-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathtest_ctc.py
166 lines (126 loc) · 6.63 KB
/
test_ctc.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
#!/usr/env/python
import pickle
import numpy as np
from numpy import testing
import theano
import theano.tensor as T
import lasagne
from lasagne.layers import *
from lasagne.nonlinearities import identity, softmax
import ctc_cost
from scribe import print_slab, prepare_print_pred
floatX = theano.config.floatX
theano.config.compute_test_value = 'off'
#theano.config.on_unused_input = 'ignore'
class Network:
x = T.itensor3('inputs') # num_batch, input_seq_len, num_inputs
y = T.imatrix('targets') # num_batch, output_seq_len
mask_x = T.imatrix('mask_inputs')
mask_y = T.imatrix('mask_targets')
def __init__(self, num_batch, input_seq_len, num_inputs, output_seq_len, num_classes):
self.num_batch = num_batch
self.input_seq_len = input_seq_len
self.num_inputs = num_inputs
self.num_features = 15
self.num_units = 50
self.output_seq_len = output_seq_len
self.num_classes = num_classes
self.num_outputs = num_classes + 1 # add blank
self.setup()
def setup(self):
# setup Lasagne Recurrent network
# The output from the network is shape
# a) output_lin_ctc is the activation before softmax (input_seq_len, batch_size, num_classes + 1)
# b) ouput_softmax is the output after softmax (batch_size, input_seq_len, num_classes + 1)
l_inp = InputLayer(shape=(self.num_batch, self.input_seq_len, self.num_inputs))
l_mask = InputLayer(shape=(self.num_batch, self.input_seq_len))
l_emb = EmbeddingLayer(l_inp, input_size=self.num_inputs, output_size=self.num_features)
l_rnn = LSTMLayer(l_inp, num_units=self.num_units, peepholes=True, mask_input=l_mask)
l_rnn_shp = ReshapeLayer(l_rnn, shape=(-1, self.num_units))
l_out = DenseLayer(l_rnn_shp, num_units=self.num_outputs, nonlinearity=identity)
l_out_shp = ReshapeLayer(l_out, shape=(-1, self.input_seq_len, self.num_outputs))
# dimshuffle to shape format (input_seq_len, batch_size, num_classes + 1)
#l_out_shp_ctc = lasagne.layers.DimshuffleLayer(l_out_shp, (1, 0, 2))
l_out_softmax = NonlinearityLayer(l_out, nonlinearity=softmax)
l_out_softmax_shp = ReshapeLayer(l_out_softmax, shape=(-1, self.input_seq_len, self.num_outputs))
# calculate grad and cost
output_lin_ctc = get_output(l_out_shp, {l_inp: self.x, l_mask: self.mask_x})
output_softmax = get_output(l_out_softmax_shp, {l_inp: self.x, l_mask: self.mask_x})
all_params = get_all_params(l_out_softmax_shp, trainable=True) # dont learn embeddinglayer
# the CTC cross entropy between y and linear output network
pseudo_cost = ctc_cost.pseudo_cost(self.y, output_lin_ctc, self.mask_y, self.mask_x)
# calculate the gradients of the CTC wrt. linar output of network
pseudo_grad = T.grad(pseudo_cost.sum() / self.num_batch, all_params)
true_cost = ctc_cost.cost(self.y, output_softmax, self.mask_y, self.mask_x)
cost = T.mean(true_cost)
shared_lr = theano.shared(lasagne.utils.floatX(0.001))
#updates = lasagne.updates.sgd(pseudo_cost_grad, all_params, learning_rate=shared_lr)
#updates = lasagne.updates.apply_nesterov_momentum(updates, all_params, momentum=0.9)
updates = lasagne.updates.rmsprop(pseudo_grad, all_params, learning_rate=shared_lr)
self.train = theano.function([self.x, self.mask_x, self.y, self.mask_y],
[output_softmax, cost], updates=updates)
self.test = theano.function([self.x, self.mask_x], [output_softmax])
if __name__ == '__main__':
#Y_hat = np.asarray(np.random.normal(
# 0, 1, (input_seq_len, num_batch, num_classes + 1)), dtype=floatX)
#Y = np.zeros((target_seq_len, num_batch), dtype='int64')
#Y[25:, :] = 1
#Y_hat_mask = np.ones((input_seq_len, num_batch), dtype=floatX)
#Y_hat_mask[-5:] = 0
# default blank symbol is the highest class index (3 in this case)
#Y_mask = np.asarray(np.ones_like(Y), dtype=floatX)
#X = np.random.random(
# (num_batch, input_seq_len)).astype('int32')
#
#y = T.imatrix('phonemes')
#x = T.imatrix() # batchsize, input_seq_len, features
with open("digit.pkl", "rb") as pkl_file:
data = pickle.load(pkl_file)
chars = data['chars']
num_samples = len(data['x'])
num_inputs = data['x'][0].shape[0]
num_classes = len(chars)
print_pred = prepare_print_pred(num_classes)
max_input_seq_len = max([x.shape[-1] for x in data['x']])
max_output_seq_len = max([len(y) for y in data['y']])
#print(num_samples, num_inputs, num_classes, max_input_seq_len, max_output_seq_len)
num_batch = 20
net = Network(num_batch, max_input_seq_len, num_inputs, max_output_seq_len, num_classes)
scale_to_int = 1024
num_epoch = 100
for epoch in range(num_epoch):
print("\n## EPOCH", epoch)
shuffle = np.random.permutation(num_samples)
cost_lst = []
for batch in range(num_samples//num_batch):
idx = shuffle[batch*num_batch:(batch+1)*num_batch]
X = np.zeros(shape=(num_batch, max_input_seq_len, num_inputs), dtype='int32')
y = np.ones(shape=(num_batch, max_output_seq_len), dtype='int32') * num_classes
mask_X = np.zeros(shape=(num_batch, max_input_seq_len), dtype=np.bool)
mask_y = np.zeros(shape=(num_batch, max_output_seq_len), dtype=np.bool)
for i in range(num_batch):
j = idx[i]
input_seq_len = data['x'][j].shape[-1]
output_seq_len = len(data['y'][j])
X[i, 0:input_seq_len] = data['x'][j].T * scale_to_int
y[i, 0:output_seq_len] = np.array(data['y'][j])
mask_X[i, 0:input_seq_len] = 1
mask_y[i, 0:output_seq_len] = 1
output_softmax, cost = net.train(X, mask_X, y, mask_y)
cost_lst.append(cost)
#testing.assert_almost_equal(pseudo_cost, pseudo_cost_old, decimal=4)
#testing.assert_array_almost_equal(pseudo_cost_val, pseudo_cost_old_val)
print(" - mean cost:", np.mean(cost_lst))
for i in range(num_batch):
j = idx[i]
pred = np.argmax(output_softmax[i], axis=-1)
pred = print_pred(pred)
true = print_pred(y[i], ignore_repeat=True)
print()
print("target :", true)
print("prediction :", pred)
print("input bitmap :")
print_slab(data['x'][j])
print("softmax firing :")
input_seq_len = data['x'][j].shape[-1]
print_slab(output_softmax[i, 0:input_seq_len].T)