您好，我在给出的默认配置下在MR只能取得0.7左右的acc
我参考这个issue下面的一些配置，如下：
环境配置：
Python 3.6.13
Tensorflow 1.11.0
Scipy 1.5.4
参数配置：
learning_rate 0.005
epochs 300
batch_size 256
input_dim 300
Hidden 128
steps 1
dropout 0.5
weight_decay 0
early_stopping -1
max_degree 3
也无法取得论文汇报的结果
这是我的运行日志
D:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\framework\dtypes.py:523: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future versio
n of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint8 = np.dtype([("qint8", np.int8, 1)])
D:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\framework\dtypes.py:524: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future versio
n of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint8 = np.dtype([("quint8", np.uint8, 1)])
D:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\framework\dtypes.py:525: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future versio
n of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint16 = np.dtype([("qint16", np.int16, 1)])
D:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\framework\dtypes.py:526: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future versio
n of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint16 = np.dtype([("quint16", np.uint16, 1)])
D:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\framework\dtypes.py:527: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future versio
n of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint32 = np.dtype([("qint32", np.int32, 1)])
D:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\framework\dtypes.py:532: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future versio
n of numpy, it will be understood as (type, (1,)) / '(1,)type'.
np_resource = np.dtype([("resource", np.ubyte, 1)])
E:\desktop\TextING-master\utils.py:82: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with differe
nt lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray
train_adj = np.array(train_adj)
E:\desktop\TextING-master\utils.py:83: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with differe
nt lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray
val_adj = np.array(val_adj)
E:\desktop\TextING-master\utils.py:84: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with differe
nt lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray
test_adj = np.array(test_adj)
E:\desktop\TextING-master\utils.py:85: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with differe
nt lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray
train_embed = np.array(train_embed)
E:\desktop\TextING-master\utils.py:86: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with differe
nt lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray
val_embed = np.array(val_embed)
E:\desktop\TextING-master\utils.py:87: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with differe
nt lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray
test_embed = np.array(test_embed)
loading training set
100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 6398/6398 [00:00<00:00, 10601.33it/s]
100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 6398/6398 [00:00<00:00, 11214.96it/s]
loading validation set
100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 710/710 [00:00<00:00, 10162.46it/s]
100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 710/710 [00:00<00:00, 12274.92it/s]
loading test set
100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3554/3554 [00:00<00:00, 10605.50it/s]
100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3554/3554 [00:00<00:00, 11161.50it/s]
build...
WARNING:tensorflow:From E:\desktop\TextING-master\metrics.py:6: softmax_cross_entropy_with_logits (from tensorflow.python.ops.nn_ops) is deprecated and will be removed in a future vers
ion.
Instructions for updating:
Future major versions of TensorFlow will allow gradients to flow
into the labels input on backprop by default.
Macro average Test Precision, Recall and F1-Score...
(0.7209362974880018, 0.7191896454698932, 0.7186335470736362, None)
Micro average Test Precision, Recall and F1-Score...
(0.7191896454698931, 0.7191896454698931, 0.7191896454698931, None)
您好，我在给出的默认配置下在MR只能取得0.7左右的acc 我参考这个issue下面的一些配置，如下： 环境配置： Python 3.6.13 Tensorflow 1.11.0 Scipy 1.5.4 参数配置： learning_rate 0.005 epochs 300 batch_size 256 input_dim 300 Hidden 128 steps 1 dropout 0.5 weight_decay 0 early_stopping -1 max_degree 3 也无法取得论文汇报的结果 这是我的运行日志 D:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\framework\dtypes.py:523: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future versio n of numpy, it will be understood as (type, (1,)) / '(1,)type'. _np_qint8 = np.dtype([("qint8", np.int8, 1)]) D:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\framework\dtypes.py:524: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future versio n of numpy, it will be understood as (type, (1,)) / '(1,)type'. _np_quint8 = np.dtype([("quint8", np.uint8, 1)]) D:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\framework\dtypes.py:525: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future versio n of numpy, it will be understood as (type, (1,)) / '(1,)type'. _np_qint16 = np.dtype([("qint16", np.int16, 1)]) D:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\framework\dtypes.py:526: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future versio n of numpy, it will be understood as (type, (1,)) / '(1,)type'. _np_quint16 = np.dtype([("quint16", np.uint16, 1)]) D:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\framework\dtypes.py:527: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future versio n of numpy, it will be understood as (type, (1,)) / '(1,)type'. _np_qint32 = np.dtype([("qint32", np.int32, 1)]) D:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\framework\dtypes.py:532: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future versio n of numpy, it will be understood as (type, (1,)) / '(1,)type'. np_resource = np.dtype([("resource", np.ubyte, 1)]) E:\desktop\TextING-master\utils.py:82: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with differe nt lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray train_adj = np.array(train_adj) E:\desktop\TextING-master\utils.py:83: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with differe nt lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray val_adj = np.array(val_adj) E:\desktop\TextING-master\utils.py:84: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with differe nt lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray test_adj = np.array(test_adj) E:\desktop\TextING-master\utils.py:85: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with differe nt lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray train_embed = np.array(train_embed) E:\desktop\TextING-master\utils.py:86: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with differe nt lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray val_embed = np.array(val_embed) E:\desktop\TextING-master\utils.py:87: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with differe nt lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray test_embed = np.array(test_embed) loading training set 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 6398/6398 [00:00<00:00, 10601.33it/s] 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 6398/6398 [00:00<00:00, 11214.96it/s] loading validation set 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 710/710 [00:00<00:00, 10162.46it/s] 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 710/710 [00:00<00:00, 12274.92it/s] loading test set 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3554/3554 [00:00<00:00, 10605.50it/s] 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3554/3554 [00:00<00:00, 11161.50it/s] build... WARNING:tensorflow:From E:\desktop\TextING-master\metrics.py:6: softmax_cross_entropy_with_logits (from tensorflow.python.ops.nn_ops) is deprecated and will be removed in a future vers ion. Instructions for updating:
Future major versions of TensorFlow will allow gradients to flow into the labels input on backprop by default.
See
tf.nn.softmax_cross_entropy_with_logits_v2.2022-05-13 10:29:41.954666: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 train start... Epoch: 0001 train_loss= 0.68974 train_acc= 0.52032 val_loss= 0.70149 val_acc= 0.50563 test_acc= 0.54811 time= 5.35665 Epoch: 0002 train_loss= 0.68966 train_acc= 0.51766 val_loss= 0.69056 val_acc= 0.50563 test_acc= 0.54446 time= 4.99060 Epoch: 0003 train_loss= 0.68522 train_acc= 0.51454 val_loss= 0.69074 val_acc= 0.52394 test_acc= 0.51379 time= 4.98833 Epoch: 0004 train_loss= 0.68161 train_acc= 0.53063 val_loss= 0.68953 val_acc= 0.50282 test_acc= 0.54615 time= 4.96672 Epoch: 0005 train_loss= 0.67835 train_acc= 0.53814 val_loss= 0.68977 val_acc= 0.52676 test_acc= 0.51041 time= 4.95613 Epoch: 0006 train_loss= 0.67787 train_acc= 0.53892 val_loss= 0.68808 val_acc= 0.53099 test_acc= 0.51379 time= 4.97210 Epoch: 0007 train_loss= 0.67805 train_acc= 0.53470 val_loss= 0.68568 val_acc= 0.54225 test_acc= 0.53433 time= 4.94400 Epoch: 0008 train_loss= 0.67743 train_acc= 0.55674 val_loss= 0.68912 val_acc= 0.51831 test_acc= 0.54727 time= 5.01859 Epoch: 0009 train_loss= 0.67182 train_acc= 0.56549 val_loss= 0.69753 val_acc= 0.52958 test_acc= 0.51407 time= 4.98467 Epoch: 0010 train_loss= 0.67658 train_acc= 0.55080 val_loss= 0.68936 val_acc= 0.52676 test_acc= 0.51519 time= 4.98866 Epoch: 0011 train_loss= 0.67274 train_acc= 0.56486 val_loss= 0.69097 val_acc= 0.53944 test_acc= 0.52729 time= 4.96491 Epoch: 0012 train_loss= 0.67346 train_acc= 0.54658 val_loss= 0.68696 val_acc= 0.53099 test_acc= 0.51491 time= 4.97670 Epoch: 0013 train_loss= 0.66519 train_acc= 0.57627 val_loss= 0.69759 val_acc= 0.57746 test_acc= 0.58301 time= 4.98898 Epoch: 0014 train_loss= 0.65975 train_acc= 0.59597 val_loss= 0.68219 val_acc= 0.59155 test_acc= 0.58188 time= 4.99568 Epoch: 0015 train_loss= 0.65782 train_acc= 0.59565 val_loss= 0.67787 val_acc= 0.57324 test_acc= 0.59764 time= 4.95875 Epoch: 0016 train_loss= 0.65551 train_acc= 0.59691 val_loss= 0.67830 val_acc= 0.58169 test_acc= 0.59313 time= 4.98974 Epoch: 0017 train_loss= 0.65227 train_acc= 0.60613 val_loss= 0.68627 val_acc= 0.56197 test_acc= 0.57147 time= 4.96651 Epoch: 0018 train_loss= 0.64660 train_acc= 0.61676 val_loss= 0.67300 val_acc= 0.60704 test_acc= 0.59876 time= 4.98168 Epoch: 0019 train_loss= 0.64479 train_acc= 0.62254 val_loss= 0.67970 val_acc= 0.57606 test_acc= 0.58497 time= 4.94029 Epoch: 0020 train_loss= 0.64635 train_acc= 0.62316 val_loss= 0.66683 val_acc= 0.61549 test_acc= 0.60411 time= 4.96796 Epoch: 0021 train_loss= 0.63985 train_acc= 0.62879 val_loss= 0.66047 val_acc= 0.59577 test_acc= 0.60355 time= 4.96473 Epoch: 0022 train_loss= 0.63489 train_acc= 0.63129 val_loss= 0.66431 val_acc= 0.60000 test_acc= 0.59623 time= 5.02756 Epoch: 0023 train_loss= 0.63856 train_acc= 0.62426 val_loss= 0.66811 val_acc= 0.58732 test_acc= 0.58554 time= 4.94330 Epoch: 0024 train_loss= 0.63843 train_acc= 0.62613 val_loss= 0.65142 val_acc= 0.60423 test_acc= 0.61480 time= 4.97187 Epoch: 0025 train_loss= 0.62820 train_acc= 0.64739 val_loss= 0.65906 val_acc= 0.61831 test_acc= 0.61367 time= 4.98656 Epoch: 0026 train_loss= 0.62406 train_acc= 0.64364 val_loss= 0.65610 val_acc= 0.61690 test_acc= 0.61621 time= 4.99516 Epoch: 0027 train_loss= 0.62658 train_acc= 0.64270 val_loss= 0.64364 val_acc= 0.63239 test_acc= 0.61227 time= 5.03854 Epoch: 0028 train_loss= 0.62254 train_acc= 0.64458 val_loss= 0.64637 val_acc= 0.63099 test_acc= 0.61790 time= 4.98452 Epoch: 0029 train_loss= 0.61804 train_acc= 0.65036 val_loss= 0.64005 val_acc= 0.63239 test_acc= 0.61649 time= 4.97380 Epoch: 0030 train_loss= 0.63158 train_acc= 0.63614 val_loss= 0.65047 val_acc= 0.61972 test_acc= 0.62324 time= 4.94871 Epoch: 0031 train_loss= 0.62685 train_acc= 0.64379 val_loss= 0.65903 val_acc= 0.61408 test_acc= 0.59539 time= 4.92983 Epoch: 0032 train_loss= 0.61852 train_acc= 0.65192 val_loss= 0.65042 val_acc= 0.62394 test_acc= 0.62183 time= 5.01659 Epoch: 0033 train_loss= 0.62141 train_acc= 0.65052 val_loss= 0.63492 val_acc= 0.64225 test_acc= 0.62859 time= 4.96573 Epoch: 0034 train_loss= 0.61643 train_acc= 0.66146 val_loss= 0.64947 val_acc= 0.62676 test_acc= 0.62409 time= 4.95609 Epoch: 0035 train_loss= 0.61474 train_acc= 0.65692 val_loss= 0.64714 val_acc= 0.64085 test_acc= 0.62212 time= 4.95162 Epoch: 0036 train_loss= 0.61268 train_acc= 0.66130 val_loss= 0.64826 val_acc= 0.63099 test_acc= 0.62971 time= 4.98868 Epoch: 0037 train_loss= 0.61732 train_acc= 0.65442 val_loss= 0.65289 val_acc= 0.61549 test_acc= 0.59651 time= 4.94732 Epoch: 0038 train_loss= 0.61634 train_acc= 0.65599 val_loss= 0.65297 val_acc= 0.64225 test_acc= 0.63534 time= 4.98268 Epoch: 0039 train_loss= 0.61135 train_acc= 0.66114 val_loss= 0.64255 val_acc= 0.62394 test_acc= 0.62465 time= 4.99166 Epoch: 0040 train_loss= 0.60466 train_acc= 0.66818 val_loss= 0.64700 val_acc= 0.62958 test_acc= 0.63253 time= 4.98893 Epoch: 0041 train_loss= 0.61299 train_acc= 0.65708 val_loss= 0.64733 val_acc= 0.63099 test_acc= 0.62521 time= 4.98567 Epoch: 0042 train_loss= 0.60736 train_acc= 0.67349 val_loss= 0.65121 val_acc= 0.63662 test_acc= 0.62577 time= 4.95973 Epoch: 0043 train_loss= 0.59990 train_acc= 0.67115 val_loss= 0.64064 val_acc= 0.63662 test_acc= 0.62999 time= 4.99592 Epoch: 0044 train_loss= 0.60069 train_acc= 0.67240 val_loss= 0.64601 val_acc= 0.64366 test_acc= 0.63844 time= 4.98468 Epoch: 0045 train_loss= 0.60442 train_acc= 0.66443 val_loss= 0.63394 val_acc= 0.64225 test_acc= 0.64012 time= 5.04814 Epoch: 0046 train_loss= 0.59759 train_acc= 0.67318 val_loss= 0.65898 val_acc= 0.63944 test_acc= 0.63478 time= 4.99514 Epoch: 0047 train_loss= 0.59201 train_acc= 0.67146 val_loss= 0.64651 val_acc= 0.63803 test_acc= 0.62943 time= 5.02191 Epoch: 0048 train_loss= 0.59398 train_acc= 0.67537 val_loss= 0.65326 val_acc= 0.63944 test_acc= 0.63225 time= 4.92983 Epoch: 0049 train_loss= 0.59927 train_acc= 0.66943 val_loss= 0.64316 val_acc= 0.65070 test_acc= 0.64575 time= 4.97969 Epoch: 0050 train_loss= 0.59338 train_acc= 0.68053 val_loss= 0.65488 val_acc= 0.63380 test_acc= 0.63844 time= 4.96904 Epoch: 0051 train_loss= 0.59872 train_acc= 0.67224 val_loss= 0.64356 val_acc= 0.65493 test_acc= 0.63618 time= 4.96250 Epoch: 0052 train_loss= 0.58985 train_acc= 0.68037 val_loss= 0.63632 val_acc= 0.65775 test_acc= 0.64547 time= 4.95216 Epoch: 0053 train_loss= 0.59337 train_acc= 0.67818 val_loss= 0.65191 val_acc= 0.64366 test_acc= 0.63900 time= 4.96043 Epoch: 0054 train_loss= 0.59341 train_acc= 0.67209 val_loss= 0.64374 val_acc= 0.64648 test_acc= 0.63900 time= 5.06845 Epoch: 0055 train_loss= 0.59038 train_acc= 0.67412 val_loss= 0.63084 val_acc= 0.65493 test_acc= 0.64800 time= 5.00462 Epoch: 0056 train_loss= 0.58991 train_acc= 0.68146 val_loss= 0.64089 val_acc= 0.66056 test_acc= 0.63759 time= 5.08230 Epoch: 0057 train_loss= 0.58881 train_acc= 0.67974 val_loss= 0.63646 val_acc= 0.65775 test_acc= 0.65504 time= 5.02537 Epoch: 0058 train_loss= 0.59118 train_acc= 0.68021 val_loss= 0.65055 val_acc= 0.64507 test_acc= 0.65110 time= 5.11517 Epoch: 0059 train_loss= 0.58563 train_acc= 0.68318 val_loss= 0.64125 val_acc= 0.64930 test_acc= 0.65307 time= 5.12387 Epoch: 0060 train_loss= 0.58107 train_acc= 0.68693 val_loss= 0.65093 val_acc= 0.65634 test_acc= 0.64885 time= 5.06449 Epoch: 0061 train_loss= 0.58328 train_acc= 0.68943 val_loss= 0.65025 val_acc= 0.64507 test_acc= 0.64434 time= 5.13727 Epoch: 0062 train_loss= 0.59260 train_acc= 0.66865 val_loss= 0.64489 val_acc= 0.64366 test_acc= 0.64631 time= 5.07548 Epoch: 0063 train_loss= 0.57940 train_acc= 0.68428 val_loss= 0.64480 val_acc= 0.63944 test_acc= 0.64547 time= 5.01047 Epoch: 0064 train_loss= 0.57667 train_acc= 0.69256 val_loss= 0.64620 val_acc= 0.64225 test_acc= 0.65869 time= 4.95010 Epoch: 0065 train_loss= 0.57287 train_acc= 0.69756 val_loss= 0.64612 val_acc= 0.65352 test_acc= 0.65926 time= 5.08206 Epoch: 0066 train_loss= 0.57957 train_acc= 0.68834 val_loss= 0.64258 val_acc= 0.65493 test_acc= 0.64885 time= 5.06247 Epoch: 0067 train_loss= 0.57854 train_acc= 0.69068 val_loss= 0.63763 val_acc= 0.64789 test_acc= 0.64828 time= 5.20110 Epoch: 0068 train_loss= 0.57406 train_acc= 0.69834 val_loss= 0.64716 val_acc= 0.65493 test_acc= 0.65869 time= 5.20409 Epoch: 0069 train_loss= 0.58027 train_acc= 0.68897 val_loss= 0.64970 val_acc= 0.64085 test_acc= 0.65138 time= 4.97971 Epoch: 0070 train_loss= 0.57864 train_acc= 0.68646 val_loss= 0.64732 val_acc= 0.64930 test_acc= 0.65616 time= 5.00182 Epoch: 0071 train_loss= 0.57718 train_acc= 0.69193 val_loss= 0.65608 val_acc= 0.63944 test_acc= 0.64800 time= 4.95645 Epoch: 0072 train_loss= 0.57407 train_acc= 0.69381 val_loss= 0.66782 val_acc= 0.63944 test_acc= 0.63675 time= 4.98421 Epoch: 0073 train_loss= 0.56996 train_acc= 0.69897 val_loss= 0.66796 val_acc= 0.65352 test_acc= 0.65053 time= 4.94778 Epoch: 0074 train_loss= 0.56765 train_acc= 0.68990 val_loss= 0.66288 val_acc= 0.64789 test_acc= 0.65138 time= 4.96218 Epoch: 0075 train_loss= 0.57340 train_acc= 0.69272 val_loss= 0.65589 val_acc= 0.64507 test_acc= 0.65813 time= 4.95720 Epoch: 0076 train_loss= 0.56416 train_acc= 0.70288 val_loss= 0.65214 val_acc= 0.63803 test_acc= 0.66826 time= 4.99068 Epoch: 0077 train_loss= 0.57238 train_acc= 0.70084 val_loss= 0.64875 val_acc= 0.63521 test_acc= 0.66545 time= 4.97100 Epoch: 0078 train_loss= 0.57246 train_acc= 0.68943 val_loss= 0.65582 val_acc= 0.64366 test_acc= 0.66770 time= 4.96927 Epoch: 0079 train_loss= 0.56804 train_acc= 0.70006 val_loss= 0.66752 val_acc= 0.64225 test_acc= 0.66629 time= 4.93702 Epoch: 0080 train_loss= 0.56916 train_acc= 0.69709 val_loss= 0.66552 val_acc= 0.64225 test_acc= 0.65869 time= 4.96917 Epoch: 0081 train_loss= 0.56625 train_acc= 0.70585 val_loss= 0.66044 val_acc= 0.63803 test_acc= 0.66432 time= 4.93676 Epoch: 0082 train_loss= 0.55204 train_acc= 0.71132 val_loss= 0.67733 val_acc= 0.65211 test_acc= 0.67023 time= 5.00409 Epoch: 0083 train_loss= 0.56306 train_acc= 0.70006 val_loss= 0.66638 val_acc= 0.64225 test_acc= 0.66826 time= 4.92743 Epoch: 0084 train_loss= 0.57073 train_acc= 0.70131 val_loss= 0.64397 val_acc= 0.63803 test_acc= 0.65954 time= 5.01360 Epoch: 0085 train_loss= 0.55809 train_acc= 0.70600 val_loss= 0.65279 val_acc= 0.65352 test_acc= 0.67501 time= 4.96874 Epoch: 0086 train_loss= 0.55939 train_acc= 0.69975 val_loss= 0.67705 val_acc= 0.63662 test_acc= 0.65476 time= 5.01061 Epoch: 0087 train_loss= 0.56485 train_acc= 0.70334 val_loss= 0.66047 val_acc= 0.65352 test_acc= 0.66320 time= 4.97470 Epoch: 0088 train_loss= 0.56820 train_acc= 0.70225 val_loss= 0.66241 val_acc= 0.65352 test_acc= 0.66207 time= 5.01859 Epoch: 0089 train_loss= 0.56629 train_acc= 0.70272 val_loss= 0.67573 val_acc= 0.64648 test_acc= 0.66263 time= 4.97581 Epoch: 0090 train_loss= 0.56202 train_acc= 0.70194 val_loss= 0.64849 val_acc= 0.65775 test_acc= 0.66685 time= 4.97205 Epoch: 0091 train_loss= 0.55850 train_acc= 0.70475 val_loss= 0.65591 val_acc= 0.65211 test_acc= 0.65954 time= 4.95616 Epoch: 0092 train_loss= 0.55660 train_acc= 0.70491 val_loss= 0.68784 val_acc= 0.64789 test_acc= 0.66123 time= 4.98866 Epoch: 0093 train_loss= 0.55643 train_acc= 0.71241 val_loss= 0.66816 val_acc= 0.65211 test_acc= 0.65476 time= 4.95595 Epoch: 0094 train_loss= 0.55211 train_acc= 0.70756 val_loss= 0.68170 val_acc= 0.65211 test_acc= 0.65025 time= 5.02158 Epoch: 0095 train_loss= 0.55504 train_acc= 0.70835 val_loss= 0.65302 val_acc= 0.66901 test_acc= 0.66882 time= 5.00530 Epoch: 0096 train_loss= 0.55029 train_acc= 0.71647 val_loss= 0.66579 val_acc= 0.66901 test_acc= 0.66911 time= 5.02058 Epoch: 0097 train_loss= 0.55393 train_acc= 0.70991 val_loss= 0.65799 val_acc= 0.65775 test_acc= 0.66404 time= 5.00297 Epoch: 0098 train_loss= 0.55430 train_acc= 0.70710 val_loss= 0.68202 val_acc= 0.66056 test_acc= 0.65701 time= 4.98578 Epoch: 0099 train_loss= 0.55447 train_acc= 0.70803 val_loss= 0.65508 val_acc= 0.66197 test_acc= 0.66798 time= 4.98626 Epoch: 0100 train_loss= 0.54873 train_acc= 0.71257 val_loss= 0.64963 val_acc= 0.65634 test_acc= 0.67361 time= 4.99864 Epoch: 0101 train_loss= 0.54908 train_acc= 0.71132 val_loss= 0.66229 val_acc= 0.66056 test_acc= 0.66770 time= 4.96474 Epoch: 0102 train_loss= 0.55235 train_acc= 0.71647 val_loss= 0.66755 val_acc= 0.66197 test_acc= 0.66939 time= 4.95418 Epoch: 0103 train_loss= 0.54913 train_acc= 0.71491 val_loss= 0.65488 val_acc= 0.66338 test_acc= 0.67839 time= 4.96672 Epoch: 0104 train_loss= 0.54182 train_acc= 0.71522 val_loss= 0.66820 val_acc= 0.65493 test_acc= 0.66460 time= 5.00501 Epoch: 0105 train_loss= 0.53851 train_acc= 0.72288 val_loss= 0.67286 val_acc= 0.65634 test_acc= 0.67248 time= 5.01858 Epoch: 0106 train_loss= 0.53727 train_acc= 0.71679 val_loss= 0.66831 val_acc= 0.66056 test_acc= 0.67304 time= 4.98966 Epoch: 0107 train_loss= 0.54169 train_acc= 0.72319 val_loss= 0.66395 val_acc= 0.66620 test_acc= 0.67361 time= 4.92732 Epoch: 0108 train_loss= 0.54057 train_acc= 0.72226 val_loss= 0.67008 val_acc= 0.66197 test_acc= 0.67164 time= 4.96099 Epoch: 0109 train_loss= 0.53815 train_acc= 0.71819 val_loss= 0.66328 val_acc= 0.64648 test_acc= 0.67389 time= 4.95283 Epoch: 0110 train_loss= 0.54955 train_acc= 0.71569 val_loss= 0.66743 val_acc= 0.65915 test_acc= 0.67304 time= 4.99247 Epoch: 0111 train_loss= 0.53546 train_acc= 0.72648 val_loss= 0.66124 val_acc= 0.66901 test_acc= 0.67136 time= 4.98678 Epoch: 0112 train_loss= 0.53669 train_acc= 0.71804 val_loss= 0.66199 val_acc= 0.66338 test_acc= 0.66601 time= 4.99452 Epoch: 0113 train_loss= 0.54023 train_acc= 0.71991 val_loss= 0.67346 val_acc= 0.64648 test_acc= 0.65982 time= 4.96174 Epoch: 0114 train_loss= 0.53957 train_acc= 0.72382 val_loss= 0.69259 val_acc= 0.64930 test_acc= 0.67445 time= 4.98375 Epoch: 0115 train_loss= 0.53671 train_acc= 0.71944 val_loss= 0.65979 val_acc= 0.65352 test_acc= 0.66995 time= 4.93533 Epoch: 0116 train_loss= 0.53831 train_acc= 0.72132 val_loss= 0.65991 val_acc= 0.65915 test_acc= 0.67614 time= 4.98467 Epoch: 0117 train_loss= 0.52836 train_acc= 0.73070 val_loss= 0.66416 val_acc= 0.65775 test_acc= 0.67895 time= 4.95901 Epoch: 0118 train_loss= 0.52833 train_acc= 0.73367 val_loss= 0.66063 val_acc= 0.66479 test_acc= 0.67164 time= 4.98524 Epoch: 0119 train_loss= 0.53873 train_acc= 0.72319 val_loss= 0.65915 val_acc= 0.65775 test_acc= 0.67333 time= 5.00176 Epoch: 0120 train_loss= 0.53022 train_acc= 0.72663 val_loss= 0.67533 val_acc= 0.66479 test_acc= 0.68346 time= 4.98169 Epoch: 0121 train_loss= 0.52038 train_acc= 0.73398 val_loss= 0.66636 val_acc= 0.67465 test_acc= 0.66742 time= 4.95211 Epoch: 0122 train_loss= 0.53495 train_acc= 0.72319 val_loss= 0.66245 val_acc= 0.66056 test_acc= 0.67839 time= 4.99012 Epoch: 0123 train_loss= 0.53108 train_acc= 0.72460 val_loss= 0.66969 val_acc= 0.67183 test_acc= 0.68008 time= 4.99652 Epoch: 0124 train_loss= 0.52125 train_acc= 0.74008 val_loss= 0.66424 val_acc= 0.66761 test_acc= 0.67614 time= 5.00263 Epoch: 0125 train_loss= 0.52038 train_acc= 0.73242 val_loss= 0.66380 val_acc= 0.66901 test_acc= 0.67811 time= 4.97313 Epoch: 0126 train_loss= 0.53327 train_acc= 0.71835 val_loss= 0.66741 val_acc= 0.65352 test_acc= 0.67642 time= 4.98268 Epoch: 0127 train_loss= 0.51911 train_acc= 0.74461 val_loss= 0.67886 val_acc= 0.67042 test_acc= 0.67051 time= 4.97729 Epoch: 0128 train_loss= 0.51586 train_acc= 0.73992 val_loss= 0.67685 val_acc= 0.66761 test_acc= 0.67670 time= 4.96672 Epoch: 0129 train_loss= 0.52237 train_acc= 0.73836 val_loss= 0.67010 val_acc= 0.66056 test_acc= 0.67698 time= 4.95643 Epoch: 0130 train_loss= 0.51809 train_acc= 0.72992 val_loss= 0.68235 val_acc= 0.67183 test_acc= 0.67586 time= 5.00884 Epoch: 0131 train_loss= 0.51905 train_acc= 0.73742 val_loss= 0.68609 val_acc= 0.66901 test_acc= 0.67980 time= 4.96295 Epoch: 0132 train_loss= 0.51128 train_acc= 0.73836 val_loss= 0.67983 val_acc= 0.66761 test_acc= 0.68796 time= 4.96868 Epoch: 0133 train_loss= 0.51202 train_acc= 0.74304 val_loss= 0.68441 val_acc= 0.66056 test_acc= 0.66714 time= 4.95504 Epoch: 0134 train_loss= 0.50597 train_acc= 0.75102 val_loss= 0.72047 val_acc= 0.66761 test_acc= 0.68008 time= 4.94378 Epoch: 0135 train_loss= 0.51088 train_acc= 0.74789 val_loss= 0.69227 val_acc= 0.66479 test_acc= 0.67811 time= 4.97969 Epoch: 0136 train_loss= 0.50734 train_acc= 0.74476 val_loss= 0.68167 val_acc= 0.66338 test_acc= 0.67530 time= 4.98000 Epoch: 0137 train_loss= 0.51246 train_acc= 0.73773 val_loss= 0.70116 val_acc= 0.66338 test_acc= 0.67755 time= 4.94073 Epoch: 0138 train_loss= 0.52333 train_acc= 0.73085 val_loss= 0.66459 val_acc= 0.66620 test_acc= 0.68458 time= 4.98419 Epoch: 0139 train_loss= 0.50838 train_acc= 0.73914 val_loss= 0.69545 val_acc= 0.66197 test_acc= 0.68036 time= 4.99279 Epoch: 0140 train_loss= 0.50828 train_acc= 0.74461 val_loss= 0.67580 val_acc= 0.66479 test_acc= 0.67952 time= 4.96573 Epoch: 0141 train_loss= 0.50259 train_acc= 0.74789 val_loss= 0.68539 val_acc= 0.67042 test_acc= 0.68852 time= 4.94604 Epoch: 0142 train_loss= 0.50153 train_acc= 0.75148 val_loss= 0.68514 val_acc= 0.67183 test_acc= 0.69584 time= 4.95781 Epoch: 0143 train_loss= 0.50346 train_acc= 0.75352 val_loss= 0.66855 val_acc= 0.67465 test_acc= 0.69668 time= 4.96714 Epoch: 0144 train_loss= 0.50636 train_acc= 0.74820 val_loss= 0.67624 val_acc= 0.65775 test_acc= 0.68852 time= 4.98181 Epoch: 0145 train_loss= 0.50029 train_acc= 0.75023 val_loss= 0.66726 val_acc= 0.66901 test_acc= 0.69443 time= 4.93978 Epoch: 0146 train_loss= 0.49692 train_acc= 0.74820 val_loss= 0.67311 val_acc= 0.67746 test_acc= 0.69443 time= 5.01160 Epoch: 0147 train_loss= 0.49912 train_acc= 0.74601 val_loss= 0.69719 val_acc= 0.67606 test_acc= 0.68008 time= 4.97319 Epoch: 0148 train_loss= 0.48868 train_acc= 0.75664 val_loss= 0.69013 val_acc= 0.67324 test_acc= 0.69105 time= 5.01582 Epoch: 0149 train_loss= 0.48671 train_acc= 0.76071 val_loss= 0.70432 val_acc= 0.67465 test_acc= 0.68317 time= 4.98900 Epoch: 0150 train_loss= 0.49042 train_acc= 0.75805 val_loss= 0.68115 val_acc= 0.67183 test_acc= 0.68824 time= 4.98156 Epoch: 0151 train_loss= 0.49113 train_acc= 0.75774 val_loss= 0.68832 val_acc= 0.67465 test_acc= 0.68346 time= 4.97770 Epoch: 0152 train_loss= 0.49713 train_acc= 0.75195 val_loss= 0.68923 val_acc= 0.66901 test_acc= 0.69612 time= 5.02513 Epoch: 0153 train_loss= 0.48653 train_acc= 0.76336 val_loss= 0.69096 val_acc= 0.67606 test_acc= 0.69246 time= 5.01855 Epoch: 0154 train_loss= 0.48528 train_acc= 0.76102 val_loss= 0.69321 val_acc= 0.68169 test_acc= 0.69387 time= 5.01182 Epoch: 0155 train_loss= 0.48941 train_acc= 0.75633 val_loss= 0.68016 val_acc= 0.67887 test_acc= 0.69612 time= 4.94686 Epoch: 0156 train_loss= 0.48943 train_acc= 0.76024 val_loss= 0.67278 val_acc= 0.67042 test_acc= 0.69077 time= 4.99375 Epoch: 0157 train_loss= 0.48270 train_acc= 0.77259 val_loss= 0.67920 val_acc= 0.67606 test_acc= 0.68768 time= 4.95768 Epoch: 0158 train_loss= 0.48653 train_acc= 0.76149 val_loss= 0.72142 val_acc= 0.66197 test_acc= 0.68036 time= 4.99765 Epoch: 0159 train_loss= 0.48814 train_acc= 0.75742 val_loss= 0.70440 val_acc= 0.66901 test_acc= 0.68965 time= 4.96672 Epoch: 0160 train_loss= 0.46947 train_acc= 0.77305 val_loss= 0.69346 val_acc= 0.66761 test_acc= 0.68796 time= 4.94904 Epoch: 0161 train_loss= 0.47349 train_acc= 0.76915 val_loss= 0.70256 val_acc= 0.66338 test_acc= 0.69865 time= 4.96292 Epoch: 0162 train_loss= 0.47093 train_acc= 0.77305 val_loss= 0.72137 val_acc= 0.66620 test_acc= 0.68739 time= 4.97870 Epoch: 0163 train_loss= 0.47031 train_acc= 0.77368 val_loss= 0.69899 val_acc= 0.68310 test_acc= 0.68768 time= 4.96324 Epoch: 0164 train_loss= 0.47593 train_acc= 0.77493 val_loss= 0.69749 val_acc= 0.67324 test_acc= 0.68908 time= 4.94618 Epoch: 0165 train_loss= 0.47000 train_acc= 0.76633 val_loss= 0.72306 val_acc= 0.65634 test_acc= 0.69133 time= 5.02387 Epoch: 0166 train_loss= 0.47706 train_acc= 0.76508 val_loss= 0.68819 val_acc= 0.66338 test_acc= 0.69133 time= 4.96905 Epoch: 0167 train_loss= 0.46394 train_acc= 0.77462 val_loss= 0.69400 val_acc= 0.66761 test_acc= 0.69471 time= 4.96885 Epoch: 0168 train_loss= 0.46237 train_acc= 0.77555 val_loss= 0.70935 val_acc= 0.67465 test_acc= 0.69584 time= 4.99166 Epoch: 0169 train_loss= 0.46857 train_acc= 0.77524 val_loss= 0.70653 val_acc= 0.67887 test_acc= 0.69133 time= 4.97471 Epoch: 0170 train_loss= 0.46968 train_acc= 0.77227 val_loss= 0.69461 val_acc= 0.67887 test_acc= 0.69133 time= 5.00219 Epoch: 0171 train_loss= 0.46122 train_acc= 0.77649 val_loss= 0.69894 val_acc= 0.67324 test_acc= 0.69668 time= 4.98766 Epoch: 0172 train_loss= 0.45405 train_acc= 0.77977 val_loss= 0.70575 val_acc= 0.66901 test_acc= 0.69612 time= 5.01872 Epoch: 0173 train_loss= 0.45008 train_acc= 0.78540 val_loss= 0.72087 val_acc= 0.66620 test_acc= 0.68852 time= 4.98065 Epoch: 0174 train_loss= 0.46957 train_acc= 0.77227 val_loss= 0.69190 val_acc= 0.67887 test_acc= 0.69105 time= 4.97172 Epoch: 0175 train_loss= 0.45999 train_acc= 0.77821 val_loss= 0.71654 val_acc= 0.67324 test_acc= 0.69752 time= 4.97172 Epoch: 0176 train_loss= 0.45098 train_acc= 0.78399 val_loss= 0.71002 val_acc= 0.67606 test_acc= 0.69865 time= 5.01759 Epoch: 0177 train_loss= 0.45315 train_acc= 0.78478 val_loss= 0.71155 val_acc= 0.68592 test_acc= 0.69246 time= 4.97595 Epoch: 0178 train_loss= 0.45232 train_acc= 0.78462 val_loss= 0.71577 val_acc= 0.67746 test_acc= 0.69415 time= 4.95676 Epoch: 0179 train_loss= 0.45345 train_acc= 0.78196 val_loss= 0.68200 val_acc= 0.68028 test_acc= 0.70062 time= 5.01559 Epoch: 0180 train_loss= 0.45020 train_acc= 0.78618 val_loss= 0.71269 val_acc= 0.67183 test_acc= 0.69668 time= 4.95874 Epoch: 0181 train_loss= 0.44114 train_acc= 0.78978 val_loss= 0.67986 val_acc= 0.67465 test_acc= 0.70990 time= 4.94629 Epoch: 0182 train_loss= 0.44462 train_acc= 0.79165 val_loss= 0.70316 val_acc= 0.69014 test_acc= 0.70006 time= 4.96818 Epoch: 0183 train_loss= 0.44734 train_acc= 0.78759 val_loss= 0.72214 val_acc= 0.67887 test_acc= 0.69415 time= 5.08965 Epoch: 0184 train_loss= 0.44845 train_acc= 0.78556 val_loss= 0.70294 val_acc= 0.68873 test_acc= 0.70146 time= 4.97769 Epoch: 0185 train_loss= 0.44025 train_acc= 0.78665 val_loss= 0.69748 val_acc= 0.69296 test_acc= 0.70287 time= 4.99449 Epoch: 0186 train_loss= 0.44616 train_acc= 0.78478 val_loss= 0.72333 val_acc= 0.67887 test_acc= 0.70343 time= 4.94402 Epoch: 0187 train_loss= 0.44732 train_acc= 0.78478 val_loss= 0.69303 val_acc= 0.68592 test_acc= 0.70709 time= 4.97866 Epoch: 0188 train_loss= 0.43935 train_acc= 0.79118 val_loss= 0.71342 val_acc= 0.68873 test_acc= 0.70174 time= 4.96473 Epoch: 0189 train_loss= 0.44050 train_acc= 0.79056 val_loss= 0.72025 val_acc= 0.68451 test_acc= 0.70315 time= 4.97172 Epoch: 0190 train_loss= 0.43303 train_acc= 0.79337 val_loss= 0.73369 val_acc= 0.69014 test_acc= 0.70371 time= 5.01559 Epoch: 0191 train_loss= 0.43840 train_acc= 0.79400 val_loss= 0.73038 val_acc= 0.67746 test_acc= 0.70118 time= 5.16619 Epoch: 0192 train_loss= 0.43250 train_acc= 0.79712 val_loss= 0.71148 val_acc= 0.68169 test_acc= 0.70231 time= 5.29484 Epoch: 0193 train_loss= 0.42585 train_acc= 0.80103 val_loss= 0.73560 val_acc= 0.68310 test_acc= 0.69977 time= 5.09471 Epoch: 0194 train_loss= 0.42801 train_acc= 0.79619 val_loss= 0.71638 val_acc= 0.68451 test_acc= 0.69977 time= 4.97954 Epoch: 0195 train_loss= 0.42988 train_acc= 0.79587 val_loss= 0.72307 val_acc= 0.68169 test_acc= 0.70174 time= 4.99864 Epoch: 0196 train_loss= 0.42701 train_acc= 0.79384 val_loss= 0.70023 val_acc= 0.68592 test_acc= 0.69921 time= 4.98510 Epoch: 0197 train_loss= 0.42972 train_acc= 0.79353 val_loss= 0.71931 val_acc= 0.68592 test_acc= 0.69865 time= 4.98434 Epoch: 0198 train_loss= 0.43490 train_acc= 0.78915 val_loss= 0.70607 val_acc= 0.68028 test_acc= 0.70597 time= 4.97593 Epoch: 0199 train_loss= 0.41755 train_acc= 0.79884 val_loss= 0.71773 val_acc= 0.68169 test_acc= 0.70737 time= 4.95259 Epoch: 0200 train_loss= 0.41566 train_acc= 0.80025 val_loss= 0.72945 val_acc= 0.68169 test_acc= 0.70625 time= 4.98809 Epoch: 0201 train_loss= 0.41714 train_acc= 0.80181 val_loss= 0.74671 val_acc= 0.68310 test_acc= 0.70934 time= 4.99951 Epoch: 0202 train_loss= 0.40958 train_acc= 0.80478 val_loss= 0.72055 val_acc= 0.69014 test_acc= 0.70484 time= 4.95974 Epoch: 0203 train_loss= 0.41111 train_acc= 0.80541 val_loss= 0.75938 val_acc= 0.67465 test_acc= 0.70343 time= 4.98591 Epoch: 0204 train_loss= 0.40949 train_acc= 0.80556 val_loss= 0.75863 val_acc= 0.69014 test_acc= 0.70203 time= 4.97869 Epoch: 0205 train_loss= 0.41858 train_acc= 0.80916 val_loss= 0.73022 val_acc= 0.69014 test_acc= 0.70597 time= 4.99465 Epoch: 0206 train_loss= 0.41316 train_acc= 0.80510 val_loss= 0.71530 val_acc= 0.69577 test_acc= 0.70006 time= 5.08947 Epoch: 0207 train_loss= 0.41662 train_acc= 0.80775 val_loss= 0.76140 val_acc= 0.68873 test_acc= 0.69809 time= 4.97380 Epoch: 0208 train_loss= 0.41664 train_acc= 0.80306 val_loss= 0.74439 val_acc= 0.68310 test_acc= 0.70287 time= 5.03718 Epoch: 0209 train_loss= 0.39808 train_acc= 0.81572 val_loss= 0.73347 val_acc= 0.69014 test_acc= 0.70878 time= 4.95867 Epoch: 0210 train_loss= 0.39934 train_acc= 0.81229 val_loss= 0.74146 val_acc= 0.69155 test_acc= 0.70822 time= 4.99159 Epoch: 0211 train_loss= 0.40220 train_acc= 0.81166 val_loss= 0.73373 val_acc= 0.68310 test_acc= 0.70681 time= 4.99915 Epoch: 0212 train_loss= 0.40562 train_acc= 0.81385 val_loss= 0.72305 val_acc= 0.68310 test_acc= 0.70343 time= 4.96572 Epoch: 0213 train_loss= 0.40094 train_acc= 0.81416 val_loss= 0.72515 val_acc= 0.69437 test_acc= 0.70540 time= 4.99764 Epoch: 0214 train_loss= 0.40088 train_acc= 0.81150 val_loss= 0.72188 val_acc= 0.69155 test_acc= 0.70653 time= 4.98059 Epoch: 0215 train_loss= 0.39861 train_acc= 0.81322 val_loss= 0.74067 val_acc= 0.69859 test_acc= 0.69781 time= 4.99095 Epoch: 0216 train_loss= 0.41137 train_acc= 0.80978 val_loss= 0.71737 val_acc= 0.66620 test_acc= 0.70737 time= 5.00592 Epoch: 0217 train_loss= 0.40234 train_acc= 0.81197 val_loss= 0.71265 val_acc= 0.69296 test_acc= 0.71159 time= 4.93235 Epoch: 0218 train_loss= 0.41343 train_acc= 0.80603 val_loss= 0.70428 val_acc= 0.68310 test_acc= 0.71272 time= 4.98467 Epoch: 0219 train_loss= 0.39096 train_acc= 0.81932 val_loss= 0.72731 val_acc= 0.69718 test_acc= 0.71159 time= 5.09538 Epoch: 0220 train_loss= 0.39463 train_acc= 0.82010 val_loss= 0.74144 val_acc= 0.69155 test_acc= 0.70850 time= 5.02708 Epoch: 0221 train_loss= 0.38607 train_acc= 0.82651 val_loss= 0.75153 val_acc= 0.68451 test_acc= 0.71159 time= 4.96146 Epoch: 0222 train_loss= 0.39322 train_acc= 0.81760 val_loss= 0.72800 val_acc= 0.68310 test_acc= 0.70709 time= 4.97271 Epoch: 0223 train_loss= 0.38014 train_acc= 0.82369 val_loss= 0.73129 val_acc= 0.68873 test_acc= 0.71075 time= 4.98902 Epoch: 0224 train_loss= 0.38032 train_acc= 0.82807 val_loss= 0.73083 val_acc= 0.68451 test_acc= 0.70990 time= 4.97375 Epoch: 0225 train_loss= 0.38770 train_acc= 0.81932 val_loss= 0.73536 val_acc= 0.69296 test_acc= 0.70822 time= 4.96573 Epoch: 0226 train_loss= 0.38128 train_acc= 0.82323 val_loss= 0.73027 val_acc= 0.68028 test_acc= 0.71047 time= 5.05168 Epoch: 0227 train_loss= 0.38259 train_acc= 0.82213 val_loss= 0.74099 val_acc= 0.68732 test_acc= 0.70934 time= 4.97670 Epoch: 0228 train_loss= 0.38452 train_acc= 0.82870 val_loss= 0.74915 val_acc= 0.68873 test_acc= 0.71244 time= 5.00163 Epoch: 0229 train_loss= 0.38967 train_acc= 0.82323 val_loss= 0.74114 val_acc= 0.69014 test_acc= 0.70793 time= 4.96084 Epoch: 0230 train_loss= 0.37379 train_acc= 0.83057 val_loss= 0.75612 val_acc= 0.69859 test_acc= 0.70934 time= 5.05948 Epoch: 0231 train_loss= 0.37851 train_acc= 0.82604 val_loss= 0.73410 val_acc= 0.67606 test_acc= 0.70034 time= 5.05633 Epoch: 0232 train_loss= 0.37561 train_acc= 0.82776 val_loss= 0.75393 val_acc= 0.68592 test_acc= 0.71216 time= 5.00786 Epoch: 0233 train_loss= 0.37317 train_acc= 0.83042 val_loss= 0.76086 val_acc= 0.68732 test_acc= 0.71328 time= 4.99045 Epoch: 0234 train_loss= 0.37371 train_acc= 0.83182 val_loss= 0.74519 val_acc= 0.69155 test_acc= 0.71047 time= 5.04354 Epoch: 0235 train_loss= 0.37036 train_acc= 0.83088 val_loss= 0.75217 val_acc= 0.69577 test_acc= 0.71666 time= 4.98900 Epoch: 0236 train_loss= 0.37925 train_acc= 0.82776 val_loss= 0.74192 val_acc= 0.70141 test_acc= 0.70962 time= 5.02807 Epoch: 0237 train_loss= 0.36757 train_acc= 0.83854 val_loss= 0.76673 val_acc= 0.69155 test_acc= 0.70906 time= 4.99865 Epoch: 0238 train_loss= 0.37365 train_acc= 0.83073 val_loss= 0.75775 val_acc= 0.68310 test_acc= 0.70625 time= 5.01260 Epoch: 0239 train_loss= 0.37368 train_acc= 0.82588 val_loss= 0.74376 val_acc= 0.69718 test_acc= 0.70765 time= 4.98861 Epoch: 0240 train_loss= 0.36138 train_acc= 0.83682 val_loss= 0.74463 val_acc= 0.69296 test_acc= 0.71553 time= 5.00717 Epoch: 0241 train_loss= 0.36988 train_acc= 0.83292 val_loss= 0.75183 val_acc= 0.69155 test_acc= 0.71581 time= 4.99652 Epoch: 0242 train_loss= 0.36990 train_acc= 0.82995 val_loss= 0.76329 val_acc= 0.68310 test_acc= 0.71469 time= 4.99119 Epoch: 0243 train_loss= 0.36588 train_acc= 0.83667 val_loss= 0.74275 val_acc= 0.69718 test_acc= 0.70456 time= 4.98871 Epoch: 0244 train_loss= 0.37020 train_acc= 0.83010 val_loss= 0.75178 val_acc= 0.69296 test_acc= 0.71131 time= 5.00562 Epoch: 0245 train_loss= 0.36515 train_acc= 0.83542 val_loss= 0.75577 val_acc= 0.69155 test_acc= 0.70878 time= 4.97261 Epoch: 0246 train_loss= 0.36080 train_acc= 0.83776 val_loss= 0.76941 val_acc= 0.69296 test_acc= 0.71244 time= 4.97107 Epoch: 0247 train_loss= 0.36693 train_acc= 0.83448 val_loss= 0.73072 val_acc= 0.69859 test_acc= 0.70878 time= 5.03037 Epoch: 0248 train_loss= 0.34992 train_acc= 0.84417 val_loss= 0.79395 val_acc= 0.70141 test_acc= 0.70653 time= 4.99265 Epoch: 0249 train_loss= 0.35761 train_acc= 0.83854 val_loss= 0.75043 val_acc= 0.70282 test_acc= 0.71187 time= 5.02531 Epoch: 0250 train_loss= 0.36541 train_acc= 0.83854 val_loss= 0.74052 val_acc= 0.69437 test_acc= 0.70456 time= 4.99217 Epoch: 0251 train_loss= 0.35093 train_acc= 0.84495 val_loss= 0.77135 val_acc= 0.69859 test_acc= 0.70906 time= 4.98319 Epoch: 0252 train_loss= 0.35845 train_acc= 0.83839 val_loss= 0.75024 val_acc= 0.69437 test_acc= 0.70765 time= 4.97015 Epoch: 0253 train_loss= 0.35670 train_acc= 0.84151 val_loss= 0.76338 val_acc= 0.70282 test_acc= 0.70737 time= 5.02670 Epoch: 0254 train_loss= 0.34835 train_acc= 0.83948 val_loss= 0.76086 val_acc= 0.70704 test_acc= 0.71244 time= 5.01911 Epoch: 0255 train_loss= 0.35931 train_acc= 0.83870 val_loss= 0.74746 val_acc= 0.69859 test_acc= 0.71609 time= 4.98206 Epoch: 0256 train_loss= 0.34308 train_acc= 0.84261 val_loss= 0.76793 val_acc= 0.70563 test_acc= 0.71947 time= 5.01887 Epoch: 0257 train_loss= 0.35208 train_acc= 0.84276 val_loss= 0.78378 val_acc= 0.70000 test_acc= 0.71216 time= 5.07395 Epoch: 0258 train_loss= 0.35098 train_acc= 0.84558 val_loss= 0.74788 val_acc= 0.70986 test_acc= 0.71694 time= 5.03354 Epoch: 0259 train_loss= 0.34421 train_acc= 0.84823 val_loss= 0.77456 val_acc= 0.70282 test_acc= 0.71835 time= 4.99489 Epoch: 0260 train_loss= 0.34757 train_acc= 0.84745 val_loss= 0.75748 val_acc= 0.69014 test_acc= 0.71750 time= 4.98999 Epoch: 0261 train_loss= 0.34313 train_acc= 0.84433 val_loss= 0.78820 val_acc= 0.70563 test_acc= 0.71778 time= 5.03606 Epoch: 0262 train_loss= 0.34806 train_acc= 0.84386 val_loss= 0.74050 val_acc= 0.71268 test_acc= 0.71750 time= 4.96273 Epoch: 0263 train_loss= 0.34410 train_acc= 0.84292 val_loss= 0.77108 val_acc= 0.70986 test_acc= 0.72060 time= 5.03164 Epoch: 0264 train_loss= 0.33754 train_acc= 0.85214 val_loss= 0.76244 val_acc= 0.72535 test_acc= 0.71919 time= 5.02158 Epoch: 0265 train_loss= 0.34625 train_acc= 0.84761 val_loss= 0.74414 val_acc= 0.71268 test_acc= 0.71525 time= 5.01659 Epoch: 0266 train_loss= 0.32765 train_acc= 0.85511 val_loss= 0.73852 val_acc= 0.72113 test_acc= 0.72200 time= 5.03354 Epoch: 0267 train_loss= 0.34910 train_acc= 0.84261 val_loss= 0.72343 val_acc= 0.70845 test_acc= 0.71750 time= 5.03354 Epoch: 0268 train_loss= 0.33532 train_acc= 0.84776 val_loss= 0.76488 val_acc= 0.71549 test_acc= 0.71806 time= 5.03055 Epoch: 0269 train_loss= 0.32657 train_acc= 0.85261 val_loss= 0.78776 val_acc= 0.70704 test_acc= 0.71131 time= 4.97796 Epoch: 0270 train_loss= 0.34652 train_acc= 0.84276 val_loss= 0.74937 val_acc= 0.70141 test_acc= 0.71750 time= 5.00198 Epoch: 0271 train_loss= 0.33882 train_acc= 0.85089 val_loss= 0.76570 val_acc= 0.71549 test_acc= 0.71806 time= 5.00761 Epoch: 0272 train_loss= 0.33545 train_acc= 0.84776 val_loss= 0.79784 val_acc= 0.70704 test_acc= 0.71356 time= 4.95475 Epoch: 0273 train_loss= 0.33411 train_acc= 0.84714 val_loss= 0.77424 val_acc= 0.71127 test_acc= 0.71525 time= 4.98905 Epoch: 0274 train_loss= 0.33638 train_acc= 0.85152 val_loss= 0.77830 val_acc= 0.71268 test_acc= 0.71778 time= 5.03374 Epoch: 0275 train_loss= 0.32634 train_acc= 0.85636 val_loss= 0.78228 val_acc= 0.70704 test_acc= 0.71159 time= 5.00214 Epoch: 0276 train_loss= 0.33080 train_acc= 0.85871 val_loss= 0.79556 val_acc= 0.70845 test_acc= 0.71525 time= 4.99417 Epoch: 0277 train_loss= 0.33437 train_acc= 0.85027 val_loss= 0.76795 val_acc= 0.71268 test_acc= 0.71300 time= 5.01273 Epoch: 0278 train_loss= 0.33273 train_acc= 0.85120 val_loss= 0.80859 val_acc= 0.70704 test_acc= 0.71497 time= 5.00562 Epoch: 0279 train_loss= 0.32521 train_acc= 0.85746 val_loss= 0.80111 val_acc= 0.70282 test_acc= 0.71497 time= 4.97071 Epoch: 0280 train_loss= 0.32394 train_acc= 0.85542 val_loss= 0.79826 val_acc= 0.70282 test_acc= 0.71187 time= 4.99963 Epoch: 0281 train_loss= 0.31955 train_acc= 0.85933 val_loss= 0.81747 val_acc= 0.72113 test_acc= 0.71750 time= 5.00562 Epoch: 0282 train_loss= 0.32885 train_acc= 0.86199 val_loss= 0.78439 val_acc= 0.71408 test_acc= 0.71300 time= 4.99764 Epoch: 0283 train_loss= 0.33212 train_acc= 0.85449 val_loss= 0.78911 val_acc= 0.70423 test_acc= 0.70906 time= 5.04252 Epoch: 0284 train_loss= 0.31630 train_acc= 0.85433 val_loss= 0.79538 val_acc= 0.71690 test_acc= 0.71525 time= 5.00301 Epoch: 0285 train_loss= 0.32232 train_acc= 0.85996 val_loss= 0.79981 val_acc= 0.69577 test_acc= 0.71553 time= 4.98567 Epoch: 0286 train_loss= 0.32676 train_acc= 0.85058 val_loss= 0.75683 val_acc= 0.71268 test_acc= 0.71216 time= 5.17485 Epoch: 0287 train_loss= 0.32191 train_acc= 0.85746 val_loss= 0.81409 val_acc= 0.70423 test_acc= 0.71187 time= 5.32049 Epoch: 0288 train_loss= 0.31612 train_acc= 0.86277 val_loss= 0.80182 val_acc= 0.70000 test_acc= 0.71609 time= 5.03054 Epoch: 0289 train_loss= 0.31940 train_acc= 0.86324 val_loss= 0.79205 val_acc= 0.71408 test_acc= 0.71497 time= 5.45430 Epoch: 0290 train_loss= 0.31313 train_acc= 0.86496 val_loss= 0.79241 val_acc= 0.70423 test_acc= 0.71919 time= 5.13028 Epoch: 0291 train_loss= 0.31354 train_acc= 0.86152 val_loss= 0.79377 val_acc= 0.70986 test_acc= 0.71947 time= 5.22304 Epoch: 0292 train_loss= 0.30692 train_acc= 0.86621 val_loss= 0.79869 val_acc= 0.71549 test_acc= 0.71356 time= 5.15921 Epoch: 0293 train_loss= 0.32157 train_acc= 0.86324 val_loss= 0.80658 val_acc= 0.69859 test_acc= 0.71328 time= 4.98767 Epoch: 0294 train_loss= 0.30807 train_acc= 0.86511 val_loss= 0.81952 val_acc= 0.70704 test_acc= 0.71356 time= 5.10336 Epoch: 0295 train_loss= 0.31901 train_acc= 0.86152 val_loss= 0.80780 val_acc= 0.69859 test_acc= 0.71047 time= 5.12929 Epoch: 0296 train_loss= 0.32257 train_acc= 0.85621 val_loss= 0.78455 val_acc= 0.70845 test_acc= 0.72228 time= 5.55216 Epoch: 0297 train_loss= 0.30977 train_acc= 0.86480 val_loss= 0.80951 val_acc= 0.70282 test_acc= 0.71244 time= 5.42651 Epoch: 0298 train_loss= 0.31464 train_acc= 0.85558 val_loss= 0.80726 val_acc= 0.70845 test_acc= 0.71581 time= 5.08067 Epoch: 0299 train_loss= 0.30763 train_acc= 0.86558 val_loss= 0.83737 val_acc= 0.71268 test_acc= 0.71328 time= 5.05651 Epoch: 0300 train_loss= 0.31558 train_acc= 0.86558 val_loss= 0.79260 val_acc= 0.70000 test_acc= 0.71244 time= 5.31176 Optimization Finished! Best epoch: 263 Test set results: cost= 0.74827 accuracy= 0.71919 Test Precision, Recall and F1-Score... precision recall f1-score support
avg / total 0.7209 0.7192 0.7186 3554
Macro average Test Precision, Recall and F1-Score... (0.7209362974880018, 0.7191896454698932, 0.7186335470736362, None) Micro average Test Precision, Recall and F1-Score... (0.7191896454698931, 0.7191896454698931, 0.7191896454698931, None)