training_log.txt

sjb@persephone:sjb_word_model$ ./build_dictionary.py
Corpus length in words: 7671252
Distinct words: 48160

Training with other_contributions.json...

sjb@persephone:sjb_word_model$ ./train.py
2017-03-03 09:33:34.909511  iteration = 000, loss = 4.3022, val_loss = 4.3441, training accuracy = 25.98%, test accuracy = 23.84%
2017-03-03 16:10:42.388690  iteration = 001, loss = 4.1311, val_loss = 4.2435, training accuracy = 27.18%, test accuracy = 24.68%
2017-03-03 22:45:11.131193  iteration = 002, loss = 4.0318, val_loss = 4.1936, training accuracy = 27.84%, test accuracy = 24.80%
2017-03-04 05:19:36.081660  iteration = 003, loss = 3.9603, val_loss = 4.1624, training accuracy = 28.58%, test accuracy = 25.10%
2017-03-04 11:53:16.442941  iteration = 004, loss = 3.9041, val_loss = 4.1412, training accuracy = 29.32%, test accuracy = 25.30%
2017-03-04 18:26:18.257271  iteration = 005, loss = 3.8578, val_loss = 4.1256, training accuracy = 29.88%, test accuracy = 25.40%
2017-03-05 00:59:49.743446  iteration = 006, loss = 3.8183, val_loss = 4.1143, training accuracy = 30.30%, test accuracy = 25.58%
2017-03-05 07:33:14.301961  iteration = 007, loss = 3.7837, val_loss = 4.1053, training accuracy = 30.70%, test accuracy = 25.62%
2017-03-05 14:07:55.396894  iteration = 008, loss = 3.7529, val_loss = 4.0987, training accuracy = 31.10%, test accuracy = 25.60%
2017-03-05 20:43:36.155004  iteration = 009, loss = 3.7253, val_loss = 4.0931, training accuracy = 31.40%, test accuracy = 25.78%
2017-03-06 03:17:30.148697  iteration = 010, loss = 3.6999, val_loss = 4.0891, training accuracy = 31.82%, test accuracy = 25.80%
2017-03-06 09:52:04.225548  iteration = 011, loss = 3.6769, val_loss = 4.0859, training accuracy = 32.14%, test accuracy = 25.86%
2017-03-06 16:29:00.474715  iteration = 012, loss = 3.6558, val_loss = 4.0834, training accuracy = 32.20%, test accuracy = 25.90%
2017-03-06 23:05:54.841411  iteration = 013, loss = 3.6357, val_loss = 4.0814, training accuracy = 32.36%, test accuracy = 25.92%
2017-03-07 05:42:47.360574  iteration = 014, loss = 3.6173, val_loss = 4.0803, training accuracy = 32.64%, test accuracy = 25.92%
2017-03-07 12:17:48.156429  iteration = 015, loss = 3.6001, val_loss = 4.0791, training accuracy = 32.78%, test accuracy = 26.14%
2017-03-07 19:01:34.112925  iteration = 016, loss = 3.5845, val_loss = 4.0784, training accuracy = 32.98%, test accuracy = 26.10%
2017-03-08 01:48:46.823677  iteration = 017, loss = 3.5682, val_loss = 4.0781, training accuracy = 33.14%, test accuracy = 26.20%
2017-03-08 08:54:57.774061  iteration = 018, loss = 3.5545, val_loss = 4.0780, training accuracy = 33.32%, test accuracy = 26.28%
2017-03-08 15:57:17.426668  iteration = 019, loss = 3.5412, val_loss = 4.0779, training accuracy = 33.48%, test accuracy = 26.20%
2017-03-08 22:34:00.009954  iteration = 020, loss = 3.5281, val_loss = 4.0781, training accuracy = 33.58%, test accuracy = 26.14%
2017-03-09 05:10:34.406713  iteration = 021, loss = 3.5162, val_loss = 4.0787, training accuracy = 33.70%, test accuracy = 26.32%
2017-03-09 11:46:06.444715  iteration = 022, loss = 3.5041, val_loss = 4.0791, training accuracy = 33.78%, test accuracy = 26.26%

Validation loss starts to increase after iteration 020
Take 020 result forward for further training
Archived in file 1_trained_model_020_33.5800.h5

Training with theresa_may_contributions.json...

sjb@persephone:sjb_word_model$ ./train.py theresa_may_contributions.json
Using TensorFlow backend.
Vectorisation...
n_sentences = 29591
x_D shape: (1063680, 3600)
y_D shape: (1063680, 48160)
train_num_vecs = 744448
val_num_vecs = 319232
Loading existing model and weights...
Calculating initial validation loss...
2017-03-09 14:23:43.384216  iteration = 000, loss = 3.8204, val_loss = 3.7061, training accuracy = 31.44%, test accuracy = 29.30%
2017-03-09 15:06:10.017937  iteration = 001, loss = 3.7094, val_loss = 3.6700, training accuracy = 31.86%, test accuracy = 30.08%
2017-03-09 15:48:34.998866  iteration = 002, loss = 3.6280, val_loss = 3.6490, training accuracy = 32.16%, test accuracy = 30.20%
2017-03-09 16:31:03.823442  iteration = 003, loss = 3.5627, val_loss = 3.6357, training accuracy = 32.46%, test accuracy = 30.30%
2017-03-09 17:13:30.850469  iteration = 004, loss = 3.5059, val_loss = 3.6265, training accuracy = 32.86%, test accuracy = 30.44%
2017-03-09 17:57:30.750933  iteration = 005, loss = 3.4555, val_loss = 3.6200, training accuracy = 33.34%, test accuracy = 30.58%
2017-03-09 18:40:14.918230  iteration = 006, loss = 3.4104, val_loss = 3.6155, training accuracy = 33.74%, test accuracy = 30.58%
2017-03-09 19:24:37.206339  iteration = 007, loss = 3.3688, val_loss = 3.6121, training accuracy = 34.06%, test accuracy = 30.56%
2017-03-09 20:14:57.143454  iteration = 008, loss = 3.3285, val_loss = 3.6105, training accuracy = 34.64%, test accuracy = 30.54%
2017-03-09 20:58:57.095716  iteration = 009, loss = 3.2925, val_loss = 3.6093, training accuracy = 35.12%, test accuracy = 30.60%
2017-03-09 21:46:40.077499  iteration = 010, loss = 3.2592, val_loss = 3.6092, training accuracy = 35.48%, test accuracy = 30.64%
2017-03-09 22:29:03.714321  iteration = 011, loss = 3.2274, val_loss = 3.6097, training accuracy = 35.86%, test accuracy = 30.78%
2017-03-09 23:18:52.879625  iteration = 012, loss = 3.1975, val_loss = 3.6104, training accuracy = 36.22%, test accuracy = 30.76%
2017-03-10 00:03:09.434216  iteration = 013, loss = 3.1672, val_loss = 3.6119, training accuracy = 36.48%, test accuracy = 30.94%
2017-03-10 00:45:33.198204  iteration = 014, loss = 3.1398, val_loss = 3.6136, training accuracy = 36.76%, test accuracy = 30.96%
2017-03-10 01:35:22.687494  iteration = 015, loss = 3.1136, val_loss = 3.6157, training accuracy = 37.24%, test accuracy = 31.04%
2017-03-10 02:19:29.418674  iteration = 016, loss = 3.0893, val_loss = 3.6180, training accuracy = 37.42%, test accuracy = 31.14%
2017-03-10 03:12:59.766789  iteration = 017, loss = 3.0649, val_loss = 3.6208, training accuracy = 37.68%, test accuracy = 31.26%
2017-03-10 03:56:51.254432  iteration = 018, loss = 3.0416, val_loss = 3.6237, training accuracy = 38.16%, test accuracy = 31.14%
2017-03-10 04:41:08.084692  iteration = 019, loss = 3.0189, val_loss = 3.6268, training accuracy = 38.34%, test accuracy = 31.20%
2017-03-10 05:24:34.569100  iteration = 020, loss = 2.9982, val_loss = 3.6303, training accuracy = 38.62%, test accuracy = 31.30%
2017-03-10 06:07:04.545781  iteration = 021, loss = 2.9779, val_loss = 3.6335, training accuracy = 38.86%, test accuracy = 31.36%
2017-03-10 06:53:45.726912  iteration = 022, loss = 2.9562, val_loss = 3.6373, training accuracy = 39.20%, test accuracy = 31.40%
2017-03-10 07:36:08.503499  iteration = 023, loss = 2.9374, val_loss = 3.6411, training accuracy = 39.70%, test accuracy = 31.38%
2017-03-10 08:29:53.374409  iteration = 024, loss = 2.9190, val_loss = 3.6445, training accuracy = 39.80%, test accuracy = 31.38%
2017-03-10 09:12:21.148180  iteration = 025, loss = 2.8995, val_loss = 3.6486, training accuracy = 40.08%, test accuracy = 31.32%
2017-03-10 10:04:12.392313  iteration = 026, loss = 2.8812, val_loss = 3.6527, training accuracy = 40.48%, test accuracy = 31.30%
2017-03-10 11:03:14.729630  iteration = 027, loss = 2.8639, val_loss = 3.6568, training accuracy = 40.74%, test accuracy = 31.28%
2017-03-10 11:53:26.416021  iteration = 028, loss = 2.8474, val_loss = 3.6610, training accuracy = 41.12%, test accuracy = 31.30%
2017-03-10 12:41:16.509741  iteration = 029, loss = 2.8281, val_loss = 3.6651, training accuracy = 41.34%, test accuracy = 31.34%
2017-03-10 13:39:59.364219  iteration = 030, loss = 2.8137, val_loss = 3.6694, training accuracy = 41.60%, test accuracy = 31.26%
2017-03-10 14:24:20.175296  iteration = 031, loss = 2.7985, val_loss = 3.6739, training accuracy = 41.80%, test accuracy = 31.36%
2017-03-10 15:08:10.700458  iteration = 032, loss = 2.7830, val_loss = 3.6781, training accuracy = 42.16%, test accuracy = 31.30%
2017-03-10 16:03:39.818522  iteration = 033, loss = 2.7668, val_loss = 3.6828, training accuracy = 42.40%, test accuracy = 31.24%
2017-03-10 16:48:04.069603  iteration = 034, loss = 2.7529, val_loss = 3.6870, training accuracy = 42.68%, test accuracy = 31.22%
2017-03-10 17:30:38.200220  iteration = 035, loss = 2.7376, val_loss = 3.6919, training accuracy = 42.96%, test accuracy = 31.18%
2017-03-10 18:15:00.245223  iteration = 036, loss = 2.7246, val_loss = 3.6966, training accuracy = 43.16%, test accuracy = 31.18%
2017-03-10 18:58:52.211302  iteration = 037, loss = 2.7092, val_loss = 3.7012, training accuracy = 43.44%, test accuracy = 31.14%
2017-03-10 19:41:27.940696  iteration = 038, loss = 2.6972, val_loss = 3.7051, training accuracy = 43.66%, test accuracy = 31.22%
2017-03-10 20:33:57.728802  iteration = 039, loss = 2.6834, val_loss = 3.7106, training accuracy = 43.92%, test accuracy = 31.18%
2017-03-10 21:18:07.737129  iteration = 040, loss = 2.6704, val_loss = 3.7152, training accuracy = 44.28%, test accuracy = 31.30%
2017-03-10 22:00:41.082111  iteration = 041, loss = 2.6577, val_loss = 3.7200, training accuracy = 44.48%, test accuracy = 31.18%
2017-03-10 22:50:57.394001  iteration = 042, loss = 2.6437, val_loss = 3.7242, training accuracy = 44.82%, test accuracy = 31.18%
2017-03-10 23:35:18.191999  iteration = 043, loss = 2.6326, val_loss = 3.7290, training accuracy = 44.86%, test accuracy = 31.18%
2017-03-11 00:17:54.014918  iteration = 044, loss = 2.6202, val_loss = 3.7341, training accuracy = 45.00%, test accuracy = 31.16%
2017-03-11 01:07:56.979884  iteration = 045, loss = 2.6067, val_loss = 3.7388, training accuracy = 45.50%, test accuracy = 31.12%
2017-03-11 01:52:09.410095  iteration = 046, loss = 2.5962, val_loss = 3.7433, training accuracy = 45.68%, test accuracy = 31.08%
2017-03-11 02:34:43.065693  iteration = 047, loss = 2.5844, val_loss = 3.7487, training accuracy = 45.86%, test accuracy = 31.10%
2017-03-11 03:25:06.974807  iteration = 048, loss = 2.5737, val_loss = 3.7531, training accuracy = 46.14%, test accuracy = 30.98%
2017-03-11 04:09:45.050473  iteration = 049, loss = 2.5620, val_loss = 3.7583, training accuracy = 46.32%, test accuracy = 31.04%
2017-03-11 04:52:22.505345  iteration = 050, loss = 2.5517, val_loss = 3.7631, training accuracy = 46.58%, test accuracy = 31.04%
2017-03-11 05:44:13.620233  iteration = 051, loss = 2.5410, val_loss = 3.7671, training accuracy = 46.68%, test accuracy = 31.06%
2017-03-11 06:31:43.675174  iteration = 052, loss = 2.5301, val_loss = 3.7727, training accuracy = 46.84%, test accuracy = 31.06%
2017-03-11 07:14:21.970689  iteration = 053, loss = 2.5217, val_loss = 3.7769, training accuracy = 47.10%, test accuracy = 31.12%
2017-03-11 08:09:07.409675  iteration = 054, loss = 2.5102, val_loss = 3.7818, training accuracy = 47.24%, test accuracy = 31.04%
2017-03-11 08:56:47.313101  iteration = 055, loss = 2.4978, val_loss = 3.7869, training accuracy = 47.52%, test accuracy = 31.08%
2017-03-11 09:39:23.597022  iteration = 056, loss = 2.4900, val_loss = 3.7904, training accuracy = 47.82%, test accuracy = 31.08%
2017-03-11 10:34:37.458578  iteration = 057, loss = 2.4799, val_loss = 3.7961, training accuracy = 48.12%, test accuracy = 31.10%
2017-03-11 11:18:59.984882  iteration = 058, loss = 2.4712, val_loss = 3.8011, training accuracy = 48.30%, test accuracy = 31.08%
2017-03-11 12:01:35.494865  iteration = 059, loss = 2.4616, val_loss = 3.8052, training accuracy = 48.62%, test accuracy = 31.06%
2017-03-11 12:52:29.746911  iteration = 060, loss = 2.4513, val_loss = 3.8106, training accuracy = 48.66%, test accuracy = 30.96%
2017-03-11 13:37:01.098392  iteration = 061, loss = 2.4418, val_loss = 3.8147, training accuracy = 48.84%, test accuracy = 30.98%
2017-03-11 14:19:37.664053  iteration = 062, loss = 2.4333, val_loss = 3.8196, training accuracy = 49.06%, test accuracy = 30.94%
2017-03-11 15:10:45.350775  iteration = 063, loss = 2.4243, val_loss = 3.8245, training accuracy = 49.40%, test accuracy = 30.94%
2017-03-11 15:55:00.077809  iteration = 064, loss = 2.4153, val_loss = 3.8297, training accuracy = 49.54%, test accuracy = 31.00%
2017-03-11 16:37:32.945550  iteration = 065, loss = 2.4070, val_loss = 3.8340, training accuracy = 49.46%, test accuracy = 31.00%
2017-03-11 17:31:09.943738  iteration = 066, loss = 2.3985, val_loss = 3.8387, training accuracy = 49.70%, test accuracy = 30.98%
2017-03-11 18:24:42.016976  iteration = 067, loss = 2.3899, val_loss = 3.8430, training accuracy = 49.76%, test accuracy = 30.98%
2017-03-11 19:07:45.105062  iteration = 068, loss = 2.3829, val_loss = 3.8475, training accuracy = 49.92%, test accuracy = 30.92%
2017-03-11 20:06:05.102587  iteration = 069, loss = 2.3740, val_loss = 3.8524, training accuracy = 50.44%, test accuracy = 30.98%
2017-03-11 20:59:17.095862  iteration = 070, loss = 2.3665, val_loss = 3.8573, training accuracy = 50.60%, test accuracy = 30.98%
2017-03-11 21:41:58.117878  iteration = 071, loss = 2.3591, val_loss = 3.8618, training accuracy = 50.88%, test accuracy = 30.96%
2017-03-11 22:36:12.176692  iteration = 072, loss = 2.3502, val_loss = 3.8658, training accuracy = 50.78%, test accuracy = 30.96%
2017-03-11 23:20:37.273925  iteration = 073, loss = 2.3418, val_loss = 3.8706, training accuracy = 51.00%, test accuracy = 30.96%
2017-03-12 00:03:13.854602  iteration = 074, loss = 2.3364, val_loss = 3.8744, training accuracy = 51.10%, test accuracy = 30.88%
2017-03-12 00:54:04.945059  iteration = 075, loss = 2.3273, val_loss = 3.8796, training accuracy = 51.26%, test accuracy = 30.92%
2017-03-12 01:39:10.176649  iteration = 076, loss = 2.3204, val_loss = 3.8841, training accuracy = 51.36%, test accuracy = 30.86%
2017-03-12 02:21:41.888668  iteration = 077, loss = 2.3133, val_loss = 3.8889, training accuracy = 51.60%, test accuracy = 30.86%
2017-03-12 03:12:23.744264  iteration = 078, loss = 2.3065, val_loss = 3.8931, training accuracy = 51.92%, test accuracy = 30.84%
2017-03-12 04:00:38.626884  iteration = 079, loss = 2.2992, val_loss = 3.8979, training accuracy = 52.16%, test accuracy = 30.92%
2017-03-12 04:43:12.376572  iteration = 080, loss = 2.2918, val_loss = 3.9023, training accuracy = 52.28%, test accuracy = 30.92%
2017-03-12 05:39:10.839388  iteration = 081, loss = 2.2857, val_loss = 3.9062, training accuracy = 52.34%, test accuracy = 30.94%
2017-03-12 06:23:47.821883  iteration = 082, loss = 2.2770, val_loss = 3.9110, training accuracy = 52.64%, test accuracy = 30.88%
2017-03-12 07:06:21.016257  iteration = 083, loss = 2.2712, val_loss = 3.9154, training accuracy = 52.64%, test accuracy = 30.72%
2017-03-12 07:58:34.435493  iteration = 084, loss = 2.2649, val_loss = 3.9202, training accuracy = 52.76%, test accuracy = 30.86%
2017-03-12 08:42:49.996973  iteration = 085, loss = 2.2584, val_loss = 3.9243, training accuracy = 52.88%, test accuracy = 30.82%
2017-03-12 09:25:23.090575  iteration = 086, loss = 2.2526, val_loss = 3.9281, training accuracy = 53.08%, test accuracy = 30.82%
2017-03-12 10:17:10.350424  iteration = 087, loss = 2.2446, val_loss = 3.9327, training accuracy = 53.18%, test accuracy = 30.90%
2017-03-12 11:01:27.304637  iteration = 088, loss = 2.2382, val_loss = 3.9367, training accuracy = 53.24%, test accuracy = 30.88%
2017-03-12 11:44:03.202525  iteration = 089, loss = 2.2335, val_loss = 3.9408, training accuracy = 53.36%, test accuracy = 30.78%
2017-03-12 12:35:00.227860  iteration = 090, loss = 2.2269, val_loss = 3.9455, training accuracy = 53.52%, test accuracy = 30.86%
2017-03-12 13:20:57.969792  iteration = 091, loss = 2.2202, val_loss = 3.9495, training accuracy = 53.56%, test accuracy = 30.82%
2017-03-12 14:03:33.031509  iteration = 092, loss = 2.2141, val_loss = 3.9540, training accuracy = 53.68%, test accuracy = 30.94%
2017-03-12 14:58:19.618811  iteration = 093, loss = 2.2086, val_loss = 3.9580, training accuracy = 53.80%, test accuracy = 30.86%
2017-03-12 15:47:14.858710  iteration = 094, loss = 2.2038, val_loss = 3.9621, training accuracy = 53.84%, test accuracy = 30.88%
2017-03-12 16:29:45.750942  iteration = 095, loss = 2.1987, val_loss = 3.9652, training accuracy = 54.08%, test accuracy = 30.72%
2017-03-12 17:26:23.160344  iteration = 096, loss = 2.1920, val_loss = 3.9703, training accuracy = 54.32%, test accuracy = 30.80%
2017-03-12 18:10:48.987543  iteration = 097, loss = 2.1867, val_loss = 3.9745, training accuracy = 54.34%, test accuracy = 30.96%
2017-03-12 18:53:21.564751  iteration = 098, loss = 2.1823, val_loss = 3.9778, training accuracy = 54.46%, test accuracy = 30.88%
2017-03-12 19:46:27.690230  iteration = 099, loss = 2.1762, val_loss = 3.9815, training accuracy = 54.42%, test accuracy = 30.94%
2017-03-12 20:30:44.682547  iteration = 100, loss = 2.1707, val_loss = 3.9863, training accuracy = 54.60%, test accuracy = 30.88%
2017-03-12 21:13:20.537342  iteration = 101, loss = 2.1636, val_loss = 3.9896, training accuracy = 54.70%, test accuracy = 30.86%
2017-03-12 22:04:31.100085  iteration = 102, loss = 2.1603, val_loss = 3.9940, training accuracy = 54.70%, test accuracy = 30.76%
2017-03-12 22:55:52.658777  iteration = 103, loss = 2.1541, val_loss = 3.9968, training accuracy = 55.00%, test accuracy = 30.86%
2017-03-12 23:38:29.285683  iteration = 104, loss = 2.1487, val_loss = 4.0015, training accuracy = 54.98%, test accuracy = 30.90%
2017-03-13 00:34:10.943843  iteration = 105, loss = 2.1438, val_loss = 4.0051, training accuracy = 55.22%, test accuracy = 30.92%
2017-03-13 01:23:00.374178  iteration = 106, loss = 2.1378, val_loss = 4.0099, training accuracy = 55.32%, test accuracy = 30.88%
2017-03-13 02:05:35.806257  iteration = 107, loss = 2.1337, val_loss = 4.0136, training accuracy = 55.44%, test accuracy = 30.84%
2017-03-13 03:00:32.739283  iteration = 108, loss = 2.1281, val_loss = 4.0168, training accuracy = 55.56%, test accuracy = 30.88%
2017-03-13 03:44:53.638815  iteration = 109, loss = 2.1246, val_loss = 4.0209, training accuracy = 55.90%, test accuracy = 30.92%
2017-03-13 04:27:26.721201  iteration = 110, loss = 2.1192, val_loss = 4.0249, training accuracy = 55.76%, test accuracy = 30.86%
2017-03-13 05:17:56.240924  iteration = 111, loss = 2.1150, val_loss = 4.0286, training accuracy = 56.00%, test accuracy = 30.82%
2017-03-13 06:02:14.989620  iteration = 112, loss = 2.1112, val_loss = 4.0325, training accuracy = 56.10%, test accuracy = 30.78%
2017-03-13 06:44:48.716893  iteration = 113, loss = 2.1042, val_loss = 4.0362, training accuracy = 56.28%, test accuracy = 30.74%
2017-03-13 07:36:02.057652  iteration = 114, loss = 2.0989, val_loss = 4.0398, training accuracy = 56.32%, test accuracy = 30.90%
2017-03-13 08:25:47.181280  iteration = 115, loss = 2.0952, val_loss = 4.0437, training accuracy = 56.64%, test accuracy = 30.86%
2017-03-13 09:08:20.071666  iteration = 116, loss = 2.0913, val_loss = 4.0462, training accuracy = 56.58%, test accuracy = 30.92%

Validation loss starts to increase after iteration 010
Archived in file 2_trained_model_010_35.4800.h5

Saved intermediate result at iteration 094
Archived in file 3_trained_model_094_53.8400.h5

PC accidentally unplugged. Restarted training

sjb@persephone:sjb_word_model$ ./train.py theresa_may_contributions.json
2017-03-13 12:02:19.202019  iteration = 000, loss = 2.0873, val_loss = 4.0504, training accuracy = 56.66%, test accuracy = 30.80%
2017-03-13 12:44:48.946873  iteration = 001, loss = 2.0822, val_loss = 4.0542, training accuracy = 56.78%, test accuracy = 30.74%
2017-03-13 13:27:19.057455  iteration = 002, loss = 2.0772, val_loss = 4.0572, training accuracy = 56.90%, test accuracy = 30.76%
2017-03-13 14:09:50.010197  iteration = 003, loss = 2.0738, val_loss = 4.0618, training accuracy = 57.14%, test accuracy = 30.74%
2017-03-13 14:52:19.241862  iteration = 004, loss = 2.0688, val_loss = 4.0653, training accuracy = 57.12%, test accuracy = 30.62%
2017-03-13 15:34:51.889627  iteration = 005, loss = 2.0665, val_loss = 4.0688, training accuracy = 57.28%, test accuracy = 30.58%
2017-03-13 16:17:21.942310  iteration = 006, loss = 2.0612, val_loss = 4.0716, training accuracy = 57.50%, test accuracy = 30.46%
2017-03-13 16:59:51.939691  iteration = 007, loss = 2.0574, val_loss = 4.0755, training accuracy = 57.52%, test accuracy = 30.58%
2017-03-13 17:42:28.302276  iteration = 008, loss = 2.0534, val_loss = 4.0789, training accuracy = 57.68%, test accuracy = 30.70%
2017-03-13 18:25:01.376031  iteration = 009, loss = 2.0482, val_loss = 4.0822, training accuracy = 57.80%, test accuracy = 30.62%
2017-03-13 19:07:30.547743  iteration = 010, loss = 2.0447, val_loss = 4.0865, training accuracy = 57.76%, test accuracy = 30.70%
2017-03-13 19:50:00.198001  iteration = 011, loss = 2.0407, val_loss = 4.0902, training accuracy = 57.86%, test accuracy = 30.58%
2017-03-13 20:32:30.436284  iteration = 012, loss = 2.0363, val_loss = 4.0929, training accuracy = 58.00%, test accuracy = 30.58%
2017-03-13 21:15:06.416320  iteration = 013, loss = 2.0335, val_loss = 4.0962, training accuracy = 58.04%, test accuracy = 30.54%
2017-03-13 21:57:32.766322  iteration = 014, loss = 2.0284, val_loss = 4.0995, training accuracy = 58.08%, test accuracy = 30.58%
2017-03-13 22:40:01.340139  iteration = 015, loss = 2.0257, val_loss = 4.1031, training accuracy = 58.26%, test accuracy = 30.50%
2017-03-13 23:22:30.784463  iteration = 016, loss = 2.0209, val_loss = 4.1057, training accuracy = 58.40%, test accuracy = 30.56%
2017-03-14 00:05:02.877517  iteration = 017, loss = 2.0180, val_loss = 4.1101, training accuracy = 58.60%, test accuracy = 30.50%
2017-03-14 00:47:36.015074  iteration = 018, loss = 2.0135, val_loss = 4.1131, training accuracy = 58.66%, test accuracy = 30.44%
2017-03-14 01:30:05.350901  iteration = 019, loss = 2.0102, val_loss = 4.1162, training accuracy = 58.80%, test accuracy = 30.44%
2017-03-14 02:12:35.458251  iteration = 020, loss = 2.0075, val_loss = 4.1188, training accuracy = 58.72%, test accuracy = 30.46%
2017-03-14 02:55:08.855988  iteration = 021, loss = 2.0031, val_loss = 4.1229, training accuracy = 58.76%, test accuracy = 30.52%
2017-03-14 03:37:40.456761  iteration = 022, loss = 2.0008, val_loss = 4.1259, training accuracy = 58.82%, test accuracy = 30.38%
2017-03-14 04:20:11.249789  iteration = 023, loss = 1.9956, val_loss = 4.1296, training accuracy = 59.00%, test accuracy = 30.58%
2017-03-14 05:02:45.515442  iteration = 024, loss = 1.9927, val_loss = 4.1320, training accuracy = 59.12%, test accuracy = 30.44%
2017-03-14 05:45:16.442687  iteration = 025, loss = 1.9888, val_loss = 4.1357, training accuracy = 59.24%, test accuracy = 30.38%
2017-03-14 06:27:45.712426  iteration = 026, loss = 1.9856, val_loss = 4.1388, training accuracy = 59.26%, test accuracy = 30.26%
2017-03-14 07:10:17.817489  iteration = 027, loss = 1.9820, val_loss = 4.1417, training accuracy = 59.42%, test accuracy = 30.38%
2017-03-14 07:52:44.968567  iteration = 028, loss = 1.9790, val_loss = 4.1444, training accuracy = 59.60%, test accuracy = 30.32%
2017-03-14 08:35:16.040024  iteration = 029, loss = 1.9730, val_loss = 4.1490, training accuracy = 59.56%, test accuracy = 30.40%
2017-03-14 09:17:47.698577  iteration = 030, loss = 1.9713, val_loss = 4.1513, training accuracy = 59.62%, test accuracy = 30.28%
2017-03-14 10:00:20.803326  iteration = 031, loss = 1.9674, val_loss = 4.1544, training accuracy = 59.72%, test accuracy = 30.24%
2017-03-14 10:42:55.189755  iteration = 032, loss = 1.9664, val_loss = 4.1569, training accuracy = 59.68%, test accuracy = 30.24%
2017-03-14 11:25:31.436218  iteration = 033, loss = 1.9612, val_loss = 4.1608, training accuracy = 59.72%, test accuracy = 30.42%
2017-03-14 12:08:00.422205  iteration = 034, loss = 1.9586, val_loss = 4.1633, training accuracy = 59.86%, test accuracy = 30.28%
2017-03-14 12:50:36.166541  iteration = 035, loss = 1.9541, val_loss = 4.1656, training accuracy = 59.94%, test accuracy = 30.30%
2017-03-14 13:33:04.514182  iteration = 036, loss = 1.9515, val_loss = 4.1685, training accuracy = 60.00%, test accuracy = 30.38%
2017-03-14 14:15:36.190233  iteration = 037, loss = 1.9483, val_loss = 4.1725, training accuracy = 60.12%, test accuracy = 30.22%
2017-03-14 14:58:05.222629  iteration = 038, loss = 1.9445, val_loss = 4.1758, training accuracy = 60.08%, test accuracy = 30.34%
2017-03-14 15:40:38.372385  iteration = 039, loss = 1.9421, val_loss = 4.1783, training accuracy = 60.14%, test accuracy = 30.26%
2017-03-14 16:23:12.033912  iteration = 040, loss = 1.9404, val_loss = 4.1814, training accuracy = 60.22%, test accuracy = 30.20%
2017-03-14 17:05:44.365954  iteration = 041, loss = 1.9358, val_loss = 4.1840, training accuracy = 60.32%, test accuracy = 30.24%
2017-03-14 17:48:15.479580  iteration = 042, loss = 1.9324, val_loss = 4.1870, training accuracy = 60.26%, test accuracy = 30.16%
2017-03-14 18:30:44.091998  iteration = 043, loss = 1.9307, val_loss = 4.1900, training accuracy = 60.48%, test accuracy = 30.20%
2017-03-14 19:13:15.214850  iteration = 044, loss = 1.9258, val_loss = 4.1933, training accuracy = 60.48%, test accuracy = 30.20%
2017-03-14 19:55:48.977310  iteration = 045, loss = 1.9242, val_loss = 4.1962, training accuracy = 60.50%, test accuracy = 30.00%
2017-03-14 20:38:21.750096  iteration = 046, loss = 1.9205, val_loss = 4.1985, training accuracy = 60.68%, test accuracy = 30.08%
2017-03-14 21:20:55.314551  iteration = 047, loss = 1.9190, val_loss = 4.2019, training accuracy = 60.74%, test accuracy = 30.12%
2017-03-14 22:03:26.718286  iteration = 048, loss = 1.9157, val_loss = 4.2047, training accuracy = 60.74%, test accuracy = 30.06%
2017-03-14 22:45:59.376021  iteration = 049, loss = 1.9122, val_loss = 4.2079, training accuracy = 60.76%, test accuracy = 30.10%
2017-03-14 23:28:31.233240  iteration = 050, loss = 1.9096, val_loss = 4.2104, training accuracy = 60.96%, test accuracy = 30.06%
2017-03-15 00:11:03.858357  iteration = 051, loss = 1.9064, val_loss = 4.2131, training accuracy = 60.80%, test accuracy = 30.14%
2017-03-15 00:53:34.129885  iteration = 052, loss = 1.9048, val_loss = 4.2160, training accuracy = 61.02%, test accuracy = 30.12%
2017-03-15 01:36:08.881765  iteration = 053, loss = 1.9007, val_loss = 4.2187, training accuracy = 60.94%, test accuracy = 30.22%
2017-03-15 02:18:41.425684  iteration = 054, loss = 1.8997, val_loss = 4.2214, training accuracy = 61.14%, test accuracy = 30.22%
2017-03-15 03:01:14.231824  iteration = 055, loss = 1.8961, val_loss = 4.2246, training accuracy = 61.06%, test accuracy = 30.00%
2017-03-15 03:43:47.664980  iteration = 056, loss = 1.8925, val_loss = 4.2269, training accuracy = 61.20%, test accuracy = 29.98%
2017-03-15 04:26:27.082931  iteration = 057, loss = 1.8894, val_loss = 4.2294, training accuracy = 61.26%, test accuracy = 30.00%
2017-03-15 05:09:11.836697  iteration = 058, loss = 1.8888, val_loss = 4.2328, training accuracy = 61.34%, test accuracy = 29.98%
2017-03-15 05:51:55.423608  iteration = 059, loss = 1.8851, val_loss = 4.2346, training accuracy = 61.38%, test accuracy = 30.08%
2017-03-15 06:34:27.624867  iteration = 060, loss = 1.8831, val_loss = 4.2382, training accuracy = 61.54%, test accuracy = 30.08%
2017-03-15 07:17:11.735224  iteration = 061, loss = 1.8787, val_loss = 4.2407, training accuracy = 61.56%, test accuracy = 30.00%
2017-03-15 07:59:47.111602  iteration = 062, loss = 1.8765, val_loss = 4.2434, training accuracy = 61.58%, test accuracy = 30.08%
2017-03-15 08:42:36.683265  iteration = 063, loss = 1.8747, val_loss = 4.2457, training accuracy = 61.74%, test accuracy = 29.98%
2017-03-15 09:25:22.796161  iteration = 064, loss = 1.8708, val_loss = 4.2481, training accuracy = 61.66%, test accuracy = 30.06%
2017-03-15 10:07:58.278205  iteration = 065, loss = 1.8677, val_loss = 4.2511, training accuracy = 61.82%, test accuracy = 29.86%
2017-03-15 10:50:46.981485  iteration = 066, loss = 1.8667, val_loss = 4.2533, training accuracy = 61.92%, test accuracy = 29.84%
2017-03-15 11:33:33.693203  iteration = 067, loss = 1.8653, val_loss = 4.2556, training accuracy = 61.92%, test accuracy = 29.88%
2017-03-15 12:16:13.578770  iteration = 068, loss = 1.8627, val_loss = 4.2589, training accuracy = 62.12%, test accuracy = 29.90%
2017-03-15 12:59:16.916691  iteration = 069, loss = 1.8580, val_loss = 4.2607, training accuracy = 62.10%, test accuracy = 30.04%
2017-03-15 13:43:24.738728  iteration = 070, loss = 1.8553, val_loss = 4.2641, training accuracy = 62.16%, test accuracy = 30.06%
2017-03-15 14:27:25.259905  iteration = 071, loss = 1.8541, val_loss = 4.2666, training accuracy = 62.08%, test accuracy = 30.02%
2017-03-15 15:11:03.918011  iteration = 072, loss = 1.8507, val_loss = 4.2689, training accuracy = 62.28%, test accuracy = 30.12%
2017-03-15 15:54:29.936567  iteration = 073, loss = 1.8499, val_loss = 4.2710, training accuracy = 62.40%, test accuracy = 30.08%
2017-03-15 16:37:46.918378  iteration = 074, loss = 1.8474, val_loss = 4.2744, training accuracy = 62.44%, test accuracy = 30.06%
2017-03-15 17:20:53.374061  iteration = 075, loss = 1.8441, val_loss = 4.2764, training accuracy = 62.46%, test accuracy = 30.02%
2017-03-15 18:03:50.116764  iteration = 076, loss = 1.8420, val_loss = 4.2790, training accuracy = 62.48%, test accuracy = 30.06%
2017-03-15 18:46:52.560518  iteration = 077, loss = 1.8405, val_loss = 4.2827, training accuracy = 62.62%, test accuracy = 30.08%
2017-03-15 19:29:51.167288  iteration = 078, loss = 1.8377, val_loss = 4.2842, training accuracy = 62.58%, test accuracy = 29.92%
2017-03-15 20:12:58.358484  iteration = 079, loss = 1.8352, val_loss = 4.2861, training accuracy = 62.56%, test accuracy = 30.06%
2017-03-15 20:55:47.315369  iteration = 080, loss = 1.8328, val_loss = 4.2899, training accuracy = 62.60%, test accuracy = 30.04%
2017-03-15 21:38:39.139355  iteration = 081, loss = 1.8312, val_loss = 4.2915, training accuracy = 62.78%, test accuracy = 30.00%
2017-03-15 22:21:16.535459  iteration = 082, loss = 1.8283, val_loss = 4.2938, training accuracy = 63.02%, test accuracy = 30.04%
2017-03-15 23:05:23.376769  iteration = 083, loss = 1.8261, val_loss = 4.2952, training accuracy = 62.86%, test accuracy = 30.04%
2017-03-15 23:48:13.616784  iteration = 084, loss = 1.8222, val_loss = 4.2984, training accuracy = 62.92%, test accuracy = 29.82%
2017-03-16 00:31:05.471467  iteration = 085, loss = 1.8229, val_loss = 4.3005, training accuracy = 62.98%, test accuracy = 29.86%
2017-03-16 01:13:49.830044  iteration = 086, loss = 1.8178, val_loss = 4.3038, training accuracy = 62.92%, test accuracy = 30.04%
2017-03-16 01:57:45.381924  iteration = 087, loss = 1.8148, val_loss = 4.3063, training accuracy = 63.06%, test accuracy = 29.94%
2017-03-16 02:41:43.148718  iteration = 088, loss = 1.8148, val_loss = 4.3088, training accuracy = 63.22%, test accuracy = 29.88%
2017-03-16 03:25:19.606644  iteration = 089, loss = 1.8139, val_loss = 4.3107, training accuracy = 63.12%, test accuracy = 29.80%
2017-03-16 04:08:40.696677  iteration = 090, loss = 1.8097, val_loss = 4.3142, training accuracy = 63.34%, test accuracy = 29.76%
2017-03-16 04:52:07.994758  iteration = 091, loss = 1.8075, val_loss = 4.3155, training accuracy = 63.50%, test accuracy = 29.86%
2017-03-16 05:35:13.754140  iteration = 092, loss = 1.8066, val_loss = 4.3178, training accuracy = 63.56%, test accuracy = 29.86%
2017-03-16 06:18:22.520419  iteration = 093, loss = 1.8027, val_loss = 4.3194, training accuracy = 63.62%, test accuracy = 29.78%
2017-03-16 07:01:37.430915  iteration = 094, loss = 1.7997, val_loss = 4.3232, training accuracy = 63.56%, test accuracy = 29.84%
2017-03-16 07:44:38.928838  iteration = 095, loss = 1.7988, val_loss = 4.3247, training accuracy = 63.66%, test accuracy = 29.76%
2017-03-16 08:27:34.748843  iteration = 096, loss = 1.7980, val_loss = 4.3270, training accuracy = 63.64%, test accuracy = 29.84%
2017-03-16 09:10:24.312971  iteration = 097, loss = 1.7959, val_loss = 4.3286, training accuracy = 63.72%, test accuracy = 29.84%
2017-03-16 09:53:10.404626  iteration = 098, loss = 1.7960, val_loss = 4.3315, training accuracy = 63.86%, test accuracy = 29.86%
2017-03-16 10:37:03.360802  iteration = 099, loss = 1.7905, val_loss = 4.3340, training accuracy = 64.00%, test accuracy = 29.76%
2017-03-16 11:20:53.057416  iteration = 100, loss = 1.7897, val_loss = 4.3356, training accuracy = 63.94%, test accuracy = 29.80%
2017-03-16 12:04:27.744902  iteration = 101, loss = 1.7870, val_loss = 4.3393, training accuracy = 63.98%, test accuracy = 29.82%
2017-03-16 12:47:53.958836  iteration = 102, loss = 1.7849, val_loss = 4.3409, training accuracy = 64.00%, test accuracy = 29.84%
2017-03-16 13:31:05.738289  iteration = 103, loss = 1.7827, val_loss = 4.3425, training accuracy = 63.92%, test accuracy = 29.84%
2017-03-16 14:14:17.428551  iteration = 104, loss = 1.7813, val_loss = 4.3454, training accuracy = 64.18%, test accuracy = 29.84%
2017-03-16 14:57:15.466527  iteration = 105, loss = 1.7806, val_loss = 4.3467, training accuracy = 64.04%, test accuracy = 29.86%
2017-03-16 15:39:58.150125  iteration = 106, loss = 1.7779, val_loss = 4.3495, training accuracy = 64.16%, test accuracy = 29.80%
2017-03-16 16:22:54.937418  iteration = 107, loss = 1.7758, val_loss = 4.3514, training accuracy = 64.38%, test accuracy = 29.76%
2017-03-16 17:05:26.434698  iteration = 108, loss = 1.7729, val_loss = 4.3538, training accuracy = 64.50%, test accuracy = 29.74%
2017-03-16 17:49:40.764638  iteration = 109, loss = 1.7705, val_loss = 4.3558, training accuracy = 64.40%, test accuracy = 29.76%
2017-03-16 18:33:36.557099  iteration = 110, loss = 1.7701, val_loss = 4.3578, training accuracy = 64.52%, test accuracy = 29.72%
2017-03-16 19:17:01.141352  iteration = 111, loss = 1.7674, val_loss = 4.3597, training accuracy = 64.74%, test accuracy = 29.70%
2017-03-16 20:00:13.363922  iteration = 112, loss = 1.7659, val_loss = 4.3624, training accuracy = 64.68%, test accuracy = 29.74%
2017-03-16 20:43:21.811332  iteration = 113, loss = 1.7624, val_loss = 4.3649, training accuracy = 64.70%, test accuracy = 29.70%
2017-03-16 21:26:36.263038  iteration = 114, loss = 1.7618, val_loss = 4.3653, training accuracy = 64.72%, test accuracy = 29.82%
2017-03-16 22:09:25.298825  iteration = 115, loss = 1.7603, val_loss = 4.3688, training accuracy = 64.76%, test accuracy = 29.60%
2017-03-16 22:52:07.051348  iteration = 116, loss = 1.7573, val_loss = 4.3707, training accuracy = 64.68%, test accuracy = 29.78%
2017-03-16 23:35:14.518885  iteration = 117, loss = 1.7561, val_loss = 4.3724, training accuracy = 64.94%, test accuracy = 29.62%
2017-03-17 00:19:11.636829  iteration = 118, loss = 1.7545, val_loss = 4.3758, training accuracy = 64.84%, test accuracy = 29.64%
2017-03-17 01:03:07.897166  iteration = 119, loss = 1.7536, val_loss = 4.3768, training accuracy = 64.80%, test accuracy = 29.66%
2017-03-17 01:46:31.406891  iteration = 120, loss = 1.7520, val_loss = 4.3791, training accuracy = 64.84%, test accuracy = 29.68%
2017-03-17 02:29:40.191431  iteration = 121, loss = 1.7474, val_loss = 4.3820, training accuracy = 64.78%, test accuracy = 29.74%
2017-03-17 03:12:47.845738  iteration = 122, loss = 1.7474, val_loss = 4.3843, training accuracy = 65.12%, test accuracy = 29.58%
2017-03-17 03:55:52.693106  iteration = 123, loss = 1.7442, val_loss = 4.3861, training accuracy = 65.18%, test accuracy = 29.54%
2017-03-17 04:38:43.900299  iteration = 124, loss = 1.7442, val_loss = 4.3875, training accuracy = 65.30%, test accuracy = 29.68%
2017-03-17 05:21:37.084716  iteration = 125, loss = 1.7415, val_loss = 4.3900, training accuracy = 65.30%, test accuracy = 29.50%
2017-03-17 06:05:08.359554  iteration = 126, loss = 1.7407, val_loss = 4.3918, training accuracy = 65.22%, test accuracy = 29.66%
2017-03-17 06:49:04.211616  iteration = 127, loss = 1.7365, val_loss = 4.3960, training accuracy = 65.46%, test accuracy = 29.58%
2017-03-17 07:32:44.280706  iteration = 128, loss = 1.7357, val_loss = 4.3961, training accuracy = 65.54%, test accuracy = 29.54%
2017-03-17 08:16:07.261268  iteration = 129, loss = 1.7349, val_loss = 4.3985, training accuracy = 65.58%, test accuracy = 29.56%
2017-03-17 08:59:14.212745  iteration = 130, loss = 1.7321, val_loss = 4.4009, training accuracy = 65.52%, test accuracy = 29.60%
2017-03-17 09:42:30.472248  iteration = 131, loss = 1.7317, val_loss = 4.4015, training accuracy = 65.66%, test accuracy = 29.64%
2017-03-17 10:25:16.399750  iteration = 132, loss = 1.7293, val_loss = 4.4041, training accuracy = 65.58%, test accuracy = 29.58%
2017-03-17 11:07:59.037180  iteration = 133, loss = 1.7264, val_loss = 4.4057, training accuracy = 65.68%, test accuracy = 29.52%
2017-03-17 11:51:54.483122  iteration = 134, loss = 1.7258, val_loss = 4.4085, training accuracy = 65.70%, test accuracy = 29.56%
2017-03-17 12:34:50.827239  iteration = 135, loss = 1.7239, val_loss = 4.4095, training accuracy = 65.78%, test accuracy = 29.70%
2017-03-17 13:19:03.224188  iteration = 136, loss = 1.7220, val_loss = 4.4129, training accuracy = 65.78%, test accuracy = 29.68%
2017-03-17 14:04:57.901602  iteration = 137, loss = 1.7220, val_loss = 4.4138, training accuracy = 65.92%, test accuracy = 29.56%
2017-03-17 14:48:34.804310  iteration = 138, loss = 1.7194, val_loss = 4.4154, training accuracy = 65.94%, test accuracy = 29.46%
2017-03-17 15:32:01.555241  iteration = 139, loss = 1.7176, val_loss = 4.4175, training accuracy = 65.92%, test accuracy = 29.66%
2017-03-17 16:15:22.333139  iteration = 140, loss = 1.7167, val_loss = 4.4191, training accuracy = 66.02%, test accuracy = 29.56%
2017-03-17 16:58:39.865426  iteration = 141, loss = 1.7154, val_loss = 4.4220, training accuracy = 65.98%, test accuracy = 29.60%
2017-03-17 17:41:39.767918  iteration = 142, loss = 1.7137, val_loss = 4.4228, training accuracy = 66.16%, test accuracy = 29.58%
2017-03-17 18:24:16.350803  iteration = 143, loss = 1.7118, val_loss = 4.4257, training accuracy = 66.28%, test accuracy = 29.50%
2017-03-17 19:07:11.122943  iteration = 144, loss = 1.7111, val_loss = 4.4265, training accuracy = 66.26%, test accuracy = 29.48%
2017-03-17 19:51:01.137308  iteration = 145, loss = 1.7080, val_loss = 4.4287, training accuracy = 66.28%, test accuracy = 29.52%
2017-03-17 20:35:14.432606  iteration = 146, loss = 1.7052, val_loss = 4.4315, training accuracy = 66.42%, test accuracy = 29.60%
2017-03-17 21:19:20.135224  iteration = 147, loss = 1.7027, val_loss = 4.4336, training accuracy = 66.54%, test accuracy = 29.54%
2017-03-17 22:03:01.840511  iteration = 148, loss = 1.7017, val_loss = 4.4346, training accuracy = 66.56%, test accuracy = 29.52%
2017-03-17 22:46:32.503230  iteration = 149, loss = 1.7006, val_loss = 4.4374, training accuracy = 66.52%, test accuracy = 29.46%
2017-03-17 23:29:47.764044  iteration = 150, loss = 1.6992, val_loss = 4.4393, training accuracy = 66.62%, test accuracy = 29.44%
2017-03-18 00:12:54.209095  iteration = 151, loss = 1.6974, val_loss = 4.4406, training accuracy = 66.72%, test accuracy = 29.54%
2017-03-18 00:55:58.882909  iteration = 152, loss = 1.6974, val_loss = 4.4427, training accuracy = 66.64%, test accuracy = 29.54%
2017-03-18 01:39:15.614883  iteration = 153, loss = 1.6934, val_loss = 4.4444, training accuracy = 66.64%, test accuracy = 29.56%
2017-03-18 02:22:29.340652  iteration = 154, loss = 1.6928, val_loss = 4.4473, training accuracy = 66.78%, test accuracy = 29.50%
2017-03-18 03:05:19.835130  iteration = 155, loss = 1.6914, val_loss = 4.4483, training accuracy = 66.92%, test accuracy = 29.44%
2017-03-18 03:48:00.182984  iteration = 156, loss = 1.6890, val_loss = 4.4509, training accuracy = 66.88%, test accuracy = 29.44%
2017-03-18 04:32:06.598785  iteration = 157, loss = 1.6879, val_loss = 4.4529, training accuracy = 67.10%, test accuracy = 29.40%
2017-03-18 05:16:27.805777  iteration = 158, loss = 1.6870, val_loss = 4.4542, training accuracy = 67.04%, test accuracy = 29.44%
2017-03-18 06:00:26.402685  iteration = 159, loss = 1.6854, val_loss = 4.4556, training accuracy = 67.08%, test accuracy = 29.50%
2017-03-18 06:43:58.862475  iteration = 160, loss = 1.6843, val_loss = 4.4579, training accuracy = 67.08%, test accuracy = 29.36%
2017-03-18 07:27:29.103010  iteration = 161, loss = 1.6805, val_loss = 4.4595, training accuracy = 67.24%, test accuracy = 29.46%
2017-03-18 08:10:43.942208  iteration = 162, loss = 1.6818, val_loss = 4.4601, training accuracy = 67.20%, test accuracy = 29.40%
2017-03-18 08:53:52.421700  iteration = 163, loss = 1.6794, val_loss = 4.4629, training accuracy = 67.30%, test accuracy = 29.34%
2017-03-18 09:37:04.426659  iteration = 164, loss = 1.6778, val_loss = 4.4655, training accuracy = 67.36%, test accuracy = 29.38%
2017-03-18 10:20:12.314664  iteration = 165, loss = 1.6742, val_loss = 4.4669, training accuracy = 67.30%, test accuracy = 29.24%
2017-03-18 11:03:11.219434  iteration = 166, loss = 1.6755, val_loss = 4.4687, training accuracy = 67.34%, test accuracy = 29.36%
2017-03-18 11:46:04.432585  iteration = 167, loss = 1.6735, val_loss = 4.4706, training accuracy = 67.38%, test accuracy = 29.36%
2017-03-18 12:28:51.033537  iteration = 168, loss = 1.6715, val_loss = 4.4714, training accuracy = 67.44%, test accuracy = 29.32%
2017-03-18 13:11:47.213516  iteration = 169, loss = 1.6694, val_loss = 4.4739, training accuracy = 67.52%, test accuracy = 29.38%
2017-03-18 13:54:26.470681  iteration = 170, loss = 1.6679, val_loss = 4.4751, training accuracy = 67.40%, test accuracy = 29.36%
2017-03-18 14:38:30.063640  iteration = 171, loss = 1.6670, val_loss = 4.4777, training accuracy = 67.46%, test accuracy = 29.32%
2017-03-18 15:28:56.497318  iteration = 172, loss = 1.6678, val_loss = 4.4776, training accuracy = 67.46%, test accuracy = 29.34%
2017-03-18 16:12:56.919706  iteration = 173, loss = 1.6638, val_loss = 4.4807, training accuracy = 67.54%, test accuracy = 29.32%
2017-03-18 16:56:42.589504  iteration = 174, loss = 1.6620, val_loss = 4.4826, training accuracy = 67.54%, test accuracy = 29.34%
2017-03-18 17:40:04.032507  iteration = 175, loss = 1.6618, val_loss = 4.4840, training accuracy = 67.56%, test accuracy = 29.34%
2017-03-18 18:23:15.431259  iteration = 176, loss = 1.6600, val_loss = 4.4863, training accuracy = 67.60%, test accuracy = 29.34%
2017-03-18 19:06:21.768740  iteration = 177, loss = 1.6596, val_loss = 4.4869, training accuracy = 67.72%, test accuracy = 29.48%
2017-03-18 19:48:55.307468  iteration = 178, loss = 1.6557, val_loss = 4.4898, training accuracy = 67.82%, test accuracy = 29.38%
2017-03-18 20:31:28.340484  iteration = 179, loss = 1.6564, val_loss = 4.4905, training accuracy = 67.90%, test accuracy = 29.26%
2017-03-18 21:14:00.465422  iteration = 180, loss = 1.6538, val_loss = 4.4912, training accuracy = 67.96%, test accuracy = 29.26%
2017-03-18 21:56:32.043658  iteration = 181, loss = 1.6521, val_loss = 4.4946, training accuracy = 67.98%, test accuracy = 29.30%
2017-03-18 22:39:57.495140  iteration = 182, loss = 1.6534, val_loss = 4.4958, training accuracy = 67.96%, test accuracy = 29.34%
2017-03-18 23:23:55.696670  iteration = 183, loss = 1.6491, val_loss = 4.4981, training accuracy = 68.00%, test accuracy = 29.28%
2017-03-19 00:07:32.301229  iteration = 184, loss = 1.6477, val_loss = 4.4995, training accuracy = 68.00%, test accuracy = 29.16%
2017-03-19 00:51:07.391676  iteration = 185, loss = 1.6486, val_loss = 4.5017, training accuracy = 68.14%, test accuracy = 29.24%
2017-03-19 01:34:37.442270  iteration = 186, loss = 1.6457, val_loss = 4.5028, training accuracy = 68.12%, test accuracy = 29.38%
2017-03-19 02:17:46.155090  iteration = 187, loss = 1.6441, val_loss = 4.5052, training accuracy = 68.32%, test accuracy = 29.18%
2017-03-19 03:00:45.061888  iteration = 188, loss = 1.6429, val_loss = 4.5070, training accuracy = 68.24%, test accuracy = 29.24%
2017-03-19 03:43:46.861482  iteration = 189, loss = 1.6422, val_loss = 4.5080, training accuracy = 68.32%, test accuracy = 29.26%
2017-03-19 04:26:44.587226  iteration = 190, loss = 1.6416, val_loss = 4.5106, training accuracy = 68.48%, test accuracy = 29.26%
2017-03-19 05:09:29.230083  iteration = 191, loss = 1.6389, val_loss = 4.5107, training accuracy = 68.46%, test accuracy = 29.22%
2017-03-19 05:52:21.352227  iteration = 192, loss = 1.6366, val_loss = 4.5132, training accuracy = 68.52%, test accuracy = 29.26%
2017-03-19 06:36:16.788328  iteration = 193, loss = 1.6366, val_loss = 4.5142, training accuracy = 68.52%, test accuracy = 29.24%
2017-03-19 07:19:07.289862  iteration = 194, loss = 1.6360, val_loss = 4.5167, training accuracy = 68.54%, test accuracy = 29.26%
2017-03-19 08:01:57.411055  iteration = 195, loss = 1.6319, val_loss = 4.5171, training accuracy = 68.62%, test accuracy = 29.20%
2017-03-19 08:44:56.301581  iteration = 196, loss = 1.6338, val_loss = 4.5181, training accuracy = 68.64%, test accuracy = 29.30%
2017-03-19 09:35:52.215767  iteration = 197, loss = 1.6314, val_loss = 4.5209, training accuracy = 68.54%, test accuracy = 29.26%
2017-03-19 10:19:48.401595  iteration = 198, loss = 1.6300, val_loss = 4.5211, training accuracy = 68.62%, test accuracy = 29.24%
2017-03-19 11:02:40.394427  iteration = 199, loss = 1.6306, val_loss = 4.5249, training accuracy = 68.70%, test accuracy = 29.26%

Continuing training with a further run
sjb@persephone:sjb_word_model$ ./train.py theresa_may_contributions.json
2017-03-29 14:56:43.633711  iteration = 000, loss = 1.6272, val_loss = 4.5248, training accuracy = 68.84%, test accuracy = 29.22%
2017-03-29 15:37:57.808422  iteration = 001, loss = 1.6280, val_loss = 4.5270, training accuracy = 68.82%, test accuracy = 29.26%
2017-03-29 16:18:00.350968  iteration = 002, loss = 1.6230, val_loss = 4.5286, training accuracy = 69.00%, test accuracy = 29.24%
2017-03-29 16:59:14.410998  iteration = 003, loss = 1.6231, val_loss = 4.5298, training accuracy = 69.06%, test accuracy = 29.20%
2017-03-29 17:39:52.716374  iteration = 004, loss = 1.6210, val_loss = 4.5320, training accuracy = 69.00%, test accuracy = 29.28%
2017-03-29 18:22:11.406566  iteration = 005, loss = 1.6209, val_loss = 4.5340, training accuracy = 68.98%, test accuracy = 29.26%
2017-03-29 19:04:31.793923  iteration = 006, loss = 1.6194, val_loss = 4.5364, training accuracy = 68.98%, test accuracy = 29.18%
2017-03-29 19:46:56.971006  iteration = 007, loss = 1.6165, val_loss = 4.5376, training accuracy = 69.04%, test accuracy = 29.22%
2017-03-29 20:29:17.243079  iteration = 008, loss = 1.6170, val_loss = 4.5381, training accuracy = 69.12%, test accuracy = 29.30%
2017-03-29 21:11:39.897375  iteration = 009, loss = 1.6152, val_loss = 4.5397, training accuracy = 69.08%, test accuracy = 29.22%
2017-03-29 21:53:58.960031  iteration = 010, loss = 1.6145, val_loss = 4.5420, training accuracy = 69.16%, test accuracy = 29.22%
2017-03-29 22:36:18.911678  iteration = 011, loss = 1.6124, val_loss = 4.5438, training accuracy = 69.26%, test accuracy = 29.26%
2017-03-29 23:18:37.901239  iteration = 012, loss = 1.6107, val_loss = 4.5449, training accuracy = 69.42%, test accuracy = 29.22%
2017-03-30 00:00:59.034587  iteration = 013, loss = 1.6117, val_loss = 4.5466, training accuracy = 69.32%, test accuracy = 29.28%
2017-03-30 00:43:19.507631  iteration = 014, loss = 1.6113, val_loss = 4.5482, training accuracy = 69.16%, test accuracy = 29.34%
2017-03-30 01:25:39.898450  iteration = 015, loss = 1.6099, val_loss = 4.5485, training accuracy = 69.40%, test accuracy = 29.32%
2017-03-30 02:07:59.095044  iteration = 016, loss = 1.6082, val_loss = 4.5504, training accuracy = 69.24%, test accuracy = 29.24%
2017-03-30 02:50:18.746073  iteration = 017, loss = 1.6057, val_loss = 4.5519, training accuracy = 69.46%, test accuracy = 29.22%
2017-03-30 03:32:41.740836  iteration = 018, loss = 1.6035, val_loss = 4.5545, training accuracy = 69.60%, test accuracy = 29.24%
2017-03-30 04:15:00.594644  iteration = 019, loss = 1.6023, val_loss = 4.5551, training accuracy = 69.40%, test accuracy = 29.28%
2017-03-30 04:57:22.150770  iteration = 020, loss = 1.6043, val_loss = 4.5557, training accuracy = 69.54%, test accuracy = 29.32%
2017-03-30 05:39:49.218507  iteration = 021, loss = 1.6008, val_loss = 4.5590, training accuracy = 69.66%, test accuracy = 29.40%
2017-03-30 06:22:08.839561  iteration = 022, loss = 1.5999, val_loss = 4.5592, training accuracy = 69.62%, test accuracy = 29.34%
2017-03-30 07:04:26.836051  iteration = 023, loss = 1.5984, val_loss = 4.5613, training accuracy = 69.66%, test accuracy = 29.28%
2017-03-30 07:46:47.711538  iteration = 024, loss = 1.5967, val_loss = 4.5630, training accuracy = 69.70%, test accuracy = 29.36%
2017-03-30 08:29:08.072011  iteration = 025, loss = 1.5956, val_loss = 4.5643, training accuracy = 69.76%, test accuracy = 29.38%
2017-03-30 09:11:33.028590  iteration = 026, loss = 1.5949, val_loss = 4.5657, training accuracy = 69.78%, test accuracy = 29.24%
2017-03-30 09:53:50.429677  iteration = 027, loss = 1.5962, val_loss = 4.5668, training accuracy = 69.64%, test accuracy = 29.38%
2017-03-30 10:36:10.396081  iteration = 028, loss = 1.5923, val_loss = 4.5688, training accuracy = 69.86%, test accuracy = 29.32%
2017-03-30 11:18:29.151028  iteration = 029, loss = 1.5925, val_loss = 4.5702, training accuracy = 69.74%, test accuracy = 29.22%
2017-03-30 12:00:13.947371  iteration = 030, loss = 1.5909, val_loss = 4.5702, training accuracy = 69.80%, test accuracy = 29.20%
2017-03-30 12:41:23.144432  iteration = 031, loss = 1.5902, val_loss = 4.5739, training accuracy = 69.90%, test accuracy = 29.34%
2017-03-30 13:21:57.599909  iteration = 032, loss = 1.5869, val_loss = 4.5755, training accuracy = 69.88%, test accuracy = 29.34%
2017-03-30 14:00:56.176677  iteration = 033, loss = 1.5881, val_loss = 4.5763, training accuracy = 69.96%, test accuracy = 29.24%
2017-03-30 14:42:15.214341  iteration = 034, loss = 1.5876, val_loss = 4.5771, training accuracy = 69.88%, test accuracy = 29.24%
2017-03-30 15:24:34.791291  iteration = 035, loss = 1.5843, val_loss = 4.5779, training accuracy = 70.02%, test accuracy = 29.20%
2017-03-30 16:06:55.049567  iteration = 036, loss = 1.5842, val_loss = 4.5806, training accuracy = 70.02%, test accuracy = 29.18%
2017-03-30 16:49:17.064358  iteration = 037, loss = 1.5808, val_loss = 4.5827, training accuracy = 70.04%, test accuracy = 29.16%
2017-03-30 17:31:42.443851  iteration = 038, loss = 1.5797, val_loss = 4.5831, training accuracy = 70.06%, test accuracy = 29.20%
2017-03-30 18:14:17.038062  iteration = 039, loss = 1.5791, val_loss = 4.5850, training accuracy = 70.18%, test accuracy = 29.26%
2017-03-30 18:57:23.924827  iteration = 040, loss = 1.5779, val_loss = 4.5871, training accuracy = 70.16%, test accuracy = 29.24%
2017-03-30 19:40:28.300175  iteration = 041, loss = 1.5778, val_loss = 4.5881, training accuracy = 70.20%, test accuracy = 29.34%
2017-03-30 20:23:18.453125  iteration = 042, loss = 1.5757, val_loss = 4.5905, training accuracy = 70.20%, test accuracy = 29.24%
2017-03-30 21:06:11.234666  iteration = 043, loss = 1.5751, val_loss = 4.5902, training accuracy = 70.10%, test accuracy = 29.18%
2017-03-30 21:48:41.513848  iteration = 044, loss = 1.5737, val_loss = 4.5925, training accuracy = 70.12%, test accuracy = 29.10%
2017-03-30 22:31:13.445784  iteration = 045, loss = 1.5738, val_loss = 4.5942, training accuracy = 70.24%, test accuracy = 29.20%
2017-03-30 23:13:39.386840  iteration = 046, loss = 1.5705, val_loss = 4.5958, training accuracy = 70.18%, test accuracy = 29.12%
2017-03-30 23:57:19.287313  iteration = 047, loss = 1.5716, val_loss = 4.5961, training accuracy = 70.24%, test accuracy = 29.24%
2017-03-31 00:40:36.604654  iteration = 048, loss = 1.5704, val_loss = 4.5973, training accuracy = 70.40%, test accuracy = 29.18%
2017-03-31 01:23:41.301341  iteration = 049, loss = 1.5684, val_loss = 4.5993, training accuracy = 70.38%, test accuracy = 29.16%
2017-03-31 02:06:52.744754  iteration = 050, loss = 1.5664, val_loss = 4.6003, training accuracy = 70.28%, test accuracy = 29.30%
2017-03-31 02:49:51.539745  iteration = 051, loss = 1.5655, val_loss = 4.6025, training accuracy = 70.44%, test accuracy = 29.22%
2017-03-31 03:32:38.971884  iteration = 052, loss = 1.5650, val_loss = 4.6028, training accuracy = 70.42%, test accuracy = 29.24%
2017-03-31 04:15:21.108205  iteration = 053, loss = 1.5627, val_loss = 4.6049, training accuracy = 70.38%, test accuracy = 29.24%
2017-03-31 04:57:56.734753  iteration = 054, loss = 1.5636, val_loss = 4.6052, training accuracy = 70.26%, test accuracy = 29.42%
2017-03-31 05:40:29.735904  iteration = 055, loss = 1.5608, val_loss = 4.6076, training accuracy = 70.40%, test accuracy = 29.26%
2017-03-31 06:23:19.272830  iteration = 056, loss = 1.5610, val_loss = 4.6085, training accuracy = 70.38%, test accuracy = 29.32%
2017-03-31 07:07:04.207868  iteration = 057, loss = 1.5594, val_loss = 4.6093, training accuracy = 70.40%, test accuracy = 29.26%
2017-03-31 07:50:36.957162  iteration = 058, loss = 1.5591, val_loss = 4.6116, training accuracy = 70.46%, test accuracy = 29.30%
2017-03-31 08:33:56.551647  iteration = 059, loss = 1.5565, val_loss = 4.6123, training accuracy = 70.50%, test accuracy = 29.18%
2017-03-31 09:17:04.000139  iteration = 060, loss = 1.5570, val_loss = 4.6146, training accuracy = 70.50%, test accuracy = 29.30%
2017-03-31 09:59:57.376208  iteration = 061, loss = 1.5549, val_loss = 4.6161, training accuracy = 70.56%, test accuracy = 29.34%
2017-03-31 10:42:45.990349  iteration = 062, loss = 1.5536, val_loss = 4.6173, training accuracy = 70.60%, test accuracy = 29.30%
2017-03-31 11:25:39.162076  iteration = 063, loss = 1.5518, val_loss = 4.6191, training accuracy = 70.50%, test accuracy = 29.26%
2017-03-31 12:08:16.275953  iteration = 064, loss = 1.5514, val_loss = 4.6204, training accuracy = 70.58%, test accuracy = 29.34%
2017-03-31 12:51:00.457371  iteration = 065, loss = 1.5526, val_loss = 4.6202, training accuracy = 70.72%, test accuracy = 29.36%
2017-03-31 13:33:44.966449  iteration = 066, loss = 1.5497, val_loss = 4.6223, training accuracy = 70.66%, test accuracy = 29.36%
2017-03-31 14:17:22.056905  iteration = 067, loss = 1.5490, val_loss = 4.6243, training accuracy = 70.56%, test accuracy = 29.24%
2017-03-31 15:00:40.705205  iteration = 068, loss = 1.5474, val_loss = 4.6250, training accuracy = 70.68%, test accuracy = 29.28%
2017-03-31 15:44:05.714728  iteration = 069, loss = 1.5461, val_loss = 4.6265, training accuracy = 70.80%, test accuracy = 29.26%
2017-03-31 16:27:14.640888  iteration = 070, loss = 1.5459, val_loss = 4.6286, training accuracy = 70.92%, test accuracy = 29.42%
2017-03-31 17:10:06.652559  iteration = 071, loss = 1.5444, val_loss = 4.6296, training accuracy = 70.82%, test accuracy = 29.36%
2017-03-31 17:52:56.910685  iteration = 072, loss = 1.5445, val_loss = 4.6311, training accuracy = 71.00%, test accuracy = 29.22%
2017-03-31 18:35:34.362907  iteration = 073, loss = 1.5428, val_loss = 4.6327, training accuracy = 71.12%, test accuracy = 29.26%
2017-03-31 19:17:55.849864  iteration = 074, loss = 1.5409, val_loss = 4.6334, training accuracy = 71.08%, test accuracy = 29.34%
2017-03-31 20:01:46.487402  iteration = 075, loss = 1.5392, val_loss = 4.6357, training accuracy = 71.14%, test accuracy = 29.36%
2017-03-31 20:45:19.786181  iteration = 076, loss = 1.5381, val_loss = 4.6362, training accuracy = 71.10%, test accuracy = 29.28%
2017-03-31 21:27:57.603595  iteration = 077, loss = 1.5390, val_loss = 4.6379, training accuracy = 71.06%, test accuracy = 29.28%
2017-03-31 22:10:44.166768  iteration = 078, loss = 1.5353, val_loss = 4.6389, training accuracy = 71.20%, test accuracy = 29.28%
2017-03-31 22:53:54.455393  iteration = 079, loss = 1.5371, val_loss = 4.6389, training accuracy = 71.10%, test accuracy = 29.22%
2017-03-31 23:37:27.729936  iteration = 080, loss = 1.5364, val_loss = 4.6409, training accuracy = 71.26%, test accuracy = 29.20%
2017-04-01 00:20:44.244454  iteration = 081, loss = 1.5330, val_loss = 4.6437, training accuracy = 71.24%, test accuracy = 29.30%
2017-04-01 01:04:04.752409  iteration = 082, loss = 1.5326, val_loss = 4.6439, training accuracy = 71.22%, test accuracy = 29.32%
2017-04-01 01:47:17.230215  iteration = 083, loss = 1.5322, val_loss = 4.6451, training accuracy = 71.32%, test accuracy = 29.20%
2017-04-01 02:30:10.122191  iteration = 084, loss = 1.5305, val_loss = 4.6468, training accuracy = 71.36%, test accuracy = 29.30%
2017-04-01 03:12:59.736884  iteration = 085, loss = 1.5297, val_loss = 4.6480, training accuracy = 71.42%, test accuracy = 29.22%
2017-04-01 03:55:55.665759  iteration = 086, loss = 1.5294, val_loss = 4.6496, training accuracy = 71.42%, test accuracy = 29.28%
2017-04-01 04:38:39.013619  iteration = 087, loss = 1.5274, val_loss = 4.6512, training accuracy = 71.36%, test accuracy = 29.16%
2017-04-01 05:21:20.005339  iteration = 088, loss = 1.5267, val_loss = 4.6523, training accuracy = 71.34%, test accuracy = 29.30%
2017-04-01 06:03:53.155965  iteration = 089, loss = 1.5272, val_loss = 4.6527, training accuracy = 71.34%, test accuracy = 29.30%
2017-04-01 06:47:40.252202  iteration = 090, loss = 1.5246, val_loss = 4.6535, training accuracy = 71.48%, test accuracy = 29.26%
2017-04-01 07:31:17.175567  iteration = 091, loss = 1.5256, val_loss = 4.6551, training accuracy = 71.40%, test accuracy = 29.24%
2017-04-01 08:14:48.324006  iteration = 092, loss = 1.5231, val_loss = 4.6580, training accuracy = 71.34%, test accuracy = 29.24%
2017-04-01 08:57:47.623197  iteration = 093, loss = 1.5224, val_loss = 4.6570, training accuracy = 71.50%, test accuracy = 29.28%
2017-04-01 09:40:51.957399  iteration = 094, loss = 1.5201, val_loss = 4.6584, training accuracy = 71.40%, test accuracy = 29.24%
2017-04-01 10:23:42.308927  iteration = 095, loss = 1.5217, val_loss = 4.6603, training accuracy = 71.64%, test accuracy = 29.26%
2017-04-01 11:06:32.087482  iteration = 096, loss = 1.5200, val_loss = 4.6630, training accuracy = 71.50%, test accuracy = 29.18%
2017-04-01 11:49:26.938237  iteration = 097, loss = 1.5185, val_loss = 4.6626, training accuracy = 71.64%, test accuracy = 29.20%
2017-04-01 12:32:22.937738  iteration = 098, loss = 1.5167, val_loss = 4.6645, training accuracy = 71.76%, test accuracy = 29.18%
2017-04-01 13:15:12.816809  iteration = 099, loss = 1.5164, val_loss = 4.6660, training accuracy = 71.76%, test accuracy = 29.22%
2017-04-01 13:58:02.707144  iteration = 100, loss = 1.5150, val_loss = 4.6677, training accuracy = 71.86%, test accuracy = 29.24%
2017-04-01 14:40:50.901567  iteration = 101, loss = 1.5139, val_loss = 4.6673, training accuracy = 71.88%, test accuracy = 29.18%
2017-04-01 15:23:32.867677  iteration = 102, loss = 1.5135, val_loss = 4.6696, training accuracy = 71.86%, test accuracy = 29.20%
2017-04-01 16:06:27.158131  iteration = 103, loss = 1.5132, val_loss = 4.6702, training accuracy = 71.90%, test accuracy = 29.24%
2017-04-01 16:49:21.979902  iteration = 104, loss = 1.5119, val_loss = 4.6723, training accuracy = 71.90%, test accuracy = 29.18%
2017-04-01 17:32:00.548600  iteration = 105, loss = 1.5107, val_loss = 4.6740, training accuracy = 72.00%, test accuracy = 29.10%
2017-04-01 18:14:31.400906  iteration = 106, loss = 1.5087, val_loss = 4.6755, training accuracy = 72.04%, test accuracy = 29.12%
2017-04-01 18:58:09.227147  iteration = 107, loss = 1.5096, val_loss = 4.6755, training accuracy = 72.02%, test accuracy = 29.14%
2017-04-01 19:41:46.100178  iteration = 108, loss = 1.5079, val_loss = 4.6762, training accuracy = 72.06%, test accuracy = 29.12%
2017-04-01 20:25:29.924131  iteration = 109, loss = 1.5075, val_loss = 4.6781, training accuracy = 72.06%, test accuracy = 29.12%
2017-04-01 21:08:49.529851  iteration = 110, loss = 1.5064, val_loss = 4.6788, training accuracy = 72.12%, test accuracy = 29.22%
2017-04-01 21:51:09.269930  iteration = 111, loss = 1.5054, val_loss = 4.6801, training accuracy = 72.24%, test accuracy = 29.10%
2017-04-01 22:33:29.320495  iteration = 112, loss = 1.5048, val_loss = 4.6806, training accuracy = 72.26%, test accuracy = 29.28%
2017-04-01 23:15:52.240399  iteration = 113, loss = 1.5038, val_loss = 4.6818, training accuracy = 72.26%, test accuracy = 29.22%
2017-04-01 23:58:30.167180  iteration = 114, loss = 1.5030, val_loss = 4.6831, training accuracy = 72.26%, test accuracy = 29.24%
2017-04-02 00:41:21.384791  iteration = 115, loss = 1.5013, val_loss = 4.6857, training accuracy = 72.18%, test accuracy = 29.34%
2017-04-02 01:24:11.307388  iteration = 116, loss = 1.5016, val_loss = 4.6868, training accuracy = 72.20%, test accuracy = 29.24%
2017-04-02 02:06:59.284077  iteration = 117, loss = 1.4996, val_loss = 4.6870, training accuracy = 72.16%, test accuracy = 29.22%
2017-04-02 02:49:54.110341  iteration = 118, loss = 1.4998, val_loss = 4.6879, training accuracy = 72.36%, test accuracy = 29.26%
2017-04-02 03:32:45.778660  iteration = 119, loss = 1.4976, val_loss = 4.6909, training accuracy = 72.36%, test accuracy = 29.20%
2017-04-02 04:15:44.984326  iteration = 120, loss = 1.4971, val_loss = 4.6906, training accuracy = 72.30%, test accuracy = 29.16%
2017-04-02 04:58:35.684213  iteration = 121, loss = 1.4963, val_loss = 4.6925, training accuracy = 72.30%, test accuracy = 29.08%
2017-04-02 05:41:26.008407  iteration = 122, loss = 1.4939, val_loss = 4.6941, training accuracy = 72.44%, test accuracy = 29.26%
2017-04-02 06:24:20.097636  iteration = 123, loss = 1.4954, val_loss = 4.6942, training accuracy = 72.40%, test accuracy = 29.22%
2017-04-02 07:07:10.237945  iteration = 124, loss = 1.4923, val_loss = 4.6962, training accuracy = 72.48%, test accuracy = 29.16%
2017-04-02 07:49:56.170383  iteration = 125, loss = 1.4921, val_loss = 4.6968, training accuracy = 72.56%, test accuracy = 29.20%
2017-04-02 08:32:45.314395  iteration = 126, loss = 1.4920, val_loss = 4.6986, training accuracy = 72.46%, test accuracy = 29.20%
2017-04-02 09:15:34.899976  iteration = 127, loss = 1.4921, val_loss = 4.6992, training accuracy = 72.52%, test accuracy = 29.16%
2017-04-02 09:58:19.130688  iteration = 128, loss = 1.4891, val_loss = 4.7003, training accuracy = 72.60%, test accuracy = 29.00%
2017-04-02 10:41:14.935296  iteration = 129, loss = 1.4886, val_loss = 4.7016, training accuracy = 72.66%, test accuracy = 29.12%
2017-04-02 11:24:01.732807  iteration = 130, loss = 1.4890, val_loss = 4.7037, training accuracy = 72.58%, test accuracy = 29.16%
2017-04-02 12:06:40.968157  iteration = 131, loss = 1.4859, val_loss = 4.7035, training accuracy = 72.54%, test accuracy = 29.22%
2017-04-02 12:49:22.532311  iteration = 132, loss = 1.4856, val_loss = 4.7043, training accuracy = 72.66%, test accuracy = 29.16%
2017-04-02 13:32:01.319967  iteration = 133, loss = 1.4840, val_loss = 4.7061, training accuracy = 72.60%, test accuracy = 29.12%
2017-04-02 14:14:43.714212  iteration = 134, loss = 1.4844, val_loss = 4.7062, training accuracy = 72.68%, test accuracy = 29.12%
2017-04-02 14:58:10.555202  iteration = 135, loss = 1.4830, val_loss = 4.7079, training accuracy = 72.68%, test accuracy = 29.24%
2017-04-02 15:41:58.843907  iteration = 136, loss = 1.4833, val_loss = 4.7091, training accuracy = 72.68%, test accuracy = 29.18%
2017-04-02 16:25:24.058106  iteration = 137, loss = 1.4816, val_loss = 4.7111, training accuracy = 72.54%, test accuracy = 29.14%
2017-04-02 17:08:29.691636  iteration = 138, loss = 1.4809, val_loss = 4.7118, training accuracy = 72.62%, test accuracy = 29.12%
2017-04-02 17:51:32.203911  iteration = 139, loss = 1.4792, val_loss = 4.7134, training accuracy = 72.74%, test accuracy = 29.06%
2017-04-02 18:34:27.396360  iteration = 140, loss = 1.4782, val_loss = 4.7150, training accuracy = 72.66%, test accuracy = 29.14%
2017-04-02 19:17:19.345842  iteration = 141, loss = 1.4771, val_loss = 4.7160, training accuracy = 72.68%, test accuracy = 29.20%
2017-04-02 20:00:14.111888  iteration = 142, loss = 1.4773, val_loss = 4.7154, training accuracy = 72.78%, test accuracy = 29.14%
2017-04-02 20:43:05.807391  iteration = 143, loss = 1.4749, val_loss = 4.7191, training accuracy = 72.74%, test accuracy = 29.22%
2017-04-02 21:26:00.663960  iteration = 144, loss = 1.4736, val_loss = 4.7186, training accuracy = 72.78%, test accuracy = 29.18%
2017-04-02 22:08:42.248514  iteration = 145, loss = 1.4746, val_loss = 4.7203, training accuracy = 72.76%, test accuracy = 29.30%
2017-04-02 22:51:13.310224  iteration = 146, loss = 1.4729, val_loss = 4.7210, training accuracy = 72.82%, test accuracy = 29.20%
2017-04-02 23:34:51.350498  iteration = 147, loss = 1.4734, val_loss = 4.7206, training accuracy = 72.78%, test accuracy = 29.20%
2017-04-03 00:18:20.696324  iteration = 148, loss = 1.4718, val_loss = 4.7232, training accuracy = 72.84%, test accuracy = 29.22%
2017-04-03 01:02:15.031614  iteration = 149, loss = 1.4711, val_loss = 4.7241, training accuracy = 72.84%, test accuracy = 29.14%
2017-04-03 01:45:38.975828  iteration = 150, loss = 1.4697, val_loss = 4.7253, training accuracy = 72.86%, test accuracy = 29.20%
2017-04-03 02:28:56.438779  iteration = 151, loss = 1.4682, val_loss = 4.7269, training accuracy = 72.98%, test accuracy = 29.22%
2017-04-03 03:11:54.174412  iteration = 152, loss = 1.4676, val_loss = 4.7275, training accuracy = 73.06%, test accuracy = 29.30%
2017-04-03 03:54:48.676823  iteration = 153, loss = 1.4668, val_loss = 4.7294, training accuracy = 73.04%, test accuracy = 29.30%
2017-04-03 04:37:38.471131  iteration = 154, loss = 1.4675, val_loss = 4.7297, training accuracy = 73.08%, test accuracy = 29.30%
2017-04-03 05:20:30.888502  iteration = 155, loss = 1.4663, val_loss = 4.7303, training accuracy = 73.16%, test accuracy = 29.18%
2017-04-03 06:03:18.588353  iteration = 156, loss = 1.4649, val_loss = 4.7322, training accuracy = 73.22%, test accuracy = 29.16%
2017-04-03 06:46:05.272615  iteration = 157, loss = 1.4627, val_loss = 4.7338, training accuracy = 73.04%, test accuracy = 29.04%
2017-04-03 07:28:54.407490  iteration = 158, loss = 1.4645, val_loss = 4.7345, training accuracy = 73.22%, test accuracy = 29.22%
2017-04-03 08:11:13.634084  iteration = 159, loss = 1.4635, val_loss = 4.7356, training accuracy = 73.20%, test accuracy = 29.20%
2017-04-03 08:53:35.058489  iteration = 160, loss = 1.4616, val_loss = 4.7377, training accuracy = 73.20%, test accuracy = 29.20%
2017-04-03 09:37:37.748320  iteration = 161, loss = 1.4606, val_loss = 4.7378, training accuracy = 73.18%, test accuracy = 29.08%
2017-04-03 10:21:23.818703  iteration = 162, loss = 1.4598, val_loss = 4.7387, training accuracy = 73.18%, test accuracy = 29.22%
2017-04-03 11:04:52.167961  iteration = 163, loss = 1.4589, val_loss = 4.7398, training accuracy = 73.12%, test accuracy = 29.12%
2017-04-03 11:47:12.478603  iteration = 164, loss = 1.4599, val_loss = 4.7396, training accuracy = 73.20%, test accuracy = 29.02%
2017-04-03 12:30:33.185429  iteration = 165, loss = 1.4571, val_loss = 4.7424, training accuracy = 73.32%, test accuracy = 29.16%
2017-04-03 13:13:47.131907  iteration = 166, loss = 1.4552, val_loss = 4.7440, training accuracy = 73.34%, test accuracy = 29.18%
2017-04-03 13:56:44.173232  iteration = 167, loss = 1.4548, val_loss = 4.7443, training accuracy = 73.34%, test accuracy = 29.22%
2017-04-03 14:39:44.237522  iteration = 168, loss = 1.4552, val_loss = 4.7468, training accuracy = 73.34%, test accuracy = 29.16%
2017-04-03 15:22:39.456224  iteration = 169, loss = 1.4549, val_loss = 4.7457, training accuracy = 73.42%, test accuracy = 29.22%
2017-04-03 16:05:32.469312  iteration = 170, loss = 1.4535, val_loss = 4.7479, training accuracy = 73.38%, test accuracy = 29.16%
2017-04-03 16:48:24.029849  iteration = 171, loss = 1.4515, val_loss = 4.7493, training accuracy = 73.42%, test accuracy = 29.14%
2017-04-03 17:31:20.232159  iteration = 172, loss = 1.4533, val_loss = 4.7493, training accuracy = 73.38%, test accuracy = 29.14%
2017-04-03 18:14:11.993778  iteration = 173, loss = 1.4508, val_loss = 4.7509, training accuracy = 73.42%, test accuracy = 29.12%
2017-04-03 18:56:56.979402  iteration = 174, loss = 1.4496, val_loss = 4.7526, training accuracy = 73.44%, test accuracy = 29.08%
2017-04-03 19:39:37.348546  iteration = 175, loss = 1.4507, val_loss = 4.7540, training accuracy = 73.48%, test accuracy = 29.06%
2017-04-03 20:22:18.369610  iteration = 176, loss = 1.4494, val_loss = 4.7545, training accuracy = 73.52%, test accuracy = 29.08%
2017-04-03 21:06:09.290594  iteration = 177, loss = 1.4474, val_loss = 4.7552, training accuracy = 73.46%, test accuracy = 29.06%
2017-04-03 21:49:58.548806  iteration = 178, loss = 1.4473, val_loss = 4.7573, training accuracy = 73.48%, test accuracy = 29.10%
2017-04-03 22:33:44.201120  iteration = 179, loss = 1.4468, val_loss = 4.7569, training accuracy = 73.58%, test accuracy = 29.04%
2017-04-03 23:17:06.518429  iteration = 180, loss = 1.4449, val_loss = 4.7574, training accuracy = 73.54%, test accuracy = 29.16%
2017-04-04 00:00:23.106554  iteration = 181, loss = 1.4439, val_loss = 4.7599, training accuracy = 73.70%, test accuracy = 29.18%
2017-04-04 00:43:43.191804  iteration = 182, loss = 1.4445, val_loss = 4.7604, training accuracy = 73.62%, test accuracy = 29.10%
2017-04-04 01:26:40.862909  iteration = 183, loss = 1.4430, val_loss = 4.7611, training accuracy = 73.70%, test accuracy = 29.08%
2017-04-04 02:09:44.137270  iteration = 184, loss = 1.4420, val_loss = 4.7634, training accuracy = 73.68%, test accuracy = 29.06%
2017-04-04 02:52:51.543755  iteration = 185, loss = 1.4412, val_loss = 4.7633, training accuracy = 73.66%, test accuracy = 29.16%
2017-04-04 03:35:59.371327  iteration = 186, loss = 1.4409, val_loss = 4.7653, training accuracy = 73.66%, test accuracy = 29.16%
2017-04-04 04:19:03.254977  iteration = 187, loss = 1.4406, val_loss = 4.7656, training accuracy = 73.68%, test accuracy = 29.10%
2017-04-04 05:01:58.044444  iteration = 188, loss = 1.4383, val_loss = 4.7676, training accuracy = 73.78%, test accuracy = 29.14%
2017-04-04 05:44:35.719525  iteration = 189, loss = 1.4370, val_loss = 4.7680, training accuracy = 73.74%, test accuracy = 29.10%
2017-04-04 06:27:34.922025  iteration = 190, loss = 1.4363, val_loss = 4.7698, training accuracy = 73.76%, test accuracy = 29.18%
2017-04-04 07:11:22.743604  iteration = 191, loss = 1.4368, val_loss = 4.7694, training accuracy = 73.70%, test accuracy = 29.26%
2017-04-04 07:55:05.961168  iteration = 192, loss = 1.4352, val_loss = 4.7710, training accuracy = 73.82%, test accuracy = 29.26%
2017-04-04 08:38:29.944600  iteration = 193, loss = 1.4362, val_loss = 4.7714, training accuracy = 73.90%, test accuracy = 29.18%
2017-04-04 09:21:31.612740  iteration = 194, loss = 1.4329, val_loss = 4.7733, training accuracy = 73.84%, test accuracy = 29.18%
2017-04-04 10:04:41.659956  iteration = 195, loss = 1.4341, val_loss = 4.7735, training accuracy = 73.88%, test accuracy = 29.18%
2017-04-04 10:47:30.490119  iteration = 196, loss = 1.4326, val_loss = 4.7743, training accuracy = 74.04%, test accuracy = 29.06%
2017-04-04 11:30:33.294450  iteration = 197, loss = 1.4323, val_loss = 4.7769, training accuracy = 73.92%, test accuracy = 29.16%
2017-04-04 12:13:35.288759  iteration = 198, loss = 1.4310, val_loss = 4.7774, training accuracy = 73.96%, test accuracy = 29.14%
2017-04-04 12:56:37.152599  iteration = 199, loss = 1.4308, val_loss = 4.7786, training accuracy = 74.02%, test accuracy = 29.16%

Continuing training with a further run
sjb@persephone:sjb_word_model$ ./train.py theresa_may_contributions.json
2017-04-04 17:25:38.789650  iteration = 000, loss = 1.4294, val_loss = 4.7781, training accuracy = 74.08%, test accuracy = 29.08%
2017-04-04 18:09:25.800651  iteration = 001, loss = 1.4293, val_loss = 4.7809, training accuracy = 73.98%, test accuracy = 29.08%
2017-04-04 18:52:37.135426  iteration = 002, loss = 1.4257, val_loss = 4.7822, training accuracy = 74.02%, test accuracy = 29.24%
2017-04-04 19:35:23.289486  iteration = 003, loss = 1.4272, val_loss = 4.7823, training accuracy = 73.98%, test accuracy = 29.18%
2017-04-04 20:18:18.794794  iteration = 004, loss = 1.4265, val_loss = 4.7844, training accuracy = 74.00%, test accuracy = 29.18%
2017-04-04 21:01:03.436429  iteration = 005, loss = 1.4262, val_loss = 4.7843, training accuracy = 74.12%, test accuracy = 29.14%
2017-04-04 21:44:06.358767  iteration = 006, loss = 1.4240, val_loss = 4.7852, training accuracy = 74.20%, test accuracy = 29.10%
2017-04-04 22:27:17.298543  iteration = 007, loss = 1.4237, val_loss = 4.7873, training accuracy = 74.14%, test accuracy = 29.16%
2017-04-04 23:10:07.200995  iteration = 008, loss = 1.4240, val_loss = 4.7876, training accuracy = 74.12%, test accuracy = 29.06%
2017-04-04 23:52:58.381638  iteration = 009, loss = 1.4231, val_loss = 4.7877, training accuracy = 74.32%, test accuracy = 29.20%
2017-04-05 00:37:03.060838  iteration = 010, loss = 1.4208, val_loss = 4.7906, training accuracy = 74.34%, test accuracy = 29.18%
2017-04-05 01:20:06.817626  iteration = 011, loss = 1.4206, val_loss = 4.7909, training accuracy = 74.24%, test accuracy = 29.22%
2017-04-05 02:04:54.820133  iteration = 012, loss = 1.4185, val_loss = 4.7920, training accuracy = 74.30%, test accuracy = 29.16%
2017-04-05 02:49:19.501043  iteration = 013, loss = 1.4197, val_loss = 4.7913, training accuracy = 74.38%, test accuracy = 29.14%
2017-04-05 03:33:35.969278  iteration = 014, loss = 1.4180, val_loss = 4.7941, training accuracy = 74.30%, test accuracy = 29.12%
2017-04-05 04:17:36.964742  iteration = 015, loss = 1.4177, val_loss = 4.7951, training accuracy = 74.40%, test accuracy = 29.14%
2017-04-05 05:01:24.149579  iteration = 016, loss = 1.4168, val_loss = 4.7957, training accuracy = 74.46%, test accuracy = 29.22%
2017-04-05 05:44:56.251005  iteration = 017, loss = 1.4162, val_loss = 4.7970, training accuracy = 74.40%, test accuracy = 29.14%
2017-04-05 06:27:38.770828  iteration = 018, loss = 1.4160, val_loss = 4.7978, training accuracy = 74.28%, test accuracy = 29.22%
2017-04-05 07:10:39.598653  iteration = 019, loss = 1.4144, val_loss = 4.7997, training accuracy = 74.42%, test accuracy = 29.12%
2017-04-05 07:53:52.595873  iteration = 020, loss = 1.4135, val_loss = 4.8004, training accuracy = 74.42%, test accuracy = 29.10%
2017-04-05 08:37:04.688782  iteration = 021, loss = 1.4143, val_loss = 4.8001, training accuracy = 74.60%, test accuracy = 29.22%

Disk got full and training crashed. Started a further run
sjb@persephone:sjb_word_model$ ./train.py theresa_may_contributions.json
2017-04-06 11:48:51.539541  iteration = 000, loss = 1.4112, val_loss = 4.8018, training accuracy = 74.48%, test accuracy = 29.22%
2017-04-06 12:32:03.808826  iteration = 001, loss = 1.4119, val_loss = 4.8021, training accuracy = 74.58%, test accuracy = 29.14%
2017-04-06 13:15:10.608905  iteration = 002, loss = 1.4119, val_loss = 4.8038, training accuracy = 74.58%, test accuracy = 29.22%
2017-04-06 13:58:12.154515  iteration = 003, loss = 1.4100, val_loss = 4.8044, training accuracy = 74.70%, test accuracy = 29.12%
2017-04-06 14:41:14.992835  iteration = 004, loss = 1.4104, val_loss = 4.8040, training accuracy = 74.76%, test accuracy = 29.14%
2017-04-06 15:24:12.348365  iteration = 005, loss = 1.4087, val_loss = 4.8064, training accuracy = 74.82%, test accuracy = 29.12%
2017-04-06 16:07:10.996928  iteration = 006, loss = 1.4087, val_loss = 4.8079, training accuracy = 74.74%, test accuracy = 29.10%
2017-04-06 16:50:14.273177  iteration = 007, loss = 1.4079, val_loss = 4.8085, training accuracy = 74.74%, test accuracy = 29.18%
2017-04-06 17:33:18.976741  iteration = 008, loss = 1.4085, val_loss = 4.8110, training accuracy = 74.86%, test accuracy = 29.16%
2017-04-06 18:16:24.131119  iteration = 009, loss = 1.4087, val_loss = 4.8098, training accuracy = 74.74%, test accuracy = 29.22%
2017-04-06 18:59:21.228858  iteration = 010, loss = 1.4058, val_loss = 4.8111, training accuracy = 74.82%, test accuracy = 29.28%
2017-04-06 19:42:20.041288  iteration = 011, loss = 1.4032, val_loss = 4.8128, training accuracy = 74.82%, test accuracy = 29.12%
2017-04-06 20:25:25.259613  iteration = 012, loss = 1.4033, val_loss = 4.8136, training accuracy = 74.88%, test accuracy = 29.22%
2017-04-06 21:08:26.184189  iteration = 013, loss = 1.4038, val_loss = 4.8141, training accuracy = 74.90%, test accuracy = 29.18%
2017-04-06 21:51:33.850609  iteration = 014, loss = 1.4032, val_loss = 4.8159, training accuracy = 74.84%, test accuracy = 29.24%
2017-04-06 22:34:41.089247  iteration = 015, loss = 1.4006, val_loss = 4.8164, training accuracy = 74.98%, test accuracy = 29.20%
2017-04-06 23:17:31.853562  iteration = 016, loss = 1.4010, val_loss = 4.8175, training accuracy = 74.92%, test accuracy = 29.24%
2017-04-07 00:00:27.104362  iteration = 017, loss = 1.4015, val_loss = 4.8175, training accuracy = 74.96%, test accuracy = 29.16%
2017-04-07 00:43:22.713383  iteration = 018, loss = 1.3995, val_loss = 4.8201, training accuracy = 74.84%, test accuracy = 29.18%
2017-04-07 01:26:25.089517  iteration = 019, loss = 1.3985, val_loss = 4.8207, training accuracy = 75.08%, test accuracy = 29.12%
2017-04-07 02:09:16.831059  iteration = 020, loss = 1.3966, val_loss = 4.8215, training accuracy = 74.98%, test accuracy = 29.14%
2017-04-07 02:51:57.215817  iteration = 021, loss = 1.3975, val_loss = 4.8227, training accuracy = 75.08%, test accuracy = 29.08%
2017-04-07 03:34:54.693007  iteration = 022, loss = 1.3968, val_loss = 4.8226, training accuracy = 75.14%, test accuracy = 29.02%
2017-04-07 04:18:37.130021  iteration = 023, loss = 1.3949, val_loss = 4.8243, training accuracy = 75.08%, test accuracy = 29.18%
2017-04-07 05:06:44.227935  iteration = 024, loss = 1.3958, val_loss = 4.8241, training accuracy = 75.24%, test accuracy = 29.12%
2017-04-07 05:52:51.390948  iteration = 025, loss = 1.3928, val_loss = 4.8261, training accuracy = 75.08%, test accuracy = 29.10%
2017-04-07 06:36:29.390126  iteration = 026, loss = 1.3954, val_loss = 4.8257, training accuracy = 75.12%, test accuracy = 29.10%
2017-04-07 07:19:59.162126  iteration = 027, loss = 1.3950, val_loss = 4.8273, training accuracy = 75.22%, test accuracy = 29.10%
2017-04-07 08:03:12.439759  iteration = 028, loss = 1.3912, val_loss = 4.8287, training accuracy = 75.16%, test accuracy = 29.22%
2017-04-07 08:46:38.579335  iteration = 029, loss = 1.3912, val_loss = 4.8302, training accuracy = 75.14%, test accuracy = 29.18%
2017-04-07 09:29:39.822697  iteration = 030, loss = 1.3925, val_loss = 4.8297, training accuracy = 75.16%, test accuracy = 29.14%
2017-04-07 10:12:39.320858  iteration = 031, loss = 1.3897, val_loss = 4.8321, training accuracy = 75.16%, test accuracy = 29.20%
2017-04-07 10:55:32.409405  iteration = 032, loss = 1.3903, val_loss = 4.8332, training accuracy = 75.20%, test accuracy = 29.26%
2017-04-07 11:38:18.792794  iteration = 033, loss = 1.3909, val_loss = 4.8343, training accuracy = 75.14%, test accuracy = 29.24%
2017-04-07 12:21:00.552110  iteration = 034, loss = 1.3877, val_loss = 4.8337, training accuracy = 75.32%, test accuracy = 29.14%
2017-04-07 13:03:49.051431  iteration = 035, loss = 1.3881, val_loss = 4.8355, training accuracy = 75.32%, test accuracy = 29.20%
2017-04-07 13:46:37.435396  iteration = 036, loss = 1.3862, val_loss = 4.8365, training accuracy = 75.30%, test accuracy = 29.12%
2017-04-07 14:29:17.171691  iteration = 037, loss = 1.3875, val_loss = 4.8360, training accuracy = 75.28%, test accuracy = 29.10%
2017-04-07 15:12:08.556899  iteration = 038, loss = 1.3878, val_loss = 4.8380, training accuracy = 75.36%, test accuracy = 29.06%
2017-04-07 15:55:56.165991  iteration = 039, loss = 1.3852, val_loss = 4.8398, training accuracy = 75.30%, test accuracy = 29.12%
2017-04-07 16:39:45.006929  iteration = 040, loss = 1.3849, val_loss = 4.8400, training accuracy = 75.34%, test accuracy = 29.14%
2017-04-07 17:23:35.090231  iteration = 041, loss = 1.3839, val_loss = 4.8391, training accuracy = 75.34%, test accuracy = 29.14%
2017-04-07 18:07:13.227346  iteration = 042, loss = 1.3825, val_loss = 4.8416, training accuracy = 75.40%, test accuracy = 29.04%
2017-04-07 18:50:54.737840  iteration = 043, loss = 1.3831, val_loss = 4.8421, training accuracy = 75.38%, test accuracy = 29.16%
2017-04-07 19:34:21.554175  iteration = 044, loss = 1.3835, val_loss = 4.8433, training accuracy = 75.42%, test accuracy = 29.14%
2017-04-07 20:16:55.193980  iteration = 045, loss = 1.3802, val_loss = 4.8440, training accuracy = 75.52%, test accuracy = 29.14%
2017-04-07 20:59:30.128070  iteration = 046, loss = 1.3820, val_loss = 4.8447, training accuracy = 75.46%, test accuracy = 29.14%
2017-04-07 21:42:29.134078  iteration = 047, loss = 1.3777, val_loss = 4.8464, training accuracy = 75.50%, test accuracy = 29.12%
2017-04-07 22:25:45.673781  iteration = 048, loss = 1.3799, val_loss = 4.8466, training accuracy = 75.54%, test accuracy = 29.12%
2017-04-07 23:08:53.742545  iteration = 049, loss = 1.3792, val_loss = 4.8495, training accuracy = 75.58%, test accuracy = 29.08%
2017-04-07 23:52:04.205464  iteration = 050, loss = 1.3779, val_loss = 4.8477, training accuracy = 75.54%, test accuracy = 29.18%
2017-04-08 00:35:14.514081  iteration = 051, loss = 1.3750, val_loss = 4.8508, training accuracy = 75.52%, test accuracy = 29.12%
2017-04-08 01:18:22.847819  iteration = 052, loss = 1.3752, val_loss = 4.8520, training accuracy = 75.58%, test accuracy = 29.14%
2017-04-08 02:01:33.554809  iteration = 053, loss = 1.3752, val_loss = 4.8509, training accuracy = 75.60%, test accuracy = 29.10%
2017-04-08 02:44:47.228906  iteration = 054, loss = 1.3745, val_loss = 4.8526, training accuracy = 75.64%, test accuracy = 29.22%
2017-04-08 03:27:57.881604  iteration = 055, loss = 1.3740, val_loss = 4.8531, training accuracy = 75.80%, test accuracy = 29.12%
2017-04-08 04:11:02.428828  iteration = 056, loss = 1.3717, val_loss = 4.8548, training accuracy = 75.68%, test accuracy = 29.12%
2017-04-08 04:54:11.418630  iteration = 057, loss = 1.3725, val_loss = 4.8550, training accuracy = 75.78%, test accuracy = 29.08%
2017-04-08 05:37:12.397975  iteration = 058, loss = 1.3726, val_loss = 4.8556, training accuracy = 75.78%, test accuracy = 29.06%
2017-04-08 06:20:21.694549  iteration = 059, loss = 1.3701, val_loss = 4.8570, training accuracy = 75.72%, test accuracy = 29.10%
2017-04-08 07:03:37.944837  iteration = 060, loss = 1.3707, val_loss = 4.8573, training accuracy = 75.78%, test accuracy = 29.18%
2017-04-08 07:46:45.903085  iteration = 061, loss = 1.3695, val_loss = 4.8585, training accuracy = 75.84%, test accuracy = 29.18%
2017-04-08 08:29:49.532655  iteration = 062, loss = 1.3690, val_loss = 4.8603, training accuracy = 75.78%, test accuracy = 29.22%
2017-04-08 09:13:28.857537  iteration = 063, loss = 1.3697, val_loss = 4.8604, training accuracy = 75.86%, test accuracy = 29.12%
2017-04-08 09:58:29.044050  iteration = 064, loss = 1.3681, val_loss = 4.8617, training accuracy = 75.96%, test accuracy = 29.08%
2017-04-08 10:44:33.043572  iteration = 065, loss = 1.3671, val_loss = 4.8620, training accuracy = 75.86%, test accuracy = 29.04%
2017-04-08 11:28:21.957719  iteration = 066, loss = 1.3664, val_loss = 4.8645, training accuracy = 76.00%, test accuracy = 29.12%
2017-04-08 12:11:34.977893  iteration = 067, loss = 1.3658, val_loss = 4.8636, training accuracy = 76.00%, test accuracy = 29.04%
2017-04-08 12:54:43.677388  iteration = 068, loss = 1.3642, val_loss = 4.8656, training accuracy = 75.86%, test accuracy = 29.08%
2017-04-08 13:37:50.511309  iteration = 069, loss = 1.3645, val_loss = 4.8652, training accuracy = 75.90%, test accuracy = 29.12%
2017-04-08 14:20:57.850053  iteration = 070, loss = 1.3655, val_loss = 4.8664, training accuracy = 75.90%, test accuracy = 28.98%
2017-04-08 15:04:03.945145  iteration = 071, loss = 1.3640, val_loss = 4.8679, training accuracy = 76.06%, test accuracy = 29.06%
2017-04-08 15:47:11.407982  iteration = 072, loss = 1.3633, val_loss = 4.8678, training accuracy = 76.10%, test accuracy = 29.10%
2017-04-08 16:30:03.199843  iteration = 073, loss = 1.3616, val_loss = 4.8696, training accuracy = 76.10%, test accuracy = 29.14%
2017-04-08 17:12:50.494339  iteration = 074, loss = 1.3604, val_loss = 4.8697, training accuracy = 76.14%, test accuracy = 29.12%
2017-04-08 17:55:44.280004  iteration = 075, loss = 1.3613, val_loss = 4.8715, training accuracy = 76.10%, test accuracy = 29.06%
2017-04-08 18:39:41.615029  iteration = 076, loss = 1.3600, val_loss = 4.8710, training accuracy = 76.14%, test accuracy = 29.02%
2017-04-08 19:24:25.689189  iteration = 077, loss = 1.3580, val_loss = 4.8727, training accuracy = 76.08%, test accuracy = 29.08%
2017-04-08 20:08:06.193431  iteration = 078, loss = 1.3596, val_loss = 4.8732, training accuracy = 76.18%, test accuracy = 29.10%
2017-04-08 20:51:29.244578  iteration = 079, loss = 1.3584, val_loss = 4.8741, training accuracy = 76.20%, test accuracy = 29.10%
2017-04-08 21:34:53.364668  iteration = 080, loss = 1.3571, val_loss = 4.8760, training accuracy = 76.18%, test accuracy = 29.06%
2017-04-08 22:18:03.620758  iteration = 081, loss = 1.3559, val_loss = 4.8767, training accuracy = 76.20%, test accuracy = 29.02%
2017-04-08 23:01:09.206340  iteration = 082, loss = 1.3563, val_loss = 4.8780, training accuracy = 76.30%, test accuracy = 29.04%
2017-04-08 23:44:20.085825  iteration = 083, loss = 1.3563, val_loss = 4.8785, training accuracy = 76.20%, test accuracy = 29.16%
2017-04-09 00:27:25.812861  iteration = 084, loss = 1.3550, val_loss = 4.8787, training accuracy = 76.28%, test accuracy = 28.98%
2017-04-09 01:10:37.302894  iteration = 085, loss = 1.3548, val_loss = 4.8796, training accuracy = 76.26%, test accuracy = 29.06%
2017-04-09 01:53:48.050061  iteration = 086, loss = 1.3520, val_loss = 4.8814, training accuracy = 76.34%, test accuracy = 29.22%
2017-04-09 02:37:02.021152  iteration = 087, loss = 1.3516, val_loss = 4.8806, training accuracy = 76.34%, test accuracy = 29.08%
2017-04-09 03:20:09.803742  iteration = 088, loss = 1.3521, val_loss = 4.8820, training accuracy = 76.32%, test accuracy = 29.04%
2017-04-09 04:03:15.688597  iteration = 089, loss = 1.3520, val_loss = 4.8842, training accuracy = 76.44%, test accuracy = 29.14%
2017-04-09 04:46:12.577883  iteration = 090, loss = 1.3511, val_loss = 4.8841, training accuracy = 76.44%, test accuracy = 29.14%
2017-04-09 05:29:14.720472  iteration = 091, loss = 1.3508, val_loss = 4.8857, training accuracy = 76.38%, test accuracy = 29.10%
2017-04-09 06:12:20.008587  iteration = 092, loss = 1.3497, val_loss = 4.8853, training accuracy = 76.40%, test accuracy = 29.04%
2017-04-09 06:55:27.891978  iteration = 093, loss = 1.3478, val_loss = 4.8866, training accuracy = 76.42%, test accuracy = 29.10%
2017-04-09 07:38:37.180077  iteration = 094, loss = 1.3492, val_loss = 4.8880, training accuracy = 76.50%, test accuracy = 29.12%
2017-04-09 08:21:43.199698  iteration = 095, loss = 1.3483, val_loss = 4.8896, training accuracy = 76.42%, test accuracy = 29.20%
2017-04-09 09:04:52.423188  iteration = 096, loss = 1.3463, val_loss = 4.8902, training accuracy = 76.54%, test accuracy = 29.24%
2017-04-09 09:48:04.909191  iteration = 097, loss = 1.3472, val_loss = 4.8904, training accuracy = 76.44%, test accuracy = 29.16%
2017-04-09 10:30:51.103722  iteration = 098, loss = 1.3462, val_loss = 4.8915, training accuracy = 76.48%, test accuracy = 29.10%
2017-04-09 11:13:39.577923  iteration = 099, loss = 1.3447, val_loss = 4.8923, training accuracy = 76.56%, test accuracy = 29.22%
2017-04-09 11:57:35.398860  iteration = 100, loss = 1.3443, val_loss = 4.8926, training accuracy = 76.56%, test accuracy = 29.14%
2017-04-09 12:40:29.923198  iteration = 101, loss = 1.3431, val_loss = 4.8944, training accuracy = 76.60%, test accuracy = 29.18%
2017-04-09 13:24:30.030134  iteration = 102, loss = 1.3436, val_loss = 4.8947, training accuracy = 76.70%, test accuracy = 29.12%
2017-04-09 14:18:15.644567  iteration = 103, loss = 1.3427, val_loss = 4.8951, training accuracy = 76.70%, test accuracy = 29.12%
2017-04-09 15:12:56.677840  iteration = 104, loss = 1.3426, val_loss = 4.8970, training accuracy = 76.62%, test accuracy = 29.10%
2017-04-09 15:56:38.155202  iteration = 105, loss = 1.3416, val_loss = 4.8971, training accuracy = 76.62%, test accuracy = 29.06%
2017-04-09 16:39:28.110418  iteration = 106, loss = 1.3414, val_loss = 4.8968, training accuracy = 76.66%, test accuracy = 29.08%
2017-04-09 17:22:28.707893  iteration = 107, loss = 1.3405, val_loss = 4.8995, training accuracy = 76.60%, test accuracy = 29.10%
2017-04-09 18:06:19.902548  iteration = 108, loss = 1.3395, val_loss = 4.8989, training accuracy = 76.62%, test accuracy = 29.06%
2017-04-09 18:50:54.060744  iteration = 109, loss = 1.3388, val_loss = 4.8991, training accuracy = 76.68%, test accuracy = 29.10%
2017-04-09 19:34:17.198082  iteration = 110, loss = 1.3400, val_loss = 4.9008, training accuracy = 76.64%, test accuracy = 29.04%
2017-04-09 20:17:36.592735  iteration = 111, loss = 1.3383, val_loss = 4.9022, training accuracy = 76.70%, test accuracy = 29.10%
2017-04-09 21:01:04.309022  iteration = 112, loss = 1.3372, val_loss = 4.9030, training accuracy = 76.76%, test accuracy = 29.10%
2017-04-09 21:44:12.388860  iteration = 113, loss = 1.3361, val_loss = 4.9021, training accuracy = 76.76%, test accuracy = 29.10%
2017-04-09 22:27:17.134169  iteration = 114, loss = 1.3357, val_loss = 4.9035, training accuracy = 76.82%, test accuracy = 29.12%
2017-04-09 23:10:23.580118  iteration = 115, loss = 1.3359, val_loss = 4.9045, training accuracy = 76.80%, test accuracy = 29.12%
2017-04-09 23:53:31.081064  iteration = 116, loss = 1.3333, val_loss = 4.9067, training accuracy = 76.74%, test accuracy = 29.04%
2017-04-10 00:36:38.895197  iteration = 117, loss = 1.3339, val_loss = 4.9062, training accuracy = 76.78%, test accuracy = 29.14%
2017-04-10 01:19:40.853003  iteration = 118, loss = 1.3349, val_loss = 4.9071, training accuracy = 76.82%, test accuracy = 29.06%
2017-04-10 02:02:47.821470  iteration = 119, loss = 1.3336, val_loss = 4.9083, training accuracy = 76.88%, test accuracy = 29.20%
2017-04-10 02:45:47.848916  iteration = 120, loss = 1.3330, val_loss = 4.9085, training accuracy = 76.84%, test accuracy = 29.10%
2017-04-10 03:28:49.940646  iteration = 121, loss = 1.3309, val_loss = 4.9101, training accuracy = 76.80%, test accuracy = 29.10%
2017-04-10 04:11:59.993969  iteration = 122, loss = 1.3312, val_loss = 4.9095, training accuracy = 76.94%, test accuracy = 29.12%
2017-04-10 04:55:02.775842  iteration = 123, loss = 1.3299, val_loss = 4.9113, training accuracy = 76.92%, test accuracy = 29.04%
2017-04-10 05:38:06.261883  iteration = 124, loss = 1.3300, val_loss = 4.9134, training accuracy = 76.94%, test accuracy = 29.04%
2017-04-10 06:21:09.994572  iteration = 125, loss = 1.3295, val_loss = 4.9132, training accuracy = 76.90%, test accuracy = 29.06%
2017-04-10 07:04:19.360383  iteration = 126, loss = 1.3292, val_loss = 4.9144, training accuracy = 77.02%, test accuracy = 29.04%
2017-04-10 07:47:21.899259  iteration = 127, loss = 1.3285, val_loss = 4.9154, training accuracy = 77.04%, test accuracy = 29.04%
2017-04-10 08:30:28.235935  iteration = 128, loss = 1.3273, val_loss = 4.9147, training accuracy = 76.96%, test accuracy = 29.04%
2017-04-10 09:13:36.800430  iteration = 129, loss = 1.3277, val_loss = 4.9154, training accuracy = 77.16%, test accuracy = 29.10%
2017-04-10 09:56:38.685164  iteration = 130, loss = 1.3249, val_loss = 4.9186, training accuracy = 77.14%, test accuracy = 29.12%
2017-04-10 10:39:45.503347  iteration = 131, loss = 1.3271, val_loss = 4.9180, training accuracy = 77.08%, test accuracy = 29.12%
2017-04-10 11:22:53.382334  iteration = 132, loss = 1.3249, val_loss = 4.9192, training accuracy = 77.14%, test accuracy = 29.08%
2017-04-10 12:06:04.328174  iteration = 133, loss = 1.3243, val_loss = 4.9208, training accuracy = 77.26%, test accuracy = 29.16%
2017-04-10 12:49:12.339903  iteration = 134, loss = 1.3240, val_loss = 4.9211, training accuracy = 77.24%, test accuracy = 29.06%
2017-04-10 13:32:17.518794  iteration = 135, loss = 1.3228, val_loss = 4.9218, training accuracy = 77.30%, test accuracy = 29.06%
2017-04-10 14:15:26.641749  iteration = 136, loss = 1.3237, val_loss = 4.9218, training accuracy = 77.22%, test accuracy = 29.18%
2017-04-10 14:58:18.806541  iteration = 137, loss = 1.3220, val_loss = 4.9225, training accuracy = 77.30%, test accuracy = 29.06%
2017-04-10 15:41:10.547482  iteration = 138, loss = 1.3203, val_loss = 4.9237, training accuracy = 77.28%, test accuracy = 29.02%
2017-04-10 16:25:05.093716  iteration = 139, loss = 1.3220, val_loss = 4.9241, training accuracy = 77.24%, test accuracy = 29.18%
2017-04-10 17:08:53.004858  iteration = 140, loss = 1.3202, val_loss = 4.9257, training accuracy = 77.24%, test accuracy = 29.08%
2017-04-10 17:54:13.348510  iteration = 141, loss = 1.3191, val_loss = 4.9271, training accuracy = 77.08%, test accuracy = 29.06%
2017-04-10 18:37:58.923504  iteration = 142, loss = 1.3214, val_loss = 4.9271, training accuracy = 77.24%, test accuracy = 29.12%
2017-04-10 19:21:26.512332  iteration = 143, loss = 1.3189, val_loss = 4.9279, training accuracy = 77.28%, test accuracy = 29.12%
2017-04-10 20:04:49.412363  iteration = 144, loss = 1.3189, val_loss = 4.9278, training accuracy = 77.34%, test accuracy = 29.10%
2017-04-10 20:47:49.681351  iteration = 145, loss = 1.3162, val_loss = 4.9301, training accuracy = 77.32%, test accuracy = 29.16%
2017-04-10 21:30:59.066342  iteration = 146, loss = 1.3162, val_loss = 4.9306, training accuracy = 77.34%, test accuracy = 29.14%
2017-04-10 22:14:04.843453  iteration = 147, loss = 1.3152, val_loss = 4.9315, training accuracy = 77.34%, test accuracy = 29.02%
2017-04-10 22:57:12.721512  iteration = 148, loss = 1.3167, val_loss = 4.9328, training accuracy = 77.34%, test accuracy = 29.08%
2017-04-10 23:40:22.238022  iteration = 149, loss = 1.3149, val_loss = 4.9324, training accuracy = 77.34%, test accuracy = 28.96%
2017-04-11 00:23:31.895268  iteration = 150, loss = 1.3157, val_loss = 4.9330, training accuracy = 77.36%, test accuracy = 29.12%
2017-04-11 01:06:44.802481  iteration = 151, loss = 1.3127, val_loss = 4.9330, training accuracy = 77.46%, test accuracy = 29.02%
2017-04-11 01:49:50.834403  iteration = 152, loss = 1.3137, val_loss = 4.9347, training accuracy = 77.50%, test accuracy = 29.08%
2017-04-11 02:32:57.778453  iteration = 153, loss = 1.3117, val_loss = 4.9352, training accuracy = 77.34%, test accuracy = 29.16%
2017-04-11 03:16:03.670750  iteration = 154, loss = 1.3138, val_loss = 4.9352, training accuracy = 77.44%, test accuracy = 29.12%
2017-04-11 03:59:08.110127  iteration = 155, loss = 1.3106, val_loss = 4.9377, training accuracy = 77.44%, test accuracy = 29.12%
2017-04-11 04:42:17.879650  iteration = 156, loss = 1.3111, val_loss = 4.9396, training accuracy = 77.40%, test accuracy = 29.08%
2017-04-11 05:25:29.389736  iteration = 157, loss = 1.3095, val_loss = 4.9397, training accuracy = 77.42%, test accuracy = 29.06%
2017-04-11 06:08:19.134942  iteration = 158, loss = 1.3100, val_loss = 4.9404, training accuracy = 77.46%, test accuracy = 29.14%
2017-04-11 06:51:12.504339  iteration = 159, loss = 1.3105, val_loss = 4.9399, training accuracy = 77.56%, test accuracy = 29.08%
2017-04-11 07:35:03.433485  iteration = 160, loss = 1.3079, val_loss = 4.9420, training accuracy = 77.44%, test accuracy = 29.06%
2017-04-11 08:18:58.323913  iteration = 161, loss = 1.3088, val_loss = 4.9407, training accuracy = 77.60%, test accuracy = 29.12%
2017-04-11 09:02:50.141261  iteration = 162, loss = 1.3082, val_loss = 4.9428, training accuracy = 77.60%, test accuracy = 29.04%
2017-04-11 09:46:38.321645  iteration = 163, loss = 1.3061, val_loss = 4.9438, training accuracy = 77.64%, test accuracy = 29.06%
2017-04-11 10:30:25.863694  iteration = 164, loss = 1.3077, val_loss = 4.9431, training accuracy = 77.64%, test accuracy = 29.12%
2017-04-11 11:13:59.054222  iteration = 165, loss = 1.3057, val_loss = 4.9456, training accuracy = 77.68%, test accuracy = 29.04%
2017-04-11 11:57:22.662123  iteration = 166, loss = 1.3067, val_loss = 4.9449, training accuracy = 77.68%, test accuracy = 29.14%
2017-04-11 12:40:31.971176  iteration = 167, loss = 1.3046, val_loss = 4.9463, training accuracy = 77.64%, test accuracy = 29.00%
2017-04-11 13:23:45.404262  iteration = 168, loss = 1.3030, val_loss = 4.9462, training accuracy = 77.70%, test accuracy = 29.10%
2017-04-11 14:07:00.546435  iteration = 169, loss = 1.3026, val_loss = 4.9477, training accuracy = 77.68%, test accuracy = 29.10%
2017-04-11 14:50:02.786280  iteration = 170, loss = 1.3031, val_loss = 4.9491, training accuracy = 77.68%, test accuracy = 29.06%
2017-04-11 15:33:01.274427  iteration = 171, loss = 1.3014, val_loss = 4.9501, training accuracy = 77.56%, test accuracy = 29.06%
2017-04-11 16:16:10.144155  iteration = 172, loss = 1.3009, val_loss = 4.9496, training accuracy = 77.64%, test accuracy = 29.06%
2017-04-11 16:59:15.329421  iteration = 173, loss = 1.3013, val_loss = 4.9507, training accuracy = 77.68%, test accuracy = 29.02%
2017-04-11 17:42:16.208483  iteration = 174, loss = 1.2995, val_loss = 4.9516, training accuracy = 77.66%, test accuracy = 29.10%
2017-04-11 18:25:25.840213  iteration = 175, loss = 1.2998, val_loss = 4.9519, training accuracy = 77.76%, test accuracy = 29.04%
2017-04-11 19:08:25.315570  iteration = 176, loss = 1.3001, val_loss = 4.9550, training accuracy = 77.80%, test accuracy = 29.08%
2017-04-11 19:51:31.525842  iteration = 177, loss = 1.3002, val_loss = 4.9540, training accuracy = 77.84%, test accuracy = 29.00%
2017-04-11 20:34:26.870735  iteration = 178, loss = 1.2983, val_loss = 4.9554, training accuracy = 77.82%, test accuracy = 29.08%
2017-04-11 21:17:35.185235  iteration = 179, loss = 1.2988, val_loss = 4.9560, training accuracy = 77.90%, test accuracy = 29.08%
2017-04-11 22:00:22.736516  iteration = 180, loss = 1.2969, val_loss = 4.9563, training accuracy = 77.92%, test accuracy = 29.12%
2017-04-11 22:43:14.395428  iteration = 181, loss = 1.2967, val_loss = 4.9559, training accuracy = 77.86%, test accuracy = 29.14%
2017-04-11 23:26:05.361706  iteration = 182, loss = 1.2958, val_loss = 4.9590, training accuracy = 77.96%, test accuracy = 29.00%
2017-04-12 00:10:01.704115  iteration = 183, loss = 1.2958, val_loss = 4.9584, training accuracy = 77.98%, test accuracy = 28.96%
2017-04-12 00:53:51.139642  iteration = 184, loss = 1.2960, val_loss = 4.9613, training accuracy = 77.88%, test accuracy = 29.02%
2017-04-12 01:49:02.627816  iteration = 185, loss = 1.2940, val_loss = 4.9610, training accuracy = 78.00%, test accuracy = 29.16%
2017-04-12 02:42:46.816211  iteration = 186, loss = 1.2952, val_loss = 4.9603, training accuracy = 78.02%, test accuracy = 29.16%
2017-04-12 03:30:10.980860  iteration = 187, loss = 1.2936, val_loss = 4.9620, training accuracy = 78.10%, test accuracy = 29.04%
2017-04-12 04:13:47.650704  iteration = 188, loss = 1.2953, val_loss = 4.9617, training accuracy = 78.04%, test accuracy = 29.00%
2017-04-12 04:57:16.559851  iteration = 189, loss = 1.2936, val_loss = 4.9634, training accuracy = 78.00%, test accuracy = 29.04%
2017-04-12 05:40:42.832946  iteration = 190, loss = 1.2912, val_loss = 4.9640, training accuracy = 78.10%, test accuracy = 29.08%
2017-04-12 06:24:08.017351  iteration = 191, loss = 1.2911, val_loss = 4.9650, training accuracy = 78.10%, test accuracy = 28.90%
2017-04-12 07:07:27.168417  iteration = 192, loss = 1.2913, val_loss = 4.9648, training accuracy = 78.06%, test accuracy = 29.00%
2017-04-12 07:50:24.898139  iteration = 193, loss = 1.2906, val_loss = 4.9678, training accuracy = 78.04%, test accuracy = 29.00%
2017-04-12 08:32:58.207061  iteration = 194, loss = 1.2907, val_loss = 4.9681, training accuracy = 78.04%, test accuracy = 29.02%
2017-04-12 09:15:49.023399  iteration = 195, loss = 1.2883, val_loss = 4.9678, training accuracy = 78.12%, test accuracy = 29.02%
2017-04-12 09:58:57.867449  iteration = 196, loss = 1.2881, val_loss = 4.9681, training accuracy = 78.08%, test accuracy = 29.02%
2017-04-12 10:46:46.574558  iteration = 197, loss = 1.2877, val_loss = 4.9693, training accuracy = 78.18%, test accuracy = 29.04%
2017-04-12 11:29:54.357608  iteration = 198, loss = 1.2870, val_loss = 4.9701, training accuracy = 78.14%, test accuracy = 29.00%
2017-04-12 12:13:12.287788  iteration = 199, loss = 1.2863, val_loss = 4.9713, training accuracy = 78.22%, test accuracy = 29.04%

Continuing training with a further run
sjb@persephone:sjb_word_model$ ./train.py theresa_may_contributions.json
2017-04-13 13:55:08.309868  iteration = 000, loss = 1.2850, val_loss = 4.9732, training accuracy = 78.18%, test accuracy = 28.96%
2017-04-13 14:39:56.327691  iteration = 001, loss = 1.2853, val_loss = 4.9728, training accuracy = 78.22%, test accuracy = 29.06%
2017-04-13 15:27:10.264002  iteration = 002, loss = 1.2848, val_loss = 4.9732, training accuracy = 78.24%, test accuracy = 28.96%
2017-04-13 16:11:59.657590  iteration = 003, loss = 1.2840, val_loss = 4.9745, training accuracy = 78.24%, test accuracy = 29.00%
2017-04-13 16:56:40.287913  iteration = 004, loss = 1.2841, val_loss = 4.9755, training accuracy = 78.10%, test accuracy = 28.98%
2017-04-13 17:39:46.147786  iteration = 005, loss = 1.2840, val_loss = 4.9762, training accuracy = 78.16%, test accuracy = 29.12%
2017-04-13 18:22:49.041014  iteration = 006, loss = 1.2826, val_loss = 4.9772, training accuracy = 78.28%, test accuracy = 29.04%
2017-04-13 19:05:37.293221  iteration = 007, loss = 1.2835, val_loss = 4.9762, training accuracy = 78.30%, test accuracy = 29.06%
2017-04-13 19:49:30.199366  iteration = 008, loss = 1.2826, val_loss = 4.9771, training accuracy = 78.38%, test accuracy = 29.00%
2017-04-13 20:38:30.476995  iteration = 009, loss = 1.2808, val_loss = 4.9788, training accuracy = 78.38%, test accuracy = 28.88%
2017-04-13 21:27:41.673864  iteration = 010, loss = 1.2819, val_loss = 4.9802, training accuracy = 78.38%, test accuracy = 28.88%
2017-04-13 22:16:45.873563  iteration = 011, loss = 1.2814, val_loss = 4.9794, training accuracy = 78.32%, test accuracy = 29.02%
2017-04-13 23:05:28.307749  iteration = 012, loss = 1.2821, val_loss = 4.9801, training accuracy = 78.42%, test accuracy = 29.00%
2017-04-13 23:51:53.173909  iteration = 013, loss = 1.2800, val_loss = 4.9818, training accuracy = 78.36%, test accuracy = 29.00%
2017-04-14 00:38:51.001813  iteration = 014, loss = 1.2785, val_loss = 4.9827, training accuracy = 78.40%, test accuracy = 28.84%
2017-04-14 01:23:50.288348  iteration = 015, loss = 1.2771, val_loss = 4.9836, training accuracy = 78.36%, test accuracy = 28.82%
2017-04-14 02:07:08.226013  iteration = 016, loss = 1.2784, val_loss = 4.9827, training accuracy = 78.40%, test accuracy = 28.86%
2017-04-14 02:50:20.155856  iteration = 017, loss = 1.2761, val_loss = 4.9850, training accuracy = 78.42%, test accuracy = 29.06%
2017-04-14 03:33:24.128834  iteration = 018, loss = 1.2751, val_loss = 4.9831, training accuracy = 78.36%, test accuracy = 28.84%
2017-04-14 04:16:18.389761  iteration = 019, loss = 1.2759, val_loss = 4.9856, training accuracy = 78.52%, test accuracy = 28.92%
2017-04-14 04:59:13.167249  iteration = 020, loss = 1.2748, val_loss = 4.9871, training accuracy = 78.48%, test accuracy = 28.94%
2017-04-14 05:42:45.818213  iteration = 021, loss = 1.2767, val_loss = 4.9865, training accuracy = 78.46%, test accuracy = 28.94%
2017-04-14 06:26:40.346720  iteration = 022, loss = 1.2742, val_loss = 4.9878, training accuracy = 78.52%, test accuracy = 28.82%
2017-04-14 07:13:27.861522  iteration = 023, loss = 1.2741, val_loss = 4.9891, training accuracy = 78.52%, test accuracy = 28.82%
2017-04-14 07:56:29.677476  iteration = 024, loss = 1.2757, val_loss = 4.9895, training accuracy = 78.46%, test accuracy = 28.86%
2017-04-14 08:40:18.868676  iteration = 025, loss = 1.2732, val_loss = 4.9891, training accuracy = 78.52%, test accuracy = 28.92%
2017-04-14 09:24:10.447168  iteration = 026, loss = 1.2713, val_loss = 4.9917, training accuracy = 78.54%, test accuracy = 28.92%
2017-04-14 10:14:07.409769  iteration = 027, loss = 1.2715, val_loss = 4.9920, training accuracy = 78.56%, test accuracy = 28.96%
2017-04-14 10:57:46.151832  iteration = 028, loss = 1.2710, val_loss = 4.9921, training accuracy = 78.62%, test accuracy = 28.96%
2017-04-14 11:41:13.489918  iteration = 029, loss = 1.2698, val_loss = 4.9926, training accuracy = 78.60%, test accuracy = 28.96%
2017-04-14 12:24:39.659089  iteration = 030, loss = 1.2713, val_loss = 4.9931, training accuracy = 78.70%, test accuracy = 28.92%
2017-04-14 13:08:03.607181  iteration = 031, loss = 1.2708, val_loss = 4.9951, training accuracy = 78.64%, test accuracy = 28.96%
2017-04-14 13:51:11.316408  iteration = 032, loss = 1.2699, val_loss = 4.9946, training accuracy = 78.62%, test accuracy = 28.86%
2017-04-14 14:34:14.852855  iteration = 033, loss = 1.2701, val_loss = 4.9961, training accuracy = 78.70%, test accuracy = 28.92%
2017-04-14 15:17:23.369787  iteration = 034, loss = 1.2667, val_loss = 4.9955, training accuracy = 78.66%, test accuracy = 28.96%
2017-04-14 16:00:15.645350  iteration = 035, loss = 1.2681, val_loss = 4.9960, training accuracy = 78.76%, test accuracy = 28.78%
2017-04-14 16:43:05.636290  iteration = 036, loss = 1.2673, val_loss = 4.9983, training accuracy = 78.72%, test accuracy = 28.88%
2017-04-14 17:26:15.545755  iteration = 037, loss = 1.2669, val_loss = 4.9996, training accuracy = 78.76%, test accuracy = 28.88%
2017-04-14 18:10:19.309866  iteration = 038, loss = 1.2674, val_loss = 5.0006, training accuracy = 78.68%, test accuracy = 28.84%
2017-04-14 19:02:35.873133  iteration = 039, loss = 1.2631, val_loss = 5.0023, training accuracy = 78.74%, test accuracy = 28.86%
2017-04-14 19:49:38.819400  iteration = 040, loss = 1.2663, val_loss = 5.0008, training accuracy = 78.70%, test accuracy = 28.80%
2017-04-14 20:35:18.302170  iteration = 041, loss = 1.2640, val_loss = 5.0026, training accuracy = 78.82%, test accuracy = 28.86%
2017-04-14 21:20:40.027915  iteration = 042, loss = 1.2631, val_loss = 5.0028, training accuracy = 78.74%, test accuracy = 28.86%
2017-04-14 22:04:29.395297  iteration = 043, loss = 1.2637, val_loss = 5.0046, training accuracy = 78.90%, test accuracy = 28.90%
2017-04-14 22:48:27.145026  iteration = 044, loss = 1.2624, val_loss = 5.0047, training accuracy = 78.86%, test accuracy = 29.00%
2017-04-14 23:32:33.978802  iteration = 045, loss = 1.2629, val_loss = 5.0046, training accuracy = 78.74%, test accuracy = 28.92%
2017-04-15 00:16:53.305842  iteration = 046, loss = 1.2633, val_loss = 5.0051, training accuracy = 78.94%, test accuracy = 28.96%
2017-04-15 01:00:17.057028  iteration = 047, loss = 1.2598, val_loss = 5.0061, training accuracy = 78.88%, test accuracy = 28.90%
2017-04-15 01:44:55.071220  iteration = 048, loss = 1.2613, val_loss = 5.0073, training accuracy = 79.06%, test accuracy = 28.82%
2017-04-15 02:28:24.404595  iteration = 049, loss = 1.2602, val_loss = 5.0070, training accuracy = 79.00%, test accuracy = 28.86%
2017-04-15 03:12:00.784853  iteration = 050, loss = 1.2597, val_loss = 5.0077, training accuracy = 78.96%, test accuracy = 28.86%
2017-04-15 03:55:42.052172  iteration = 051, loss = 1.2595, val_loss = 5.0091, training accuracy = 79.04%, test accuracy = 28.78%
2017-04-15 04:38:40.078409  iteration = 052, loss = 1.2591, val_loss = 5.0096, training accuracy = 79.08%, test accuracy = 28.86%
2017-04-15 05:22:01.186223  iteration = 053, loss = 1.2592, val_loss = 5.0094, training accuracy = 79.16%, test accuracy = 28.94%
2017-04-15 06:06:03.901649  iteration = 054, loss = 1.2583, val_loss = 5.0112, training accuracy = 79.06%, test accuracy = 28.88%
2017-04-15 06:58:04.993743  iteration = 055, loss = 1.2574, val_loss = 5.0120, training accuracy = 79.14%, test accuracy = 28.84%
2017-04-15 07:47:05.157015  iteration = 056, loss = 1.2567, val_loss = 5.0128, training accuracy = 79.10%, test accuracy = 28.90%
2017-04-15 08:32:31.775058  iteration = 057, loss = 1.2549, val_loss = 5.0132, training accuracy = 79.14%, test accuracy = 28.86%
2017-04-15 09:16:40.405533  iteration = 058, loss = 1.2559, val_loss = 5.0134, training accuracy = 79.12%, test accuracy = 28.86%
2017-04-15 10:00:41.902668  iteration = 059, loss = 1.2547, val_loss = 5.0145, training accuracy = 79.22%, test accuracy = 28.82%
2017-04-15 10:44:12.725124  iteration = 060, loss = 1.2543, val_loss = 5.0161, training accuracy = 79.16%, test accuracy = 28.82%
2017-04-15 11:27:20.483402  iteration = 061, loss = 1.2537, val_loss = 5.0167, training accuracy = 79.26%, test accuracy = 28.78%
2017-04-15 12:10:33.235228  iteration = 062, loss = 1.2529, val_loss = 5.0167, training accuracy = 79.14%, test accuracy = 28.92%
2017-04-15 12:53:39.292116  iteration = 063, loss = 1.2541, val_loss = 5.0181, training accuracy = 79.12%, test accuracy = 28.94%
2017-04-15 13:36:46.579626  iteration = 064, loss = 1.2532, val_loss = 5.0182, training accuracy = 79.16%, test accuracy = 28.86%
2017-04-15 14:19:54.846050  iteration = 065, loss = 1.2522, val_loss = 5.0182, training accuracy = 79.26%, test accuracy = 28.92%
2017-04-15 15:03:04.440484  iteration = 066, loss = 1.2527, val_loss = 5.0194, training accuracy = 79.30%, test accuracy = 28.92%
2017-04-15 15:46:13.686560  iteration = 067, loss = 1.2506, val_loss = 5.0210, training accuracy = 79.28%, test accuracy = 28.88%
2017-04-15 16:29:05.872407  iteration = 068, loss = 1.2510, val_loss = 5.0205, training accuracy = 79.28%, test accuracy = 28.84%
2017-04-15 17:12:02.363070  iteration = 069, loss = 1.2509, val_loss = 5.0211, training accuracy = 79.30%, test accuracy = 29.02%
2017-04-15 17:55:37.667269  iteration = 070, loss = 1.2516, val_loss = 5.0218, training accuracy = 79.30%, test accuracy = 28.92%
2017-04-15 18:39:28.992141  iteration = 071, loss = 1.2480, val_loss = 5.0231, training accuracy = 79.24%, test accuracy = 29.02%
2017-04-15 19:23:14.181977  iteration = 072, loss = 1.2488, val_loss = 5.0236, training accuracy = 79.38%, test accuracy = 29.00%
2017-04-15 20:06:05.643133  iteration = 073, loss = 1.2498, val_loss = 5.0245, training accuracy = 79.36%, test accuracy = 28.94%
2017-04-15 20:50:03.785156  iteration = 074, loss = 1.2480, val_loss = 5.0255, training accuracy = 79.38%, test accuracy = 28.90%
2017-04-15 21:34:01.836913  iteration = 075, loss = 1.2470, val_loss = 5.0252, training accuracy = 79.38%, test accuracy = 28.88%
2017-04-15 22:24:58.359208  iteration = 076, loss = 1.2469, val_loss = 5.0271, training accuracy = 79.52%, test accuracy = 28.88%
2017-04-15 23:14:39.902254  iteration = 077, loss = 1.2463, val_loss = 5.0270, training accuracy = 79.42%, test accuracy = 29.02%
2017-04-16 00:01:25.060210  iteration = 078, loss = 1.2474, val_loss = 5.0274, training accuracy = 79.38%, test accuracy = 28.86%
2017-04-16 00:49:52.016812  iteration = 079, loss = 1.2450, val_loss = 5.0289, training accuracy = 79.46%, test accuracy = 28.92%
2017-04-16 01:34:50.052578  iteration = 080, loss = 1.2449, val_loss = 5.0300, training accuracy = 79.48%, test accuracy = 29.00%
2017-04-16 02:19:49.156154  iteration = 081, loss = 1.2426, val_loss = 5.0295, training accuracy = 79.50%, test accuracy = 28.94%
2017-04-16 03:04:32.344583  iteration = 082, loss = 1.2446, val_loss = 5.0304, training accuracy = 79.56%, test accuracy = 28.92%
2017-04-16 03:48:51.031592  iteration = 083, loss = 1.2428, val_loss = 5.0311, training accuracy = 79.50%, test accuracy = 28.86%
2017-04-16 04:33:08.629384  iteration = 084, loss = 1.2438, val_loss = 5.0317, training accuracy = 79.52%, test accuracy = 28.94%
2017-04-16 05:17:15.815720  iteration = 085, loss = 1.2428, val_loss = 5.0339, training accuracy = 79.58%, test accuracy = 28.84%
2017-04-16 06:00:42.937741  iteration = 086, loss = 1.2420, val_loss = 5.0336, training accuracy = 79.54%, test accuracy = 28.94%
2017-04-16 06:45:47.730576  iteration = 087, loss = 1.2410, val_loss = 5.0346, training accuracy = 79.54%, test accuracy = 28.92%
2017-04-16 07:29:06.948178  iteration = 088, loss = 1.2414, val_loss = 5.0351, training accuracy = 79.62%, test accuracy = 28.94%
2017-04-16 08:12:13.766920  iteration = 089, loss = 1.2399, val_loss = 5.0367, training accuracy = 79.86%, test accuracy = 28.86%
2017-04-16 08:55:18.756333  iteration = 090, loss = 1.2401, val_loss = 5.0360, training accuracy = 79.62%, test accuracy = 28.88%
2017-04-16 09:38:25.466301  iteration = 091, loss = 1.2391, val_loss = 5.0379, training accuracy = 79.72%, test accuracy = 28.86%
2017-04-16 10:21:29.606463  iteration = 092, loss = 1.2394, val_loss = 5.0386, training accuracy = 79.70%, test accuracy = 28.96%
2017-04-16 11:04:37.165022  iteration = 093, loss = 1.2391, val_loss = 5.0376, training accuracy = 79.88%, test accuracy = 28.88%
2017-04-16 11:47:43.205404  iteration = 094, loss = 1.2382, val_loss = 5.0389, training accuracy = 79.78%, test accuracy = 28.86%
2017-04-16 12:30:48.559095  iteration = 095, loss = 1.2375, val_loss = 5.0387, training accuracy = 79.72%, test accuracy = 28.82%
2017-04-16 13:13:59.368589  iteration = 096, loss = 1.2359, val_loss = 5.0397, training accuracy = 79.74%, test accuracy = 28.92%
2017-04-16 13:57:03.487543  iteration = 097, loss = 1.2368, val_loss = 5.0407, training accuracy = 79.84%, test accuracy = 28.90%
2017-04-16 14:40:07.430863  iteration = 098, loss = 1.2360, val_loss = 5.0409, training accuracy = 79.80%, test accuracy = 28.92%
2017-04-16 15:22:56.593465  iteration = 099, loss = 1.2340, val_loss = 5.0420, training accuracy = 79.86%, test accuracy = 28.92%
2017-04-16 16:05:47.385586  iteration = 100, loss = 1.2349, val_loss = 5.0416, training accuracy = 79.88%, test accuracy = 28.86%
2017-04-16 16:49:24.548107  iteration = 101, loss = 1.2349, val_loss = 5.0436, training accuracy = 79.90%, test accuracy = 28.92%
2017-04-16 17:33:19.323965  iteration = 102, loss = 1.2342, val_loss = 5.0445, training accuracy = 80.06%, test accuracy = 28.88%
2017-04-16 18:17:14.203609  iteration = 103, loss = 1.2325, val_loss = 5.0457, training accuracy = 80.06%, test accuracy = 28.86%
2017-04-16 19:01:12.183604  iteration = 104, loss = 1.2338, val_loss = 5.0452, training accuracy = 80.10%, test accuracy = 28.88%
2017-04-16 19:46:16.605872  iteration = 105, loss = 1.2311, val_loss = 5.0467, training accuracy = 80.12%, test accuracy = 28.92%
2017-04-16 20:29:50.846654  iteration = 106, loss = 1.2322, val_loss = 5.0456, training accuracy = 80.06%, test accuracy = 28.98%
2017-04-16 21:13:21.812857  iteration = 107, loss = 1.2307, val_loss = 5.0471, training accuracy = 80.10%, test accuracy = 28.98%
2017-04-16 21:56:45.815415  iteration = 108, loss = 1.2295, val_loss = 5.0471, training accuracy = 80.14%, test accuracy = 28.94%
2017-04-16 22:39:51.175981  iteration = 109, loss = 1.2302, val_loss = 5.0494, training accuracy = 80.20%, test accuracy = 29.00%
2017-04-16 23:22:56.149112  iteration = 110, loss = 1.2322, val_loss = 5.0499, training accuracy = 80.12%, test accuracy = 28.94%
2017-04-17 00:06:03.458593  iteration = 111, loss = 1.2315, val_loss = 5.0493, training accuracy = 80.22%, test accuracy = 28.86%
2017-04-17 00:49:07.965993  iteration = 112, loss = 1.2279, val_loss = 5.0513, training accuracy = 80.22%, test accuracy = 28.96%
2017-04-17 01:32:13.162321  iteration = 113, loss = 1.2299, val_loss = 5.0514, training accuracy = 80.16%, test accuracy = 28.92%
2017-04-17 02:15:17.144775  iteration = 114, loss = 1.2290, val_loss = 5.0525, training accuracy = 80.22%, test accuracy = 28.92%
2017-04-17 02:58:23.974911  iteration = 115, loss = 1.2282, val_loss = 5.0525, training accuracy = 80.26%, test accuracy = 28.96%
2017-04-17 03:41:30.070469  iteration = 116, loss = 1.2275, val_loss = 5.0532, training accuracy = 80.26%, test accuracy = 28.90%
2017-04-17 04:24:33.933954  iteration = 117, loss = 1.2264, val_loss = 5.0541, training accuracy = 80.28%, test accuracy = 28.86%
2017-04-17 05:07:22.467968  iteration = 118, loss = 1.2267, val_loss = 5.0545, training accuracy = 80.22%, test accuracy = 28.74%
2017-04-17 05:50:07.979374  iteration = 119, loss = 1.2264, val_loss = 5.0536, training accuracy = 80.26%, test accuracy = 28.80%
2017-04-17 06:33:34.523040  iteration = 120, loss = 1.2239, val_loss = 5.0560, training accuracy = 80.36%, test accuracy = 28.86%
2017-04-17 07:17:20.665612  iteration = 121, loss = 1.2261, val_loss = 5.0555, training accuracy = 80.32%, test accuracy = 28.76%
2017-04-17 08:08:42.708092  iteration = 122, loss = 1.2255, val_loss = 5.0558, training accuracy = 80.28%, test accuracy = 28.92%
2017-04-17 09:01:47.035022  iteration = 123, loss = 1.2252, val_loss = 5.0565, training accuracy = 80.34%, test accuracy = 28.86%
2017-04-17 09:49:37.019743  iteration = 124, loss = 1.2215, val_loss = 5.0598, training accuracy = 80.40%, test accuracy = 28.82%
2017-04-17 10:34:50.222303  iteration = 125, loss = 1.2234, val_loss = 5.0582, training accuracy = 80.40%, test accuracy = 28.80%
2017-04-17 11:19:55.877761  iteration = 126, loss = 1.2229, val_loss = 5.0592, training accuracy = 80.34%, test accuracy = 28.88%
2017-04-17 12:04:50.042410  iteration = 127, loss = 1.2222, val_loss = 5.0595, training accuracy = 80.46%, test accuracy = 28.76%
2017-04-17 12:49:48.782344  iteration = 128, loss = 1.2223, val_loss = 5.0604, training accuracy = 80.34%, test accuracy = 28.84%
2017-04-17 13:34:48.742278  iteration = 129, loss = 1.2228, val_loss = 5.0614, training accuracy = 80.48%, test accuracy = 28.74%
2017-04-17 14:19:57.616242  iteration = 130, loss = 1.2197, val_loss = 5.0623, training accuracy = 80.48%, test accuracy = 28.84%
2017-04-17 15:03:10.870752  iteration = 131, loss = 1.2203, val_loss = 5.0644, training accuracy = 80.48%, test accuracy = 28.84%
2017-04-17 15:46:13.282617  iteration = 132, loss = 1.2201, val_loss = 5.0615, training accuracy = 80.50%, test accuracy = 28.84%
2017-04-17 16:30:06.618203  iteration = 133, loss = 1.2197, val_loss = 5.0633, training accuracy = 80.52%, test accuracy = 28.96%
2017-04-17 17:22:16.553604  iteration = 134, loss = 1.2202, val_loss = 5.0638, training accuracy = 80.56%, test accuracy = 28.92%
2017-04-17 18:06:17.161224  iteration = 135, loss = 1.2183, val_loss = 5.0665, training accuracy = 80.52%, test accuracy = 28.84%
2017-04-17 18:49:43.670107  iteration = 136, loss = 1.2180, val_loss = 5.0661, training accuracy = 80.54%, test accuracy = 28.94%
2017-04-17 19:33:04.554024  iteration = 137, loss = 1.2175, val_loss = 5.0671, training accuracy = 80.52%, test accuracy = 28.88%
2017-04-17 20:16:12.479829  iteration = 138, loss = 1.2167, val_loss = 5.0674, training accuracy = 80.50%, test accuracy = 28.90%
2017-04-17 20:59:19.331559  iteration = 139, loss = 1.2182, val_loss = 5.0674, training accuracy = 80.54%, test accuracy = 28.86%
2017-04-17 21:42:07.090966  iteration = 140, loss = 1.2150, val_loss = 5.0691, training accuracy = 80.52%, test accuracy = 28.82%
2017-04-17 22:24:58.758850  iteration = 141, loss = 1.2138, val_loss = 5.0716, training accuracy = 80.56%, test accuracy = 28.76%
2017-04-17 23:08:27.089965  iteration = 142, loss = 1.2161, val_loss = 5.0692, training accuracy = 80.56%, test accuracy = 28.74%
2017-04-17 23:52:09.674154  iteration = 143, loss = 1.2136, val_loss = 5.0700, training accuracy = 80.54%, test accuracy = 28.66%
2017-04-18 00:35:58.019149  iteration = 144, loss = 1.2150, val_loss = 5.0719, training accuracy = 80.58%, test accuracy = 28.78%
2017-04-18 01:19:33.317595  iteration = 145, loss = 1.2146, val_loss = 5.0707, training accuracy = 80.64%, test accuracy = 28.74%
2017-04-18 02:03:01.863803  iteration = 146, loss = 1.2141, val_loss = 5.0711, training accuracy = 80.62%, test accuracy = 28.86%
2017-04-18 02:46:29.259108  iteration = 147, loss = 1.2137, val_loss = 5.0728, training accuracy = 80.62%, test accuracy = 28.88%
2017-04-18 03:29:51.520072  iteration = 148, loss = 1.2138, val_loss = 5.0741, training accuracy = 80.72%, test accuracy = 28.86%
2017-04-18 04:13:02.678255  iteration = 149, loss = 1.2125, val_loss = 5.0738, training accuracy = 80.68%, test accuracy = 28.80%
2017-04-18 04:56:09.196331  iteration = 150, loss = 1.2128, val_loss = 5.0748, training accuracy = 80.74%, test accuracy = 28.78%
2017-04-18 05:39:13.139553  iteration = 151, loss = 1.2115, val_loss = 5.0758, training accuracy = 80.80%, test accuracy = 28.84%
2017-04-18 06:22:01.006788  iteration = 152, loss = 1.2099, val_loss = 5.0760, training accuracy = 80.72%, test accuracy = 28.86%
2017-04-18 07:04:50.254321  iteration = 153, loss = 1.2098, val_loss = 5.0771, training accuracy = 80.70%, test accuracy = 28.84%
2017-04-18 07:47:41.266728  iteration = 154, loss = 1.2091, val_loss = 5.0791, training accuracy = 80.86%, test accuracy = 28.82%
2017-04-18 08:31:30.503348  iteration = 155, loss = 1.2106, val_loss = 5.0776, training accuracy = 80.84%, test accuracy = 28.78%
2017-04-18 09:21:51.480327  iteration = 156, loss = 1.2084, val_loss = 5.0792, training accuracy = 80.86%, test accuracy = 28.82%
2017-04-18 10:11:28.075976  iteration = 157, loss = 1.2083, val_loss = 5.0788, training accuracy = 80.86%, test accuracy = 28.78%
2017-04-18 11:01:14.594828  iteration = 158, loss = 1.2095, val_loss = 5.0815, training accuracy = 80.88%, test accuracy = 28.90%
2017-04-18 11:50:32.638867  iteration = 159, loss = 1.2084, val_loss = 5.0804, training accuracy = 80.84%, test accuracy = 28.84%
2017-04-18 12:37:52.278904  iteration = 160, loss = 1.2070, val_loss = 5.0818, training accuracy = 80.76%, test accuracy = 28.78%
2017-04-18 13:25:32.195018  iteration = 161, loss = 1.2075, val_loss = 5.0822, training accuracy = 80.88%, test accuracy = 28.88%
2017-04-18 14:10:48.902410  iteration = 162, loss = 1.2071, val_loss = 5.0838, training accuracy = 80.88%, test accuracy = 28.80%
2017-04-18 14:55:29.108389  iteration = 163, loss = 1.2049, val_loss = 5.0832, training accuracy = 80.84%, test accuracy = 28.84%
2017-04-18 15:39:23.110840  iteration = 164, loss = 1.2061, val_loss = 5.0839, training accuracy = 80.88%, test accuracy = 28.86%
2017-04-18 16:23:14.066228  iteration = 165, loss = 1.2064, val_loss = 5.0837, training accuracy = 80.88%, test accuracy = 28.84%
2017-04-18 17:07:00.507652  iteration = 166, loss = 1.2038, val_loss = 5.0848, training accuracy = 80.96%, test accuracy = 28.80%
2017-04-18 17:50:52.942115  iteration = 167, loss = 1.2030, val_loss = 5.0852, training accuracy = 80.90%, test accuracy = 28.88%
2017-04-18 18:34:27.740462  iteration = 168, loss = 1.2048, val_loss = 5.0864, training accuracy = 80.96%, test accuracy = 28.84%
2017-04-18 19:17:57.430213  iteration = 169, loss = 1.2029, val_loss = 5.0870, training accuracy = 80.98%, test accuracy = 28.84%
2017-04-18 20:01:12.591958  iteration = 170, loss = 1.2032, val_loss = 5.0875, training accuracy = 81.00%, test accuracy = 28.84%
2017-04-18 20:44:31.938892  iteration = 171, loss = 1.2028, val_loss = 5.0870, training accuracy = 80.96%, test accuracy = 28.84%
2017-04-18 21:27:50.396030  iteration = 172, loss = 1.2019, val_loss = 5.0887, training accuracy = 80.98%, test accuracy = 28.82%
2017-04-18 22:10:53.251978  iteration = 173, loss = 1.2000, val_loss = 5.0902, training accuracy = 81.06%, test accuracy = 28.92%
2017-04-18 22:53:59.583394  iteration = 174, loss = 1.2009, val_loss = 5.0905, training accuracy = 81.10%, test accuracy = 28.84%
2017-04-18 23:36:48.301578  iteration = 175, loss = 1.2017, val_loss = 5.0905, training accuracy = 81.00%, test accuracy = 28.80%
2017-04-19 00:19:30.112847  iteration = 176, loss = 1.2007, val_loss = 5.0917, training accuracy = 81.12%, test accuracy = 28.80%
2017-04-19 01:02:10.572833  iteration = 177, loss = 1.1977, val_loss = 5.0930, training accuracy = 81.14%, test accuracy = 28.76%
2017-04-19 01:46:08.655630  iteration = 178, loss = 1.1991, val_loss = 5.0933, training accuracy = 81.18%, test accuracy = 28.82%
2017-04-19 02:30:05.553414  iteration = 179, loss = 1.1971, val_loss = 5.0939, training accuracy = 81.20%, test accuracy = 28.84%
2017-04-19 03:12:55.880244  iteration = 180, loss = 1.1986, val_loss = 5.0952, training accuracy = 81.06%, test accuracy = 28.84%
2017-04-19 03:55:50.030526  iteration = 181, loss = 1.1977, val_loss = 5.0942, training accuracy = 81.16%, test accuracy = 28.84%
2017-04-19 04:41:40.456190  iteration = 182, loss = 1.1982, val_loss = 5.0963, training accuracy = 81.28%, test accuracy = 28.88%
2017-04-19 05:25:54.038544  iteration = 183, loss = 1.1977, val_loss = 5.0955, training accuracy = 81.20%, test accuracy = 28.80%
2017-04-19 06:17:43.678809  iteration = 184, loss = 1.1973, val_loss = 5.0973, training accuracy = 81.22%, test accuracy = 28.70%
2017-04-19 07:01:04.822839  iteration = 185, loss = 1.1975, val_loss = 5.0961, training accuracy = 81.20%, test accuracy = 28.80%
2017-04-19 07:44:37.037002  iteration = 186, loss = 1.1958, val_loss = 5.0979, training accuracy = 81.18%, test accuracy = 28.90%
2017-04-19 08:28:05.516556  iteration = 187, loss = 1.1935, val_loss = 5.0977, training accuracy = 81.34%, test accuracy = 28.78%
2017-04-19 09:11:35.295916  iteration = 188, loss = 1.1954, val_loss = 5.0988, training accuracy = 81.22%, test accuracy = 28.86%
2017-04-19 09:54:53.104098  iteration = 189, loss = 1.1946, val_loss = 5.0998, training accuracy = 81.38%, test accuracy = 28.80%
2017-04-19 10:38:11.534085  iteration = 190, loss = 1.1929, val_loss = 5.1011, training accuracy = 81.30%, test accuracy = 28.88%
2017-04-19 11:21:34.787118  iteration = 191, loss = 1.1951, val_loss = 5.1002, training accuracy = 81.32%, test accuracy = 28.78%
2017-04-19 12:04:42.706420  iteration = 192, loss = 1.1944, val_loss = 5.1009, training accuracy = 81.30%, test accuracy = 28.82%
2017-04-19 12:47:54.347596  iteration = 193, loss = 1.1916, val_loss = 5.1018, training accuracy = 81.36%, test accuracy = 28.70%
2017-04-19 13:31:03.427738  iteration = 194, loss = 1.1925, val_loss = 5.1029, training accuracy = 81.42%, test accuracy = 28.82%
2017-04-19 14:14:05.444771  iteration = 195, loss = 1.1928, val_loss = 5.1027, training accuracy = 81.38%, test accuracy = 28.72%
2017-04-19 14:57:04.527807  iteration = 196, loss = 1.1921, val_loss = 5.1047, training accuracy = 81.36%, test accuracy = 28.72%
2017-04-19 15:41:11.488731  iteration = 197, loss = 1.1907, val_loss = 5.1050, training accuracy = 81.36%, test accuracy = 28.78%
2017-04-19 16:25:15.689266  iteration = 198, loss = 1.1902, val_loss = 5.1037, training accuracy = 81.46%, test accuracy = 28.72%
2017-04-19 17:11:48.179117  iteration = 199, loss = 1.1896, val_loss = 5.1045, training accuracy = 81.34%, test accuracy = 28.78%

Continuing training with a further run
sjb@persephone:sjb_word_model$ ./train.py theresa_may_contributions.json
2017-04-20 23:27:42.798093  iteration = 000, loss = 1.1890, val_loss = 5.1042, training accuracy = 81.34%, test accuracy = 28.78%
2017-04-21 00:29:06.712063  iteration = 001, loss = 1.1909, val_loss = 5.1065, training accuracy = 81.50%, test accuracy = 28.76%
2017-04-21 01:44:57.588345  iteration = 002, loss = 1.1883, val_loss = 5.1074, training accuracy = 81.40%, test accuracy = 28.76%
2017-04-21 02:29:00.652485  iteration = 003, loss = 1.1888, val_loss = 5.1080, training accuracy = 81.36%, test accuracy = 28.84%
2017-04-21 03:27:30.160791  iteration = 004, loss = 1.1875, val_loss = 5.1103, training accuracy = 81.52%, test accuracy = 28.78%
2017-04-21 04:13:02.847078  iteration = 005, loss = 1.1878, val_loss = 5.1093, training accuracy = 81.48%, test accuracy = 28.74%
2017-04-21 04:56:59.855404  iteration = 006, loss = 1.1872, val_loss = 5.1100, training accuracy = 81.50%, test accuracy = 28.72%
2017-04-21 05:40:52.203097  iteration = 007, loss = 1.1880, val_loss = 5.1104, training accuracy = 81.54%, test accuracy = 28.78%
2017-04-21 06:31:10.477251  iteration = 008, loss = 1.1856, val_loss = 5.1100, training accuracy = 81.54%, test accuracy = 28.72%
2017-04-21 07:23:20.557165  iteration = 009, loss = 1.1860, val_loss = 5.1114, training accuracy = 81.54%, test accuracy = 28.70%
2017-04-21 08:16:39.208832  iteration = 010, loss = 1.1859, val_loss = 5.1113, training accuracy = 81.66%, test accuracy = 28.72%
2017-04-21 09:11:09.899240  iteration = 011, loss = 1.1857, val_loss = 5.1121, training accuracy = 81.52%, test accuracy = 28.66%
2017-04-21 10:04:02.642610  iteration = 012, loss = 1.1839, val_loss = 5.1137, training accuracy = 81.62%, test accuracy = 28.70%
2017-04-21 10:47:11.787985  iteration = 013, loss = 1.1841, val_loss = 5.1141, training accuracy = 81.62%, test accuracy = 28.68%
2017-04-21 11:31:56.598151  iteration = 014, loss = 1.1832, val_loss = 5.1145, training accuracy = 81.56%, test accuracy = 28.66%
2017-04-21 12:14:29.943442  iteration = 015, loss = 1.1820, val_loss = 5.1160, training accuracy = 81.66%, test accuracy = 28.62%
2017-04-21 12:56:34.499290  iteration = 016, loss = 1.1806, val_loss = 5.1173, training accuracy = 81.60%, test accuracy = 28.70%
2017-04-21 13:39:52.976313  iteration = 017, loss = 1.1827, val_loss = 5.1156, training accuracy = 81.64%, test accuracy = 28.72%
2017-04-21 14:30:17.339139  iteration = 018, loss = 1.1821, val_loss = 5.1174, training accuracy = 81.58%, test accuracy = 28.76%
2017-04-21 15:21:31.357765  iteration = 019, loss = 1.1800, val_loss = 5.1181, training accuracy = 81.68%, test accuracy = 28.70%
2017-04-21 16:07:00.114264  iteration = 020, loss = 1.1812, val_loss = 5.1184, training accuracy = 81.62%, test accuracy = 28.76%
2017-04-21 16:52:33.895391  iteration = 021, loss = 1.1807, val_loss = 5.1186, training accuracy = 81.74%, test accuracy = 28.74%
2017-04-21 17:27:49.876095  iteration = 022, loss = 1.1800, val_loss = 5.1194, training accuracy = 81.70%, test accuracy = 28.68%
2017-04-21 18:05:14.198724  iteration = 023, loss = 1.1796, val_loss = 5.1200, training accuracy = 81.70%, test accuracy = 28.64%
2017-04-21 18:42:18.718087  iteration = 024, loss = 1.1787, val_loss = 5.1203, training accuracy = 81.78%, test accuracy = 28.70%
2017-04-21 19:26:13.064012  iteration = 025, loss = 1.1788, val_loss = 5.1219, training accuracy = 81.72%, test accuracy = 28.70%
2017-04-21 20:09:56.075598  iteration = 026, loss = 1.1791, val_loss = 5.1220, training accuracy = 81.74%, test accuracy = 28.72%
2017-04-21 20:53:45.561287  iteration = 027, loss = 1.1776, val_loss = 5.1233, training accuracy = 81.76%, test accuracy = 28.70%
2017-04-21 21:37:17.902175  iteration = 028, loss = 1.1768, val_loss = 5.1234, training accuracy = 81.82%, test accuracy = 28.78%
2017-04-21 22:20:45.358564  iteration = 029, loss = 1.1769, val_loss = 5.1239, training accuracy = 81.86%, test accuracy = 28.80%
2017-04-21 23:04:06.907274  iteration = 030, loss = 1.1766, val_loss = 5.1240, training accuracy = 81.80%, test accuracy = 28.82%
2017-04-21 23:47:11.647726  iteration = 031, loss = 1.1769, val_loss = 5.1251, training accuracy = 81.92%, test accuracy = 28.90%
2017-04-22 00:30:17.693995  iteration = 032, loss = 1.1760, val_loss = 5.1256, training accuracy = 81.86%, test accuracy = 28.84%
2017-04-22 01:13:15.698464  iteration = 033, loss = 1.1754, val_loss = 5.1258, training accuracy = 81.82%, test accuracy = 28.80%
2017-04-22 01:56:20.184535  iteration = 034, loss = 1.1753, val_loss = 5.1270, training accuracy = 81.94%, test accuracy = 28.88%
2017-04-22 02:39:48.135913  iteration = 035, loss = 1.1740, val_loss = 5.1270, training accuracy = 81.84%, test accuracy = 28.78%
2017-04-22 03:22:17.321724  iteration = 036, loss = 1.1742, val_loss = 5.1285, training accuracy = 81.84%, test accuracy = 28.74%
2017-04-22 04:04:44.562536  iteration = 037, loss = 1.1754, val_loss = 5.1300, training accuracy = 81.92%, test accuracy = 28.78%
2017-04-22 04:47:28.251398  iteration = 038, loss = 1.1720, val_loss = 5.1304, training accuracy = 81.98%, test accuracy = 28.64%
2017-04-22 05:30:42.160048  iteration = 039, loss = 1.1730, val_loss = 5.1294, training accuracy = 81.86%, test accuracy = 28.80%
2017-04-22 06:13:45.385583  iteration = 040, loss = 1.1733, val_loss = 5.1296, training accuracy = 81.96%, test accuracy = 28.74%
2017-04-22 06:56:52.488307  iteration = 041, loss = 1.1726, val_loss = 5.1305, training accuracy = 82.02%, test accuracy = 28.68%
2017-04-22 07:39:38.282783  iteration = 042, loss = 1.1727, val_loss = 5.1294, training accuracy = 82.08%, test accuracy = 28.78%
2017-04-22 08:23:17.297073  iteration = 043, loss = 1.1715, val_loss = 5.1326, training accuracy = 82.06%, test accuracy = 28.72%
2017-04-22 09:07:20.876633  iteration = 044, loss = 1.1698, val_loss = 5.1344, training accuracy = 82.04%, test accuracy = 28.90%
2017-04-22 09:57:18.850847  iteration = 045, loss = 1.1702, val_loss = 5.1342, training accuracy = 82.10%, test accuracy = 28.78%
2017-04-22 10:53:37.235721  iteration = 046, loss = 1.1702, val_loss = 5.1348, training accuracy = 82.12%, test accuracy = 28.74%
2017-04-22 11:47:04.619690  iteration = 047, loss = 1.1705, val_loss = 5.1341, training accuracy = 82.14%, test accuracy = 28.80%
2017-04-22 12:36:38.829130  iteration = 048, loss = 1.1692, val_loss = 5.1340, training accuracy = 82.08%, test accuracy = 28.76%
2017-04-22 13:21:15.292909  iteration = 049, loss = 1.1684, val_loss = 5.1370, training accuracy = 82.12%, test accuracy = 28.74%
2017-04-22 14:04:20.522410  iteration = 050, loss = 1.1677, val_loss = 5.1371, training accuracy = 82.18%, test accuracy = 28.76%
2017-04-22 14:48:22.896566  iteration = 051, loss = 1.1677, val_loss = 5.1360, training accuracy = 82.20%, test accuracy = 28.90%
2017-04-22 15:32:33.290095  iteration = 052, loss = 1.1676, val_loss = 5.1370, training accuracy = 82.18%, test accuracy = 28.84%
2017-04-22 16:19:39.795404  iteration = 053, loss = 1.1663, val_loss = 5.1375, training accuracy = 82.24%, test accuracy = 28.90%
2017-04-22 17:03:11.290264  iteration = 054, loss = 1.1662, val_loss = 5.1394, training accuracy = 82.26%, test accuracy = 28.82%
2017-04-22 17:46:35.306699  iteration = 055, loss = 1.1660, val_loss = 5.1388, training accuracy = 82.32%, test accuracy = 28.86%
2017-04-22 18:29:47.517426  iteration = 056, loss = 1.1656, val_loss = 5.1400, training accuracy = 82.26%, test accuracy = 28.76%
2017-04-22 19:14:13.384295  iteration = 057, loss = 1.1668, val_loss = 5.1384, training accuracy = 82.34%, test accuracy = 28.86%
2017-04-22 19:57:25.952063  iteration = 058, loss = 1.1639, val_loss = 5.1410, training accuracy = 82.34%, test accuracy = 28.88%
2017-04-22 20:42:00.990478  iteration = 059, loss = 1.1661, val_loss = 5.1414, training accuracy = 82.32%, test accuracy = 28.78%
2017-04-22 21:25:24.336626  iteration = 060, loss = 1.1634, val_loss = 5.1416, training accuracy = 82.30%, test accuracy = 28.84%
2017-04-22 22:09:02.893257  iteration = 061, loss = 1.1613, val_loss = 5.1436, training accuracy = 82.18%, test accuracy = 28.84%
2017-04-22 22:51:56.898666  iteration = 062, loss = 1.1637, val_loss = 5.1426, training accuracy = 82.24%, test accuracy = 28.86%
2017-04-22 23:36:08.137567  iteration = 063, loss = 1.1618, val_loss = 5.1442, training accuracy = 82.38%, test accuracy = 28.88%
2017-04-23 00:20:13.335317  iteration = 064, loss = 1.1640, val_loss = 5.1442, training accuracy = 82.36%, test accuracy = 28.82%
2017-04-23 01:17:28.552999  iteration = 065, loss = 1.1620, val_loss = 5.1447, training accuracy = 82.34%, test accuracy = 28.88%
2017-04-23 02:16:13.890772  iteration = 066, loss = 1.1611, val_loss = 5.1443, training accuracy = 82.38%, test accuracy = 28.80%
2017-04-23 03:03:51.546380  iteration = 067, loss = 1.1605, val_loss = 5.1464, training accuracy = 82.34%, test accuracy = 28.88%
2017-04-23 03:46:54.705840  iteration = 068, loss = 1.1608, val_loss = 5.1471, training accuracy = 82.40%, test accuracy = 28.68%
2017-04-23 04:31:10.034472  iteration = 069, loss = 1.1606, val_loss = 5.1469, training accuracy = 82.40%, test accuracy = 28.74%
2017-04-23 05:28:02.271235  iteration = 070, loss = 1.1621, val_loss = 5.1474, training accuracy = 82.42%, test accuracy = 28.72%
2017-04-23 06:15:11.425440  iteration = 071, loss = 1.1596, val_loss = 5.1489, training accuracy = 82.50%, test accuracy = 28.70%
2017-04-23 06:58:22.400406  iteration = 072, loss = 1.1581, val_loss = 5.1495, training accuracy = 82.48%, test accuracy = 28.74%
2017-04-23 07:41:12.639360  iteration = 073, loss = 1.1587, val_loss = 5.1492, training accuracy = 82.48%, test accuracy = 28.70%
2017-04-23 08:25:03.868093  iteration = 074, loss = 1.1573, val_loss = 5.1516, training accuracy = 82.52%, test accuracy = 28.78%
2017-04-23 09:09:07.056746  iteration = 075, loss = 1.1592, val_loss = 5.1510, training accuracy = 82.54%, test accuracy = 28.84%
2017-04-23 10:06:17.232817  iteration = 076, loss = 1.1574, val_loss = 5.1516, training accuracy = 82.54%, test accuracy = 28.70%
2017-04-23 10:58:08.295653  iteration = 077, loss = 1.1568, val_loss = 5.1518, training accuracy = 82.52%, test accuracy = 28.76%
2017-04-23 11:46:48.940123  iteration = 078, loss = 1.1574, val_loss = 5.1521, training accuracy = 82.52%, test accuracy = 28.74%
2017-04-23 12:33:58.218035  iteration = 079, loss = 1.1565, val_loss = 5.1540, training accuracy = 82.62%, test accuracy = 28.78%
2017-04-23 13:17:55.726630  iteration = 080, loss = 1.1571, val_loss = 5.1540, training accuracy = 82.58%, test accuracy = 28.76%
2017-04-23 14:14:25.449800  iteration = 081, loss = 1.1545, val_loss = 5.1546, training accuracy = 82.68%, test accuracy = 28.80%
2017-04-23 15:02:26.002181  iteration = 082, loss = 1.1558, val_loss = 5.1548, training accuracy = 82.66%, test accuracy = 28.80%
2017-04-23 15:46:34.096507  iteration = 083, loss = 1.1537, val_loss = 5.1561, training accuracy = 82.72%, test accuracy = 28.78%
2017-04-23 16:29:45.417988  iteration = 084, loss = 1.1535, val_loss = 5.1553, training accuracy = 82.64%, test accuracy = 28.76%
2017-04-23 17:12:49.116435  iteration = 085, loss = 1.1528, val_loss = 5.1568, training accuracy = 82.70%, test accuracy = 28.78%
2017-04-23 17:56:27.821139  iteration = 086, loss = 1.1554, val_loss = 5.1557, training accuracy = 82.78%, test accuracy = 28.82%
2017-04-23 18:39:35.867415  iteration = 087, loss = 1.1533, val_loss = 5.1573, training accuracy = 82.74%, test accuracy = 28.70%
2017-04-23 19:22:25.745264  iteration = 088, loss = 1.1532, val_loss = 5.1576, training accuracy = 82.78%, test accuracy = 28.74%
2017-04-23 20:05:16.621153  iteration = 089, loss = 1.1500, val_loss = 5.1590, training accuracy = 82.74%, test accuracy = 28.80%
2017-04-23 20:48:03.357114  iteration = 090, loss = 1.1529, val_loss = 5.1599, training accuracy = 82.70%, test accuracy = 28.74%
2017-04-23 21:31:59.551281  iteration = 091, loss = 1.1518, val_loss = 5.1587, training accuracy = 82.74%, test accuracy = 28.78%
2017-04-23 22:27:48.238377  iteration = 092, loss = 1.1513, val_loss = 5.1603, training accuracy = 82.74%, test accuracy = 28.76%
2017-04-23 23:27:39.654618  iteration = 093, loss = 1.1513, val_loss = 5.1607, training accuracy = 82.72%, test accuracy = 28.72%
2017-04-24 00:14:31.750402  iteration = 094, loss = 1.1493, val_loss = 5.1626, training accuracy = 82.74%, test accuracy = 28.66%
2017-04-24 00:58:11.435055  iteration = 095, loss = 1.1493, val_loss = 5.1623, training accuracy = 82.82%, test accuracy = 28.72%
2017-04-24 01:42:12.256931  iteration = 096, loss = 1.1509, val_loss = 5.1620, training accuracy = 82.80%, test accuracy = 28.70%
2017-04-24 02:27:47.077361  iteration = 097, loss = 1.1481, val_loss = 5.1617, training accuracy = 82.82%, test accuracy = 28.72%
2017-04-24 03:15:45.507341  iteration = 098, loss = 1.1495, val_loss = 5.1635, training accuracy = 82.82%, test accuracy = 28.66%
2017-04-24 03:59:11.097818  iteration = 099, loss = 1.1488, val_loss = 5.1638, training accuracy = 82.86%, test accuracy = 28.72%
2017-04-24 04:42:27.380777  iteration = 100, loss = 1.1480, val_loss = 5.1631, training accuracy = 82.94%, test accuracy = 28.72%
2017-04-24 05:25:39.814076  iteration = 101, loss = 1.1489, val_loss = 5.1649, training accuracy = 82.80%, test accuracy = 28.70%
2017-04-24 06:08:36.962208  iteration = 102, loss = 1.1485, val_loss = 5.1652, training accuracy = 82.78%, test accuracy = 28.68%
2017-04-24 06:51:30.843094  iteration = 103, loss = 1.1477, val_loss = 5.1654, training accuracy = 82.86%, test accuracy = 28.76%
2017-04-24 07:35:34.897545  iteration = 104, loss = 1.1479, val_loss = 5.1662, training accuracy = 82.86%, test accuracy = 28.78%
2017-04-24 08:19:39.512396  iteration = 105, loss = 1.1463, val_loss = 5.1667, training accuracy = 82.92%, test accuracy = 28.80%
2017-04-24 09:06:51.232773  iteration = 106, loss = 1.1453, val_loss = 5.1679, training accuracy = 82.88%, test accuracy = 28.72%
2017-04-24 09:53:38.619897  iteration = 107, loss = 1.1459, val_loss = 5.1688, training accuracy = 83.04%, test accuracy = 28.82%
2017-04-24 10:38:16.293783  iteration = 108, loss = 1.1457, val_loss = 5.1682, training accuracy = 83.06%, test accuracy = 28.70%
2017-04-24 11:21:58.976752  iteration = 109, loss = 1.1448, val_loss = 5.1697, training accuracy = 82.98%, test accuracy = 28.64%
2017-04-24 12:05:44.880678  iteration = 110, loss = 1.1440, val_loss = 5.1700, training accuracy = 83.06%, test accuracy = 28.80%
2017-04-24 12:49:58.586698  iteration = 111, loss = 1.1447, val_loss = 5.1701, training accuracy = 83.04%, test accuracy = 28.78%
2017-04-24 13:33:07.388283  iteration = 112, loss = 1.1439, val_loss = 5.1699, training accuracy = 83.08%, test accuracy = 28.80%
2017-04-24 14:18:06.525694  iteration = 113, loss = 1.1422, val_loss = 5.1719, training accuracy = 83.10%, test accuracy = 28.76%
2017-04-24 15:02:24.212619  iteration = 114, loss = 1.1431, val_loss = 5.1726, training accuracy = 83.16%, test accuracy = 28.74%
2017-04-24 15:45:36.545731  iteration = 115, loss = 1.1410, val_loss = 5.1726, training accuracy = 83.16%, test accuracy = 28.78%
2017-04-24 16:28:29.040996  iteration = 116, loss = 1.1428, val_loss = 5.1736, training accuracy = 83.12%, test accuracy = 28.74%
2017-04-24 17:12:42.965119  iteration = 117, loss = 1.1398, val_loss = 5.1762, training accuracy = 83.12%, test accuracy = 28.74%
2017-04-24 17:56:50.365319  iteration = 118, loss = 1.1418, val_loss = 5.1743, training accuracy = 83.18%, test accuracy = 28.80%
2017-04-24 18:44:17.721986  iteration = 119, loss = 1.1396, val_loss = 5.1761, training accuracy = 83.12%, test accuracy = 28.70%
2017-04-24 19:34:23.149115  iteration = 120, loss = 1.1418, val_loss = 5.1767, training accuracy = 83.22%, test accuracy = 28.80%
2017-04-24 20:21:45.548150  iteration = 121, loss = 1.1394, val_loss = 5.1774, training accuracy = 83.30%, test accuracy = 28.80%
2017-04-24 21:07:58.391411  iteration = 122, loss = 1.1398, val_loss = 5.1767, training accuracy = 83.22%, test accuracy = 28.86%
2017-04-24 21:53:28.181802  iteration = 123, loss = 1.1385, val_loss = 5.1771, training accuracy = 83.28%, test accuracy = 28.88%
2017-04-24 22:38:04.763452  iteration = 124, loss = 1.1393, val_loss = 5.1769, training accuracy = 83.30%, test accuracy = 28.80%
2017-04-24 23:23:17.258198  iteration = 125, loss = 1.1388, val_loss = 5.1787, training accuracy = 83.28%, test accuracy = 28.92%
2017-04-25 00:07:44.478233  iteration = 126, loss = 1.1367, val_loss = 5.1802, training accuracy = 83.28%, test accuracy = 28.80%
2017-04-25 00:50:34.666732  iteration = 127, loss = 1.1381, val_loss = 5.1790, training accuracy = 83.38%, test accuracy = 28.76%
2017-04-25 01:34:43.593517  iteration = 128, loss = 1.1360, val_loss = 5.1806, training accuracy = 83.30%, test accuracy = 28.82%
2017-04-25 02:18:38.610202  iteration = 129, loss = 1.1363, val_loss = 5.1818, training accuracy = 83.34%, test accuracy = 28.76%
2017-04-25 03:15:01.108800  iteration = 130, loss = 1.1358, val_loss = 5.1820, training accuracy = 83.40%, test accuracy = 28.88%
2017-04-25 04:11:44.848561  iteration = 131, loss = 1.1366, val_loss = 5.1810, training accuracy = 83.30%, test accuracy = 28.78%
2017-04-25 04:59:29.172419  iteration = 132, loss = 1.1350, val_loss = 5.1835, training accuracy = 83.42%, test accuracy = 28.78%
2017-04-25 05:42:18.457621  iteration = 133, loss = 1.1345, val_loss = 5.1848, training accuracy = 83.30%, test accuracy = 28.70%
2017-04-25 06:26:24.180712  iteration = 134, loss = 1.1340, val_loss = 5.1845, training accuracy = 83.38%, test accuracy = 28.80%
2017-04-25 07:10:23.841252  iteration = 135, loss = 1.1350, val_loss = 5.1848, training accuracy = 83.38%, test accuracy = 28.78%
2017-04-25 08:06:37.803252  iteration = 136, loss = 1.1335, val_loss = 5.1864, training accuracy = 83.30%, test accuracy = 28.78%
2017-04-25 08:57:21.130896  iteration = 137, loss = 1.1336, val_loss = 5.1844, training accuracy = 83.46%, test accuracy = 28.74%
2017-04-25 09:44:24.797476  iteration = 138, loss = 1.1340, val_loss = 5.1868, training accuracy = 83.48%, test accuracy = 28.80%
2017-04-25 10:31:24.527149  iteration = 139, loss = 1.1340, val_loss = 5.1872, training accuracy = 83.54%, test accuracy = 28.76%
2017-04-25 11:15:30.949067  iteration = 140, loss = 1.1330, val_loss = 5.1864, training accuracy = 83.42%, test accuracy = 28.80%
2017-04-25 11:58:13.968117  iteration = 141, loss = 1.1315, val_loss = 5.1862, training accuracy = 83.48%, test accuracy = 28.68%
2017-04-25 12:41:54.825376  iteration = 142, loss = 1.1320, val_loss = 5.1862, training accuracy = 83.50%, test accuracy = 28.72%
2017-04-25 13:31:30.079578  iteration = 143, loss = 1.1320, val_loss = 5.1887, training accuracy = 83.50%, test accuracy = 28.64%
2017-04-25 14:10:13.765924  iteration = 144, loss = 1.1332, val_loss = 5.1880, training accuracy = 83.62%, test accuracy = 28.62%
2017-04-25 14:49:16.481138  iteration = 145, loss = 1.1319, val_loss = 5.1901, training accuracy = 83.56%, test accuracy = 28.64%
2017-04-25 15:32:28.057313  iteration = 146, loss = 1.1296, val_loss = 5.1896, training accuracy = 83.60%, test accuracy = 28.70%
2017-04-25 16:14:15.340974  iteration = 147, loss = 1.1304, val_loss = 5.1901, training accuracy = 83.56%, test accuracy = 28.72%
2017-04-25 16:55:42.959423  iteration = 148, loss = 1.1305, val_loss = 5.1911, training accuracy = 83.64%, test accuracy = 28.62%
2017-04-25 17:38:04.894166  iteration = 149, loss = 1.1289, val_loss = 5.1936, training accuracy = 83.56%, test accuracy = 28.72%
2017-04-25 18:20:40.297276  iteration = 150, loss = 1.1284, val_loss = 5.1922, training accuracy = 83.58%, test accuracy = 28.70%
2017-04-25 19:03:13.871684  iteration = 151, loss = 1.1284, val_loss = 5.1930, training accuracy = 83.56%, test accuracy = 28.72%
2017-04-25 19:45:58.730403  iteration = 152, loss = 1.1271, val_loss = 5.1933, training accuracy = 83.58%, test accuracy = 28.62%
2017-04-25 20:28:37.339842  iteration = 153, loss = 1.1298, val_loss = 5.1933, training accuracy = 83.60%, test accuracy = 28.62%
2017-04-25 21:11:10.606173  iteration = 154, loss = 1.1274, val_loss = 5.1947, training accuracy = 83.62%, test accuracy = 28.70%
2017-04-25 21:54:53.938387  iteration = 155, loss = 1.1279, val_loss = 5.1942, training accuracy = 83.60%, test accuracy = 28.62%
2017-04-25 22:38:32.609222  iteration = 156, loss = 1.1265, val_loss = 5.1954, training accuracy = 83.60%, test accuracy = 28.68%
2017-04-25 23:22:05.569769  iteration = 157, loss = 1.1254, val_loss = 5.1968, training accuracy = 83.68%, test accuracy = 28.60%
2017-04-26 00:05:18.669997  iteration = 158, loss = 1.1259, val_loss = 5.1986, training accuracy = 83.74%, test accuracy = 28.60%
2017-04-26 00:48:17.984582  iteration = 159, loss = 1.1282, val_loss = 5.1976, training accuracy = 83.66%, test accuracy = 28.80%
2017-04-26 01:31:12.711139  iteration = 160, loss = 1.1257, val_loss = 5.1976, training accuracy = 83.68%, test accuracy = 28.64%
2017-04-26 02:14:03.513253  iteration = 161, loss = 1.1238, val_loss = 5.1989, training accuracy = 83.70%, test accuracy = 28.62%
2017-04-26 02:58:07.172794  iteration = 162, loss = 1.1249, val_loss = 5.1980, training accuracy = 83.72%, test accuracy = 28.58%
2017-04-26 03:45:27.560200  iteration = 163, loss = 1.1230, val_loss = 5.1997, training accuracy = 83.70%, test accuracy = 28.70%
2017-04-26 04:31:49.253553  iteration = 164, loss = 1.1246, val_loss = 5.1987, training accuracy = 83.76%, test accuracy = 28.64%
2017-04-26 05:15:26.004239  iteration = 165, loss = 1.1240, val_loss = 5.1991, training accuracy = 83.70%, test accuracy = 28.64%
2017-04-26 05:58:57.398603  iteration = 166, loss = 1.1220, val_loss = 5.2009, training accuracy = 83.76%, test accuracy = 28.64%
2017-04-26 06:42:15.212880  iteration = 167, loss = 1.1229, val_loss = 5.2009, training accuracy = 83.80%, test accuracy = 28.66%
2017-04-26 07:25:16.506377  iteration = 168, loss = 1.1223, val_loss = 5.2017, training accuracy = 83.88%, test accuracy = 28.62%
2017-04-26 08:08:23.315531  iteration = 169, loss = 1.1223, val_loss = 5.2024, training accuracy = 83.94%, test accuracy = 28.66%
2017-04-26 08:51:29.294726  iteration = 170, loss = 1.1201, val_loss = 5.2031, training accuracy = 83.94%, test accuracy = 28.68%
2017-04-26 09:34:18.558195  iteration = 171, loss = 1.1201, val_loss = 5.2024, training accuracy = 84.00%, test accuracy = 28.62%
2017-04-26 10:17:13.206773  iteration = 172, loss = 1.1199, val_loss = 5.2052, training accuracy = 83.90%, test accuracy = 28.60%
2017-04-26 11:00:02.749965  iteration = 173, loss = 1.1204, val_loss = 5.2050, training accuracy = 83.94%, test accuracy = 28.70%
2017-04-26 11:47:21.718163  iteration = 174, loss = 1.1207, val_loss = 5.2046, training accuracy = 83.82%, test accuracy = 28.72%
2017-04-26 12:34:38.528173  iteration = 175, loss = 1.1195, val_loss = 5.2067, training accuracy = 83.86%, test accuracy = 28.62%
2017-04-26 13:19:31.764763  iteration = 176, loss = 1.1194, val_loss = 5.2056, training accuracy = 83.86%, test accuracy = 28.56%
2017-04-26 13:58:33.464244  iteration = 177, loss = 1.1196, val_loss = 5.2074, training accuracy = 83.90%, test accuracy = 28.70%
2017-04-26 14:39:31.439031  iteration = 178, loss = 1.1197, val_loss = 5.2073, training accuracy = 83.88%, test accuracy = 28.68%
2017-04-26 15:20:32.090167  iteration = 179, loss = 1.1184, val_loss = 5.2090, training accuracy = 83.88%, test accuracy = 28.62%
2017-04-26 15:55:58.733825  iteration = 180, loss = 1.1179, val_loss = 5.2091, training accuracy = 83.96%, test accuracy = 28.64%
2017-04-26 16:31:31.590174  iteration = 181, loss = 1.1186, val_loss = 5.2080, training accuracy = 84.02%, test accuracy = 28.64%
2017-04-26 17:06:12.911770  iteration = 182, loss = 1.1183, val_loss = 5.2106, training accuracy = 84.08%, test accuracy = 28.64%
2017-04-26 17:40:24.650014  iteration = 183, loss = 1.1173, val_loss = 5.2101, training accuracy = 84.02%, test accuracy = 28.66%
2017-04-26 18:17:17.370205  iteration = 184, loss = 1.1169, val_loss = 5.2095, training accuracy = 84.04%, test accuracy = 28.64%
2017-04-26 18:59:48.346633  iteration = 185, loss = 1.1161, val_loss = 5.2131, training accuracy = 84.00%, test accuracy = 28.64%
2017-04-26 19:42:18.700957  iteration = 186, loss = 1.1153, val_loss = 5.2128, training accuracy = 84.02%, test accuracy = 28.62%
2017-04-26 20:24:48.751079  iteration = 187, loss = 1.1161, val_loss = 5.2121, training accuracy = 83.98%, test accuracy = 28.72%
2017-04-26 21:07:15.623629  iteration = 188, loss = 1.1153, val_loss = 5.2138, training accuracy = 84.00%, test accuracy = 28.68%
2017-04-26 21:49:43.953165  iteration = 189, loss = 1.1137, val_loss = 5.2141, training accuracy = 84.06%, test accuracy = 28.64%
2017-04-26 22:32:07.696483  iteration = 190, loss = 1.1143, val_loss = 5.2126, training accuracy = 84.10%, test accuracy = 28.60%
2017-04-26 23:16:26.579903  iteration = 191, loss = 1.1145, val_loss = 5.2148, training accuracy = 84.08%, test accuracy = 28.62%
2017-04-27 00:00:57.002193  iteration = 192, loss = 1.1145, val_loss = 5.2158, training accuracy = 84.08%, test accuracy = 28.68%
2017-04-27 00:45:00.018508  iteration = 193, loss = 1.1132, val_loss = 5.2161, training accuracy = 84.10%, test accuracy = 28.64%
2017-04-27 01:28:40.562877  iteration = 194, loss = 1.1135, val_loss = 5.2157, training accuracy = 84.04%, test accuracy = 28.64%
2017-04-27 02:12:05.162562  iteration = 195, loss = 1.1117, val_loss = 5.2167, training accuracy = 84.16%, test accuracy = 28.64%
2017-04-27 02:55:27.680344  iteration = 196, loss = 1.1124, val_loss = 5.2179, training accuracy = 84.14%, test accuracy = 28.52%
2017-04-27 03:38:40.009641  iteration = 197, loss = 1.1113, val_loss = 5.2183, training accuracy = 84.14%, test accuracy = 28.54%
2017-04-27 04:22:11.270109  iteration = 198, loss = 1.1128, val_loss = 5.2181, training accuracy = 84.18%, test accuracy = 28.50%
2017-04-27 05:05:09.651766  iteration = 199, loss = 1.1103, val_loss = 5.2191, training accuracy = 84.16%, test accuracy = 28.60%

Continuing training with a further run
sjb@persephone:sjb_word_model$ ./train.py theresa_may_contributions.json
2017-05-08 20:43:20.497287  iteration = 000, loss = 1.1120, val_loss = 5.2198, training accuracy = 84.02%, test accuracy = 28.60%
2017-05-08 21:32:43.185121  iteration = 001, loss = 1.1100, val_loss = 5.2196, training accuracy = 84.10%, test accuracy = 28.66%
2017-05-08 22:19:02.764886  iteration = 002, loss = 1.1087, val_loss = 5.2201, training accuracy = 84.20%, test accuracy = 28.58%
2017-05-08 23:12:03.922208  iteration = 003, loss = 1.1108, val_loss = 5.2208, training accuracy = 84.20%, test accuracy = 28.62%
2017-05-09 00:01:40.845772  iteration = 004, loss = 1.1086, val_loss = 5.2224, training accuracy = 84.12%, test accuracy = 28.58%
2017-05-09 00:47:16.383430  iteration = 005, loss = 1.1076, val_loss = 5.2229, training accuracy = 84.14%, test accuracy = 28.60%
2017-05-09 01:33:06.300147  iteration = 006, loss = 1.1099, val_loss = 5.2228, training accuracy = 84.12%, test accuracy = 28.56%
2017-05-09 02:23:44.704949  iteration = 007, loss = 1.1095, val_loss = 5.2223, training accuracy = 84.14%, test accuracy = 28.58%
2017-05-09 03:16:50.013370  iteration = 008, loss = 1.1074, val_loss = 5.2228, training accuracy = 84.20%, test accuracy = 28.54%
2017-05-09 04:02:51.489483  iteration = 009, loss = 1.1074, val_loss = 5.2243, training accuracy = 84.18%, test accuracy = 28.58%
2017-05-09 04:46:49.954621  iteration = 010, loss = 1.1076, val_loss = 5.2241, training accuracy = 84.24%, test accuracy = 28.68%
2017-05-09 05:29:52.330280  iteration = 011, loss = 1.1070, val_loss = 5.2232, training accuracy = 84.20%, test accuracy = 28.66%
2017-05-09 06:13:58.037605  iteration = 012, loss = 1.1077, val_loss = 5.2245, training accuracy = 84.24%, test accuracy = 28.60%
2017-05-09 06:58:15.797566  iteration = 013, loss = 1.1052, val_loss = 5.2262, training accuracy = 84.28%, test accuracy = 28.64%
2017-05-09 07:49:05.199644  iteration = 014, loss = 1.1060, val_loss = 5.2260, training accuracy = 84.20%, test accuracy = 28.64%
2017-05-09 08:36:56.356267  iteration = 015, loss = 1.1062, val_loss = 5.2268, training accuracy = 84.24%, test accuracy = 28.60%
2017-05-09 09:20:22.078965  iteration = 016, loss = 1.1066, val_loss = 5.2276, training accuracy = 84.26%, test accuracy = 28.62%
2017-05-09 10:03:45.054693  iteration = 017, loss = 1.1051, val_loss = 5.2277, training accuracy = 84.28%, test accuracy = 28.62%
2017-05-09 10:47:13.104192  iteration = 018, loss = 1.1049, val_loss = 5.2283, training accuracy = 84.32%, test accuracy = 28.62%
2017-05-09 11:28:45.337638  iteration = 019, loss = 1.1046, val_loss = 5.2291, training accuracy = 84.30%, test accuracy = 28.66%
2017-05-09 12:21:05.952354  iteration = 020, loss = 1.1022, val_loss = 5.2285, training accuracy = 84.30%, test accuracy = 28.60%
2017-05-09 13:01:59.009123  iteration = 021, loss = 1.1055, val_loss = 5.2288, training accuracy = 84.30%, test accuracy = 28.60%
2017-05-09 13:57:35.560798  iteration = 022, loss = 1.1034, val_loss = 5.2306, training accuracy = 84.36%, test accuracy = 28.58%
2017-05-09 14:38:09.606132  iteration = 023, loss = 1.1020, val_loss = 5.2300, training accuracy = 84.34%, test accuracy = 28.60%
2017-05-09 15:22:18.909710  iteration = 024, loss = 1.1016, val_loss = 5.2305, training accuracy = 84.24%, test accuracy = 28.54%
2017-05-09 16:03:52.748299  iteration = 025, loss = 1.1035, val_loss = 5.2326, training accuracy = 84.26%, test accuracy = 28.58%
2017-05-09 16:46:47.544785  iteration = 026, loss = 1.1010, val_loss = 5.2324, training accuracy = 84.32%, test accuracy = 28.50%
2017-05-09 17:30:39.872227  iteration = 027, loss = 1.1026, val_loss = 5.2330, training accuracy = 84.38%, test accuracy = 28.58%
2017-05-09 18:14:26.552760  iteration = 028, loss = 1.1025, val_loss = 5.2333, training accuracy = 84.34%, test accuracy = 28.60%
2017-05-09 18:57:47.452295  iteration = 029, loss = 1.1012, val_loss = 5.2340, training accuracy = 84.38%, test accuracy = 28.52%
2017-05-09 19:41:23.107300  iteration = 030, loss = 1.0998, val_loss = 5.2340, training accuracy = 84.36%, test accuracy = 28.48%
2017-05-09 20:24:53.766366  iteration = 031, loss = 1.0999, val_loss = 5.2335, training accuracy = 84.40%, test accuracy = 28.50%
2017-05-09 21:07:52.879579  iteration = 032, loss = 1.1000, val_loss = 5.2362, training accuracy = 84.30%, test accuracy = 28.60%
2017-05-09 21:51:55.144411  iteration = 033, loss = 1.0998, val_loss = 5.2350, training accuracy = 84.38%, test accuracy = 28.58%
2017-05-09 22:36:19.694483  iteration = 034, loss = 1.1001, val_loss = 5.2364, training accuracy = 84.38%, test accuracy = 28.54%
2017-05-09 23:32:54.786582  iteration = 035, loss = 1.0983, val_loss = 5.2369, training accuracy = 84.36%, test accuracy = 28.58%
2017-05-10 00:26:08.089464  iteration = 036, loss = 1.0979, val_loss = 5.2371, training accuracy = 84.38%, test accuracy = 28.54%
2017-05-10 01:17:31.794222  iteration = 037, loss = 1.0987, val_loss = 5.2392, training accuracy = 84.48%, test accuracy = 28.50%
2017-05-10 02:06:28.639071  iteration = 038, loss = 1.0976, val_loss = 5.2388, training accuracy = 84.42%, test accuracy = 28.56%
2017-05-10 02:53:18.602832  iteration = 039, loss = 1.0962, val_loss = 5.2400, training accuracy = 84.40%, test accuracy = 28.64%
2017-05-10 03:37:43.630366  iteration = 040, loss = 1.0958, val_loss = 5.2402, training accuracy = 84.44%, test accuracy = 28.58%
2017-05-10 04:22:00.340620  iteration = 041, loss = 1.0963, val_loss = 5.2408, training accuracy = 84.42%, test accuracy = 28.48%
2017-05-10 05:08:55.958843  iteration = 042, loss = 1.0958, val_loss = 5.2413, training accuracy = 84.40%, test accuracy = 28.60%
2017-05-10 06:05:28.484944  iteration = 043, loss = 1.0966, val_loss = 5.2401, training accuracy = 84.52%, test accuracy = 28.52%
2017-05-10 06:57:58.992564  iteration = 044, loss = 1.0938, val_loss = 5.2421, training accuracy = 84.48%, test accuracy = 28.54%
2017-05-10 07:47:38.082399  iteration = 045, loss = 1.0961, val_loss = 5.2421, training accuracy = 84.48%, test accuracy = 28.44%
2017-05-10 08:34:26.464760  iteration = 046, loss = 1.0944, val_loss = 5.2431, training accuracy = 84.40%, test accuracy = 28.42%
2017-05-10 09:18:44.901081  iteration = 047, loss = 1.0920, val_loss = 5.2434, training accuracy = 84.50%, test accuracy = 28.46%
2017-05-10 10:01:52.270430  iteration = 048, loss = 1.0952, val_loss = 5.2431, training accuracy = 84.62%, test accuracy = 28.44%
2017-05-10 10:46:23.635609  iteration = 049, loss = 1.0935, val_loss = 5.2448, training accuracy = 84.66%, test accuracy = 28.42%
2017-05-10 11:28:06.284030  iteration = 050, loss = 1.0943, val_loss = 5.2445, training accuracy = 84.60%, test accuracy = 28.46%
2017-05-10 12:16:40.943531  iteration = 051, loss = 1.0930, val_loss = 5.2452, training accuracy = 84.64%, test accuracy = 28.48%
2017-05-10 13:12:46.187842  iteration = 052, loss = 1.0937, val_loss = 5.2454, training accuracy = 84.64%, test accuracy = 28.48%
2017-05-10 13:59:58.566047  iteration = 053, loss = 1.0915, val_loss = 5.2476, training accuracy = 84.62%, test accuracy = 28.52%
2017-05-10 14:45:35.037609  iteration = 054, loss = 1.0919, val_loss = 5.2476, training accuracy = 84.70%, test accuracy = 28.58%
2017-05-10 15:29:57.088177  iteration = 055, loss = 1.0909, val_loss = 5.2471, training accuracy = 84.64%, test accuracy = 28.42%
2017-05-10 16:43:37.057468  iteration = 056, loss = 1.0911, val_loss = 5.2482, training accuracy = 84.62%, test accuracy = 28.50%
2017-05-10 17:30:29.694909  iteration = 057, loss = 1.0896, val_loss = 5.2501, training accuracy = 84.66%, test accuracy = 28.52%
2017-05-10 18:16:07.152132  iteration = 058, loss = 1.0928, val_loss = 5.2480, training accuracy = 84.70%, test accuracy = 28.44%
2017-05-10 19:00:25.519606  iteration = 059, loss = 1.0905, val_loss = 5.2497, training accuracy = 84.62%, test accuracy = 28.56%
2017-05-10 19:44:45.615809  iteration = 060, loss = 1.0891, val_loss = 5.2488, training accuracy = 84.72%, test accuracy = 28.46%
2017-05-10 20:35:46.793888  iteration = 061, loss = 1.0891, val_loss = 5.2514, training accuracy = 84.70%, test accuracy = 28.46%
2017-05-10 21:24:31.939862  iteration = 062, loss = 1.0903, val_loss = 5.2507, training accuracy = 84.66%, test accuracy = 28.46%
2017-05-10 22:08:41.211482  iteration = 063, loss = 1.0878, val_loss = 5.2524, training accuracy = 84.68%, test accuracy = 28.48%
2017-05-10 22:52:48.113866  iteration = 064, loss = 1.0890, val_loss = 5.2523, training accuracy = 84.76%, test accuracy = 28.52%
2017-05-10 23:45:14.255710  iteration = 065, loss = 1.0892, val_loss = 5.2520, training accuracy = 84.74%, test accuracy = 28.46%
2017-05-11 00:33:55.009319  iteration = 066, loss = 1.0888, val_loss = 5.2513, training accuracy = 84.80%, test accuracy = 28.50%
2017-05-11 01:19:51.415776  iteration = 067, loss = 1.0879, val_loss = 5.2543, training accuracy = 84.80%, test accuracy = 28.44%
2017-05-11 02:05:27.174019  iteration = 068, loss = 1.0861, val_loss = 5.2554, training accuracy = 84.76%, test accuracy = 28.54%
2017-05-11 02:49:45.883760  iteration = 069, loss = 1.0864, val_loss = 5.2542, training accuracy = 84.80%, test accuracy = 28.52%
2017-05-11 03:41:43.703296  iteration = 070, loss = 1.0862, val_loss = 5.2535, training accuracy = 84.84%, test accuracy = 28.48%
2017-05-11 04:29:04.420451  iteration = 071, loss = 1.0866, val_loss = 5.2553, training accuracy = 84.76%, test accuracy = 28.54%
2017-05-11 05:15:01.036389  iteration = 072, loss = 1.0868, val_loss = 5.2546, training accuracy = 84.80%, test accuracy = 28.48%
2017-05-11 06:02:09.694348  iteration = 073, loss = 1.0844, val_loss = 5.2575, training accuracy = 84.82%, test accuracy = 28.50%
2017-05-11 06:54:08.376563  iteration = 074, loss = 1.0855, val_loss = 5.2563, training accuracy = 84.86%, test accuracy = 28.44%
2017-05-11 07:41:21.877861  iteration = 075, loss = 1.0835, val_loss = 5.2574, training accuracy = 84.88%, test accuracy = 28.58%
2017-05-11 08:25:28.677693  iteration = 076, loss = 1.0842, val_loss = 5.2575, training accuracy = 84.90%, test accuracy = 28.56%
2017-05-11 09:10:23.249620  iteration = 077, loss = 1.0833, val_loss = 5.2576, training accuracy = 84.90%, test accuracy = 28.48%
2017-05-11 09:54:57.220968  iteration = 078, loss = 1.0822, val_loss = 5.2587, training accuracy = 84.94%, test accuracy = 28.46%
2017-05-11 10:37:54.479904  iteration = 079, loss = 1.0840, val_loss = 5.2592, training accuracy = 84.86%, test accuracy = 28.46%
2017-05-11 11:22:24.931226  iteration = 080, loss = 1.0818, val_loss = 5.2601, training accuracy = 84.90%, test accuracy = 28.36%
2017-05-11 12:15:55.145823  iteration = 081, loss = 1.0834, val_loss = 5.2613, training accuracy = 84.90%, test accuracy = 28.46%
2017-05-11 13:03:37.816875  iteration = 082, loss = 1.0834, val_loss = 5.2599, training accuracy = 84.90%, test accuracy = 28.50%
2017-05-11 13:48:14.318173  iteration = 083, loss = 1.0818, val_loss = 5.2613, training accuracy = 84.96%, test accuracy = 28.54%
2017-05-11 14:32:35.003409  iteration = 084, loss = 1.0824, val_loss = 5.2615, training accuracy = 84.96%, test accuracy = 28.46%
2017-05-11 15:17:01.668392  iteration = 085, loss = 1.0806, val_loss = 5.2617, training accuracy = 84.94%, test accuracy = 28.46%
2017-05-11 16:00:14.174260  iteration = 086, loss = 1.0813, val_loss = 5.2634, training accuracy = 84.92%, test accuracy = 28.54%
2017-05-11 16:43:47.280873  iteration = 087, loss = 1.0799, val_loss = 5.2630, training accuracy = 85.00%, test accuracy = 28.48%
2017-05-11 17:28:09.824163  iteration = 088, loss = 1.0800, val_loss = 5.2634, training accuracy = 85.04%, test accuracy = 28.44%
2017-05-11 18:21:11.816555  iteration = 089, loss = 1.0803, val_loss = 5.2634, training accuracy = 85.10%, test accuracy = 28.40%
2017-05-11 19:10:15.321850  iteration = 090, loss = 1.0798, val_loss = 5.2650, training accuracy = 85.12%, test accuracy = 28.44%
2017-05-11 19:55:37.633598  iteration = 091, loss = 1.0812, val_loss = 5.2654, training accuracy = 85.16%, test accuracy = 28.50%
2017-05-11 20:40:51.752245  iteration = 092, loss = 1.0782, val_loss = 5.2667, training accuracy = 85.08%, test accuracy = 28.44%
2017-05-11 21:24:31.552083  iteration = 093, loss = 1.0798, val_loss = 5.2665, training accuracy = 85.10%, test accuracy = 28.50%
2017-05-11 22:08:38.462506  iteration = 094, loss = 1.0785, val_loss = 5.2665, training accuracy = 85.12%, test accuracy = 28.40%
2017-05-11 22:59:49.913691  iteration = 095, loss = 1.0782, val_loss = 5.2674, training accuracy = 85.06%, test accuracy = 28.42%
2017-05-11 23:52:04.942188  iteration = 096, loss = 1.0764, val_loss = 5.2683, training accuracy = 85.12%, test accuracy = 28.46%
2017-05-12 00:38:23.445817  iteration = 097, loss = 1.0801, val_loss = 5.2673, training accuracy = 85.14%, test accuracy = 28.34%
2017-05-12 01:23:08.357053  iteration = 098, loss = 1.0784, val_loss = 5.2685, training accuracy = 85.24%, test accuracy = 28.42%
2017-05-12 02:07:15.146781  iteration = 099, loss = 1.0780, val_loss = 5.2687, training accuracy = 85.16%, test accuracy = 28.38%
2017-05-12 02:57:17.563597  iteration = 100, loss = 1.0761, val_loss = 5.2683, training accuracy = 85.16%, test accuracy = 28.38%
2017-05-12 03:48:50.555081  iteration = 101, loss = 1.0764, val_loss = 5.2699, training accuracy = 85.32%, test accuracy = 28.48%
2017-05-12 04:36:21.022385  iteration = 102, loss = 1.0761, val_loss = 5.2724, training accuracy = 85.20%, test accuracy = 28.46%
2017-05-12 05:20:55.243835  iteration = 103, loss = 1.0755, val_loss = 5.2721, training accuracy = 85.22%, test accuracy = 28.50%
2017-05-12 06:05:15.443737  iteration = 104, loss = 1.0752, val_loss = 5.2720, training accuracy = 85.18%, test accuracy = 28.52%
2017-05-12 06:57:18.136587  iteration = 105, loss = 1.0752, val_loss = 5.2731, training accuracy = 85.18%, test accuracy = 28.46%
2017-05-12 07:47:06.845344  iteration = 106, loss = 1.0739, val_loss = 5.2737, training accuracy = 85.24%, test accuracy = 28.46%
2017-05-12 08:35:07.993349  iteration = 107, loss = 1.0746, val_loss = 5.2724, training accuracy = 85.24%, test accuracy = 28.44%
2017-05-12 09:21:12.870444  iteration = 108, loss = 1.0736, val_loss = 5.2738, training accuracy = 85.24%, test accuracy = 28.44%
2017-05-12 10:05:23.196759  iteration = 109, loss = 1.0728, val_loss = 5.2747, training accuracy = 85.36%, test accuracy = 28.38%
2017-05-12 10:55:58.613069  iteration = 110, loss = 1.0741, val_loss = 5.2750, training accuracy = 85.30%, test accuracy = 28.48%
2017-05-12 11:48:04.173409  iteration = 111, loss = 1.0747, val_loss = 5.2750, training accuracy = 85.16%, test accuracy = 28.48%
2017-05-12 12:33:37.540833  iteration = 112, loss = 1.0726, val_loss = 5.2767, training accuracy = 85.32%, test accuracy = 28.38%
2017-05-12 13:17:41.253433  iteration = 113, loss = 1.0730, val_loss = 5.2760, training accuracy = 85.28%, test accuracy = 28.42%
2017-05-12 14:01:54.117830  iteration = 114, loss = 1.0717, val_loss = 5.2770, training accuracy = 85.32%, test accuracy = 28.40%
2017-05-12 14:53:05.328918  iteration = 115, loss = 1.0729, val_loss = 5.2766, training accuracy = 85.34%, test accuracy = 28.42%
2017-05-12 15:40:48.724085  iteration = 116, loss = 1.0712, val_loss = 5.2784, training accuracy = 85.28%, test accuracy = 28.46%
2017-05-12 16:25:00.414251  iteration = 117, loss = 1.0719, val_loss = 5.2774, training accuracy = 85.24%, test accuracy = 28.48%
2017-05-12 17:09:50.811674  iteration = 118, loss = 1.0707, val_loss = 5.2798, training accuracy = 85.36%, test accuracy = 28.36%
2017-05-12 18:00:41.949927  iteration = 119, loss = 1.0707, val_loss = 5.2804, training accuracy = 85.36%, test accuracy = 28.40%
2017-05-12 18:50:55.384321  iteration = 120, loss = 1.0719, val_loss = 5.2797, training accuracy = 85.38%, test accuracy = 28.44%
2017-05-12 19:36:57.037812  iteration = 121, loss = 1.0700, val_loss = 5.2807, training accuracy = 85.38%, test accuracy = 28.48%
2017-05-12 20:21:28.758212  iteration = 122, loss = 1.0702, val_loss = 5.2800, training accuracy = 85.46%, test accuracy = 28.48%
2017-05-12 21:05:39.233504  iteration = 123, loss = 1.0696, val_loss = 5.2811, training accuracy = 85.42%, test accuracy = 28.42%
2017-05-12 21:56:59.418173  iteration = 124, loss = 1.0704, val_loss = 5.2786, training accuracy = 85.38%, test accuracy = 28.44%
2017-05-12 22:48:23.076697  iteration = 125, loss = 1.0685, val_loss = 5.2822, training accuracy = 85.46%, test accuracy = 28.44%
2017-05-12 23:34:56.122009  iteration = 126, loss = 1.0679, val_loss = 5.2829, training accuracy = 85.44%, test accuracy = 28.32%
2017-05-13 00:18:59.975449  iteration = 127, loss = 1.0687, val_loss = 5.2835, training accuracy = 85.50%, test accuracy = 28.44%
2017-05-13 01:03:06.073745  iteration = 128, loss = 1.0670, val_loss = 5.2846, training accuracy = 85.56%, test accuracy = 28.42%
2017-05-13 01:47:16.414475  iteration = 129, loss = 1.0654, val_loss = 5.2842, training accuracy = 85.58%, test accuracy = 28.44%
2017-05-13 02:39:24.914801  iteration = 130, loss = 1.0660, val_loss = 5.2851, training accuracy = 85.48%, test accuracy = 28.44%
2017-05-13 03:27:50.031105  iteration = 131, loss = 1.0676, val_loss = 5.2852, training accuracy = 85.58%, test accuracy = 28.46%
2017-05-13 04:13:22.549560  iteration = 132, loss = 1.0674, val_loss = 5.2858, training accuracy = 85.60%, test accuracy = 28.48%
2017-05-13 04:57:31.663074  iteration = 133, loss = 1.0661, val_loss = 5.2851, training accuracy = 85.56%, test accuracy = 28.52%
2017-05-13 05:41:38.408240  iteration = 134, loss = 1.0669, val_loss = 5.2864, training accuracy = 85.52%, test accuracy = 28.38%
2017-05-13 06:33:26.807565  iteration = 135, loss = 1.0667, val_loss = 5.2877, training accuracy = 85.56%, test accuracy = 28.42%
2017-05-13 07:23:51.659086  iteration = 136, loss = 1.0662, val_loss = 5.2858, training accuracy = 85.58%, test accuracy = 28.38%
2017-05-13 08:09:44.978035  iteration = 137, loss = 1.0656, val_loss = 5.2862, training accuracy = 85.58%, test accuracy = 28.46%
2017-05-13 08:54:01.140759  iteration = 138, loss = 1.0638, val_loss = 5.2869, training accuracy = 85.58%, test accuracy = 28.46%
2017-05-13 09:38:17.119632  iteration = 139, loss = 1.0650, val_loss = 5.2881, training accuracy = 85.54%, test accuracy = 28.44%
2017-05-13 10:30:20.386704  iteration = 140, loss = 1.0648, val_loss = 5.2894, training accuracy = 85.50%, test accuracy = 28.46%
2017-05-13 11:18:28.357323  iteration = 141, loss = 1.0638, val_loss = 5.2899, training accuracy = 85.58%, test accuracy = 28.46%
2017-05-13 12:04:07.723600  iteration = 142, loss = 1.0636, val_loss = 5.2912, training accuracy = 85.52%, test accuracy = 28.48%
2017-05-13 12:48:17.660445  iteration = 143, loss = 1.0621, val_loss = 5.2903, training accuracy = 85.56%, test accuracy = 28.36%
2017-05-13 13:32:45.783250  iteration = 144, loss = 1.0634, val_loss = 5.2902, training accuracy = 85.54%, test accuracy = 28.34%
2017-05-13 14:25:00.932992  iteration = 145, loss = 1.0620, val_loss = 5.2930, training accuracy = 85.62%, test accuracy = 28.32%
2017-05-13 15:10:31.926266  iteration = 146, loss = 1.0611, val_loss = 5.2928, training accuracy = 85.62%, test accuracy = 28.46%
2017-05-13 15:56:37.752807  iteration = 147, loss = 1.0616, val_loss = 5.2926, training accuracy = 85.52%, test accuracy = 28.42%
2017-05-13 16:44:54.366122  iteration = 148, loss = 1.0622, val_loss = 5.2932, training accuracy = 85.58%, test accuracy = 28.40%
2017-05-13 17:36:37.026327  iteration = 149, loss = 1.0625, val_loss = 5.2936, training accuracy = 85.66%, test accuracy = 28.40%
2017-05-13 18:23:49.992995  iteration = 150, loss = 1.0610, val_loss = 5.2945, training accuracy = 85.72%, test accuracy = 28.38%
2017-05-13 19:08:13.774596  iteration = 151, loss = 1.0607, val_loss = 5.2940, training accuracy = 85.84%, test accuracy = 28.46%
2017-05-13 19:51:30.921625  iteration = 152, loss = 1.0601, val_loss = 5.2943, training accuracy = 85.68%, test accuracy = 28.32%
2017-05-13 20:35:43.381238  iteration = 153, loss = 1.0602, val_loss = 5.2954, training accuracy = 85.72%, test accuracy = 28.30%
2017-05-13 21:25:30.583790  iteration = 154, loss = 1.0583, val_loss = 5.2977, training accuracy = 85.72%, test accuracy = 28.38%
2017-05-13 22:18:03.223037  iteration = 155, loss = 1.0605, val_loss = 5.2960, training accuracy = 85.66%, test accuracy = 28.36%
2017-05-13 23:04:08.387132  iteration = 156, loss = 1.0603, val_loss = 5.2957, training accuracy = 85.70%, test accuracy = 28.38%
2017-05-13 23:48:15.266395  iteration = 157, loss = 1.0581, val_loss = 5.2987, training accuracy = 85.70%, test accuracy = 28.34%
2017-05-14 00:33:00.526895  iteration = 158, loss = 1.0583, val_loss = 5.2973, training accuracy = 85.74%, test accuracy = 28.40%
2017-05-14 01:17:45.853786  iteration = 159, loss = 1.0572, val_loss = 5.2988, training accuracy = 85.72%, test accuracy = 28.30%
2017-05-14 02:10:21.975004  iteration = 160, loss = 1.0581, val_loss = 5.3001, training accuracy = 85.70%, test accuracy = 28.38%
2017-05-14 02:55:28.013961  iteration = 161, loss = 1.0580, val_loss = 5.3001, training accuracy = 85.82%, test accuracy = 28.36%
2017-05-14 03:39:45.561314  iteration = 162, loss = 1.0561, val_loss = 5.3010, training accuracy = 85.78%, test accuracy = 28.44%
2017-05-14 04:23:22.289936  iteration = 163, loss = 1.0567, val_loss = 5.3012, training accuracy = 85.76%, test accuracy = 28.42%
2017-05-14 05:07:33.618563  iteration = 164, loss = 1.0562, val_loss = 5.3009, training accuracy = 85.84%, test accuracy = 28.38%
2017-05-14 05:58:56.295297  iteration = 165, loss = 1.0552, val_loss = 5.3017, training accuracy = 85.78%, test accuracy = 28.42%
2017-05-14 06:47:23.816959  iteration = 166, loss = 1.0543, val_loss = 5.3016, training accuracy = 85.84%, test accuracy = 28.50%
2017-05-14 07:33:30.441909  iteration = 167, loss = 1.0556, val_loss = 5.3013, training accuracy = 85.86%, test accuracy = 28.38%
2017-05-14 08:25:29.408881  iteration = 168, loss = 1.0549, val_loss = 5.3027, training accuracy = 85.76%, test accuracy = 28.48%
2017-05-14 09:19:03.693272  iteration = 169, loss = 1.0537, val_loss = 5.3043, training accuracy = 85.78%, test accuracy = 28.28%
2017-05-14 10:05:00.723191  iteration = 170, loss = 1.0546, val_loss = 5.3041, training accuracy = 85.80%, test accuracy = 28.48%
2017-05-14 10:50:47.552761  iteration = 171, loss = 1.0539, val_loss = 5.3038, training accuracy = 85.84%, test accuracy = 28.30%
2017-05-14 11:42:59.388818  iteration = 172, loss = 1.0536, val_loss = 5.3050, training accuracy = 85.80%, test accuracy = 28.40%
2017-05-14 12:35:11.953096  iteration = 173, loss = 1.0525, val_loss = 5.3055, training accuracy = 85.86%, test accuracy = 28.46%
2017-05-14 13:21:29.014200  iteration = 174, loss = 1.0508, val_loss = 5.3058, training accuracy = 85.92%, test accuracy = 28.44%
2017-05-14 14:11:20.014996  iteration = 175, loss = 1.0527, val_loss = 5.3073, training accuracy = 85.92%, test accuracy = 28.38%
2017-05-14 15:07:49.907966  iteration = 176, loss = 1.0516, val_loss = 5.3059, training accuracy = 85.92%, test accuracy = 28.38%
2017-05-14 15:55:09.626916  iteration = 177, loss = 1.0521, val_loss = 5.3056, training accuracy = 85.86%, test accuracy = 28.38%
2017-05-14 16:40:57.239551  iteration = 178, loss = 1.0515, val_loss = 5.3091, training accuracy = 85.78%, test accuracy = 28.40%
2017-05-14 17:31:15.338591  iteration = 179, loss = 1.0511, val_loss = 5.3080, training accuracy = 85.86%, test accuracy = 28.32%
2017-05-14 18:20:26.074270  iteration = 180, loss = 1.0479, val_loss = 5.3086, training accuracy = 85.94%, test accuracy = 28.36%
2017-05-14 19:06:28.118228  iteration = 181, loss = 1.0525, val_loss = 5.3075, training accuracy = 85.90%, test accuracy = 28.38%
2017-05-14 19:58:22.177591  iteration = 182, loss = 1.0501, val_loss = 5.3106, training accuracy = 85.90%, test accuracy = 28.38%
2017-05-14 20:54:25.822856  iteration = 183, loss = 1.0501, val_loss = 5.3098, training accuracy = 85.92%, test accuracy = 28.42%
2017-05-14 21:40:12.965822  iteration = 184, loss = 1.0497, val_loss = 5.3108, training accuracy = 85.92%, test accuracy = 28.42%
2017-05-14 22:26:33.377689  iteration = 185, loss = 1.0504, val_loss = 5.3112, training accuracy = 86.04%, test accuracy = 28.40%
2017-05-14 23:19:01.105429  iteration = 186, loss = 1.0493, val_loss = 5.3121, training accuracy = 86.02%, test accuracy = 28.32%
2017-05-15 00:06:55.248937  iteration = 187, loss = 1.0504, val_loss = 5.3115, training accuracy = 86.04%, test accuracy = 28.34%
2017-05-15 00:50:59.704543  iteration = 188, loss = 1.0486, val_loss = 5.3127, training accuracy = 85.90%, test accuracy = 28.36%
2017-05-15 01:40:16.366787  iteration = 189, loss = 1.0471, val_loss = 5.3118, training accuracy = 85.98%, test accuracy = 28.30%
2017-05-15 02:32:49.069861  iteration = 190, loss = 1.0478, val_loss = 5.3126, training accuracy = 85.98%, test accuracy = 28.42%
2017-05-15 03:18:15.889219  iteration = 191, loss = 1.0481, val_loss = 5.3141, training accuracy = 86.00%, test accuracy = 28.44%
2017-05-15 04:02:45.432554  iteration = 192, loss = 1.0479, val_loss = 5.3134, training accuracy = 86.02%, test accuracy = 28.38%
2017-05-15 04:45:55.264223  iteration = 193, loss = 1.0468, val_loss = 5.3147, training accuracy = 86.10%, test accuracy = 28.44%
2017-05-15 05:30:06.601049  iteration = 194, loss = 1.0459, val_loss = 5.3143, training accuracy = 86.10%, test accuracy = 28.40%
2017-05-15 06:21:38.199023  iteration = 195, loss = 1.0458, val_loss = 5.3141, training accuracy = 86.06%, test accuracy = 28.40%
2017-05-15 07:10:10.670589  iteration = 196, loss = 1.0461, val_loss = 5.3171, training accuracy = 86.08%, test accuracy = 28.40%
2017-05-15 07:55:48.613941  iteration = 197, loss = 1.0456, val_loss = 5.3161, training accuracy = 86.00%, test accuracy = 28.34%
2017-05-15 08:39:56.326925  iteration = 198, loss = 1.0456, val_loss = 5.3162, training accuracy = 86.06%, test accuracy = 28.36%
2017-05-15 09:26:45.077034  iteration = 199, loss = 1.0466, val_loss = 5.3164, training accuracy = 86.24%, test accuracy = 28.40%

Continuing training with a further run
sjb@persephone:sjb_word_model$ ./train.py theresa_may_contributions.json
2017-05-15 13:57:03.362095  iteration = 000, loss = 1.0449, val_loss = 5.3179, training accuracy = 86.06%, test accuracy = 28.34%
2017-05-15 14:42:08.062884  iteration = 001, loss = 1.0449, val_loss = 5.3202, training accuracy = 86.08%, test accuracy = 28.32%
2017-05-15 15:26:21.077693  iteration = 002, loss = 1.0451, val_loss = 5.3202, training accuracy = 86.06%, test accuracy = 28.34%
2017-05-15 16:18:01.050117  iteration = 003, loss = 1.0438, val_loss = 5.3201, training accuracy = 86.12%, test accuracy = 28.42%
2017-05-15 17:05:20.351982  iteration = 004, loss = 1.0444, val_loss = 5.3181, training accuracy = 86.14%, test accuracy = 28.38%
2017-05-15 17:56:52.775248  iteration = 005, loss = 1.0452, val_loss = 5.3184, training accuracy = 86.20%, test accuracy = 28.34%
2017-05-15 18:48:56.478706  iteration = 006, loss = 1.0432, val_loss = 5.3197, training accuracy = 86.24%, test accuracy = 28.30%
2017-05-15 19:34:49.762857  iteration = 007, loss = 1.0432, val_loss = 5.3205, training accuracy = 86.26%, test accuracy = 28.44%
2017-05-15 20:21:39.230315  iteration = 008, loss = 1.0429, val_loss = 5.3213, training accuracy = 86.22%, test accuracy = 28.40%
2017-05-15 21:14:30.688490  iteration = 009, loss = 1.0410, val_loss = 5.3219, training accuracy = 86.22%, test accuracy = 28.38%
2017-05-15 22:02:14.728399  iteration = 010, loss = 1.0432, val_loss = 5.3220, training accuracy = 86.32%, test accuracy = 28.44%
2017-05-15 22:47:56.154129  iteration = 011, loss = 1.0426, val_loss = 5.3231, training accuracy = 86.34%, test accuracy = 28.30%
2017-05-15 23:35:25.220987  iteration = 012, loss = 1.0425, val_loss = 5.3215, training accuracy = 86.20%, test accuracy = 28.34%
2017-05-16 00:23:27.509292  iteration = 013, loss = 1.0408, val_loss = 5.3234, training accuracy = 86.30%, test accuracy = 28.24%
2017-05-16 01:09:26.017065  iteration = 014, loss = 1.0399, val_loss = 5.3244, training accuracy = 86.40%, test accuracy = 28.30%
2017-05-16 01:54:16.589732  iteration = 015, loss = 1.0407, val_loss = 5.3262, training accuracy = 86.30%, test accuracy = 28.50%
2017-05-16 02:39:25.570566  iteration = 016, loss = 1.0401, val_loss = 5.3260, training accuracy = 86.42%, test accuracy = 28.42%
2017-05-16 03:23:52.497284  iteration = 017, loss = 1.0404, val_loss = 5.3268, training accuracy = 86.46%, test accuracy = 28.42%
2017-05-16 04:15:58.084333  iteration = 018, loss = 1.0397, val_loss = 5.3261, training accuracy = 86.40%, test accuracy = 28.34%
2017-05-16 05:05:09.338693  iteration = 019, loss = 1.0394, val_loss = 5.3256, training accuracy = 86.38%, test accuracy = 28.28%
2017-05-16 05:51:13.648327  iteration = 020, loss = 1.0390, val_loss = 5.3275, training accuracy = 86.40%, test accuracy = 28.40%
2017-05-16 06:35:02.930657  iteration = 021, loss = 1.0397, val_loss = 5.3267, training accuracy = 86.48%, test accuracy = 28.22%
2017-05-16 07:18:57.405577  iteration = 022, loss = 1.0378, val_loss = 5.3303, training accuracy = 86.50%, test accuracy = 28.32%
2017-05-16 08:01:56.455367  iteration = 023, loss = 1.0386, val_loss = 5.3285, training accuracy = 86.50%, test accuracy = 28.24%
2017-05-16 08:45:54.067748  iteration = 024, loss = 1.0396, val_loss = 5.3276, training accuracy = 86.54%, test accuracy = 28.32%
2017-05-16 09:37:03.425705  iteration = 025, loss = 1.0380, val_loss = 5.3292, training accuracy = 86.56%, test accuracy = 28.30%
2017-05-16 10:28:16.954563  iteration = 026, loss = 1.0372, val_loss = 5.3292, training accuracy = 86.56%, test accuracy = 28.32%
2017-05-16 11:10:38.925473  iteration = 027, loss = 1.0381, val_loss = 5.3285, training accuracy = 86.56%, test accuracy = 28.30%
2017-05-16 11:56:05.131561  iteration = 028, loss = 1.0392, val_loss = 5.3298, training accuracy = 86.58%, test accuracy = 28.32%
2017-05-16 12:46:23.307253  iteration = 029, loss = 1.0377, val_loss = 5.3304, training accuracy = 86.52%, test accuracy = 28.30%
2017-05-16 13:36:12.229345  iteration = 030, loss = 1.0367, val_loss = 5.3327, training accuracy = 86.50%, test accuracy = 28.38%
2017-05-16 14:19:09.714395  iteration = 031, loss = 1.0358, val_loss = 5.3325, training accuracy = 86.64%, test accuracy = 28.38%
2017-05-16 15:03:20.375219  iteration = 032, loss = 1.0357, val_loss = 5.3327, training accuracy = 86.54%, test accuracy = 28.36%
2017-05-16 15:47:26.625304  iteration = 033, loss = 1.0346, val_loss = 5.3316, training accuracy = 86.46%, test accuracy = 28.32%
2017-05-16 16:23:23.267661  iteration = 034, loss = 1.0339, val_loss = 5.3350, training accuracy = 86.56%, test accuracy = 28.34%
2017-05-16 17:02:28.210389  iteration = 035, loss = 1.0350, val_loss = 5.3345, training accuracy = 86.54%, test accuracy = 28.34%
2017-05-16 17:39:59.788766  iteration = 036, loss = 1.0331, val_loss = 5.3341, training accuracy = 86.64%, test accuracy = 28.30%
2017-05-16 18:20:54.296552  iteration = 037, loss = 1.0332, val_loss = 5.3356, training accuracy = 86.64%, test accuracy = 28.32%
2017-05-16 18:56:25.339954  iteration = 038, loss = 1.0338, val_loss = 5.3352, training accuracy = 86.56%, test accuracy = 28.34%
2017-05-16 19:34:35.229964  iteration = 039, loss = 1.0331, val_loss = 5.3356, training accuracy = 86.58%, test accuracy = 28.36%
2017-05-16 20:17:27.865556  iteration = 040, loss = 1.0335, val_loss = 5.3355, training accuracy = 86.72%, test accuracy = 28.34%
2017-05-16 21:00:19.989063  iteration = 041, loss = 1.0317, val_loss = 5.3368, training accuracy = 86.68%, test accuracy = 28.34%
2017-05-16 21:44:15.632051  iteration = 042, loss = 1.0332, val_loss = 5.3362, training accuracy = 86.62%, test accuracy = 28.30%
2017-05-16 22:34:54.515946  iteration = 043, loss = 1.0325, val_loss = 5.3375, training accuracy = 86.68%, test accuracy = 28.36%
2017-05-16 23:27:23.309855  iteration = 044, loss = 1.0328, val_loss = 5.3388, training accuracy = 86.68%, test accuracy = 28.30%
2017-05-17 00:17:35.317078  iteration = 045, loss = 1.0326, val_loss = 5.3385, training accuracy = 86.72%, test accuracy = 28.30%
2017-05-17 01:03:53.477150  iteration = 046, loss = 1.0330, val_loss = 5.3383, training accuracy = 86.58%, test accuracy = 28.32%
2017-05-17 01:49:34.664027  iteration = 047, loss = 1.0316, val_loss = 5.3376, training accuracy = 86.70%, test accuracy = 28.30%
2017-05-17 02:39:33.103400  iteration = 048, loss = 1.0325, val_loss = 5.3399, training accuracy = 86.78%, test accuracy = 28.30%
2017-05-17 03:32:34.111073  iteration = 049, loss = 1.0315, val_loss = 5.3419, training accuracy = 86.78%, test accuracy = 28.32%
2017-05-17 04:20:44.298374  iteration = 050, loss = 1.0298, val_loss = 5.3404, training accuracy = 86.86%, test accuracy = 28.40%
2017-05-17 05:06:15.898756  iteration = 051, loss = 1.0304, val_loss = 5.3419, training accuracy = 86.80%, test accuracy = 28.38%
2017-05-17 05:51:41.499372  iteration = 052, loss = 1.0304, val_loss = 5.3420, training accuracy = 86.82%, test accuracy = 28.38%
2017-05-17 06:34:58.026157  iteration = 053, loss = 1.0294, val_loss = 5.3412, training accuracy = 86.76%, test accuracy = 28.38%
2017-05-17 07:19:29.541899  iteration = 054, loss = 1.0303, val_loss = 5.3435, training accuracy = 86.84%, test accuracy = 28.40%
2017-05-17 08:11:59.426143  iteration = 055, loss = 1.0292, val_loss = 5.3422, training accuracy = 86.88%, test accuracy = 28.36%
2017-05-17 09:03:02.117644  iteration = 056, loss = 1.0290, val_loss = 5.3423, training accuracy = 86.90%, test accuracy = 28.42%
2017-05-17 09:50:18.211660  iteration = 057, loss = 1.0269, val_loss = 5.3443, training accuracy = 86.92%, test accuracy = 28.40%
2017-05-17 10:33:11.956075  iteration = 058, loss = 1.0286, val_loss = 5.3446, training accuracy = 86.76%, test accuracy = 28.30%
2017-05-17 11:19:03.619870  iteration = 059, loss = 1.0282, val_loss = 5.3439, training accuracy = 86.82%, test accuracy = 28.34%
2017-05-17 12:08:55.650874  iteration = 060, loss = 1.0267, val_loss = 5.3449, training accuracy = 86.84%, test accuracy = 28.36%
2017-05-17 13:00:25.054420  iteration = 061, loss = 1.0265, val_loss = 5.3451, training accuracy = 86.98%, test accuracy = 28.42%
2017-05-17 13:55:02.781594  iteration = 062, loss = 1.0253, val_loss = 5.3460, training accuracy = 86.94%, test accuracy = 28.48%
2017-05-17 14:39:46.105632  iteration = 063, loss = 1.0270, val_loss = 5.3452, training accuracy = 86.90%, test accuracy = 28.36%
2017-05-17 15:22:24.960084  iteration = 064, loss = 1.0256, val_loss = 5.3466, training accuracy = 86.92%, test accuracy = 28.30%
2017-05-17 16:08:03.906109  iteration = 065, loss = 1.0265, val_loss = 5.3462, training accuracy = 87.02%, test accuracy = 28.32%
2017-05-17 16:56:39.084549  iteration = 066, loss = 1.0259, val_loss = 5.3472, training accuracy = 86.98%, test accuracy = 28.24%
2017-05-17 17:49:13.460775  iteration = 067, loss = 1.0254, val_loss = 5.3471, training accuracy = 86.98%, test accuracy = 28.40%
2017-05-17 18:35:33.119245  iteration = 068, loss = 1.0269, val_loss = 5.3490, training accuracy = 87.04%, test accuracy = 28.30%
2017-05-17 19:21:35.433709  iteration = 069, loss = 1.0267, val_loss = 5.3474, training accuracy = 87.00%, test accuracy = 28.32%
2017-05-17 20:14:02.997649  iteration = 070, loss = 1.0248, val_loss = 5.3523, training accuracy = 87.04%, test accuracy = 28.26%
2017-05-17 20:59:46.951840  iteration = 071, loss = 1.0255, val_loss = 5.3499, training accuracy = 87.04%, test accuracy = 28.30%
2017-05-17 21:47:59.502345  iteration = 072, loss = 1.0248, val_loss = 5.3506, training accuracy = 87.02%, test accuracy = 28.36%
2017-05-17 22:41:56.540878  iteration = 073, loss = 1.0253, val_loss = 5.3484, training accuracy = 86.96%, test accuracy = 28.32%
2017-05-17 23:29:34.860355  iteration = 074, loss = 1.0236, val_loss = 5.3526, training accuracy = 86.96%, test accuracy = 28.36%
2017-05-18 00:16:57.914451  iteration = 075, loss = 1.0227, val_loss = 5.3531, training accuracy = 87.04%, test accuracy = 28.26%
2017-05-18 01:10:34.465446  iteration = 076, loss = 1.0254, val_loss = 5.3517, training accuracy = 86.98%, test accuracy = 28.30%
2017-05-18 01:58:41.528171  iteration = 077, loss = 1.0220, val_loss = 5.3530, training accuracy = 87.00%, test accuracy = 28.28%
2017-05-18 02:45:44.742402  iteration = 078, loss = 1.0241, val_loss = 5.3538, training accuracy = 87.10%, test accuracy = 28.26%
2017-05-18 03:37:50.011938  iteration = 079, loss = 1.0213, val_loss = 5.3540, training accuracy = 87.00%, test accuracy = 28.20%
2017-05-18 04:26:03.267694  iteration = 080, loss = 1.0221, val_loss = 5.3534, training accuracy = 86.92%, test accuracy = 28.24%
2017-05-18 05:12:08.692567  iteration = 081, loss = 1.0232, val_loss = 5.3542, training accuracy = 87.06%, test accuracy = 28.30%
2017-05-18 05:59:08.448781  iteration = 082, loss = 1.0220, val_loss = 5.3556, training accuracy = 87.04%, test accuracy = 28.38%
2017-05-18 06:50:56.712888  iteration = 083, loss = 1.0217, val_loss = 5.3554, training accuracy = 87.04%, test accuracy = 28.38%
2017-05-18 07:39:36.412826  iteration = 084, loss = 1.0211, val_loss = 5.3562, training accuracy = 87.06%, test accuracy = 28.34%
2017-05-18 08:25:23.844326  iteration = 085, loss = 1.0203, val_loss = 5.3564, training accuracy = 86.98%, test accuracy = 28.36%
2017-05-18 09:09:26.071722  iteration = 086, loss = 1.0198, val_loss = 5.3565, training accuracy = 86.98%, test accuracy = 28.38%
2017-05-18 10:01:06.868600  iteration = 087, loss = 1.0228, val_loss = 5.3567, training accuracy = 87.02%, test accuracy = 28.34%
2017-05-18 10:51:23.420591  iteration = 088, loss = 1.0203, val_loss = 5.3573, training accuracy = 87.00%, test accuracy = 28.34%
2017-05-18 11:37:18.832712  iteration = 089, loss = 1.0218, val_loss = 5.3578, training accuracy = 87.16%, test accuracy = 28.34%
2017-05-18 12:21:17.780640  iteration = 090, loss = 1.0201, val_loss = 5.3584, training accuracy = 87.06%, test accuracy = 28.32%
2017-05-18 13:11:07.511507  iteration = 091, loss = 1.0199, val_loss = 5.3582, training accuracy = 87.16%, test accuracy = 28.42%
2017-05-18 13:55:12.139509  iteration = 092, loss = 1.0188, val_loss = 5.3602, training accuracy = 87.12%, test accuracy = 28.34%
2017-05-18 14:40:16.472744  iteration = 093, loss = 1.0209, val_loss = 5.3584, training accuracy = 87.14%, test accuracy = 28.38%
2017-05-18 15:28:58.656008  iteration = 094, loss = 1.0173, val_loss = 5.3609, training accuracy = 87.18%, test accuracy = 28.42%
2017-05-18 16:14:29.497680  iteration = 095, loss = 1.0180, val_loss = 5.3622, training accuracy = 87.18%, test accuracy = 28.40%
2017-05-18 16:57:53.213323  iteration = 096, loss = 1.0180, val_loss = 5.3603, training accuracy = 87.12%, test accuracy = 28.36%
2017-05-18 17:41:51.992567  iteration = 097, loss = 1.0153, val_loss = 5.3610, training accuracy = 87.18%, test accuracy = 28.36%
2017-05-18 18:26:57.461993  iteration = 098, loss = 1.0189, val_loss = 5.3601, training accuracy = 87.12%, test accuracy = 28.36%
2017-05-18 19:18:58.732557  iteration = 099, loss = 1.0163, val_loss = 5.3641, training accuracy = 87.18%, test accuracy = 28.36%
2017-05-18 20:08:16.217720  iteration = 100, loss = 1.0162, val_loss = 5.3616, training accuracy = 87.08%, test accuracy = 28.26%
2017-05-18 20:54:19.687998  iteration = 101, loss = 1.0151, val_loss = 5.3638, training accuracy = 87.18%, test accuracy = 28.44%
2017-05-18 21:40:27.612856  iteration = 102, loss = 1.0170, val_loss = 5.3638, training accuracy = 87.20%, test accuracy = 28.38%
2017-05-18 22:24:32.838075  iteration = 103, loss = 1.0138, val_loss = 5.3647, training accuracy = 87.08%, test accuracy = 28.40%
2017-05-18 23:08:22.493318  iteration = 104, loss = 1.0139, val_loss = 5.3654, training accuracy = 87.22%, test accuracy = 28.30%
2017-05-18 23:59:36.556724  iteration = 105, loss = 1.0138, val_loss = 5.3647, training accuracy = 87.18%, test accuracy = 28.36%
2017-05-19 00:52:54.126063  iteration = 106, loss = 1.0151, val_loss = 5.3651, training accuracy = 87.24%, test accuracy = 28.42%
2017-05-19 01:45:03.433591  iteration = 107, loss = 1.0162, val_loss = 5.3656, training accuracy = 87.24%, test accuracy = 28.34%
2017-05-19 02:35:29.802542  iteration = 108, loss = 1.0145, val_loss = 5.3675, training accuracy = 87.26%, test accuracy = 28.34%
2017-05-19 03:22:46.909755  iteration = 109, loss = 1.0132, val_loss = 5.3678, training accuracy = 87.34%, test accuracy = 28.36%
2017-05-19 04:08:33.704030  iteration = 110, loss = 1.0148, val_loss = 5.3681, training accuracy = 87.26%, test accuracy = 28.26%
2017-05-19 04:52:32.051639  iteration = 111, loss = 1.0131, val_loss = 5.3674, training accuracy = 87.14%, test accuracy = 28.34%
2017-05-19 05:36:26.324358  iteration = 112, loss = 1.0127, val_loss = 5.3687, training accuracy = 87.26%, test accuracy = 28.32%
2017-05-19 06:32:27.529931  iteration = 113, loss = 1.0129, val_loss = 5.3682, training accuracy = 87.22%, test accuracy = 28.32%
2017-05-19 07:24:40.236261  iteration = 114, loss = 1.0125, val_loss = 5.3684, training accuracy = 87.18%, test accuracy = 28.30%
2017-05-19 08:12:47.584673  iteration = 115, loss = 1.0133, val_loss = 5.3691, training accuracy = 87.22%, test accuracy = 28.30%
2017-05-19 08:57:58.407821  iteration = 116, loss = 1.0134, val_loss = 5.3688, training accuracy = 87.24%, test accuracy = 28.32%
2017-05-19 09:41:48.015965  iteration = 117, loss = 1.0127, val_loss = 5.3694, training accuracy = 87.22%, test accuracy = 28.34%
2017-05-19 10:25:42.561214  iteration = 118, loss = 1.0103, val_loss = 5.3733, training accuracy = 87.22%, test accuracy = 28.28%
2017-05-19 11:16:40.169197  iteration = 119, loss = 1.0092, val_loss = 5.3722, training accuracy = 87.22%, test accuracy = 28.36%
2017-05-19 12:01:00.307840  iteration = 120, loss = 1.0111, val_loss = 5.3714, training accuracy = 87.22%, test accuracy = 28.38%
2017-05-19 12:46:07.512836  iteration = 121, loss = 1.0102, val_loss = 5.3730, training accuracy = 87.34%, test accuracy = 28.36%
2017-05-19 13:29:57.633230  iteration = 122, loss = 1.0106, val_loss = 5.3722, training accuracy = 87.32%, test accuracy = 28.38%
2017-05-19 14:20:59.535898  iteration = 123, loss = 1.0096, val_loss = 5.3740, training accuracy = 87.28%, test accuracy = 28.40%
2017-05-19 15:12:45.252346  iteration = 124, loss = 1.0111, val_loss = 5.3722, training accuracy = 87.28%, test accuracy = 28.40%
2017-05-19 16:00:37.025704  iteration = 125, loss = 1.0072, val_loss = 5.3736, training accuracy = 87.24%, test accuracy = 28.42%
2017-05-19 16:44:43.394344  iteration = 126, loss = 1.0084, val_loss = 5.3748, training accuracy = 87.28%, test accuracy = 28.26%
2017-05-19 17:28:14.228485  iteration = 127, loss = 1.0102, val_loss = 5.3751, training accuracy = 87.28%, test accuracy = 28.36%
2017-05-19 18:13:34.387980  iteration = 128, loss = 1.0088, val_loss = 5.3744, training accuracy = 87.30%, test accuracy = 28.44%
2017-05-19 18:57:31.266291  iteration = 129, loss = 1.0077, val_loss = 5.3764, training accuracy = 87.34%, test accuracy = 28.36%
2017-05-19 19:41:29.283057  iteration = 130, loss = 1.0069, val_loss = 5.3764, training accuracy = 87.32%, test accuracy = 28.38%
2017-05-19 20:27:10.842251  iteration = 131, loss = 1.0089, val_loss = 5.3761, training accuracy = 87.32%, test accuracy = 28.30%
2017-05-19 21:14:16.100061  iteration = 132, loss = 1.0056, val_loss = 5.3767, training accuracy = 87.40%, test accuracy = 28.32%
2017-05-19 22:02:38.258475  iteration = 133, loss = 1.0075, val_loss = 5.3764, training accuracy = 87.36%, test accuracy = 28.36%
2017-05-19 22:52:22.608503  iteration = 134, loss = 1.0073, val_loss = 5.3779, training accuracy = 87.38%, test accuracy = 28.30%
2017-05-19 23:40:09.154885  iteration = 135, loss = 1.0082, val_loss = 5.3783, training accuracy = 87.34%, test accuracy = 28.34%
2017-05-20 00:27:21.748489  iteration = 136, loss = 1.0080, val_loss = 5.3808, training accuracy = 87.36%, test accuracy = 28.32%
2017-05-20 01:11:46.929258  iteration = 137, loss = 1.0054, val_loss = 5.3803, training accuracy = 87.34%, test accuracy = 28.34%
2017-05-20 01:55:10.407729  iteration = 138, loss = 1.0057, val_loss = 5.3804, training accuracy = 87.44%, test accuracy = 28.38%
2017-05-20 02:39:54.507067  iteration = 139, loss = 1.0053, val_loss = 5.3791, training accuracy = 87.52%, test accuracy = 28.32%
2017-05-20 03:24:11.832024  iteration = 140, loss = 1.0053, val_loss = 5.3806, training accuracy = 87.42%, test accuracy = 28.40%
2017-05-20 04:11:33.111721  iteration = 141, loss = 1.0051, val_loss = 5.3818, training accuracy = 87.42%, test accuracy = 28.34%
2017-05-20 05:03:04.935133  iteration = 142, loss = 1.0061, val_loss = 5.3813, training accuracy = 87.48%, test accuracy = 28.30%
2017-05-20 05:51:19.191352  iteration = 143, loss = 1.0043, val_loss = 5.3821, training accuracy = 87.36%, test accuracy = 28.42%
2017-05-20 06:37:18.102797  iteration = 144, loss = 1.0042, val_loss = 5.3828, training accuracy = 87.40%, test accuracy = 28.40%
2017-05-20 07:21:17.204940  iteration = 145, loss = 1.0051, val_loss = 5.3835, training accuracy = 87.42%, test accuracy = 28.34%
2017-05-20 08:05:13.093556  iteration = 146, loss = 1.0026, val_loss = 5.3852, training accuracy = 87.42%, test accuracy = 28.32%
2017-05-20 08:57:55.746783  iteration = 147, loss = 1.0041, val_loss = 5.3820, training accuracy = 87.44%, test accuracy = 28.38%
2017-05-20 09:47:00.590534  iteration = 148, loss = 1.0024, val_loss = 5.3831, training accuracy = 87.40%, test accuracy = 28.32%
2017-05-20 10:32:04.246254  iteration = 149, loss = 1.0032, val_loss = 5.3839, training accuracy = 87.38%, test accuracy = 28.32%
2017-05-20 11:17:04.338476  iteration = 150, loss = 1.0046, val_loss = 5.3842, training accuracy = 87.42%, test accuracy = 28.36%
2017-05-20 12:01:05.503030  iteration = 151, loss = 1.0028, val_loss = 5.3851, training accuracy = 87.40%, test accuracy = 28.40%
2017-05-20 12:53:08.649246  iteration = 152, loss = 1.0020, val_loss = 5.3856, training accuracy = 87.50%, test accuracy = 28.34%
2017-05-20 13:45:11.745537  iteration = 153, loss = 1.0016, val_loss = 5.3859, training accuracy = 87.44%, test accuracy = 28.34%
2017-05-20 14:32:34.981250  iteration = 154, loss = 1.0021, val_loss = 5.3855, training accuracy = 87.52%, test accuracy = 28.36%
2017-05-20 15:16:21.128509  iteration = 155, loss = 1.0027, val_loss = 5.3867, training accuracy = 87.50%, test accuracy = 28.32%
2017-05-20 16:00:14.551298  iteration = 156, loss = 1.0004, val_loss = 5.3872, training accuracy = 87.44%, test accuracy = 28.36%
2017-05-20 16:45:42.403983  iteration = 157, loss = 1.0025, val_loss = 5.3868, training accuracy = 87.46%, test accuracy = 28.30%
2017-05-20 17:37:38.464654  iteration = 158, loss = 1.0002, val_loss = 5.3879, training accuracy = 87.48%, test accuracy = 28.38%
2017-05-20 18:29:04.691384  iteration = 159, loss = 1.0002, val_loss = 5.3875, training accuracy = 87.48%, test accuracy = 28.32%
2017-05-20 19:14:46.715544  iteration = 160, loss = 1.0010, val_loss = 5.3873, training accuracy = 87.50%, test accuracy = 28.34%
2017-05-20 19:58:25.811585  iteration = 161, loss = 1.0015, val_loss = 5.3883, training accuracy = 87.50%, test accuracy = 28.32%
2017-05-20 20:42:09.387817  iteration = 162, loss = 0.9990, val_loss = 5.3873, training accuracy = 87.52%, test accuracy = 28.28%
2017-05-20 21:27:16.370111  iteration = 163, loss = 0.9988, val_loss = 5.3901, training accuracy = 87.48%, test accuracy = 28.32%
2017-05-20 22:13:15.938495  iteration = 164, loss = 0.9995, val_loss = 5.3898, training accuracy = 87.50%, test accuracy = 28.30%
2017-05-20 23:05:59.505792  iteration = 165, loss = 0.9998, val_loss = 5.3908, training accuracy = 87.46%, test accuracy = 28.30%
2017-05-20 23:51:24.908272  iteration = 166, loss = 0.9987, val_loss = 5.3897, training accuracy = 87.52%, test accuracy = 28.30%
2017-05-21 00:35:26.122516  iteration = 167, loss = 0.9979, val_loss = 5.3918, training accuracy = 87.48%, test accuracy = 28.30%
2017-05-21 01:27:32.219114  iteration = 168, loss = 0.9973, val_loss = 5.3925, training accuracy = 87.60%, test accuracy = 28.36%
2017-05-21 02:18:59.024519  iteration = 169, loss = 0.9984, val_loss = 5.3917, training accuracy = 87.56%, test accuracy = 28.30%
2017-05-21 03:07:02.306428  iteration = 170, loss = 0.9986, val_loss = 5.3937, training accuracy = 87.54%, test accuracy = 28.28%
2017-05-21 03:52:35.029147  iteration = 171, loss = 0.9961, val_loss = 5.3936, training accuracy = 87.50%, test accuracy = 28.20%
2017-05-21 04:38:30.756571  iteration = 172, loss = 0.9968, val_loss = 5.3927, training accuracy = 87.54%, test accuracy = 28.20%
2017-05-21 05:30:33.524705  iteration = 173, loss = 0.9954, val_loss = 5.3960, training accuracy = 87.56%, test accuracy = 28.20%
2017-05-21 06:20:26.430119  iteration = 174, loss = 0.9949, val_loss = 5.3937, training accuracy = 87.52%, test accuracy = 28.32%
2017-05-21 07:06:25.148188  iteration = 175, loss = 0.9965, val_loss = 5.3947, training accuracy = 87.62%, test accuracy = 28.36%
2017-05-21 07:50:56.234239  iteration = 176, loss = 0.9955, val_loss = 5.3957, training accuracy = 87.60%, test accuracy = 28.30%
2017-05-21 08:42:57.319563  iteration = 177, loss = 0.9948, val_loss = 5.3967, training accuracy = 87.58%, test accuracy = 28.34%
2017-05-21 09:35:44.245190  iteration = 178, loss = 0.9974, val_loss = 5.3955, training accuracy = 87.62%, test accuracy = 28.28%
2017-05-21 10:23:56.187215  iteration = 179, loss = 0.9952, val_loss = 5.3988, training accuracy = 87.60%, test accuracy = 28.26%
2017-05-21 11:10:19.710494  iteration = 180, loss = 0.9945, val_loss = 5.3964, training accuracy = 87.62%, test accuracy = 28.26%
2017-05-21 11:56:08.975093  iteration = 181, loss = 0.9966, val_loss = 5.3973, training accuracy = 87.66%, test accuracy = 28.20%
2017-05-21 12:48:12.870186  iteration = 182, loss = 0.9947, val_loss = 5.3967, training accuracy = 87.62%, test accuracy = 28.32%
2017-05-21 13:39:07.267174  iteration = 183, loss = 0.9945, val_loss = 5.3985, training accuracy = 87.64%, test accuracy = 28.30%
2017-05-21 14:25:03.855313  iteration = 184, loss = 0.9940, val_loss = 5.3980, training accuracy = 87.62%, test accuracy = 28.30%
2017-05-21 15:09:42.998147  iteration = 185, loss = 0.9951, val_loss = 5.3980, training accuracy = 87.66%, test accuracy = 28.40%
2017-05-21 16:01:19.028770  iteration = 186, loss = 0.9943, val_loss = 5.3996, training accuracy = 87.64%, test accuracy = 28.40%
2017-05-21 16:54:19.458062  iteration = 187, loss = 0.9922, val_loss = 5.3993, training accuracy = 87.76%, test accuracy = 28.34%
2017-05-21 17:42:03.209703  iteration = 188, loss = 0.9932, val_loss = 5.4010, training accuracy = 87.62%, test accuracy = 28.30%
2017-05-21 18:28:11.505243  iteration = 189, loss = 0.9941, val_loss = 5.4009, training accuracy = 87.68%, test accuracy = 28.34%
2017-05-21 19:13:00.928842  iteration = 190, loss = 0.9917, val_loss = 5.4001, training accuracy = 87.74%, test accuracy = 28.28%
2017-05-21 19:58:08.146344  iteration = 191, loss = 0.9919, val_loss = 5.4010, training accuracy = 87.70%, test accuracy = 28.34%
2017-05-21 20:44:23.579867  iteration = 192, loss = 0.9933, val_loss = 5.4023, training accuracy = 87.66%, test accuracy = 28.28%
2017-05-21 21:36:38.645639  iteration = 193, loss = 0.9923, val_loss = 5.4018, training accuracy = 87.76%, test accuracy = 28.36%
2017-05-21 22:30:13.739943  iteration = 194, loss = 0.9916, val_loss = 5.4016, training accuracy = 87.76%, test accuracy = 28.34%
2017-05-21 23:18:40.906456  iteration = 195, loss = 0.9909, val_loss = 5.4028, training accuracy = 87.76%, test accuracy = 28.30%
2017-05-22 00:03:50.537612  iteration = 196, loss = 0.9893, val_loss = 5.4037, training accuracy = 87.72%, test accuracy = 28.26%
2017-05-22 00:47:42.782301  iteration = 197, loss = 0.9914, val_loss = 5.4036, training accuracy = 87.80%, test accuracy = 28.26%
2017-05-22 01:30:44.646716  iteration = 198, loss = 0.9910, val_loss = 5.4036, training accuracy = 87.80%, test accuracy = 28.32%
2017-05-22 02:14:34.482524  iteration = 199, loss = 0.9893, val_loss = 5.4058, training accuracy = 87.78%, test accuracy = 28.26%

Continuing training with a further run
sjb@persephone:sjb_word_model$ ./train.py theresa_may_contributions.json
2017-05-22 23:31:58.204790  iteration = 000, loss = 0.9882, val_loss = 5.4068, training accuracy = 87.80%, test accuracy = 28.40%
2017-05-23 00:21:46.665191  iteration = 001, loss = 0.9901, val_loss = 5.4054, training accuracy = 87.82%, test accuracy = 28.32%
2017-05-23 01:07:46.006372  iteration = 002, loss = 0.9895, val_loss = 5.4050, training accuracy = 87.82%, test accuracy = 28.34%
2017-05-23 02:00:29.932167  iteration = 003, loss = 0.9901, val_loss = 5.4052, training accuracy = 87.90%, test accuracy = 28.30%
2017-05-23 02:54:52.230885  iteration = 004, loss = 0.9901, val_loss = 5.4067, training accuracy = 87.88%, test accuracy = 28.36%
2017-05-23 03:40:38.770164  iteration = 005, loss = 0.9887, val_loss = 5.4067, training accuracy = 87.90%, test accuracy = 28.36%
2017-05-23 04:31:03.350346  iteration = 006, loss = 0.9877, val_loss = 5.4108, training accuracy = 87.84%, test accuracy = 28.36%
2017-05-23 05:22:51.775288  iteration = 007, loss = 0.9874, val_loss = 5.4089, training accuracy = 87.88%, test accuracy = 28.42%
2017-05-23 06:10:27.345155  iteration = 008, loss = 0.9887, val_loss = 5.4085, training accuracy = 87.90%, test accuracy = 28.34%
2017-05-23 06:55:27.949378  iteration = 009, loss = 0.9874, val_loss = 5.4096, training accuracy = 87.92%, test accuracy = 28.26%
2017-05-23 07:40:58.758417  iteration = 010, loss = 0.9857, val_loss = 5.4101, training accuracy = 87.88%, test accuracy = 28.38%
2017-05-23 08:25:07.443072  iteration = 011, loss = 0.9869, val_loss = 5.4090, training accuracy = 87.94%, test accuracy = 28.30%
2017-05-23 09:14:34.951352  iteration = 012, loss = 0.9875, val_loss = 5.4099, training accuracy = 88.00%, test accuracy = 28.30%
2017-05-23 10:07:10.007563  iteration = 013, loss = 0.9866, val_loss = 5.4091, training accuracy = 87.90%, test accuracy = 28.34%
2017-05-23 10:55:17.418625  iteration = 014, loss = 0.9866, val_loss = 5.4110, training accuracy = 87.94%, test accuracy = 28.36%
2017-05-23 11:40:28.669633  iteration = 015, loss = 0.9867, val_loss = 5.4121, training accuracy = 87.88%, test accuracy = 28.32%
2017-05-23 12:24:17.463251  iteration = 016, loss = 0.9849, val_loss = 5.4116, training accuracy = 88.00%, test accuracy = 28.28%
2017-05-23 13:08:04.124140  iteration = 017, loss = 0.9856, val_loss = 5.4128, training accuracy = 87.92%, test accuracy = 28.30%
2017-05-23 14:00:28.611688  iteration = 018, loss = 0.9848, val_loss = 5.4143, training accuracy = 87.94%, test accuracy = 28.32%
2017-05-23 14:48:20.789802  iteration = 019, loss = 0.9844, val_loss = 5.4124, training accuracy = 87.98%, test accuracy = 28.36%
2017-05-23 15:36:09.162839  iteration = 020, loss = 0.9851, val_loss = 5.4117, training accuracy = 88.00%, test accuracy = 28.30%
2017-05-23 16:21:07.156571  iteration = 021, loss = 0.9843, val_loss = 5.4137, training accuracy = 87.98%, test accuracy = 28.34%
2017-05-23 17:04:58.321505  iteration = 022, loss = 0.9835, val_loss = 5.4145, training accuracy = 87.88%, test accuracy = 28.28%
2017-05-23 17:49:03.063036  iteration = 023, loss = 0.9833, val_loss = 5.4157, training accuracy = 87.92%, test accuracy = 28.26%
2017-05-23 18:41:54.356279  iteration = 024, loss = 0.9846, val_loss = 5.4155, training accuracy = 87.94%, test accuracy = 28.22%
2017-05-23 19:29:38.647845  iteration = 025, loss = 0.9836, val_loss = 5.4153, training accuracy = 87.82%, test accuracy = 28.22%
2017-05-23 20:15:42.260410  iteration = 026, loss = 0.9822, val_loss = 5.4180, training accuracy = 87.92%, test accuracy = 28.28%
2017-05-23 20:59:20.620916  iteration = 027, loss = 0.9830, val_loss = 5.4155, training accuracy = 87.96%, test accuracy = 28.32%
2017-05-23 21:43:10.947332  iteration = 028, loss = 0.9823, val_loss = 5.4166, training accuracy = 88.06%, test accuracy = 28.26%
2017-05-23 22:26:55.684882  iteration = 029, loss = 0.9836, val_loss = 5.4170, training accuracy = 87.94%, test accuracy = 28.34%
2017-05-23 23:19:06.420149  iteration = 030, loss = 0.9819, val_loss = 5.4176, training accuracy = 88.12%, test accuracy = 28.30%
2017-05-24 00:06:42.408712  iteration = 031, loss = 0.9821, val_loss = 5.4160, training accuracy = 88.00%, test accuracy = 28.30%
2017-05-24 00:51:58.211011  iteration = 032, loss = 0.9826, val_loss = 5.4193, training accuracy = 88.04%, test accuracy = 28.28%
2017-05-24 01:34:56.564286  iteration = 033, loss = 0.9817, val_loss = 5.4178, training accuracy = 88.04%, test accuracy = 28.34%
2017-05-24 02:19:44.554371  iteration = 034, loss = 0.9803, val_loss = 5.4212, training accuracy = 88.06%, test accuracy = 28.34%
2017-05-24 03:03:36.292931  iteration = 035, loss = 0.9825, val_loss = 5.4189, training accuracy = 88.10%, test accuracy = 28.32%
2017-05-24 03:55:08.110782  iteration = 036, loss = 0.9816, val_loss = 5.4209, training accuracy = 87.98%, test accuracy = 28.32%
2017-05-24 04:47:41.102636  iteration = 037, loss = 0.9793, val_loss = 5.4210, training accuracy = 88.04%, test accuracy = 28.34%
2017-05-24 05:33:38.545932  iteration = 038, loss = 0.9798, val_loss = 5.4221, training accuracy = 88.04%, test accuracy = 28.36%
2017-05-24 06:17:28.473423  iteration = 039, loss = 0.9811, val_loss = 5.4211, training accuracy = 88.04%, test accuracy = 28.30%
2017-05-24 07:01:10.409375  iteration = 040, loss = 0.9792, val_loss = 5.4230, training accuracy = 88.12%, test accuracy = 28.32%
2017-05-24 07:53:36.382232  iteration = 041, loss = 0.9796, val_loss = 5.4225, training accuracy = 88.10%, test accuracy = 28.32%
2017-05-24 08:42:37.079456  iteration = 042, loss = 0.9794, val_loss = 5.4242, training accuracy = 88.04%, test accuracy = 28.26%
2017-05-24 09:28:25.605602  iteration = 043, loss = 0.9794, val_loss = 5.4225, training accuracy = 88.04%, test accuracy = 28.32%
2017-05-24 10:13:44.852524  iteration = 044, loss = 0.9798, val_loss = 5.4241, training accuracy = 88.06%, test accuracy = 28.34%
2017-05-24 11:06:17.064595  iteration = 045, loss = 0.9775, val_loss = 5.4244, training accuracy = 88.12%, test accuracy = 28.26%
2017-05-24 11:59:17.636419  iteration = 046, loss = 0.9790, val_loss = 5.4232, training accuracy = 88.12%, test accuracy = 28.28%
2017-05-24 12:46:59.727352  iteration = 047, loss = 0.9762, val_loss = 5.4243, training accuracy = 88.24%, test accuracy = 28.16%
2017-05-24 13:32:30.378170  iteration = 048, loss = 0.9765, val_loss = 5.4267, training accuracy = 88.16%, test accuracy = 28.32%
2017-05-24 14:17:07.960986  iteration = 049, loss = 0.9770, val_loss = 5.4240, training accuracy = 88.12%, test accuracy = 28.24%
2017-05-24 15:01:00.133248  iteration = 050, loss = 0.9773, val_loss = 5.4268, training accuracy = 88.12%, test accuracy = 28.30%
2017-05-24 15:53:05.866620  iteration = 051, loss = 0.9765, val_loss = 5.4276, training accuracy = 88.10%, test accuracy = 28.24%
2017-05-24 16:43:38.042306  iteration = 052, loss = 0.9764, val_loss = 5.4276, training accuracy = 88.20%, test accuracy = 28.24%
2017-05-24 17:29:39.154302  iteration = 053, loss = 0.9764, val_loss = 5.4285, training accuracy = 88.18%, test accuracy = 28.22%
2017-05-24 18:14:13.087349  iteration = 054, loss = 0.9758, val_loss = 5.4271, training accuracy = 88.18%, test accuracy = 28.20%
2017-05-24 18:57:54.422590  iteration = 055, loss = 0.9761, val_loss = 5.4275, training accuracy = 88.14%, test accuracy = 28.18%
2017-05-24 19:50:16.437121  iteration = 056, loss = 0.9748, val_loss = 5.4292, training accuracy = 88.20%, test accuracy = 28.30%
2017-05-24 20:40:06.403224  iteration = 057, loss = 0.9762, val_loss = 5.4282, training accuracy = 88.18%, test accuracy = 28.26%
2017-05-24 21:26:27.669424  iteration = 058, loss = 0.9755, val_loss = 5.4297, training accuracy = 88.18%, test accuracy = 28.30%
2017-05-24 22:12:08.212493  iteration = 059, loss = 0.9753, val_loss = 5.4317, training accuracy = 88.24%, test accuracy = 28.20%
2017-05-24 22:55:59.646195  iteration = 060, loss = 0.9739, val_loss = 5.4295, training accuracy = 88.16%, test accuracy = 28.20%
2017-05-24 23:47:52.518181  iteration = 061, loss = 0.9752, val_loss = 5.4294, training accuracy = 88.26%, test accuracy = 28.26%
2017-05-25 00:38:54.711431  iteration = 062, loss = 0.9754, val_loss = 5.4308, training accuracy = 88.30%, test accuracy = 28.28%
2017-05-25 01:24:29.044398  iteration = 063, loss = 0.9737, val_loss = 5.4300, training accuracy = 88.22%, test accuracy = 28.26%
2017-05-25 02:10:01.947374  iteration = 064, loss = 0.9726, val_loss = 5.4336, training accuracy = 88.22%, test accuracy = 28.26%
2017-05-25 02:53:29.529775  iteration = 065, loss = 0.9733, val_loss = 5.4314, training accuracy = 88.22%, test accuracy = 28.22%
2017-05-25 03:37:24.039388  iteration = 066, loss = 0.9733, val_loss = 5.4315, training accuracy = 88.28%, test accuracy = 28.16%
2017-05-25 04:29:22.338915  iteration = 067, loss = 0.9742, val_loss = 5.4326, training accuracy = 88.28%, test accuracy = 28.24%
2017-05-25 05:19:37.769487  iteration = 068, loss = 0.9736, val_loss = 5.4328, training accuracy = 88.26%, test accuracy = 28.24%
2017-05-25 06:05:28.514622  iteration = 069, loss = 0.9722, val_loss = 5.4333, training accuracy = 88.32%, test accuracy = 28.30%
2017-05-25 06:50:26.799562  iteration = 070, loss = 0.9721, val_loss = 5.4350, training accuracy = 88.34%, test accuracy = 28.24%
2017-05-25 07:38:23.605614  iteration = 071, loss = 0.9723, val_loss = 5.4344, training accuracy = 88.26%, test accuracy = 28.18%
2017-05-25 08:31:22.189515  iteration = 072, loss = 0.9706, val_loss = 5.4387, training accuracy = 88.30%, test accuracy = 28.22%
2017-05-25 09:19:22.870762  iteration = 073, loss = 0.9706, val_loss = 5.4360, training accuracy = 88.32%, test accuracy = 28.26%
2017-05-25 10:05:03.384191  iteration = 074, loss = 0.9709, val_loss = 5.4370, training accuracy = 88.34%, test accuracy = 28.26%
2017-05-25 10:51:00.989924  iteration = 075, loss = 0.9711, val_loss = 5.4357, training accuracy = 88.36%, test accuracy = 28.24%
2017-05-25 11:43:31.159685  iteration = 076, loss = 0.9708, val_loss = 5.4374, training accuracy = 88.34%, test accuracy = 28.28%
2017-05-25 12:37:05.100508  iteration = 077, loss = 0.9723, val_loss = 5.4366, training accuracy = 88.32%, test accuracy = 28.34%
2017-05-25 13:25:41.212006  iteration = 078, loss = 0.9694, val_loss = 5.4396, training accuracy = 88.34%, test accuracy = 28.26%
2017-05-25 14:11:45.398753  iteration = 079, loss = 0.9706, val_loss = 5.4386, training accuracy = 88.36%, test accuracy = 28.26%
2017-05-25 14:58:54.944666  iteration = 080, loss = 0.9696, val_loss = 5.4385, training accuracy = 88.34%, test accuracy = 28.26%
2017-05-25 15:44:40.037301  iteration = 081, loss = 0.9697, val_loss = 5.4400, training accuracy = 88.36%, test accuracy = 28.24%
2017-05-25 16:27:41.595710  iteration = 082, loss = 0.9684, val_loss = 5.4396, training accuracy = 88.36%, test accuracy = 28.24%
2017-05-25 17:10:37.990840  iteration = 083, loss = 0.9691, val_loss = 5.4398, training accuracy = 88.36%, test accuracy = 28.24%
2017-05-25 17:54:30.144872  iteration = 084, loss = 0.9688, val_loss = 5.4398, training accuracy = 88.36%, test accuracy = 28.38%
2017-05-25 18:48:54.834547  iteration = 085, loss = 0.9680, val_loss = 5.4413, training accuracy = 88.36%, test accuracy = 28.34%
2017-05-25 19:46:53.765544  iteration = 086, loss = 0.9693, val_loss = 5.4418, training accuracy = 88.30%, test accuracy = 28.34%
2017-05-25 20:34:02.087238  iteration = 087, loss = 0.9679, val_loss = 5.4413, training accuracy = 88.36%, test accuracy = 28.36%
2017-05-25 21:20:24.376365  iteration = 088, loss = 0.9680, val_loss = 5.4402, training accuracy = 88.36%, test accuracy = 28.38%
2017-05-25 22:04:28.008335  iteration = 089, loss = 0.9694, val_loss = 5.4406, training accuracy = 88.40%, test accuracy = 28.34%
2017-05-25 23:01:00.236581  iteration = 090, loss = 0.9681, val_loss = 5.4415, training accuracy = 88.34%, test accuracy = 28.32%
2017-05-25 23:56:25.890886  iteration = 091, loss = 0.9667, val_loss = 5.4427, training accuracy = 88.34%, test accuracy = 28.38%
2017-05-26 00:44:25.588424  iteration = 092, loss = 0.9653, val_loss = 5.4442, training accuracy = 88.36%, test accuracy = 28.28%
2017-05-26 01:29:52.144222  iteration = 093, loss = 0.9680, val_loss = 5.4439, training accuracy = 88.44%, test accuracy = 28.22%
2017-05-26 02:15:53.736369  iteration = 094, loss = 0.9663, val_loss = 5.4440, training accuracy = 88.44%, test accuracy = 28.22%
2017-05-26 02:59:45.547335  iteration = 095, loss = 0.9665, val_loss = 5.4428, training accuracy = 88.42%, test accuracy = 28.18%
2017-05-26 03:52:01.443822  iteration = 096, loss = 0.9660, val_loss = 5.4431, training accuracy = 88.52%, test accuracy = 28.28%
2017-05-26 04:43:59.026782  iteration = 097, loss = 0.9669, val_loss = 5.4437, training accuracy = 88.54%, test accuracy = 28.26%
2017-05-26 05:30:00.216742  iteration = 098, loss = 0.9659, val_loss = 5.4452, training accuracy = 88.48%, test accuracy = 28.24%
2017-05-26 06:15:09.899417  iteration = 099, loss = 0.9650, val_loss = 5.4474, training accuracy = 88.58%, test accuracy = 28.20%
2017-05-26 06:58:52.760588  iteration = 100, loss = 0.9636, val_loss = 5.4467, training accuracy = 88.52%, test accuracy = 28.22%
2017-05-26 07:51:13.222604  iteration = 101, loss = 0.9666, val_loss = 5.4471, training accuracy = 88.54%, test accuracy = 28.28%
2017-05-26 08:43:41.937808  iteration = 102, loss = 0.9657, val_loss = 5.4460, training accuracy = 88.52%, test accuracy = 28.20%
2017-05-26 09:28:50.404956  iteration = 103, loss = 0.9656, val_loss = 5.4483, training accuracy = 88.56%, test accuracy = 28.18%
2017-05-26 10:12:37.869065  iteration = 104, loss = 0.9643, val_loss = 5.4482, training accuracy = 88.50%, test accuracy = 28.20%
2017-05-26 11:03:33.646342  iteration = 105, loss = 0.9617, val_loss = 5.4503, training accuracy = 88.54%, test accuracy = 28.12%
2017-05-26 11:58:16.638345  iteration = 106, loss = 0.9631, val_loss = 5.4483, training accuracy = 88.60%, test accuracy = 28.26%
2017-05-26 12:44:14.018165  iteration = 107, loss = 0.9630, val_loss = 5.4493, training accuracy = 88.50%, test accuracy = 28.30%
2017-05-26 13:30:08.966217  iteration = 108, loss = 0.9636, val_loss = 5.4479, training accuracy = 88.54%, test accuracy = 28.22%
2017-05-26 14:13:51.073791  iteration = 109, loss = 0.9643, val_loss = 5.4496, training accuracy = 88.56%, test accuracy = 28.22%
2017-05-26 15:05:18.980641  iteration = 110, loss = 0.9636, val_loss = 5.4503, training accuracy = 88.52%, test accuracy = 28.26%
2017-05-26 15:53:15.077054  iteration = 111, loss = 0.9611, val_loss = 5.4513, training accuracy = 88.58%, test accuracy = 28.22%
2017-05-26 16:38:42.416153  iteration = 112, loss = 0.9631, val_loss = 5.4499, training accuracy = 88.62%, test accuracy = 28.16%
2017-05-26 17:23:35.742835  iteration = 113, loss = 0.9630, val_loss = 5.4499, training accuracy = 88.60%, test accuracy = 28.22%
2017-05-26 18:05:50.902539  iteration = 114, loss = 0.9631, val_loss = 5.4514, training accuracy = 88.56%, test accuracy = 28.20%
2017-05-26 18:57:29.949024  iteration = 115, loss = 0.9621, val_loss = 5.4523, training accuracy = 88.62%, test accuracy = 28.22%
2017-05-26 19:43:11.439944  iteration = 116, loss = 0.9619, val_loss = 5.4559, training accuracy = 88.58%, test accuracy = 28.24%
2017-05-26 20:27:04.011728  iteration = 117, loss = 0.9608, val_loss = 5.4530, training accuracy = 88.64%, test accuracy = 28.18%
2017-05-26 21:13:22.721587  iteration = 118, loss = 0.9609, val_loss = 5.4526, training accuracy = 88.64%, test accuracy = 28.24%
2017-05-26 22:06:47.028565  iteration = 119, loss = 0.9609, val_loss = 5.4535, training accuracy = 88.62%, test accuracy = 28.28%
2017-05-26 22:54:15.693327  iteration = 120, loss = 0.9606, val_loss = 5.4546, training accuracy = 88.68%, test accuracy = 28.20%
2017-05-26 23:39:15.167875  iteration = 121, loss = 0.9591, val_loss = 5.4544, training accuracy = 88.70%, test accuracy = 28.24%
2017-05-27 00:23:14.921517  iteration = 122, loss = 0.9600, val_loss = 5.4566, training accuracy = 88.64%, test accuracy = 28.20%
2017-05-27 01:16:09.547332  iteration = 123, loss = 0.9598, val_loss = 5.4555, training accuracy = 88.68%, test accuracy = 28.20%
2017-05-27 02:04:53.165134  iteration = 124, loss = 0.9611, val_loss = 5.4563, training accuracy = 88.68%, test accuracy = 28.34%
2017-05-27 02:50:35.650757  iteration = 125, loss = 0.9599, val_loss = 5.4563, training accuracy = 88.72%, test accuracy = 28.22%
2017-05-27 03:36:26.626858  iteration = 126, loss = 0.9612, val_loss = 5.4552, training accuracy = 88.74%, test accuracy = 28.28%
2017-05-27 04:27:50.207500  iteration = 127, loss = 0.9591, val_loss = 5.4550, training accuracy = 88.76%, test accuracy = 28.22%
2017-05-27 05:11:22.527468  iteration = 128, loss = 0.9597, val_loss = 5.4552, training accuracy = 88.74%, test accuracy = 28.34%
2017-05-27 05:57:05.369693  iteration = 129, loss = 0.9586, val_loss = 5.4579, training accuracy = 88.76%, test accuracy = 28.28%
2017-05-27 06:49:16.674933  iteration = 130, loss = 0.9586, val_loss = 5.4590, training accuracy = 88.72%, test accuracy = 28.28%
2017-05-27 07:44:07.846487  iteration = 131, loss = 0.9582, val_loss = 5.4587, training accuracy = 88.80%, test accuracy = 28.22%
2017-05-27 08:31:29.603633  iteration = 132, loss = 0.9577, val_loss = 5.4583, training accuracy = 88.76%, test accuracy = 28.26%
2017-05-27 09:17:40.370892  iteration = 133, loss = 0.9572, val_loss = 5.4583, training accuracy = 88.70%, test accuracy = 28.26%
2017-05-27 10:03:23.341609  iteration = 134, loss = 0.9569, val_loss = 5.4594, training accuracy = 88.72%, test accuracy = 28.30%
2017-05-27 10:47:41.253623  iteration = 135, loss = 0.9568, val_loss = 5.4600, training accuracy = 88.76%, test accuracy = 28.24%
2017-05-27 11:33:34.186551  iteration = 136, loss = 0.9559, val_loss = 5.4595, training accuracy = 88.72%, test accuracy = 28.24%
2017-05-27 12:16:49.822437  iteration = 137, loss = 0.9571, val_loss = 5.4601, training accuracy = 88.78%, test accuracy = 28.24%
2017-05-27 13:02:03.306250  iteration = 138, loss = 0.9568, val_loss = 5.4621, training accuracy = 88.84%, test accuracy = 28.22%
2017-05-27 13:54:17.850966  iteration = 139, loss = 0.9563, val_loss = 5.4616, training accuracy = 88.86%, test accuracy = 28.22%
2017-05-27 14:48:39.376068  iteration = 140, loss = 0.9550, val_loss = 5.4631, training accuracy = 88.80%, test accuracy = 28.14%
2017-05-27 15:34:42.374520  iteration = 141, loss = 0.9567, val_loss = 5.4612, training accuracy = 88.80%, test accuracy = 28.14%
2017-05-27 16:20:20.755762  iteration = 142, loss = 0.9552, val_loss = 5.4630, training accuracy = 88.90%, test accuracy = 28.12%
2017-05-27 17:05:17.087241  iteration = 143, loss = 0.9562, val_loss = 5.4642, training accuracy = 88.90%, test accuracy = 28.14%
2017-05-27 17:57:28.600642  iteration = 144, loss = 0.9536, val_loss = 5.4637, training accuracy = 88.80%, test accuracy = 28.20%
2017-05-27 18:45:06.144257  iteration = 145, loss = 0.9548, val_loss = 5.4650, training accuracy = 88.78%, test accuracy = 28.20%
2017-05-27 19:30:54.747768  iteration = 146, loss = 0.9547, val_loss = 5.4644, training accuracy = 88.84%, test accuracy = 28.20%
2017-05-27 20:14:39.136677  iteration = 147, loss = 0.9558, val_loss = 5.4644, training accuracy = 88.90%, test accuracy = 28.24%
2017-05-27 21:05:18.754484  iteration = 148, loss = 0.9540, val_loss = 5.4643, training accuracy = 88.90%, test accuracy = 28.24%
2017-05-27 21:56:05.824622  iteration = 149, loss = 0.9548, val_loss = 5.4652, training accuracy = 88.88%, test accuracy = 28.28%
2017-05-27 22:40:56.263950  iteration = 150, loss = 0.9541, val_loss = 5.4653, training accuracy = 88.92%, test accuracy = 28.20%
2017-05-27 23:25:32.029722  iteration = 151, loss = 0.9527, val_loss = 5.4662, training accuracy = 88.90%, test accuracy = 28.20%
2017-05-28 00:18:11.869477  iteration = 152, loss = 0.9522, val_loss = 5.4661, training accuracy = 88.92%, test accuracy = 28.16%
2017-05-28 01:07:42.190605  iteration = 153, loss = 0.9536, val_loss = 5.4671, training accuracy = 89.04%, test accuracy = 28.20%
2017-05-28 01:54:00.625130  iteration = 154, loss = 0.9525, val_loss = 5.4677, training accuracy = 88.90%, test accuracy = 28.20%
2017-05-28 02:38:30.502290  iteration = 155, loss = 0.9523, val_loss = 5.4690, training accuracy = 88.94%, test accuracy = 28.20%
2017-05-28 03:31:03.106633  iteration = 156, loss = 0.9532, val_loss = 5.4680, training accuracy = 89.04%, test accuracy = 28.24%
2017-05-28 04:15:58.435420  iteration = 157, loss = 0.9507, val_loss = 5.4682, training accuracy = 88.94%, test accuracy = 28.24%
2017-05-28 05:01:33.082100  iteration = 158, loss = 0.9534, val_loss = 5.4678, training accuracy = 88.92%, test accuracy = 28.24%
2017-05-28 05:45:56.247707  iteration = 159, loss = 0.9518, val_loss = 5.4682, training accuracy = 88.92%, test accuracy = 28.20%
2017-05-28 06:31:00.604956  iteration = 160, loss = 0.9505, val_loss = 5.4704, training accuracy = 88.98%, test accuracy = 28.20%
2017-05-28 07:23:18.488915  iteration = 161, loss = 0.9513, val_loss = 5.4705, training accuracy = 88.92%, test accuracy = 28.28%
2017-05-28 08:17:30.529384  iteration = 162, loss = 0.9515, val_loss = 5.4699, training accuracy = 88.92%, test accuracy = 28.28%
2017-05-28 09:03:28.501541  iteration = 163, loss = 0.9511, val_loss = 5.4707, training accuracy = 88.96%, test accuracy = 28.24%
2017-05-28 09:49:18.965084  iteration = 164, loss = 0.9490, val_loss = 5.4720, training accuracy = 88.90%, test accuracy = 28.28%
2017-05-28 10:39:39.160852  iteration = 165, loss = 0.9522, val_loss = 5.4716, training accuracy = 88.98%, test accuracy = 28.26%
2017-05-28 11:32:36.623050  iteration = 166, loss = 0.9509, val_loss = 5.4717, training accuracy = 89.00%, test accuracy = 28.26%
2017-05-28 12:21:28.979415  iteration = 167, loss = 0.9513, val_loss = 5.4720, training accuracy = 88.96%, test accuracy = 28.24%
2017-05-28 13:07:29.690095  iteration = 168, loss = 0.9496, val_loss = 5.4739, training accuracy = 88.90%, test accuracy = 28.20%
2017-05-28 13:51:26.161929  iteration = 169, loss = 0.9493, val_loss = 5.4724, training accuracy = 89.02%, test accuracy = 28.32%
2017-05-28 14:43:05.017359  iteration = 170, loss = 0.9488, val_loss = 5.4761, training accuracy = 89.04%, test accuracy = 28.18%
2017-05-28 15:34:03.230069  iteration = 171, loss = 0.9488, val_loss = 5.4725, training accuracy = 89.04%, test accuracy = 28.16%
2017-05-28 16:19:49.666606  iteration = 172, loss = 0.9478, val_loss = 5.4738, training accuracy = 89.02%, test accuracy = 28.22%
2017-05-28 17:04:42.019480  iteration = 173, loss = 0.9482, val_loss = 5.4742, training accuracy = 89.00%, test accuracy = 28.24%
2017-05-28 17:55:31.184736  iteration = 174, loss = 0.9486, val_loss = 5.4746, training accuracy = 88.98%, test accuracy = 28.26%
2017-05-28 18:50:08.648382  iteration = 175, loss = 0.9469, val_loss = 5.4746, training accuracy = 88.90%, test accuracy = 28.20%
2017-05-28 19:39:06.394463  iteration = 176, loss = 0.9481, val_loss = 5.4781, training accuracy = 89.00%, test accuracy = 28.20%
2017-05-28 20:24:23.992765  iteration = 177, loss = 0.9478, val_loss = 5.4788, training accuracy = 89.04%, test accuracy = 28.26%
2017-05-28 21:08:24.698464  iteration = 178, loss = 0.9484, val_loss = 5.4755, training accuracy = 89.02%, test accuracy = 28.18%
2017-05-28 22:01:28.042277  iteration = 179, loss = 0.9480, val_loss = 5.4771, training accuracy = 89.06%, test accuracy = 28.16%
2017-05-28 22:48:02.596526  iteration = 180, loss = 0.9478, val_loss = 5.4764, training accuracy = 89.10%, test accuracy = 28.16%
2017-05-28 23:33:22.474679  iteration = 181, loss = 0.9455, val_loss = 5.4790, training accuracy = 89.08%, test accuracy = 28.22%
2017-05-29 00:19:31.515935  iteration = 182, loss = 0.9448, val_loss = 5.4781, training accuracy = 89.10%, test accuracy = 28.10%
2017-05-29 01:12:15.639416  iteration = 183, loss = 0.9468, val_loss = 5.4780, training accuracy = 89.10%, test accuracy = 28.12%
2017-05-29 02:02:40.727335  iteration = 184, loss = 0.9452, val_loss = 5.4793, training accuracy = 89.06%, test accuracy = 28.20%
2017-05-29 02:46:23.898978  iteration = 185, loss = 0.9449, val_loss = 5.4804, training accuracy = 89.10%, test accuracy = 28.12%
2017-05-29 03:30:25.175085  iteration = 186, loss = 0.9441, val_loss = 5.4802, training accuracy = 89.16%, test accuracy = 28.20%
2017-05-29 04:22:44.561197  iteration = 187, loss = 0.9454, val_loss = 5.4795, training accuracy = 89.16%, test accuracy = 28.14%
2017-05-29 05:10:47.410963  iteration = 188, loss = 0.9444, val_loss = 5.4806, training accuracy = 89.16%, test accuracy = 28.18%
2017-05-29 05:56:54.048218  iteration = 189, loss = 0.9446, val_loss = 5.4790, training accuracy = 89.06%, test accuracy = 28.12%
2017-05-29 06:41:51.357177  iteration = 190, loss = 0.9451, val_loss = 5.4810, training accuracy = 89.18%, test accuracy = 28.18%
2017-05-29 07:34:04.952424  iteration = 191, loss = 0.9452, val_loss = 5.4823, training accuracy = 89.20%, test accuracy = 28.20%
2017-05-29 08:24:24.393616  iteration = 192, loss = 0.9437, val_loss = 5.4823, training accuracy = 89.20%, test accuracy = 28.26%
2017-05-29 09:10:19.409421  iteration = 193, loss = 0.9438, val_loss = 5.4814, training accuracy = 89.12%, test accuracy = 28.16%
2017-05-29 09:55:19.923846  iteration = 194, loss = 0.9435, val_loss = 5.4838, training accuracy = 89.18%, test accuracy = 28.16%
2017-05-29 10:40:47.267513  iteration = 195, loss = 0.9435, val_loss = 5.4833, training accuracy = 89.20%, test accuracy = 28.22%
2017-05-29 11:26:35.423409  iteration = 196, loss = 0.9442, val_loss = 5.4834, training accuracy = 89.12%, test accuracy = 28.14%
2017-05-29 12:18:20.118052  iteration = 197, loss = 0.9438, val_loss = 5.4838, training accuracy = 89.24%, test accuracy = 28.06%
2017-05-29 13:11:19.990845  iteration = 198, loss = 0.9433, val_loss = 5.4838, training accuracy = 89.28%, test accuracy = 28.16%
2017-05-29 14:00:30.281172  iteration = 199, loss = 0.9426, val_loss = 5.4842, training accuracy = 89.18%, test accuracy = 28.26%

Continuing training with a further run
sjb@persephone:sjb_word_model$ ./train.py theresa_may_contributions.json
2017-05-30 11:21:51.432232  iteration = 000, loss = 0.9419, val_loss = 5.4852, training accuracy = 89.30%, test accuracy = 28.26%
2017-05-30 12:07:28.640807  iteration = 001, loss = 0.9420, val_loss = 5.4849, training accuracy = 89.24%, test accuracy = 28.20%
2017-05-30 12:52:13.173819  iteration = 002, loss = 0.9407, val_loss = 5.4883, training accuracy = 89.22%, test accuracy = 28.12%
2017-05-30 13:37:08.982544  iteration = 003, loss = 0.9414, val_loss = 5.4863, training accuracy = 89.20%, test accuracy = 28.18%
2017-05-30 14:20:31.441170  iteration = 004, loss = 0.9423, val_loss = 5.4877, training accuracy = 89.16%, test accuracy = 28.14%
2017-05-30 15:03:56.709480  iteration = 005, loss = 0.9419, val_loss = 5.4875, training accuracy = 89.12%, test accuracy = 28.16%
2017-05-30 15:47:55.708851  iteration = 006, loss = 0.9421, val_loss = 5.4870, training accuracy = 89.22%, test accuracy = 28.16%
2017-05-30 16:40:15.105576  iteration = 007, loss = 0.9424, val_loss = 5.4875, training accuracy = 89.22%, test accuracy = 28.10%
2017-05-30 17:30:51.967111  iteration = 008, loss = 0.9407, val_loss = 5.4890, training accuracy = 89.28%, test accuracy = 28.08%
2017-05-30 18:17:42.546393  iteration = 009, loss = 0.9402, val_loss = 5.4891, training accuracy = 89.24%, test accuracy = 28.14%
2017-05-30 19:03:33.263533  iteration = 010, loss = 0.9398, val_loss = 5.4880, training accuracy = 89.24%, test accuracy = 28.14%
2017-05-30 19:48:01.989427  iteration = 011, loss = 0.9386, val_loss = 5.4901, training accuracy = 89.26%, test accuracy = 28.08%
2017-05-30 20:31:50.293279  iteration = 012, loss = 0.9389, val_loss = 5.4905, training accuracy = 89.26%, test accuracy = 28.20%
2017-05-30 21:22:37.243807  iteration = 013, loss = 0.9394, val_loss = 5.4903, training accuracy = 89.34%, test accuracy = 28.10%
2017-05-30 22:06:24.329283  iteration = 014, loss = 0.9379, val_loss = 5.4916, training accuracy = 89.30%, test accuracy = 28.16%
2017-05-30 22:51:45.099032  iteration = 015, loss = 0.9389, val_loss = 5.4913, training accuracy = 89.30%, test accuracy = 28.14%
2017-05-30 23:35:36.539554  iteration = 016, loss = 0.9393, val_loss = 5.4909, training accuracy = 89.36%, test accuracy = 28.06%
2017-05-31 00:27:27.406457  iteration = 017, loss = 0.9395, val_loss = 5.4913, training accuracy = 89.26%, test accuracy = 28.14%
2017-05-31 01:20:09.279802  iteration = 018, loss = 0.9384, val_loss = 5.4932, training accuracy = 89.30%, test accuracy = 28.18%
2017-05-31 02:08:33.703997  iteration = 019, loss = 0.9384, val_loss = 5.4939, training accuracy = 89.26%, test accuracy = 28.10%
2017-05-31 02:55:44.043939  iteration = 020, loss = 0.9380, val_loss = 5.4940, training accuracy = 89.30%, test accuracy = 28.14%
2017-05-31 03:40:51.305926  iteration = 021, loss = 0.9370, val_loss = 5.4929, training accuracy = 89.28%, test accuracy = 28.16%
2017-05-31 04:24:51.884477  iteration = 022, loss = 0.9370, val_loss = 5.4935, training accuracy = 89.30%, test accuracy = 28.10%
2017-05-31 05:08:56.487857  iteration = 023, loss = 0.9371, val_loss = 5.4944, training accuracy = 89.28%, test accuracy = 28.10%
2017-05-31 06:00:47.971887  iteration = 024, loss = 0.9392, val_loss = 5.4950, training accuracy = 89.32%, test accuracy = 28.18%
2017-05-31 06:47:47.019402  iteration = 025, loss = 0.9367, val_loss = 5.4946, training accuracy = 89.28%, test accuracy = 28.06%
2017-05-31 07:33:30.968116  iteration = 026, loss = 0.9370, val_loss = 5.4950, training accuracy = 89.34%, test accuracy = 28.12%
2017-05-31 08:18:14.537870  iteration = 027, loss = 0.9370, val_loss = 5.4945, training accuracy = 89.28%, test accuracy = 28.16%
2017-05-31 09:04:58.687989  iteration = 028, loss = 0.9371, val_loss = 5.4946, training accuracy = 89.32%, test accuracy = 28.04%
2017-05-31 09:51:50.105790  iteration = 029, loss = 0.9362, val_loss = 5.4952, training accuracy = 89.36%, test accuracy = 28.06%
2017-05-31 10:37:46.813502  iteration = 030, loss = 0.9348, val_loss = 5.4952, training accuracy = 89.32%, test accuracy = 28.16%
2017-05-31 11:22:46.474102  iteration = 031, loss = 0.9366, val_loss = 5.4969, training accuracy = 89.38%, test accuracy = 28.18%
2017-05-31 12:05:54.569802  iteration = 032, loss = 0.9354, val_loss = 5.4963, training accuracy = 89.34%, test accuracy = 28.14%
2017-05-31 12:48:59.292446  iteration = 033, loss = 0.9341, val_loss = 5.4974, training accuracy = 89.40%, test accuracy = 28.10%
2017-05-31 13:32:05.197863  iteration = 034, loss = 0.9334, val_loss = 5.4988, training accuracy = 89.40%, test accuracy = 28.10%
2017-05-31 14:14:04.689972  iteration = 035, loss = 0.9352, val_loss = 5.5002, training accuracy = 89.38%, test accuracy = 28.12%
2017-05-31 14:54:51.804555  iteration = 036, loss = 0.9351, val_loss = 5.4976, training accuracy = 89.36%, test accuracy = 28.04%
2017-05-31 15:36:20.267254  iteration = 037, loss = 0.9339, val_loss = 5.4968, training accuracy = 89.40%, test accuracy = 28.12%
2017-05-31 16:17:52.630263  iteration = 038, loss = 0.9329, val_loss = 5.4996, training accuracy = 89.38%, test accuracy = 28.18%
2017-05-31 17:01:38.884399  iteration = 039, loss = 0.9337, val_loss = 5.5006, training accuracy = 89.44%, test accuracy = 28.14%
2017-05-31 17:52:21.925472  iteration = 040, loss = 0.9323, val_loss = 5.4992, training accuracy = 89.48%, test accuracy = 28.14%
2017-05-31 18:35:30.919128  iteration = 041, loss = 0.9323, val_loss = 5.5009, training accuracy = 89.44%, test accuracy = 28.08%
2017-05-31 19:18:35.187696  iteration = 042, loss = 0.9321, val_loss = 5.5020, training accuracy = 89.40%, test accuracy = 28.12%
2017-05-31 20:01:43.693235  iteration = 043, loss = 0.9332, val_loss = 5.5015, training accuracy = 89.50%, test accuracy = 28.14%
2017-05-31 20:44:57.953332  iteration = 044, loss = 0.9334, val_loss = 5.5015, training accuracy = 89.42%, test accuracy = 28.18%
2017-05-31 21:28:09.682277  iteration = 045, loss = 0.9321, val_loss = 5.5027, training accuracy = 89.52%, test accuracy = 28.18%
2017-05-31 22:12:25.184231  iteration = 046, loss = 0.9326, val_loss = 5.5027, training accuracy = 89.46%, test accuracy = 28.24%
2017-05-31 22:58:17.804482  iteration = 047, loss = 0.9316, val_loss = 5.5031, training accuracy = 89.50%, test accuracy = 28.16%
2017-05-31 23:45:17.106538  iteration = 048, loss = 0.9321, val_loss = 5.5043, training accuracy = 89.44%, test accuracy = 28.18%
2017-06-01 00:31:52.923428  iteration = 049, loss = 0.9311, val_loss = 5.5030, training accuracy = 89.52%, test accuracy = 28.10%
2017-06-01 01:18:12.446508  iteration = 050, loss = 0.9313, val_loss = 5.5041, training accuracy = 89.48%, test accuracy = 28.06%
2017-06-01 02:02:20.296971  iteration = 051, loss = 0.9327, val_loss = 5.5042, training accuracy = 89.46%, test accuracy = 28.08%
2017-06-01 02:46:31.252487  iteration = 052, loss = 0.9305, val_loss = 5.5043, training accuracy = 89.46%, test accuracy = 28.12%
2017-06-01 03:40:00.221285  iteration = 053, loss = 0.9292, val_loss = 5.5030, training accuracy = 89.50%, test accuracy = 28.06%
2017-06-01 04:27:29.413691  iteration = 054, loss = 0.9290, val_loss = 5.5065, training accuracy = 89.48%, test accuracy = 28.02%
2017-06-01 05:10:55.106485  iteration = 055, loss = 0.9321, val_loss = 5.5046, training accuracy = 89.50%, test accuracy = 28.00%
2017-06-01 05:54:12.029166  iteration = 056, loss = 0.9280, val_loss = 5.5064, training accuracy = 89.60%, test accuracy = 28.08%
2017-06-01 06:37:23.279830  iteration = 057, loss = 0.9287, val_loss = 5.5063, training accuracy = 89.54%, test accuracy = 28.08%
2017-06-01 07:20:54.626454  iteration = 058, loss = 0.9310, val_loss = 5.5070, training accuracy = 89.60%, test accuracy = 28.06%
2017-06-01 08:05:36.097076  iteration = 059, loss = 0.9290, val_loss = 5.5089, training accuracy = 89.62%, test accuracy = 28.04%
2017-06-01 08:50:44.293561  iteration = 060, loss = 0.9304, val_loss = 5.5055, training accuracy = 89.60%, test accuracy = 28.06%
2017-06-01 09:36:29.683794  iteration = 061, loss = 0.9293, val_loss = 5.5076, training accuracy = 89.56%, test accuracy = 27.96%
2017-06-01 10:22:26.060870  iteration = 062, loss = 0.9291, val_loss = 5.5098, training accuracy = 89.52%, test accuracy = 28.00%
2017-06-01 11:06:40.559033  iteration = 063, loss = 0.9286, val_loss = 5.5084, training accuracy = 89.56%, test accuracy = 28.10%
2017-06-01 11:50:52.458484  iteration = 064, loss = 0.9288, val_loss = 5.5083, training accuracy = 89.58%, test accuracy = 28.04%
2017-06-01 12:34:54.194826  iteration = 065, loss = 0.9296, val_loss = 5.5097, training accuracy = 89.56%, test accuracy = 28.02%
2017-06-01 13:26:43.073293  iteration = 066, loss = 0.9290, val_loss = 5.5111, training accuracy = 89.52%, test accuracy = 28.12%
2017-06-01 14:18:32.518353  iteration = 067, loss = 0.9271, val_loss = 5.5106, training accuracy = 89.56%, test accuracy = 28.02%
2017-06-01 15:06:16.456088  iteration = 068, loss = 0.9288, val_loss = 5.5099, training accuracy = 89.56%, test accuracy = 28.06%
2017-06-01 15:51:55.493921  iteration = 069, loss = 0.9283, val_loss = 5.5116, training accuracy = 89.58%, test accuracy = 28.08%
2017-06-01 16:35:54.200521  iteration = 070, loss = 0.9275, val_loss = 5.5129, training accuracy = 89.62%, test accuracy = 28.08%
2017-06-01 17:28:46.615775  iteration = 071, loss = 0.9263, val_loss = 5.5116, training accuracy = 89.62%, test accuracy = 28.08%
2017-06-01 18:21:28.402524  iteration = 072, loss = 0.9265, val_loss = 5.5129, training accuracy = 89.62%, test accuracy = 28.04%
2017-06-01 19:09:17.340234  iteration = 073, loss = 0.9266, val_loss = 5.5110, training accuracy = 89.66%, test accuracy = 28.08%
2017-06-01 19:54:12.111316  iteration = 074, loss = 0.9261, val_loss = 5.5138, training accuracy = 89.62%, test accuracy = 28.00%
2017-06-01 20:38:03.182763  iteration = 075, loss = 0.9251, val_loss = 5.5144, training accuracy = 89.60%, test accuracy = 28.04%
2017-06-01 21:21:00.230919  iteration = 076, loss = 0.9268, val_loss = 5.5160, training accuracy = 89.60%, test accuracy = 28.04%
2017-06-01 22:04:57.240112  iteration = 077, loss = 0.9251, val_loss = 5.5140, training accuracy = 89.70%, test accuracy = 28.08%
2017-06-01 22:57:42.820559  iteration = 078, loss = 0.9260, val_loss = 5.5133, training accuracy = 89.74%, test accuracy = 28.10%
2017-06-01 23:46:00.864005  iteration = 079, loss = 0.9256, val_loss = 5.5138, training accuracy = 89.72%, test accuracy = 28.10%
2017-06-02 00:32:13.920501  iteration = 080, loss = 0.9255, val_loss = 5.5143, training accuracy = 89.76%, test accuracy = 28.02%
2017-06-02 01:15:58.885863  iteration = 081, loss = 0.9248, val_loss = 5.5155, training accuracy = 89.72%, test accuracy = 28.04%
2017-06-02 01:59:56.504637  iteration = 082, loss = 0.9255, val_loss = 5.5150, training accuracy = 89.76%, test accuracy = 28.06%
2017-06-02 02:51:43.232144  iteration = 083, loss = 0.9254, val_loss = 5.5158, training accuracy = 89.70%, test accuracy = 28.16%
2017-06-02 03:41:36.171178  iteration = 084, loss = 0.9238, val_loss = 5.5154, training accuracy = 89.74%, test accuracy = 27.98%
2017-06-02 04:27:40.338488  iteration = 085, loss = 0.9242, val_loss = 5.5173, training accuracy = 89.80%, test accuracy = 28.00%
2017-06-02 05:13:54.121333  iteration = 086, loss = 0.9233, val_loss = 5.5164, training accuracy = 89.82%, test accuracy = 27.94%
2017-06-02 06:06:53.220411  iteration = 087, loss = 0.9243, val_loss = 5.5172, training accuracy = 89.76%, test accuracy = 28.04%
2017-06-02 06:59:19.479659  iteration = 088, loss = 0.9253, val_loss = 5.5162, training accuracy = 89.84%, test accuracy = 28.10%
2017-06-02 07:47:03.375920  iteration = 089, loss = 0.9226, val_loss = 5.5191, training accuracy = 89.86%, test accuracy = 28.14%
2017-06-02 08:33:24.460306  iteration = 090, loss = 0.9240, val_loss = 5.5194, training accuracy = 89.80%, test accuracy = 28.08%
2017-06-02 09:21:27.021395  iteration = 091, loss = 0.9217, val_loss = 5.5207, training accuracy = 89.88%, test accuracy = 28.06%
2017-06-02 10:14:15.631957  iteration = 092, loss = 0.9226, val_loss = 5.5186, training accuracy = 89.82%, test accuracy = 28.06%
2017-06-02 11:04:32.494788  iteration = 093, loss = 0.9216, val_loss = 5.5211, training accuracy = 89.76%, test accuracy = 28.06%
2017-06-02 11:50:40.061411  iteration = 094, loss = 0.9229, val_loss = 5.5192, training accuracy = 89.82%, test accuracy = 28.06%
2017-06-02 12:37:02.906195  iteration = 095, loss = 0.9226, val_loss = 5.5198, training accuracy = 89.88%, test accuracy = 28.12%
2017-06-02 13:24:36.627985  iteration = 096, loss = 0.9218, val_loss = 5.5200, training accuracy = 89.84%, test accuracy = 28.16%
2017-06-02 14:17:14.109411  iteration = 097, loss = 0.9219, val_loss = 5.5198, training accuracy = 89.82%, test accuracy = 28.10%
2017-06-02 15:05:08.762885  iteration = 098, loss = 0.9214, val_loss = 5.5217, training accuracy = 89.94%, test accuracy = 28.12%
2017-06-02 15:49:19.036603  iteration = 099, loss = 0.9221, val_loss = 5.5222, training accuracy = 89.92%, test accuracy = 28.02%
2017-06-02 16:33:14.434933  iteration = 100, loss = 0.9206, val_loss = 5.5230, training accuracy = 89.90%, test accuracy = 27.98%
2017-06-02 17:25:43.270009  iteration = 101, loss = 0.9213, val_loss = 5.5210, training accuracy = 89.86%, test accuracy = 28.08%
2017-06-02 18:09:37.501091  iteration = 102, loss = 0.9199, val_loss = 5.5234, training accuracy = 89.88%, test accuracy = 28.08%
2017-06-02 18:53:27.538898  iteration = 103, loss = 0.9191, val_loss = 5.5253, training accuracy = 89.88%, test accuracy = 28.04%
2017-06-02 19:45:39.860881  iteration = 104, loss = 0.9201, val_loss = 5.5244, training accuracy = 89.86%, test accuracy = 28.14%
2017-06-02 20:38:39.958697  iteration = 105, loss = 0.9192, val_loss = 5.5251, training accuracy = 89.86%, test accuracy = 28.14%
2017-06-02 21:26:51.207467  iteration = 106, loss = 0.9192, val_loss = 5.5239, training accuracy = 89.92%, test accuracy = 28.20%
2017-06-02 22:12:31.309102  iteration = 107, loss = 0.9198, val_loss = 5.5243, training accuracy = 89.90%, test accuracy = 28.06%
2017-06-02 22:56:34.357910  iteration = 108, loss = 0.9210, val_loss = 5.5248, training accuracy = 89.88%, test accuracy = 28.00%
2017-06-02 23:48:40.476322  iteration = 109, loss = 0.9199, val_loss = 5.5264, training accuracy = 89.88%, test accuracy = 27.96%
2017-06-03 00:35:04.815404  iteration = 110, loss = 0.9189, val_loss = 5.5274, training accuracy = 89.90%, test accuracy = 27.96%
2017-06-03 01:19:06.973305  iteration = 111, loss = 0.9193, val_loss = 5.5267, training accuracy = 89.92%, test accuracy = 28.00%
2017-06-03 02:01:59.182973  iteration = 112, loss = 0.9198, val_loss = 5.5263, training accuracy = 89.90%, test accuracy = 28.00%
2017-06-03 02:44:56.925482  iteration = 113, loss = 0.9187, val_loss = 5.5260, training accuracy = 89.92%, test accuracy = 28.02%
2017-06-03 03:28:46.028530  iteration = 114, loss = 0.9187, val_loss = 5.5265, training accuracy = 89.94%, test accuracy = 28.00%
2017-06-03 04:12:53.031343  iteration = 115, loss = 0.9181, val_loss = 5.5265, training accuracy = 89.98%, test accuracy = 27.88%
2017-06-03 05:03:35.813021  iteration = 116, loss = 0.9183, val_loss = 5.5275, training accuracy = 89.98%, test accuracy = 28.08%
2017-06-03 05:47:34.060802  iteration = 117, loss = 0.9192, val_loss = 5.5287, training accuracy = 90.00%, test accuracy = 28.06%
2017-06-03 06:33:46.508355  iteration = 118, loss = 0.9164, val_loss = 5.5285, training accuracy = 89.96%, test accuracy = 28.08%
2017-06-03 07:18:35.339671  iteration = 119, loss = 0.9169, val_loss = 5.5295, training accuracy = 89.94%, test accuracy = 28.04%
2017-06-03 08:03:11.046933  iteration = 120, loss = 0.9178, val_loss = 5.5273, training accuracy = 89.94%, test accuracy = 28.08%
2017-06-03 08:46:12.111161  iteration = 121, loss = 0.9173, val_loss = 5.5288, training accuracy = 89.98%, test accuracy = 28.02%
2017-06-03 09:29:16.516631  iteration = 122, loss = 0.9169, val_loss = 5.5283, training accuracy = 89.94%, test accuracy = 28.02%
2017-06-03 10:15:07.707152  iteration = 123, loss = 0.9167, val_loss = 5.5300, training accuracy = 89.94%, test accuracy = 28.02%
2017-06-03 11:05:56.132201  iteration = 124, loss = 0.9160, val_loss = 5.5311, training accuracy = 90.02%, test accuracy = 28.00%
2017-06-03 11:56:15.702772  iteration = 125, loss = 0.9172, val_loss = 5.5321, training accuracy = 89.92%, test accuracy = 28.00%
2017-06-03 12:44:01.122502  iteration = 126, loss = 0.9159, val_loss = 5.5319, training accuracy = 89.98%, test accuracy = 28.00%
2017-06-03 13:29:47.788511  iteration = 127, loss = 0.9154, val_loss = 5.5292, training accuracy = 89.96%, test accuracy = 28.00%
2017-06-03 14:15:09.692134  iteration = 128, loss = 0.9151, val_loss = 5.5340, training accuracy = 90.00%, test accuracy = 28.06%
2017-06-03 14:59:12.722062  iteration = 129, loss = 0.9154, val_loss = 5.5324, training accuracy = 89.96%, test accuracy = 28.12%
2017-06-03 15:51:13.529079  iteration = 130, loss = 0.9135, val_loss = 5.5345, training accuracy = 90.04%, test accuracy = 28.10%
2017-06-03 16:41:16.679944  iteration = 131, loss = 0.9149, val_loss = 5.5339, training accuracy = 90.06%, test accuracy = 27.96%
2017-06-03 17:26:53.748762  iteration = 132, loss = 0.9148, val_loss = 5.5337, training accuracy = 90.08%, test accuracy = 28.04%
2017-06-03 18:09:50.620957  iteration = 133, loss = 0.9145, val_loss = 5.5339, training accuracy = 90.02%, test accuracy = 28.02%
2017-06-03 18:53:43.329199  iteration = 134, loss = 0.9144, val_loss = 5.5348, training accuracy = 90.04%, test accuracy = 28.06%
2017-06-03 19:45:41.877377  iteration = 135, loss = 0.9158, val_loss = 5.5329, training accuracy = 90.02%, test accuracy = 28.06%
2017-06-03 20:37:22.838612  iteration = 136, loss = 0.9135, val_loss = 5.5361, training accuracy = 90.06%, test accuracy = 28.00%
2017-06-03 21:25:46.571028  iteration = 137, loss = 0.9143, val_loss = 5.5345, training accuracy = 90.02%, test accuracy = 28.00%
2017-06-03 22:11:24.878641  iteration = 138, loss = 0.9138, val_loss = 5.5361, training accuracy = 90.00%, test accuracy = 28.00%
2017-06-03 22:56:51.589326  iteration = 139, loss = 0.9135, val_loss = 5.5374, training accuracy = 90.02%, test accuracy = 27.96%
2017-06-03 23:40:55.584242  iteration = 140, loss = 0.9120, val_loss = 5.5372, training accuracy = 90.02%, test accuracy = 27.94%
2017-06-04 00:33:22.988136  iteration = 141, loss = 0.9128, val_loss = 5.5369, training accuracy = 89.98%, test accuracy = 27.96%
2017-06-04 01:20:51.006124  iteration = 142, loss = 0.9130, val_loss = 5.5377, training accuracy = 90.00%, test accuracy = 27.98%
2017-06-04 02:03:49.325682  iteration = 143, loss = 0.9123, val_loss = 5.5369, training accuracy = 89.96%, test accuracy = 28.00%
2017-06-04 02:46:59.490744  iteration = 144, loss = 0.9111, val_loss = 5.5393, training accuracy = 90.04%, test accuracy = 28.02%
2017-06-04 03:30:51.359804  iteration = 145, loss = 0.9128, val_loss = 5.5379, training accuracy = 90.02%, test accuracy = 28.10%
2017-06-04 04:22:51.623927  iteration = 146, loss = 0.9131, val_loss = 5.5380, training accuracy = 90.04%, test accuracy = 28.04%
2017-06-04 05:11:45.402789  iteration = 147, loss = 0.9136, val_loss = 5.5399, training accuracy = 90.06%, test accuracy = 27.98%
2017-06-04 05:57:05.508060  iteration = 148, loss = 0.9120, val_loss = 5.5385, training accuracy = 90.00%, test accuracy = 28.04%
2017-06-04 06:41:02.836888  iteration = 149, loss = 0.9104, val_loss = 5.5392, training accuracy = 90.04%, test accuracy = 28.02%
2017-06-04 07:30:11.335926  iteration = 150, loss = 0.9116, val_loss = 5.5400, training accuracy = 90.06%, test accuracy = 28.00%
2017-06-04 08:22:32.110350  iteration = 151, loss = 0.9098, val_loss = 5.5414, training accuracy = 90.08%, test accuracy = 28.02%
2017-06-04 09:10:14.645167  iteration = 152, loss = 0.9119, val_loss = 5.5391, training accuracy = 90.08%, test accuracy = 28.06%
2017-06-04 09:55:47.496709  iteration = 153, loss = 0.9109, val_loss = 5.5410, training accuracy = 90.08%, test accuracy = 28.10%
2017-06-04 10:41:52.589152  iteration = 154, loss = 0.9108, val_loss = 5.5419, training accuracy = 90.06%, test accuracy = 28.08%
2017-06-04 11:28:25.331155  iteration = 155, loss = 0.9123, val_loss = 5.5417, training accuracy = 90.04%, test accuracy = 28.06%
2017-06-04 12:11:52.231964  iteration = 156, loss = 0.9087, val_loss = 5.5420, training accuracy = 90.16%, test accuracy = 28.04%
2017-06-04 12:55:42.991279  iteration = 157, loss = 0.9096, val_loss = 5.5416, training accuracy = 90.14%, test accuracy = 28.10%
2017-06-04 13:48:08.248707  iteration = 158, loss = 0.9097, val_loss = 5.5433, training accuracy = 90.18%, test accuracy = 28.12%
2017-06-04 14:39:12.205008  iteration = 159, loss = 0.9101, val_loss = 5.5418, training accuracy = 90.16%, test accuracy = 28.06%
2017-06-04 15:23:12.660853  iteration = 160, loss = 0.9086, val_loss = 5.5450, training accuracy = 90.12%, test accuracy = 28.06%
2017-06-04 16:06:42.929538  iteration = 161, loss = 0.9075, val_loss = 5.5427, training accuracy = 90.14%, test accuracy = 28.06%
2017-06-04 16:51:44.564726  iteration = 162, loss = 0.9082, val_loss = 5.5435, training accuracy = 90.14%, test accuracy = 28.04%
2017-06-04 17:38:32.531324  iteration = 163, loss = 0.9089, val_loss = 5.5470, training accuracy = 90.16%, test accuracy = 28.10%
2017-06-04 18:25:53.778374  iteration = 164, loss = 0.9080, val_loss = 5.5456, training accuracy = 90.18%, test accuracy = 28.04%
2017-06-04 19:12:49.918923  iteration = 165, loss = 0.9080, val_loss = 5.5451, training accuracy = 90.18%, test accuracy = 28.06%
2017-06-04 19:58:00.679137  iteration = 166, loss = 0.9082, val_loss = 5.5464, training accuracy = 90.20%, test accuracy = 28.02%
2017-06-04 20:42:39.250174  iteration = 167, loss = 0.9076, val_loss = 5.5466, training accuracy = 90.18%, test accuracy = 28.04%
2017-06-04 21:27:14.441210  iteration = 168, loss = 0.9076, val_loss = 5.5466, training accuracy = 90.16%, test accuracy = 27.92%
2017-06-04 22:12:14.546398  iteration = 169, loss = 0.9081, val_loss = 5.5476, training accuracy = 90.24%, test accuracy = 28.04%
2017-06-04 22:57:28.173106  iteration = 170, loss = 0.9074, val_loss = 5.5474, training accuracy = 90.12%, test accuracy = 27.94%
2017-06-04 23:42:47.661672  iteration = 171, loss = 0.9080, val_loss = 5.5468, training accuracy = 90.20%, test accuracy = 28.02%
2017-06-05 00:27:51.899773  iteration = 172, loss = 0.9086, val_loss = 5.5456, training accuracy = 90.22%, test accuracy = 27.94%
2017-06-05 01:13:21.329280  iteration = 173, loss = 0.9067, val_loss = 5.5476, training accuracy = 90.28%, test accuracy = 27.98%
2017-06-05 01:58:40.868961  iteration = 174, loss = 0.9063, val_loss = 5.5473, training accuracy = 90.18%, test accuracy = 27.96%
2017-06-05 02:42:13.189896  iteration = 175, loss = 0.9063, val_loss = 5.5487, training accuracy = 90.22%, test accuracy = 28.00%
2017-06-05 03:25:34.470189  iteration = 176, loss = 0.9055, val_loss = 5.5513, training accuracy = 90.20%, test accuracy = 28.02%
2017-06-05 04:09:30.562579  iteration = 177, loss = 0.9053, val_loss = 5.5501, training accuracy = 90.16%, test accuracy = 28.02%
2017-06-05 04:54:38.103583  iteration = 178, loss = 0.9061, val_loss = 5.5489, training accuracy = 90.20%, test accuracy = 28.02%
2017-06-05 05:46:16.791139  iteration = 179, loss = 0.9053, val_loss = 5.5510, training accuracy = 90.22%, test accuracy = 28.06%
2017-06-05 06:35:57.175612  iteration = 180, loss = 0.9048, val_loss = 5.5509, training accuracy = 90.22%, test accuracy = 28.08%
2017-06-05 07:21:42.954951  iteration = 181, loss = 0.9056, val_loss = 5.5516, training accuracy = 90.14%, test accuracy = 28.06%
2017-06-05 08:04:37.808923  iteration = 182, loss = 0.9038, val_loss = 5.5502, training accuracy = 90.22%, test accuracy = 28.12%
2017-06-05 08:47:57.786461  iteration = 183, loss = 0.9042, val_loss = 5.5503, training accuracy = 90.26%, test accuracy = 28.08%
2017-06-05 09:31:51.470572  iteration = 184, loss = 0.9037, val_loss = 5.5548, training accuracy = 90.22%, test accuracy = 28.06%
2017-06-05 10:18:23.033644  iteration = 185, loss = 0.9035, val_loss = 5.5547, training accuracy = 90.32%, test accuracy = 27.98%
2017-06-05 11:10:32.680811  iteration = 186, loss = 0.9035, val_loss = 5.5528, training accuracy = 90.26%, test accuracy = 28.10%
2017-06-05 11:57:48.569295  iteration = 187, loss = 0.9035, val_loss = 5.5536, training accuracy = 90.26%, test accuracy = 28.00%
2017-06-05 12:41:11.054988  iteration = 188, loss = 0.9041, val_loss = 5.5528, training accuracy = 90.26%, test accuracy = 28.00%
2017-06-05 13:24:39.633599  iteration = 189, loss = 0.9024, val_loss = 5.5555, training accuracy = 90.28%, test accuracy = 28.04%
2017-06-05 14:12:11.604238  iteration = 190, loss = 0.9040, val_loss = 5.5545, training accuracy = 90.24%, test accuracy = 28.04%
2017-06-05 14:55:36.950723  iteration = 191, loss = 0.9029, val_loss = 5.5551, training accuracy = 90.30%, test accuracy = 28.00%
2017-06-05 15:39:27.920921  iteration = 192, loss = 0.9028, val_loss = 5.5558, training accuracy = 90.32%, test accuracy = 27.98%
2017-06-05 16:23:23.281354  iteration = 193, loss = 0.9037, val_loss = 5.5555, training accuracy = 90.26%, test accuracy = 28.00%
2017-06-05 17:13:42.429325  iteration = 194, loss = 0.9020, val_loss = 5.5553, training accuracy = 90.26%, test accuracy = 28.04%
2017-06-05 17:57:18.150294  iteration = 195, loss = 0.9021, val_loss = 5.5566, training accuracy = 90.26%, test accuracy = 28.02%
2017-06-05 18:41:07.346840  iteration = 196, loss = 0.9021, val_loss = 5.5568, training accuracy = 90.28%, test accuracy = 28.08%
2017-06-05 19:31:09.546366  iteration = 197, loss = 0.9021, val_loss = 5.5566, training accuracy = 90.28%, test accuracy = 28.04%
2017-06-05 20:23:46.287270  iteration = 198, loss = 0.9034, val_loss = 5.5556, training accuracy = 90.28%, test accuracy = 28.04%
2017-06-05 21:11:44.518423  iteration = 199, loss = 0.9025, val_loss = 5.5568, training accuracy = 90.30%, test accuracy = 28.00%

Continuing training with a further run
sjb@persephone:sjb_word_model$ ./train.py theresa_may_contributions.json
2017-06-05 22:03:08.330868  iteration = 000, loss = 0.9016, val_loss = 5.5561, training accuracy = 90.34%, test accuracy = 28.06%
2017-06-05 22:49:39.109063  iteration = 001, loss = 0.9015, val_loss = 5.5588, training accuracy = 90.34%, test accuracy = 28.08%
2017-06-05 23:34:59.339384  iteration = 002, loss = 0.8997, val_loss = 5.5585, training accuracy = 90.36%, test accuracy = 28.06%
2017-06-06 00:20:33.830759  iteration = 003, loss = 0.9008, val_loss = 5.5576, training accuracy = 90.34%, test accuracy = 28.04%
2017-06-06 01:03:58.432306  iteration = 004, loss = 0.9001, val_loss = 5.5599, training accuracy = 90.34%, test accuracy = 28.14%
2017-06-06 01:47:36.074987  iteration = 005, loss = 0.9004, val_loss = 5.5596, training accuracy = 90.36%, test accuracy = 28.10%
2017-06-06 02:31:52.671109  iteration = 006, loss = 0.8996, val_loss = 5.5598, training accuracy = 90.30%, test accuracy = 28.06%
2017-06-06 03:23:02.894531  iteration = 007, loss = 0.9006, val_loss = 5.5605, training accuracy = 90.38%, test accuracy = 28.08%
2017-06-06 04:10:58.563000  iteration = 008, loss = 0.9001, val_loss = 5.5608, training accuracy = 90.34%, test accuracy = 28.08%
2017-06-06 04:56:33.706728  iteration = 009, loss = 0.9005, val_loss = 5.5597, training accuracy = 90.30%, test accuracy = 28.04%
2017-06-06 05:40:19.912549  iteration = 010, loss = 0.9008, val_loss = 5.5614, training accuracy = 90.38%, test accuracy = 28.12%
2017-06-06 06:24:08.744570  iteration = 011, loss = 0.8986, val_loss = 5.5606, training accuracy = 90.32%, test accuracy = 28.08%
2017-06-06 07:06:57.196625  iteration = 012, loss = 0.9003, val_loss = 5.5613, training accuracy = 90.34%, test accuracy = 27.94%
2017-06-06 07:59:45.450719  iteration = 013, loss = 0.8978, val_loss = 5.5630, training accuracy = 90.30%, test accuracy = 28.10%
2017-06-06 08:49:48.676368  iteration = 014, loss = 0.8992, val_loss = 5.5627, training accuracy = 90.36%, test accuracy = 28.06%
2017-06-06 09:35:16.446970  iteration = 015, loss = 0.8970, val_loss = 5.5629, training accuracy = 90.34%, test accuracy = 27.96%
2017-06-06 10:19:02.276902  iteration = 016, loss = 0.8977, val_loss = 5.5644, training accuracy = 90.38%, test accuracy = 28.06%
2017-06-06 11:02:47.817234  iteration = 017, loss = 0.8985, val_loss = 5.5647, training accuracy = 90.34%, test accuracy = 28.08%
2017-06-06 11:52:36.774312  iteration = 018, loss = 0.8981, val_loss = 5.5629, training accuracy = 90.22%, test accuracy = 28.12%
2017-06-06 12:44:35.925885  iteration = 019, loss = 0.8987, val_loss = 5.5631, training accuracy = 90.40%, test accuracy = 28.12%
2017-06-06 13:32:46.112916  iteration = 020, loss = 0.8971, val_loss = 5.5633, training accuracy = 90.36%, test accuracy = 28.10%
2017-06-06 14:16:51.617382  iteration = 021, loss = 0.8955, val_loss = 5.5650, training accuracy = 90.36%, test accuracy = 28.10%
2017-06-06 15:00:37.351097  iteration = 022, loss = 0.8970, val_loss = 5.5654, training accuracy = 90.28%, test accuracy = 28.14%
2017-06-06 15:44:23.864222  iteration = 023, loss = 0.8971, val_loss = 5.5643, training accuracy = 90.32%, test accuracy = 28.08%
2017-06-06 16:27:14.240746  iteration = 024, loss = 0.8977, val_loss = 5.5669, training accuracy = 90.38%, test accuracy = 28.04%
2017-06-06 17:10:55.664870  iteration = 025, loss = 0.8958, val_loss = 5.5662, training accuracy = 90.32%, test accuracy = 28.08%
2017-06-06 18:02:47.626178  iteration = 026, loss = 0.8956, val_loss = 5.5679, training accuracy = 90.30%, test accuracy = 28.02%
2017-06-06 18:50:37.760035  iteration = 027, loss = 0.8965, val_loss = 5.5666, training accuracy = 90.24%, test accuracy = 28.00%
2017-06-06 19:34:45.961847  iteration = 028, loss = 0.8974, val_loss = 5.5655, training accuracy = 90.24%, test accuracy = 28.02%
2017-06-06 20:17:43.676550  iteration = 029, loss = 0.8957, val_loss = 5.5684, training accuracy = 90.32%, test accuracy = 28.04%
2017-06-06 21:01:23.062106  iteration = 030, loss = 0.8974, val_loss = 5.5699, training accuracy = 90.34%, test accuracy = 28.02%
2017-06-06 21:53:19.431770  iteration = 031, loss = 0.8942, val_loss = 5.5703, training accuracy = 90.34%, test accuracy = 28.02%
2017-06-06 22:42:21.937113  iteration = 032, loss = 0.8956, val_loss = 5.5693, training accuracy = 90.30%, test accuracy = 28.10%
2017-06-06 23:27:24.918751  iteration = 033, loss = 0.8940, val_loss = 5.5691, training accuracy = 90.34%, test accuracy = 28.06%
2017-06-07 00:11:14.513222  iteration = 034, loss = 0.8940, val_loss = 5.5688, training accuracy = 90.36%, test accuracy = 28.02%
2017-06-07 01:02:42.060994  iteration = 035, loss = 0.8934, val_loss = 5.5716, training accuracy = 90.34%, test accuracy = 28.00%
2017-06-07 01:48:42.341594  iteration = 036, loss = 0.8951, val_loss = 5.5711, training accuracy = 90.32%, test accuracy = 28.04%
2017-06-07 02:34:07.687315  iteration = 037, loss = 0.8923, val_loss = 5.5707, training accuracy = 90.30%, test accuracy = 28.00%
2017-06-07 03:17:34.088908  iteration = 038, loss = 0.8944, val_loss = 5.5710, training accuracy = 90.36%, test accuracy = 28.02%
2017-06-07 04:01:12.780794  iteration = 039, loss = 0.8946, val_loss = 5.5696, training accuracy = 90.28%, test accuracy = 28.04%
2017-06-07 04:44:58.001671  iteration = 040, loss = 0.8945, val_loss = 5.5694, training accuracy = 90.34%, test accuracy = 28.00%
2017-06-07 05:35:57.121819  iteration = 041, loss = 0.8933, val_loss = 5.5706, training accuracy = 90.40%, test accuracy = 28.04%
2017-06-07 06:23:14.253092  iteration = 042, loss = 0.8936, val_loss = 5.5729, training accuracy = 90.32%, test accuracy = 27.98%
2017-06-07 07:10:41.547409  iteration = 043, loss = 0.8936, val_loss = 5.5745, training accuracy = 90.42%, test accuracy = 28.08%
2017-06-07 07:56:28.443684  iteration = 044, loss = 0.8929, val_loss = 5.5714, training accuracy = 90.34%, test accuracy = 28.02%
2017-06-07 08:40:11.275734  iteration = 045, loss = 0.8943, val_loss = 5.5744, training accuracy = 90.40%, test accuracy = 28.04%
2017-06-07 09:23:55.744153  iteration = 046, loss = 0.8913, val_loss = 5.5743, training accuracy = 90.34%, test accuracy = 28.06%
2017-06-07 10:14:50.087914  iteration = 047, loss = 0.8916, val_loss = 5.5735, training accuracy = 90.40%, test accuracy = 28.04%
2017-06-07 11:04:51.177788  iteration = 048, loss = 0.8932, val_loss = 5.5738, training accuracy = 90.38%, test accuracy = 28.04%
2017-06-07 11:47:33.117192  iteration = 049, loss = 0.8927, val_loss = 5.5746, training accuracy = 90.36%, test accuracy = 28.02%
2017-06-07 12:30:13.928984  iteration = 050, loss = 0.8920, val_loss = 5.5755, training accuracy = 90.36%, test accuracy = 28.04%
2017-06-07 13:12:57.610433  iteration = 051, loss = 0.8913, val_loss = 5.5759, training accuracy = 90.36%, test accuracy = 28.00%
2017-06-07 13:55:57.616529  iteration = 052, loss = 0.8930, val_loss = 5.5752, training accuracy = 90.46%, test accuracy = 28.00%
2017-06-07 14:39:03.376650  iteration = 053, loss = 0.8911, val_loss = 5.5777, training accuracy = 90.42%, test accuracy = 28.02%
2017-06-07 15:21:50.158717  iteration = 054, loss = 0.8911, val_loss = 5.5773, training accuracy = 90.38%, test accuracy = 28.10%
2017-06-07 16:05:18.165405  iteration = 055, loss = 0.8910, val_loss = 5.5786, training accuracy = 90.36%, test accuracy = 28.00%
2017-06-07 16:48:02.403472  iteration = 056, loss = 0.8900, val_loss = 5.5769, training accuracy = 90.40%, test accuracy = 28.02%
2017-06-07 17:31:48.994170  iteration = 057, loss = 0.8910, val_loss = 5.5784, training accuracy = 90.42%, test accuracy = 28.06%
2017-06-07 18:23:51.781201  iteration = 058, loss = 0.8910, val_loss = 5.5772, training accuracy = 90.42%, test accuracy = 28.16%
2017-06-07 19:13:25.007539  iteration = 059, loss = 0.8911, val_loss = 5.5783, training accuracy = 90.38%, test accuracy = 28.14%
2017-06-07 19:58:35.794213  iteration = 060, loss = 0.8902, val_loss = 5.5791, training accuracy = 90.42%, test accuracy = 28.16%
2017-06-07 20:42:22.128640  iteration = 061, loss = 0.8895, val_loss = 5.5783, training accuracy = 90.38%, test accuracy = 28.04%
2017-06-07 21:35:46.789415  iteration = 062, loss = 0.8892, val_loss = 5.5798, training accuracy = 90.44%, test accuracy = 28.06%
2017-06-07 22:21:26.333752  iteration = 063, loss = 0.8886, val_loss = 5.5791, training accuracy = 90.46%, test accuracy = 28.02%
2017-06-07 23:04:22.106607  iteration = 064, loss = 0.8899, val_loss = 5.5809, training accuracy = 90.44%, test accuracy = 28.00%
2017-06-07 23:47:20.266830  iteration = 065, loss = 0.8883, val_loss = 5.5798, training accuracy = 90.42%, test accuracy = 28.04%
2017-06-08 00:31:03.569372  iteration = 066, loss = 0.8885, val_loss = 5.5810, training accuracy = 90.50%, test accuracy = 28.04%
2017-06-08 01:22:58.459620  iteration = 067, loss = 0.8877, val_loss = 5.5811, training accuracy = 90.46%, test accuracy = 28.10%
2017-06-08 02:13:19.799502  iteration = 068, loss = 0.8899, val_loss = 5.5801, training accuracy = 90.50%, test accuracy = 28.06%
2017-06-08 02:59:06.797658  iteration = 069, loss = 0.8894, val_loss = 5.5825, training accuracy = 90.50%, test accuracy = 28.06%
2017-06-08 03:44:38.940281  iteration = 070, loss = 0.8892, val_loss = 5.5806, training accuracy = 90.42%, test accuracy = 28.16%
2017-06-08 04:36:42.436625  iteration = 071, loss = 0.8875, val_loss = 5.5824, training accuracy = 90.46%, test accuracy = 28.06%
2017-06-08 05:26:53.260317  iteration = 072, loss = 0.8870, val_loss = 5.5815, training accuracy = 90.46%, test accuracy = 28.02%
2017-06-08 06:12:50.771006  iteration = 073, loss = 0.8877, val_loss = 5.5842, training accuracy = 90.50%, test accuracy = 28.02%
2017-06-08 07:06:02.335741  iteration = 074, loss = 0.8875, val_loss = 5.5836, training accuracy = 90.48%, test accuracy = 28.00%
2017-06-08 07:59:27.250196  iteration = 075, loss = 0.8858, val_loss = 5.5849, training accuracy = 90.52%, test accuracy = 28.08%
2017-06-08 08:44:58.111852  iteration = 076, loss = 0.8885, val_loss = 5.5838, training accuracy = 90.46%, test accuracy = 28.10%
2017-06-08 09:36:53.672024  iteration = 077, loss = 0.8877, val_loss = 5.5841, training accuracy = 90.52%, test accuracy = 28.14%
2017-06-08 10:33:16.379784  iteration = 078, loss = 0.8865, val_loss = 5.5850, training accuracy = 90.54%, test accuracy = 28.14%
2017-06-08 11:22:43.322385  iteration = 079, loss = 0.8887, val_loss = 5.5841, training accuracy = 90.50%, test accuracy = 28.00%
2017-06-08 12:17:09.686373  iteration = 080, loss = 0.8886, val_loss = 5.5855, training accuracy = 90.50%, test accuracy = 28.04%
2017-06-08 13:07:38.107498  iteration = 081, loss = 0.8875, val_loss = 5.5845, training accuracy = 90.50%, test accuracy = 28.06%
2017-06-08 14:01:32.561616  iteration = 082, loss = 0.8875, val_loss = 5.5849, training accuracy = 90.52%, test accuracy = 28.02%
2017-06-08 14:58:49.954997  iteration = 083, loss = 0.8850, val_loss = 5.5852, training accuracy = 90.56%, test accuracy = 28.00%
2017-06-08 15:44:37.333844  iteration = 084, loss = 0.8872, val_loss = 5.5845, training accuracy = 90.48%, test accuracy = 28.06%
2017-06-08 16:38:42.806104  iteration = 085, loss = 0.8860, val_loss = 5.5868, training accuracy = 90.54%, test accuracy = 28.08%
2017-06-08 17:29:36.163448  iteration = 086, loss = 0.8870, val_loss = 5.5864, training accuracy = 90.52%, test accuracy = 28.08%
2017-06-08 18:24:29.409756  iteration = 087, loss = 0.8860, val_loss = 5.5891, training accuracy = 90.52%, test accuracy = 28.02%
2017-06-08 19:20:03.217896  iteration = 088, loss = 0.8853, val_loss = 5.5875, training accuracy = 90.52%, test accuracy = 27.96%
2017-06-08 20:07:30.259664  iteration = 089, loss = 0.8855, val_loss = 5.5881, training accuracy = 90.46%, test accuracy = 27.96%
2017-06-08 20:54:36.028015  iteration = 090, loss = 0.8848, val_loss = 5.5878, training accuracy = 90.56%, test accuracy = 27.98%
2017-06-08 21:47:07.113599  iteration = 091, loss = 0.8837, val_loss = 5.5905, training accuracy = 90.54%, test accuracy = 28.12%
2017-06-08 22:34:49.768285  iteration = 092, loss = 0.8839, val_loss = 5.5896, training accuracy = 90.54%, test accuracy = 28.02%
2017-06-08 23:20:04.211315  iteration = 093, loss = 0.8830, val_loss = 5.5908, training accuracy = 90.56%, test accuracy = 28.04%
2017-06-09 00:03:33.074078  iteration = 094, loss = 0.8848, val_loss = 5.5902, training accuracy = 90.48%, test accuracy = 28.12%
2017-06-09 00:47:12.222674  iteration = 095, loss = 0.8848, val_loss = 5.5876, training accuracy = 90.54%, test accuracy = 28.08%
2017-06-09 01:38:23.290917  iteration = 096, loss = 0.8855, val_loss = 5.5907, training accuracy = 90.50%, test accuracy = 28.04%
2017-06-09 02:30:18.189388  iteration = 097, loss = 0.8828, val_loss = 5.5921, training accuracy = 90.54%, test accuracy = 28.06%
2017-06-09 03:18:05.663509  iteration = 098, loss = 0.8824, val_loss = 5.5916, training accuracy = 90.66%, test accuracy = 28.14%
2017-06-09 04:01:48.807913  iteration = 099, loss = 0.8851, val_loss = 5.5907, training accuracy = 90.60%, test accuracy = 28.04%
2017-06-09 04:45:40.276774  iteration = 100, loss = 0.8830, val_loss = 5.5920, training accuracy = 90.54%, test accuracy = 28.02%
2017-06-09 05:37:55.771302  iteration = 101, loss = 0.8808, val_loss = 5.5924, training accuracy = 90.60%, test accuracy = 27.96%
2017-06-09 06:27:40.019748  iteration = 102, loss = 0.8839, val_loss = 5.5924, training accuracy = 90.62%, test accuracy = 28.04%
2017-06-09 07:14:58.349291  iteration = 103, loss = 0.8830, val_loss = 5.5939, training accuracy = 90.64%, test accuracy = 27.96%
2017-06-09 07:59:32.870899  iteration = 104, loss = 0.8826, val_loss = 5.5922, training accuracy = 90.62%, test accuracy = 28.00%
2017-06-09 08:43:12.980730  iteration = 105, loss = 0.8838, val_loss = 5.5928, training accuracy = 90.60%, test accuracy = 28.00%
2017-06-09 09:34:52.578374  iteration = 106, loss = 0.8817, val_loss = 5.5941, training accuracy = 90.56%, test accuracy = 28.02%
2017-06-09 10:24:47.442627  iteration = 107, loss = 0.8816, val_loss = 5.5960, training accuracy = 90.62%, test accuracy = 28.06%
2017-06-09 11:10:29.754809  iteration = 108, loss = 0.8817, val_loss = 5.5951, training accuracy = 90.62%, test accuracy = 28.08%
2017-06-09 11:55:12.333391  iteration = 109, loss = 0.8823, val_loss = 5.5948, training accuracy = 90.60%, test accuracy = 28.10%
2017-06-09 12:40:43.961608  iteration = 110, loss = 0.8813, val_loss = 5.5964, training accuracy = 90.64%, test accuracy = 27.98%
2017-06-09 13:24:51.046692  iteration = 111, loss = 0.8807, val_loss = 5.5961, training accuracy = 90.68%, test accuracy = 28.00%
2017-06-09 14:13:48.198427  iteration = 112, loss = 0.8814, val_loss = 5.5952, training accuracy = 90.68%, test accuracy = 27.98%
2017-06-09 15:03:07.829649  iteration = 113, loss = 0.8813, val_loss = 5.5958, training accuracy = 90.72%, test accuracy = 28.04%
2017-06-09 15:50:34.586929  iteration = 114, loss = 0.8815, val_loss = 5.5956, training accuracy = 90.60%, test accuracy = 28.02%
2017-06-09 16:35:58.095968  iteration = 115, loss = 0.8801, val_loss = 5.5985, training accuracy = 90.68%, test accuracy = 28.00%
2017-06-09 17:19:34.006822  iteration = 116, loss = 0.8796, val_loss = 5.5958, training accuracy = 90.66%, test accuracy = 28.04%
2017-06-09 18:02:26.183282  iteration = 117, loss = 0.8787, val_loss = 5.5973, training accuracy = 90.68%, test accuracy = 27.94%
2017-06-09 18:46:15.040147  iteration = 118, loss = 0.8799, val_loss = 5.6003, training accuracy = 90.60%, test accuracy = 27.98%
2017-06-09 19:29:13.700967  iteration = 119, loss = 0.8784, val_loss = 5.5987, training accuracy = 90.68%, test accuracy = 27.98%
2017-06-09 20:12:58.787516  iteration = 120, loss = 0.8788, val_loss = 5.5985, training accuracy = 90.66%, test accuracy = 27.96%
2017-06-09 21:05:11.836730  iteration = 121, loss = 0.8781, val_loss = 5.5984, training accuracy = 90.68%, test accuracy = 28.00%
2017-06-09 21:54:35.697811  iteration = 122, loss = 0.8778, val_loss = 5.6000, training accuracy = 90.66%, test accuracy = 27.98%
2017-06-09 22:40:43.153293  iteration = 123, loss = 0.8787, val_loss = 5.6002, training accuracy = 90.66%, test accuracy = 27.90%
2017-06-09 23:25:02.762924  iteration = 124, loss = 0.8801, val_loss = 5.6005, training accuracy = 90.68%, test accuracy = 27.94%
2017-06-10 00:08:53.317434  iteration = 125, loss = 0.8783, val_loss = 5.6014, training accuracy = 90.66%, test accuracy = 28.00%
2017-06-10 01:00:36.775701  iteration = 126, loss = 0.8783, val_loss = 5.6011, training accuracy = 90.72%, test accuracy = 28.02%
2017-06-10 01:49:38.795373  iteration = 127, loss = 0.8770, val_loss = 5.6018, training accuracy = 90.72%, test accuracy = 28.00%
2017-06-10 02:34:00.915642  iteration = 128, loss = 0.8788, val_loss = 5.6031, training accuracy = 90.66%, test accuracy = 27.98%
2017-06-10 03:19:04.386778  iteration = 129, loss = 0.8788, val_loss = 5.5995, training accuracy = 90.60%, test accuracy = 28.00%
2017-06-10 04:06:20.063148  iteration = 130, loss = 0.8784, val_loss = 5.6012, training accuracy = 90.68%, test accuracy = 27.96%
2017-06-10 04:49:31.705868  iteration = 131, loss = 0.8768, val_loss = 5.6019, training accuracy = 90.72%, test accuracy = 28.06%
2017-06-10 05:33:13.640893  iteration = 132, loss = 0.8767, val_loss = 5.6030, training accuracy = 90.70%, test accuracy = 27.94%
2017-06-10 06:25:14.889334  iteration = 133, loss = 0.8757, val_loss = 5.6023, training accuracy = 90.68%, test accuracy = 28.08%
2017-06-10 07:15:13.836216  iteration = 134, loss = 0.8767, val_loss = 5.6037, training accuracy = 90.74%, test accuracy = 28.04%
2017-06-10 08:01:35.469634  iteration = 135, loss = 0.8761, val_loss = 5.6048, training accuracy = 90.70%, test accuracy = 28.04%
2017-06-10 08:46:07.849641  iteration = 136, loss = 0.8764, val_loss = 5.6021, training accuracy = 90.72%, test accuracy = 28.04%
2017-06-10 09:35:05.896150  iteration = 137, loss = 0.8769, val_loss = 5.6043, training accuracy = 90.64%, test accuracy = 28.04%
2017-06-10 10:27:34.488822  iteration = 138, loss = 0.8766, val_loss = 5.6042, training accuracy = 90.74%, test accuracy = 28.00%
2017-06-10 11:14:06.769883  iteration = 139, loss = 0.8761, val_loss = 5.6056, training accuracy = 90.76%, test accuracy = 28.04%
2017-06-10 11:57:42.889327  iteration = 140, loss = 0.8781, val_loss = 5.6032, training accuracy = 90.70%, test accuracy = 28.00%
2017-06-10 12:41:43.288488  iteration = 141, loss = 0.8762, val_loss = 5.6071, training accuracy = 90.78%, test accuracy = 28.00%
2017-06-10 13:33:16.011004  iteration = 142, loss = 0.8757, val_loss = 5.6053, training accuracy = 90.74%, test accuracy = 28.06%
2017-06-10 14:19:09.586018  iteration = 143, loss = 0.8771, val_loss = 5.6064, training accuracy = 90.68%, test accuracy = 28.08%
2017-06-10 15:04:57.807295  iteration = 144, loss = 0.8747, val_loss = 5.6074, training accuracy = 90.76%, test accuracy = 28.08%
2017-06-10 15:52:54.866315  iteration = 145, loss = 0.8763, val_loss = 5.6065, training accuracy = 90.80%, test accuracy = 28.12%
2017-06-10 16:45:06.570026  iteration = 146, loss = 0.8750, val_loss = 5.6072, training accuracy = 90.76%, test accuracy = 27.96%
2017-06-10 17:32:04.903712  iteration = 147, loss = 0.8751, val_loss = 5.6077, training accuracy = 90.78%, test accuracy = 28.02%
2017-06-10 18:17:28.737830  iteration = 148, loss = 0.8744, val_loss = 5.6100, training accuracy = 90.84%, test accuracy = 28.02%
2017-06-10 19:02:13.967134  iteration = 149, loss = 0.8749, val_loss = 5.6081, training accuracy = 90.78%, test accuracy = 28.06%
2017-06-10 19:54:10.998134  iteration = 150, loss = 0.8743, val_loss = 5.6089, training accuracy = 90.76%, test accuracy = 28.06%
2017-06-10 20:42:02.731251  iteration = 151, loss = 0.8728, val_loss = 5.6095, training accuracy = 90.74%, test accuracy = 28.06%
2017-06-10 21:26:13.513539  iteration = 152, loss = 0.8726, val_loss = 5.6121, training accuracy = 90.64%, test accuracy = 28.04%
2017-06-10 22:10:01.033544  iteration = 153, loss = 0.8726, val_loss = 5.6097, training accuracy = 90.78%, test accuracy = 28.02%
2017-06-10 23:01:29.768808  iteration = 154, loss = 0.8737, val_loss = 5.6092, training accuracy = 90.74%, test accuracy = 28.04%
2017-06-10 23:45:02.875498  iteration = 155, loss = 0.8741, val_loss = 5.6108, training accuracy = 90.76%, test accuracy = 28.16%
2017-06-11 00:28:52.941878  iteration = 156, loss = 0.8745, val_loss = 5.6111, training accuracy = 90.80%, test accuracy = 28.04%
2017-06-11 01:19:29.889022  iteration = 157, loss = 0.8725, val_loss = 5.6096, training accuracy = 90.78%, test accuracy = 28.02%
2017-06-11 02:11:20.165729  iteration = 158, loss = 0.8734, val_loss = 5.6119, training accuracy = 90.72%, test accuracy = 28.10%
2017-06-11 02:57:19.713417  iteration = 159, loss = 0.8731, val_loss = 5.6125, training accuracy = 90.74%, test accuracy = 28.14%
2017-06-11 03:41:06.356977  iteration = 160, loss = 0.8719, val_loss = 5.6113, training accuracy = 90.68%, test accuracy = 28.14%
2017-06-11 04:24:56.987405  iteration = 161, loss = 0.8725, val_loss = 5.6123, training accuracy = 90.80%, test accuracy = 28.10%
2017-06-11 05:16:53.855583  iteration = 162, loss = 0.8720, val_loss = 5.6134, training accuracy = 90.78%, test accuracy = 28.04%
2017-06-11 06:03:23.816748  iteration = 163, loss = 0.8722, val_loss = 5.6132, training accuracy = 90.80%, test accuracy = 28.04%
2017-06-11 06:47:09.759357  iteration = 164, loss = 0.8734, val_loss = 5.6126, training accuracy = 90.76%, test accuracy = 28.00%
2017-06-11 07:30:50.546548  iteration = 165, loss = 0.8721, val_loss = 5.6139, training accuracy = 90.80%, test accuracy = 28.08%
2017-06-11 08:23:39.228967  iteration = 166, loss = 0.8713, val_loss = 5.6134, training accuracy = 90.80%, test accuracy = 28.12%
2017-06-11 09:12:22.267460  iteration = 167, loss = 0.8710, val_loss = 5.6161, training accuracy = 90.82%, test accuracy = 28.14%
2017-06-11 09:58:09.584472  iteration = 168, loss = 0.8713, val_loss = 5.6151, training accuracy = 90.86%, test accuracy = 28.10%
2017-06-11 10:41:49.043128  iteration = 169, loss = 0.8694, val_loss = 5.6153, training accuracy = 90.88%, test accuracy = 28.08%
2017-06-11 11:25:30.215817  iteration = 170, loss = 0.8735, val_loss = 5.6153, training accuracy = 90.84%, test accuracy = 28.06%
2017-06-11 12:16:55.144459  iteration = 171, loss = 0.8715, val_loss = 5.6145, training accuracy = 90.86%, test accuracy = 28.06%
2017-06-11 13:05:29.028026  iteration = 172, loss = 0.8706, val_loss = 5.6157, training accuracy = 90.88%, test accuracy = 28.02%
2017-06-11 13:51:35.967779  iteration = 173, loss = 0.8703, val_loss = 5.6166, training accuracy = 90.84%, test accuracy = 28.12%
2017-06-11 14:35:17.048981  iteration = 174, loss = 0.8693, val_loss = 5.6174, training accuracy = 90.86%, test accuracy = 28.14%
2017-06-11 15:18:57.040419  iteration = 175, loss = 0.8699, val_loss = 5.6171, training accuracy = 90.88%, test accuracy = 28.08%
2017-06-11 16:10:58.615499  iteration = 176, loss = 0.8701, val_loss = 5.6166, training accuracy = 90.84%, test accuracy = 28.14%
2017-06-11 17:01:59.356171  iteration = 177, loss = 0.8688, val_loss = 5.6171, training accuracy = 90.84%, test accuracy = 28.06%
2017-06-11 17:48:01.457354  iteration = 178, loss = 0.8687, val_loss = 5.6180, training accuracy = 90.84%, test accuracy = 28.08%
2017-06-11 18:31:43.046831  iteration = 179, loss = 0.8704, val_loss = 5.6185, training accuracy = 90.92%, test accuracy = 28.14%
2017-06-11 19:16:09.808749  iteration = 180, loss = 0.8689, val_loss = 5.6207, training accuracy = 90.86%, test accuracy = 28.20%
2017-06-11 20:07:36.463071  iteration = 181, loss = 0.8685, val_loss = 5.6190, training accuracy = 90.88%, test accuracy = 28.16%
2017-06-11 20:53:53.823396  iteration = 182, loss = 0.8685, val_loss = 5.6190, training accuracy = 90.88%, test accuracy = 28.08%
2017-06-11 21:39:04.557383  iteration = 183, loss = 0.8678, val_loss = 5.6205, training accuracy = 90.88%, test accuracy = 28.12%
2017-06-11 22:22:53.220928  iteration = 184, loss = 0.8688, val_loss = 5.6215, training accuracy = 90.86%, test accuracy = 28.12%
2017-06-11 23:14:46.485594  iteration = 185, loss = 0.8680, val_loss = 5.6190, training accuracy = 90.94%, test accuracy = 28.10%
2017-06-12 00:04:25.365771  iteration = 186, loss = 0.8675, val_loss = 5.6206, training accuracy = 90.92%, test accuracy = 28.12%
2017-06-12 00:49:45.512744  iteration = 187, loss = 0.8667, val_loss = 5.6231, training accuracy = 90.88%, test accuracy = 28.08%
2017-06-12 01:33:24.452433  iteration = 188, loss = 0.8698, val_loss = 5.6201, training accuracy = 90.84%, test accuracy = 28.12%
2017-06-12 02:16:16.176456  iteration = 189, loss = 0.8675, val_loss = 5.6225, training accuracy = 90.94%, test accuracy = 28.14%
2017-06-12 03:00:04.110922  iteration = 190, loss = 0.8673, val_loss = 5.6203, training accuracy = 90.92%, test accuracy = 28.10%
2017-06-12 03:52:34.435641  iteration = 191, loss = 0.8674, val_loss = 5.6230, training accuracy = 90.92%, test accuracy = 28.02%
2017-06-12 04:40:46.748757  iteration = 192, loss = 0.8683, val_loss = 5.6211, training accuracy = 90.98%, test accuracy = 28.12%
2017-06-12 05:24:33.005163  iteration = 193, loss = 0.8666, val_loss = 5.6221, training accuracy = 90.92%, test accuracy = 28.08%
2017-06-12 06:08:14.785237  iteration = 194, loss = 0.8664, val_loss = 5.6215, training accuracy = 90.94%, test accuracy = 28.08%
2017-06-12 07:00:20.481029  iteration = 195, loss = 0.8671, val_loss = 5.6237, training accuracy = 91.00%, test accuracy = 28.04%
2017-06-12 07:49:16.756742  iteration = 196, loss = 0.8667, val_loss = 5.6214, training accuracy = 91.00%, test accuracy = 27.96%
2017-06-12 08:33:10.604688  iteration = 197, loss = 0.8664, val_loss = 5.6229, training accuracy = 91.04%, test accuracy = 28.10%
2017-06-12 09:16:55.245860  iteration = 198, loss = 0.8674, val_loss = 5.6234, training accuracy = 91.10%, test accuracy = 28.04%
2017-06-12 10:08:34.179152  iteration = 199, loss = 0.8654, val_loss = 5.6248, training accuracy = 90.98%, test accuracy = 28.04%

Continuing training with a further run
sjb@persephone:sjb_word_model$ ./train.py theresa_may_contributions.json
2017-06-12 15:39:05.149531  iteration = 000, loss = 0.8655, val_loss = 5.6257, training accuracy = 91.06%, test accuracy = 28.00%
2017-06-12 16:28:29.891174  iteration = 001, loss = 0.8659, val_loss = 5.6241, training accuracy = 91.10%, test accuracy = 28.12%
2017-06-12 17:14:05.592575  iteration = 002, loss = 0.8665, val_loss = 5.6242, training accuracy = 91.02%, test accuracy = 28.04%
2017-06-12 18:07:44.371458  iteration = 003, loss = 0.8670, val_loss = 5.6254, training accuracy = 91.02%, test accuracy = 28.04%
2017-06-12 18:57:37.430848  iteration = 004, loss = 0.8658, val_loss = 5.6262, training accuracy = 90.98%, test accuracy = 28.02%
2017-06-12 19:44:35.040540  iteration = 005, loss = 0.8644, val_loss = 5.6304, training accuracy = 91.00%, test accuracy = 27.96%
2017-06-12 20:38:52.611405  iteration = 006, loss = 0.8641, val_loss = 5.6267, training accuracy = 91.06%, test accuracy = 28.02%
2017-06-12 21:29:32.627889  iteration = 007, loss = 0.8663, val_loss = 5.6271, training accuracy = 91.04%, test accuracy = 27.96%
2017-06-12 22:15:30.719325  iteration = 008, loss = 0.8657, val_loss = 5.6267, training accuracy = 91.00%, test accuracy = 28.10%
2017-06-12 23:03:42.751457  iteration = 009, loss = 0.8628, val_loss = 5.6293, training accuracy = 91.00%, test accuracy = 28.02%
2017-06-12 23:56:01.821039  iteration = 010, loss = 0.8643, val_loss = 5.6299, training accuracy = 91.02%, test accuracy = 28.04%
2017-06-13 00:44:12.739612  iteration = 011, loss = 0.8638, val_loss = 5.6289, training accuracy = 91.08%, test accuracy = 27.98%
2017-06-13 01:30:00.392997  iteration = 012, loss = 0.8631, val_loss = 5.6284, training accuracy = 91.08%, test accuracy = 28.06%
2017-06-13 02:15:36.465376  iteration = 013, loss = 0.8639, val_loss = 5.6281, training accuracy = 91.08%, test accuracy = 28.06%
2017-06-13 02:59:23.584698  iteration = 014, loss = 0.8634, val_loss = 5.6296, training accuracy = 91.06%, test accuracy = 28.16%
2017-06-13 03:43:27.982881  iteration = 015, loss = 0.8632, val_loss = 5.6275, training accuracy = 91.04%, test accuracy = 28.10%
2017-06-13 04:34:28.528142  iteration = 016, loss = 0.8638, val_loss = 5.6321, training accuracy = 91.08%, test accuracy = 28.10%
2017-06-13 05:26:57.938729  iteration = 017, loss = 0.8628, val_loss = 5.6300, training accuracy = 91.08%, test accuracy = 28.04%
2017-06-13 06:13:01.432009  iteration = 018, loss = 0.8639, val_loss = 5.6282, training accuracy = 91.10%, test accuracy = 28.06%
2017-06-13 06:58:41.815620  iteration = 019, loss = 0.8623, val_loss = 5.6308, training accuracy = 91.06%, test accuracy = 28.08%
2017-06-13 07:44:21.925875  iteration = 020, loss = 0.8622, val_loss = 5.6321, training accuracy = 91.08%, test accuracy = 28.10%
2017-06-13 08:31:08.664743  iteration = 021, loss = 0.8623, val_loss = 5.6315, training accuracy = 91.08%, test accuracy = 28.06%
2017-06-13 09:22:57.829416  iteration = 022, loss = 0.8639, val_loss = 5.6321, training accuracy = 91.06%, test accuracy = 28.00%
2017-06-13 10:11:33.009866  iteration = 023, loss = 0.8638, val_loss = 5.6310, training accuracy = 91.12%, test accuracy = 28.04%
2017-06-13 10:55:55.074720  iteration = 024, loss = 0.8623, val_loss = 5.6324, training accuracy = 91.14%, test accuracy = 28.10%
2017-06-13 11:38:10.794396  iteration = 025, loss = 0.8613, val_loss = 5.6326, training accuracy = 91.08%, test accuracy = 28.02%
2017-06-13 12:28:39.592629  iteration = 026, loss = 0.8624, val_loss = 5.6308, training accuracy = 91.08%, test accuracy = 28.08%
2017-06-13 13:15:16.088450  iteration = 027, loss = 0.8629, val_loss = 5.6317, training accuracy = 91.10%, test accuracy = 28.12%
2017-06-13 13:58:46.050410  iteration = 028, loss = 0.8614, val_loss = 5.6347, training accuracy = 91.14%, test accuracy = 28.08%
2017-06-13 14:44:34.038377  iteration = 029, loss = 0.8609, val_loss = 5.6337, training accuracy = 91.18%, test accuracy = 28.14%
2017-06-13 15:35:45.146149  iteration = 030, loss = 0.8606, val_loss = 5.6350, training accuracy = 91.20%, test accuracy = 28.10%
2017-06-13 16:25:03.396204  iteration = 031, loss = 0.8611, val_loss = 5.6358, training accuracy = 91.16%, test accuracy = 28.06%
2017-06-13 17:13:13.088945  iteration = 032, loss = 0.8613, val_loss = 5.6352, training accuracy = 91.16%, test accuracy = 28.12%
2017-06-13 18:08:35.525871  iteration = 033, loss = 0.8610, val_loss = 5.6360, training accuracy = 91.18%, test accuracy = 28.12%
2017-06-13 18:54:35.088512  iteration = 034, loss = 0.8597, val_loss = 5.6361, training accuracy = 91.12%, test accuracy = 28.18%
2017-06-13 19:42:05.042334  iteration = 035, loss = 0.8603, val_loss = 5.6366, training accuracy = 91.14%, test accuracy = 28.18%
2017-06-13 20:35:28.515031  iteration = 036, loss = 0.8600, val_loss = 5.6368, training accuracy = 91.20%, test accuracy = 28.20%
2017-06-13 21:29:08.756604  iteration = 037, loss = 0.8603, val_loss = 5.6336, training accuracy = 91.20%, test accuracy = 28.16%
2017-06-13 22:21:21.255295  iteration = 038, loss = 0.8598, val_loss = 5.6376, training accuracy = 91.18%, test accuracy = 28.18%
2017-06-13 23:12:38.724301  iteration = 039, loss = 0.8589, val_loss = 5.6368, training accuracy = 91.20%, test accuracy = 28.24%
2017-06-14 00:02:35.288458  iteration = 040, loss = 0.8592, val_loss = 5.6373, training accuracy = 91.18%, test accuracy = 28.20%
2017-06-14 00:48:17.046657  iteration = 041, loss = 0.8595, val_loss = 5.6374, training accuracy = 91.10%, test accuracy = 28.22%
2017-06-14 01:33:10.893362  iteration = 042, loss = 0.8595, val_loss = 5.6364, training accuracy = 91.22%, test accuracy = 28.16%
2017-06-14 02:25:33.793362  iteration = 043, loss = 0.8585, val_loss = 5.6381, training accuracy = 91.14%, test accuracy = 28.16%
2017-06-14 03:17:43.020186  iteration = 044, loss = 0.8581, val_loss = 5.6375, training accuracy = 91.14%, test accuracy = 28.30%
2017-06-14 04:03:56.205054  iteration = 045, loss = 0.8594, val_loss = 5.6403, training accuracy = 91.16%, test accuracy = 28.20%
2017-06-14 04:49:56.001238  iteration = 046, loss = 0.8587, val_loss = 5.6385, training accuracy = 91.26%, test accuracy = 28.22%
2017-06-14 05:40:07.200496  iteration = 047, loss = 0.8587, val_loss = 5.6392, training accuracy = 91.16%, test accuracy = 28.16%
2017-06-14 06:31:51.175363  iteration = 048, loss = 0.8592, val_loss = 5.6386, training accuracy = 91.20%, test accuracy = 28.12%
2017-06-14 07:19:47.576982  iteration = 049, loss = 0.8582, val_loss = 5.6396, training accuracy = 91.22%, test accuracy = 28.18%
2017-06-14 08:05:55.554741  iteration = 050, loss = 0.8569, val_loss = 5.6399, training accuracy = 91.26%, test accuracy = 28.18%
2017-06-14 08:50:25.162404  iteration = 051, loss = 0.8575, val_loss = 5.6408, training accuracy = 91.22%, test accuracy = 28.16%
2017-06-14 09:42:57.117550  iteration = 052, loss = 0.8576, val_loss = 5.6424, training accuracy = 91.20%, test accuracy = 28.22%
2017-06-14 10:34:18.103238  iteration = 053, loss = 0.8584, val_loss = 5.6395, training accuracy = 91.26%, test accuracy = 28.16%
2017-06-14 11:17:56.754315  iteration = 054, loss = 0.8564, val_loss = 5.6425, training accuracy = 91.28%, test accuracy = 28.08%
2017-06-14 12:01:54.988159  iteration = 055, loss = 0.8571, val_loss = 5.6435, training accuracy = 91.28%, test accuracy = 28.10%
2017-06-14 12:54:02.409124  iteration = 056, loss = 0.8564, val_loss = 5.6432, training accuracy = 91.28%, test accuracy = 28.10%
2017-06-14 13:45:26.361928  iteration = 057, loss = 0.8573, val_loss = 5.6415, training accuracy = 91.30%, test accuracy = 28.04%
2017-06-14 14:31:27.487030  iteration = 058, loss = 0.8558, val_loss = 5.6439, training accuracy = 91.32%, test accuracy = 28.02%
2017-06-14 15:14:01.176945  iteration = 059, loss = 0.8555, val_loss = 5.6434, training accuracy = 91.30%, test accuracy = 28.10%
Stopped for Athene to be rebooted

Continuing training with a further run
sjb@persephone:sjb_word_model$ ./train.py theresa_may_contributions.json
2017-06-14 22:03:56.125893  iteration = 000, loss = 0.8561, val_loss = 5.6439, training accuracy = 91.30%, test accuracy = 28.06%
2017-06-14 22:47:04.144561  iteration = 001, loss = 0.8548, val_loss = 5.6441, training accuracy = 91.28%, test accuracy = 28.14%
2017-06-14 23:30:36.920769  iteration = 002, loss = 0.8564, val_loss = 5.6442, training accuracy = 91.32%, test accuracy = 28.20%
2017-06-15 00:14:22.876289  iteration = 003, loss = 0.8563, val_loss = 5.6449, training accuracy = 91.28%, test accuracy = 28.18%
2017-06-15 00:57:58.174808  iteration = 004, loss = 0.8561, val_loss = 5.6446, training accuracy = 91.32%, test accuracy = 28.10%
2017-06-15 01:41:53.875217  iteration = 005, loss = 0.8548, val_loss = 5.6454, training accuracy = 91.30%, test accuracy = 28.06%
2017-06-15 02:26:32.170286  iteration = 006, loss = 0.8546, val_loss = 5.6453, training accuracy = 91.26%, test accuracy = 28.06%
2017-06-15 03:11:33.953265  iteration = 007, loss = 0.8552, val_loss = 5.6455, training accuracy = 91.32%, test accuracy = 28.12%
2017-06-15 03:57:29.613115  iteration = 008, loss = 0.8531, val_loss = 5.6477, training accuracy = 91.28%, test accuracy = 28.20%
2017-06-15 04:43:23.835079  iteration = 009, loss = 0.8541, val_loss = 5.6462, training accuracy = 91.30%, test accuracy = 28.16%
2017-06-15 05:28:56.373480  iteration = 010, loss = 0.8548, val_loss = 5.6471, training accuracy = 91.32%, test accuracy = 28.12%
2017-06-15 06:14:09.672611  iteration = 011, loss = 0.8541, val_loss = 5.6469, training accuracy = 91.34%, test accuracy = 28.10%
2017-06-15 06:57:44.931618  iteration = 012, loss = 0.8538, val_loss = 5.6465, training accuracy = 91.32%, test accuracy = 28.18%
2017-06-15 07:42:54.257261  iteration = 013, loss = 0.8528, val_loss = 5.6500, training accuracy = 91.28%, test accuracy = 28.12%
2017-06-15 08:33:44.757126  iteration = 014, loss = 0.8548, val_loss = 5.6482, training accuracy = 91.28%, test accuracy = 28.20%
2017-06-15 09:26:50.805852  iteration = 015, loss = 0.8542, val_loss = 5.6490, training accuracy = 91.28%, test accuracy = 28.14%
2017-06-15 10:14:56.703753  iteration = 016, loss = 0.8535, val_loss = 5.6483, training accuracy = 91.32%, test accuracy = 28.22%
2017-06-15 11:01:00.512899  iteration = 017, loss = 0.8528, val_loss = 5.6483, training accuracy = 91.36%, test accuracy = 28.18%
2017-06-15 11:44:47.746725  iteration = 018, loss = 0.8537, val_loss = 5.6499, training accuracy = 91.32%, test accuracy = 28.24%
2017-06-15 12:28:31.032414  iteration = 019, loss = 0.8537, val_loss = 5.6489, training accuracy = 91.34%, test accuracy = 28.16%
2017-06-15 13:19:22.385692  iteration = 020, loss = 0.8539, val_loss = 5.6491, training accuracy = 91.36%, test accuracy = 28.24%
2017-06-15 14:02:52.917158  iteration = 021, loss = 0.8538, val_loss = 5.6501, training accuracy = 91.26%, test accuracy = 28.12%
2017-06-15 14:45:44.575614  iteration = 022, loss = 0.8530, val_loss = 5.6503, training accuracy = 91.32%, test accuracy = 28.20%
2017-06-15 15:39:13.980205  iteration = 023, loss = 0.8523, val_loss = 5.6516, training accuracy = 91.32%, test accuracy = 28.26%
2017-06-15 16:33:28.414985  iteration = 024, loss = 0.8524, val_loss = 5.6517, training accuracy = 91.34%, test accuracy = 28.14%
2017-06-15 17:20:47.031908  iteration = 025, loss = 0.8524, val_loss = 5.6501, training accuracy = 91.34%, test accuracy = 28.10%
2017-06-15 18:08:16.757561  iteration = 026, loss = 0.8515, val_loss = 5.6518, training accuracy = 91.28%, test accuracy = 28.20%
2017-06-15 18:56:55.760180  iteration = 027, loss = 0.8502, val_loss = 5.6537, training accuracy = 91.30%, test accuracy = 28.22%
2017-06-15 19:55:19.591916  iteration = 028, loss = 0.8522, val_loss = 5.6521, training accuracy = 91.32%, test accuracy = 28.22%
2017-06-15 20:42:41.884271  iteration = 029, loss = 0.8506, val_loss = 5.6529, training accuracy = 91.34%, test accuracy = 28.16%
2017-06-15 21:39:30.207136  iteration = 030, loss = 0.8517, val_loss = 5.6539, training accuracy = 91.30%, test accuracy = 28.20%
2017-06-15 22:33:21.513032  iteration = 031, loss = 0.8503, val_loss = 5.6536, training accuracy = 91.36%, test accuracy = 28.16%
2017-06-15 23:20:30.522433  iteration = 032, loss = 0.8509, val_loss = 5.6531, training accuracy = 91.38%, test accuracy = 28.12%
2017-06-16 00:05:40.886830  iteration = 033, loss = 0.8500, val_loss = 5.6542, training accuracy = 91.40%, test accuracy = 28.14%
2017-06-16 00:50:16.853926  iteration = 034, loss = 0.8505, val_loss = 5.6536, training accuracy = 91.38%, test accuracy = 28.16%
2017-06-16 01:42:24.016571  iteration = 035, loss = 0.8495, val_loss = 5.6547, training accuracy = 91.44%, test accuracy = 28.10%
2017-06-16 02:32:01.572187  iteration = 036, loss = 0.8516, val_loss = 5.6568, training accuracy = 91.46%, test accuracy = 28.12%
2017-06-16 03:17:53.135189  iteration = 037, loss = 0.8508, val_loss = 5.6543, training accuracy = 91.44%, test accuracy = 28.14%
2017-06-16 04:03:33.274658  iteration = 038, loss = 0.8483, val_loss = 5.6562, training accuracy = 91.42%, test accuracy = 28.14%
2017-06-16 04:47:55.819224  iteration = 039, loss = 0.8506, val_loss = 5.6557, training accuracy = 91.40%, test accuracy = 28.12%
2017-06-16 05:33:28.565004  iteration = 040, loss = 0.8500, val_loss = 5.6582, training accuracy = 91.44%, test accuracy = 28.14%
2017-06-16 06:25:20.581371  iteration = 041, loss = 0.8492, val_loss = 5.6568, training accuracy = 91.42%, test accuracy = 28.14%
2017-06-16 07:15:12.688981  iteration = 042, loss = 0.8491, val_loss = 5.6563, training accuracy = 91.44%, test accuracy = 28.12%
2017-06-16 08:00:46.540928  iteration = 043, loss = 0.8490, val_loss = 5.6594, training accuracy = 91.44%, test accuracy = 28.14%
2017-06-16 08:44:37.784810  iteration = 044, loss = 0.8488, val_loss = 5.6581, training accuracy = 91.42%, test accuracy = 28.14%
2017-06-16 09:34:25.628332  iteration = 045, loss = 0.8500, val_loss = 5.6566, training accuracy = 91.46%, test accuracy = 28.12%
2017-06-16 10:27:19.267048  iteration = 046, loss = 0.8473, val_loss = 5.6591, training accuracy = 91.44%, test accuracy = 28.16%
2017-06-16 11:15:01.161275  iteration = 047, loss = 0.8471, val_loss = 5.6592, training accuracy = 91.42%, test accuracy = 28.14%
2017-06-16 11:58:20.502549  iteration = 048, loss = 0.8474, val_loss = 5.6593, training accuracy = 91.48%, test accuracy = 28.10%
2017-06-16 12:49:14.555265  iteration = 049, loss = 0.8492, val_loss = 5.6602, training accuracy = 91.50%, test accuracy = 28.12%
2017-06-16 13:38:25.202723  iteration = 050, loss = 0.8473, val_loss = 5.6595, training accuracy = 91.50%, test accuracy = 28.14%
2017-06-16 14:20:03.726760  iteration = 051, loss = 0.8482, val_loss = 5.6575, training accuracy = 91.42%, test accuracy = 28.10%
2017-06-16 15:08:21.780632  iteration = 052, loss = 0.8489, val_loss = 5.6583, training accuracy = 91.48%, test accuracy = 28.10%
2017-06-16 15:58:54.146248  iteration = 053, loss = 0.8474, val_loss = 5.6596, training accuracy = 91.48%, test accuracy = 28.16%
2017-06-16 16:44:04.912118  iteration = 054, loss = 0.8497, val_loss = 5.6596, training accuracy = 91.46%, test accuracy = 28.18%
2017-06-16 17:31:06.858162  iteration = 055, loss = 0.8471, val_loss = 5.6623, training accuracy = 91.44%, test accuracy = 28.20%
2017-06-16 18:23:57.944928  iteration = 056, loss = 0.8475, val_loss = 5.6609, training accuracy = 91.46%, test accuracy = 28.22%
2017-06-16 19:12:44.667036  iteration = 057, loss = 0.8477, val_loss = 5.6619, training accuracy = 91.46%, test accuracy = 28.18%
2017-06-16 19:58:47.673896  iteration = 058, loss = 0.8478, val_loss = 5.6616, training accuracy = 91.50%, test accuracy = 28.18%
2017-06-16 20:51:19.023237  iteration = 059, loss = 0.8468, val_loss = 5.6629, training accuracy = 91.40%, test accuracy = 28.12%
2017-06-16 21:41:22.323474  iteration = 060, loss = 0.8462, val_loss = 5.6637, training accuracy = 91.48%, test accuracy = 28.10%
2017-06-16 22:25:06.521850  iteration = 061, loss = 0.8461, val_loss = 5.6619, training accuracy = 91.42%, test accuracy = 28.16%
2017-06-16 23:09:02.822619  iteration = 062, loss = 0.8467, val_loss = 5.6633, training accuracy = 91.44%, test accuracy = 28.16%
2017-06-17 00:00:39.784381  iteration = 063, loss = 0.8440, val_loss = 5.6642, training accuracy = 91.48%, test accuracy = 28.12%
2017-06-17 00:51:14.033341  iteration = 064, loss = 0.8461, val_loss = 5.6652, training accuracy = 91.48%, test accuracy = 28.16%
2017-06-17 01:39:33.368192  iteration = 065, loss = 0.8461, val_loss = 5.6633, training accuracy = 91.48%, test accuracy = 28.18%
2017-06-17 02:25:11.015264  iteration = 066, loss = 0.8448, val_loss = 5.6650, training accuracy = 91.54%, test accuracy = 28.16%
2017-06-17 03:09:03.951847  iteration = 067, loss = 0.8446, val_loss = 5.6650, training accuracy = 91.58%, test accuracy = 28.12%
2017-06-17 04:01:19.105987  iteration = 068, loss = 0.8443, val_loss = 5.6655, training accuracy = 91.50%, test accuracy = 28.12%
2017-06-17 04:51:38.741608  iteration = 069, loss = 0.8439, val_loss = 5.6649, training accuracy = 91.50%, test accuracy = 28.22%
2017-06-17 05:39:23.719666  iteration = 070, loss = 0.8458, val_loss = 5.6653, training accuracy = 91.56%, test accuracy = 28.22%
2017-06-17 06:25:04.028635  iteration = 071, loss = 0.8460, val_loss = 5.6657, training accuracy = 91.54%, test accuracy = 28.18%
2017-06-17 07:09:04.527166  iteration = 072, loss = 0.8436, val_loss = 5.6656, training accuracy = 91.52%, test accuracy = 28.14%
2017-06-17 08:01:10.619199  iteration = 073, loss = 0.8459, val_loss = 5.6651, training accuracy = 91.50%, test accuracy = 28.16%
2017-06-17 08:50:51.644884  iteration = 074, loss = 0.8455, val_loss = 5.6661, training accuracy = 91.52%, test accuracy = 28.16%
2017-06-17 09:36:47.485885  iteration = 075, loss = 0.8439, val_loss = 5.6688, training accuracy = 91.56%, test accuracy = 28.20%
2017-06-17 10:21:05.345663  iteration = 076, loss = 0.8429, val_loss = 5.6676, training accuracy = 91.50%, test accuracy = 28.18%
2017-06-17 11:10:42.625862  iteration = 077, loss = 0.8444, val_loss = 5.6691, training accuracy = 91.58%, test accuracy = 28.18%
2017-06-17 12:02:08.135549  iteration = 078, loss = 0.8447, val_loss = 5.6669, training accuracy = 91.56%, test accuracy = 28.24%
2017-06-17 12:51:35.875635  iteration = 079, loss = 0.8434, val_loss = 5.6685, training accuracy = 91.50%, test accuracy = 28.16%
2017-06-17 13:37:37.560058  iteration = 080, loss = 0.8434, val_loss = 5.6690, training accuracy = 91.58%, test accuracy = 28.20%
2017-06-17 14:22:14.110488  iteration = 081, loss = 0.8443, val_loss = 5.6677, training accuracy = 91.56%, test accuracy = 28.18%
2017-06-17 15:06:09.053496  iteration = 082, loss = 0.8431, val_loss = 5.6687, training accuracy = 91.56%, test accuracy = 28.12%
2017-06-17 15:58:01.726807  iteration = 083, loss = 0.8414, val_loss = 5.6699, training accuracy = 91.56%, test accuracy = 28.14%
2017-06-17 16:48:38.937412  iteration = 084, loss = 0.8418, val_loss = 5.6701, training accuracy = 91.58%, test accuracy = 28.14%
2017-06-17 17:34:49.983236  iteration = 085, loss = 0.8434, val_loss = 5.6679, training accuracy = 91.58%, test accuracy = 28.16%
2017-06-17 18:18:46.722768  iteration = 086, loss = 0.8422, val_loss = 5.6720, training accuracy = 91.50%, test accuracy = 28.20%
2017-06-17 19:02:56.639372  iteration = 087, loss = 0.8425, val_loss = 5.6706, training accuracy = 91.58%, test accuracy = 28.12%
2017-06-17 19:55:20.019972  iteration = 088, loss = 0.8428, val_loss = 5.6716, training accuracy = 91.54%, test accuracy = 28.14%
2017-06-17 20:45:18.906874  iteration = 089, loss = 0.8425, val_loss = 5.6719, training accuracy = 91.56%, test accuracy = 28.10%
2017-06-17 21:31:27.517645  iteration = 090, loss = 0.8415, val_loss = 5.6708, training accuracy = 91.52%, test accuracy = 28.08%
2017-06-17 22:16:30.256038  iteration = 091, loss = 0.8399, val_loss = 5.6724, training accuracy = 91.58%, test accuracy = 28.18%
2017-06-17 23:00:16.811263  iteration = 092, loss = 0.8425, val_loss = 5.6710, training accuracy = 91.54%, test accuracy = 28.10%
2017-06-17 23:44:19.788725  iteration = 093, loss = 0.8422, val_loss = 5.6730, training accuracy = 91.54%, test accuracy = 28.06%
2017-06-18 00:32:56.941282  iteration = 094, loss = 0.8408, val_loss = 5.6718, training accuracy = 91.52%, test accuracy = 28.14%
2017-06-18 01:23:06.486099  iteration = 095, loss = 0.8423, val_loss = 5.6745, training accuracy = 91.50%, test accuracy = 28.08%
2017-06-18 02:11:24.813313  iteration = 096, loss = 0.8397, val_loss = 5.6735, training accuracy = 91.54%, test accuracy = 28.08%
2017-06-18 02:58:23.440950  iteration = 097, loss = 0.8406, val_loss = 5.6726, training accuracy = 91.68%, test accuracy = 28.12%
2017-06-18 03:43:59.601437  iteration = 098, loss = 0.8419, val_loss = 5.6740, training accuracy = 91.60%, test accuracy = 28.12%
2017-06-18 04:29:16.369906  iteration = 099, loss = 0.8401, val_loss = 5.6734, training accuracy = 91.62%, test accuracy = 28.14%
2017-06-18 05:13:13.725597  iteration = 100, loss = 0.8415, val_loss = 5.6726, training accuracy = 91.60%, test accuracy = 28.12%
2017-06-18 06:05:49.660980  iteration = 101, loss = 0.8403, val_loss = 5.6745, training accuracy = 91.58%, test accuracy = 28.08%
2017-06-18 06:53:47.949144  iteration = 102, loss = 0.8401, val_loss = 5.6766, training accuracy = 91.62%, test accuracy = 28.10%
2017-06-18 07:40:01.817117  iteration = 103, loss = 0.8401, val_loss = 5.6760, training accuracy = 91.60%, test accuracy = 28.18%
2017-06-18 08:26:12.964637  iteration = 104, loss = 0.8381, val_loss = 5.6758, training accuracy = 91.66%, test accuracy = 28.14%
2017-06-18 09:10:02.591703  iteration = 105, loss = 0.8403, val_loss = 5.6771, training accuracy = 91.62%, test accuracy = 28.08%
2017-06-18 09:53:49.092423  iteration = 106, loss = 0.8399, val_loss = 5.6746, training accuracy = 91.62%, test accuracy = 28.14%
2017-06-18 10:37:42.099762  iteration = 107, loss = 0.8387, val_loss = 5.6757, training accuracy = 91.62%, test accuracy = 28.10%
Stopped to re-organise directories

Continuing training with a further run
sjb@persephone:sjb_word_model$ ./train.py theresa_may_contributions.json
2017-06-18 11:27:31.145809  iteration = 000, loss = 0.8391, val_loss = 5.6762, training accuracy = 91.66%, test accuracy = 28.12%
2017-06-18 12:16:00.397795  iteration = 001, loss = 0.8383, val_loss = 5.6778, training accuracy = 91.60%, test accuracy = 28.12%
2017-06-18 13:01:54.238534  iteration = 002, loss = 0.8390, val_loss = 5.6771, training accuracy = 91.64%, test accuracy = 28.08%
2017-06-18 13:45:57.434029  iteration = 003, loss = 0.8377, val_loss = 5.6776, training accuracy = 91.64%, test accuracy = 28.18%
2017-06-18 14:37:54.596785  iteration = 004, loss = 0.8387, val_loss = 5.6784, training accuracy = 91.64%, test accuracy = 28.16%
2017-06-18 15:31:03.656096  iteration = 005, loss = 0.8388, val_loss = 5.6777, training accuracy = 91.58%, test accuracy = 28.18%
2017-06-18 16:19:05.688904  iteration = 006, loss = 0.8372, val_loss = 5.6791, training accuracy = 91.66%, test accuracy = 28.10%
2017-06-18 17:03:01.020011  iteration = 007, loss = 0.8400, val_loss = 5.6790, training accuracy = 91.60%, test accuracy = 28.10%