Using ChatGPT to query your videos - ChatGPT, Chroma, LangChain, Python - Tutorial

preview_player
Показать описание
In this tutorial, we will explore how to use OpenAI's Whisper API to query specific youtube videos using Python, ChatGPT, OpenAI's APIs, Chroma and LangChain. To get started, we will build a Python script that downloads the video and chunks it down. With the video audio data we use Whisper API to get the actual content.
From there on we will create the embeddings based on the chunks. The distance between the query vector and the chunks is factor on which the model answers.

Saving the embeddings to a vector database allows us to
query our own data without being limited to pre-trained models.

The usage of LangChain is another way to ask detailed questions on the merged chunks of text.

The possibilities with this technology are endless! What's your opinion on the AI technology?

Timestamps:

00:00 introduction and recap of my previous videos
01:00 technology stack I use for this video
01:39 code explanation
06:10 results of the script
07:58 hint on how to change the code to get more results of where you have to start watching
08:30 outro

You can access the tutorial on Github at

My previous videos:

Рекомендации по теме
Комментарии
Автор

Awesome video! Thanks for doing a tutorial on this, I want to learn more about LangChain

MatthiasMayer
Автор

Wow this is excellent and as an entrepreneur and non-developer I see many potential applications for this. Would love to chat to you further

henripereira
welcome to shbcf.ru