Upstage Project

A JAX-based simulator for autonomous driving research.

Dataset to assess the disentanglement properties of unsupervised learning methods

🏛️A research-friendly codebase for fast experimentation of single-agent reinforcement learning in JAX • End-to-End JAX RL

DreamerV3 implementation of Curious Replay, a method for prioritizing experience replay that is tailored to model-based reinforcement learning agents.

megastep helps you build 1-million FPS reinforcement learning environments on a single GPU

TensorFlow's Visualization Toolkit

🕹️ A diverse suite of scalable reinforcement learning environments in JAX

An open source toolkit for Distributed Deep Reinforcement Learning on real and simulated robots.

🤗 LeRobot: End-to-end Learning for Real-World Robotics in Pytorch

PIX is an image processing library in JAX, for JAX.

JAX-accelerated Meta-Reinforcement Learning Environments Inspired by XLand and MiniGrid 🏎️

A collection of high-quality models for the MuJoCo physics engine, curated by Google DeepMind.

OpenSpiel is a collection of environments and algorithms for research in general reinforcement learning and search/planning in games.

MuJoCo simulation code

[ICLR 2023] Choreographer: a model-based agent that discovers and learns unsupervised skills in latent imagination, and it's able to efficiently coordinate and adapt the skills to solve downstream …

Differentiable Rasteriser implemented in JAX. Reference: https://github.com/erwincoumans/tinyrenderer, https://github.com/ssloy/tinyrenderer/wiki; PR: https://github.com/google/brax/pull/367

common in-memory tensor structure

Differentiable Rasteriser implemented in JAX. Reference: https://github.com/erwincoumans/tinyrenderer, https://github.com/ssloy/tinyrenderer/wiki; PR: https://github.com/google/brax/pull/367

A massively parallel, high-level programming language

Evaluation Code repository for the paper "ModuLoRA: Finetuning 3-Bit LLMs on Consumer GPUs by Integrating with Modular Quantizers". (2023 TMLR Submission)

Finetuning Large Language Models on One Consumer GPU in Under 4 Bits

Tile primitives for speedy kernels

ROS interface for MuJoCo simulator

An open-source reinforcement learning environment build toolkit for ROS and Gazebo.