Accelerating MoE with IO and Tile-aware Optimizations

Triton-based implementation of Sparse Mixture of Experts.

Pytorch Distributed native training library for LLMs/VLMs with OOTB Hugging Face support

Interactive web application for visualizing U.S. prevailing wages based on official data from OFLC, with region-wise wage breakdowns by job titles and level for LCA/PERM filings. Built with HTML, C…

A framework for few-shot evaluation of language models.

Hand-Rolled GPU communications library

Qwen3-omni is a natively end-to-end, omni-modal LLM developed by the Qwen team at Alibaba Cloud, capable of understanding text, audio, images, and video, as well as generating speech in real time.

Trainable H-Net Package

Ship correct and fast LLM kernels to PyTorch

User-friendly AI Interface (Supports Ollama, OpenAI API, ...)

Muon is an optimizer for hidden layers in neural networks

State-of-the-art Image & Video CLIP, Multimodal Large Language Models, and More!

An umbrella of radically different directions that we can interface with the web.

Reference PyTorch implementation and models for DINOv3

gpt-oss-120b and gpt-oss-20b are two open-weight language models by OpenAI

Renderer for the harmony response format to be used with gpt-oss

Song world.execute (me) ; Presented in HTML

An AI agent system for solving International Mathematical Olympiad (IMO) problems using Google's Gemini, OpenAI, and XAI APIs.

Official PyTorch implementation for ICLR2025 paper "Scaling up Masked Diffusion Models on Text"

paper list, tutorial, and nano code snippet for Diffusion Large Language Models.

CN Zoomer Values

诺亚盘古大模型研发背后的真正的心酸与黑暗的故事。

Virtual whiteboard for sketching hand-drawn like diagrams

A demonstration of text/GUI bi-directional editing via an LSP server

Get your documents ready for gen AI

Convert unstructured documentsto markdown using the Docling.

A Python-embedded DSL that makes it easy to write fast, scalable ML kernels with minimal boilerplate.