zeux / calmLinks
CUDA/Metal accelerated language model inference
☆589Updated last month
Alternatives and similar repositories for calm
Users that are interested in calm are comparing it to the libraries listed below
Sorting:
- Yet Another Language Model: LLM inference in C++/CUDA, no libraries except for I/O☆383Updated 3 weeks ago
- Flash Attention in ~100 lines of CUDA (forward pass only)☆853Updated 5 months ago
- Perplexity GPU Kernels☆375Updated 2 weeks ago
- FP16xINT4 LLM inference kernel that can achieve near-ideal ~4x speedups up to medium batchsizes of 16-32 tokens.☆846Updated 9 months ago
- kernels, of the mega variety☆406Updated 3 weeks ago
- [MLSys'25] QServe: W4A8KV4 Quantization and System Co-design for Efficient LLM Serving; [MLSys'25] LServe: Efficient Long-sequence LLM Se…☆708Updated 3 months ago
- Mirage: Automatically Generating Fast GPU Kernels without Programming in Triton/CUDA☆1,322Updated this week
- A throughput-oriented high-performance serving framework for LLMs☆832Updated 3 weeks ago
- Fast low-bit matmul kernels in Triton☆323Updated last week
- Fast CUDA matrix multiplication from scratch☆751Updated last year
- Distributed Compiler Based on Triton for Parallel Systems☆846Updated last week
- BitBLAS is a library to support mixed-precision matrix multiplications, especially for quantized LLM deployment.☆637Updated last month
- Fastest kernels written from scratch☆284Updated 2 months ago
- An efficient GPU support for LLM inference with x-bit quantization (e.g. FP6,FP5).☆252Updated 8 months ago
- Repository for the QUIK project, enabling the use of 4bit kernels for generative inference - EMNLP 2024☆180Updated last year
- Applied AI experiments and examples for PyTorch☆277Updated 3 weeks ago
- High-Performance SGEMM on CUDA devices☆96Updated 5 months ago
- Cataloging released Triton kernels.☆238Updated 5 months ago
- ☆221Updated this week
- A subset of PyTorch's neural network modules, written in Python using OpenAI's Triton.☆558Updated last week
- Domain-specific language designed to streamline the development of high-performance GPU/CPU/Accelerators kernels☆1,314Updated this week
- [MLSys'24] Atom: Low-bit Quantization for Efficient and Accurate LLM Serving☆310Updated 11 months ago
- Step-by-step optimization of CUDA SGEMM☆344Updated 3 years ago
- GPTQ inference Triton kernel☆302Updated 2 years ago
- Dynamic Memory Management for Serving LLMs without PagedAttention☆397Updated 3 weeks ago
- flash attention tutorial written in python, triton, cuda, cutlass☆377Updated last month
- Code for the paper "QMoE: Practical Sub-1-Bit Compression of Trillion-Parameter Models".☆277Updated last year
- A simple high performance CUDA GEMM implementation.☆382Updated last year
- Serving multiple LoRA finetuned LLM as one☆1,066Updated last year
- Code for Neurips24 paper: QuaRot, an end-to-end 4-bit inference of large language models.☆398Updated 7 months ago