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

dynamic distpatch of fp8 kernels (#14245)

Browse Source 
Signed-off-by: Jeff Daily <jeff.daily@amd.com>
This commit is contained in:
Jeff Daily
2025-03-11 07:54:56 -07:00
committed by GitHub
parent 08a1a1121d
commit a1c8f3796c
25 changed files with 292 additions and 159 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
csrc/quantization/vectorization.cuh
View File

@@ -4,7 +4,6 @@
*/
// Include both AMD and NVIDIA fp8 types to avoid circular import
// TODO(luka/varun) use FP8_TYPE instead after refactoring
#include <c10/util/Float8_e4m3fnuz.h>
#include <c10/util/Float8_e4m3fn.h>