def dnn_available():
if dnn_available.avail is None:
dev = active_device_number()
if device_properties(dev)['major'] < 3:
dnn_available.msg = "Device not supported by cuDNN"
dnn_available.avail = False
else:
dnn_available.msg = "Can not find the cuDNN library"
dnn_available.avail = theano.gof.cmodule.GCC_compiler.try_flags(
["-l", "cudnn"])
return dnn_available.avail
def setUp(self):
super(TestGpuCumsum, self).setUp()
# Fetch some useful properties on the device
cuda = theano.sandbox.cuda
device_id = cuda.use.device_number
if device_id is None:
cuda.use("gpu",
force=False,
default_to_move_computation_to_gpu=False,
move_shared_float32_to_gpu=False,
enable_cuda=False,
test_driver=True)
device_id = cuda.use.device_number
cuda_ndarray = theano.sandbox.cuda.cuda_ndarray.cuda_ndarray
prop = cuda_ndarray.device_properties(device_id)
self.max_threads_dim0 = prop['maxThreadsDim0']
self.max_grid_size1 = prop['maxGridSize1']
def test_conv_grads():
if cuda.device_properties(cuda.active_device_number())['major'] < 3:
ops = [gemm_op]
else:
ops = [gemm_op, dnn_op]
for mode in 'valid', 'full':
for bs in [1, 5]:
for ch in [4]:
for nf in [3]:
for rImg1 in [2, 5]:
for rImg2 in [2, 8]:
for rFlt1 in [1, 2]:
for rFlt2 in [1, 2]:
for subsample in (1, 1), (1, 2), (2, 2):
for op in ops:
yield (conv_grad, mode, bs, ch, nf,
rImg1, rImg2, rFlt1, rFlt2,
subsample, op)
def test_conv_grads():
if cuda.device_properties(cuda.active_device_number())['major'] < 3:
ops = [gemm_op]
else:
ops = [gemm_op, dnn_op]
for mode in 'valid', 'full':
for bs in [1, 5]:
for ch in [4]:
for nf in [3]:
for rImg1 in [2, 5]:
for rImg2 in [2, 8]:
for rFlt1 in [1, 2]:
for rFlt2 in [1, 2]:
for subsample in (1, 1), (1, 2), (2, 2):
for op in ops:
yield (conv_grad, mode, bs, ch, nf,
rImg1, rImg2, rFlt1, rFlt2,
subsample, op)
def setUp(self):
super(TestGpuCumsum, self).setUp()
# Fetch some useful properties on the device
cuda = theano.sandbox.cuda
device_id = cuda.use.device_number
if device_id is None:
cuda.use("gpu",
force=False,
default_to_move_computation_to_gpu=False,
move_shared_float32_to_gpu=False,
enable_cuda=False,
test_driver=True)
device_id = cuda.use.device_number
cuda_ndarray = theano.sandbox.cuda.cuda_ndarray.cuda_ndarray
prop = cuda_ndarray.device_properties(device_id)
self.max_threads_dim0 = prop['maxThreadsDim0']
self.max_grid_size1 = prop['maxGridSize1']
def dnn_available():
if dnn_available.avail is None:
dev = active_device_number()
if device_properties(dev)['major'] < 3:
dnn_available.msg = "Device not supported by cuDNN"
dnn_available.avail = False
else:
preambule = """
#include <stdio.h>
#include <cuda.h>
#include <cudnn.h>
#include <cudnn_helper.h>
"""
body = """
cudnnHandle_t _handle = NULL;
cudnnStatus_t err;
if ((err = cudnnCreate(&_handle)) != CUDNN_STATUS_SUCCESS) {
fprintf(stderr, "could not create cuDNN handle: %s",
cudnnGetErrorString(err));
return 1;
}
"""
comp, run, out, err = gof.cmodule.GCC_compiler.try_flags(
[
"-l", "cudnn", "-I" + os.path.dirname(__file__),
"-I" + os.path.join(theano.config.cuda.root, 'include'),
"-L" + os.path.join(theano.config.cuda.root, 'lib64')
],
preambule=preambule,
body=body,
try_run=True,
output=True)
dnn_available.avail = comp and run
if dnn_available.avail:
dnn_available.msg = "cuDNN should work"
else:
dnn_available.msg = (
"Theano is not able to use cuDNN. We got this error: \n" +
err)
return dnn_available.avail
def dnn_available():
if dnn_available.avail is None:
dev = active_device_number()
if device_properties(dev)['major'] < 3:
dnn_available.msg = "Device not supported by cuDNN"
dnn_available.avail = False
else:
preambule = """
#include <stdio.h>
#include <cuda.h>
#include <cudnn.h>
#include <cudnn_helper.h>
"""
body = """
cudnnHandle_t _handle = NULL;
cudnnStatus_t err;
if ((err = cudnnCreate(&_handle)) != CUDNN_STATUS_SUCCESS) {
fprintf(stderr, "could not create cuDNN handle: %s",
cudnnGetErrorString(err));
return 1;
}
"""
comp, run, out, err = gof.cmodule.GCC_compiler.try_flags(
["-l", "cudnn", "-I" + os.path.dirname(__file__),
"-I" + os.path.join(theano.config.cuda.root, 'include'),
"-L" + os.path.join(theano.config.cuda.root, 'lib64')],
preambule=preambule, body=body,
try_run=True, output=True)
dnn_available.avail = comp and run
if dnn_available.avail:
dnn_available.msg = "cuDNN should work"
else:
dnn_available.msg = (
"Theano is not able to use cuDNN. We got this error: \n" +
str(err))
return dnn_available.avail
def test_dnn_subsample():
if cuda.device_properties(cuda.active_device_number())['major'] < 3:
raise SkipTest('Current GPU too old')
for t in _test_subsample(GpuDnnConv, theano_mode.including('cudnn')):
yield t
def test_dnn_full():
if cuda.device_properties(cuda.active_device_number())['major'] < 3:
raise SkipTest('Current GPU too old')
for t in _test_full(GpuDnnConv, mode=theano_mode.including("cudnn")):
yield t
cuda_tensor4 = cuda_ndarray.CudaNdarrayType([False] * 4)
device_id = theano.sandbox.cuda.use.device_number
if device_id is None:
cuda_ndarray.shared_constructor(numpy.zeros(2, dtype='float32'))
device_id = theano.sandbox.cuda.use.device_number
if device_id is None:
cuda.use("gpu",
force=False,
default_to_move_computation_to_gpu=False,
move_shared_float32_to_gpu=False,
enable_cuda=False,
test_driver=True)
device_id = theano.sandbox.cuda.use.device_number
cuda_ndarray = theano.sandbox.cuda.cuda_ndarray.cuda_ndarray
device_prop = cuda_ndarray.device_properties(device_id)
def py_conv_valid_numpy(img, kern):
assert img.shape[1] == kern.shape[1]
outshp = (img.shape[0], kern.shape[0],
img.shape[2] - kern.shape[2] + 1,
img.shape[3] - kern.shape[3] + 1)
out = numpy.zeros(outshp, dtype='float32')
for b in xrange(out.shape[0]):
for k in xrange(out.shape[1]):
for rr in xrange(out.shape[2]):
for cc in xrange(out.shape[3]):
#rr, cc is the upper-left corner of img patches
imgpatch = img[b, :, rr:rr + kern.shape[2],
cc:cc + kern.shape[3]]