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

update for fp8

Browse Source 
Signed-off-by: yewentao256 <zhyanwentao@126.com>
This commit is contained in:
yewentao256
2025-08-27 21:36:03 +00:00
parent 57f2f26a05
commit e92676ef4e
12 changed files with 13 additions and 13 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
csrc/rocm/attention.cu
View File

@@ -23,7 +23,7 @@
#include <algorithm>
#include "../attention/dtype_fp8.cuh"
#include "../quantization/fp8/amd/quant_utils.cuh"
#include "../quantization/w8a8/fp8/amd/quant_utils.cuh"
#if defined(__HIPCC__) && \
(defined(__gfx90a__) || defined(__gfx942__) || defined(__gfx950__))

2
csrc/rocm/skinny_gemms.cu
View File

@@ -11,7 +11,7 @@
#include "../cuda_compat.h"
#include "dispatch_utils.h"
#include "quantization/fp8/common.cuh"
#include "quantization/w8a8/fp8/common.cuh"
#if defined(__HIPCC__) && \
(defined(__gfx90a__) || defined(__gfx942__) || defined(__gfx950__))