Why Google Summer Of Code?

Manish Kumar Thota
4 min readMay 29, 2024

--

Google Summer of Code (GSoC) is a global, online program that brings new contributors into open-source software development. GSoC Contributors work with an open-source organization on a 12+ week programming project under the guidance of mentors.

It is not an internship or a job but much more than that. It offers participants the unique opportunity to gain firsthand experience by either contributing to an existing GSoC project by developing new features or creating a Minimum Viable Product (MVP) from scratch, which can later be developed into a full product.

Open source is at the heart of innovation. Imagine if we didn’t have Hugging Face and other open-source large language models and resources; we would likely remain mere consumers of closed models like ChatGPT. Thanks to the open-source community, with contributions from individuals like Andrej Karpathy with llama.cpp and the development of LLaMA models, open source continues to drive technological advancements.

Google Summer of Code is at the forefront of these developments, pushing the boundaries of open source further by organizing GSoC every year. It fosters a community of contributors who play a crucial role in the ongoing evolution of technology.

Acceptance Rate

Celebrating 20th anniversary of Google Summer of Code (GSoC)

Highlighting significant results from this year’s application period:

  • 43,984 applicants from 172 countries
  • 9,107 proposals submitted by 6,518 applicants
  • 1,220 GSoC contributors accepted from 73 countries
  • Over 2,800 mentors and organization administrators
  • 34 mentoring organizations are participating in their 16th-20th GSoC!

So, approximately 2.77% of the applicants were accepted into Google Summer of Code 2024.

GSoC’2024

How do I get into GSoC?

The information is just one click away, with numerous resources available online on how to get into one of the organizations. However, in this blog, I will share my personal journey. How did I apply, what steps did I take, and how did I succeed?

Typically, people start preparing 2–3 months prior to applying. They choose the organization and the tech stack they want to work with, and they reach out to mentors and project stakeholders to discuss and establish common grounds on the project perspective. It’s important to start writing proposals early; don’t wait for the deadline. Write an early draft and share it with your prospective mentor to get feedback. Early reviews can put you ahead of others and work in your favor.

Once you have a clear understanding of the project requirements, it’s just a matter of finalizing your proposal. Make sure to give your 100% when writing the proposal. It should not only be aesthetically pleasing but also showcase your technical skills and genuine interest in the project.

What Project Will I Be Working On?

Organization: Red Hen Lab

Context: Imagine you need to find a specific video from among many saved on your system. These videos are scattered across multiple drives, making it time-consuming and frustrating to locate the one you need.

Objective: This project aims to develop an innovative system to enhance video annotation capabilities. By integrating a multimodal vision and language model with spatial-temporal analysis, we will automatically tag videos with relevant entities based on their content. This tagging system will make it significantly easier and faster to retrieve the videos you need.

Approach:

  • Model Utilization: We will use the LLaVA-v1.5–13b model with Video Adapter, known for its rapid inference and entity detection in video frames.
  • Annotation Extraction: The system will automatically extract annotations from videos and compile them into a user-friendly CSV format.
  • Core Functionality: A Pydantic API will extract boolean values for annotation entities from enhanced language model responses, enabling the concurrent processing of multiple videos.
  • User Interface: The entire workflow, from video and JSON input to CSV output, will be accessible via a Gradio interface on the Hugging Face platform. This ensures ease of use and broad accessibility.

By implementing this solution, we can save time and improve the organization of video files, making it much easier to manage and retrieve the content you need.

The official coding period has begun, and it’s going to be exciting to contribute! There’s much more to share in the coming weeks, so stay tuned for future blog updates.

References:

https://summerofcode.withgoogle.com/

https://summerofcode.withgoogle.com/programs/2024/projects/rYosJtqh

https://developers.google.com/open-source/gsoc/help/student-advice

--

--