mirage-project / mirageLinks
Mirage Persistent Kernel: Compiling LLMs into a MegaKernel
☆1,842Updated this week
Alternatives and similar repositories for mirage
Users that are interested in mirage are comparing it to the libraries listed below
Sorting:
- Domain-specific language designed to streamline the development of high-performance GPU/CPU/Accelerators kernels☆1,793Updated this week
- Tile primitives for speedy kernels☆2,734Updated last week
- Distributed Compiler based on Triton for Parallel Systems☆1,148Updated last week
- Flash Attention in ~100 lines of CUDA (forward pass only)☆940Updated 9 months ago
- FlashInfer: Kernel Library for LLM Serving☆3,829Updated this week
- Puzzles for learning Triton☆2,008Updated 10 months ago
- A Quirky Assortment of CuTe Kernels☆582Updated last week
- Fast CUDA matrix multiplication from scratch☆865Updated last month
- A throughput-oriented high-performance serving framework for LLMs☆893Updated 2 weeks ago
- A fast communication-overlapping library for tensor/expert parallelism on GPUs.☆1,133Updated last month
- depyf is a tool to help you understand and adapt to PyTorch compiler torch.compile.☆736Updated 5 months ago
- [MLSys'25] QServe: W4A8KV4 Quantization and System Co-design for Efficient LLM Serving; [MLSys'25] LServe: Efficient Long-sequence LLM Se…☆758Updated 6 months ago
- KernelBench: Can LLMs Write GPU Kernels? - Benchmark with Torch -> CUDA problems☆581Updated 2 weeks ago
- BitBLAS is a library to support mixed-precision matrix multiplications, especially for quantized LLM deployment.☆678Updated last month
- Perplexity GPU Kernels☆476Updated last week
- kernels, of the mega variety☆502Updated last week
- Puzzles for learning Triton, play it with minimal environment configuration!☆527Updated last week
- A unified library of state-of-the-art model optimization techniques like quantization, pruning, distillation, speculative decoding, etc. …☆1,422Updated this week
- FlagGems is an operator library for large language models implemented in the Triton Language.☆684Updated last week
- FP16xINT4 LLM inference kernel that can achieve near-ideal ~4x speedups up to medium batchsizes of 16-32 tokens.☆902Updated last year
- A PyTorch Native LLM Training Framework☆874Updated 2 weeks ago
- A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit floating point (FP8) precision on Hopper, Ada and Bla…☆2,755Updated this week
- 🚀 Efficient implementations of state-of-the-art linear attention models☆3,404Updated last week
- A subset of PyTorch's neural network modules, written in Python using OpenAI's Triton.☆576Updated last month
- Yet Another Language Model: LLM inference in C++/CUDA, no libraries except for I/O☆500Updated 2 weeks ago
- Low-bit LLM inference on CPU/NPU with lookup table☆862Updated 3 months ago
- Materials for learning SGLang☆590Updated last month
- PyTorch native quantization and sparsity for training and inference☆2,384Updated this week
- flash attention tutorial written in python, triton, cuda, cutlass☆425Updated 4 months ago
- 🐳 Efficient Triton implementations for "Native Sparse Attention: Hardware-Aligned and Natively Trainable Sparse Attention"☆877Updated 6 months ago