-
Notifications
You must be signed in to change notification settings - Fork 2
/
map_small.py
214 lines (173 loc) · 5.86 KB
/
map_small.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
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
import os
import numpy as np
import keras
from keras.layers import Input, Dense, Activation, BatchNormalization
from keras.constraints import maxnorm
from keras.layers.core import Dropout
from keras.optimizers import SGD
from keras.models import Model, Sequential
import numpy as np
from keras.models import load_model
import os,sys
from sklearn import preprocessing
import pickle, logging
from keras.callbacks import *
import random
inp_dim=711
out_dim = 66
hidden = int(sys.argv[1])
exp_name = sys.argv[2]
context_flag = 1
context = 9
arch = str(hidden) + '_6layerReLu'
logfile_name = exp_name + '/logs/log_' + arch + '.log'
g = open(logfile_name,'w')
g.close()
model_dir = exp_name + '/models/'
test_dir = exp_name + '/test/' + arch
resynth_dir = exp_name + '/resynth/' + arch
validation_dir = exp_name + '/validation/' + arch
save_model = 1
for k in [model_dir, test_dir, resynth_dir, validation_dir]:
if not os.path.exists(k):
os.makedirs(k)
# Contexts
def make_contexts(nparray, window):
context_frames = np.asarray(zip(*[nparray[n:] for n in range(window)]))
temp = []
for f in context_frames:
temp.append(np.concatenate(([f[n] for n in range(window)])))
return np.asarray(temp)
# Declare train and test files
files_train = []
files_test = []
train_file = 'files.train'
test_file = 'files.test'
f = open(train_file)
for line in f:
line = line.split('\n')[0]
files_train.append(line)
f.close()
f = open(test_file)
for line in f:
line = line.split('\n')[0]
files_test.append(line)
# Load train and validation data
train_input = []
train_output = []
valid_input = []
valid_output = []
valid_files = []
for train_file in files_train:
A = np.loadtxt('/home/sirisha.rallabandi/data/tts_stuff/input_full/' + train_file + '.lab')
i_l = len(A)
B = np.loadtxt('/home/sirisha.rallabandi/data/tts_stuff/output_full/' + train_file + '.ccoeffs')
o_l = len(B)
if i_l == o_l:
if context_flag:
A = make_contexts(A, context)
B = B[(context-1)/2:-(context-1)/2]
for (a,b) in zip(A,B):
train_input.append(a)
train_output.append(b)
else:
print "Discarded ", train_file
for valid_file in files_test:
A = np.loadtxt('/home/sirisha.rallabandi/data/tts_stuff/input_full/' + valid_file + '.lab')
i_l = len(A)
B = np.loadtxt('/home/sirisha.rallabandi/data/tts_stuff/output_full/' + valid_file + '.ccoeffs')
o_l = len(B)
if i_l == o_l:
valid_input.append(A)
valid_output.append(B)
valid_files.append(valid_file)
else:
print "Discarded ", valid_file
num_valid = len(valid_files)
valid_data = zip(valid_input, valid_output, valid_files)
random.shuffle(valid_data)
test_data = valid_data[0:int(num_valid/2)]
valid_data = valid_data[int(num_valid)/2+1:]
train_input = np.array(train_input)
train_output = np.array(train_output)
class LoggingCallback(Callback):
"""Callback that logs message at end of epoch.
"""
def __init__(self, print_fcn="print"):
Callback.__init__(self)
self.print_fcn = print_fcn
def on_epoch_end(self, epoch, logs={}):
pass
# If first epoch, remove the log file
if epoch == 0:
g = open(logfile_name,'w')
g.close()
# Log the progress
msg = "{Epoch: %i} %s" % (epoch, ", ".join("%s: %f" % (k, v) for k, v in logs.items()))
self.print_fcn(msg)
with open(logfile_name,'a') as g:
g.write(msg + '\n')
#test_model(self.model,input_scaler,output_scaler, epoch)
#Save the model every 5 epochs
if epoch % 15 == 1 and save_model:
print self.model
self.model.save(model_dir + '_' + arch + '.h5')
def test_model():
# Test each file
for (inp, out, fname) in valid_data:
#inp = input_scaler.transform(inp)
pred = model.predict(inp)
pred = output_scaler.inverse_transform(pred)
np.savetxt(validation_dir + '/' + fname + '.ccoeffs', pred)
np.savetxt(resynth_dir + '/' + fname + '.ccoeffs', out)
for (inp, out, fname) in test_data:
#inp = input_scaler.transform(inp)
pred = model.predict(inp)
pred = output_scaler.inverse_transform(pred)
np.savetxt(test_dir + '/' + fname + '.ccoeffs', pred)
np.savetxt(resynth_dir + '/' + fname + '.ccoeffs', out)
input_scaler = preprocessing.StandardScaler().fit(train_input)
output_scaler = preprocessing.StandardScaler().fit(train_output)
#train_input = input_scaler.transform(train_input)
train_output = output_scaler.transform(train_output)
def train_model():
global model
# Create the model
model = Sequential()
# INPUT LAYER
model.add(Dropout(0.0, input_shape=(inp_dim,)))
model.add(Dense(inp_dim,activation='relu'))
model.add(Dropout(0.2))
# HIDDEN 1
model.add(Dense(hidden, activation='relu'))
model.add(Dropout(0.2))
# HIDDEN 2
model.add(Dense(hidden, activation='relu'))
model.add(Dropout(0.2))
# HIDDEN 3
model.add(Dense(hidden, activation='relu'))
model.add(Dropout(0.2))
# HIDDEN 4
model.add(Dense(hidden, activation='relu'))
model.add(Dropout(0.2))
# HIDDEN 5
model.add(Dense(hidden, activation='relu'))
model.add(Dropout(0.2))
# HIDDEN 6
model.add(Dense(hidden, activation='relu'))
model.add(Dropout(0.2))
model.add(Dense(out_dim, activation='relu'))
# Compile the model
sgd = SGD(lr=0.01, momentum=0.9, decay=1e-6, nesterov=False)
model.compile(optimizer=sgd, loss='mse')
model.summary()
model.fit(train_input,train_output,epochs=40, batch_size=256, shuffle=True,callbacks=[LoggingCallback(logging.info)])
train_model()
test_model()
# Resynthesize
cmd = "./do_synth synth_world " + resynth_dir + ' files.test'
os.system(cmd)
cmd = "./do_synth synth_world " + test_dir + ' files.test'
os.system(cmd)
cmd = "./do_synth synth_world " + validation_dir + ' files.test'
os.system(cmd)