def test_mnist():
"""
Test the mnist.yaml file from the maxout
paper on random input
"""
skip_if_no_gpu()
train = load_train_file(
os.path.join(pylearn2.__path__[0], "scripts/papers/maxout/mnist.yaml"))
# Load fake MNIST data
init_value = control.load_data
control.load_data = [False]
train.dataset = MNIST(which_set='train',
one_hot=1,
axes=['c', 0, 1, 'b'],
start=0,
stop=100)
train.algorithm._set_monitoring_dataset(train.dataset)
control.load_data = init_value
# Train shortly and prevent saving
train.algorithm.termination_criterion = EpochCounter(max_epochs=1)
train.extensions.pop(0)
train.save_freq = 0
train.main_loop()
def setUp(self):
"""
Set up a test image and filter to re-use
"""
skip_if_no_gpu()
self.image = \
numpy.random.rand(16, 3, 3, 1).astype(theano.config.floatX)
self.image_tensor = tensor.tensor4()
self.filters_values = numpy.ones((2, 2, 16, 2, 2, 1, 16),
dtype=theano.config.floatX)
self.filters = sharedX(self.filters_values, name='filters')
self.local = Local(self.filters, (3, 3), 1)
def setUp(self):
"""
Set up a test image and filter to re-use
"""
skip_if_no_gpu()
self.image = \
numpy.random.rand(16, 3, 3, 1).astype(theano.config.floatX)
self.image_tensor = tensor.tensor4()
self.filters_values = numpy.ones(
(2, 2, 16, 2, 2, 1, 16), dtype=theano.config.floatX
)
self.filters = sharedX(self.filters_values, name='filters')
self.local = Local(self.filters, (3, 3), 1)
def setUp(self):
"""
Set up a test image and filter to re-use.
"""
skip_if_no_gpu()
if not dnn_available():
raise SkipTest('Skipping tests cause cudnn is not available')
self.orig_floatX = theano.config.floatX
theano.config.floatX = 'float32'
self.image = np.random.rand(1, 1, 3, 3).astype(theano.config.floatX)
self.image_tensor = tensor.tensor4()
self.input_space = Conv2DSpace((3, 3), 1, axes=('b', 'c', 0, 1))
self.filters_values = np.ones((1, 1, 2, 2), dtype=theano.config.floatX)
self.filters = sharedX(self.filters_values, name='filters')
self.batch_size = 1
self.cudnn2d = Cudnn2D(self.filters, self.batch_size, self.input_space)
def setUp(self):
"""
Set up a test image and filter to re-use.
"""
skip_if_no_gpu()
if not dnn_available():
raise SkipTest('Skipping tests cause cudnn is not available')
self.orig_floatX = theano.config.floatX
theano.config.floatX = 'float32'
self.image = np.random.rand(1, 1, 3, 3).astype(theano.config.floatX)
self.image_tensor = tensor.tensor4()
self.input_space = Conv2DSpace((3, 3), 1, axes=('b', 'c', 0, 1))
self.filters_values = np.ones(
(1, 1, 2, 2), dtype=theano.config.floatX
)
self.filters = sharedX(self.filters_values, name='filters')
self.batch_size = 1
self.cudnn2d = Cudnn2D(self.filters, self.batch_size, self.input_space)
def test_mnist():
"""
Test the mnist.yaml file from the maxout
paper on random input
"""
skip_if_no_gpu()
train = load_train_file(os.path.join(pylearn2.__path__[0],
"scripts/papers/maxout/mnist.yaml"))
# Load fake MNIST data
init_value = control.load_data
control.load_data = [False]
train.dataset = MNIST(which_set='train', one_hot=1,
axes=['c', 0, 1, 'b'], start=0, stop=100)
train.algorithm._set_monitoring_dataset(train.dataset)
control.load_data = init_value
# Train shortly and prevent saving
train.algorithm.termination_criterion = EpochCounter(max_epochs=1)
train.extensions.pop(0)
train.save_freq = 0
train.main_loop()
__authors__ = "David Warde-Farley, Ian Goodfellow"
__copyright__ = "Copyright 2010-2012, Universite de Montreal"
__credits__ = ["David Warde-Farley", "Ian Goodfellow"]
__license__ = "3-clause BSD"
__maintainer__ = "David Warde-Farley"
__email__ = "wardefar@iro"
from pylearn2.testing.skip import skip_if_no_gpu
skip_if_no_gpu()
import numpy as np
import warnings
from theano import function
from theano.sandbox.cuda import gpu_from_host
from theano.sandbox.cuda import host_from_gpu
from theano.sandbox.rng_mrg import MRG_RandomStreams
from theano import shared
from theano import tensor as T
from theano.tensor import as_tensor_variable
from theano.tensor.nnet.conv import conv2d
from pylearn2.sandbox.cuda_convnet.img_acts import ImageActs
def test_match_full_conv():
# Tests that running ImageActs with no padding is the same as running
# theano's conv2D in full mode after flipping the kernel and tranposing
# the output and input channels
# In other words, if convolution computes H=XK, we now compute
# R=HK^T
__authors__ = "David Warde-Farley, Ian Goodfellow"
__copyright__ = "Copyright 2010-2012, Universite de Montreal"
__credits__ = ["David Warde-Farley", "Ian Goodfellow"]
__license__ = "3-clause BSD"
__maintainer__ = "David Warde-Farley"
__email__ = "wardefar@iro"
from pylearn2.testing.skip import skip_if_no_gpu
skip_if_no_gpu()
import numpy as np
import warnings
from theano import function
from theano.sandbox.cuda import gpu_from_host
from theano.sandbox.cuda import host_from_gpu
from theano.sandbox.rng_mrg import MRG_RandomStreams
from theano import shared
from theano import tensor as T
from theano.tensor import as_tensor_variable
from theano.tensor.nnet.conv import conv2d
from pylearn2.sandbox.cuda_convnet.img_acts import ImageActs
def test_match_full_conv():
# Tests that running ImageActs with no padding is the same as running
# theano's conv2D in full mode after flipping the kernel and tranposing
# the output and input channels
# In other words, if convolution computes H=XK, we now compute
def test_biglayer():
"""Test a crash during Theano compilation. It would be too long to
redo this test without depending on Pylearn2. So we put it
here.
"""
skip_if_no_gpu()
yaml_string = """
!obj:pylearn2.train.Train {
dataset: &train
!obj:pylearn2.testing.datasets.random_one_hot_topological_dense_design_matrix {
rng: !obj:numpy.random.RandomState { seed: [2014, 6, 6] },
shape: &input_shape [%(xsize)i, %(ysize)i],
channels: 4,
axes: ['c', 0, 1, 'b'],
num_examples: 128,
num_classes: 10
},
model: !obj:pylearn2.models.mlp.MLP {
batch_size: 128,
layers: [
!obj:pylearn2.models.mlp.FlattenerLayer {
raw_layer: !obj:pylearn2.models.mlp.CompositeLayer {
layer_name: 'h0',
layers: [
!obj:pylearn2.models.mlp.MLP {
layer_name: 'h1',
layers: [
!obj:pylearn2.models.maxout.MaxoutConvC01B {
layer_name: 'conv00',
tied_b: 1,
W_lr_scale: .05,
b_lr_scale: .05,
num_channels: 16,
num_pieces: 1,
kernel_shape: [1, 1],
pool_shape: [4, 4],
pool_stride: [4, 4],
irange: .005,
max_kernel_norm: 0.9,
}
]},
!obj:pylearn2.models.maxout.Maxout {
layer_name: 'max0',
W_lr_scale: .1,
b_lr_scale: .1,
num_units: 16,
irange: .005,
max_col_norm: 1.9365,
num_pieces: 1,
}
]
}
},
!obj:pylearn2.models.mlp.Softmax {
max_col_norm: 1.9365,
layer_name: 'y',
n_classes: 10,
irange: .005
}
],
input_space: !obj:pylearn2.space.Conv2DSpace {
shape: *input_shape,
num_channels: 4,
axes: ['c', 0, 1, 'b'],
},
},
algorithm: !obj:pylearn2.training_algorithms.sgd.SGD {
learning_rate: .05,
learning_rule: !obj:pylearn2.training_algorithms.learning_rule.Momentum {
init_momentum: 0.5,
},
monitoring_dataset:
{
'train': *train
},
termination_criterion: !obj:pylearn2.termination_criteria.EpochCounter {
max_epochs: 3
},
},
extensions: [
!obj:pylearn2.training_algorithms.learning_rule.MomentumAdjustor {
start: 1,
saturate: 250,
final_momentum: .7
}
]
}
"""
try:
orig_floatX = theano.config.floatX
theano.config.floatX = 'float32'
theano.sandbox.cuda.use('gpu')
x_size, y_size = 4, 4
parameters = {'xsize': x_size, 'ysize': y_size}
test = yaml_parse.load(yaml_string % parameters)
test.main_loop()
finally:
theano.config.floatX = orig_floatX
theano.sandbox.cuda.unuse()
