tlc-pack / libflash_attnLinks
Standalone Flash Attention v2 kernel without libtorch dependency
☆111Updated 10 months ago
Alternatives and similar repositories for libflash_attn
Users that are interested in libflash_attn are comparing it to the libraries listed below
Sorting:
- A standalone GEMM kernel for fp16 activation and quantized weight, extracted from FasterTransformer☆94Updated 3 weeks ago
- ⚡️Write HGEMM from scratch using Tensor Cores with WMMA, MMA and CuTe API, Achieve Peak⚡️ Performance.☆93Updated 2 months ago
- 使用 cutlass 仓库在 ada 架构上实现 fp8 的 flash attention☆74Updated 11 months ago
- Performance of the C++ interface of flash attention and flash attention v2 in large language model (LLM) inference scenarios.☆39Updated 5 months ago
- High-speed GEMV kernels, at most 2.7x speedup compared to pytorch baseline.☆113Updated last year
- We invite you to visit and follow our new repository at https://github.com/microsoft/TileFusion. TiledCUDA is a highly efficient kernel …☆183Updated 6 months ago
- Benchmark code for the "Online normalizer calculation for softmax" paper☆95Updated 7 years ago
- Several optimization methods of half-precision general matrix vector multiplication (HGEMV) using CUDA core.☆63Updated 10 months ago
- ☆96Updated 10 months ago
- TileFusion is an experimental C++ macro kernel template library that elevates the abstraction level in CUDA C for tile processing.☆93Updated last month
- ☆60Updated 3 months ago
- llama INT4 cuda inference with AWQ☆54Updated 6 months ago
- ☆31Updated 6 months ago
- ☆91Updated 2 months ago
- Implement Flash Attention using Cute.☆92Updated 7 months ago
- Multiple GEMM operators are constructed with cutlass to support LLM inference.☆18Updated this week
- ☆37Updated last year
- ☆92Updated 4 months ago
- GPTQ inference TVM kernel☆40Updated last year
- Decoding Attention is specially optimized for MHA, MQA, GQA and MLA using CUDA core for the decoding stage of LLM inference.☆40Updated last month
- ☆51Updated 2 weeks ago
- play gemm with tvm☆91Updated 2 years ago
- ⚡️FFPA: Extend FlashAttention-2 with Split-D, achieve ~O(1) SRAM complexity for large headdim, 1.8x~3x↑ vs SDPA.🎉☆194Updated 2 months ago
- ☆11Updated 5 months ago
- Triton adapter for Ascend. Mirror of https://gitee.com/ascend/triton-ascend☆61Updated last week
- ☆149Updated 6 months ago
- DeeperGEMM: crazy optimized version☆71Updated 3 months ago
- CUDA 8-bit Tensor Core Matrix Multiplication based on m16n16k16 WMMA API☆31Updated last year
- This is a demo how to write a high performance convolution run on apple silicon☆54Updated 3 years ago
- PyTorch bindings for CUTLASS grouped GEMM.☆107Updated 2 months ago