def parse_config_cnn(self, arguments, nnet_spec, conv_nnet_spec):
self.parse_config_dnn(arguments, nnet_spec)
# parse convolutional layer structure
self.conv_layer_configs = parse_conv_spec(conv_nnet_spec, self.batch_size)
# parse convolutional layer activation
# parse activation function, including maxout
if arguments.has_key('conv_activation'):
self.conv_activation_text = arguments['conv_activation']
self.conv_activation = parse_activation(arguments['conv_activation'])
# maxout not supported yet
# whether we use the fast version of convolution
if arguments.has_key('use_fast'):
self.use_fast = string_2_bool(arguments['use_fast'])
def read_data_args(data_spec):
elements = data_spec.split(",")
pfile_path_list = glob.glob(elements[0])
dataset_args = {}
# default settings
dataset_args['type'] = 'pickle'
dataset_args['random'] = False
dataset_args['stream'] = False
dataset_args['partition'] = 1024 * 1024 * 600 # by default the partition size is 600m if stream is True
# the type of the data: pickle, pfile TO-DO: HDF5
if '.pickle' in data_spec:
dataset_args['type'] = 'pickle'
elif '.pfile' in data_spec:
dataset_args['type'] = 'pfile'
elif '.scp' in data_spec:
dataset_args['type'] = 'kaldi'
else:
dataset_args['type'] = ''
for i in range(1, len(elements)):
element = elements[i]
arg_value = element.split("=")
value = arg_value[1]
key = arg_value[0]
if key == 'partition':
dataset_args['partition'] = 1024 * 1024 * int(value.replace('m',''))
elif key == 'stream':
dataset_args['stream'] = string_2_bool(value) # not supported for now
elif key == 'random':
dataset_args['random'] = string_2_bool(value)
elif key == 'label':
dataset_args['label'] = value
else:
dataset_args[key] = value # left context & right context; maybe different
return pfile_path_list, dataset_args
