube video content using LangChain framework and LLM from Hugging Face

License

Notifications You must be signed in to change notification settings

Nidheesh-Panchal/YouTubeSearch

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
 
 
 
 
 
 
 
 

Repository files navigation

YouTubeSearch

Using the Machine Learning Engineering for Produciton (MLOps) playlist from the channel DeepLearning.AI, instructed by Andrew Ng.

Create a chatbot that allows the user to search for specific questions about the topics covered in the video without trying to search for any description or specific time stamp to watch the video again.

The chatbot is created using LangChain framework which uses google/flan-t5-large LLM and sentence-transformers/all-MiniLM-L6-v2 embeddings model from Hugging Face, and Qdrant vector store.

The notebook with chatbot created is located at "./src/notebooks/YouTubeSearch.ipynb"

Run:

Change the value of "query" variable to get answer. The source of the answer is also returned by the conversation chain.

Screenshot

Image

Future work

The current project carries no-cost, but can be integrated with OpenAI GPT 3.5 turbo to have much better interpretation of questions and well formed answers. The chatbot can be hosted on Hugging Face Hub or a server along with a UI for better experience. The vector store can be further extended with ability to add more youtube video links.

About

Create a chatbot for a YouTube video content using LangChain framework and LLM from Hugging Face

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published