--------[20_09_2019 13:25:28]-------- Random Grid Search Search 1 of 500 momentum0.96, features=[96, 192, 192], dropout_rate=0.6 kernel=Tuple{Int64,Int64}[(7, 1), (5, 1), (2, 6)], pooldims=Tuple{Int64,Int64}[(2, 1), (2, 1)], learning_rate=0.03 [13:26:39] INIT Loss(val): 0.141007 Accuarcy: 0.117381 [13:29:14] Epoch 2: Loss(train): 0.088968 Loss(val): 0.087040 [13:30:22] Epoch 4: Loss(train): 0.078608 Loss(val): 0.076951 [13:31:29] Epoch 6: Loss(train): 0.076482 Loss(val): 0.074798 [13:32:37] Epoch 8: Loss(train): 0.074540 Loss(val): 0.072936 [13:33:44] Epoch 10: Loss(train): 0.073011 Loss(val): 0.071681 [13:34:52] Epoch 12: Loss(train): 0.071422 Loss(val): 0.070381 [13:36:01] Epoch 14: Loss(train): 0.070299 Loss(val): 0.069362 [13:37:10] Epoch 16: Loss(train): 0.069983 Loss(val): 0.069102 [13:38:18] Epoch 18: Loss(train): 0.069733 Loss(val): 0.068830 [13:39:27] Epoch 20: Loss(train): 0.068419 Loss(val): 0.067956 [13:40:35] Epoch 22: Loss(train): 0.068390 Loss(val): 0.067870 [13:41:44] Epoch 24: Loss(train): 0.068358 Loss(val): 0.067915 [13:42:53] Epoch 26: Loss(train): 0.068200 Loss(val): 0.067803 [13:44:01] Epoch 28: Loss(train): 0.067958 Loss(val): 0.067655 [13:45:10] Epoch 30: Loss(train): 0.067752 Loss(val): 0.067437 [13:46:18] Epoch 32: Loss(train): 0.067372 Loss(val): 0.067225 [13:47:28] Epoch 34: Loss(train): 0.067107 Loss(val): 0.067024 [13:48:38] Epoch 36: Loss(train): 0.066756 Loss(val): 0.066762 [13:49:50] Epoch 38: Loss(train): 0.066352 Loss(val): 0.066501 [13:50:59] Epoch 40: Loss(train): 0.066226 Loss(val): 0.066400 [13:51:06] FINAL(40) Loss(val): 0.066400 Accuarcy: 0.621310 Search 2 of 500 momentum0.99, features=[32, 32, 32], dropout_rate=0.4 kernel=Tuple{Int64,Int64}[(5, 1), (3, 1), (3, 6)], pooldims=Tuple{Int64,Int64}[(3, 1), (3, 1)], learning_rate=1.0 [13:51:25] INIT Loss(val): 0.116734 Accuarcy: 0.086190 [13:52:24] Epoch 2: Loss(train): 0.111209 Loss(val): 0.112811 [13:52:52] Epoch 4: Loss(train): 0.070398 Loss(val): 0.072423 [13:53:22] Epoch 6: Loss(train): 0.068424 Loss(val): 0.070707 [13:53:52] Epoch 8: Loss(train): 0.067158 Loss(val): 0.069694 [13:54:21] Epoch 10: Loss(train): 0.066686 Loss(val): 0.069351 [13:54:50] Epoch 12: Loss(train): 0.066295 Loss(val): 0.069087 [13:55:20] Epoch 14: Loss(train): 0.066166 Loss(val): 0.069011 [13:55:49] Epoch 16: Loss(train): 0.066101 Loss(val): 0.068977 [13:56:18] Epoch 18: Loss(train): 0.066013 Loss(val): 0.068943 [13:56:46] Epoch 20: Loss(train): 0.065961 Loss(val): 0.068933 [13:57:14] Epoch 22: Loss(train): 0.065942 Loss(val): 0.068932 Early stopping with Loss(train) 0.065942 at epoch 22 (Accuracy: 0.091786) Search 3 of 500 momentum0.98, features=[32, 64, 128], dropout_rate=0.3 kernel=Tuple{Int64,Int64}[(5, 1), (5, 1), (2, 6)], pooldims=Tuple{Int64,Int64}[(3, 1), (3, 1)], learning_rate=0.3 [13:57:28] INIT Loss(val): 0.158934 Accuarcy: 0.094031 [13:59:04] Epoch 2: Loss(train): 0.079493 Loss(val): 0.079803 [14:00:10] Epoch 4: Loss(train): 0.071139 Loss(val): 0.070190 [14:01:15] Epoch 6: Loss(train): 0.057562 Loss(val): 0.059708 [14:02:20] Epoch 8: Loss(train): 0.042016 Loss(val): 0.042341 [14:03:25] Epoch 10: Loss(train): 0.035010 Loss(val): 0.034012 [14:04:32] Epoch 12: Loss(train): 0.031262 Loss(val): 0.030622 [14:05:40] Epoch 14: Loss(train): 0.031100 Loss(val): 0.031548 [14:06:43] Epoch 16: Loss(train): 0.027940 Loss(val): 0.026886 [14:07:49] Epoch 18: Loss(train): 0.026650 Loss(val): 0.026341 [14:08:55] Epoch 20: Loss(train): 0.025911 Loss(val): 0.024959 [14:09:59] Epoch 22: Loss(train): 0.023871 Loss(val): 0.022819 [14:11:04] Epoch 24: Loss(train): 0.022636 Loss(val): 0.021778 [14:12:08] Epoch 26: Loss(train): 0.020553 Loss(val): 0.020178 [14:13:13] Epoch 28: Loss(train): 0.020170 Loss(val): 0.019662 [14:14:21] Epoch 30: Loss(train): 0.020618 Loss(val): 0.019873 [14:15:29] Epoch 32: Loss(train): 0.021250 Loss(val): 0.020526 [14:16:36] Epoch 34: Loss(train): 0.020209 Loss(val): 0.019461 [14:17:50] Epoch 36: Loss(train): 0.019272 Loss(val): 0.018871 [14:19:04] Epoch 38: Loss(train): 0.019579 Loss(val): 0.019146 [14:20:18] Epoch 40: Loss(train): 0.020248 Loss(val): 0.019587 [14:20:28] FINAL(40) Loss(val): 0.019587 Accuarcy: 0.607517 Search 4 of 500 momentum0.98, features=[32, 32, 32], dropout_rate=0.3 kernel=Tuple{Int64,Int64}[(5, 1), (5, 1), (2, 6)], pooldims=Tuple{Int64,Int64}[(3, 1), (3, 1)], learning_rate=0.003 [14:20:51] INIT Loss(val): 0.128640 Accuarcy: 0.094643 [14:21:39] Epoch 2: Loss(train): 0.056761 Loss(val): 0.056647 [14:22:08] Epoch 4: Loss(train): 0.050836 Loss(val): 0.050852 [14:22:37] Epoch 6: Loss(train): 0.048547 Loss(val): 0.048590 [14:23:06] Epoch 8: Loss(train): 0.047363 Loss(val): 0.047409 [14:23:35] Epoch 10: Loss(train): 0.046506 Loss(val): 0.046528 [14:24:05] Epoch 12: Loss(train): 0.046048 Loss(val): 0.046113 [14:24:34] Epoch 14: Loss(train): 0.045531 Loss(val): 0.045707 [14:25:05] Epoch 16: Loss(train): 0.045208 Loss(val): 0.045390 [14:25:35] Epoch 18: Loss(train): 0.044927 Loss(val): 0.045131 [14:26:06] Epoch 20: Loss(train): 0.044599 Loss(val): 0.044836 [14:26:36] Epoch 22: Loss(train): 0.044361 Loss(val): 0.044653 [14:27:05] Epoch 24: Loss(train): 0.044173 Loss(val): 0.044473 [14:27:36] Epoch 26: Loss(train): 0.043907 Loss(val): 0.044196 [14:28:06] Epoch 28: Loss(train): 0.043688 Loss(val): 0.044030 [14:28:36] Epoch 30: Loss(train): 0.043458 Loss(val): 0.043854 [14:29:07] Epoch 32: Loss(train): 0.043282 Loss(val): 0.043654 [14:29:37] Epoch 34: Loss(train): 0.043138 Loss(val): 0.043506 [14:30:07] Epoch 36: Loss(train): 0.042960 Loss(val): 0.043359 [14:30:38] Epoch 38: Loss(train): 0.042856 Loss(val): 0.043255 Early stopping with Loss(train) 0.044059 at epoch 38 (Accuracy: 0.575901) Search 5 of 500 momentum0.94, features=[64, 64, 64], dropout_rate=0.1 kernel=Tuple{Int64,Int64}[(5, 1), (3, 1), (2, 6)], pooldims=Tuple{Int64,Int64}[(3, 1), (3, 1)], learning_rate=0.003 [14:30:53] INIT Loss(val): 0.119816 Accuarcy: 0.091463 [14:34:37] Epoch 2: Loss(train): 0.068172 Loss(val): 0.066521 [14:38:47] Epoch 4: Loss(train): 0.061371 Loss(val): 0.060421 [14:44:35] Epoch 6: Loss(train): 0.058998 Loss(val): 0.058359 [14:50:21] Epoch 8: Loss(train): 0.057713 Loss(val): 0.057312 [14:54:55] Epoch 10: Loss(train): 0.056964 Loss(val): 0.056694 [14:59:13] Epoch 12: Loss(train): 0.056202 Loss(val): 0.056133 [15:03:30] Epoch 14: Loss(train): 0.055774 Loss(val): 0.055825 [15:07:46] Epoch 16: Loss(train): 0.055417 Loss(val): 0.055541 [15:09:41] Epoch 18: Loss(train): 0.055157 Loss(val): 0.055343 [15:11:36] Epoch 20: Loss(train): 0.054948 Loss(val): 0.055184 [15:14:40] Epoch 22: Loss(train): 0.054736 Loss(val): 0.055040 [15:17:10] Epoch 24: Loss(train): 0.054518 Loss(val): 0.054851 [15:19:29] Epoch 26: Loss(train): 0.054388 Loss(val): 0.054757 [15:21:38] Epoch 28: Loss(train): 0.054141 Loss(val): 0.054547 [15:23:56] Epoch 30: Loss(train): 0.054011 Loss(val): 0.054442 [15:25:57] Epoch 32: Loss(train): 0.053802 Loss(val): 0.054259 [15:27:59] Epoch 34: Loss(train): 0.053625 Loss(val): 0.054112 [15:30:03] Epoch 36: Loss(train): 0.053405 Loss(val): 0.053942 [15:32:05] Epoch 38: Loss(train): 0.053237 Loss(val): 0.053814 Early stopping with Loss(train) 0.054171 at epoch 38 (Accuracy: 0.534439) Search 6 of 500 momentum0.94, features=[96, 192, 192], dropout_rate=0.3 kernel=Tuple{Int64,Int64}[(3, 1), (3, 1), (2, 6)], pooldims=Tuple{Int64,Int64}[(2, 1), (2, 1)], learning_rate=0.1 [15:32:54] INIT Loss(val): 0.138280 Accuarcy: 0.103418 [15:34:06] Epoch 2: Loss(train): 0.083781 Loss(val): 0.084471 [15:35:16] Epoch 4: Loss(train): 0.076121 Loss(val): 0.077370 [15:36:23] Epoch 6: Loss(train): 0.074109 Loss(val): 0.073338 [15:37:30] Epoch 8: Loss(train): 0.070528 Loss(val): 0.070370 [15:38:36] Epoch 10: Loss(train): 0.067579 Loss(val): 0.067635 [15:39:43] Epoch 12: Loss(train): 0.065495 Loss(val): 0.065627 [15:40:51] Epoch 14: Loss(train): 0.064575 Loss(val): 0.064573 [15:42:01] Epoch 16: Loss(train): 0.062974 Loss(val): 0.063009 [15:43:23] Epoch 18: Loss(train): 0.061456 Loss(val): 0.061671 [15:44:49] Epoch 20: Loss(train): 0.060464 Loss(val): 0.060732 [15:46:07] Epoch 22: Loss(train): 0.059721 Loss(val): 0.060011 [15:47:29] Epoch 24: Loss(train): 0.058930 Loss(val): 0.059211 [15:48:42] Epoch 26: Loss(train): 0.058132 Loss(val): 0.058528 [15:50:06] Epoch 28: Loss(train): 0.057695 Loss(val): 0.058175 [15:51:18] Epoch 30: Loss(train): 0.056704 Loss(val): 0.057436 [15:52:28] Epoch 32: Loss(train): 0.056475 Loss(val): 0.057144 [15:53:48] Epoch 34: Loss(train): 0.055826 Loss(val): 0.056582 Early stopping with Loss(train) 0.056777 at epoch 35 (Accuracy: 0.632602) Search 7 of 500 momentum0.9, features=[32, 64, 128], dropout_rate=0.4 kernel=Tuple{Int64,Int64}[(5, 1), (3, 1), (3, 6)], pooldims=Tuple{Int64,Int64}[(2, 1), (2, 1)], learning_rate=0.003 [15:54:46] INIT Loss(val): 0.146946 Accuarcy: 0.092534 [15:59:34] Epoch 2: Loss(train): 0.080755 Loss(val): 0.080996 [16:04:53] Epoch 4: Loss(train): 0.072572 Loss(val): 0.072945 [16:10:04] Epoch 6: Loss(train): 0.069757 Loss(val): 0.070359 [16:16:05] Epoch 8: Loss(train): 0.068311 Loss(val): 0.069053 [16:18:27] Epoch 10: Loss(train): 0.067540 Loss(val): 0.068310 [16:20:27] Epoch 12: Loss(train): 0.066874 Loss(val): 0.067751 [16:21:51] Epoch 14: Loss(train): 0.066258 Loss(val): 0.067226 [16:23:24] Epoch 16: Loss(train): 0.065859 Loss(val): 0.066875 [16:24:42] Epoch 18: Loss(train): 0.065511 Loss(val): 0.066519 [16:26:03] Epoch 20: Loss(train): 0.065213 Loss(val): 0.066262 [16:27:34] Epoch 22: Loss(train): 0.064957 Loss(val): 0.066055 [16:28:56] Epoch 24: Loss(train): 0.064687 Loss(val): 0.065852 [16:30:15] Epoch 26: Loss(train): 0.064495 Loss(val): 0.065710 [16:31:42] Epoch 28: Loss(train): 0.064351 Loss(val): 0.065567 [16:33:03] Epoch 30: Loss(train): 0.064181 Loss(val): 0.065445 Early stopping with Loss(train) 0.067298 at epoch 31 (Accuracy: 0.446395) Search 8 of 500 momentum0.96, features=[32, 32, 32], dropout_rate=0.6 kernel=Tuple{Int64,Int64}[(5, 1), (5, 1), (2, 6)], pooldims=Tuple{Int64,Int64}[(3, 1), (3, 1)], learning_rate=0.003 [16:34:07] INIT Loss(val): 0.114290 Accuarcy: 0.088316 [16:34:46] Epoch 2: Loss(train): 0.064952 Loss(val): 0.064766 [16:35:23] Epoch 4: Loss(train): 0.055647 Loss(val): 0.055572 [16:36:02] Epoch 6: Loss(train): 0.052844 Loss(val): 0.052606 [16:36:38] Epoch 8: Loss(train): 0.050985 Loss(val): 0.050845 [16:37:12] Epoch 10: Loss(train): 0.049845 Loss(val): 0.049633 [16:37:47] Epoch 12: Loss(train): 0.049152 Loss(val): 0.048910 [16:38:22] Epoch 14: Loss(train): 0.048669 Loss(val): 0.048351 [16:38:57] Epoch 16: Loss(train): 0.048184 Loss(val): 0.047886 [16:39:31] Epoch 18: Loss(train): 0.047724 Loss(val): 0.047523 [16:40:06] Epoch 20: Loss(train): 0.047366 Loss(val): 0.047177 [16:40:42] Epoch 22: Loss(train): 0.047168 Loss(val): 0.046973 [16:41:17] Epoch 24: Loss(train): 0.046858 Loss(val): 0.046676 [16:41:52] Epoch 26: Loss(train): 0.046654 Loss(val): 0.046519 [16:42:27] Epoch 28: Loss(train): 0.046444 Loss(val): 0.046329 [16:43:03] Epoch 30: Loss(train): 0.046292 Loss(val): 0.046197 [16:43:39] Epoch 32: Loss(train): 0.046148 Loss(val): 0.046083 [16:44:14] Epoch 34: Loss(train): 0.045966 Loss(val): 0.045960 [16:44:50] Epoch 36: Loss(train): 0.045872 Loss(val): 0.045834 [16:45:40] Epoch 38: Loss(train): 0.045775 Loss(val): 0.045743 Early stopping with Loss(train) 0.048965 at epoch 38 (Accuracy: 0.502738) Search 9 of 500 momentum0.9, features=[32, 64, 128], dropout_rate=0.6 kernel=Tuple{Int64,Int64}[(7, 1), (5, 1), (2, 6)], pooldims=Tuple{Int64,Int64}[(3, 1), (3, 1)], learning_rate=0.3 [16:46:09] INIT Loss(val): 0.128704 Accuarcy: 0.095255 [16:47:37] Epoch 2: Loss(train): 0.068261 Loss(val): 0.066425 [16:48:59] Epoch 4: Loss(train): 0.059768 Loss(val): 0.058380 [16:49:56] Epoch 6: Loss(train): 0.056013 Loss(val): 0.054921 [16:50:56] Epoch 8: Loss(train): 0.052184 Loss(val): 0.051936 [16:52:12] Epoch 10: Loss(train): 0.050930 Loss(val): 0.050823 [16:53:14] Epoch 12: Loss(train): 0.050832 Loss(val): 0.051081 [16:54:05] Epoch 14: Loss(train): 0.047966 Loss(val): 0.048353 [16:54:58] Epoch 16: Loss(train): 0.046510 Loss(val): 0.046471 [16:55:53] Epoch 18: Loss(train): 0.045391 Loss(val): 0.045162 Early stopping with Loss(train) 0.047088 at epoch 19 (Accuracy: 0.514779) Search 10 of 500 momentum0.99, features=[32, 32, 32], dropout_rate=0.6 kernel=Tuple{Int64,Int64}[(7, 1), (7, 1), (2, 6)], pooldims=Tuple{Int64,Int64}[(3, 1), (3, 1)], learning_rate=0.1 [16:56:41] INIT Loss(val): 0.155156 Accuarcy: 0.107092 [16:57:23] Epoch 2: Loss(train): 0.082692 Loss(val): 0.082872 [16:58:03] Epoch 4: Loss(train): 0.070038 Loss(val): 0.068951 [16:58:43] Epoch 6: Loss(train): 0.052024 Loss(val): 0.051050 [16:59:23] Epoch 8: Loss(train): 0.042558 Loss(val): 0.042033 [17:00:01] Epoch 10: Loss(train): 0.033804 Loss(val): 0.032005 [17:00:41] Epoch 12: Loss(train): 0.028679 Loss(val): 0.028131 [17:01:25] Epoch 14: Loss(train): 0.026270 Loss(val): 0.025863 [17:02:07] Epoch 16: Loss(train): 0.025516 Loss(val): 0.024859 [17:02:46] Epoch 18: Loss(train): 0.024369 Loss(val): 0.024071 [17:03:26] Epoch 20: Loss(train): 0.022684 Loss(val): 0.022117 [17:04:06] Epoch 22: Loss(train): 0.022049 Loss(val): 0.020969 [17:04:45] Epoch 24: Loss(train): 0.021334 Loss(val): 0.020421 [17:05:26] Epoch 26: Loss(train): 0.021068 Loss(val): 0.020358 [17:06:06] Epoch 28: Loss(train): 0.021064 Loss(val): 0.020323 [17:06:45] Epoch 30: Loss(train): 0.021924 Loss(val): 0.020800 [17:07:25] Epoch 32: Loss(train): 0.022128 Loss(val): 0.021133 [17:08:05] Epoch 34: Loss(train): 0.021443 Loss(val): 0.020385 [17:08:44] Epoch 36: Loss(train): 0.023132 Loss(val): 0.021738 [17:09:23] Epoch 38: Loss(train): 0.024656 Loss(val): 0.023247 [17:10:02] Epoch 40: Loss(train): 0.022748 Loss(val): 0.021955 [17:10:07] FINAL(40) Loss(val): 0.021955 Accuarcy: 0.579932 Search 11 of 500 momentum0.99, features=[32, 64, 128], dropout_rate=0.3 kernel=Tuple{Int64,Int64}[(5, 1), (5, 1), (2, 6)], pooldims=Tuple{Int64,Int64}[(3, 1), (3, 1)], learning_rate=0.003 [17:10:22] INIT Loss(val): 0.131512 Accuarcy: 0.094524 [17:11:17] Epoch 2: Loss(train): 0.067090 Loss(val): 0.065792 [17:12:12] Epoch 4: Loss(train): 0.061073 Loss(val): 0.059899 [17:13:05] Epoch 6: Loss(train): 0.059854 Loss(val): 0.059079 [17:14:02] Epoch 8: Loss(train): 0.059035 Loss(val): 0.058605 [17:14:57] Epoch 10: Loss(train): 0.058039 Loss(val): 0.057734 [17:15:54] Epoch 12: Loss(train): 0.056114 Loss(val): 0.055818 [17:17:03] Epoch 14: Loss(train): 0.055465 Loss(val): 0.055276 [17:18:32] Epoch 16: Loss(train): 0.055368 Loss(val): 0.055433 [17:19:59] Epoch 18: Loss(train): 0.055454 Loss(val): 0.055520 [17:21:32] Epoch 20: Loss(train): 0.055490 Loss(val): 0.055287 [17:22:53] Epoch 22: Loss(train): 0.054614 Loss(val): 0.054396 [17:23:49] Epoch 24: Loss(train): 0.053453 Loss(val): 0.053415 [17:25:15] Epoch 26: Loss(train): 0.053183 Loss(val): 0.053106 [17:26:20] Epoch 28: Loss(train): 0.053469 Loss(val): 0.053248 [17:27:23] Epoch 30: Loss(train): 0.054454 Loss(val): 0.053836 [17:28:23] Epoch 32: Loss(train): 0.054597 Loss(val): 0.053841 [17:29:24] Epoch 34: Loss(train): 0.053575 Loss(val): 0.053139 [17:30:23] Epoch 36: Loss(train): 0.052531 Loss(val): 0.052516 [17:31:27] Epoch 38: Loss(train): 0.052137 Loss(val): 0.052324 [17:32:28] Epoch 40: Loss(train): 0.051984 Loss(val): 0.052194 [17:32:35] FINAL(40) Loss(val): 0.052194 Accuarcy: 0.657959 Search 12 of 500 momentum0.98, features=[64, 64, 64], dropout_rate=0.1 kernel=Tuple{Int64,Int64}[(3, 1), (3, 1), (3, 6)], pooldims=Tuple{Int64,Int64}[(3, 1), (3, 1)], learning_rate=0.1 [17:32:56] INIT Loss(val): 0.152510 Accuarcy: 0.094932 [17:34:22] Epoch 2: Loss(train): 0.076679 Loss(val): 0.078237 [17:35:44] Epoch 4: Loss(train): 0.059022 Loss(val): 0.060126 [17:40:56] Epoch 6: Loss(train): 0.050557 Loss(val): 0.051844 [17:49:29] Epoch 8: Loss(train): 0.045205 Loss(val): 0.045557 [18:00:08] Epoch 10: Loss(train): 0.040994 Loss(val): 0.041337 [18:08:11] Epoch 12: Loss(train): 0.038653 Loss(val): 0.038964 [18:15:40] Epoch 14: Loss(train): 0.036438 Loss(val): 0.036957 [18:25:46] Epoch 16: Loss(train): 0.035700 Loss(val): 0.035639 [18:35:47] Epoch 18: Loss(train): 0.034637 Loss(val): 0.034805 [18:43:48] Epoch 20: Loss(train): 0.031340 Loss(val): 0.031079 [18:51:14] Epoch 22: Loss(train): 0.029792 Loss(val): 0.029546 [19:02:50] Epoch 24: Loss(train): 0.028174 Loss(val): 0.027946 [19:11:45] Epoch 26: Loss(train): 0.027586 Loss(val): 0.027259 [19:19:33] Epoch 28: Loss(train): 0.027456 Loss(val): 0.027579 [19:28:45] Epoch 30: Loss(train): 0.027956 Loss(val): 0.027891 [19:40:09] Epoch 32: Loss(train): 0.028770 Loss(val): 0.028286 Early stopping with Loss(train) 0.028975 at epoch 33 (Accuracy: 0.584439) Search 13 of 500 momentum0.99, features=[64, 64, 64], dropout_rate=0.1 kernel=Tuple{Int64,Int64}[(5, 1), (3, 1), (3, 6)], pooldims=Tuple{Int64,Int64}[(2, 1), (2, 1)], learning_rate=0.001 [19:46:57] INIT Loss(val): 0.145452 Accuarcy: 0.113588 [19:50:44] Epoch 2: Loss(train): 0.089347 Loss(val): 0.088077 [19:54:42] Epoch 4: Loss(train): 0.063581 Loss(val): 0.064125 [19:59:21] Epoch 6: Loss(train): 0.061616 Loss(val): 0.062470 [20:06:13] Epoch 8: Loss(train): 0.061005 Loss(val): 0.061986 [20:11:36] Epoch 10: Loss(train): 0.060453 Loss(val): 0.061392 [20:16:11] Epoch 12: Loss(train): 0.059992 Loss(val): 0.060942 [20:20:27] Epoch 14: Loss(train): 0.059627 Loss(val): 0.060763 [20:24:29] Epoch 16: Loss(train): 0.059573 Loss(val): 0.060785 [20:28:31] Epoch 18: Loss(train): 0.059532 Loss(val): 0.060666 [20:34:27] Epoch 20: Loss(train): 0.058798 Loss(val): 0.059861 [20:41:03] Epoch 22: Loss(train): 0.058062 Loss(val): 0.059021 [20:46:14] Epoch 24: Loss(train): 0.057693 Loss(val): 0.058506 [20:50:45] Epoch 26: Loss(train): 0.057584 Loss(val): 0.058362 [20:54:59] Epoch 28: Loss(train): 0.057410 Loss(val): 0.058215 [20:59:00] Epoch 30: Loss(train): 0.057174 Loss(val): 0.057987 [21:03:40] Epoch 32: Loss(train): 0.057070 Loss(val): 0.057808 [21:10:36] Epoch 34: Loss(train): 0.057235 Loss(val): 0.057774 [21:16:08] Epoch 36: Loss(train): 0.057474 Loss(val): 0.057816 [21:21:06] Epoch 38: Loss(train): 0.057530 Loss(val): 0.057807 [21:25:37] Epoch 40: Loss(train): 0.057105 Loss(val): 0.057592 [21:26:04] FINAL(40) Loss(val): 0.057592 Accuarcy: 0.591735 Search 14 of 500 momentum0.9, features=[32, 32, 32], dropout_rate=0.6 kernel=Tuple{Int64,Int64}[(7, 1), (5, 1), (3, 6)], pooldims=Tuple{Int64,Int64}[(3, 1), (3, 1)], learning_rate=0.01 [21:26:59] INIT Loss(val): 0.122847 Accuarcy: 0.101803 [21:29:56] Epoch 2: Loss(train): 0.064045 Loss(val): 0.063516 [21:33:02] Epoch 4: Loss(train): 0.056780 Loss(val): 0.055908 [21:36:53] Epoch 6: Loss(train): 0.054376 Loss(val): 0.053519 [21:42:25] Epoch 8: Loss(train): 0.052633 Loss(val): 0.051863 [21:47:10] Epoch 10: Loss(train): 0.051527 Loss(val): 0.050836 [21:51:13] Epoch 12: Loss(train): 0.050602 Loss(val): 0.049957 [21:54:56] Epoch 14: Loss(train): 0.050027 Loss(val): 0.049464 [21:58:20] Epoch 16: Loss(train): 0.049429 Loss(val): 0.048945 [22:01:33] Epoch 18: Loss(train): 0.048902 Loss(val): 0.048480 [22:04:43] Epoch 20: Loss(train): 0.048485 Loss(val): 0.048132 [22:07:56] Epoch 22: Loss(train): 0.048165 Loss(val): 0.047822 [22:11:08] Epoch 24: Loss(train): 0.047914 Loss(val): 0.047628 [22:14:30] Epoch 26: Loss(train): 0.047709 Loss(val): 0.047426 [22:17:49] Epoch 28: Loss(train): 0.047584 Loss(val): 0.047305 [22:21:12] Epoch 30: Loss(train): 0.047354 Loss(val): 0.047143 Early stopping with Loss(train) 0.050260 at epoch 31 (Accuracy: 0.503571) Search 15 of 500 momentum0.98, features=[32, 64, 128], dropout_rate=0.1 kernel=Tuple{Int64,Int64}[(5, 1), (5, 1), (3, 6)], pooldims=Tuple{Int64,Int64}[(3, 1), (3, 1)], learning_rate=0.001 [22:23:39] INIT Loss(val): 0.218097 Accuarcy: 0.093946 [22:31:36] Epoch 2: Loss(train): 0.072229 Loss(val): 0.072008 [22:42:38] Epoch 4: Loss(train): 0.066523 Loss(val): 0.066956 [22:51:18] Epoch 6: Loss(train): 0.064689 Loss(val): 0.065551 [22:59:03] Epoch 8: Loss(train): 0.063820 Loss(val): 0.064851 [23:08:26] Epoch 10: Loss(train): 0.063226 Loss(val): 0.064352 [23:17:54] Epoch 12: Loss(train): 0.062768 Loss(val): 0.063999 [23:26:17] Epoch 14: Loss(train): 0.062558 Loss(val): 0.063801 [23:34:03] Epoch 16: Loss(train): 0.062267 Loss(val): 0.063589 [23:44:32] Epoch 18: Loss(train): 0.061969 Loss(val): 0.063301 [23:53:31] Epoch 20: Loss(train): 0.061649 Loss(val): 0.062996 [00:01:31] Epoch 22: Loss(train): 0.061327 Loss(val): 0.062696 [00:11:05] Epoch 24: Loss(train): 0.060975 Loss(val): 0.062371 [00:20:49] Epoch 26: Loss(train): 0.060806 Loss(val): 0.062130 [00:29:09] Epoch 28: Loss(train): 0.060544 Loss(val): 0.061854 [00:37:04] Epoch 30: Loss(train): 0.060328 Loss(val): 0.061611 [00:48:06] Epoch 32: Loss(train): 0.060126 Loss(val): 0.061430 [00:57:20] Epoch 34: Loss(train): 0.059966 Loss(val): 0.061301 [01:05:24] Epoch 36: Loss(train): 0.059804 Loss(val): 0.061190 [01:16:04] Epoch 38: Loss(train): 0.059661 Loss(val): 0.061111 [01:25:57] Epoch 40: Loss(train): 0.059557 Loss(val): 0.061046 [01:27:07] FINAL(40) Loss(val): 0.061046 Accuarcy: 0.573963 Search 16 of 500 momentum0.94, features=[64, 64, 64], dropout_rate=0.8 kernel=Tuple{Int64,Int64}[(5, 1), (3, 1), (3, 6)], pooldims=Tuple{Int64,Int64}[(2, 1), (2, 1)], learning_rate=0.001 [01:29:00] INIT Loss(val): 0.151834 Accuarcy: 0.108827 [01:33:32] Epoch 2: Loss(train): 0.110386 Loss(val): 0.113633 [01:38:13] Epoch 4: Loss(train): 0.106996 Loss(val): 0.110476 [01:44:49] Epoch 6: Loss(train): 0.100696 Loss(val): 0.104549 [01:52:04] Epoch 8: Loss(train): 0.094171 Loss(val): 0.098184 [01:58:26] Epoch 10: Loss(train): 0.089803 Loss(val): 0.093782 [02:03:44] Epoch 12: Loss(train): 0.087074 Loss(val): 0.091006 [02:08:33] Epoch 14: Loss(train): 0.085113 Loss(val): 0.088931 [02:13:20] Epoch 16: Loss(train): 0.083500 Loss(val): 0.087246 [02:21:19] Epoch 18: Loss(train): 0.081911 Loss(val): 0.085563 [02:28:13] Epoch 20: Loss(train): 0.080556 Loss(val): 0.084120 [02:34:03] Epoch 22: Loss(train): 0.079362 Loss(val): 0.082816 [02:39:08] Epoch 24: Loss(train): 0.078311 Loss(val): 0.081671 [02:43:58] Epoch 26: Loss(train): 0.077400 Loss(val): 0.080661 [02:49:40] Epoch 28: Loss(train): 0.076570 Loss(val): 0.079715 [02:57:52] Epoch 30: Loss(train): 0.075883 Loss(val): 0.078951 [03:04:07] Epoch 32: Loss(train): 0.075348 Loss(val): 0.078346 [03:09:39] Epoch 34: Loss(train): 0.074812 Loss(val): 0.077732 [03:14:37] Epoch 36: Loss(train): 0.074405 Loss(val): 0.077256 [03:19:32] Epoch 38: Loss(train): 0.074032 Loss(val): 0.076837 [03:27:14] Epoch 40: Loss(train): 0.073680 Loss(val): 0.076428 [03:28:09] FINAL(40) Loss(val): 0.076428 Accuarcy: 0.274915 Search 17 of 500 momentum0.9, features=[64, 64, 64], dropout_rate=0.4 kernel=Tuple{Int64,Int64}[(5, 1), (3, 1), (3, 6)], pooldims=Tuple{Int64,Int64}[(3, 1), (3, 1)], learning_rate=0.03 [03:29:53] INIT Loss(val): 0.131165 Accuarcy: 0.093435 [03:41:05] Epoch 2: Loss(train): 0.068231 Loss(val): 0.066406 [03:51:07] Epoch 4: Loss(train): 0.064709 Loss(val): 0.062722 [04:06:03] Epoch 6: Loss(train): 0.062242 Loss(val): 0.060431 [04:17:27] Epoch 8: Loss(train): 0.060373 Loss(val): 0.058780 [04:28:16] Epoch 10: Loss(train): 0.059085 Loss(val): 0.057533 [04:43:05] Epoch 12: Loss(train): 0.058318 Loss(val): 0.056778 [04:53:42] Epoch 14: Loss(train): 0.057390 Loss(val): 0.055989 [05:06:46] Epoch 16: Loss(train): 0.056481 Loss(val): 0.055205 [05:19:57] Epoch 18: Loss(train): 0.055879 Loss(val): 0.054648 [05:30:13] Epoch 20: Loss(train): 0.055364 Loss(val): 0.054170 [05:45:34] Epoch 22: Loss(train): 0.054641 Loss(val): 0.053574 [05:56:57] Epoch 24: Loss(train): 0.054031 Loss(val): 0.053062 [06:07:12] Epoch 26: Loss(train): 0.053433 Loss(val): 0.052634 Early stopping with Loss(train) 0.054989 at epoch 27 (Accuracy: 0.557653) Search 18 of 500 momentum0.92, features=[32, 32, 32], dropout_rate=0.3 kernel=Tuple{Int64,Int64}[(7, 1), (7, 1), (2, 6)], pooldims=Tuple{Int64,Int64}[(3, 1), (3, 1)], learning_rate=0.001 [06:15:45] INIT Loss(val): 0.118689 Accuarcy: 0.101990 [06:19:38] Epoch 2: Loss(train): 0.068195 Loss(val): 0.069483 [06:23:52] Epoch 4: Loss(train): 0.062983 Loss(val): 0.063813 [06:28:34] Epoch 6: Loss(train): 0.060712 Loss(val): 0.061474 [06:34:45] Epoch 8: Loss(train): 0.059224 Loss(val): 0.060002 [06:39:41] Epoch 10: Loss(train): 0.058208 Loss(val): 0.059001 [06:44:28] Epoch 12: Loss(train): 0.057323 Loss(val): 0.058152 [06:48:47] Epoch 14: Loss(train): 0.056700 Loss(val): 0.057535 [06:52:56] Epoch 16: Loss(train): 0.056110 Loss(val): 0.056953 [06:57:00] Epoch 18: Loss(train): 0.055600 Loss(val): 0.056486 [07:01:11] Epoch 20: Loss(train): 0.055162 Loss(val): 0.056064 [07:07:25] Epoch 22: Loss(train): 0.054831 Loss(val): 0.055762 [07:12:41] Epoch 24: Loss(train): 0.054532 Loss(val): 0.055485 [07:17:21] Epoch 26: Loss(train): 0.054293 Loss(val): 0.055247 [07:21:44] Epoch 28: Loss(train): 0.054086 Loss(val): 0.055038 [07:25:57] Epoch 30: Loss(train): 0.053895 Loss(val): 0.054867 [07:30:12] Epoch 32: Loss(train): 0.053734 Loss(val): 0.054717 [07:35:30] Epoch 34: Loss(train): 0.053598 Loss(val): 0.054584 [07:41:27] Epoch 36: Loss(train): 0.053484 Loss(val): 0.054487 [07:46:25] Epoch 38: Loss(train): 0.053388 Loss(val): 0.054393 [07:50:59] Epoch 40: Loss(train): 0.053292 Loss(val): 0.054313 [07:51:26] FINAL(40) Loss(val): 0.054313 Accuarcy: 0.432704 Search 19 of 500 momentum0.9, features=[96, 192, 192], dropout_rate=0.4 kernel=Tuple{Int64,Int64}[(5, 1), (5, 1), (2, 6)], pooldims=Tuple{Int64,Int64}[(3, 1), (3, 1)], learning_rate=1.0 [07:52:19] INIT Loss(val): 0.174489 Accuarcy: 0.104456 [08:05:37] Epoch 2: Loss(train): 0.102096 Loss(val): 0.102915 [08:21:51] Epoch 4: Loss(train): 0.066749 Loss(val): 0.065971 [08:35:13] Epoch 6: Loss(train): 0.055856 Loss(val): 0.055964 [08:52:20] Epoch 8: Loss(train): 0.047794 Loss(val): 0.047813 [09:05:42] Epoch 10: Loss(train): 0.041983 Loss(val): 0.042211 [09:23:11] Epoch 12: Loss(train): 0.037630 Loss(val): 0.038138 [09:36:57] Epoch 14: Loss(train): 0.034115 Loss(val): 0.034479 [09:55:56] Epoch 16: Loss(train): 0.032334 Loss(val): 0.032522 [10:09:49] Epoch 18: Loss(train): 0.030284 Loss(val): 0.030510 [10:29:29] Epoch 20: Loss(train): 0.028992 Loss(val): 0.029348 [10:43:10] Epoch 22: Loss(train): 0.028101 Loss(val): 0.028161 [11:02:49] Epoch 24: Loss(train): 0.027532 Loss(val): 0.027852 Early stopping with Loss(train) 0.028088 at epoch 25 (Accuracy: 0.567806) Search 20 of 500 momentum0.9, features=[96, 192, 192], dropout_rate=0.6 kernel=Tuple{Int64,Int64}[(3, 1), (3, 1), (3, 6)], pooldims=Tuple{Int64,Int64}[(2, 1), (2, 1)], learning_rate=0.3 [11:12:09] INIT Loss(val): 0.158469 Accuarcy: 0.092806 [11:25:25] Epoch 2: Loss(train): 0.092062 Loss(val): 0.091566 [11:38:25] Epoch 4: Loss(train): 0.081868 Loss(val): 0.081678 [11:49:01] Epoch 6: Loss(train): 0.076118 Loss(val): 0.075529 [12:04:29] Epoch 8: Loss(train): 0.073110 Loss(val): 0.072846 [12:16:23] Epoch 10: Loss(train): 0.069156 Loss(val): 0.069085 [12:30:09] Epoch 12: Loss(train): 0.066856 Loss(val): 0.067039 [12:44:01] Epoch 14: Loss(train): 0.065261 Loss(val): 0.065455 [12:54:52] Epoch 16: Loss(train): 0.063509 Loss(val): 0.064036 [13:10:55] Epoch 18: Loss(train): 0.061315 Loss(val): 0.061605 [13:22:43] Epoch 20: Loss(train): 0.059638 Loss(val): 0.059675 [13:36:46] Epoch 22: Loss(train): 0.058390 Loss(val): 0.058582 [13:50:21] Epoch 24: Loss(train): 0.057187 Loss(val): 0.057440 [14:01:25] Epoch 26: Loss(train): 0.056083 Loss(val): 0.056261 [14:12:57] Epoch 28: Loss(train): 0.055120 Loss(val): 0.055430 [14:25:54] Epoch 30: Loss(train): 0.054959 Loss(val): 0.055094 [14:40:39] Epoch 32: Loss(train): 0.054236 Loss(val): 0.054509 [14:52:26] Epoch 34: Loss(train): 0.053323 Loss(val): 0.053676 [15:04:28] Epoch 36: Loss(train): 0.052686 Loss(val): 0.053072 [15:18:03] Epoch 38: Loss(train): 0.051935 Loss(val): 0.052419 [15:29:22] Epoch 40: Loss(train): 0.051245 Loss(val): 0.051794 [15:30:12] FINAL(40) Loss(val): 0.051794 Accuarcy: 0.634048 Search 21 of 500 momentum0.99, features=[96, 192, 192], dropout_rate=0.3 kernel=Tuple{Int64,Int64}[(7, 1), (7, 1), (3, 6)], pooldims=Tuple{Int64,Int64}[(3, 1), (3, 1)], learning_rate=0.1 [15:31:35] INIT Loss(val): 0.245048 Accuarcy: 0.072585 [15:49:17] Epoch 2: Loss(train): 0.106915 Loss(val): 0.105428 [16:05:12] Epoch 4: Loss(train): 0.089091 Loss(val): 0.088908 [16:23:07] Epoch 6: Loss(train): 0.075282 Loss(val): 0.074277 [16:40:46] Epoch 8: Loss(train): 0.062850 Loss(val): 0.062624 [16:58:29] Epoch 10: Loss(train): 0.056189 Loss(val): 0.056083 [17:17:16] Epoch 12: Loss(train): 0.049655 Loss(val): 0.049626 [17:34:08] Epoch 14: Loss(train): 0.045409 Loss(val): 0.045518 [17:54:19] Epoch 16: Loss(train): 0.042162 Loss(val): 0.042254 [18:10:02] Epoch 18: Loss(train): 0.039818 Loss(val): 0.040065 [18:30:42] Epoch 20: Loss(train): 0.038585 Loss(val): 0.038927 [18:46:52] Epoch 22: Loss(train): 0.038263 Loss(val): 0.038389 [19:07:10] Epoch 24: Loss(train): 0.034990 Loss(val): 0.035632 [19:25:45] Epoch 26: Loss(train): 0.034374 Loss(val): 0.035082 [19:43:48] Epoch 28: Loss(train): 0.033463 Loss(val): 0.033584 [20:03:44] Epoch 30: Loss(train): 0.032233 Loss(val): 0.032437 [20:20:20] Epoch 32: Loss(train): 0.033065 Loss(val): 0.032645 [20:41:38] Epoch 34: Loss(train): 0.032210 Loss(val): 0.032403 [20:58:04] Epoch 36: Loss(train): 0.031725 Loss(val): 0.032099 [21:18:34] Epoch 38: Loss(train): 0.030471 Loss(val): 0.030934 [21:36:54] Epoch 40: Loss(train): 0.030312 Loss(val): 0.030813 [21:38:37] FINAL(40) Loss(val): 0.030813 Accuarcy: 0.600442 Search 22 of 500 momentum0.98, features=[96, 192, 192], dropout_rate=0.1 kernel=Tuple{Int64,Int64}[(7, 1), (7, 1), (2, 6)], pooldims=Tuple{Int64,Int64}[(3, 1), (3, 1)], learning_rate=1.0 [21:41:08] INIT Loss(val): 0.164662 Accuarcy: 0.097636 [21:57:03] Epoch 2: Loss(train): 20.748877 Loss(val): 20.759123 [22:12:28] Epoch 4: Loss(train): 27.568346 Loss(val): 27.566154 [22:30:34] Epoch 6: Loss(train): 28.220102 Loss(val): 28.217909 [22:48:10] Epoch 8: Loss(train): 67.155548 Loss(val): 67.165779 [23:07:08] Epoch 10: Loss(train): 67.068855 Loss(val): 67.079079 [23:23:42] Epoch 12: Loss(train): 66.994072 Loss(val): 67.004318 [23:43:22] Epoch 14: Loss(train): 66.931145 Loss(val): 66.941414 [23:59:21] Epoch 16: Loss(train): 66.875031 Loss(val): 66.885284 [00:19:03] Epoch 18: Loss(train): 66.825371 Loss(val): 66.835571 [00:36:41] Epoch 20: Loss(train): 66.781197 Loss(val): 66.791435 [00:54:54] Epoch 22: Loss(train): 66.741692 Loss(val): 66.751991 [01:15:18] Epoch 24: Loss(train): 66.706535 Loss(val): 66.716827 [01:32:32] Epoch 26: Loss(train): 66.675232 Loss(val): 66.685478 [01:54:11] Epoch 28: Loss(train): 66.647171 Loss(val): 66.657463 [02:12:09] Epoch 30: Loss(train): 66.621796 Loss(val): 66.632065 [02:32:33] Epoch 32: Loss(train): 66.598854 Loss(val): 66.609077 [02:53:21] Epoch 34: Loss(train): 66.578270 Loss(val): 66.588493 [03:10:47] Epoch 36: Loss(train): 66.560196 Loss(val): 66.570473 [03:33:05] Epoch 38: Loss(train): 66.544395 Loss(val): 66.554642 [03:51:21] Epoch 40: Loss(train): 66.530609 Loss(val): 66.540833 [03:53:19] FINAL(40) Loss(val): 66.540833 Accuarcy: 0.080357 Search 23 of 500 momentum0.99, features=[64, 64, 64], dropout_rate=0.3 kernel=Tuple{Int64,Int64}[(7, 1), (7, 1), (3, 6)], pooldims=Tuple{Int64,Int64}[(2, 1), (2, 1)], learning_rate=0.01 [03:56:53] INIT Loss(val): 0.156261 Accuarcy: 0.086173 [04:05:45] Epoch 2: Loss(train): 0.079105 Loss(val): 0.076430 [04:12:50] Epoch 4: Loss(train): 0.067922 Loss(val): 0.066296 [04:19:22] Epoch 6: Loss(train): 0.062268 Loss(val): 0.061061 [04:29:48] Epoch 8: Loss(train): 0.059802 Loss(val): 0.059280 [04:39:00] Epoch 10: Loss(train): 0.059741 Loss(val): 0.059402 [04:46:13] Epoch 12: Loss(train): 0.058659 Loss(val): 0.058159 [04:52:49] Epoch 14: Loss(train): 0.057998 Loss(val): 0.057408 [05:03:34] Epoch 16: Loss(train): 0.057697 Loss(val): 0.057173 [05:12:41] Epoch 18: Loss(train): 0.058854 Loss(val): 0.058166 [05:19:47] Epoch 20: Loss(train): 0.056203 Loss(val): 0.056039 [05:26:30] Epoch 22: Loss(train): 0.056706 Loss(val): 0.056652 [05:37:03] Epoch 24: Loss(train): 0.057241 Loss(val): 0.057240 [05:45:12] Epoch 26: Loss(train): 0.056111 Loss(val): 0.055896 [05:52:35] Epoch 28: Loss(train): 0.054545 Loss(val): 0.054455 [05:59:21] Epoch 30: Loss(train): 0.053666 Loss(val): 0.053996 [06:06:18] Epoch 32: Loss(train): 0.055091 Loss(val): 0.055624 [06:13:18] Epoch 34: Loss(train): 0.054184 Loss(val): 0.054805 [06:20:29] Epoch 36: Loss(train): 0.052324 Loss(val): 0.052880 [06:30:01] Epoch 38: Loss(train): 0.051825 Loss(val): 0.052219 [06:38:57] Epoch 40: Loss(train): 0.050568 Loss(val): 0.051316 [06:39:40] FINAL(40) Loss(val): 0.051316 Accuarcy: 0.649643 Search 24 of 500 momentum0.96, features=[32, 64, 128], dropout_rate=0.8 kernel=Tuple{Int64,Int64}[(3, 1), (3, 1), (3, 6)], pooldims=Tuple{Int64,Int64}[(3, 1), (3, 1)], learning_rate=0.03 [06:41:03] INIT Loss(val): 0.188385 Accuarcy: 0.083690 [06:52:54] Epoch 2: Loss(train): 0.072349 Loss(val): 0.071118 [07:08:32] Epoch 4: Loss(train): 0.064639 Loss(val): 0.063467 [07:21:35] Epoch 6: Loss(train): 0.061816 Loss(val): 0.060563 [07:37:04] Epoch 8: Loss(train): 0.058359 Loss(val): 0.057653 [07:50:18] Epoch 10: Loss(train): 0.057401 Loss(val): 0.056310 [08:04:59] Epoch 12: Loss(train): 0.056189 Loss(val): 0.055490 [08:19:04] Epoch 14: Loss(train): 0.054788 Loss(val): 0.054267 [08:33:18] Epoch 16: Loss(train): 0.054138 Loss(val): 0.053642 [08:48:22] Epoch 18: Loss(train): 0.053486 Loss(val): 0.052964 [09:01:45] Epoch 20: Loss(train): 0.053176 Loss(val): 0.052770 [09:19:46] Epoch 22: Loss(train): 0.052395 Loss(val): 0.052135 [09:32:18] Epoch 24: Loss(train): 0.052084 Loss(val): 0.051754 [09:50:48] Epoch 26: Loss(train): 0.051454 Loss(val): 0.051278 [10:03:51] Epoch 28: Loss(train): 0.051378 Loss(val): 0.051049 [10:22:26] Epoch 30: Loss(train): 0.051123 Loss(val): 0.050865 [10:35:48] Epoch 32: Loss(train): 0.050678 Loss(val): 0.050489 [10:54:01] Epoch 34: Loss(train): 0.050485 Loss(val): 0.050431 [11:07:16] Epoch 36: Loss(train): 0.050130 Loss(val): 0.050041 [11:25:31] Epoch 38: Loss(train): 0.050138 Loss(val): 0.049954 [11:39:24] Epoch 40: Loss(train): 0.049971 Loss(val): 0.049950 [11:40:54] FINAL(40) Loss(val): 0.049950 Accuarcy: 0.656633 Search 25 of 500 momentum0.94, features=[96, 192, 192], dropout_rate=0.4 kernel=Tuple{Int64,Int64}[(7, 1), (7, 1), (2, 6)], pooldims=Tuple{Int64,Int64}[(2, 1), (2, 1)], learning_rate=0.3 [11:43:33] INIT Loss(val): 0.206394 Accuarcy: 0.087228 [11:52:00] Epoch 2: Loss(train): 0.101289 Loss(val): 0.103436 [11:59:34] Epoch 4: Loss(train): 0.080375 Loss(val): 0.080668 [12:05:42] Epoch 6: Loss(train): 0.071989 Loss(val): 0.073021 [12:11:22] Epoch 8: Loss(train): 0.067130 Loss(val): 0.066621 [12:17:03] Epoch 10: Loss(train): 0.062482 Loss(val): 0.062284 [12:26:05] Epoch 12: Loss(train): 0.059126 Loss(val): 0.059354 [12:33:42] Epoch 14: Loss(train): 0.054605 Loss(val): 0.055075 [12:40:10] Epoch 16: Loss(train): 0.052491 Loss(val): 0.053066 [12:46:25] Epoch 18: Loss(train): 0.050310 Loss(val): 0.050966 [12:52:38] Epoch 20: Loss(train): 0.048141 Loss(val): 0.048782 [12:59:55] Epoch 22: Loss(train): 0.046192 Loss(val): 0.047045 [13:06:18] Epoch 24: Loss(train): 0.044195 Loss(val): 0.045356 [13:12:35] Epoch 26: Loss(train): 0.043286 Loss(val): 0.044734 [13:18:26] Epoch 28: Loss(train): 0.042307 Loss(val): 0.043518 [13:24:05] Epoch 30: Loss(train): 0.041010 Loss(val): 0.042333 [13:31:27] Epoch 32: Loss(train): 0.040678 Loss(val): 0.042079 [13:38:43] Epoch 34: Loss(train): 0.039944 Loss(val): 0.041404 [13:45:18] Epoch 36: Loss(train): 0.039505 Loss(val): 0.040942 [13:51:09] Epoch 38: Loss(train): 0.039001 Loss(val): 0.040541 [13:56:48] Epoch 40: Loss(train): 0.038627 Loss(val): 0.040171 [13:57:14] FINAL(40) Loss(val): 0.040171 Accuarcy: 0.615595 Search 26 of 500 momentum0.99, features=[64, 64, 64], dropout_rate=0.1 kernel=Tuple{Int64,Int64}[(5, 1), (5, 1), (3, 6)], pooldims=Tuple{Int64,Int64}[(2, 1), (2, 1)], learning_rate=0.03 [13:58:10] INIT Loss(val): 0.150567 Accuarcy: 0.075561 [14:05:36] Epoch 2: Loss(train): 0.109244 Loss(val): 0.109046 [14:13:33] Epoch 4: Loss(train): 0.069782 Loss(val): 0.068761 [14:21:46] Epoch 6: Loss(train): 0.057850 Loss(val): 0.057034 [14:32:51] Epoch 8: Loss(train): 0.053643 Loss(val): 0.053023 [14:41:22] Epoch 10: Loss(train): 0.051602 Loss(val): 0.051370 [14:49:15] Epoch 12: Loss(train): 0.051565 Loss(val): 0.051038 [14:56:52] Epoch 14: Loss(train): 0.050527 Loss(val): 0.050176 [15:08:13] Epoch 16: Loss(train): 0.047978 Loss(val): 0.047917 [15:16:48] Epoch 18: Loss(train): 0.046836 Loss(val): 0.046800 [15:24:29] Epoch 20: Loss(train): 0.049216 Loss(val): 0.049681 [15:34:18] Epoch 22: Loss(train): 0.044710 Loss(val): 0.045216 [15:43:20] Epoch 24: Loss(train): 0.046563 Loss(val): 0.047376 [15:51:33] Epoch 26: Loss(train): 0.042391 Loss(val): 0.042913 [15:59:26] Epoch 28: Loss(train): 0.041383 Loss(val): 0.041802 [16:09:13] Epoch 30: Loss(train): 0.040634 Loss(val): 0.041065 [16:18:52] Epoch 32: Loss(train): 0.039923 Loss(val): 0.040496 [16:27:03] Epoch 34: Loss(train): 0.040688 Loss(val): 0.042053 [16:34:46] Epoch 36: Loss(train): 0.038398 Loss(val): 0.039507 [16:45:25] Epoch 38: Loss(train): 0.037850 Loss(val): 0.038930 [16:53:54] Epoch 40: Loss(train): 0.038935 Loss(val): 0.039747 [16:54:57] FINAL(40) Loss(val): 0.039747 Accuarcy: 0.614592 Search 27 of 500 momentum0.98, features=[96, 192, 192], dropout_rate=0.1 kernel=Tuple{Int64,Int64}[(3, 1), (3, 1), (2, 6)], pooldims=Tuple{Int64,Int64}[(2, 1), (2, 1)], learning_rate=0.03 [16:56:34] INIT Loss(val): 0.134213 Accuarcy: 0.091514 [17:02:28] Epoch 2: Loss(train): 0.090860 Loss(val): 0.090545 [17:08:42] Epoch 4: Loss(train): 0.078672 Loss(val): 0.076943 [17:16:48] Epoch 6: Loss(train): 0.069592 Loss(val): 0.068936 [17:23:32] Epoch 8: Loss(train): 0.066346 Loss(val): 0.065861 [17:29:47] Epoch 10: Loss(train): 0.064795 Loss(val): 0.064491 [17:35:36] Epoch 12: Loss(train): 0.063546 Loss(val): 0.063492 [17:43:11] Epoch 14: Loss(train): 0.062957 Loss(val): 0.062895 [17:50:18] Epoch 16: Loss(train): 0.062168 Loss(val): 0.062227 [17:56:49] Epoch 18: Loss(train): 0.061885 Loss(val): 0.061934 [18:02:49] Epoch 20: Loss(train): 0.061532 Loss(val): 0.061760 Early stopping with Loss(train) 0.061861 at epoch 20 (Accuracy: 0.608146) Search 28 of 500 momentum0.9, features=[64, 64, 64], dropout_rate=0.4 kernel=Tuple{Int64,Int64}[(3, 1), (3, 1), (3, 6)], pooldims=Tuple{Int64,Int64}[(2, 1), (2, 1)], learning_rate=0.003 [18:04:05] INIT Loss(val): 0.128106 Accuarcy: 0.102738 [18:12:15] Epoch 2: Loss(train): 0.074689 Loss(val): 0.075390 [18:24:22] Epoch 4: Loss(train): 0.068198 Loss(val): 0.068720 [18:33:08] Epoch 6: Loss(train): 0.066180 Loss(val): 0.066643 [18:40:59] Epoch 8: Loss(train): 0.064997 Loss(val): 0.065444 [18:53:03] Epoch 10: Loss(train): 0.064207 Loss(val): 0.064619 [19:02:48] Epoch 12: Loss(train): 0.063483 Loss(val): 0.063943 [19:10:44] Epoch 14: Loss(train): 0.063037 Loss(val): 0.063462 [19:21:21] Epoch 16: Loss(train): 0.062571 Loss(val): 0.063062 [19:32:27] Epoch 18: Loss(train): 0.062316 Loss(val): 0.062758 [19:41:00] Epoch 20: Loss(train): 0.061978 Loss(val): 0.062450 [19:49:41] Epoch 22: Loss(train): 0.061717 Loss(val): 0.062215 [20:02:18] Epoch 24: Loss(train): 0.061504 Loss(val): 0.062028 [20:11:17] Epoch 26: Loss(train): 0.061352 Loss(val): 0.061835 [20:19:18] Epoch 28: Loss(train): 0.061151 Loss(val): 0.061692 [20:31:40] Epoch 30: Loss(train): 0.061015 Loss(val): 0.061562 [20:41:43] Epoch 32: Loss(train): 0.060934 Loss(val): 0.061451 [20:49:44] Epoch 34: Loss(train): 0.060828 Loss(val): 0.061372 [21:01:36] Epoch 36: Loss(train): 0.060729 Loss(val): 0.061282 [21:11:39] Epoch 38: Loss(train): 0.060642 Loss(val): 0.061227 [21:20:03] Epoch 40: Loss(train): 0.060570 Loss(val): 0.061159 [21:20:56] FINAL(40) Loss(val): 0.061159 Accuarcy: 0.525493 Search 29 of 500 momentum0.92, features=[32, 32, 32], dropout_rate=0.4 kernel=Tuple{Int64,Int64}[(3, 1), (3, 1), (3, 6)], pooldims=Tuple{Int64,Int64}[(3, 1), (3, 1)], learning_rate=1.0 [21:22:36] INIT Loss(val): 0.137907 Accuarcy: 0.095374 [21:31:00] Epoch 2: Loss(train): 0.065825 Loss(val): 0.066324 [21:40:00] Epoch 4: Loss(train): 0.058934 Loss(val): 0.057341 [21:47:02] Epoch 6: Loss(train): 0.051880 Loss(val): 0.052709 [21:52:58] Epoch 8: Loss(train): 0.044297 Loss(val): 0.045645 [21:58:54] Epoch 10: Loss(train): 0.041006 Loss(val): 0.042122 [22:04:54] Epoch 12: Loss(train): 0.040873 Loss(val): 0.042592 [22:10:53] Epoch 14: Loss(train): 0.037040 Loss(val): 0.038307 [22:17:16] Epoch 16: Loss(train): 0.039913 Loss(val): 0.042390 [22:24:29] Epoch 18: Loss(train): 0.031755 Loss(val): 0.033696 [22:32:00] Epoch 20: Loss(train): 0.030639 Loss(val): 0.030783 [22:39:06] Epoch 22: Loss(train): 0.028381 Loss(val): 0.028629 [22:45:35] Epoch 24: Loss(train): 0.027468 Loss(val): 0.027725 [22:51:38] Epoch 26: Loss(train): 0.025523 Loss(val): 0.025599 [23:00:27] Epoch 28: Loss(train): 0.025676 Loss(val): 0.025638 [23:07:41] Epoch 30: Loss(train): 0.025141 Loss(val): 0.025392 [23:14:21] Epoch 32: Loss(train): 0.024193 Loss(val): 0.024012 [23:20:33] Epoch 34: Loss(train): 0.023470 Loss(val): 0.023429 [23:28:02] Epoch 36: Loss(train): 0.022927 Loss(val): 0.022909 [23:36:20] Epoch 38: Loss(train): 0.022175 Loss(val): 0.022574 [23:43:15] Epoch 40: Loss(train): 0.021259 Loss(val): 0.021576 [23:43:50] FINAL(40) Loss(val): 0.021576 Accuarcy: 0.537568 Search 30 of 500 momentum0.96, features=[96, 192, 192], dropout_rate=0.3 kernel=Tuple{Int64,Int64}[(3, 1), (3, 1), (2, 6)], pooldims=Tuple{Int64,Int64}[(3, 1), (3, 1)], learning_rate=1.0 [23:44:59] INIT Loss(val): 0.157926 Accuarcy: 0.096905 [00:07:35] Epoch 2: Loss(train): 0.083815 Loss(val): 0.085143 [00:30:44] Epoch 4: Loss(train): 0.066922 Loss(val): 0.068693 [00:54:18] Epoch 6: Loss(train): 0.062800 Loss(val): 0.064619 [01:19:21] Epoch 8: Loss(train): 0.055423 Loss(val): 0.056209 [01:45:26] Epoch 10: Loss(train): 0.051950 Loss(val): 0.052295 [02:10:56] Epoch 12: Loss(train): 0.048478 Loss(val): 0.050152 [02:34:46] Epoch 14: Loss(train): 0.049978 Loss(val): 0.049653 [03:01:13] Epoch 16: Loss(train): 0.043648 Loss(val): 0.045337 [03:29:02] Epoch 18: Loss(train): 0.045327 Loss(val): 0.048296 [03:56:01] Epoch 20: Loss(train): 0.045312 Loss(val): 0.048873 [04:21:49] Epoch 22: Loss(train): 0.040705 Loss(val): 0.040873 [04:45:42] Epoch 24: Loss(train): 0.046981 Loss(val): 0.049900 [05:12:34] Epoch 26: Loss(train): 0.038865 Loss(val): 0.040929 [05:38:34] Epoch 28: Loss(train): 0.030650 Loss(val): 0.030977 [06:00:05] Epoch 30: Loss(train): 0.030971 Loss(val): 0.032735 [06:25:37] Epoch 32: Loss(train): 0.032108 Loss(val): 0.030270 [06:50:02] Epoch 34: Loss(train): 0.029535 Loss(val): 0.028256 [07:13:34] Epoch 36: Loss(train): 0.025526 Loss(val): 0.024374 [07:38:32] Epoch 38: Loss(train): 0.025246 Loss(val): 0.023842 [08:03:25] Epoch 40: Loss(train): 0.024390 Loss(val): 0.023580 [08:05:38] FINAL(40) Loss(val): 0.023580 Accuarcy: 0.518469 Search 31 of 500 momentum0.98, features=[64, 64, 64], dropout_rate=0.6 kernel=Tuple{Int64,Int64}[(5, 1), (5, 1), (3, 6)], pooldims=Tuple{Int64,Int64}[(2, 1), (2, 1)], learning_rate=0.001 [08:08:47] INIT Loss(val): 0.136395 Accuarcy: 0.102109 [08:16:59] Epoch 2: Loss(train): 0.083085 Loss(val): 0.084971 [08:28:48] Epoch 4: Loss(train): 0.070578 Loss(val): 0.071192 [08:38:26] Epoch 6: Loss(train): 0.067818 Loss(val): 0.068311 [08:47:17] Epoch 8: Loss(train): 0.066219 Loss(val): 0.066654 [08:58:49] Epoch 10: Loss(train): 0.065115 Loss(val): 0.065624 [09:08:46] Epoch 12: Loss(train): 0.064355 Loss(val): 0.064817 [09:17:31] Epoch 14: Loss(train): 0.063803 Loss(val): 0.064206 [09:28:50] Epoch 16: Loss(train): 0.063242 Loss(val): 0.063749 [09:38:48] Epoch 18: Loss(train): 0.062930 Loss(val): 0.063295 [09:47:45] Epoch 20: Loss(train): 0.062665 Loss(val): 0.063019 [09:58:15] Epoch 22: Loss(train): 0.062384 Loss(val): 0.062730 [10:09:26] Epoch 24: Loss(train): 0.062142 Loss(val): 0.062500 [10:18:33] Epoch 26: Loss(train): 0.061943 Loss(val): 0.062314 [10:30:07] Epoch 28: Loss(train): 0.061784 Loss(val): 0.062140 [10:40:47] Epoch 30: Loss(train): 0.061690 Loss(val): 0.062006 [10:50:03] Epoch 32: Loss(train): 0.061544 Loss(val): 0.061859 [11:02:31] Epoch 34: Loss(train): 0.061405 Loss(val): 0.061765 [11:13:54] Epoch 36: Loss(train): 0.061310 Loss(val): 0.061659 Early stopping with Loss(train) 0.065227 at epoch 36 (Accuracy: 0.477653) Search 32 of 500 momentum0.92, features=[32, 64, 128], dropout_rate=0.1 kernel=Tuple{Int64,Int64}[(5, 1), (3, 1), (3, 6)], pooldims=Tuple{Int64,Int64}[(3, 1), (3, 1)], learning_rate=0.001 [11:16:41] INIT Loss(val): 0.156245 Accuarcy: 0.081514 [11:35:09] Epoch 2: Loss(train): 0.075818 Loss(val): 0.075976 [11:54:19] Epoch 4: Loss(train): 0.069581 Loss(val): 0.069867 [12:16:03] Epoch 6: Loss(train): 0.067491 Loss(val): 0.067784 [12:33:26] Epoch 8: Loss(train): 0.066309 Loss(val): 0.066613 [12:55:22] Epoch 10: Loss(train): 0.065570 Loss(val): 0.065932 [13:16:01] Epoch 12: Loss(train): 0.064942 Loss(val): 0.065361 [13:34:16] Epoch 14: Loss(train): 0.064421 Loss(val): 0.064918 [13:50:08] Epoch 16: Loss(train): 0.064045 Loss(val): 0.064583 [14:10:29] Epoch 18: Loss(train): 0.063691 Loss(val): 0.064300 [14:27:48] Epoch 20: Loss(train): 0.063414 Loss(val): 0.064062 [14:47:52] Epoch 22: Loss(train): 0.063163 Loss(val): 0.063851 [15:05:02] Epoch 24: Loss(train): 0.062957 Loss(val): 0.063692 [15:25:31] Epoch 26: Loss(train): 0.062760 Loss(val): 0.063534 [15:44:27] Epoch 28: Loss(train): 0.062596 Loss(val): 0.063417 [16:02:11] Epoch 30: Loss(train): 0.062460 Loss(val): 0.063311 [16:22:44] Epoch 32: Loss(train): 0.062348 Loss(val): 0.063210 [16:40:25] Epoch 34: Loss(train): 0.062251 Loss(val): 0.063149 [17:00:11] Epoch 36: Loss(train): 0.062162 Loss(val): 0.063097 [17:19:44] Epoch 38: Loss(train): 0.062111 Loss(val): 0.063032 [17:37:13] Epoch 40: Loss(train): 0.062047 Loss(val): 0.062991 [17:39:26] FINAL(40) Loss(val): 0.062991 Accuarcy: 0.500272 Search 33 of 500 momentum0.94, features=[64, 64, 64], dropout_rate=0.3 kernel=Tuple{Int64,Int64}[(5, 1), (5, 1), (2, 6)], pooldims=Tuple{Int64,Int64}[(3, 1), (3, 1)], learning_rate=0.001 [17:44:54] INIT Loss(val): 0.129508 Accuarcy: 0.087432 [18:04:53] Epoch 2: Loss(train): 0.072966 Loss(val): 0.074566 [18:29:05] Epoch 4: Loss(train): 0.067318 Loss(val): 0.068274 [18:53:29] Epoch 6: Loss(train): 0.064862 Loss(val): 0.065613 [19:14:04] Epoch 8: Loss(train): 0.063560 Loss(val): 0.064150 [19:39:54] Epoch 10: Loss(train): 0.062483 Loss(val): 0.063086 [20:05:05] Epoch 12: Loss(train): 0.061835 Loss(val): 0.062398 [20:28:37] Epoch 14: Loss(train): 0.061281 Loss(val): 0.061852 [20:50:12] Epoch 16: Loss(train): 0.060817 Loss(val): 0.061377 [21:16:16] Epoch 18: Loss(train): 0.060441 Loss(val): 0.061011 [21:35:17] Epoch 20: Loss(train): 0.060146 Loss(val): 0.060714 [21:58:43] Epoch 22: Loss(train): 0.059869 Loss(val): 0.060446 [22:20:18] Epoch 24: Loss(train): 0.059664 Loss(val): 0.060259 [22:42:47] Epoch 26: Loss(train): 0.059494 Loss(val): 0.060083 [23:06:12] Epoch 28: Loss(train): 0.059329 Loss(val): 0.059940 Early stopping with Loss(train) 0.062314 at epoch 29 (Accuracy: 0.436071) Search 34 of 500 momentum0.94, features=[64, 64, 64], dropout_rate=0.8 kernel=Tuple{Int64,Int64}[(5, 1), (3, 1), (2, 6)], pooldims=Tuple{Int64,Int64}[(2, 1), (2, 1)], learning_rate=0.01 [23:23:32] INIT Loss(val): 0.161144 Accuarcy: 0.107738 [23:47:42] Epoch 2: Loss(train): 0.079291 Loss(val): 0.079141 [00:12:49] Epoch 4: Loss(train): 0.067550 Loss(val): 0.066014 [00:39:19] Epoch 6: Loss(train): 0.063289 Loss(val): 0.061791 [01:05:04] Epoch 8: Loss(train): 0.060796 Loss(val): 0.059554 [01:31:19] Epoch 10: Loss(train): 0.059412 Loss(val): 0.058274 [01:59:17] Epoch 12: Loss(train): 0.058747 Loss(val): 0.057489 [02:28:58] Epoch 14: Loss(train): 0.058412 Loss(val): 0.057003 [02:58:46] Epoch 16: Loss(train): 0.057484 Loss(val): 0.056236 [03:28:32] Epoch 18: Loss(train): 0.057060 Loss(val): 0.055838 [03:57:59] Epoch 20: Loss(train): 0.056660 Loss(val): 0.055544 [04:26:58] Epoch 22: Loss(train): 0.056213 Loss(val): 0.055160 [04:56:08] Epoch 24: Loss(train): 0.056014 Loss(val): 0.054934 [05:21:16] Epoch 26: Loss(train): 0.055638 Loss(val): 0.054669 [05:48:43] Epoch 28: Loss(train): 0.055279 Loss(val): 0.054408 [06:13:58] Epoch 30: Loss(train): 0.055161 Loss(val): 0.054242 [06:41:24] Epoch 32: Loss(train): 0.054907 Loss(val): 0.054066 [07:08:20] Epoch 34: Loss(train): 0.054621 Loss(val): 0.053844 [07:35:37] Epoch 36: Loss(train): 0.054479 Loss(val): 0.053736 [08:03:00] Epoch 38: Loss(train): 0.054394 Loss(val): 0.053631 [08:29:53] Epoch 40: Loss(train): 0.054193 Loss(val): 0.053513 [08:32:42] FINAL(40) Loss(val): 0.053513 Accuarcy: 0.610051 Search 35 of 500 momentum0.94, features=[64, 64, 64], dropout_rate=0.3 kernel=Tuple{Int64,Int64}[(5, 1), (3, 1), (3, 6)], pooldims=Tuple{Int64,Int64}[(3, 1), (3, 1)], learning_rate=0.1 [08:36:22] INIT Loss(val): 0.123761 Accuarcy: 0.091531 [08:57:47] Epoch 2: Loss(train): 0.065259 Loss(val): 0.065832 [09:19:41] Epoch 4: Loss(train): 0.059903 Loss(val): 0.058281 [09:44:11] Epoch 6: Loss(train): 0.056566 Loss(val): 0.054878 [10:10:43] Epoch 8: Loss(train): 0.053432 Loss(val): 0.052172 [10:35:42] Epoch 10: Loss(train): 0.051702 Loss(val): 0.050658 [10:57:51] Epoch 12: Loss(train): 0.050034 Loss(val): 0.049270 [11:24:31] Epoch 14: Loss(train): 0.048701 Loss(val): 0.048126 [11:51:11] Epoch 16: Loss(train): 0.047392 Loss(val): 0.047040 [12:17:09] Epoch 18: Loss(train): 0.046799 Loss(val): 0.046500 [12:40:51] Epoch 20: Loss(train): 0.045793 Loss(val): 0.045674 [13:05:28] Epoch 22: Loss(train): 0.045341 Loss(val): 0.045252 [13:25:44] Epoch 24: Loss(train): 0.044809 Loss(val): 0.044760 [13:50:57] Epoch 26: Loss(train): 0.044566 Loss(val): 0.044537 [14:14:24] Epoch 28: Loss(train): 0.043879 Loss(val): 0.043859 [14:36:11] Epoch 30: Loss(train): 0.043258 Loss(val): 0.043268 [15:00:39] Epoch 32: Loss(train): 0.042763 Loss(val): 0.042793 [15:25:03] Epoch 34: Loss(train): 0.042519 Loss(val): 0.042491 [15:48:44] Epoch 36: Loss(train): 0.042028 Loss(val): 0.041963 [16:10:14] Epoch 38: Loss(train): 0.041775 Loss(val): 0.041745 [16:34:21] Epoch 40: Loss(train): 0.041268 Loss(val): 0.041277 [16:37:30] FINAL(40) Loss(val): 0.041277 Accuarcy: 0.626071 Search 36 of 500 momentum0.94, features=[32, 32, 32], dropout_rate=0.4 kernel=Tuple{Int64,Int64}[(5, 1), (3, 1), (2, 6)], pooldims=Tuple{Int64,Int64}[(2, 1), (2, 1)], learning_rate=0.3 [16:42:54] INIT Loss(val): 0.113207 Accuarcy: 0.083673 [17:05:24] Epoch 2: Loss(train): 0.071602 Loss(val): 0.071321 [17:30:50] Epoch 4: Loss(train): 0.060653 Loss(val): 0.061529 [17:55:40] Epoch 6: Loss(train): 0.050651 Loss(val): 0.051918 [18:18:57] Epoch 8: Loss(train): 0.046450 Loss(val): 0.047496 [18:46:06] Epoch 10: Loss(train): 0.042359 Loss(val): 0.043354 [19:13:08] Epoch 12: Loss(train): 0.038272 Loss(val): 0.038781 [19:39:21] Epoch 14: Loss(train): 0.039356 Loss(val): 0.039668 [20:03:45] Epoch 16: Loss(train): 0.036349 Loss(val): 0.036034 [20:27:40] Epoch 18: Loss(train): 0.034223 Loss(val): 0.034151 [20:55:17] Epoch 20: Loss(train): 0.032276 Loss(val): 0.032745 [21:16:30] Epoch 22: Loss(train): 0.029983 Loss(val): 0.030295 [21:42:49] Epoch 24: Loss(train): 0.028562 Loss(val): 0.028459 [22:06:22] Epoch 26: Loss(train): 0.027147 Loss(val): 0.026862 [22:29:11] Epoch 28: Loss(train): 0.025605 Loss(val): 0.025301 [22:54:25] Epoch 30: Loss(train): 0.025205 Loss(val): 0.025065 [23:19:33] Epoch 32: Loss(train): 0.024206 Loss(val): 0.024281 [23:44:35] Epoch 34: Loss(train): 0.023469 Loss(val): 0.023604 [00:07:42] Epoch 36: Loss(train): 0.022483 Loss(val): 0.022621 [00:31:31] Epoch 38: Loss(train): 0.022307 Loss(val): 0.022504 [00:57:02] Epoch 40: Loss(train): 0.021628 Loss(val): 0.021826 [01:00:15] FINAL(40) Loss(val): 0.021826 Accuarcy: 0.606854 Search 37 of 500 momentum0.92, features=[64, 64, 64], dropout_rate=0.8 kernel=Tuple{Int64,Int64}[(3, 1), (3, 1), (2, 6)], pooldims=Tuple{Int64,Int64}[(3, 1), (3, 1)], learning_rate=1.0 [01:05:55] INIT Loss(val): 0.118011 Accuarcy: 0.093418 [01:28:59] Epoch 2: Loss(train): 0.071947 Loss(val): 0.071658 [01:54:44] Epoch 4: Loss(train): 0.057903 Loss(val): 0.058991 [02:20:57] Epoch 6: Loss(train): 0.048510 Loss(val): 0.047729 [02:46:19] Epoch 8: Loss(train): 0.042900 Loss(val): 0.043739 [03:11:28] Epoch 10: Loss(train): 0.036034 Loss(val): 0.035340 [03:38:58] Epoch 12: Loss(train): 0.035133 Loss(val): 0.036576 [04:06:04] Epoch 14: Loss(train): 0.031527 Loss(val): 0.032155 [04:33:18] Epoch 16: Loss(train): 0.032142 Loss(val): 0.033449 [04:56:07] Epoch 18: Loss(train): 0.029012 Loss(val): 0.028829 [05:20:51] Epoch 20: Loss(train): 0.027290 Loss(val): 0.027411 [05:43:24] Epoch 22: Loss(train): 0.026393 Loss(val): 0.025477 [06:09:01] Epoch 24: Loss(train): 0.025136 Loss(val): 0.024777 [06:34:51] Epoch 26: Loss(train): 0.023614 Loss(val): 0.022565 [07:00:12] Epoch 28: Loss(train): 0.023541 Loss(val): 0.023284 [07:24:35] Epoch 30: Loss(train): 0.023083 Loss(val): 0.022527 [07:48:19] Epoch 32: Loss(train): 0.022078 Loss(val): 0.021276 [08:12:51] Epoch 34: Loss(train): 0.021870 Loss(val): 0.021672 [08:36:53] Epoch 36: Loss(train): 0.021393 Loss(val): 0.021033 [09:00:56] Epoch 38: Loss(train): 0.021274 Loss(val): 0.020903 [09:27:55] Epoch 40: Loss(train): 0.022459 Loss(val): 0.021333 [09:31:11] FINAL(40) Loss(val): 0.021333 Accuarcy: 0.573265 Search 38 of 500 momentum0.92, features=[96, 192, 192], dropout_rate=0.8 kernel=Tuple{Int64,Int64}[(7, 1), (7, 1), (3, 6)], pooldims=Tuple{Int64,Int64}[(2, 1), (2, 1)], learning_rate=0.01 [09:40:13] INIT Loss(val): 0.177381 Accuarcy: 0.082194 [10:02:06] Epoch 2: Loss(train): 0.100145 Loss(val): 0.101442 [10:28:32] Epoch 4: Loss(train): 0.090080 Loss(val): 0.090478 [10:54:33] Epoch 6: Loss(train): 0.086961 Loss(val): 0.087106 [11:18:52] Epoch 8: Loss(train): 0.085055 Loss(val): 0.085218 [11:42:03] Epoch 10: Loss(train): 0.084008 Loss(val): 0.084082 [12:08:51] Epoch 12: Loss(train): 0.083061 Loss(val): 0.083309 [12:34:18] Epoch 14: Loss(train): 0.082237 Loss(val): 0.082573 [12:54:57] Epoch 16: Loss(train): 0.082143 Loss(val): 0.082355 [13:18:12] Epoch 18: Loss(train): 0.081369 Loss(val): 0.081832 [13:40:38] Epoch 20: Loss(train): 0.081240 Loss(val): 0.081622 [14:05:08] Epoch 22: Loss(train): 0.080846 Loss(val): 0.081271 [14:28:44] Epoch 24: Loss(train): 0.080775 Loss(val): 0.081209 Early stopping with Loss(train) 0.087498 at epoch 24 (Accuracy: 0.458163) Search 39 of 500 momentum0.94, features=[64, 64, 64], dropout_rate=0.1 kernel=Tuple{Int64,Int64}[(3, 1), (3, 1), (3, 6)], pooldims=Tuple{Int64,Int64}[(2, 1), (2, 1)], learning_rate=0.003 [14:32:44] INIT Loss(val): 0.131732 Accuarcy: 0.093946 [14:43:04] Epoch 2: Loss(train): 0.081255 Loss(val): 0.079859 [14:57:19] Epoch 4: Loss(train): 0.068332 Loss(val): 0.067841 [15:09:13] Epoch 6: Loss(train): 0.064528 Loss(val): 0.064704 [15:21:16] Epoch 8: Loss(train): 0.062921 Loss(val): 0.063261 [15:34:46] Epoch 10: Loss(train): 0.062030 Loss(val): 0.062498 [15:45:50] Epoch 12: Loss(train): 0.061265 Loss(val): 0.061891 [16:00:28] Epoch 14: Loss(train): 0.060731 Loss(val): 0.061435 [16:12:25] Epoch 16: Loss(train): 0.060351 Loss(val): 0.061083 [16:24:49] Epoch 18: Loss(train): 0.060006 Loss(val): 0.060862 [16:37:32] Epoch 20: Loss(train): 0.059765 Loss(val): 0.060644 [16:48:56] Epoch 22: Loss(train): 0.059551 Loss(val): 0.060483 [17:01:56] Epoch 24: Loss(train): 0.059328 Loss(val): 0.060275 [17:14:41] Epoch 26: Loss(train): 0.059153 Loss(val): 0.060095 [17:26:03] Epoch 28: Loss(train): 0.058965 Loss(val): 0.059951 [17:40:07] Epoch 30: Loss(train): 0.058867 Loss(val): 0.059819 [17:52:28] Epoch 32: Loss(train): 0.058637 Loss(val): 0.059653 [18:05:08] Epoch 34: Loss(train): 0.058543 Loss(val): 0.059542 [18:19:54] Epoch 36: Loss(train): 0.058402 Loss(val): 0.059439 [18:30:58] Epoch 38: Loss(train): 0.058293 Loss(val): 0.059341 Early stopping with Loss(train) 0.059225 at epoch 38 (Accuracy: 0.544405) Search 40 of 500 momentum0.92, features=[32, 32, 32], dropout_rate=0.1 kernel=Tuple{Int64,Int64}[(5, 1), (5, 1), (2, 6)], pooldims=Tuple{Int64,Int64}[(3, 1), (3, 1)], learning_rate=0.1 [18:34:47] INIT Loss(val): 0.124025 Accuarcy: 0.088776 [18:47:33] Epoch 2: Loss(train): 0.057543 Loss(val): 0.058215 [18:57:06] Epoch 4: Loss(train): 0.051879 Loss(val): 0.052554 [19:05:25] Epoch 6: Loss(train): 0.048303 Loss(val): 0.048884 [19:19:12] Epoch 8: Loss(train): 0.047418 Loss(val): 0.046097 [19:28:58] Epoch 10: Loss(train): 0.046238 Loss(val): 0.044915 [19:37:28] Epoch 12: Loss(train): 0.044154 Loss(val): 0.043053 [19:51:17] Epoch 14: Loss(train): 0.042655 Loss(val): 0.041790 [20:01:13] Epoch 16: Loss(train): 0.041199 Loss(val): 0.040503 [20:09:50] Epoch 18: Loss(train): 0.040215 Loss(val): 0.039571 [20:23:19] Epoch 20: Loss(train): 0.039205 Loss(val): 0.038591 [20:33:47] Epoch 22: Loss(train): 0.038505 Loss(val): 0.037907 [20:42:15] Epoch 24: Loss(train): 0.037870 Loss(val): 0.037287 [20:50:49] Epoch 26: Loss(train): 0.037377 Loss(val): 0.036808 [20:59:19] Epoch 28: Loss(train): 0.036958 Loss(val): 0.036429 [21:08:32] Epoch 30: Loss(train): 0.036559 Loss(val): 0.036016 [21:21:13] Epoch 32: Loss(train): 0.036171 Loss(val): 0.035656 [21:30:26] Epoch 34: Loss(train): 0.035821 Loss(val): 0.035373 [21:39:16] Epoch 36: Loss(train): 0.035724 Loss(val): 0.035267 [21:51:46] Epoch 38: Loss(train): 0.035684 Loss(val): 0.035258 [22:01:20] Epoch 40: Loss(train): 0.035597 Loss(val): 0.035128 [22:02:18] FINAL(40) Loss(val): 0.035128 Accuarcy: 0.636701 Search 41 of 500 momentum0.98, features=[64, 64, 64], dropout_rate=0.1 kernel=Tuple{Int64,Int64}[(3, 1), (3, 1), (3, 6)], pooldims=Tuple{Int64,Int64}[(3, 1), (3, 1)], learning_rate=0.003 [22:03:51] INIT Loss(val): 0.116976 Accuarcy: 0.093078 [22:29:47] Epoch 2: Loss(train): 0.064875 Loss(val): 0.063949 [22:57:17] Epoch 4: Loss(train): 0.059843 Loss(val): 0.058989 [23:24:07] Epoch 6: Loss(train): 0.057889 Loss(val): 0.057287 [23:50:54] Epoch 8: Loss(train): 0.056778 Loss(val): 0.056329 [00:17:41] Epoch 10: Loss(train): 0.056209 Loss(val): 0.055885 [00:43:47] Epoch 12: Loss(train): 0.055611 Loss(val): 0.055446 [01:11:43] Epoch 14: Loss(train): 0.055447 Loss(val): 0.055392 [01:40:33] Epoch 16: Loss(train): 0.055111 Loss(val): 0.055106 [02:09:24] Epoch 18: Loss(train): 0.054887 Loss(val): 0.054910 [02:39:30] Epoch 20: Loss(train): 0.054918 Loss(val): 0.054874 [03:09:24] Epoch 22: Loss(train): 0.054555 Loss(val): 0.054594 [03:38:37] Epoch 24: Loss(train): 0.054447 Loss(val): 0.054464 [04:07:54] Epoch 26: Loss(train): 0.054165 Loss(val): 0.054282 [04:34:20] Epoch 28: Loss(train): 0.053983 Loss(val): 0.054111 [05:01:39] Epoch 30: Loss(train): 0.053737 Loss(val): 0.053914 [05:27:48] Epoch 32: Loss(train): 0.053491 Loss(val): 0.053724 [05:55:10] Epoch 34: Loss(train): 0.053195 Loss(val): 0.053495 [06:23:22] Epoch 36: Loss(train): 0.052964 Loss(val): 0.053314 [06:51:28] Epoch 38: Loss(train): 0.052682 Loss(val): 0.053067 [07:19:10] Epoch 40: Loss(train): 0.052387 Loss(val): 0.052851 [07:23:01] FINAL(40) Loss(val): 0.052851 Accuarcy: 0.607602 Search 42 of 500 momentum0.96, features=[32, 64, 128], dropout_rate=0.1 kernel=Tuple{Int64,Int64}[(7, 1), (7, 1), (3, 6)], pooldims=Tuple{Int64,Int64}[(2, 1), (2, 1)], learning_rate=0.001 [07:29:16] INIT Loss(val): 0.176981 Accuarcy: 0.089626 [08:08:38] Epoch 2: Loss(train): 0.084734 Loss(val): 0.084191 [08:51:50] Epoch 4: Loss(train): 0.074645 Loss(val): 0.074708 [09:34:55] Epoch 6: Loss(train): 0.071603 Loss(val): 0.072190 [10:19:23] Epoch 8: Loss(train): 0.070126 Loss(val): 0.070948 [11:06:15] Epoch 10: Loss(train): 0.069323 Loss(val): 0.070290