xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2025-12-09 11:45:59 +08:00
Code Issues Packages Projects Releases Wiki Activity

[Kernel][Bugfix] Fixup some warnings in nvfp4_blockwise_moe when CUDA < 12.8 (#20324)

Browse Source 
Signed-off-by: Tyler Michael Smith <tyler@neuralmagic.com>
This commit is contained in:
Tyler Michael Smith 2025-07-01 21:05:47 -04:00 committed by GitHub
parent 3abfe22154
commit 3be8d312a2
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 2 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
csrc/quantization/fp4/nvfp4_blockwise_moe_kernel.cu
View File

@ -335,8 +335,10 @@ void run_fp4_blockwise_scaled_group_mm(
TORCH_CHECK(status == cutlass::Status::kSuccess, "Failed to run GEMM");
}
#if defined ENABLE_NVFP4 && ENABLE_NVFP4
constexpr auto FLOAT4_E2M1X2 = at::ScalarType::Byte;
constexpr auto SF_DTYPE = at::ScalarType::Float8_e4m3fn;
#endif
#define CHECK_TYPE(x, st, m) \
TORCH_CHECK(x.scalar_type() == st, ": Inconsistency of Tensor type:", m)