cnn_param_file = arguments['cnn_param_file']
cnn_cfg_file = arguments['cnn_cfg_file']
# network structure
cfg = cPickle.load(smart_open(cnn_cfg_file, 'r'))
conv_configs = cfg.conv_layer_configs
conv_layer_number = len(conv_configs)
for i in xrange(conv_layer_number):
conv_configs[i]['activation'] = cfg.conv_activation
# whether to use the fast mode
use_fast = cfg.use_fast
if arguments.has_key('use_fast'):
use_fast = string2bool(arguments['use_fast'])
kaldiread = KaldiReadIn(in_scp_file)
kaldiwrite = KaldiWriteOut(out_ark_file)
log('> ... setting up the CNN convolution layers')
input_shape_train = conv_configs[0]['input_shape']
input_shape_1 = (input_shape_train[1], input_shape_train[2],
input_shape_train[3])
rng = numpy.random.RandomState(123)
theano_rng = RandomStreams(rng.randint(2**30))
cnn = CNN_Forward(numpy_rng=rng,
theano_rng=theano_rng,
conv_layer_configs=conv_configs,
use_fast=use_fast)
_file2nnet(cnn.conv_layers,
layer_index = int(arguments['layer_index'])
# network structure
cfg = cPickle.load(open(cnn_cfg_file, 'r'))
conv_configs = cfg.conv_layer_configs
conv_layer_number = len(conv_configs)
for i in xrange(conv_layer_number):
conv_configs[i]['activation'] = cfg.conv_activation
# whether to use the fast mode
use_fast = cfg.use_fast
if arguments.has_key('use_fast'):
use_fast = string_2_bool(arguments['use_fast'])
kaldiread = KaldiReadIn(in_scp_file)
extra_kaldiread = KaldiReadIn(extra_in_scp_file)
kaldiwrite = KaldiWriteOut(out_ark_file)
log('> ... setting up the CNN convolution layers')
input_shape_train = conv_configs[0]['input_shape']
input_shape_1 = (input_shape_train[1], input_shape_train[2],
input_shape_train[3])
rng = numpy.random.RandomState(89677)
theano_rng = RandomStreams(rng.randint(2**30))
cfg.init_activation()
log('> ... setting up the CNN layers')
cnn = CNN_Forward(numpy_rng=rng, theano_rng=theano_rng, cfg=cfg)
_file2nnet(cnn.layers,
layer_index = int(arguments['layer_index'])
# network structure
cfg = cPickle.load(open(cnn_cfg_file,'r'))
conv_configs = cfg.conv_layer_configs
conv_layer_number = len(conv_configs)
for i in xrange(conv_layer_number):
conv_configs[i]['activation'] = cfg.conv_activation
# whether to use the fast mode
use_fast = cfg.use_fast
if arguments.has_key('use_fast'):
use_fast = string_2_bool(arguments['use_fast'])
kaldiread = KaldiReadIn(in_scp_file)
kaldiwrite = KaldiWriteOut(out_ark_file)
log('> ... setting up the CNN convolution layers')
input_shape_train = conv_configs[0]['input_shape']
input_shape_1 = (input_shape_train[1], input_shape_train[2], input_shape_train[3])
rng = numpy.random.RandomState(89677)
theano_rng = RandomStreams(rng.randint(2 ** 30))
cfg.init_activation()
cnn = CNN_Forward(numpy_rng = rng, theano_rng=theano_rng, conv_layer_configs = conv_configs, use_fast = use_fast)
#cnn = CNNV(numpy_rng = rng, theano_rng=theano_rng, cfg=cfg)
_file2nnet(cnn.conv_layers, set_layer_num = len(conv_configs), filename=cnn_param_file)
out_function = cnn.build_out_function()
for arg in required_arguments:
if arguments.has_key(arg) == False:
print "Error: the argument %s has to be specified" % (arg); exit(1)
# mandatory arguments
in_scp_file = arguments['in_scp_file']
out_ark_file = arguments['out_ark_file']
extra_in_scp_file = arguments['extra_in_scp_file']
lstm_param_file = arguments['lstm_param_file']
lstm_cfg_file = arguments['lstm_cfg_file']
layer_index = int(arguments['layer_index'])
# network structure
cfg = cPickle.load(open(lstm_cfg_file,'r'))
kaldiread = KaldiReadIn(in_scp_file)
extra_kaldiread = KaldiReadIn(extra_in_scp_file)
kaldiwrite = KaldiWriteOut(out_ark_file)
log('> ... setting up the ATTEND LSTM layers')
rng = numpy.random.RandomState(89677)
theano_rng = RandomStreams(rng.randint(2 ** 30))
cfg.init_activation()
lstm = PhaseATTENDLSTM_Forward(numpy_rng=rng, lstm_layer_configs = cfg, n_ins = cfg.n_ins)
_file2nnet(layers = lstm.lstm_layers, set_layer_num = lstm.lstm_layer_num, filename=lstm_param_file)
out_function = lstm.build_out_function()
log('> ... setting up the DNN layers')
dnn = DNNV(numpy_rng = rng, theano_rng = theano_rng, cfg = cfg, input=lstm.lstm_layers[-1].output)
_file2nnet(layers = dnn.layers, set_layer_num = len(dnn.layers)+lstm.lstm_layer_num, filename = lstm_param_file, start_layer = lstm.lstm_layer_num)
out_function2 = dnn.build_extract_feat_function(layer_index)
