Skip to content

Instantly share code, notes, and snippets.

@YASH-GU24

YASH-GU24/GSoC work.md

Created September 12, 2022 08:26
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save YASH-GU24/9ce35e5c40ac8a2efbcaa1989a23da05 to your computer and use it in GitHub Desktop.
Save YASH-GU24/9ce35e5c40ac8a2efbcaa1989a23da05 to your computer and use it in GitHub Desktop.
Download ZIP
This gist contains Information about the work done by me as a part of GSoC 2022
Raw
GSoC work.md

Project Description:-

This project aims to make more examples and tutorials for weaviate. As any beginner who wants to use weaviate, These examples and tutorials are the best way to start learning. So it's better that we have a good number of these examples. The project can use various datasets available at sites like Kaggle.com or https://paperswithcode.com/datasets. These examples will not only help the beginners but also help people who already are using weaviate as by watching these they can come up with a new similar idea of their own to use weaviate.

Demos made as a part of GSoC:

  • Movie Search Engine:- In this demo I used the dataset of over 48,000+ movies scraped from the IMBD website. The demo had various functionalities like filtered searching or semantic searching. We could also sort the results on the basis of title, rating, etc.

    The PR link of the demo can be found here:- weaviate/weaviate-examples#22

  • Audio Genre Classifier:- This Demo demonstrates the usage of the weaviate img2vec module, Which is a module that converts images to vectors using a neural network and then allows us to perform various operations. The dataset consisted of 10 music genre classes each having 100 spectrogram images. We first load spectrograms of all the classes into weaviate, Then perform image classification on a new spectrogram to find its genre.

    The PR link of the demo can be found here:- weaviate/weaviate-examples#29

  • Caption Search:- This is a demo example to show how to perform caption search using weaviate. We can load captions of any YouTube video and then perform a Q&A search on its captions. After that, we will get the timestamp of the particular spot where the particular question is asked. This example used youtube-transcript-api to fetch captions.

    The PR link of the demo can be found here:- weaviate/weaviate-examples#28

  • Improvements in existing examples:- There were improvements made in 6 previous examples. The improvements mainly consisted in modifying the docker files and changing the process of data importing.

    The PR link of the demo can be found here:- weaviate/weaviate-examples#29

All the work done in the above demos was merged in this repository: https://github.com/semi-technologies/weaviate-examples

@TanmayAgarwal123
Copy link

I would love to work on this project, can you assign me
Regards,
Tanmay Agarwal

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment