def __init__(self, compute_grad=True):
if not ctc_available():
raise RuntimeError('Baidu CTC is not available and '
'GpuConnectionistTemporalClassification Op '
'can not be constructed.')
self.compute_grad = compute_grad
# Return only the cost. Gradient will be returned by grad()
self.default_output = 0
gof.COp.__init__(self, self.func_file, self.func_name)
def __init__(self, compute_grad=True):
if not ctc_available():
raise RuntimeError('Baidu CTC is not available and '
'GpuConnectionistTemporalClassification Op '
'can not be constructed.')
self.compute_grad = compute_grad
# Return only the cost. Gradient will be returned by grad()
self.default_output = 0
gof.COp.__init__(self, self.func_file, self.func_name)
def setUp(self):
if not ctc_available():
self.skipTest('Optional library warp-ctc not available')
def setUp(self):
if not ctc_available():
self.skipTest('Optional library warp-ctc not available')
def setup_method(self):
if not ctc_available():
pytest.skip("Optional library warp-ctc not available")
from theano.tensor.nnet.ctc import (
ctc,
ctc_available,
ConnectionistTemporalClassification,
)
from tests import unittest_tools as utt
from tests.gpuarray.config import mode_with_gpu, mode_without_gpu
from tests.tensor.nnet.test_ctc import (
setup_torch_case,
setup_ctc_case,
setup_grad_case,
)
@pytest.mark.skipif(not ctc_available(),
reason="Optional library warp-ctc not available")
class TestCTC:
def check_ctc(self, activations, labels, input_length, expected_costs,
expected_grads):
# Create symbolic variables
t_activations = theano.shared(activations, name="activations")
t_activation_times = theano.shared(input_length,
name="activation_times")
t_labels = theano.shared(labels, name="labels")
inputs = [t_activations, t_labels, t_activation_times]
# Execute several tests for each test case
self.check_expected_values(t_activations, t_labels, t_activation_times,
expected_costs, expected_grads)
import theano
import theano.gpuarray
import theano.tensor as tt
from tests import unittest_tools as utt
from tests.gpuarray.config import mode_with_gpu, mode_without_gpu
from tests.tensor.nnet.test_ctc import setup_ctc_case, setup_grad_case, setup_torch_case
from theano.gpuarray.ctc import GpuConnectionistTemporalClassification, gpu_ctc
from theano.tensor.nnet.ctc import (
ConnectionistTemporalClassification,
ctc,
ctc_available,
)
@pytest.mark.skipif(
not ctc_available(), reason="Optional library warp-ctc not available"
)
class TestCTC:
def check_ctc(
self, activations, labels, input_length, expected_costs, expected_grads
):
# Create symbolic variables
t_activations = theano.shared(activations, name="activations")
t_activation_times = theano.shared(input_length, name="activation_times")
t_labels = theano.shared(labels, name="labels")
inputs = [t_activations, t_labels, t_activation_times]
# Execute several tests for each test case
self.check_expected_values(
t_activations, t_labels, t_activation_times, expected_costs, expected_grads
def setUp(self):
if theano.config.mode == "FAST_COMPILE" or theano.config.cxx == "":
raise SkipTest("We need a c compiler")
if not ctc_available():
self.skipTest('Optional library warp-ctc not available')
def setup_method(self):
if theano.config.mode == "FAST_COMPILE" or theano.config.cxx == "":
pytest.skip("We need a c compiler")
if not ctc_available():
pytest.skip("Optional library warp-ctc not available")
def setUp(self):
if theano.config.mode == "FAST_COMPILE" or theano.config.cxx == "":
raise SkipTest("We need a c compiler")
if not ctc_available():
self.skipTest('Optional library warp-ctc not available')