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

[ROCm] Enable fused_silu_mul_block_quant on ROCm (#38817)

Browse Source 
Signed-off-by: Gregory Shtrasberg <Gregory.Shtrasberg@amd.com>
This commit is contained in:
Gregory Shtrasberg
2026-04-08 11:23:32 -05:00
committed by GitHub
parent d74a306c4b
commit 56c976c1b5
7 changed files with 28 additions and 29 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
CMakeLists.txt
View File

@@ -299,6 +299,7 @@ set(VLLM_EXT_SRC
"csrc/quantization/w8a8/int8/scaled_quant.cu"
"csrc/quantization/w8a8/fp8/common.cu"
"csrc/quantization/fused_kernels/fused_layernorm_dynamic_per_token_quant.cu"
"csrc/quantization/fused_kernels/fused_silu_mul_block_quant.cu"
"csrc/quantization/gguf/gguf_kernel.cu"
"csrc/quantization/activation_kernels.cu"
"csrc/cuda_utils_kernels.cu"
@@ -340,8 +341,7 @@ if(VLLM_GPU_LANG STREQUAL "CUDA")
list(APPEND VLLM_EXT_SRC
"csrc/quantization/awq/gemm_kernels.cu"
"csrc/cutlass_extensions/common.cpp"
"csrc/quantization/fused_kernels/fused_silu_mul_block_quant.cu")
"csrc/cutlass_extensions/common.cpp")
set_gencode_flags_for_srcs(
SRCS "${VLLM_EXT_SRC}"