mirror of
https://github.com/PaddlePaddle/FastDeploy.git
synced 2025-10-06 09:07:10 +08:00
支持w4afp8 (#3324)
This commit is contained in:
@@ -494,6 +494,8 @@ elif paddle.is_compiled_with_cuda():
|
||||
if cc >= 90 and nvcc_version >= 12.0:
|
||||
# Hopper optmized mla
|
||||
sources += find_end_files("gpu_ops/mla_attn", ".cu")
|
||||
os.system("python utils/auto_gen_w4afp8_gemm_kernel.py")
|
||||
sources += find_end_files("gpu_ops/w4afp8_gemm", ".cu")
|
||||
|
||||
setup(
|
||||
name="fastdeploy_ops",
|
||||
|
Reference in New Issue
Block a user