4M: Massively Multimodal Masked Modeling

🔋 Datasets with baselines for offline multi-agent reinforcement learning.

Scalable neural net training via automatic normalization in the modular norm.

Large language modeling applied to T-cell receptor (TCR) sequences.

Official implementation of Unique3D: High-Quality and Efficient 3D Mesh Generation from a Single Image

[NeurIPS'21 Outstanding Paper] Library for reliable evaluation on RL and ML benchmarks, even with only a handful of seeds.

DIAMOND (DIffusion As a Model Of eNvironment Dreams) is a reinforcement learning agent trained in a diffusion world model.

Repository for Meta Chameleon, a mixed-modal early-fusion foundation model from FAIR.

A Python framework for high performance GPU simulation and graphics

Scenic: A Jax Library for Computer Vision Research and Beyond

Image-to-Image Translation in PyTorch

Protein structure diffusion model for unconditional protein generation and motif scaffolding

maximal update parametrization (µP)

Neighborhood Attention Extension. Bringing attention to a neighborhood near you!

A memory profiler for Linux.

Simple implementation of muP, based on Spectral Condition for Feature Learning

ClickHouse® is a real-time analytics DBMS

Self-contained, minimalistic implementation of diffusion models with Pytorch.

Pax is a Jax-based machine learning framework for training large scale models. Pax allows for advanced and fully configurable experimentation and parallelization, and has demonstrated industry lead…

Minimal implementation of scalable rectified flow transformers, based on SD3's approach

A simple, performant and scalable Jax LLM!

Kolmogorov Arnold Networks

A minimal GPU design in Verilog to learn how GPUs work from the ground up