Unable to install flash_attn on H100 with CUDA 12.5

[ghadiaravi13]

Hi,

I am trying to install flash_attn on H100 with CUDA driver 12.5, and I already installed torch with cu12.6

torch.cuda.is_available() = True on my system

However, I get the below error when trying to install flash_attn:

`
[9/9] /home1/apps/nvidia/Linux_aarch64/24.7/cuda/12.5/bin/nvcc --generate-dependencies-with-compile --dependency-output /work/10198/ghadiaravi13/ls6/flash_attn_wheels-1.0.9/build/temp.linux-aarch64-cpython-39/csrc/flash_attn/src/fmha_fwd_hdim64.o.d -I/work/10198/ghadiaravi13/ls6/flash_attn_wheels-1.0.9/csrc/flash_attn -I/work/10198/ghadiaravi13/ls6/flash_attn_wheels-1.0.9/csrc/flash_attn/src -I/work/10198/ghadiaravi13/ls6/flash_attn_wheels-1.0.9/csrc/flash_attn/cutlass/include -I/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/torch/include -I/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/torch/include/torch/csrc/api/include -I/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/torch/include/TH -I/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/torch/include/THC -I/home1/apps/nvidia/Linux_aarch64/24.7/cuda/12.5/include -I/work/10198/ghadiaravi13/vista/eff_llm_env_vista/include -I/usr/include/python3.9 -c -c /work/10198/ghadiaravi13/ls6/flash_attn_wheels-1.0.9/csrc/flash_attn/src/fmha_fwd_hdim64.cu -o /work/10198/ghadiaravi13/ls6/flash_attn_wheels-1.0.9/build/temp.linux-aarch64-cpython-39/csrc/flash_attn/src/fmha_fwd_hdim64.o -D__CUDA_NO_HALF_OPERATORS__ -D__CUDA_NO_HALF_CONVERSIONS__ -D__CUDA_NO_BFLOAT16_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ --expt-relaxed-constexpr --compiler-options ''"'"'-fPIC'"'"'' -O3 -std=c++17 -U__CUDA_NO_HALF_OPERATORS__ -U__CUDA_NO_HALF_CONVERSIONS__ -U__CUDA_NO_HALF2_OPERATORS__ -U__CUDA_NO_BFLOAT16_CONVERSIONS__ --expt-relaxed-constexpr --expt-extended-lambda --use_fast_math --ptxas-options=-v -lineinfo -gencode arch=compute_75,code=sm_75 -gencode arch=compute_80,code=sm_80 -gencode arch=compute_90,code=sm_90 --threads 4 -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="gcc"' '-DPYBIND11_STDLIB="libstdcpp"' '-DPYBIND11_BUILD_ABI="cxxabi1016"' -DTORCH_EXTENSION_NAME=flash_attn_cuda -D_GLIBCXX_USE_CXX11_ABI=1 -ccbin /home1/apps/nvidia/Linux_aarch64/24.7/compilers/bin/nvc
FAILED: /work/10198/ghadiaravi13/ls6/flash_attn_wheels-1.0.9/build/temp.linux-aarch64-cpython-39/csrc/flash_attn/src/fmha_fwd_hdim64.o
/home1/apps/nvidia/Linux_aarch64/24.7/cuda/12.5/bin/nvcc --generate-dependencies-with-compile --dependency-output /work/10198/ghadiaravi13/ls6/flash_attn_wheels-1.0.9/build/temp.linux-aarch64-cpython-39/csrc/flash_attn/src/fmha_fwd_hdim64.o.d -I/work/10198/ghadiaravi13/ls6/flash_attn_wheels-1.0.9/csrc/flash_attn -I/work/10198/ghadiaravi13/ls6/flash_attn_wheels-1.0.9/csrc/flash_attn/src -I/work/10198/ghadiaravi13/ls6/flash_attn_wheels-1.0.9/csrc/flash_attn/cutlass/include -I/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/torch/include -I/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/torch/include/torch/csrc/api/include -I/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/torch/include/TH -I/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/torch/include/THC -I/home1/apps/nvidia/Linux_aarch64/24.7/cuda/12.5/include -I/work/10198/ghadiaravi13/vista/eff_llm_env_vista/include -I/usr/include/python3.9 -c -c /work/10198/ghadiaravi13/ls6/flash_attn_wheels-1.0.9/csrc/flash_attn/src/fmha_fwd_hdim64.cu -o /work/10198/ghadiaravi13/ls6/flash_attn_wheels-1.0.9/build/temp.linux-aarch64-cpython-39/csrc/flash_attn/src/fmha_fwd_hdim64.o -D__CUDA_NO_HALF_OPERATORS -D__CUDA_NO_HALF_CONVERSIONS_ -D__CUDA_NO_BFLOAT16_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ --expt-relaxed-constexpr --compiler-options ''"'"'-fPIC'"'"'' -O3 -std=c++17 -U__CUDA_NO_HALF_OPERATORS__ -U__CUDA_NO_HALF_CONVERSIONS__ -U__CUDA_NO_HALF2_OPERATORS__ -U__CUDA_NO_BFLOAT16_CONVERSIONS__ --expt-relaxed-constexpr --expt-extended-lambda --use_fast_math --ptxas-options=-v -lineinfo -gencode arch=compute_75,code=sm_75 -gencode arch=compute_80,code=sm_80 -gencode arch=compute_90,code=sm_90 --threads 4 -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1016"' -DTORCH_EXTENSION_NAME=flash_attn_cuda -D_GLIBCXX_USE_CXX11_ABI=1 -ccbin /home1/apps/nvidia/Linux_aarch64/24.7/compilers/bin/nvc
nvcc fatal : Unsupported NVHPC compiler found. nvc++ is the only NVHPC compiler that is supported.
ninja: build stopped: subcommand failed.
Traceback (most recent call last):
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/torch/utils/cpp_extension.py", line 2219, in _run_ninja_build
subprocess.run(
File "/usr/lib64/python3.9/subprocess.py", line 528, in run
raise CalledProcessError(retcode, process.args,
subprocess.CalledProcessError: Command '['ninja', '-v']' returned non-zero exit status 1.

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
File "/work/10198/ghadiaravi13/ls6/flash_attn_wheels-1.0.9/setup.py", line 257, in
setup(
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/init.py", line 117, in setup
return distutils.core.setup(**attrs)
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/_distutils/core.py", line 186, in setup
return run_commands(dist)
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/_distutils/core.py", line 202, in run_commands
dist.run_commands()
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/_distutils/dist.py", line 983, in run_commands
self.run_command(cmd)
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/dist.py", line 993, in run_command
super().run_command(command)
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/_distutils/dist.py", line 1002, in run_command
cmd_obj.run()
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/command/install.py", line 109, in run
self.do_egg_install()
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/command/install.py", line 167, in do_egg_install
self.run_command('bdist_egg')
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/_distutils/cmd.py", line 339, in run_command
self.distribution.run_command(command)
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/dist.py", line 993, in run_command
super().run_command(command)
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/_distutils/dist.py", line 1002, in run_command
cmd_obj.run()
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/command/bdist_egg.py", line 177, in run
cmd = self.call_command('install_lib', warn_dir=False)
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/command/bdist_egg.py", line 163, in call_command
self.run_command(cmdname)
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/_distutils/cmd.py", line 339, in run_command
self.distribution.run_command(command)
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/dist.py", line 993, in run_command
super().run_command(command)
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/_distutils/dist.py", line 1002, in run_command
cmd_obj.run()
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/command/install_lib.py", line 19, in run
self.build()
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/_distutils/command/install_lib.py", line 110, in build
self.run_command('build_ext')
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/_distutils/cmd.py", line 339, in run_command
self.distribution.run_command(command)
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/dist.py", line 993, in run_command
super().run_command(command)
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/_distutils/dist.py", line 1002, in run_command
cmd_obj.run()
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/command/build_ext.py", line 99, in run
_build_ext.run(self)
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/_distutils/command/build_ext.py", line 365, in run
self.build_extensions()
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/torch/utils/cpp_extension.py", line 909, in build_extensions
build_ext.build_extensions(self)
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/_distutils/command/build_ext.py", line 481, in build_extensions
self._build_extensions_serial()
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/_distutils/command/build_ext.py", line 507, in _build_extensions_serial
self.build_extension(ext)
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/command/build_ext.py", line 264, in build_extension
_build_ext.build_extension(self, ext)
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/setuptools/_distutils/command/build_ext.py", line 562, in build_extension
objects = self.compiler.compile(
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/torch/utils/cpp_extension.py", line 722, in unix_wrap_ninja_compile
_write_ninja_file_and_compile_objects(
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/torch/utils/cpp_extension.py", line 1879, in _write_ninja_file_and_compile_objects
_run_ninja_build(
File "/work/10198/ghadiaravi13/vista/eff_llm_env_vista/lib64/python3.9/site-packages/torch/utils/cpp_extension.py", line 2235, in _run_ninja_build
raise RuntimeError(message) from e
RuntimeError: Error compiling objects for extension
`

I also tried to build Pytorch using nvc++ compiler (instead of the default g++) but am unable to do so due to some other errors.

Can someone help suggest if they have faced similar error earlier?