In [1]:
%load_ext autoreload
%autoreload 2
import cPickle as pickle
import os; import sys; sys.path.append('..')
import gp
import gp.nets as nets
from nolearn.lasagne.visualize import plot_loss
from nolearn.lasagne.visualize import plot_conv_weights
from nolearn.lasagne.visualize import plot_conv_activity
from nolearn.lasagne.visualize import plot_occlusion
from matplotlib.pyplot import imshow
import matplotlib.pyplot as plt
%matplotlib inline
Using gpu device 0: GeForce GTX TITAN (CNMeM is disabled, CuDNN 4007)
/home/d/nolearn/local/lib/python2.7/site-packages/theano/tensor/signal/downsample.py:6: UserWarning: downsample module has been moved to the theano.tensor.signal.pool module.
"downsample module has been moved to the theano.tensor.signal.pool module.")
In [2]:
PATCH_PATH = ('cylinder2_rgb')
In [3]:
X_train, y_train, X_test, y_test = gp.Patch.load_rgb(PATCH_PATH)
Loaded /home/d/patches//cylinder2_rgb/ in 0.154430866241 seconds.
---------------------------------------------------------------------------
MemoryError Traceback (most recent call last)
<ipython-input-3-85b5c04080c1> in <module>()
----> 1 X_train, y_train, X_test, y_test = gp.Patch.load_rgb(PATCH_PATH)
/home/d/Projects/gp/gp/patch.pyc in load_rgb(PATCH_PATH, patch_size, verbose)
652 print 'Loaded', PATCH_PATH, 'in', time.time()-t0, 'seconds.'
653
--> 654 return training['rgb'], training_targets['targets'].astype(np.uint8), test['rgb'], test_targets['targets'].astype(np.uint8)
655
656
/home/d/nolearn/local/lib/python2.7/site-packages/numpy/lib/npyio.pyc in __getitem__(self, key)
222 return format.read_array(bytes,
223 allow_pickle=self.allow_pickle,
--> 224 pickle_kwargs=self.pickle_kwargs)
225 else:
226 return self.zip.read(key)
/home/d/nolearn/local/lib/python2.7/site-packages/numpy/lib/format.pyc in read_array(fp, allow_pickle, pickle_kwargs)
659 max_read_count = BUFFER_SIZE // min(BUFFER_SIZE, dtype.itemsize)
660
--> 661 array = numpy.empty(count, dtype=dtype)
662 for i in range(0, count, max_read_count):
663 read_count = min(max_read_count, count - i)
MemoryError:
In [5]:
gp.Util.view_rgba(X_train[200], y_train[200])
In [6]:
cnn = nets.RGBNetPlus()
CNN configuration:
Our CNN with image, prob, merged_array as RGB.
This includes dropout. This also includes more layers.
In [7]:
cnn = cnn.fit(X_train, y_train)
# Neural Network with 170898 learnable parameters
## Layer information
# name size
--- -------- --------
0 input 3x75x75
1 conv1 64x73x73
2 pool1 64x36x36
3 dropout1 64x36x36
4 conv2 48x34x34
5 pool2 48x17x17
6 dropout2 48x17x17
7 conv3 48x15x15
8 pool3 48x7x7
9 dropout3 48x7x7
10 conv4 48x5x5
11 pool4 48x2x2
12 dropout4 48x2x2
13 hidden5 512
14 dropout5 512
15 output 2
epoch trn loss val loss trn/val valid acc dur
------- ---------- ---------- --------- ----------- ------
1 0.27873 0.24810 1.12346 0.90708 85.21s
2 0.22028 0.21378 1.03039 0.92169 85.39s
3 0.20895 0.19618 1.06504 0.92948 85.35s
4 0.20218 0.18869 1.07149 0.93113 85.54s
5 0.19603 0.18712 1.04762 0.93278 85.19s
6 0.19252 0.17948 1.07266 0.93471 85.36s
7 0.18680 0.18103 1.03187 0.93485 85.81s
8 0.18355 0.17366 1.05697 0.93760 85.25s
9 0.18008 0.17935 1.00404 0.93499 85.89s
10 0.17775 0.17329 1.02572 0.93753 85.70s
11 0.17482 0.16781 1.04179 0.93904 85.53s
12 0.17185 0.16516 1.04050 0.94083 85.55s
13 0.16969 0.16125 1.05236 0.94227 85.48s
14 0.16738 0.15513 1.07893 0.94365 85.47s
15 0.16517 0.15445 1.06935 0.94370 85.33s
16 0.16307 0.15639 1.04275 0.94255 88.01s
17 0.15990 0.15819 1.01080 0.94285 88.01s
18 0.15896 0.15170 1.04790 0.94440 87.99s
19 0.15779 0.15301 1.03126 0.94463 88.01s
20 0.15562 0.14927 1.04253 0.94667 88.06s
21 0.15482 0.14582 1.06171 0.94775 87.97s
22 0.15275 0.14795 1.03241 0.94587 88.04s
23 0.15160 0.14472 1.04755 0.94837 88.10s
24 0.14991 0.14482 1.03513 0.94713 87.98s
25 0.14916 0.14039 1.06246 0.94989 87.98s
26 0.14720 0.13808 1.06604 0.95005 88.03s
27 0.14622 0.14942 0.97857 0.94526 87.97s
28 0.14548 0.14103 1.03159 0.95000 87.96s
29 0.14381 0.14236 1.01016 0.94883 87.96s
30 0.14271 0.13701 1.04157 0.95035 87.97s
31 0.14201 0.13296 1.06806 0.95287 88.03s
32 0.14151 0.13404 1.05570 0.95183 88.03s
33 0.14043 0.14000 1.00312 0.95060 87.92s
34 0.13968 0.13116 1.06492 0.95236 87.94s
35 0.13935 0.13645 1.02123 0.94975 87.90s
36 0.13744 0.13205 1.04083 0.95278 87.90s
37 0.13860 0.13221 1.04836 0.95293 87.90s
38 0.13679 0.13110 1.04339 0.95293 87.91s
39 0.13643 0.12808 1.06517 0.95435 87.95s
40 0.13472 0.12677 1.06273 0.95628 88.00s
41 0.13432 0.12534 1.07166 0.95654 87.98s
42 0.13399 0.13469 0.99477 0.95135 88.00s
43 0.13286 0.12418 1.06986 0.95598 87.91s
44 0.13330 0.12261 1.08721 0.95658 88.51s
45 0.13157 0.12413 1.06000 0.95603 88.21s
46 0.13216 0.12720 1.03895 0.95440 87.91s
47 0.13219 0.12497 1.05773 0.95557 87.95s
48 0.12981 0.12224 1.06190 0.95639 87.89s
49 0.13045 0.11843 1.10149 0.95834 87.93s
50 0.13020 0.12251 1.06277 0.95713 88.01s
51 0.12953 0.12313 1.05193 0.95779 87.97s
52 0.12875 0.13102 0.98264 0.95271 87.94s
53 0.12982 0.11812 1.09902 0.95880 88.01s
54 0.12883 0.13252 0.97212 0.95328 87.98s
55 0.12780 0.11614 1.10041 0.95873 88.00s
56 0.12697 0.12227 1.03841 0.95660 87.91s
57 0.12640 0.12053 1.04872 0.95805 87.96s
58 0.12676 0.11708 1.08272 0.95850 87.90s
59 0.12671 0.11728 1.08040 0.95803 87.91s
60 0.12536 0.11572 1.08328 0.95991 87.90s
61 0.12638 0.11775 1.07330 0.95832 88.11s
62 0.12479 0.11929 1.04612 0.95924 88.01s
63 0.12464 0.11665 1.06851 0.95825 87.97s
64 0.12448 0.11715 1.06252 0.95825 88.03s
65 0.12422 0.11772 1.05525 0.95835 87.97s
66 0.12473 0.11723 1.06398 0.95747 87.95s
67 0.12333 0.10759 1.14625 0.96266 87.95s
68 0.12308 0.10927 1.12638 0.96219 87.97s
69 0.12341 0.10928 1.12930 0.96059 88.05s
70 0.12205 0.11428 1.06803 0.95933 87.91s
71 0.12139 0.10694 1.13509 0.96304 87.96s
72 0.12248 0.11420 1.07248 0.95848 88.00s
73 0.12090 0.10778 1.12174 0.96204 87.97s
74 0.12179 0.11176 1.08977 0.96068 87.94s
75 0.12174 0.10840 1.12306 0.96084 87.97s
76 0.11956 0.11055 1.08150 0.96094 87.96s
77 0.12128 0.11066 1.09591 0.96190 87.93s
78 0.12193 0.12199 0.99955 0.95511 87.91s
79 0.11955 0.11008 1.08601 0.96249 88.06s
80 0.11889 0.10867 1.09409 0.96112 87.78s
81 0.12036 0.10981 1.09613 0.95910 88.40s
82 0.11815 0.10691 1.10516 0.96194 87.97s
83 0.11913 0.10299 1.15676 0.96304 88.10s
84 0.12028 0.10571 1.13787 0.96311 88.38s
85 0.11984 0.10599 1.13065 0.96240 88.20s
86 0.12083 0.11013 1.09712 0.96027 88.18s
87 0.11944 0.10531 1.13415 0.96314 88.24s
88 0.11928 0.10611 1.12418 0.96341 88.14s
89 0.11856 0.10435 1.13615 0.96295 88.16s
90 0.11754 0.10387 1.13163 0.96328 87.86s
91 0.11918 0.10522 1.13266 0.96265 87.79s
92 0.12018 0.11795 1.01892 0.95818 87.74s
93 0.11932 0.10478 1.13879 0.96362 87.84s
94 0.11769 0.11232 1.04774 0.95866 87.89s
95 0.11661 0.12023 0.96990 0.95466 87.82s
96 0.11826 0.10466 1.12997 0.96348 87.84s
97 0.11685 0.10239 1.14119 0.96396 87.81s
98 0.11778 0.10560 1.11529 0.96250 87.82s
99 0.11739 0.10404 1.12830 0.96371 87.80s
100 0.11740 0.10840 1.08304 0.96229 87.75s
101 0.11804 0.10348 1.14074 0.96334 87.93s
102 0.11711 0.11157 1.04969 0.95921 87.79s
103 0.11746 0.11358 1.03422 0.95942 87.79s
104 0.11670 0.11401 1.02362 0.95727 87.75s
105 0.11728 0.10741 1.09189 0.96020 87.80s
106 0.11614 0.10605 1.09508 0.96123 87.76s
107 0.11614 0.11139 1.04267 0.95889 87.81s
108 0.11693 0.10825 1.08020 0.96197 87.75s
109 0.11699 0.10569 1.10684 0.96224 87.79s
110 0.11614 0.10662 1.08923 0.96362 87.86s
111 0.11670 0.10488 1.11271 0.96148 87.82s
112 0.11533 0.10844 1.06359 0.95880 88.05s
113 0.11553 0.10357 1.11557 0.96245 88.14s
114 0.11588 0.10877 1.06534 0.96066 88.17s
115 0.11504 0.10295 1.11740 0.96203 88.19s
116 0.11445 0.11089 1.03213 0.95846 88.11s
117 0.11585 0.11174 1.03679 0.95832 87.96s
118 0.11523 0.10163 1.13376 0.96426 88.14s
119 0.11479 0.10017 1.14597 0.96446 88.46s
120 0.11556 0.10223 1.13043 0.96376 87.67s
121 0.11522 0.11072 1.04061 0.95905 87.73s
122 0.11562 0.10751 1.07548 0.96094 87.73s
123 0.11472 0.10589 1.08345 0.96259 87.77s
124 0.11380 0.10267 1.10835 0.96323 87.81s
125 0.11361 0.10009 1.13502 0.96302 87.71s
126 0.11414 0.10427 1.09468 0.96169 87.72s
127 0.11391 0.10139 1.12358 0.96313 87.70s
128 0.11353 0.10267 1.10572 0.96155 87.70s
129 0.11449 0.10288 1.11277 0.96231 87.76s
130 0.11382 0.10352 1.09953 0.96176 87.66s
131 0.11229 0.09834 1.14183 0.96561 87.70s
132 0.11387 0.10316 1.10388 0.96172 87.68s
133 0.11395 0.09652 1.18049 0.96706 87.80s
134 0.11275 0.10374 1.08684 0.96204 87.67s
135 0.11410 0.09922 1.14994 0.96421 87.78s
136 0.11186 0.09727 1.14997 0.96504 87.67s
137 0.11306 0.10658 1.06079 0.95990 87.71s
138 0.11324 0.11294 1.00265 0.95763 87.72s
139 0.11259 0.10676 1.05464 0.95903 87.74s
140 0.11319 0.10458 1.08233 0.96100 87.67s
141 0.11388 0.10107 1.12669 0.96291 87.76s
142 0.11374 0.09748 1.16683 0.96564 87.89s
143 0.11160 0.09714 1.14885 0.96520 88.05s
144 0.11273 0.10164 1.10907 0.96398 88.11s
145 0.11269 0.09863 1.14257 0.96188 88.06s
146 0.11333 0.10072 1.12523 0.96298 88.06s
147 0.11225 0.10062 1.11562 0.96336 88.18s
148 0.11279 0.09849 1.14515 0.96421 88.09s
149 0.11223 0.09417 1.19178 0.96678 87.78s
150 0.11246 0.10348 1.08677 0.96282 87.80s
151 0.11390 0.09674 1.17731 0.96543 87.77s
152 0.11126 0.10163 1.09474 0.96181 87.72s
153 0.11333 0.10610 1.06810 0.96153 87.71s
154 0.11340 0.10453 1.08487 0.96018 87.85s
155 0.11220 0.09665 1.16081 0.96577 87.75s
156 0.11291 0.09826 1.14919 0.96477 88.48s
157 0.11104 0.09501 1.16870 0.96554 87.66s
158 0.11170 0.09877 1.13096 0.96375 87.63s
159 0.11054 0.09419 1.17360 0.96577 87.63s
160 0.11267 0.09678 1.16410 0.96426 87.67s
161 0.11188 0.10307 1.08548 0.96192 87.65s
162 0.11354 0.09650 1.17665 0.96460 87.62s
163 0.11099 0.09911 1.11982 0.96389 87.70s
164 0.11342 0.10436 1.08677 0.95995 87.65s
165 0.11273 0.10274 1.09730 0.96153 87.84s
166 0.11107 0.09575 1.15991 0.96454 88.01s
167 0.11032 0.09792 1.12663 0.96440 87.67s
168 0.11281 0.10221 1.10364 0.96437 87.89s
169 0.11233 0.09512 1.18090 0.96490 88.03s
170 0.11176 0.09817 1.13849 0.96398 88.16s
171 0.11111 0.09643 1.15230 0.96744 87.83s
172 0.11069 0.09861 1.12248 0.96323 87.67s
173 0.11183 0.09406 1.18899 0.96713 87.74s
174 0.11125 0.10226 1.08788 0.96109 87.71s
175 0.11212 0.09774 1.14707 0.96481 87.60s
176 0.11116 0.09206 1.20751 0.96735 87.72s
177 0.11129 0.09769 1.13926 0.96405 87.66s
178 0.11172 0.10101 1.10600 0.96279 87.68s
179 0.11136 0.10132 1.09911 0.96247 87.66s
180 0.10992 0.09109 1.20678 0.96873 87.66s
181 0.11181 0.10102 1.10674 0.96291 87.67s
182 0.11094 0.09705 1.14318 0.96442 87.66s
183 0.11084 0.09992 1.10923 0.96238 87.64s
184 0.11037 0.09826 1.12319 0.96483 87.68s
185 0.11168 0.09477 1.17841 0.96628 87.60s
186 0.11068 0.09943 1.11317 0.96256 87.73s
187 0.10973 0.09934 1.10460 0.96295 87.73s
188 0.11091 0.09932 1.11668 0.96344 87.78s
189 0.11029 0.09700 1.13712 0.96410 87.62s
190 0.10987 0.09851 1.11527 0.96357 87.68s
191 0.11203 0.09808 1.14230 0.96259 87.67s
192 0.11056 0.09770 1.13166 0.96344 87.61s
193 0.10879 0.09858 1.10358 0.96435 88.25s
194 0.10996 0.10071 1.09186 0.96194 87.54s
195 0.10979 0.10038 1.09373 0.96204 87.53s
196 0.11133 0.09382 1.18663 0.96586 87.46s
197 0.11033 0.10253 1.07608 0.96105 87.40s
198 0.10918 0.09598 1.13749 0.96437 87.29s
199 0.11217 0.09665 1.16058 0.96359 87.30s
200 0.11061 0.09704 1.13982 0.96543 87.30s
201 0.11160 0.09970 1.11932 0.96231 87.30s
202 0.11154 0.09351 1.19284 0.96513 87.28s
203 0.10879 0.09594 1.13391 0.96520 87.26s
204 0.10833 0.09741 1.11203 0.96355 87.30s
205 0.10936 0.09755 1.12109 0.96265 87.28s
206 0.11006 0.10119 1.08770 0.96041 87.24s
207 0.11089 0.09190 1.20665 0.96767 87.27s
208 0.10975 0.10024 1.09491 0.96130 87.31s
209 0.11088 0.09741 1.13828 0.96272 87.25s
210 0.10963 0.09351 1.17238 0.96534 87.31s
211 0.10961 0.09893 1.10796 0.96392 87.27s
212 0.10896 0.09567 1.13887 0.96580 87.27s
213 0.11136 0.10210 1.09068 0.96075 87.27s
214 0.11019 0.09670 1.13946 0.96509 87.25s
215 0.11018 0.10514 1.04792 0.95850 87.37s
216 0.10962 0.10262 1.06817 0.96052 87.24s
217 0.10997 0.09629 1.14202 0.96538 87.37s
218 0.11016 0.09568 1.15144 0.96470 87.29s
219 0.10899 0.09122 1.19492 0.96887 87.28s
220 0.11012 0.09870 1.11569 0.96226 87.28s
221 0.11060 0.10767 1.02720 0.95788 87.27s
222 0.10932 0.09036 1.20980 0.96653 87.25s
223 0.11104 0.09686 1.14632 0.96339 89.34s
224 0.10962 0.09760 1.12313 0.96396 89.53s
225 0.10974 0.09269 1.18393 0.96788 87.17s
226 0.11045 0.09986 1.10603 0.96376 88.87s
227 0.10903 0.10089 1.08064 0.96101 91.11s
228 0.11160 0.09985 1.11764 0.96330 90.98s
229 0.10979 0.10216 1.07471 0.96027 91.10s
230 0.10959 0.09297 1.17883 0.96607 91.04s
231 0.10980 0.09335 1.17621 0.96626 91.15s
232 0.10899 0.09292 1.17298 0.96635 90.92s
233 0.10925 0.09832 1.11117 0.96380 91.13s
234 0.10945 0.09876 1.10822 0.96382 91.08s
235 0.11001 0.10460 1.05169 0.96321 91.31s
236 0.10871 0.09710 1.11950 0.96430 91.12s
237 0.11033 0.08782 1.25624 0.96836 91.04s
238 0.11034 0.10451 1.05580 0.95747 91.18s
239 0.10864 0.09363 1.16022 0.96543 91.15s
240 0.11002 0.09937 1.10714 0.96254 91.05s
241 0.11008 0.09282 1.18598 0.96559 91.01s
242 0.10974 0.09102 1.20569 0.96609 91.26s
243 0.10959 0.08955 1.22378 0.96722 91.01s
244 0.10915 0.09057 1.20510 0.96680 91.02s
245 0.10927 0.09838 1.11067 0.96160 91.04s
246 0.11064 0.09091 1.21705 0.96618 91.33s
247 0.10952 0.09396 1.16555 0.96548 91.24s
248 0.10990 0.09016 1.21892 0.96713 91.32s
249 0.10953 0.09599 1.14114 0.96396 90.82s
250 0.10793 0.09366 1.15232 0.96607 91.18s
251 0.10817 0.09580 1.12912 0.96485 90.89s
252 0.10798 0.08925 1.20978 0.96889 91.01s
253 0.10747 0.09204 1.16759 0.96720 90.96s
254 0.10872 0.09079 1.19752 0.96596 90.60s
255 0.10800 0.09467 1.14078 0.96579 91.04s
256 0.10945 0.09296 1.17734 0.96657 90.97s
257 0.10987 0.09352 1.17475 0.96422 91.04s
258 0.10943 0.10105 1.08293 0.96185 91.02s
259 0.10830 0.09405 1.15148 0.96444 90.95s
260 0.10867 0.09113 1.19251 0.96579 90.86s
261 0.10933 0.09158 1.19385 0.96628 91.11s
262 0.10870 0.08697 1.24984 0.96903 90.68s
263 0.10863 0.09234 1.17643 0.96598 90.86s
264 0.10833 0.09010 1.20231 0.96612 90.88s
265 0.10977 0.09423 1.16486 0.96302 91.33s
266 0.10903 0.09155 1.19095 0.96612 91.13s
267 0.10865 0.10031 1.08317 0.96059 91.14s
268 0.10922 0.09361 1.16679 0.96502 90.98s
269 0.10919 0.09764 1.11834 0.96188 90.99s
270 0.10822 0.08652 1.25083 0.96857 91.08s
271 0.10663 0.09375 1.13738 0.96511 90.76s
272 0.10768 0.09264 1.16239 0.96566 90.46s
273 0.10856 0.09083 1.19528 0.96637 90.66s
274 0.10833 0.09551 1.13432 0.96408 90.78s
275 0.10991 0.09381 1.17164 0.96472 91.04s
276 0.10919 0.09545 1.14396 0.96229 90.96s
277 0.10876 0.08902 1.22178 0.96799 91.18s
278 0.10942 0.09653 1.13354 0.96362 90.98s
279 0.10819 0.09556 1.13223 0.96385 90.88s
280 0.10834 0.09020 1.20111 0.96602 90.98s
281 0.10831 0.09489 1.14147 0.96323 91.12s
282 0.10758 0.08662 1.24196 0.96811 91.29s
283 0.10981 0.08798 1.24806 0.96924 90.83s
284 0.10970 0.09177 1.19535 0.96541 90.84s
285 0.10781 0.09365 1.15121 0.96327 91.16s
286 0.10824 0.09625 1.12453 0.96259 90.28s
287 0.10926 0.10000 1.09255 0.96075 90.41s
288 0.10988 0.08572 1.28181 0.96960 90.56s
289 0.10998 0.09504 1.15725 0.96472 90.86s
290 0.10821 0.09525 1.13605 0.96433 90.85s
291 0.10839 0.08858 1.22359 0.96660 90.98s
292 0.10852 0.09876 1.09880 0.96011 90.87s
293 0.11037 0.09029 1.22245 0.96529 91.06s
294 0.10643 0.09127 1.16620 0.96566 91.01s
295 0.10767 0.09258 1.16302 0.96561 91.00s
296 0.10908 0.09562 1.14076 0.96368 90.82s
297 0.10814 0.09652 1.12040 0.96297 90.86s
298 0.10895 0.08925 1.22074 0.96655 90.80s
299 0.10647 0.09604 1.10862 0.96371 90.72s
300 0.10940 0.08920 1.22638 0.96717 90.86s
301 0.10835 0.09169 1.18165 0.96634 90.62s
302 0.10674 0.09938 1.07400 0.96234 90.75s
303 0.10892 0.08941 1.21817 0.96557 90.93s
304 0.10888 0.09026 1.20628 0.96648 90.75s
305 0.10746 0.09705 1.10730 0.96321 90.95s
306 0.10860 0.09380 1.15786 0.96462 90.84s
307 0.10750 0.09228 1.16485 0.96430 90.80s
308 0.10952 0.09123 1.20047 0.96582 90.81s
309 0.10712 0.09750 1.09858 0.96133 90.87s
310 0.10819 0.09132 1.18470 0.96566 90.66s
311 0.10846 0.09156 1.18452 0.96589 90.98s
312 0.10945 0.09141 1.19729 0.96655 90.72s
313 0.10669 0.09996 1.06735 0.96011 90.63s
314 0.10925 0.09559 1.14286 0.96376 90.54s
315 0.10755 0.08935 1.20364 0.96660 90.76s
316 0.10751 0.09516 1.12975 0.96325 90.77s
317 0.10839 0.09289 1.16691 0.96552 90.81s
318 0.10755 0.09451 1.13803 0.96403 90.65s
319 0.10715 0.09014 1.18862 0.96653 90.70s
320 0.10783 0.09141 1.17957 0.96545 90.81s
321 0.10717 0.09161 1.16992 0.96609 90.89s
322 0.11064 0.10749 1.02929 0.95756 90.81s
323 0.10752 0.09273 1.15951 0.96596 91.06s
324 0.10774 0.08995 1.19783 0.96511 90.93s
325 0.10726 0.09406 1.14033 0.96525 91.04s
326 0.10729 0.09083 1.18119 0.96465 90.95s
327 0.10836 0.09002 1.20372 0.96752 90.73s
328 0.10838 0.08840 1.22610 0.96795 90.96s
329 0.10657 0.09304 1.14550 0.96421 90.74s
330 0.10843 0.08852 1.22495 0.96861 91.00s
331 0.10799 0.08800 1.22705 0.96648 90.91s
332 0.10988 0.09074 1.21093 0.96667 90.74s
333 0.10867 0.09046 1.20131 0.96639 90.90s
334 0.10636 0.09243 1.15073 0.96511 90.53s
335 0.10711 0.09199 1.16434 0.96621 90.68s
336 0.10817 0.09087 1.19037 0.96678 90.91s
337 0.10842 0.09990 1.08526 0.96169 90.68s
338 0.10901 0.09471 1.15096 0.96495 90.93s
Early stopping.
Best valid loss was 0.085722 at epoch 288.
Loaded parameters to layer 'conv1' (shape 64x3x3x3).
Loaded parameters to layer 'conv1' (shape 64).
Loaded parameters to layer 'conv2' (shape 48x64x3x3).
Loaded parameters to layer 'conv2' (shape 48).
Loaded parameters to layer 'conv3' (shape 48x48x3x3).
Loaded parameters to layer 'conv3' (shape 48).
Loaded parameters to layer 'conv4' (shape 48x48x3x3).
Loaded parameters to layer 'conv4' (shape 48).
Loaded parameters to layer 'hidden5' (shape 192x512).
Loaded parameters to layer 'hidden5' (shape 512).
Loaded parameters to layer 'output' (shape 512x2).
Loaded parameters to layer 'output' (shape 2).
In [7]:
test_accuracy = cnn.score(X_test, y_test)
In [8]:
test_accuracy
Out[8]:
0.9195725534308211
In [8]:
plot_loss(cnn)
Out[8]:
<module 'matplotlib.pyplot' from '/n/home05/haehn/nolearncox/lib/python2.7/site-packages/matplotlib-1.5.2-py2.7-linux-x86_64.egg/matplotlib/pyplot.pyc'>
In [9]:
# store CNN
sys.setrecursionlimit(1000000000)
with open(os.path.expanduser('~/Projects/gp/nets/RGBPlus_FULL.p'), 'wb') as f:
pickle.dump(cnn, f, -1)
In [ ]:
Content source: VCG/gp
Similar notebooks: