Stars
Train to 94% on CIFAR-10 in <6.3 seconds on a single A100. Or ~95.79% in ~110 seconds (or less!)
GPUGrants - a list of GPU grants that I can think of
Official inference framework for 1-bit LLMs
Rust library for generating vector embeddings, reranking. Re-write of qdrant/fastembed.
Local first semantic and hybrid BM25 grep / search tool for use by AI and humans!
Hindsight: Agent Memory That Works Like Human Memory
Atomic secret provisioning for NixOS based on sops
🌈 React for interactive command-line apps
CPM.cu is a lightweight, high-performance CUDA implementation for LLMs, optimized for end-device inference and featuring cutting-edge techniques in sparse architecture, speculative sampling and qua…
Inference RWKV v5, v6 and v7 with Qualcomm AI Engine Direct SDK
Reliable model swapping for any local OpenAI/Anthropic compatible server - llama.cpp, vllm, etc
A tool for parsing, dumping and modifying data in Radeon PowerPlay tables
Composable Kernel: Performance Portable Programming Model for Machine Learning Tensor Operators
Development repository for the Triton language and compiler
A retargetable MLIR-based machine learning compiler and runtime toolkit.
A port of the RWKV v7 language model, implemented with the Burn deep learning framework
RimSort is an open source mod manager for the video game RimWorld. There is support for Linux, Mac, and Windows, built from the ground up to be a reliable, community-managed alternative to RimPy Mo…
Tensor computation with WebGPU acceleration
Efficient RWKV inference engine. RWKV7 7.2B fp16 decoding 10250 tps @ single 5090.
A calm, CLI-native way to semantically grep everything, like code, images, pdfs and more.
You like pytorch? You like micrograd? You love tinygrad! ❤️