TensorRT-LLMs/cpp/kernels/fmha_v2/test_sm75_fmhca.sh
qsang-nv 0fd59d64ab
infra: open source fmha v2 kernels (#4185)
* add fmha repo

Signed-off-by: Qidi Sang <200703406+qsang-nv@users.noreply.github.com>

* fix format

Signed-off-by: Qidi Sang <200703406+qsang-nv@users.noreply.github.com>

* fix code style

Signed-off-by: Qidi Sang <200703406+qsang-nv@users.noreply.github.com>

* fix header

Signed-off-by: Qidi Sang <200703406+qsang-nv@users.noreply.github.com>

* fix header kernel_traits.h

Signed-off-by: Qidi Sang <200703406+qsang-nv@users.noreply.github.com>

* add .gitignore file

Signed-off-by: Qidi Sang <200703406+qsang-nv@users.noreply.github.com>

* add SLIDING_WINDOW_ATTENTION

Signed-off-by: Qidi Sang <200703406+qsang-nv@users.noreply.github.com>

* fix style

Signed-off-by: Qidi Sang <200703406+qsang-nv@users.noreply.github.com>

* fix format

Signed-off-by: Qidi Sang <200703406+qsang-nv@users.noreply.github.com>

* update setup.py

Signed-off-by: Qidi Sang <200703406+qsang-nv@users.noreply.github.com>

* update build_wheel.py

Signed-off-by: Qidi Sang <200703406+qsang-nv@users.noreply.github.com>

---------

Signed-off-by: Qidi Sang <200703406+qsang-nv@users.noreply.github.com>
Signed-off-by: qsang-nv <200703406+qsang-nv@users.noreply.github.com>
2025-05-15 10:56:34 +08:00

21 lines
782 B
Bash
Executable File

# COMMAND="$CUDA_PATH/compute-sanitizer/compute-sanitizer"
COMMAND=
$COMMAND ./bin/fmhca.exe -b 1 -s-q 4096 -min-s 4096 -d 40
$COMMAND ./bin/fmhca.exe -b 1 -s-q 4096 -min-s 4096 -d 80
$COMMAND ./bin/fmhca.exe -b 4 -s-q 4096 -min-s 4096 -d 40
$COMMAND ./bin/fmhca.exe -b 4 -s-q 4096 -min-s 4096 -d 80
$COMMAND ./bin/fmhca.exe -b 1 -s-q 2304 -min-s 2304 -d 40
$COMMAND ./bin/fmhca.exe -b 1 -s-q 2304 -min-s 2304 -d 80
$COMMAND ./bin/fmhca.exe -b 4 -s-q 2304 -min-s 2304 -d 40
$COMMAND ./bin/fmhca.exe -b 4 -s-q 2304 -min-s 2304 -d 80
$COMMAND ./bin/fmhca.exe -b 1 -s-q 1024 -min-s 1024 -d 40
$COMMAND ./bin/fmhca.exe -b 1 -s-q 1024 -min-s 1024 -d 80
$COMMAND ./bin/fmhca.exe -b 4 -s-q 1024 -min-s 1024 -d 40
$COMMAND ./bin/fmhca.exe -b 4 -s-q 1024 -min-s 1024 -d 80