Alternatives and similar repositories for libflash_attn:

Users that are interested in libflash_attn are comparing it to the libraries listed below

• tlc-pack / cutlass_fpA_intB_gemm
  A standalone GEMM kernel for fp16 activation and quantized weight, extracted from FasterTransformer
  ☆91Updated 2 weeks ago
• weishengying / cutlass_flash_atten_fp8
  使用 cutlass 仓库在 ada 架构上实现 fp8 的 flash attention
  ☆63Updated 8 months ago
• wangsiping97 / FastGEMV
  High-speed GEMV kernels, at most 2.7x speedup compared to pytorch baseline.
  ☆103Updated 9 months ago
• Bruce-Lee-LY / flash_attention_inference
  Performance of the C++ interface of flash attention and flash attention v2 in large language model (LLM) inference scenarios.
  ☆35Updated last month
• xlite-dev / hgemm-tensorcores-mma
  ⚡️Write HGEMM from scratch using Tensor Cores with WMMA, MMA and CuTe API, Achieve Peak⚡️ Performance.
  ☆68Updated 2 weeks ago
• AlibabaPAI / FLASHNN
  ☆91Updated 7 months ago
• TiledTensor / TiledCUDA
  We invite you to visit and follow our new repository at https://github.com/microsoft/TileFusion. TiledCUDA is a highly efficient kernel …
  ☆180Updated 2 months ago
• NVIDIA / online-softmax
  Benchmark code for the "Online normalizer calculation for softmax" paper
  ☆90Updated 6 years ago
• luliyucoordinate / cute-flash-attention
  Implement Flash Attention using Cute.
  ☆74Updated 3 months ago
• microsoft / TileFusion
  TileFusion is an experimental C++ macro kernel template library that elevates the abstraction level in CUDA C for tile processing.
  ☆78Updated this week
• CalebDu / Awesome-Cute
  ☆52Updated 2 months ago
• BBuf / tensorrt-llm-moe
  ☆28Updated 2 months ago
• Bruce-Lee-LY / cuda_hgemv
  Several optimization methods of half-precision general matrix vector multiplication (HGEMV) using CUDA core.
  ☆60Updated 7 months ago
• zeroine / cutlass-cute-sample
  ☆29Updated 11 months ago
• tgale96 / grouped_gemm
  PyTorch bindings for CUTLASS grouped GEMM.
  ☆80Updated 5 months ago
• ankan-ban / llama_cu_awq
  llama INT4 cuda inference with AWQ
  ☆54Updated 2 months ago
• InternLM / turbomind
  ☆78Updated 2 weeks ago
• Bruce-Lee-LY / decoding_attention
  Decoding Attention is specially optimized for MHA, MQA, GQA and MLA using CUDA core for the decoding stage of LLM inference.
  ☆35Updated last week
• LeiWang1999 / AutoGPTQ.tvm
  GPTQ inference TVM kernel
  ☆38Updated 11 months ago
• LeiWang1999 / tvm_gpu_gemm
  play gemm with tvm
  ☆90Updated last year
• pzhao-eng / FlashMLA
  ☆43Updated 3 weeks ago
• OpenPPL / ppl.llm.kernel.cuda
  ☆148Updated 3 months ago
• ColfaxResearch / cfx-article-src
  ☆97Updated last month
• xlite-dev / ffpa-attn-mma
  📚FFPA(Split-D): Yet another Faster Flash Attention with O(1) GPU SRAM complexity large headdim, 1.8x~3x↑🎉 faster than SDPA EA.
  ☆164Updated last week
• OpenPPL / ppl.pmx
  ☆58Updated 4 months ago
• ademeure / DeeperGEMM
  DeeperGEMM: crazy optimized version
  ☆65Updated last week
• yifuwang / symm-mem-recipes
  ☆64Updated 3 months ago
• Bruce-Lee-LY / cutlass_gemm
  Multiple GEMM operators are constructed with cutlass to support LLM inference.
  ☆17Updated 6 months ago
• caiwanxianhust / FasterLLaMA
  使用 CUDA C++ 实现的 llama 模型推理框架
  ☆49Updated 5 months ago
• gty111 / GEMM_MMA
  Optimize GEMM with tensorcore step by step
  ☆25Updated last year