-
Notifications
You must be signed in to change notification settings - Fork 28
/
utils.py
executable file
·262 lines (211 loc) · 7.06 KB
/
utils.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
''' Code partly from https://github.com/ShichenLiu/CondenseNet/blob/master/utils.py '''
import torch
from torch.autograd import Variable
import os
import sys
import time
def get_num_gen(gen):
return sum(1 for x in gen)
def is_leaf(model):
return get_num_gen(model.children()) == 0
def get_layer_info(layer):
layer_str = str(layer)
type_name = layer_str[:layer_str.find('(')].strip()
return type_name
def get_layer_param(model):
import operator
import functools
return sum([functools.reduce(operator.mul, i.size(), 1) for i in model.parameters()])
def measure_layer(layer, x):
global count_ops, count_params
delta_ops = 0
delta_params = 0
multi_add = 1
type_name = get_layer_info(layer)
# ops_conv
if type_name in ['Conv2d']:
out_h = int((x.size()[2] + 2 * layer.padding[0] - layer.kernel_size[0]) /
layer.stride[0] + 1)
out_w = int((x.size()[3] + 2 * layer.padding[1] - layer.kernel_size[1]) /
layer.stride[1] + 1)
delta_ops = layer.in_channels * layer.out_channels * layer.kernel_size[0] * \
layer.kernel_size[1] * out_h * out_w / layer.groups * multi_add
delta_params = get_layer_param(layer)
# ops_nonlinearity
elif type_name in ['ReLU']:
delta_ops = x.numel() / x.size(0)
delta_params = get_layer_param(layer)
# ops_pooling
elif type_name in ['AvgPool2d']:
in_w = x.size()[2]
kernel_ops = layer.kernel_size * layer.kernel_size
out_w = int((in_w + 2 * layer.padding - layer.kernel_size) / layer.stride + 1)
out_h = int((in_w + 2 * layer.padding - layer.kernel_size) / layer.stride + 1)
delta_ops = x.size()[1] * out_w * out_h * kernel_ops
delta_params = get_layer_param(layer)
elif type_name in ['AdaptiveAvgPool2d']:
delta_ops = x.size()[1] * x.size()[2] * x.size()[3]
delta_params = get_layer_param(layer)
# ops_linear
elif type_name in ['Linear']:
weight_ops = layer.weight.numel() * multi_add
bias_ops = layer.bias.numel()
delta_ops = weight_ops + bias_ops
delta_params = get_layer_param(layer)
# ops_nothing
elif type_name in ['BatchNorm2d', 'Dropout2d', 'DropChannel', 'Dropout']:
delta_params = get_layer_param(layer)
# unknown layer type
else:
delta_params = get_layer_param(layer)
count_ops += delta_ops
count_params += delta_params
return
def measure_model(model, H, W):
global count_ops, count_params
count_ops = 0
count_params = 0
data = Variable(torch.zeros(1, 3, H, W))
def should_measure(x):
return is_leaf(x)
def modify_forward(model):
for child in model.children():
if should_measure(child):
def new_forward(m):
def lambda_forward(x):
measure_layer(m, x)
return m.old_forward(x)
return lambda_forward
child.old_forward = child.forward
child.forward = new_forward(child)
else:
modify_forward(child)
def restore_forward(model):
for child in model.children():
# leaf node
if is_leaf(child) and hasattr(child, 'old_forward'):
child.forward = child.old_forward
child.old_forward = None
else:
restore_forward(child)
modify_forward(model)
model.forward(data)
restore_forward(model)
return count_ops, count_params
class AverageMeter(object):
"""Computes and stores the average and current value"""
def __init__(self):
self.reset()
def reset(self):
self.val = 0
self.avg = 0
self.sum = 0
self.count = 0
def update(self, val, n=1):
self.val = val
self.sum += val * n
self.count += n
if self.count > 0:
self.avg = self.sum / self.count
def accumulate(self, val, n=1):
self.sum += val
self.count += n
if self.count > 0:
self.avg = self.sum / self.count
def accuracy(output, target, topk=(1,)):
"""Computes the precision@k for the specified values of k"""
batch_size = target.size(0)
num = output.size(1)
target_topk = []
appendices = []
for k in topk:
if k <= num:
target_topk.append(k)
else:
appendices.append([0.0])
topk = target_topk
maxk = max(topk)
_, pred = output.topk(maxk, 1, True, True)
pred = pred.t()
correct = pred.eq(target.view(1, -1).expand_as(pred))
res = []
for k in topk:
correct_k = correct[:k].view(-1).float().sum(0)
res.append(correct_k.mul_(100.0 / batch_size))
return res + appendices
def process_state_dict(state_dict):
# process state dict so that it can be loaded by normal models
for k in list(state_dict.keys()):
state_dict[k.replace('module.', '')] = state_dict.pop(k)
return state_dict
# Custom progress bar
_, term_width = os.popen('stty size', 'r').read().split()
term_width = int(term_width)
TOTAL_BAR_LENGTH = 40.
last_time = time.time()
begin_time = last_time
def format_time(seconds):
days = int(seconds / 3600/24)
seconds = seconds - days*3600*24
hours = int(seconds / 3600)
seconds = seconds - hours*3600
minutes = int(seconds / 60)
seconds = seconds - minutes*60
secondsf = int(seconds)
seconds = seconds - secondsf
millis = int(seconds*1000)
f = ''
i = 1
if days > 0:
f += str(days) + 'D'
i += 1
if hours > 0 and i <= 2:
f += str(hours) + 'h'
i += 1
if minutes > 0 and i <= 2:
f += str(minutes) + 'm'
i += 1
if secondsf > 0 and i <= 2:
f += str(secondsf) + 's'
i += 1
if millis > 0 and i <= 2:
f += str(millis) + 'ms'
i += 1
if f == '':
f = '0ms'
return f
def progress_bar(current, total, msg=None):
global last_time, begin_time
if current == 0:
begin_time = time.time() # Reset for new bar.
cur_len = int(TOTAL_BAR_LENGTH*current/total)
rest_len = int(TOTAL_BAR_LENGTH - cur_len) - 1
sys.stdout.write(' [')
for i in range(cur_len):
sys.stdout.write('=')
sys.stdout.write('>')
for i in range(rest_len):
sys.stdout.write('.')
sys.stdout.write(']')
cur_time = time.time()
step_time = cur_time - last_time
last_time = cur_time
tot_time = cur_time - begin_time
L = []
L.append(' Step: %s' % format_time(step_time))
L.append(' | Tot: %s' % format_time(tot_time))
if msg:
L.append(' | ' + msg)
msg = ''.join(L)
sys.stdout.write(msg)
for i in range(term_width-int(TOTAL_BAR_LENGTH)-len(msg)-3):
sys.stdout.write(' ')
# Go back to the center of the bar.
for i in range(term_width-int(TOTAL_BAR_LENGTH/2)+2):
sys.stdout.write('\b')
sys.stdout.write(' %d/%d ' % (current+1, total))
if current < total-1:
sys.stdout.write('\r')
else:
sys.stdout.write('\n')
sys.stdout.flush()