-
Massachusetts Institute of Technology
- Cambridge, MA
- https://avik-pal.github.io
- @avikpal1410
Highlights
Stars
Fast and simple fluid simulator in Julia
Production-tested AI infrastructure tools for efficient AGI development and community-driven innovation
MLGym A New Framework and Benchmark for Advancing AI Research Agents
Legible, Scalable, Reproducible Foundation Models with Named Tensors and Jax
Home for "How To Scale Your Model", a short blog-style textbook about scaling LLMs on TPUs
A 1D analogue of the MNIST dataset for measuring spatial biases and answering Science of Deep Learning questions.
🧑🏫 60+ Implementations/tutorials of deep learning papers with side-by-side notes 📝; including transformers (original, xl, switch, feedback, vit, ...), optimizers (adam, adabelief, sophia, ...), ga…
Fast analytic raytracing around Kerr black holes
A retargetable MLIR-based machine learning compiler and runtime toolkit.
An AI-powered research assistant that performs iterative, deep research on any topic by combining search engines, web scraping, and large language models. The goal of this repo is to provide the si…
The simplest, fastest repository for training/finetuning medium-sized GPTs.
The Torch-MLIR project aims to provide first class support from the PyTorch ecosystem to the MLIR ecosystem.
Repository containing the EFfective Field theORy surrogaTe
Any model. Any hardware. Zero compromise. Built with @ziglang / @openxla / MLIR / @bazelbuild
Clean, minimal, accessible reproduction of DeepSeek R1-Zero
Enabling PyTorch on XLA Devices (e.g. Google TPU)
A concise but complete full-attention transformer with a set of promising experimental features from various papers
A template for writing CS papers with latex -- includes CI, todonotes, ...
JAX - A curated list of resources https://github.com/google/jax
a language for fast, portable data-parallel computation
Play atmospheric modelling like it's LEGO.
JAX implementation of OpenAI's Whisper model for up to 70x speed-up on TPU.
Official repository for reproducing the paper "Constants of motion network" (NeurIPS 2022)
An open-source, cross-platform terminal for seamless workflows
User-friendly AI Interface (Supports Ollama, OpenAI API, ...)
Documentation with Documenter.jl and VitePress