/
train_bigdata_class.py
379 lines (335 loc) · 12 KB
/
train_bigdata_class.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
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
#!/usr/bin/env python
# -*- coding: utf-8 -*-
from __future__ import print_function
import argparse
import datetime
import json
import multiprocessing
import os
import random
import sys
import threading
import time
import linecache
import csv
import pyximport
pyximport.install()
import gc
import numpy as np
from PIL import Image
import six
import six.moves.cPickle as pickle
from six.moves import queue
from chainer import computational_graph as c
from chainer import cuda
from chainer import optimizers
import cyfuncs
import tools
parser = argparse.ArgumentParser(description='Chainer example: MNIST')
parser.add_argument('--gpu', '-g', default=-1, type=int,
help='GPU ID (negative value indicates CPU)')
parser.add_argument('trainfile', help='Path to train file name')
parser.add_argument('testfile', help='Path to test file name')
parser.add_argument('--experiment_name', '-n', default='experiment', type=str,
help='experiment name')
parser.add_argument('--epoch', '-E', default=2000, type=int,
help='Number of epochs to learn')
parser.add_argument('--batchsize', '-B', type=int, default=20000,
help='Learning minibatch size')
parser.add_argument('--arch', '-a', default='dnn_4',
help='dnn architecture \
(snn, dnn_4)')
parser.add_argument('--input', '-in', default=60, type=int,
help='input node number')
parser.add_argument('--hidden', '-hn', default=100, type=int,
help='hidden node number')
parser.add_argument('--channel', '-c', default=1, type=int,
help='data channel')
parser.add_argument('--loaderjob', '-j', default=2, type=int,
help='Number of parallel data loading processes')
args = parser.parse_args()
if args.gpu >= 0:
cuda.check_cuda_available()
xp = cuda.cupy if args.gpu >= 0 else np
#assert 50000 % args.batchsize == 0
# Prepare model
if args.arch == 'snn':
import snn
model = snn.RClassification_ShallowNN(args.input, args.hidden)
print ('model is snn')
elif args.arch == 'dnn_4':
import dnn_4
model = dnn_4.Classification_DNN(args.input, args.hidden)
print ('model is dnn4')
elif args.arch == 'dnn_5':
import dnn_5
model = dnn_5.Classification_DNN(args.input, args.hidden)
print ('model is dnn5')
elif args.arch == 'dnn_6':
import dnn_6
model = dnn_6.Classification_DNN(args.input, args.hidden)
print ('model is dnn6')
elif args.arch == 'cnn_5':
import cnn_5
model = cnn_5.Classification_CNN(args.channel)
elif args.arch == 'cnn_6':
import cnn_6
model = cnn_6.Classification_CNN(args.channel)
else:
raise ValueError('Invalid architecture name')
if args.gpu >= 0:
cuda.get_device(args.gpu).use()
model.to_gpu()
print ("model to gpu")
# Setup optimizer
optimizer = optimizers.MomentumSGD(lr=0.01, momentum=0.9)
optimizer.setup(model)
folder = './train_result/' + args.experiment_name + '/'
t_folder = './teacher_data/'
if os.path.isdir(folder) == True:
print ('this experiment name is existed')
print ('please change experiment name')
raw_input()
else:
print ('make experiment folder')
os.makedirs(folder)
trainfile = t_folder + args.trainfile
testfile = t_folder + args.testfile
with open(folder + 'settings.txt', 'wb') as o:
o.write('epoch:' + str(args.epoch) + '\n')
o.write('modelname:' + str(model.modelname) + '\n')
o.write('input:' + str(model.input_num) + '\n')
o.write('hidden:' + str(model.hidden_num) + '\n')
o.write('layer_num:' + str(model.layer_num) + '\n')
o.write('channel_num:' + str(args.channel) + '\n')
o.write('batchsize:' + str(args.batchsize) + '\n')
o.write(args.trainfile + ':' + args.testfile + '\n')
N = sum(1 for line in open(trainfile))
print ('N = ', N)
N_test = sum(1 for line in open(testfile))
print ('N_test = ', N_test)
output_num = 1
# ------------------------------------------------------------------------------
# This example consists of three threads: data feeder, logger and trainer.
# These communicate with each other via Queue.
data_q = queue.Queue(maxsize=1)
res_q = queue.Queue()
def read_data(path, num):
# Data loading routine
line = linecache.getline(path, num)
line = line.rstrip().split(",")
linecache.clearcache()
return line
def read_batch(path, randlist):
batch = []
for i in randlist:
line = linecache.getline(path, i+1)
line = line.rstrip().split(",")
batch.append(line)
linecache.clearcache()
return batch
def read_batch2(path, randlist):
batch = []
randlist = np.sort(randlist)
f = open(path, 'rb')
reader = csv.reader(f)
i = 0
for k, row in enumerate(reader):
if k == randlist[i]:
batch.append(row)
i+=1
if i == len(randlist):
break
f.close()
return batch
def sprit_data(data):
inputlist = data[:model.input_num]
outputlist = data[-output_num-2:-2]
inputlist = np.array(inputlist).astype(np.float32)
outputlist = np.array(outputlist).astype(np.float32)
return inputlist, outputlist
def sprit_batch(listbatch):
batch = np.array(listbatch).astype(np.float32)
try:
x_batch = batch[:, :-output_num-2]
y_batch = batch[:, -3]
y_batch = np.array(y_batch).astype(np.int32)
except:
print (batch.shape)
print ("error!")
raw_input()
return x_batch, y_batch
def batchToChannel(batch,bsize,input_num):
batch = np.reshape(batch,(bsize,-1,input_num))
return batch
epoch_count=0
def feed_data():
# Data feeder
global epoch_count
count = 0
pool = multiprocessing.Pool(args.loaderjob)
data_q.put('train')
for epoch in six.moves.range(1, 1 + args.epoch):
epoch_count = epoch
print('epoch', epoch, file=sys.stderr)
print('learning rate', optimizer.lr, file=sys.stderr)
perm = np.random.permutation(N)
for i in range(0, N, args.batchsize):
batch = pool.apply_async(cyfuncs.read_batch2, (trainfile, perm[i:i + args.batchsize]))
x_batch, y_batch = sprit_batch(batch.get())
data_q.put((x_batch.copy(), y_batch.copy()))
del batch, x_batch, y_batch
gc.collect()
count += 1
if count % 100 == 0:
data_q.put('val')
for l in range(0, N_test, args.batchsize):
val_batch = pool.apply_async(cyfuncs.read_batch2, (testfile, range(l, l + args.batchsize)))
val_x_batch, val_y_batch = sprit_batch(val_batch.get())
data_q.put((val_x_batch.copy(), val_y_batch.copy()))
del val_batch, val_x_batch, val_y_batch
gc.collect()
data_q.put('train')
optimizer.lr *= 0.97
pool.close()
pool.join()
data_q.put('end')
def log_result():
# Logger
global train_loss_list, mean_error_list
train_count = 0
train_cur_loss = 0
train_cur_accuracy = 0
begin_at = time.time()
val_begin_at = None
while True:
result = res_q.get()
if result == 'end':
print(file=sys.stderr)
break
elif result == 'train':
print(file=sys.stderr)
train = True
if val_begin_at is not None:
begin_at += time.time() - val_begin_at
val_begin_at = None
continue
elif result == 'val':
print(file=sys.stderr)
train = False
val_count = val_loss = val_accuracy = 0
val_begin_at = time.time()
continue
loss, accuracy = result
if train:
train_count += 1
duration = time.time() - begin_at
throughput = train_count * args.batchsize / duration
sys.stderr.write(
'\rtrain {} updates ({} samples) time: {} ({} images/sec)'
.format(train_count, train_count * args.batchsize,
datetime.timedelta(seconds=duration), throughput))
train_cur_loss += loss
train_cur_accuracy += accuracy
if train_count % 10 == 0:
mean_loss = train_cur_loss / 10
train_loss_list.append(mean_loss)
mean_error = 1 - train_cur_accuracy / 10
mean_error_list.append(mean_error)
tools.listToCsv(folder+'trainloss.csv',train_loss_list,mean_error_list)
print(file=sys.stderr)
print(json.dumps({'type': 'train', 'iteration': train_count,
'error': mean_error, 'loss': mean_loss}))
sys.stdout.flush()
train_cur_loss = 0
train_cur_accuracy = 0
else:
val_count += args.batchsize
duration = time.time() - val_begin_at
throughput = val_count / duration
sys.stderr.write(
'\rval {} batches ({} samples) time: {} ({} images/sec)'
.format(val_count / args.batchsize, val_count,
datetime.timedelta(seconds=duration), throughput))
val_loss += loss
val_accuracy += accuracy
if val_count == 10000:
mean_loss = val_loss * args.batchsize / 10000
mean_error = 1 - val_accuracy * args.batchsize / 10000
print(file=sys.stderr)
print(json.dumps({'type': 'val', 'iteration': train_count,
'error': mean_error, 'loss': mean_loss}))
sys.stdout.flush()
def train_loop():
# Trainer
graph_generated = False
while True:
while data_q.empty():
time.sleep(0.1)
inp = data_q.get()
if inp == 'end': # quit
res_q.put('end')
break
elif inp == 'train': # restart training
res_q.put('train')
train = True
continue
elif inp == 'val': # start validation
res_q.put('val')
pickle.dump(model, open(folder+'model', 'wb'), -1)
train = False
continue
x = xp.asarray(inp[0])
y = xp.asarray(inp[1])
if train:
optimizer.zero_grads()
loss, accuracy = model.forward(x, y)
loss.backward()
optimizer.update()
if not graph_generated:
with open('graph.dot', 'w') as o:
o.write(c.build_computational_graph((loss,), False).dump())
with open('graph.wo_split.dot', 'w') as o:
o.write(c.build_computational_graph((loss,), True).dump())
print('generated graph')
graph_generated = True
else:
loss, accuracy = model.forward(x, y, train=False)
if epoch_count % 2 == 0:
print ('save model')
model.to_cpu()
with open(folder + 'model_' + str(epoch_count), 'wb') as o:
pickle.dump(model, o)
model.to_gpu()#もう一度GPUに戻すのか?
optimizer.setup(model)
res_q.put((float(loss.data),
float(accuracy.data)))
del loss, accuracy, x, y
train_loss_list = []
mean_error_list = []
# Invoke threads
feeder = threading.Thread(target=feed_data)
feeder.daemon = True
feeder.start()
logger = threading.Thread(target=log_result)
logger.daemon = True
logger.start()
train_loop()
feeder.join()
logger.join()
with open(folder + 'loss.csv', 'wb') as oc:
odata = []
odata.append(train_loss_list)
odata.append(test_loss_list)
odata = np.array(odata).transpose()
writer = csv.writer(oc)
writer.writerows(odata)
print ('save loss.csv')
if args.gpu >= 0:
print ('model to cpu')
model.to_cpu()
#pickle.dump(model, open("model", 'wb'), -1)
with open(folder + 'final_model', 'wb') as o:
pickle.dump(model, o)
print ("model saved")
print ("finished!!!")