A highly customizable macOS status bar replacement

AeroSpace is an i3-like tiling window manager for macOS

Codex skill and Rust DataFusion converter for exporting Nsight Systems reports to Perfetto JSON

FlowBlock: Wavefront-Parallel Decoding Framework for Self-Correcting Diffusion Language Models

FA4-based Relative Attention Kernel developed by TML and Colfax

Helpful tools and examples for working with flex-attention

Real-time serving of full-duplex interaction models: per-tick deadline scheduling + KV-budget admission control. Systems research.

The official implementation of DeLS-Spec.

A high-performance inference engine for AI models

🔥 [ICML'26] ParisKV: Fast and Drift-Robust KV-Cache Retrieval for Long-Context LLMs

Scaling Mixture-of-Experts Video Pretraining for Embodied Intelligence

CUDA Templates and Python DSLs for High-Performance Linear Algebra

SwiftRDMA -- Exposing RDMA NIC Resources for Software-Defined RDMA Scheduling

SGLang is a high-performance serving framework for large language models and multimodal models.

ElasticMM: Elastic and Efficient MLLM Serving System

Ask the oracle when you're stuck. Invoke GPT-5 Pro with a custom context and files.

Rust bindings for the [3FS USRBIO API](https://github.com/deepseek-ai/3FS/blob/main/src/lib/api/UsrbIo.md).

Algorithms for latent compaction

personal skills

Implementation for paper "Forcing-KV: Hybrid KV Cache Compression for Efficient Autoregressive Video Diffusion Models".

DeepSpec: a full-stack codebase for training and evaluating speculative decoding algorithms

Using a swizzled hierarchical layout for GEMM