-
Notifications
You must be signed in to change notification settings - Fork 1
/
prune_cifar.py
211 lines (160 loc) · 8.51 KB
/
prune_cifar.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
import torch
from torch.autograd import Variable
from torchvision import models
#import cv2
import argparse
import sys
import time
import math
import copy
import logging
import json
import numpy as np
import torchvision
import torch.nn.functional as F
# from sklearn.metrics.pairwise import rbf_kernel
from scipy.spatial.distance import pdist, squareform
import scipy.spatial as sp
# from sklearn.decomposition import PCA
# from sklearn.decomposition import KernelPCA
# from Spectral_Clustering.spectral_clustering import Spectral_Clustering
# from same_size_dbscan import Same_Size_DBSCAN
from gkp_utils import *
logger = logging.getLogger("Test")
logger.info('All modules imported')
class ModifiedResNet(torch.nn.Module):
def __init__(self, model, original_model, pruning_rate = 0.4375, setting = None):
super(ModifiedResNet, self).__init__()
self.conv1 = model.conv_1_3x3
self.bn1 = model.bn_1
self.layer1 = model.stage_1
self.layer2 = model.stage_2
self.layer3 = model.stage_3
self.avgpool = model.avgpool
self.linear = model.classifier
self.original_model = original_model
self.kernel_gcd = float('inf')
# self.model = model
self.pruning_rate = pruning_rate
self.pruning_target_layers = setting['prune_params']['pruning_target_layers']
self.grouping_target_layers = setting['prune_params']['grouping_target_layers']
self.n_clusters = setting['prune_params']['n_clusters']
self.clustering_method = setting['prune_params']['clustering_method']
self.pruning_strategy = setting['prune_params']['pruning_strategy']
logger.info(f"clustering_method: {self.clustering_method}; n_clusters: {self.n_clusters}; pruning_target_layers: {self.pruning_target_layers}; grouping_target_layers: {self.grouping_target_layers}.")
logger.info(f"pruning_strategy: {self.pruning_strategy}; pruning_rate: {self.pruning_rate}.")
if self.pruning_strategy =='smooth_cost_beam':
logger.info(f"\tmetric: {setting['prune_params']['metric']}; inner_outer_balancer: {setting['prune_params']['inner_outer_balancer']}; cost_smooth_balancer: {setting['prune_params']['cost_smooth_balancer']}; eval_kept_kernel_number: {setting['prune_params']['eval_kept_kernel_number']}; beam_width: {setting['prune_params']['beam_width']}; smoothness_check_step: {setting['prune_params']['smoothness_check_step']}.")
for current_block, layer, sublayer, modules, submodule in self.gen_unpruned_block(model):
modules[sublayer] = self.prune_block(current_block, submodule, (layer, sublayer), setting = setting)
def gen_unpruned_block(self, model):
# imagenet_target_layers = [4, 5, 6, 7]
# tiny_imagenet_target_layers = [2, 3, 4]
# cifar10_target_layers = [2, 3, 4]
#
# if self.dataset == 'imagenet':
# target_layers = imagenet_target_layers
# elif self.dataset == 'tiny_imagenet':
# target_layers = tiny_imagenet_target_layers
# elif self.dataset == 'cifar10':
# target_layers = cifar10_target_layers
# else:
# logger.error(f'Invalid dataset input {dataset}.')
# sys.exit(0)
for layer, (name, modules) in enumerate(model._modules.items()):
if layer in self.pruning_target_layers:
for sublayer, (name, submodule) in enumerate(modules._modules.items()):
current_block = modules[sublayer]
yield current_block, layer, sublayer, modules, submodule
def prune_block(self, current_block, submodule, block_info, setting = None):
layer, sublayer = block_info
block_out_list = []
block_layer_list = []
# if pruned_flag:
# block_in_planes = None
# block_planes = None
# else:
block_in_planes = submodule.conv_a.in_channels
block_planes = submodule.conv_a.out_channels
block_prune_masks = []
block_candidate_methods_list = []
block_preserved_kernel_index = []
block_layer_info = []
for subsublayer, (name, module) in enumerate(submodule._modules.items()):
if isinstance(module, torch.nn.modules.conv.Conv2d):
new_conv = make_new_conv(module)
old_weights = module.weight.cuda()
old_out_channels, old_in_channels, old_kernel_size, old_kernel_size = old_weights.data.size()
old_weights = old_weights.data.cpu().numpy()
original_shape = old_weights.shape
self.kernel_gcd = min(self.kernel_gcd, original_shape[0])
old_weights_float = torch.from_numpy(old_weights).float()
layer_info = (layer, sublayer, subsublayer)
# if not pruned_flag:
block_layer_info.append(layer_info)
old_weights = old_weights.reshape(old_out_channels, old_in_channels*old_kernel_size*old_kernel_size)
preferred_permutation_matrix, preferred_clustering_method = get_cluster_permutation_matrix(old_weights, old_out_channels, n_clusters = self.n_clusters, clustering_method = self.clustering_method)
clustering_info = (preferred_permutation_matrix, preferred_clustering_method)
block_candidate_methods_list.append(clustering_info)
logger.info(f'Layer {layer}-{sublayer}-{subsublayer}; Shape {original_shape} -> {old_weights.shape}; Method: {preferred_clustering_method}.')
block_out_index = get_out_index(preferred_permutation_matrix.transpose(1,0)).cuda()
block_out_index = Variable(block_out_index)
block_out_list.append(block_out_index)
new_weights = np.dot(preferred_permutation_matrix, old_weights)
new_weights = new_weights.reshape(old_out_channels, old_in_channels, old_kernel_size, old_kernel_size)
new_weights = Variable(torch.from_numpy(new_weights)).cuda()
if subsublayer == 0:
conv_num = 0
elif subsublayer == 2:
conv_num = 1
else:
logger.error(f'subsublayer: {subsublayer} is neither 0 or 2.')
sys.exit(0)
new_conv, new_conv_prune_mask, new_conv_preserved_kernel_index = prune_kernels(current_block, conv_num, new_conv, new_weights, old_out_channels,
pruning_rate = self.pruning_rate,
n_clusters = self.n_clusters,
pruning_strategy = setting['prune_params']['pruning_strategy'],
setting = setting
)
block_prune_masks.append(new_conv_prune_mask)
block_layer_list.append(new_conv)
block_preserved_kernel_index.append(new_conv_preserved_kernel_index)
return NewBasicblock(submodule, submodule.bn_a, submodule.bn_b, block_out_list, block_layer_list, block_prune_masks, block_candidate_methods_list, block_preserved_kernel_index, in_planes = block_in_planes, planes = block_planes)
def forward(self, x):
x = F.relu(self.bn1(self.conv1(x)))
x = self.layer1(x)
x = self.layer2(x)
x = self.layer3(x)
x = self.avgpool(x)
x = x.view(x.size(0), -1)
x = self.linear(x)
return x
class NewBasicblock(torch.nn.Module):
expansion = 1
def __init__(self, module, bn1, bn2, out_list, layer_list, prune_mask, candidate_methods_list, preserved_kernel_index, in_planes = None, planes = None, stride=1, shortcut=None):
super(NewBasicblock, self).__init__()
self.out_list = out_list
self.conv1 = layer_list[0]
self.bn1 = bn1
self.conv2 = layer_list[1]
self.bn2 = bn2
self.shortcut = shortcut
self.prune_mask = prune_mask
self.candidate_methods_list = candidate_methods_list
self.preserved_kernel_index = preserved_kernel_index
if stride != 1 or in_planes != self.expansion * planes:
self.shortcut = module.downsample
def forward(self, x):
residual = x
out = self.conv1(x)
out = torch.index_select(out, 1, self.out_list[0])
out = self.bn1(out)
out = F.relu(out)
out = self.conv2(out)
out = torch.index_select(out, 1, self.out_list[1])
out = self.bn2(out)
if self.shortcut is not None:
residual = self.shortcut(x)
out += residual
out = F.relu(out)
return out