-
Notifications
You must be signed in to change notification settings - Fork 0
/
gradient_descent_output.txt
223 lines (192 loc) · 6.61 KB
/
gradient_descent_output.txt
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
Training on device cpu.
Global parameters:
batch_size = 256
n_epoch = 30
loss_fn = CrossEntropyLoss()
seed = 265
=========================================================
Current parameters:
lr = 0.01
mom = 0
decay = 0
--------- Using Pytorch's SGD ---------
15:27:35.605270 | Epoch 1 | Training loss 0.68093
15:27:40.688064 | Epoch 5 | Training loss 0.54943
15:27:47.054778 | Epoch 10 | Training loss 0.46524
15:27:53.257568 | Epoch 15 | Training loss 0.42144
15:27:59.615822 | Epoch 20 | Training loss 0.38399
15:28:04.971721 | Epoch 25 | Training loss 0.35094
15:28:10.766444 | Epoch 30 | Training loss 0.31793
--- Accuracies ---
Training
Accuracy: 0.88
Validation
Accuracy: 0.83
--------- Using manual update ----------
15:28:13.014504 | Epoch 1 | Training loss 0.68093
15:28:19.681324 | Epoch 5 | Training loss 0.54943
15:28:26.238622 | Epoch 10 | Training loss 0.46524
15:28:32.575133 | Epoch 15 | Training loss 0.42144
15:28:39.409641 | Epoch 20 | Training loss 0.38399
15:28:47.866196 | Epoch 25 | Training loss 0.35094
15:28:54.901863 | Epoch 30 | Training loss 0.31793
--- Accuracies ---
Training
Accuracy: 0.88
Validation
Accuracy: 0.83
=========================================================
Current parameters:
lr = 0.01
mom = 0
decay = 0.01
--------- Using Pytorch's SGD ---------
15:28:56.584109 | Epoch 1 | Training loss 0.68103
15:29:02.281089 | Epoch 5 | Training loss 0.55376
15:29:09.971498 | Epoch 10 | Training loss 0.47008
15:29:19.032535 | Epoch 15 | Training loss 0.42922
15:29:28.241046 | Epoch 20 | Training loss 0.39408
15:29:36.335902 | Epoch 25 | Training loss 0.36438
15:29:44.158214 | Epoch 30 | Training loss 0.33567
--- Accuracies ---
Training
Accuracy: 0.87
Validation
Accuracy: 0.83
--------- Using manual update ----------
15:29:47.163674 | Epoch 1 | Training loss 0.68103
15:29:56.388759 | Epoch 5 | Training loss 0.55376
15:30:05.207449 | Epoch 10 | Training loss 0.47008
15:30:13.610137 | Epoch 15 | Training loss 0.42922
15:30:21.929857 | Epoch 20 | Training loss 0.39408
15:30:30.404753 | Epoch 25 | Training loss 0.36438
15:30:39.078477 | Epoch 30 | Training loss 0.33567
--- Accuracies ---
Training
Accuracy: 0.87
Validation
Accuracy: 0.83
=========================================================
Current parameters:
lr = 0.01
mom = 0.9
decay = 0
--------- Using Pytorch's SGD ---------
15:30:41.131709 | Epoch 1 | Training loss 0.61787
15:30:46.561217 | Epoch 5 | Training loss 0.36053
15:30:53.532891 | Epoch 10 | Training loss 0.22891
15:31:00.975329 | Epoch 15 | Training loss 0.25503
15:31:07.768451 | Epoch 20 | Training loss 0.15373
15:31:14.622177 | Epoch 25 | Training loss 0.16546
15:31:21.671228 | Epoch 30 | Training loss 0.07555
--- Accuracies ---
Training
Accuracy: 0.97
Validation
Accuracy: 0.83
--------- Using manual update ----------
15:31:23.931476 | Epoch 1 | Training loss 0.61787
15:31:30.849225 | Epoch 5 | Training loss 0.36053
15:31:39.036897 | Epoch 10 | Training loss 0.22891
15:31:47.561861 | Epoch 15 | Training loss 0.25503
15:31:56.124781 | Epoch 20 | Training loss 0.15373
15:32:04.366268 | Epoch 25 | Training loss 0.16546
15:32:14.073238 | Epoch 30 | Training loss 0.07555
--- Accuracies ---
Training
Accuracy: 0.97
Validation
Accuracy: 0.83
=========================================================
Current parameters:
lr = 0.01
mom = 0.9
decay = 0.01
--------- Using Pytorch's SGD ---------
15:32:16.341249 | Epoch 1 | Training loss 0.61972
15:32:23.350286 | Epoch 5 | Training loss 0.37262
15:32:31.482353 | Epoch 10 | Training loss 0.26217
15:32:40.051164 | Epoch 15 | Training loss 0.20043
15:32:49.058091 | Epoch 20 | Training loss 0.19053
15:32:57.452878 | Epoch 25 | Training loss 0.15727
15:33:05.959197 | Epoch 30 | Training loss 0.08699
--- Accuracies ---
Training
Accuracy: 0.96
Validation
Accuracy: 0.84
--------- Using manual update ----------
15:33:08.830914 | Epoch 1 | Training loss 0.61972
15:33:17.195033 | Epoch 5 | Training loss 0.37262
15:33:27.630327 | Epoch 10 | Training loss 0.26217
15:33:39.487334 | Epoch 15 | Training loss 0.20043
15:33:49.997598 | Epoch 20 | Training loss 0.19053
15:33:59.917693 | Epoch 25 | Training loss 0.15727
15:34:09.844068 | Epoch 30 | Training loss 0.08699
--- Accuracies ---
Training
Accuracy: 0.96
Validation
Accuracy: 0.84
=========================================================
Current parameters:
lr = 0.01
mom = 0.9
decay = 0.001
--------- Using Pytorch's SGD ---------
15:34:11.804257 | Epoch 1 | Training loss 0.61805
15:34:17.670192 | Epoch 5 | Training loss 0.36084
15:34:25.666121 | Epoch 10 | Training loss 0.22181
15:34:33.149762 | Epoch 15 | Training loss 0.24899
15:34:41.312123 | Epoch 20 | Training loss 0.11224
15:34:49.423027 | Epoch 25 | Training loss 0.12177
15:34:57.055142 | Epoch 30 | Training loss 0.05682
--- Accuracies ---
Training
Accuracy: 0.98
Validation
Accuracy: 0.83
--------- Using manual update ----------
15:34:59.260751 | Epoch 1 | Training loss 0.61805
15:35:06.841876 | Epoch 5 | Training loss 0.36084
15:35:16.691645 | Epoch 10 | Training loss 0.22181
15:35:28.186123 | Epoch 15 | Training loss 0.24899
15:35:38.790608 | Epoch 20 | Training loss 0.11224
15:35:48.491566 | Epoch 25 | Training loss 0.12177
15:35:57.542884 | Epoch 30 | Training loss 0.05682
--- Accuracies ---
Training
Accuracy: 0.98
Validation
Accuracy: 0.83
=========================================================
Current parameters:
lr = 0.01
mom = 0.8
decay = 0.01
--------- Using Pytorch's SGD ---------
15:35:59.448439 | Epoch 1 | Training loss 0.64125
15:36:05.807860 | Epoch 5 | Training loss 0.40867
15:36:15.054189 | Epoch 10 | Training loss 0.31498
15:36:25.051344 | Epoch 15 | Training loss 0.23144
15:36:32.358912 | Epoch 20 | Training loss 0.18008
15:36:39.851710 | Epoch 25 | Training loss 0.19677
15:36:47.477805 | Epoch 30 | Training loss 0.14508
--- Accuracies ---
Training
Accuracy: 0.88
Validation
Accuracy: 0.80
--------- Using manual update ----------
15:36:49.914018 | Epoch 1 | Training loss 0.64125
15:36:57.391801 | Epoch 5 | Training loss 0.40867
15:37:06.490094 | Epoch 10 | Training loss 0.31498
15:37:16.265842 | Epoch 15 | Training loss 0.23144
15:37:25.813427 | Epoch 20 | Training loss 0.18008
15:37:34.834043 | Epoch 25 | Training loss 0.19677
15:37:44.176163 | Epoch 30 | Training loss 0.14508
--- Accuracies ---
Training
Accuracy: 0.88
Validation
Accuracy: 0.80