ROCm / flash-attentionLinks
Fast and memory-efficient exact attention
☆192Updated last week
Alternatives and similar repositories for flash-attention
Users that are interested in flash-attention are comparing it to the libraries listed below
Sorting:
- Development repository for the Triton language and compiler☆131Updated this week
- An efficient GPU support for LLM inference with x-bit quantization (e.g. FP6,FP5).☆265Updated 2 months ago
- A safetensors extension to efficiently store sparse quantized tensors on disk☆164Updated this week
- BitBLAS is a library to support mixed-precision matrix multiplications, especially for quantized LLM deployment.☆678Updated last month
- Fast low-bit matmul kernels in Triton☆373Updated last week
- A high-throughput and memory-efficient inference and serving engine for LLMs☆102Updated this week
- Fast Hadamard transform in CUDA, with a PyTorch interface☆239Updated last month
- QQQ is an innovative and hardware-optimized W4A8 quantization solution for LLMs.☆140Updated last month
- ☆45Updated this week
- AI Tensor Engine for ROCm☆281Updated this week
- [MLSys'24] Atom: Low-bit Quantization for Efficient and Accurate LLM Serving☆320Updated last year
- GPTQ inference Triton kernel☆309Updated 2 years ago
- Applied AI experiments and examples for PyTorch☆296Updated last month
- PyTorch bindings for CUTLASS grouped GEMM.☆123Updated 4 months ago
- ☆199Updated 4 months ago
- [NeurIPS 2024] KVQuant: Towards 10 Million Context Length LLM Inference with KV Cache Quantization☆383Updated last year
- ☆153Updated 3 months ago
- Ahead of Time (AOT) Triton Math Library☆76Updated 2 weeks ago
- [ICLR'25] Fast Inference of MoE Models with CPU-GPU Orchestration☆234Updated 10 months ago
- High-speed GEMV kernels, at most 2.7x speedup compared to pytorch baseline.☆115Updated last year
- FP16xINT4 LLM inference kernel that can achieve near-ideal ~4x speedups up to medium batchsizes of 16-32 tokens.☆902Updated last year
- This repository contains integer operators on GPUs for PyTorch.☆218Updated 2 years ago
- Code for Neurips24 paper: QuaRot, an end-to-end 4-bit inference of large language models.☆429Updated 10 months ago
- 8-bit CUDA functions for PyTorch☆62Updated last week
- Boosting 4-bit inference kernels with 2:4 Sparsity☆82Updated last year
- [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
- ☆90Updated 10 months ago
- Flash-LLM: Enabling Cost-Effective and Highly-Efficient Large Generative Model Inference with Unstructured Sparsity☆221Updated 2 years ago
- Reorder-based post-training quantization for large language model☆195Updated 2 years ago
- An easy-to-use package for implementing SmoothQuant for LLMs☆106Updated 5 months ago