1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
|
##############################################################################
# Copyright (c) 2013-2018, Lawrence Livermore National Security, LLC.
# Produced at the Lawrence Livermore National Laboratory.
#
# This file is part of Spack.
# Created by Todd Gamblin, tgamblin@llnl.gov, All rights reserved.
# LLNL-CODE-647188
#
# For details, see https://github.com/spack/spack
# Please also see the NOTICE and LICENSE files for our notice and the LGPL.
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU Lesser General Public License (as
# published by the Free Software Foundation) version 2.1, February 1999.
#
# This program is distributed in the hope that it will be useful, but
# WITHOUT ANY WARRANTY; without even the IMPLIED WARRANTY OF
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the terms and
# conditions of the GNU Lesser General Public License for more details.
#
# You should have received a copy of the GNU Lesser General Public
# License along with this program; if not, write to the Free Software
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
##############################################################################
import os
import sys
from spack import *
class Lbann(CMakePackage):
"""LBANN: Livermore Big Artificial Neural Network Toolkit. A distributed
memory, HPC-optimized, model and data parallel training toolkit for deep
neural networks."""
homepage = "http://software.llnl.gov/lbann/"
url = "https://github.com/LLNL/lbann/archive/v0.91.tar.gz"
git = "https://github.com/LLNL/lbann.git"
version('develop', branch='develop')
version('0.93', '1913a25a53d4025fa04c16f14afdaa55')
version('0.92', 'c0eb1595a7c74640e96f280beb497564')
version('0.91', '83b0ec9cd0b7625d41dfb06d2abd4134')
variant('gpu', default=False, description='Builds with support for GPUs via CUDA and cuDNN')
variant('nccl', default=False, description='Builds with support for NCCL communication lib')
variant('opencv', default=True, description='Builds with support for image processing routines with OpenCV')
variant('seq_init', default=False, description='Force serial initialization of weight matrices.')
variant('dtype', default='float',
description='Type for floating point representation of weights',
values=('float', 'double'))
variant('build_type', default='Release',
description='The build type to build',
values=('Debug', 'Release'))
# It seems that there is a need for one statement per version bounds
depends_on('hydrogen +openmp_blas +shared +int64', when='@0.95:')
depends_on('hydrogen +openmp_blas +shared +int64', when='@:0.90')
depends_on('hydrogen +openmp_blas +shared +int64 build_type=Debug',
when=('build_type=Debug' '@0.95:'))
depends_on('hydrogen +openmp_blas +shared +int64 build_type=Debug',
when=('build_type=Debug' '@:0.90'))
depends_on('hydrogen +openmp_blas +shared +int64 +cuda',
when=('+gpu' '@0.95:'))
depends_on('hydrogen +openmp_blas +shared +int64 +cuda',
when=('+gpu' '@:0.90'))
depends_on('hydrogen +openmp_blas +shared +int64 +cuda build_type=Debug',
when=('build_type=Debug' '@0.95:' '+gpu'))
depends_on('hydrogen +openmp_blas +shared +int64 +cuda build_type=Debug',
when=('build_type=Debug' '@:0.90' '+gpu'))
depends_on('elemental +openmp_blas +shared +int64', when=('@0.91:0.94'))
depends_on('elemental +openmp_blas +shared +int64 build_type=Debug',
when=('build_type=Debug' '@0.91:0.94'))
depends_on('cuda', when='+gpu')
depends_on('cudnn', when='+gpu')
depends_on('cub', when='+gpu')
depends_on('mpi', when='~gpu')
depends_on('mpi +cuda', when='+gpu')
depends_on('hwloc ~pci ~libxml2')
# LBANN wraps OpenCV calls in OpenMP parallel loops, build without OpenMP
# Additionally disable video related options, they incorrectly link in a
# bad OpenMP library when building with clang or Intel compilers
depends_on('opencv@3.2.0: +core +highgui +imgproc +jpeg +png +tiff +zlib '
'+fast-math +powerpc +vsx ~calib3d ~cuda ~dnn ~eigen'
'~features2d ~flann ~gtk ~ipp ~ipp_iw ~jasper ~java ~lapack ~ml'
'~openmp ~opencl ~opencl_svm ~openclamdblas ~openclamdfft'
'~pthreads_pf ~python ~qt ~stitching ~superres ~ts ~video'
'~videostab ~videoio ~vtk', when='+opencv')
depends_on('protobuf@3.0.2:')
depends_on('cnpy')
depends_on('nccl', when='+gpu +nccl')
@property
def common_config_args(self):
spec = self.spec
# Environment variables
cppflags = []
cppflags.append('-DLBANN_SET_EL_RNG -ldl')
return [
'-DCMAKE_INSTALL_MESSAGE=LAZY',
'-DCMAKE_CXX_FLAGS=%s' % ' '.join(cppflags),
'-DLBANN_VERSION=spack',
'-DCNPY_DIR={0}'.format(spec['cnpy'].prefix),
]
# Get any recent versions or non-numeric version
# Note that develop > numeric and non-develop < numeric
@when('@:0.90,0.94:')
def cmake_args(self):
spec = self.spec
args = self.common_config_args
args.extend([
'-DLBANN_WITH_TOPO_AWARE:BOOL=%s' % ('+gpu +nccl' in spec),
'-DLBANN_SEQUENTIAL_INITIALIZATION:BOOL=%s' %
('+seq_init' in spec),
'-DLBANN_WITH_TBINF=OFF',
'-DLBANN_WITH_VTUNE=OFF',
'-DLBANN_DATATYPE={0}'.format(spec.variants['dtype'].value),
'-DLBANN_VERBOSE=0'])
if self.spec.satisfies('@:0.90') or self.spec.satisfies('@0.95:'):
args.extend([
'-DHydrogen_DIR={0}/CMake/hydrogen'.format(
spec['hydrogen'].prefix)])
elif self.spec.satisfies('@0.94'):
args.extend([
'-DElemental_DIR={0}/CMake/elemental'.format(
spec['elemental'].prefix)])
# Add support for OpenMP
if (self.spec.satisfies('%clang')):
if (sys.platform == 'darwin'):
clang = self.compiler.cc
clang_bin = os.path.dirname(clang)
clang_root = os.path.dirname(clang_bin)
args.extend([
'-DOpenMP_CXX_FLAGS=-fopenmp=libomp',
'-DOpenMP_CXX_LIB_NAMES=libomp',
'-DOpenMP_libomp_LIBRARY={0}/lib/libomp.dylib'.format(
clang_root)])
if '+opencv' in spec:
args.extend(['-DOpenCV_DIR:STRING={0}'.format(
spec['opencv'].prefix)])
if '+gpu' in spec:
args.extend([
'-DLBANN_WITH_CUDA:BOOL=%s' % ('+gpu' in spec),
'-DLBANN_WITH_SOFTMAX_CUDA:BOOL=%s' % ('+gpu' in spec),
'-DCUDA_TOOLKIT_ROOT_DIR={0}'.format(
spec['cuda'].prefix)])
args.extend([
'-DLBANN_WITH_CUDNN:BOOL=%s' % ('+gpu' in spec),
'-DcuDNN_DIR={0}'.format(
spec['cudnn'].prefix)])
args.extend(['-DCUB_DIR={0}'.format(
spec['cub'].prefix)])
if '+nccl' in spec:
args.extend([
'-DLBANN_WITH_NCCL:BOOL=%s' % ('+gpu +nccl' in spec),
'-DNCCL_DIR={0}'.format(
spec['nccl'].prefix)])
return args
@when('@0.91:0.93')
def cmake_args(self):
spec = self.spec
args = self.common_config_args
args.extend([
'-DWITH_CUDA:BOOL=%s' % ('+gpu' in spec),
'-DWITH_CUDNN:BOOL=%s' % ('+gpu' in spec),
'-DELEMENTAL_USE_CUBLAS:BOOL=%s' % (
'+cublas' in spec['elemental']),
'-DWITH_TBINF=OFF',
'-DWITH_VTUNE=OFF',
'-DElemental_DIR={0}'.format(spec['elemental'].prefix),
'-DELEMENTAL_MATH_LIBS={0}'.format(
spec['elemental'].libs),
'-DSEQ_INIT:BOOL=%s' % ('+seq_init' in spec),
'-DVERBOSE=0',
'-DLBANN_HOME=.'])
if spec.variants['dtype'].value == 'float':
args.extend(['-DDATATYPE=4'])
elif spec.variants['dtype'].value == 'double':
args.extend(['-DDATATYPE=8'])
if '+opencv' in spec:
args.extend(['-DOpenCV_DIR:STRING={0}'.format(
spec['opencv'].prefix)])
if '+cudnn' in spec:
args.extend(['-DcuDNN_DIR={0}'.format(
spec['cudnn'].prefix)])
if '+cub' in spec:
args.extend(['-DCUB_DIR={0}'.format(
spec['cub'].prefix)])
return args
|
