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

[ROCm] support Radeon™ 7900 series (gfx1100) without using flash-attention (#2768)

Browse Source
This commit is contained in:
Hongxia Yang
2024-02-11 02:14:37 -05:00
committed by GitHub
parent 3711811b1d
commit 0580aab02f
4 changed files with 60 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
setup.py
View File

@@ -24,7 +24,7 @@ MAIN_CUDA_VERSION = "12.1"
# Supported NVIDIA GPU architectures.
NVIDIA_SUPPORTED_ARCHS = {"7.0", "7.5", "8.0", "8.6", "8.9", "9.0"}
ROCM_SUPPORTED_ARCHS = {"gfx90a", "gfx942"}
ROCM_SUPPORTED_ARCHS = {"gfx90a", "gfx942", "gfx1100"}
# SUPPORTED_ARCHS = NVIDIA_SUPPORTED_ARCHS.union(ROCM_SUPPORTED_ARCHS)