biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Bugfix][Build/CI] Fix sparse CUTLASS compilation on CUDA [12.0, 12.2) (#11311)

Browse Source 
Signed-off-by: Tyler Michael Smith <tyler@neuralmagic.com>
This commit is contained in:
Tyler Michael Smith
2024-12-18 21:43:30 -05:00
committed by GitHub
parent fdea8ec167
commit 5a9da2e6e9
12 changed files with 89 additions and 20 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
csrc/sparse/cutlass/sparse_compressor_c3x.cu
View File

@@ -2,6 +2,7 @@
// clang-format off
#include <cudaTypedefs.h>
#if defined CUDA_VERSION && CUDA_VERSION >= 12020
#include "sparse_scaled_mm_c3x.cuh"
#include "cutlass/numeric_conversion.h"
@@ -160,4 +161,5 @@ bool cutlass_sparse_compress_sm90(torch::Tensor& a_nzs, torch::Tensor& a_meta,
return cutlass_sparse_compress<int8_t, int32_t>(a_nzs, a_meta, a);
}
return false;
}
}
#endif