Skip to content
View inclinedadarsh's full-sized avatar
🚆
Shipping
🚆
Shipping
114 followers · 32 following

Achievements

Achievement: Pull Sharkx3Achievement: YOLOAchievement: StarstruckAchievement: Pair ExtraordinaireAchievement: Quickdraw

Achievements

Achievement: Pull Sharkx3Achievement: YOLOAchievement: StarstruckAchievement: Pair ExtraordinaireAchievement: Quickdraw

Highlights

  • Pro

Block or report inclinedadarsh

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
inclinedadarsh/README.md

Hey there 👋, I'm Adarsh Dubey

Building ML infrastructure and full-stack applications. Recently wrapped up GSoC 2025 with Google DeepMind.

🚀 What I Do

I work at the intersection of machine learning and software engineering — building tools that make ML more accessible and creating robust full-stack applications. Currently focused on fine-tuning infrastructure, deep learning from scratch, and scalable web platforms.

🔨 Recent Work

Gemma Facet

  • No-code platform for fine-tuning Google DeepMind's Gemma models
  • Handles dataset preprocessing, model training, and exports to multiple formats (GGUF, quantized, etc.).
  • Full pipeline from data upload to deployment-ready models.
  • Built during GSoC 2025.

IncliArray:

  • N-dimensional array library with autograd in C++
  • Minimal implementation with reverse-mode automatic differentiation, broadcasting, and stride-aware memory management.
  • Think NumPy meets autograd, but in C++.

Inclinet:

  • Deep learning library from scratch
  • Pure Python/NumPy implementation of neural networks.
  • Layers, optimizers, loss functions, all built from first principles.

Anon Forum:

  • Anonymous social platform for my college
  • Full-stack app with Google OAuth verification but zero personal data storage.
  • FastAPI backend, Next.js frontend, deployed on AWS.

💼 Experience

Open Source Developer @ Google DeepMind (GSoC 2025)
Built core microservices for dataset preprocessing and model export pipelines. Worked on the Gemma Facet project, creating tools used by the ML community for fine-tuning multimodal models.

🛠️ Tech Stack

Languages: Python, C/C++, JavaScript/TypeScript
ML/AI: PyTorch, Keras, Hugging Face, Unsloth
Web: Next.js, React, FastAPI, Node.js
Databases: PostgreSQL, MongoDB
Cloud & Tools: Docker, GCP, AWS, Git

📝 Writing

I write about ML, web development, and building stuff at adarshdubey.com

Recent posts:

📫 Connect

Currently pursuing B.Tech in AI & Data Science at K. K. Wagh Institute • Open to interesting projects and collaborations

Pinned Loading

  1. gemma-facet/cloud-services gemma-facet/cloud-services Public

    Supabase for LLM fine-tuning -- go from idea to deployed model in hours, not days (Google Summer of Code '25)

    Python 11 4

  2. inclinet inclinet Public

    Implementation of deep learning library from scratch in Python and Numpy

    Python 8

  3. incliarray incliarray Public

    Minimal NumPy-like N-dimensional float array library in C++.

    C++ 6

  4. helix helix Public

    L1 Cache for your LLM!

    TypeScript 10 1

  5. anon anon Public

    TypeScript 6 2

  6. gemma-facet/frontend-ui gemma-facet/frontend-ui Public

    Customize and manage multimodal datasets and fine tuning for DeepMind's Gemma SLM (GSoC '25)

    TypeScript 6 4