DeepGEMM: clean and efficient FP8 GEMM kernels with fine-grained scaling

Expert Parallelism Load Balancer

A bidirectional pipeline parallelism algorithm for computation-communication overlap in V3/R1 training.

Production-tested AI infrastructure tools for efficient AGI development and community-driven innovation

Analyze computation-communication overlap in V3/R1.

Integrate the DeepSeek API into popular softwares

DeepSeek-VL2: Mixture-of-Experts Vision-Language Models for Advanced Multimodal Understanding

🤗 LeRobot: Making AI for Robotics more accessible with end-to-end learning

Retrieval and Retrieval-augmented LLMs

Barkour Robot: Agile Quadruped Robots by Google DeepMind

PantoMatrix: Co-Speech Talking Head and Gestures Generation

VILA - a multi-image visual language model with training, inference and evaluation recipe, deployable from cloud to edge (Jetson Orin and laptops)

Mobile-Agent: Autonomous Multi-Modal Mobile Device Agent with Visual Perception

Universal Manipulation Interface: In-The-Wild Robot Teaching Without In-The-Wild Robots

A natural language interface for computers

Awesome-LLM-3D: a curated list of Multi-modal Large Language Model in 3D world Resources

Mobile ALOHA: Learning Bimanual Mobile Manipulation with Low-Cost Whole-Body Teleoperation

Imitation Learning algorithms with Co-traing for Mobile ALOHA: ACT, Diffusion Policy, VINN

Mac compatible Ollama Voice

Papers on integrating large language models with embodied AI

The most powerful and modular stable diffusion GUI with a graph/nodes interface.

Benchmark baseline for retrieval qa applications

one-click deepfake (face swap)

MotionGPT: Human Motion as a Foreign Language, a unified motion-language generation model using LLMs

4-Bit Finetuning of Large Language Models on One Consumer GPU