Stars
🚀 Efficient implementations of state-of-the-art linear attention models
A collection of notebooks/recipes showcasing some fun and effective ways of using Claude.
Build compute kernels and load them from the Hub.
DiffuCoder: Understanding and Improving Masked Diffusion Models for Code Generation
A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit and 4-bit floating point (FP8 and FP4) precision on Hopper, Ada and Blackwell GPUs, to provide better performance…
AirLLM 70B inference with single 4GB GPU
deepbeepmeep / Wan2GP
Forked from Wan-Video/Wan2.1A fast AI Video Generator for the GPU Poor. Supports Wan 2.1/2.2, Qwen Image, Hunyuan Video, LTX Video and Flux.
A fast, single-binary qBittorrent web UI: manage multiple instances, automate torrent workflows, and cross-seed across trackers.
A unified library of SOTA model optimization techniques like quantization, pruning, distillation, speculative decoding, etc. It compresses deep learning models for downstream deployment frameworks …
The slightly more awesome standard unix password manager for teams
Sharing both practical insights and theoretical knowledge about LLM evaluation that we gathered while managing the Open LLM Leaderboard and designing lighteval!
A time traveling resource monitor for modern Linux systems
Not UFO in the sky, but an ultra fold in Neovim.
Collective communications library with various primitives for multi-machine training.
a plugin to make your hyprland cursor more realistic, also adds shake to find
Seemless interface of using PyTOrch distributed with Jupyter notebooks
SGLang is a high-performance serving framework for large language models and multimodal models.
Continuous Thought Machines, because thought takes time and reasoning is a process.
Library for reading and processing ML training data.
Pax is a Jax-based machine learning framework for training large scale models. Pax allows for advanced and fully configurable experimentation and parallelization, and has demonstrated industry lead…
A user-friendly tool chain that enables the seamless execution of ONNX models using JAX as the backend.
JAX - A curated list of resources https://github.com/google/jax