Add QNNPACK package (#13549)
This commit is contained in:
parent
c43e6839d4
commit
573789f067
2 changed files with 85 additions and 1 deletions
|
@ -122,7 +122,6 @@ class PyTorch(PythonPackage):
|
||||||
depends_on('mkl', when='+mkldnn')
|
depends_on('mkl', when='+mkldnn')
|
||||||
# TODO: add dependency: https://github.com/Maratyszcza/NNPACK
|
# TODO: add dependency: https://github.com/Maratyszcza/NNPACK
|
||||||
depends_on('nnpack', when='+nnpack')
|
depends_on('nnpack', when='+nnpack')
|
||||||
# TODO: add dependency: https://github.com/pytorch/QNNPACK
|
|
||||||
depends_on('qnnpack', when='+qnnpack')
|
depends_on('qnnpack', when='+qnnpack')
|
||||||
depends_on('mpi', when='+distributed')
|
depends_on('mpi', when='+distributed')
|
||||||
depends_on('nccl', when='+nccl')
|
depends_on('nccl', when='+nccl')
|
||||||
|
@ -201,6 +200,8 @@ def enable_or_disable(variant, keyword='USE', var=None, newer=False):
|
||||||
|
|
||||||
enable_or_disable('nnpack')
|
enable_or_disable('nnpack')
|
||||||
enable_or_disable('qnnpack')
|
enable_or_disable('qnnpack')
|
||||||
|
# Never use vendored copy of QNNPACK
|
||||||
|
env.set('USE_PYTORCH_QNNPACK=OFF')
|
||||||
enable_or_disable('distributed')
|
enable_or_disable('distributed')
|
||||||
|
|
||||||
enable_or_disable('nccl')
|
enable_or_disable('nccl')
|
||||||
|
|
83
var/spack/repos/builtin/packages/qnnpack/package.py
Normal file
83
var/spack/repos/builtin/packages/qnnpack/package.py
Normal file
|
@ -0,0 +1,83 @@
|
||||||
|
# Copyright 2013-2019 Lawrence Livermore National Security, LLC and other
|
||||||
|
# Spack Project Developers. See the top-level COPYRIGHT file for details.
|
||||||
|
#
|
||||||
|
# SPDX-License-Identifier: (Apache-2.0 OR MIT)
|
||||||
|
|
||||||
|
from spack import *
|
||||||
|
|
||||||
|
|
||||||
|
class Qnnpack(CMakePackage):
|
||||||
|
"""QNNPACK (Quantized Neural Networks PACKage) is a mobile-optimized
|
||||||
|
library for low-precision high-performance neural network inference.
|
||||||
|
QNNPACK provides implementation of common neural network operators on
|
||||||
|
quantized 8-bit tensors."""
|
||||||
|
|
||||||
|
homepage = "https://github.com/pytorch/QNNPACK"
|
||||||
|
git = "https://github.com/pytorch/QNNPACK.git"
|
||||||
|
|
||||||
|
version('master', branch='master')
|
||||||
|
|
||||||
|
depends_on('cmake@3.5:', type='build')
|
||||||
|
|
||||||
|
resource(
|
||||||
|
name='cpuinfo',
|
||||||
|
git='https://github.com/Maratyszcza/cpuinfo.git',
|
||||||
|
destination='deps',
|
||||||
|
placement='cpuinfo'
|
||||||
|
)
|
||||||
|
resource(
|
||||||
|
name='fp16',
|
||||||
|
git='https://github.com/Maratyszcza/FP16.git',
|
||||||
|
destination='deps',
|
||||||
|
placement='fp16'
|
||||||
|
)
|
||||||
|
resource(
|
||||||
|
name='fxdiv',
|
||||||
|
git='https://github.com/Maratyszcza/FXdiv.git',
|
||||||
|
destination='deps',
|
||||||
|
placement='fxdiv'
|
||||||
|
)
|
||||||
|
resource(
|
||||||
|
name='googlebenchmark',
|
||||||
|
url='https://github.com/google/benchmark/archive/v1.4.1.zip',
|
||||||
|
sha256='61ae07eb5d4a0b02753419eb17a82b7d322786bb36ab62bd3df331a4d47c00a7',
|
||||||
|
destination='deps',
|
||||||
|
placement='googlebenchmark',
|
||||||
|
)
|
||||||
|
resource(
|
||||||
|
name='googletest',
|
||||||
|
url='https://github.com/google/googletest/archive/release-1.8.0.zip',
|
||||||
|
sha256='f3ed3b58511efd272eb074a3a6d6fb79d7c2e6a0e374323d1e6bcbcc1ef141bf',
|
||||||
|
destination='deps',
|
||||||
|
placement='googletest',
|
||||||
|
)
|
||||||
|
resource(
|
||||||
|
name='psimd',
|
||||||
|
git='https://github.com/Maratyszcza/psimd.git',
|
||||||
|
destination='deps',
|
||||||
|
placement='psimd'
|
||||||
|
)
|
||||||
|
resource(
|
||||||
|
name='pthreadpool',
|
||||||
|
git='https://github.com/Maratyszcza/pthreadpool.git',
|
||||||
|
destination='deps',
|
||||||
|
placement='pthreadpool'
|
||||||
|
)
|
||||||
|
|
||||||
|
def cmake_args(self):
|
||||||
|
return [
|
||||||
|
'-DCPUINFO_SOURCE_DIR={0}'.format(
|
||||||
|
join_path(self.stage.source_path, 'deps/cpuinfo')),
|
||||||
|
'-DFP16_SOURCE_DIR={0}'.format(
|
||||||
|
join_path(self.stage.source_path, 'deps/fp16')),
|
||||||
|
'-DFXDIV_SOURCE_DIR={0}'.format(
|
||||||
|
join_path(self.stage.source_path, 'deps/fxdiv')),
|
||||||
|
'-DPSIMD_SOURCE_DIR={0}'.format(
|
||||||
|
join_path(self.stage.source_path, 'deps/psimd')),
|
||||||
|
'-DPTHREADPOOL_SOURCE_DIR={0}'.format(
|
||||||
|
join_path(self.stage.source_path, 'deps/pthreadpool')),
|
||||||
|
'-DGOOGLEBENCHMARK_SOURCE_DIR={0}'.format(
|
||||||
|
join_path(self.stage.source_path, 'deps/googlebenchmark')),
|
||||||
|
'-DGOOGLETEST_SOURCE_DIR={0}'.format(
|
||||||
|
join_path(self.stage.source_path, 'deps/googletest')),
|
||||||
|
]
|
Loading…
Reference in a new issue