Pinned Loading

1. decoding_attention decoding_attention Public

   Decoding Attention is specially optimized for multi head attention (MHA) using CUDA core for the decoding stage of LLM inference.

   C++ 22 1

2. flash_attention_inference flash_attention_inference Public

   Performance of the C++ interface of flash attention and flash attention v2 in large language model (LLM) inference scenarios.

   C++ 22 3

3. cuda_hgemm cuda_hgemm Public

   Several optimization methods of half-precision general matrix multiplication (HGEMM) using tensor core with WMMA API and MMA PTX instruction.

   Cuda 290 66

4. cuda_hook cuda_hook Public

   Hooked CUDA-related dynamic libraries by using automated code generation tools.

   C 135 35

5. cuda_hgemv cuda_hgemv Public

   Several optimization methods of half-precision general matrix vector multiplication (HGEMV) using CUDA core.

   Cuda 48 4

6. cutlass_gemm cutlass_gemm Public

   Multiple GEMM operators are constructed with cutlass to support LLM inference.

   C++ 4 1