Heterogeneous GPU Sharing on Kubernetes

Application Kernel for Containers

This repository has implementations of various alternatives to backpropagation for training neural networks.

Code for SkipNet: Learning Dynamic Routing in Convolutional Networks (ECCV 2018)

nanoGPT: A minimal (nanomal?) repository containing code for building, training, and running nanoGPT

a streamlined pipeline for converting videos into 3D Gaussian Splatting models

Developing and Comparing Vision-based Algorithms for Vision-based Agile Flight

ROS counterpart for the FlightForge UAV simulator

Autonomous Flight Demo

A framework for photorealistic hardware-in-the-loop agile flight simulation using Unity3D and ROS. Developed by MIT AERA group.

WebGL 3D Gaussian Splat Viewer

Converts 3D Gaussian Splat (.ply / .spz) data to USD using the ParticleField3DGaussianSplat schema

PyBullet Gym environments for single and multi-agent reinforcement learning of quadcopter control

Codes accompanying the paper "RODE: Learning Roles to Decompose Multi-Agent Tasks (ICLR 2021, https://arxiv.org/abs/2010.01523). RODE is a scalable role-based multi-agent learning method which effe…

[RSS 2023] Diffusion Policy Visuomotor Policy Learning via Action Diffusion

Massively parallel rigidbody physics simulation on accelerator hardware.

A generative world for general-purpose robotics & embodied AI learning.

TensorDict is a pytorch dedicated tensor container.

Using DDPG and ConvLSTM to control a drone to avoid obstacle in AirSim

Developing and Comparing Vision-based Algorithms for Vision-based Agile Flight

This project involves autonomous navigation and inventory validation using the Swift Pico Drone in a simulated warehouse environment built with Gazebo. The drone utilizes ROS2 for real-time communi…

This project aims to transfer the data between two devices without any physical connection with the help of light.

A ros2 policy execution pipeline package that provides tools to deploy ONNX models

ncdu fork that can compare and diff results

slime is an LLM post-training framework for RL Scaling.

LLM Inference server

Writing & Making Operating System and Kernel parts so simple like Hello World Programs, Starting from writing Bootloaders, Hello World Kernel, GDT, IDT, Terminal, Keyboard/Mouse, Memory Manager, HD…

Small x86/ARM Operating System built from scratch