notebook.community

Edit and run 





In [1]:



    


import warnings
warnings.filterwarnings('ignore')



    











In [2]:



    


%matplotlib inline
%pylab inline



    


















    






Populating the interactive namespace from numpy and matplotlib

















In [3]:



    


import matplotlib.pylab as plt
# https://docs.scipy.org/doc/numpy/reference/routines.math.html
import numpy as np



    











In [4]:



    


from datetime import tzinfo, timedelta, datetime



    











In [5]:



    


from distutils.version import StrictVersion



    











In [6]:



    


import sklearn

assert StrictVersion(sklearn.__version__ ) >= StrictVersion('0.18.1')

sklearn.__version__



    


















    
Out[6]:








'0.18.1'

















In [7]:



    


import tensorflow as tf
tf.logging.set_verbosity(tf.logging.ERROR)

assert StrictVersion(tf.__version__) >= StrictVersion('1.1.0')

tf.__version__



    


















    
Out[7]:








'1.2.1'

















In [8]:



    


import keras

assert StrictVersion(keras.__version__) >= StrictVersion('2.0.0')

keras.__version__



    


















    






Using TensorFlow backend.










    
Out[8]:








'2.0.6'

















In [9]:



    


# !curl -O https://raw.githubusercontent.com/DJCordhose/speed-limit-signs/master/data/speed-limit-signs.zip
# !curl -O https://raw.githubusercontent.com/DJCordhose/speed-limit-signs/master/data/augmented-signs.zip



    











In [10]:



    


# https://docs.python.org/3/library/zipfile.html
# from zipfile import ZipFile
# zip = ZipFile(r'speed-limit-signs.zip')
# zip.extractall('.')
# zip = ZipFile(r'augmented-signs.zip')
# zip.extractall('.')



    











In [11]:



    


# !ls -l speed-limit-signs



    











In [12]:



    


# !ls -l augmented-signs



    











In [13]:



    


import os
import skimage.data
import skimage.transform
from keras.utils.np_utils import to_categorical
import numpy as np

def load_data(data_dir, type=".ppm"):
    num_categories = 6

    # Get all subdirectories of data_dir. Each represents a label.
    directories = [d for d in os.listdir(data_dir) 
                   if os.path.isdir(os.path.join(data_dir, d))]
    # Loop through the label directories and collect the data in
    # two lists, labels and images.
    labels = []
    images = []
    for d in directories:
        label_dir = os.path.join(data_dir, d)
        file_names = [os.path.join(label_dir, f) for f in os.listdir(label_dir) if f.endswith(type)]
        # For each label, load it's images and add them to the images list.
        # And add the label number (i.e. directory name) to the labels list.
        for f in file_names:
            images.append(skimage.data.imread(f))
            labels.append(int(d))
    images64 = [skimage.transform.resize(image, (64, 64)) for image in images]
    y = np.array(labels)
    y = to_categorical(y, num_categories)
    X = np.array(images64)
    return X, y



    











In [33]:



    


# Depends on harware GPU architecture, set as high as possible (this works well on K80)
BATCH_SIZE = 500



    











In [14]:



    


# Load datasets.
ROOT_PATH = "./"



    











In [15]:



    


original_dir = os.path.join(ROOT_PATH, "speed-limit-signs")
original_images, original_labels = load_data(original_dir, type=".ppm")



    











In [16]:



    


data_dir = os.path.join(ROOT_PATH, "augmented-signs")
augmented_images, augmented_labels = load_data(data_dir, type=".png")



    











In [40]:



    


# Turn this around if you want the large training set using augmented data or the original one

# X, y = original_images, original_labels
X, y = augmented_images, augmented_labels



    











In [41]:



    


# Same as above

# evaluation_X, evaluation_y = augmented_images, augmented_labels
evaluation_X, evaluation_y = original_images, original_labels



    











In [19]:



    


from sklearn.model_selection import train_test_split



    











In [20]:



    


checkpoint_callback = keras.callbacks.ModelCheckpoint('../tmp/model-checkpoints/weights.epoch-{epoch:02d}-val_loss-{val_loss:.2f}.hdf5');



    











In [42]:



    


early_stopping_callback = keras.callbacks.EarlyStopping(monitor='val_loss', patience=30, verbose=1)



    











In [43]:



    


# https://github.com/tensorflow/tensorflow/blob/master/tensorflow/tensorboard/README.md
# https://keras.io/callbacks/#tensorboard
# http://stackoverflow.com/questions/42112260/how-do-i-use-the-tensorboard-callback-of-keras
tb_callback = keras.callbacks.TensorBoard(log_dir='../tmp/tf_log')
#                                          histogram_freq=1, write_graph=True, write_images=True)
#                                          histogram_freq=1, write_graph=True, write_images=True)
# tbCallBack = keras.callbacks.TensorBoard(log_dir='./logs', histogram_freq=0, batch_size=32, write_graph=True, write_grads=False, write_images=False, embeddings_freq=0, embeddings_layer_names=None, embeddings_metadata=None)
# To start tensorboard
# tensorboard --logdir=/mnt/c/Users/olive/Development/ml/tf_log
# open http://localhost:6006



    











In [44]:



    


# we want to distribute our different classes equally over test and train, this works using stratify
# https://github.com/amueller/scipy-2017-sklearn/blob/master/notebooks/04.Training_and_Testing_Data.ipynb
# http://scikit-learn.org/stable/modules/generated/sklearn.model_selection.train_test_split.html

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42, stratify=y, )



    











In [45]:



    


X_train.shape, y_train.shape



    


















    
Out[45]:








((3032, 64, 64, 3), (3032, 6))

















In [46]:



    


from keras.models import Model
from keras.layers import Dense, Dropout, Activation, Flatten, Input
from keras.layers import Convolution2D, MaxPooling2D

# drop_out = 0.9
# drop_out = 0.75
drop_out = 0.5
# drop_out = 0.25
# drop_out = 0.0

# input tensor for a 3-channel 64x64 image
inputs = Input(shape=(64, 64, 3))

# one block of convolutional layers
x = Convolution2D(64, 3, 3, activation='relu')(inputs)
# x = Dropout(drop_out)(x)
x = Convolution2D(64, 3, 3, activation='relu')(x)
# x = Dropout(drop_out)(x)
x = Convolution2D(64, 3, 3, activation='relu')(x)
x = MaxPooling2D(pool_size=(2, 2))(x)
x = Dropout(drop_out)(x)

# one more block
x = Convolution2D(128, 3, 3, activation='relu')(x)
# x = Dropout(drop_out)(x)
x = Convolution2D(128, 3, 3, activation='relu')(x)
x = MaxPooling2D(pool_size=(2, 2))(x)
x = Dropout(drop_out)(x)

# one more block
x = Convolution2D(256, 3, 3, activation='relu')(x)
x = MaxPooling2D(pool_size=(2, 2))(x)
x = Dropout(drop_out)(x)

x = Flatten()(x)
x = Dense(256, activation='relu')(x)
x = Dropout(drop_out)(x)

# softmax activation, 6 categories
predictions = Dense(6, activation='softmax')(x)
model = Model(input=inputs, output=predictions)
model.summary()
model.compile(optimizer='rmsprop',
              loss='categorical_crossentropy',
              metrics=['accuracy'])



    


















    






_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
input_3 (InputLayer)         (None, 64, 64, 3)         0         
_________________________________________________________________
conv2d_13 (Conv2D)           (None, 62, 62, 64)        1792      
_________________________________________________________________
conv2d_14 (Conv2D)           (None, 60, 60, 64)        36928     
_________________________________________________________________
conv2d_15 (Conv2D)           (None, 58, 58, 64)        36928     
_________________________________________________________________
max_pooling2d_7 (MaxPooling2 (None, 29, 29, 64)        0         
_________________________________________________________________
dropout_9 (Dropout)          (None, 29, 29, 64)        0         
_________________________________________________________________
conv2d_16 (Conv2D)           (None, 27, 27, 128)       73856     
_________________________________________________________________
conv2d_17 (Conv2D)           (None, 25, 25, 128)       147584    
_________________________________________________________________
max_pooling2d_8 (MaxPooling2 (None, 12, 12, 128)       0         
_________________________________________________________________
dropout_10 (Dropout)         (None, 12, 12, 128)       0         
_________________________________________________________________
conv2d_18 (Conv2D)           (None, 10, 10, 256)       295168    
_________________________________________________________________
max_pooling2d_9 (MaxPooling2 (None, 5, 5, 256)         0         
_________________________________________________________________
dropout_11 (Dropout)         (None, 5, 5, 256)         0         
_________________________________________________________________
flatten_3 (Flatten)          (None, 6400)              0         
_________________________________________________________________
dense_5 (Dense)              (None, 256)               1638656   
_________________________________________________________________
dropout_12 (Dropout)         (None, 256)               0         
_________________________________________________________________
dense_6 (Dense)              (None, 6)                 1542      
=================================================================
Total params: 2,232,454
Trainable params: 2,232,454
Non-trainable params: 0
_________________________________________________________________

















In [47]:



    


!rm -rf ../tmp/tf_log
!rm -rf ../tmp/model-checkpoints

!mkdir ../tmp/model-checkpoints
!mkdir ../tmp/tf_log



    











In [ ]:



    


# Running on a GPU bach size might be critical depdendng on the GPU memory available
# more is desirable, but we might end up using 50 only 
print(datetime.utcnow().isoformat())
# BE CAREFUL, validation data is always the last data sets and not shuffled
# https://keras.io/getting-started/faq/#how-is-the-validation-split-computed
model.fit(X_train, y_train, epochs=1000, batch_size=BATCH_SIZE, validation_split=0.3, 
#           callbacks=[tb_callback, early_stopping_callback])
          callbacks=[tb_callback])
# model.fit(X_train, y_train, epochs=50, batch_size=200, validation_split=0.3)
print(datetime.utcnow().isoformat())



    


















    






2017-07-31T08:01:02.689115
Train on 2122 samples, validate on 910 samples
Epoch 1/1000
2122/2122 [==============================] - 11s - loss: 2.0733 - acc: 0.1786 - val_loss: 1.7863 - val_acc: 0.1714
Epoch 2/1000
2122/2122 [==============================] - 5s - loss: 1.7719 - acc: 0.1989 - val_loss: 1.7685 - val_acc: 0.2143
Epoch 3/1000
2122/2122 [==============================] - 5s - loss: 1.7892 - acc: 0.2092 - val_loss: 1.7748 - val_acc: 0.2143
Epoch 4/1000
2122/2122 [==============================] - 5s - loss: 1.7691 - acc: 0.2022 - val_loss: 1.7801 - val_acc: 0.2143
Epoch 5/1000
2122/2122 [==============================] - 5s - loss: 1.7605 - acc: 0.2031 - val_loss: 1.7649 - val_acc: 0.2121
Epoch 6/1000
2122/2122 [==============================] - 5s - loss: 1.7514 - acc: 0.2050 - val_loss: 1.7859 - val_acc: 0.1714
Epoch 7/1000
2122/2122 [==============================] - 5s - loss: 1.7761 - acc: 0.1956 - val_loss: 1.7741 - val_acc: 0.2121
Epoch 8/1000
2122/2122 [==============================] - 5s - loss: 1.7623 - acc: 0.2102 - val_loss: 1.7838 - val_acc: 0.2143
Epoch 9/1000
2122/2122 [==============================] - 4s - loss: 1.7676 - acc: 0.2220 - val_loss: 1.7786 - val_acc: 0.1714
Epoch 10/1000
2122/2122 [==============================] - 4s - loss: 1.7640 - acc: 0.2125 - val_loss: 1.7511 - val_acc: 0.2143
Epoch 11/1000
2122/2122 [==============================] - 5s - loss: 1.7726 - acc: 0.2196 - val_loss: 1.7410 - val_acc: 0.2132
Epoch 12/1000
2122/2122 [==============================] - 5s - loss: 1.7555 - acc: 0.2281 - val_loss: 1.7841 - val_acc: 0.1714
Epoch 13/1000
2122/2122 [==============================] - 4s - loss: 1.7686 - acc: 0.2234 - val_loss: 1.7807 - val_acc: 0.1846
Epoch 14/1000
2122/2122 [==============================] - 4s - loss: 1.7639 - acc: 0.2135 - val_loss: 1.7516 - val_acc: 0.2143
Epoch 15/1000
2122/2122 [==============================] - 4s - loss: 1.8463 - acc: 0.2342 - val_loss: 1.7482 - val_acc: 0.2143
Epoch 16/1000
2122/2122 [==============================] - 4s - loss: 1.7414 - acc: 0.2267 - val_loss: 1.7888 - val_acc: 0.1527
Epoch 17/1000
2122/2122 [==============================] - 5s - loss: 1.7522 - acc: 0.2465 - val_loss: 1.7079 - val_acc: 0.2659
Epoch 18/1000
2122/2122 [==============================] - 4s - loss: 1.8638 - acc: 0.2399 - val_loss: 1.7745 - val_acc: 0.1681
Epoch 19/1000
2122/2122 [==============================] - 5s - loss: 1.7426 - acc: 0.2427 - val_loss: 1.7738 - val_acc: 0.2099
Epoch 20/1000
2122/2122 [==============================] - 5s - loss: 1.7258 - acc: 0.2729 - val_loss: 1.7125 - val_acc: 0.2549
Epoch 21/1000
2122/2122 [==============================] - 5s - loss: 1.6884 - acc: 0.2733 - val_loss: 1.7324 - val_acc: 0.2198
Epoch 22/1000
2122/2122 [==============================] - 5s - loss: 1.7356 - acc: 0.2196 - val_loss: 1.7293 - val_acc: 0.2209
Epoch 23/1000
2122/2122 [==============================] - 5s - loss: 1.6921 - acc: 0.2686 - val_loss: 1.6968 - val_acc: 0.3110
Epoch 24/1000
2122/2122 [==============================] - 5s - loss: 1.7153 - acc: 0.2573 - val_loss: 1.6833 - val_acc: 0.2868
Epoch 25/1000
2122/2122 [==============================] - 5s - loss: 1.6696 - acc: 0.2813 - val_loss: 1.6582 - val_acc: 0.2945
Epoch 26/1000
2122/2122 [==============================] - 5s - loss: 1.6627 - acc: 0.2804 - val_loss: 1.7625 - val_acc: 0.2088
Epoch 27/1000
2122/2122 [==============================] - 5s - loss: 1.7794 - acc: 0.2488 - val_loss: 1.7151 - val_acc: 0.2538
Epoch 28/1000
2122/2122 [==============================] - 5s - loss: 1.6361 - acc: 0.3186 - val_loss: 1.6741 - val_acc: 0.2934
Epoch 29/1000
2122/2122 [==============================] - 5s - loss: 1.7005 - acc: 0.2974 - val_loss: 1.7424 - val_acc: 0.2593
Epoch 30/1000
2122/2122 [==============================] - 5s - loss: 1.7009 - acc: 0.2771 - val_loss: 1.6733 - val_acc: 0.2923
Epoch 31/1000
2122/2122 [==============================] - 5s - loss: 1.6734 - acc: 0.2959 - val_loss: 1.6733 - val_acc: 0.2681
Epoch 32/1000
2122/2122 [==============================] - 5s - loss: 1.6731 - acc: 0.2870 - val_loss: 1.7102 - val_acc: 0.2560
Epoch 33/1000
2122/2122 [==============================] - 5s - loss: 1.6351 - acc: 0.3153 - val_loss: 1.6647 - val_acc: 0.2934
Epoch 34/1000
2122/2122 [==============================] - 5s - loss: 1.6929 - acc: 0.2879 - val_loss: 1.6734 - val_acc: 0.3033
Epoch 35/1000
2122/2122 [==============================] - 5s - loss: 1.6160 - acc: 0.3341 - val_loss: 1.7149 - val_acc: 0.2484
Epoch 36/1000
2122/2122 [==============================] - 5s - loss: 1.6264 - acc: 0.3157 - val_loss: 1.6419 - val_acc: 0.3022
Epoch 37/1000
2122/2122 [==============================] - 5s - loss: 1.6406 - acc: 0.3087 - val_loss: 1.6568 - val_acc: 0.2637
Epoch 38/1000
2122/2122 [==============================] - 5s - loss: 1.6336 - acc: 0.2983 - val_loss: 1.6273 - val_acc: 0.2747
Epoch 39/1000
2122/2122 [==============================] - 5s - loss: 1.5541 - acc: 0.3563 - val_loss: 1.6117 - val_acc: 0.3297
Epoch 40/1000
2122/2122 [==============================] - 5s - loss: 1.6701 - acc: 0.2790 - val_loss: 1.6939 - val_acc: 0.2824
Epoch 41/1000
2122/2122 [==============================] - 5s - loss: 1.6069 - acc: 0.3402 - val_loss: 1.6290 - val_acc: 0.3099
Epoch 42/1000
2122/2122 [==============================] - 5s - loss: 1.6464 - acc: 0.3148 - val_loss: 1.6152 - val_acc: 0.3374
Epoch 43/1000
2122/2122 [==============================] - 5s - loss: 1.5801 - acc: 0.3369 - val_loss: 1.6672 - val_acc: 0.2824
Epoch 44/1000
2122/2122 [==============================] - 5s - loss: 1.5991 - acc: 0.3421 - val_loss: 1.6162 - val_acc: 0.3143
Epoch 45/1000
2122/2122 [==============================] - 5s - loss: 1.5678 - acc: 0.3605 - val_loss: 1.7133 - val_acc: 0.2780
Epoch 46/1000
2122/2122 [==============================] - 5s - loss: 1.6607 - acc: 0.3157 - val_loss: 1.6462 - val_acc: 0.3154
Epoch 47/1000
2122/2122 [==============================] - 5s - loss: 1.5687 - acc: 0.3671 - val_loss: 1.6336 - val_acc: 0.3385
Epoch 48/1000
2122/2122 [==============================] - 5s - loss: 1.5068 - acc: 0.3935 - val_loss: 1.6595 - val_acc: 0.3110
Epoch 49/1000
2122/2122 [==============================] - 5s - loss: 1.5543 - acc: 0.3699 - val_loss: 1.6562 - val_acc: 0.3132
Epoch 50/1000
2122/2122 [==============================] - 5s - loss: 1.4890 - acc: 0.3907 - val_loss: 1.7138 - val_acc: 0.3319
Epoch 51/1000
2122/2122 [==============================] - 5s - loss: 1.7004 - acc: 0.3139 - val_loss: 1.6428 - val_acc: 0.3176
Epoch 52/1000
2122/2122 [==============================] - 5s - loss: 1.5374 - acc: 0.3709 - val_loss: 1.5409 - val_acc: 0.3571
Epoch 53/1000
2122/2122 [==============================] - 5s - loss: 1.4809 - acc: 0.3954 - val_loss: 1.5962 - val_acc: 0.3736
Epoch 54/1000
2122/2122 [==============================] - 5s - loss: 1.4298 - acc: 0.4345 - val_loss: 1.6113 - val_acc: 0.3725
Epoch 55/1000
2122/2122 [==============================] - 5s - loss: 1.5906 - acc: 0.3662 - val_loss: 1.5895 - val_acc: 0.3659
Epoch 56/1000
2122/2122 [==============================] - 5s - loss: 1.3905 - acc: 0.4472 - val_loss: 1.5955 - val_acc: 0.3330
Epoch 57/1000
2122/2122 [==============================] - 5s - loss: 1.4354 - acc: 0.4369 - val_loss: 1.6889 - val_acc: 0.2714
Epoch 58/1000
2122/2122 [==============================] - 5s - loss: 1.4908 - acc: 0.3987 - val_loss: 1.5237 - val_acc: 0.4011
Epoch 59/1000
2122/2122 [==============================] - 5s - loss: 1.3554 - acc: 0.4680 - val_loss: 1.5650 - val_acc: 0.3484
Epoch 60/1000
2122/2122 [==============================] - 5s - loss: 1.3956 - acc: 0.4538 - val_loss: 1.5337 - val_acc: 0.4000
Epoch 61/1000
2122/2122 [==============================] - 5s - loss: 1.3562 - acc: 0.4713 - val_loss: 1.7518 - val_acc: 0.3714
Epoch 62/1000
2122/2122 [==============================] - 5s - loss: 1.6335 - acc: 0.3770 - val_loss: 1.5066 - val_acc: 0.4176
Epoch 63/1000
2122/2122 [==============================] - 5s - loss: 1.3046 - acc: 0.5005 - val_loss: 1.4179 - val_acc: 0.4484
Epoch 64/1000
2122/2122 [==============================] - 5s - loss: 1.3107 - acc: 0.4991 - val_loss: 1.4374 - val_acc: 0.4538
Epoch 65/1000
2122/2122 [==============================] - 5s - loss: 1.3248 - acc: 0.4840 - val_loss: 1.5045 - val_acc: 0.4000
Epoch 66/1000
2122/2122 [==============================] - 5s - loss: 1.3647 - acc: 0.4623 - val_loss: 1.4641 - val_acc: 0.4363
Epoch 67/1000
2122/2122 [==============================] - 5s - loss: 1.2402 - acc: 0.5156 - val_loss: 1.5779 - val_acc: 0.4011
Epoch 68/1000
2122/2122 [==============================] - 5s - loss: 1.3095 - acc: 0.5028 - val_loss: 1.4768 - val_acc: 0.4396
Epoch 69/1000
2122/2122 [==============================] - 5s - loss: 1.2280 - acc: 0.5353 - val_loss: 1.5805 - val_acc: 0.4198
Epoch 70/1000
2122/2122 [==============================] - 5s - loss: 1.3755 - acc: 0.4661 - val_loss: 1.5133 - val_acc: 0.3769
Epoch 71/1000
2122/2122 [==============================] - 5s - loss: 1.2544 - acc: 0.5207 - val_loss: 1.4487 - val_acc: 0.4495
Epoch 72/1000
2122/2122 [==============================] - 5s - loss: 1.2072 - acc: 0.5410 - val_loss: 1.4196 - val_acc: 0.4484
Epoch 73/1000
2122/2122 [==============================] - 5s - loss: 1.2220 - acc: 0.5344 - val_loss: 1.5881 - val_acc: 0.3736
Epoch 74/1000
2122/2122 [==============================] - 5s - loss: 1.1879 - acc: 0.5462 - val_loss: 1.3030 - val_acc: 0.4967
Epoch 75/1000
2122/2122 [==============================] - 5s - loss: 1.0874 - acc: 0.5796 - val_loss: 1.5210 - val_acc: 0.4033
Epoch 76/1000
2122/2122 [==============================] - 5s - loss: 1.3469 - acc: 0.5000 - val_loss: 1.4409 - val_acc: 0.4780
Epoch 77/1000
2122/2122 [==============================] - 5s - loss: 1.1481 - acc: 0.5688 - val_loss: 1.2886 - val_acc: 0.5055
Epoch 78/1000
2122/2122 [==============================] - 5s - loss: 1.0268 - acc: 0.6122 - val_loss: 1.3051 - val_acc: 0.5187
Epoch 79/1000
2122/2122 [==============================] - 5s - loss: 1.1301 - acc: 0.5740 - val_loss: 1.3162 - val_acc: 0.5220
Epoch 80/1000
2122/2122 [==============================] - 5s - loss: 1.0009 - acc: 0.6183 - val_loss: 1.3805 - val_acc: 0.5066
Epoch 81/1000
2122/2122 [==============================] - 5s - loss: 1.1061 - acc: 0.5990 - val_loss: 1.2421 - val_acc: 0.5516
Epoch 82/1000
2122/2122 [==============================] - 5s - loss: 0.9402 - acc: 0.6480 - val_loss: 1.3101 - val_acc: 0.5275
Epoch 83/1000
2122/2122 [==============================] - 5s - loss: 1.0409 - acc: 0.6117 - val_loss: 1.2847 - val_acc: 0.5593
Epoch 84/1000
2122/2122 [==============================] - 5s - loss: 1.0430 - acc: 0.6051 - val_loss: 1.2639 - val_acc: 0.5637
Epoch 85/1000
2122/2122 [==============================] - 5s - loss: 0.9947 - acc: 0.6367 - val_loss: 1.2323 - val_acc: 0.5363
Epoch 86/1000
2122/2122 [==============================] - 5s - loss: 0.8479 - acc: 0.6833 - val_loss: 1.2744 - val_acc: 0.5714
Epoch 87/1000
2122/2122 [==============================] - 5s - loss: 0.9752 - acc: 0.6362 - val_loss: 1.3410 - val_acc: 0.5099
Epoch 88/1000
2122/2122 [==============================] - 5s - loss: 0.9109 - acc: 0.6588 - val_loss: 1.2569 - val_acc: 0.5538
Epoch 89/1000
2122/2122 [==============================] - 5s - loss: 0.8678 - acc: 0.6795 - val_loss: 1.2543 - val_acc: 0.5802
Epoch 90/1000
2122/2122 [==============================] - 5s - loss: 0.7991 - acc: 0.7041 - val_loss: 1.1824 - val_acc: 0.6000
Epoch 91/1000
2122/2122 [==============================] - 5s - loss: 0.7838 - acc: 0.7172 - val_loss: 1.4967 - val_acc: 0.4813
Epoch 92/1000
2122/2122 [==============================] - 5s - loss: 1.0603 - acc: 0.6079 - val_loss: 1.2724 - val_acc: 0.5505
Epoch 93/1000
2122/2122 [==============================] - 5s - loss: 0.7818 - acc: 0.7121 - val_loss: 1.0499 - val_acc: 0.6527
Epoch 94/1000
2122/2122 [==============================] - 5s - loss: 0.7311 - acc: 0.7300 - val_loss: 1.2250 - val_acc: 0.5440
Epoch 95/1000
2122/2122 [==============================] - 5s - loss: 0.7738 - acc: 0.7243 - val_loss: 1.0249 - val_acc: 0.6560
Epoch 96/1000
2122/2122 [==============================] - 5s - loss: 0.6776 - acc: 0.7625 - val_loss: 1.4949 - val_acc: 0.5560
Epoch 97/1000
2122/2122 [==============================] - 5s - loss: 1.2786 - acc: 0.5697 - val_loss: 1.0799 - val_acc: 0.6275
Epoch 98/1000
2122/2122 [==============================] - 5s - loss: 0.6050 - acc: 0.7931 - val_loss: 1.1502 - val_acc: 0.6505
Epoch 99/1000
2122/2122 [==============================] - 5s - loss: 0.5540 - acc: 0.8030 - val_loss: 1.0369 - val_acc: 0.6648
Epoch 100/1000
2122/2122 [==============================] - 5s - loss: 0.5930 - acc: 0.7804 - val_loss: 1.3092 - val_acc: 0.5802
Epoch 101/1000
2122/2122 [==============================] - 5s - loss: 0.6203 - acc: 0.7804 - val_loss: 0.9756 - val_acc: 0.6934
Epoch 102/1000
2122/2122 [==============================] - 5s - loss: 0.6102 - acc: 0.7912 - val_loss: 1.3093 - val_acc: 0.5033
Epoch 103/1000
2122/2122 [==============================] - 5s - loss: 0.8406 - acc: 0.6909 - val_loss: 1.0090 - val_acc: 0.6648
Epoch 104/1000
2122/2122 [==============================] - 5s - loss: 0.4653 - acc: 0.8369 - val_loss: 0.9602 - val_acc: 0.6692
Epoch 105/1000
2122/2122 [==============================] - 5s - loss: 0.4744 - acc: 0.8327 - val_loss: 1.1248 - val_acc: 0.6396
Epoch 106/1000
2122/2122 [==============================] - 5s - loss: 0.7676 - acc: 0.7262 - val_loss: 0.9850 - val_acc: 0.6813
Epoch 107/1000
2122/2122 [==============================] - 5s - loss: 0.4585 - acc: 0.8459 - val_loss: 1.0241 - val_acc: 0.6879
Epoch 108/1000
2122/2122 [==============================] - 5s - loss: 0.5747 - acc: 0.7922 - val_loss: 0.9704 - val_acc: 0.7000
Epoch 109/1000
2122/2122 [==============================] - 5s - loss: 0.4286 - acc: 0.8525 - val_loss: 1.1442 - val_acc: 0.6901
Epoch 110/1000
2122/2122 [==============================] - 5s - loss: 0.5270 - acc: 0.8091 - val_loss: 1.0107 - val_acc: 0.6747
Epoch 111/1000
2122/2122 [==============================] - 5s - loss: 0.4730 - acc: 0.8313 - val_loss: 1.3124 - val_acc: 0.6253
Epoch 112/1000
2122/2122 [==============================] - 5s - loss: 1.0159 - acc: 0.6635 - val_loss: 0.9841 - val_acc: 0.7132
Epoch 113/1000
2122/2122 [==============================] - 5s - loss: 0.3370 - acc: 0.8926 - val_loss: 1.0407 - val_acc: 0.7132
Epoch 114/1000
2122/2122 [==============================] - 5s - loss: 0.4804 - acc: 0.8313 - val_loss: 1.2082 - val_acc: 0.6505
Epoch 115/1000
2122/2122 [==============================] - 5s - loss: 0.4322 - acc: 0.8511 - val_loss: 1.1618 - val_acc: 0.7044
Epoch 116/1000
2122/2122 [==============================] - 5s - loss: 0.4207 - acc: 0.8633 - val_loss: 1.0850 - val_acc: 0.7143
Epoch 117/1000
2122/2122 [==============================] - 5s - loss: 0.5687 - acc: 0.8030 - val_loss: 0.9143 - val_acc: 0.7022
Epoch 118/1000
2122/2122 [==============================] - 5s - loss: 0.3527 - acc: 0.8704 - val_loss: 0.9402 - val_acc: 0.7363
Epoch 119/1000
2122/2122 [==============================] - 5s - loss: 0.3907 - acc: 0.8567 - val_loss: 1.0222 - val_acc: 0.7110
Epoch 120/1000
2122/2122 [==============================] - 5s - loss: 0.2902 - acc: 0.9001 - val_loss: 0.9634 - val_acc: 0.7341
Epoch 121/1000
2122/2122 [==============================] - 5s - loss: 0.5921 - acc: 0.8082 - val_loss: 0.8234 - val_acc: 0.7538
Epoch 122/1000
2122/2122 [==============================] - 5s - loss: 0.2272 - acc: 0.9251 - val_loss: 0.9049 - val_acc: 0.7593
Epoch 123/1000
2122/2122 [==============================] - 5s - loss: 0.3100 - acc: 0.8992 - val_loss: 1.4526 - val_acc: 0.5516
Epoch 124/1000
2122/2122 [==============================] - 5s - loss: 0.6251 - acc: 0.7828 - val_loss: 0.8358 - val_acc: 0.7615
Epoch 125/1000
2122/2122 [==============================] - 5s - loss: 0.2358 - acc: 0.9232 - val_loss: 1.0937 - val_acc: 0.7363
Epoch 126/1000
2122/2122 [==============================] - 5s - loss: 0.2074 - acc: 0.9321 - val_loss: 1.1891 - val_acc: 0.7110
Epoch 127/1000
2122/2122 [==============================] - 5s - loss: 1.1642 - acc: 0.6692 - val_loss: 1.5434 - val_acc: 0.4714
Epoch 128/1000
2122/2122 [==============================] - 5s - loss: 0.6588 - acc: 0.7955 - val_loss: 0.8177 - val_acc: 0.7451
Epoch 129/1000
2122/2122 [==============================] - 5s - loss: 0.2091 - acc: 0.9321 - val_loss: 0.8719 - val_acc: 0.7681
Epoch 130/1000
2122/2122 [==============================] - 5s - loss: 0.1785 - acc: 0.9378 - val_loss: 0.9471 - val_acc: 0.7626
Epoch 131/1000
2122/2122 [==============================] - 5s - loss: 0.1768 - acc: 0.9345 - val_loss: 1.1168 - val_acc: 0.7220
Epoch 132/1000
2122/2122 [==============================] - 5s - loss: 0.2881 - acc: 0.9025 - val_loss: 1.0266 - val_acc: 0.7626
Epoch 133/1000
2122/2122 [==============================] - 5s - loss: 0.2407 - acc: 0.9156 - val_loss: 0.9112 - val_acc: 0.7769
Epoch 134/1000
2122/2122 [==============================] - 5s - loss: 0.7425 - acc: 0.7705 - val_loss: 0.9645 - val_acc: 0.7275
Epoch 135/1000
2122/2122 [==============================] - 5s - loss: 0.2888 - acc: 0.9062 - val_loss: 0.9015 - val_acc: 0.7648
Epoch 136/1000
2122/2122 [==============================] - 5s - loss: 0.1559 - acc: 0.9453 - val_loss: 0.9186 - val_acc: 0.7703
Epoch 137/1000
2122/2122 [==============================] - 5s - loss: 0.2037 - acc: 0.9340 - val_loss: 0.9983 - val_acc: 0.7187
Epoch 138/1000
2122/2122 [==============================] - 5s - loss: 0.2329 - acc: 0.9114 - val_loss: 0.9863 - val_acc: 0.7659
Epoch 139/1000
2122/2122 [==============================] - 5s - loss: 0.2733 - acc: 0.9166 - val_loss: 1.5255 - val_acc: 0.5901
Epoch 140/1000
2122/2122 [==============================] - 5s - loss: 1.0432 - acc: 0.6993 - val_loss: 0.8462 - val_acc: 0.7516
Epoch 141/1000
2122/2122 [==============================] - 5s - loss: 0.2213 - acc: 0.9260 - val_loss: 0.9278 - val_acc: 0.7505
Epoch 142/1000
2122/2122 [==============================] - 5s - loss: 0.1437 - acc: 0.9538 - val_loss: 1.0300 - val_acc: 0.7813
Epoch 143/1000
2122/2122 [==============================] - 5s - loss: 0.1220 - acc: 0.9599 - val_loss: 0.9533 - val_acc: 0.7758
Epoch 144/1000
2122/2122 [==============================] - 5s - loss: 0.1617 - acc: 0.9425 - val_loss: 1.0881 - val_acc: 0.7615
Epoch 145/1000
2122/2122 [==============================] - 5s - loss: 0.1865 - acc: 0.9340 - val_loss: 0.9661 - val_acc: 0.7593
Epoch 146/1000
2122/2122 [==============================] - 5s - loss: 0.3706 - acc: 0.8850 - val_loss: 1.5584 - val_acc: 0.5527
Epoch 147/1000
2122/2122 [==============================] - 5s - loss: 0.5413 - acc: 0.8256 - val_loss: 0.8050 - val_acc: 0.7846
Epoch 148/1000
2122/2122 [==============================] - 5s - loss: 0.1117 - acc: 0.9642 - val_loss: 0.8860 - val_acc: 0.7912
Epoch 149/1000
2122/2122 [==============================] - 5s - loss: 0.1134 - acc: 0.9590 - val_loss: 0.9188 - val_acc: 0.8033
Epoch 150/1000
2122/2122 [==============================] - 5s - loss: 0.0997 - acc: 0.9689 - val_loss: 0.8895 - val_acc: 0.7890
Epoch 151/1000
2122/2122 [==============================] - 5s - loss: 1.0251 - acc: 0.7969 - val_loss: 1.4251 - val_acc: 0.5341
Epoch 152/1000
2122/2122 [==============================] - 5s - loss: 0.5939 - acc: 0.8247 - val_loss: 0.8434 - val_acc: 0.7714
Epoch 153/1000
2122/2122 [==============================] - 5s - loss: 0.1496 - acc: 0.9472 - val_loss: 0.8577 - val_acc: 0.7890
Epoch 154/1000
2122/2122 [==============================] - 5s - loss: 0.1205 - acc: 0.9590 - val_loss: 0.8439 - val_acc: 0.8099
Epoch 155/1000
2122/2122 [==============================] - 5s - loss: 0.1074 - acc: 0.9665 - val_loss: 1.0740 - val_acc: 0.7571
Epoch 156/1000
2122/2122 [==============================] - 5s - loss: 0.2289 - acc: 0.9133 - val_loss: 0.9154 - val_acc: 0.7824
Epoch 157/1000
2122/2122 [==============================] - 5s - loss: 0.1078 - acc: 0.9637 - val_loss: 0.9972 - val_acc: 0.7571
Epoch 158/1000
2122/2122 [==============================] - 5s - loss: 0.1659 - acc: 0.9411 - val_loss: 1.2672 - val_acc: 0.7330
Epoch 159/1000
2122/2122 [==============================] - 5s - loss: 0.6821 - acc: 0.7992 - val_loss: 1.0385 - val_acc: 0.6945
Epoch 160/1000
2122/2122 [==============================] - 5s - loss: 0.2033 - acc: 0.9345 - val_loss: 0.8160 - val_acc: 0.7813
Epoch 161/1000
2122/2122 [==============================] - 5s - loss: 0.0926 - acc: 0.9713 - val_loss: 0.8474 - val_acc: 0.7857
Epoch 162/1000
2122/2122 [==============================] - 5s - loss: 0.1068 - acc: 0.9632 - val_loss: 0.9958 - val_acc: 0.7725
Epoch 163/1000
2122/2122 [==============================] - 5s - loss: 0.0930 - acc: 0.9656 - val_loss: 1.0231 - val_acc: 0.7604
Epoch 164/1000
2122/2122 [==============================] - 5s - loss: 0.1886 - acc: 0.9336 - val_loss: 1.0647 - val_acc: 0.7659
Epoch 165/1000
2122/2122 [==============================] - 5s - loss: 0.1352 - acc: 0.9548 - val_loss: 1.0603 - val_acc: 0.7626
Epoch 166/1000
2122/2122 [==============================] - 5s - loss: 0.1973 - acc: 0.9354 - val_loss: 1.3391 - val_acc: 0.7187
Epoch 167/1000
2122/2122 [==============================] - 5s - loss: 1.0092 - acc: 0.7036 - val_loss: 0.9710 - val_acc: 0.7286
Epoch 168/1000
2122/2122 [==============================] - 5s - loss: 0.2190 - acc: 0.9284 - val_loss: 0.8257 - val_acc: 0.7989
Epoch 169/1000
2122/2122 [==============================] - 5s - loss: 0.0945 - acc: 0.9675 - val_loss: 0.8964 - val_acc: 0.7956
Epoch 170/1000
2122/2122 [==============================] - 5s - loss: 0.0841 - acc: 0.9717 - val_loss: 1.0708 - val_acc: 0.7780
Epoch 171/1000
2122/2122 [==============================] - 5s - loss: 0.0939 - acc: 0.9703 - val_loss: 0.9862 - val_acc: 0.7989
Epoch 172/1000
2122/2122 [==============================] - 5s - loss: 0.0743 - acc: 0.9746 - val_loss: 0.9448 - val_acc: 0.7934
Epoch 173/1000
2122/2122 [==============================] - 5s - loss: 0.0820 - acc: 0.9722 - val_loss: 1.0648 - val_acc: 0.7626
Epoch 174/1000
2122/2122 [==============================] - 5s - loss: 0.4257 - acc: 0.8808 - val_loss: 0.9226 - val_acc: 0.7538
Epoch 175/1000
2122/2122 [==============================] - 5s - loss: 0.1487 - acc: 0.9604 - val_loss: 1.1731 - val_acc: 0.7505
Epoch 176/1000
2122/2122 [==============================] - 5s - loss: 0.2133 - acc: 0.9336 - val_loss: 0.7353 - val_acc: 0.8044
Epoch 177/1000
2122/2122 [==============================] - 5s - loss: 0.0754 - acc: 0.9755 - val_loss: 0.8807 - val_acc: 0.8022
Epoch 178/1000
2122/2122 [==============================] - 5s - loss: 0.0603 - acc: 0.9802 - val_loss: 1.1137 - val_acc: 0.7571
Epoch 179/1000
2122/2122 [==============================] - 4s - loss: 0.0401 - acc: 0.9882 - val_loss: 0.5282 - val_acc: 0.8978
Epoch 653/1000
2122/2122 [==============================] - 4s - loss: 0.0323 - acc: 0.9925 - val_loss: 0.4912 - val_acc: 0.8912
Epoch 654/1000
2122/2122 [==============================] - 4s - loss: 0.0294 - acc: 0.9925 - val_loss: 0.6029 - val_acc: 0.8934
Epoch 655/1000
2122/2122 [==============================] - 4s - loss: 0.0206 - acc: 0.9943 - val_loss: 0.5523 - val_acc: 0.8835
Epoch 656/1000
2122/2122 [==============================] - 4s - loss: 0.0119 - acc: 0.9967 - val_loss: 0.5767 - val_acc: 0.8978
Epoch 657/1000
2122/2122 [==============================] - 4s - loss: 0.0590 - acc: 0.9840 - val_loss: 0.6430 - val_acc: 0.8802
Epoch 658/1000
2122/2122 [==============================] - 4s - loss: 0.0416 - acc: 0.9906 - val_loss: 0.6781 - val_acc: 0.8725
Epoch 659/1000
2122/2122 [==============================] - 4s - loss: 0.0248 - acc: 0.9939 - val_loss: 0.6159 - val_acc: 0.8769
Epoch 660/1000
2122/2122 [==============================] - 4s - loss: 0.0184 - acc: 0.9958 - val_loss: 0.6953 - val_acc: 0.8923
Epoch 661/1000
2122/2122 [==============================] - 4s - loss: 0.0218 - acc: 0.9948 - val_loss: 0.5205 - val_acc: 0.8780
Epoch 662/1000
2122/2122 [==============================] - 4s - loss: 0.0422 - acc: 0.9901 - val_loss: 0.9193 - val_acc: 0.8593
Epoch 663/1000
2122/2122 [==============================] - 4s - loss: 0.1543 - acc: 0.9703 - val_loss: 0.5666 - val_acc: 0.8582
Epoch 664/1000
2122/2122 [==============================] - 4s - loss: 0.0520 - acc: 0.9854 - val_loss: 0.6745 - val_acc: 0.8637
Epoch 665/1000
2122/2122 [==============================] - 4s - loss: 0.0191 - acc: 0.9943 - val_loss: 0.5815 - val_acc: 0.8857
Epoch 666/1000
2122/2122 [==============================] - 4s - loss: 0.0232 - acc: 0.9948 - val_loss: 0.6544 - val_acc: 0.8736
Epoch 667/1000
2122/2122 [==============================] - 4s - loss: 0.0399 - acc: 0.9920 - val_loss: 0.5372 - val_acc: 0.8857
Epoch 668/1000
2122/2122 [==============================] - 4s - loss: 0.0413 - acc: 0.9925 - val_loss: 0.6564 - val_acc: 0.8824
Epoch 669/1000
2122/2122 [==============================] - 4s - loss: 0.0276 - acc: 0.9920 - val_loss: 0.5529 - val_acc: 0.8659
Epoch 670/1000
2122/2122 [==============================] - 4s - loss: 0.0511 - acc: 0.9844 - val_loss: 0.7785 - val_acc: 0.7747
Epoch 671/1000
2122/2122 [==============================] - 4s - loss: 0.2743 - acc: 0.9397 - val_loss: 0.6896 - val_acc: 0.8659
Epoch 672/1000
2122/2122 [==============================] - 4s - loss: 0.0356 - acc: 0.9896 - val_loss: 0.8057 - val_acc: 0.8769
Epoch 673/1000
2122/2122 [==============================] - 4s - loss: 0.0313 - acc: 0.9920 - val_loss: 0.8801 - val_acc: 0.8681
Epoch 674/1000
2122/2122 [==============================] - 4s - loss: 0.0226 - acc: 0.9939 - val_loss: 0.6982 - val_acc: 0.8868
Epoch 675/1000
2122/2122 [==============================] - 4s - loss: 0.0168 - acc: 0.9958 - val_loss: 0.6568 - val_acc: 0.8945
Epoch 676/1000
2122/2122 [==============================] - 4s - loss: 0.0411 - acc: 0.9877 - val_loss: 1.2169 - val_acc: 0.8462
Epoch 677/1000
2122/2122 [==============================] - 4s - loss: 0.0755 - acc: 0.9854 - val_loss: 0.7424 - val_acc: 0.8505
Epoch 678/1000
2122/2122 [==============================] - 4s - loss: 0.0629 - acc: 0.9849 - val_loss: 0.6958 - val_acc: 0.8879
Epoch 679/1000
2122/2122 [==============================] - 4s - loss: 0.0250 - acc: 0.9948 - val_loss: 0.8445 - val_acc: 0.8648
Epoch 680/1000
2122/2122 [==============================] - 4s - loss: 0.0512 - acc: 0.9910 - val_loss: 0.7688 - val_acc: 0.8626
Epoch 681/1000
2122/2122 [==============================] - 4s - loss: 0.0500 - acc: 0.9901 - val_loss: 0.6642 - val_acc: 0.8725
Epoch 682/1000
2122/2122 [==============================] - 4s - loss: 0.0479 - acc: 0.9901 - val_loss: 0.5543 - val_acc: 0.8890
Epoch 683/1000
2122/2122 [==============================] - 4s - loss: 0.0241 - acc: 0.9948 - val_loss: 0.5894 - val_acc: 0.8846
Epoch 684/1000
2122/2122 [==============================] - 4s - loss: 0.0137 - acc: 0.9976 - val_loss: 0.6659 - val_acc: 0.8846
Epoch 685/1000
2122/2122 [==============================] - 4s - loss: 0.1001 - acc: 0.9741 - val_loss: 1.7921 - val_acc: 0.8088
Epoch 686/1000
2122/2122 [==============================] - 4s - loss: 0.5408 - acc: 0.9265 - val_loss: 0.7470 - val_acc: 0.8648
Epoch 687/1000
2122/2122 [==============================] - 4s - loss: 0.0459 - acc: 0.9906 - val_loss: 0.6344 - val_acc: 0.8934
Epoch 688/1000
2122/2122 [==============================] - 4s - loss: 0.0091 - acc: 0.9972 - val_loss: 0.7032 - val_acc: 0.8890
Epoch 689/1000
2122/2122 [==============================] - 4s - loss: 0.0225 - acc: 0.9925 - val_loss: 0.6388 - val_acc: 0.8901
Epoch 690/1000
2122/2122 [==============================] - 4s - loss: 0.0043 - acc: 0.9986 - val_loss: 0.7043 - val_acc: 0.8857
Epoch 691/1000
2122/2122 [==============================] - 4s - loss: 0.0483 - acc: 0.9915 - val_loss: 0.9176 - val_acc: 0.8374
Epoch 692/1000
2122/2122 [==============================] - 4s - loss: 0.1306 - acc: 0.9746 - val_loss: 0.5784 - val_acc: 0.8703
Epoch 693/1000
2122/2122 [==============================] - 4s - loss: 0.0439 - acc: 0.9896 - val_loss: 0.6436 - val_acc: 0.8912
Epoch 694/1000
2122/2122 [==============================] - 4s - loss: 0.0290 - acc: 0.9929 - val_loss: 0.6667 - val_acc: 0.8703
Epoch 695/1000
2122/2122 [==============================] - 4s - loss: 0.0149 - acc: 0.9958 - val_loss: 0.7134 - val_acc: 0.8868
Epoch 696/1000
2122/2122 [==============================] - 4s - loss: 0.0146 - acc: 0.9958 - val_loss: 0.7339 - val_acc: 0.8868
Epoch 697/1000
2122/2122 [==============================] - 4s - loss: 0.0378 - acc: 0.9915 - val_loss: 0.6810 - val_acc: 0.8813
Epoch 698/1000
2122/2122 [==============================] - 4s - loss: 0.2001 - acc: 0.9651 - val_loss: 0.4662 - val_acc: 0.8626
Epoch 699/1000
2122/2122 [==============================] - 4s - loss: 0.0627 - acc: 0.9811 - val_loss: 0.6683 - val_acc: 0.8923
Epoch 700/1000
2122/2122 [==============================] - 4s - loss: 0.0304 - acc: 0.9920 - val_loss: 0.6130 - val_acc: 0.8934
Epoch 701/1000
2122/2122 [==============================] - 4s - loss: 0.0293 - acc: 0.9925 - val_loss: 0.7149 - val_acc: 0.8989
Epoch 702/1000
2122/2122 [==============================] - 4s - loss: 0.0884 - acc: 0.9755 - val_loss: 0.7661 - val_acc: 0.8121
Epoch 703/1000
2122/2122 [==============================] - 4s - loss: 0.1013 - acc: 0.9746 - val_loss: 0.6927 - val_acc: 0.8912
Epoch 704/1000
2122/2122 [==============================] - 4s - loss: 0.0261 - acc: 0.9929 - val_loss: 0.5776 - val_acc: 0.8978
Epoch 705/1000
2122/2122 [==============================] - 4s - loss: 0.0371 - acc: 0.9929 - val_loss: 0.4703 - val_acc: 0.8923
Epoch 706/1000
2122/2122 [==============================] - 4s - loss: 0.0574 - acc: 0.9877 - val_loss: 0.9979 - val_acc: 0.8176
Epoch 707/1000
2122/2122 [==============================] - 5s - loss: 0.2148 - acc: 0.9571 - val_loss: 0.5535 - val_acc: 0.8813
Epoch 708/1000
2122/2122 [==============================] - 4s - loss: 0.0326 - acc: 0.9910 - val_loss: 0.6420 - val_acc: 0.8769
Epoch 709/1000
2122/2122 [==============================] - 4s - loss: 0.0263 - acc: 0.9943 - val_loss: 0.5771 - val_acc: 0.8923
Epoch 710/1000
2122/2122 [==============================] - 4s - loss: 0.0202 - acc: 0.9948 - val_loss: 0.6271 - val_acc: 0.8846
Epoch 711/1000
2122/2122 [==============================] - 4s - loss: 0.0089 - acc: 0.9976 - val_loss: 0.6157 - val_acc: 0.8956
Epoch 712/1000
2122/2122 [==============================] - 4s - loss: 0.0331 - acc: 0.9929 - val_loss: 0.6160 - val_acc: 0.8989
Epoch 713/1000
2122/2122 [==============================] - 4s - loss: 0.0049 - acc: 0.9986 - val_loss: 0.6234 - val_acc: 0.8901
Epoch 714/1000
2122/2122 [==============================] - 4s - loss: 0.0914 - acc: 0.9811 - val_loss: 1.0217 - val_acc: 0.6604
Epoch 715/1000
2122/2122 [==============================] - 4s - loss: 0.1868 - acc: 0.9387 - val_loss: 0.4867 - val_acc: 0.8901
Epoch 716/1000
2122/2122 [==============================] - 4s - loss: 0.0120 - acc: 0.9962 - val_loss: 0.7128 - val_acc: 0.8681
Epoch 717/1000
2122/2122 [==============================] - 4s - loss: 0.0374 - acc: 0.9929 - val_loss: 0.4692 - val_acc: 0.8659
Epoch 718/1000
2122/2122 [==============================] - 4s - loss: 0.0563 - acc: 0.9882 - val_loss: 0.5675 - val_acc: 0.8879
Epoch 719/1000
2122/2122 [==============================] - 4s - loss: 0.0256 - acc: 0.9920 - val_loss: 0.7678 - val_acc: 0.8802
Epoch 720/1000
2122/2122 [==============================] - 4s - loss: 0.1631 - acc: 0.9760 - val_loss: 0.6057 - val_acc: 0.8385
Epoch 721/1000
2122/2122 [==============================] - 4s - loss: 0.1371 - acc: 0.9637 - val_loss: 0.5911 - val_acc: 0.8824
Epoch 722/1000
2122/2122 [==============================] - 4s - loss: 0.0157 - acc: 0.9986 - val_loss: 0.4833 - val_acc: 0.8802
Epoch 723/1000
2122/2122 [==============================] - 4s - loss: 0.0423 - acc: 0.9859 - val_loss: 0.6945 - val_acc: 0.8725
Epoch 724/1000
2122/2122 [==============================] - 4s - loss: 0.0765 - acc: 0.9807 - val_loss: 0.6684 - val_acc: 0.8835
Epoch 725/1000
2122/2122 [==============================] - 4s - loss: 0.0244 - acc: 0.9929 - val_loss: 0.6843 - val_acc: 0.8890
Epoch 726/1000
2122/2122 [==============================] - 4s - loss: 0.0317 - acc: 0.9925 - val_loss: 0.6328 - val_acc: 0.8857
Epoch 727/1000
2122/2122 [==============================] - 4s - loss: 0.0279 - acc: 0.9929 - val_loss: 0.6653 - val_acc: 0.8780
Epoch 728/1000
2122/2122 [==============================] - 4s - loss: 0.0181 - acc: 0.9953 - val_loss: 0.6554 - val_acc: 0.8901
Epoch 729/1000
2122/2122 [==============================] - 4s - loss: 0.0385 - acc: 0.9929 - val_loss: 0.7656 - val_acc: 0.8857
Epoch 730/1000
2122/2122 [==============================] - 4s - loss: 0.1185 - acc: 0.9727 - val_loss: 0.7800 - val_acc: 0.8813
Epoch 731/1000
2122/2122 [==============================] - 4s - loss: 0.0288 - acc: 0.9929 - val_loss: 0.7477 - val_acc: 0.8813
Epoch 732/1000
2122/2122 [==============================] - 4s - loss: 0.0798 - acc: 0.9826 - val_loss: 0.6671 - val_acc: 0.8527
Epoch 733/1000
2122/2122 [==============================] - 4s - loss: 0.0794 - acc: 0.9811 - val_loss: 0.5057 - val_acc: 0.8692
Epoch 734/1000
2122/2122 [==============================] - 4s - loss: 0.0585 - acc: 0.9873 - val_loss: 0.4677 - val_acc: 0.8725
Epoch 735/1000
2122/2122 [==============================] - 4s - loss: 0.0534 - acc: 0.9844 - val_loss: 0.5877 - val_acc: 0.8868
Epoch 736/1000
2122/2122 [==============================] - 4s - loss: 0.0085 - acc: 0.9958 - val_loss: 0.5865 - val_acc: 0.8780
Epoch 737/1000
2122/2122 [==============================] - 4s - loss: 0.0311 - acc: 0.9934 - val_loss: 0.5884 - val_acc: 0.8802
Epoch 738/1000
2122/2122 [==============================] - 4s - loss: 0.0158 - acc: 0.9929 - val_loss: 0.7783 - val_acc: 0.8747
Epoch 739/1000
2122/2122 [==============================] - 4s - loss: 0.0554 - acc: 0.9868 - val_loss: 0.5599 - val_acc: 0.8758
Epoch 740/1000
2122/2122 [==============================] - 4s - loss: 0.0701 - acc: 0.9844 - val_loss: 0.5317 - val_acc: 0.8934
Epoch 741/1000
2122/2122 [==============================] - 4s - loss: 0.0227 - acc: 0.9939 - val_loss: 0.6753 - val_acc: 0.9000
Epoch 742/1000
2122/2122 [==============================] - 4s - loss: 0.0266 - acc: 0.9948 - val_loss: 0.6301 - val_acc: 0.8868
Epoch 743/1000
2122/2122 [==============================] - 4s - loss: 0.0559 - acc: 0.9877 - val_loss: 0.7189 - val_acc: 0.8341
Epoch 744/1000
2122/2122 [==============================] - 4s - loss: 0.0915 - acc: 0.9821 - val_loss: 0.5821 - val_acc: 0.8978
Epoch 745/1000
2122/2122 [==============================] - 4s - loss: 0.0557 - acc: 0.9873 - val_loss: 0.7629 - val_acc: 0.8692
Epoch 746/1000
2122/2122 [==============================] - 4s - loss: 0.0275 - acc: 0.9915 - val_loss: 0.6353 - val_acc: 0.8956
Epoch 747/1000
2122/2122 [==============================] - 4s - loss: 0.0225 - acc: 0.9953 - val_loss: 0.7225 - val_acc: 0.8890
Epoch 748/1000
2122/2122 [==============================] - 4s - loss: 0.0314 - acc: 0.9910 - val_loss: 0.6187 - val_acc: 0.8626
Epoch 749/1000
2122/2122 [==============================] - 4s - loss: 0.1641 - acc: 0.9623 - val_loss: 0.6320 - val_acc: 0.8747
Epoch 750/1000
2122/2122 [==============================] - 4s - loss: 0.0267 - acc: 0.9939 - val_loss: 0.6846 - val_acc: 0.8747
Epoch 751/1000
2122/2122 [==============================] - 4s - loss: 0.0544 - acc: 0.9887 - val_loss: 0.5887 - val_acc: 0.8912
Epoch 752/1000
2122/2122 [==============================] - 4s - loss: 0.0226 - acc: 0.9948 - val_loss: 0.7546 - val_acc: 0.8868
Epoch 753/1000
2122/2122 [==============================] - 4s - loss: 0.0427 - acc: 0.9934 - val_loss: 0.6292 - val_acc: 0.8758
Epoch 754/1000
2122/2122 [==============================] - 4s - loss: 0.0345 - acc: 0.9929 - val_loss: 0.5617 - val_acc: 0.8989
Epoch 755/1000
2122/2122 [==============================] - 4s - loss: 0.0386 - acc: 0.9901 - val_loss: 0.6414 - val_acc: 0.8912
Epoch 756/1000
2122/2122 [==============================] - 4s - loss: 0.0376 - acc: 0.9915 - val_loss: 0.8681 - val_acc: 0.8901
Epoch 757/1000
2122/2122 [==============================] - 4s - loss: 0.0161 - acc: 0.9948 - val_loss: 1.0573 - val_acc: 0.8703
Epoch 758/1000
2122/2122 [==============================] - 4s - loss: 0.0501 - acc: 0.9877 - val_loss: 0.5251 - val_acc: 0.9011
Epoch 759/1000
2122/2122 [==============================] - 4s - loss: 0.0327 - acc: 0.9920 - val_loss: 0.5631 - val_acc: 0.9000
Epoch 760/1000
2122/2122 [==============================] - 4s - loss: 0.0325 - acc: 0.9882 - val_loss: 0.5112 - val_acc: 0.9055
Epoch 761/1000
2122/2122 [==============================] - 4s - loss: 0.0144 - acc: 0.9953 - val_loss: 0.5272 - val_acc: 0.8989
Epoch 762/1000
2122/2122 [==============================] - 4s - loss: 0.0978 - acc: 0.9816 - val_loss: 0.7810 - val_acc: 0.8571
Epoch 763/1000
2122/2122 [==============================] - 4s - loss: 0.0654 - acc: 0.9826 - val_loss: 0.6039 - val_acc: 0.8626
Epoch 764/1000
2122/2122 [==============================] - 4s - loss: 0.0368 - acc: 0.9939 - val_loss: 0.5613 - val_acc: 0.8934
Epoch 765/1000
2122/2122 [==============================] - 4s - loss: 0.0281 - acc: 0.9939 - val_loss: 0.5638 - val_acc: 0.9165
Epoch 766/1000
2122/2122 [==============================] - 4s - loss: 0.0324 - acc: 0.9915 - val_loss: 0.5379 - val_acc: 0.8846
Epoch 767/1000
2122/2122 [==============================] - 4s - loss: 0.0222 - acc: 0.9934 - val_loss: 0.4852 - val_acc: 0.9121
Epoch 768/1000
2122/2122 [==============================] - 4s - loss: 0.0317 - acc: 0.9915 - val_loss: 0.6505 - val_acc: 0.8879
Epoch 769/1000
2122/2122 [==============================] - 4s - loss: 0.0207 - acc: 0.9934 - val_loss: 0.6171 - val_acc: 0.9011
Epoch 770/1000
2122/2122 [==============================] - 4s - loss: 0.1052 - acc: 0.9783 - val_loss: 0.4904 - val_acc: 0.8989
Epoch 771/1000
2122/2122 [==============================] - 4s - loss: 0.0311 - acc: 0.9929 - val_loss: 0.5827 - val_acc: 0.9011
Epoch 772/1000
2122/2122 [==============================] - 4s - loss: 0.0344 - acc: 0.9896 - val_loss: 0.4710 - val_acc: 0.8780
Epoch 773/1000
2122/2122 [==============================] - 4s - loss: 0.0563 - acc: 0.9877 - val_loss: 0.4691 - val_acc: 0.9022
Epoch 774/1000
2122/2122 [==============================] - 4s - loss: 0.0260 - acc: 0.9948 - val_loss: 0.6753 - val_acc: 0.8791
Epoch 775/1000
2122/2122 [==============================] - 4s - loss: 0.0851 - acc: 0.9892 - val_loss: 0.5345 - val_acc: 0.9033
Epoch 776/1000
2122/2122 [==============================] - 4s - loss: 0.0279 - acc: 0.9943 - val_loss: 0.5075 - val_acc: 0.9044
Epoch 777/1000
2122/2122 [==============================] - 4s - loss: 0.0334 - acc: 0.9925 - val_loss: 0.5151 - val_acc: 0.8604
Epoch 778/1000
2122/2122 [==============================] - 4s - loss: 0.0503 - acc: 0.9896 - val_loss: 0.4101 - val_acc: 0.8901
Epoch 779/1000
2122/2122 [==============================] - 4s - loss: 0.0606 - acc: 0.9840 - val_loss: 0.5046 - val_acc: 0.8956
Epoch 780/1000
2122/2122 [==============================] - 4s - loss: 0.0150 - acc: 0.9953 - val_loss: 0.8793 - val_acc: 0.8429
Epoch 781/1000
2122/2122 [==============================] - 5s - loss: 0.3604 - acc: 0.9321 - val_loss: 0.5779 - val_acc: 0.8725
Epoch 782/1000
2122/2122 [==============================] - 4s - loss: 0.0233 - acc: 0.9910 - val_loss: 0.5943 - val_acc: 0.8923
Epoch 783/1000
2122/2122 [==============================] - 4s - loss: 0.0137 - acc: 0.9943 - val_loss: 0.6954 - val_acc: 0.8923
Epoch 784/1000
2122/2122 [==============================] - 4s - loss: 0.0213 - acc: 0.9948 - val_loss: 0.6612 - val_acc: 0.9055
Epoch 785/1000
2122/2122 [==============================] - 4s - loss: 0.0267 - acc: 0.9929 - val_loss: 0.6047 - val_acc: 0.8945
Epoch 786/1000
2122/2122 [==============================] - 4s - loss: 0.0642 - acc: 0.9849 - val_loss: 0.6074 - val_acc: 0.8275
Epoch 787/1000
2122/2122 [==============================] - 4s - loss: 0.1137 - acc: 0.9755 - val_loss: 1.0097 - val_acc: 0.8011
Epoch 788/1000
2122/2122 [==============================] - 4s - loss: 0.2114 - acc: 0.9623 - val_loss: 0.5250 - val_acc: 0.8879
Epoch 789/1000
2122/2122 [==============================] - 4s - loss: 0.0333 - acc: 0.9925 - val_loss: 0.4567 - val_acc: 0.8945
Epoch 790/1000
2122/2122 [==============================] - 4s - loss: 0.0222 - acc: 0.9953 - val_loss: 0.5160 - val_acc: 0.8835
Epoch 791/1000
2122/2122 [==============================] - 4s - loss: 0.0129 - acc: 0.9972 - val_loss: 0.4712 - val_acc: 0.8934
Epoch 792/1000
2122/2122 [==============================] - 4s - loss: 0.0252 - acc: 0.9925 - val_loss: 0.5441 - val_acc: 0.8945
Epoch 793/1000
2122/2122 [==============================] - 4s - loss: 0.0130 - acc: 0.9967 - val_loss: 0.5722 - val_acc: 0.8868
Epoch 794/1000
2122/2122 [==============================] - 4s - loss: 0.0247 - acc: 0.9943 - val_loss: 0.7876 - val_acc: 0.8516
Epoch 795/1000
2122/2122 [==============================] - 4s - loss: 0.0613 - acc: 0.9854 - val_loss: 0.5239 - val_acc: 0.8923
Epoch 796/1000
2122/2122 [==============================] - 4s - loss: 0.0372 - acc: 0.9929 - val_loss: 0.5120 - val_acc: 0.8538
Epoch 797/1000
2122/2122 [==============================] - 4s - loss: 0.0883 - acc: 0.9741 - val_loss: 0.7404 - val_acc: 0.8253
Epoch 798/1000
2122/2122 [==============================] - 4s - loss: 0.0754 - acc: 0.9821 - val_loss: 0.9085 - val_acc: 0.8363
Epoch 799/1000
2122/2122 [==============================] - 4s - loss: 0.1276 - acc: 0.9760 - val_loss: 0.5846 - val_acc: 0.8615
Epoch 800/1000
2122/2122 [==============================] - 4s - loss: 0.0172 - acc: 0.9939 - val_loss: 0.5879 - val_acc: 0.8945
Epoch 801/1000
2122/2122 [==============================] - 4s - loss: 0.0037 - acc: 0.9986 - val_loss: 0.6848 - val_acc: 0.8846
Epoch 802/1000
2122/2122 [==============================] - 4s - loss: 0.0398 - acc: 0.9934 - val_loss: 0.5685 - val_acc: 0.9011
Epoch 803/1000
2122/2122 [==============================] - 4s - loss: 0.0403 - acc: 0.9906 - val_loss: 0.6414 - val_acc: 0.8824
Epoch 804/1000
2122/2122 [==============================] - 4s - loss: 0.0268 - acc: 0.9925 - val_loss: 0.6941 - val_acc: 0.8846
Epoch 805/1000
2122/2122 [==============================] - 4s - loss: 0.0205 - acc: 0.9948 - val_loss: 0.5748 - val_acc: 0.9033
Epoch 806/1000
2122/2122 [==============================] - 4s - loss: 0.0083 - acc: 0.9976 - val_loss: 0.7020 - val_acc: 0.8879
Epoch 807/1000
2122/2122 [==============================] - 4s - loss: 0.1932 - acc: 0.9614 - val_loss: 0.4902 - val_acc: 0.8824
Epoch 808/1000
2122/2122 [==============================] - 4s - loss: 0.0374 - acc: 0.9887 - val_loss: 0.6242 - val_acc: 0.8813
Epoch 809/1000
2122/2122 [==============================] - 4s - loss: 0.0710 - acc: 0.9844 - val_loss: 0.6405 - val_acc: 0.8901
Epoch 810/1000
2122/2122 [==============================] - 4s - loss: 0.0257 - acc: 0.9925 - val_loss: 0.6598 - val_acc: 0.8582
Epoch 811/1000
2122/2122 [==============================] - 4s - loss: 0.0235 - acc: 0.9915 - val_loss: 0.5846 - val_acc: 0.9033
Epoch 812/1000
2122/2122 [==============================] - 4s - loss: 0.0300 - acc: 0.9925 - val_loss: 0.5583 - val_acc: 0.8989
Epoch 813/1000
2122/2122 [==============================] - 4s - loss: 0.0353 - acc: 0.9929 - val_loss: 0.6974 - val_acc: 0.8835
Epoch 814/1000
2122/2122 [==============================] - 4s - loss: 0.0513 - acc: 0.9882 - val_loss: 0.7402 - val_acc: 0.8429
Epoch 815/1000
2122/2122 [==============================] - 4s - loss: 0.1730 - acc: 0.9548 - val_loss: 0.6599 - val_acc: 0.8857
Epoch 816/1000
2122/2122 [==============================] - 4s - loss: 0.0303 - acc: 0.9943 - val_loss: 0.6682 - val_acc: 0.8989
Epoch 817/1000
2122/2122 [==============================] - 4s - loss: 0.0248 - acc: 0.9939 - val_loss: 0.5205 - val_acc: 0.9011
Epoch 818/1000
2122/2122 [==============================] - 4s - loss: 0.0128 - acc: 0.9962 - val_loss: 0.5055 - val_acc: 0.9055
Epoch 819/1000
2122/2122 [==============================] - 4s - loss: 0.0198 - acc: 0.9958 - val_loss: 0.5397 - val_acc: 0.9044
Epoch 820/1000
2122/2122 [==============================] - 4s - loss: 0.0304 - acc: 0.9943 - val_loss: 0.8066 - val_acc: 0.8824
Epoch 821/1000
2122/2122 [==============================] - 4s - loss: 0.0518 - acc: 0.9906 - val_loss: 0.6759 - val_acc: 0.8615
Epoch 822/1000
2122/2122 [==============================] - 4s - loss: 0.0609 - acc: 0.9877 - val_loss: 0.7012 - val_acc: 0.8736
Epoch 823/1000
2122/2122 [==============================] - 4s - loss: 0.0110 - acc: 0.9958 - val_loss: 0.7699 - val_acc: 0.8890
Epoch 824/1000
2122/2122 [==============================] - 4s - loss: 0.0468 - acc: 0.9896 - val_loss: 0.8800 - val_acc: 0.8440
Epoch 825/1000
2122/2122 [==============================] - 4s - loss: 0.0538 - acc: 0.9863 - val_loss: 0.8777 - val_acc: 0.8879
Epoch 826/1000
2122/2122 [==============================] - 4s - loss: 0.0530 - acc: 0.9915 - val_loss: 0.6385 - val_acc: 0.8626
Epoch 827/1000
2122/2122 [==============================] - 4s - loss: 0.0581 - acc: 0.9859 - val_loss: 0.7030 - val_acc: 0.8835
Epoch 828/1000
2122/2122 [==============================] - 4s - loss: 0.0472 - acc: 0.9920 - val_loss: 0.8236 - val_acc: 0.8934
Epoch 829/1000
2122/2122 [==============================] - 4s - loss: 0.0679 - acc: 0.9896 - val_loss: 0.5028 - val_acc: 0.9055
Epoch 830/1000
2122/2122 [==============================] - 4s - loss: 0.0196 - acc: 0.9962 - val_loss: 0.5649 - val_acc: 0.8879
Epoch 831/1000
2122/2122 [==============================] - 4s - loss: 0.0291 - acc: 0.9910 - val_loss: 0.6065 - val_acc: 0.8824
Epoch 832/1000
2122/2122 [==============================] - 4s - loss: 0.0332 - acc: 0.9943 - val_loss: 0.5196 - val_acc: 0.9088
Epoch 833/1000
2122/2122 [==============================] - 4s - loss: 0.0376 - acc: 0.9925 - val_loss: 0.6286 - val_acc: 0.8330
Epoch 834/1000
2122/2122 [==============================] - 4s - loss: 0.2857 - acc: 0.9449 - val_loss: 0.5909 - val_acc: 0.8912
Epoch 835/1000
2122/2122 [==============================] - 4s - loss: 0.0193 - acc: 0.9948 - val_loss: 0.6637 - val_acc: 0.8846
Epoch 836/1000
2122/2122 [==============================] - 4s - loss: 0.0245 - acc: 0.9925 - val_loss: 0.6382 - val_acc: 0.8802
Epoch 837/1000
2122/2122 [==============================] - 4s - loss: 0.0546 - acc: 0.9887 - val_loss: 0.5045 - val_acc: 0.8857
Epoch 838/1000
2122/2122 [==============================] - 4s - loss: 0.0223 - acc: 0.9943 - val_loss: 0.5730 - val_acc: 0.8846
Epoch 839/1000
2122/2122 [==============================] - 4s - loss: 0.0505 - acc: 0.9877 - val_loss: 0.5582 - val_acc: 0.8890
Epoch 840/1000
2122/2122 [==============================] - 4s - loss: 0.0347 - acc: 0.9920 - val_loss: 0.5955 - val_acc: 0.9121
Epoch 841/1000
2122/2122 [==============================] - 4s - loss: 0.0457 - acc: 0.9939 - val_loss: 0.5056 - val_acc: 0.9154
Epoch 842/1000
2122/2122 [==============================] - 4s - loss: 0.0629 - acc: 0.9868 - val_loss: 0.4984 - val_acc: 0.8890
Epoch 843/1000
2122/2122 [==============================] - 4s - loss: 0.0313 - acc: 0.9925 - val_loss: 0.6419 - val_acc: 0.8648
Epoch 844/1000
2122/2122 [==============================] - 4s - loss: 0.1061 - acc: 0.9684 - val_loss: 1.1979 - val_acc: 0.7813
Epoch 845/1000
2122/2122 [==============================] - 4s - loss: 0.2484 - acc: 0.9453 - val_loss: 0.4516 - val_acc: 0.8967
Epoch 846/1000
2122/2122 [==============================] - 4s - loss: 0.0250 - acc: 0.9948 - val_loss: 0.5191 - val_acc: 0.9044
Epoch 847/1000
2122/2122 [==============================] - 4s - loss: 0.0195 - acc: 0.9972 - val_loss: 0.5208 - val_acc: 0.9110
Epoch 848/1000
2122/2122 [==============================] - 4s - loss: 0.0284 - acc: 0.9943 - val_loss: 0.4636 - val_acc: 0.8758
Epoch 849/1000
2122/2122 [==============================] - 4s - loss: 0.0159 - acc: 0.9948 - val_loss: 0.4844 - val_acc: 0.9088
Epoch 850/1000
2122/2122 [==============================] - 4s - loss: 0.0123 - acc: 0.9972 - val_loss: 0.5336 - val_acc: 0.8857
Epoch 851/1000
2122/2122 [==============================] - 4s - loss: 0.0564 - acc: 0.9868 - val_loss: 0.5998 - val_acc: 0.8626
Epoch 852/1000
2122/2122 [==============================] - 4s - loss: 0.0129 - acc: 0.9967 - val_loss: 0.6570 - val_acc: 0.8967
Epoch 853/1000
2122/2122 [==============================] - 4s - loss: 0.0195 - acc: 0.9958 - val_loss: 0.6402 - val_acc: 0.9055
Epoch 854/1000
2122/2122 [==============================] - 4s - loss: 0.0239 - acc: 0.9943 - val_loss: 0.8484 - val_acc: 0.8330
Epoch 855/1000
2122/2122 [==============================] - 5s - loss: 0.2936 - acc: 0.9406 - val_loss: 0.4993 - val_acc: 0.8813
Epoch 856/1000
2122/2122 [==============================] - 4s - loss: 0.0546 - acc: 0.9859 - val_loss: 0.6301 - val_acc: 0.8989
Epoch 857/1000
2122/2122 [==============================] - 4s - loss: 0.0336 - acc: 0.9939 - val_loss: 0.7225 - val_acc: 0.8945
Epoch 858/1000
2122/2122 [==============================] - 4s - loss: 0.0512 - acc: 0.9882 - val_loss: 0.6017 - val_acc: 0.8879
Epoch 859/1000
2122/2122 [==============================] - 4s - loss: 0.0606 - acc: 0.9873 - val_loss: 0.4461 - val_acc: 0.9000
Epoch 860/1000
2122/2122 [==============================] - 4s - loss: 0.0301 - acc: 0.9929 - val_loss: 0.6296 - val_acc: 0.8945
Epoch 861/1000
2122/2122 [==============================] - 4s - loss: 0.0257 - acc: 0.9925 - val_loss: 0.6122 - val_acc: 0.8912
Epoch 862/1000
2122/2122 [==============================] - 4s - loss: 0.0390 - acc: 0.9910 - val_loss: 0.8844 - val_acc: 0.8813
Epoch 863/1000
2122/2122 [==============================] - 4s - loss: 0.0393 - acc: 0.9906 - val_loss: 0.8141 - val_acc: 0.8659
Epoch 864/1000
2122/2122 [==============================] - 4s - loss: 0.0677 - acc: 0.9854 - val_loss: 0.9916 - val_acc: 0.8363
Epoch 865/1000
2122/2122 [==============================] - 4s - loss: 0.1008 - acc: 0.9783 - val_loss: 0.5468 - val_acc: 0.8967
Epoch 866/1000
2122/2122 [==============================] - 4s - loss: 0.0399 - acc: 0.9896 - val_loss: 0.6595 - val_acc: 0.8890
Epoch 867/1000
2122/2122 [==============================] - 4s - loss: 0.0114 - acc: 0.9972 - val_loss: 0.6803 - val_acc: 0.8945
Epoch 868/1000
2122/2122 [==============================] - 4s - loss: 0.0141 - acc: 0.9958 - val_loss: 1.1736 - val_acc: 0.8505
Epoch 869/1000
2122/2122 [==============================] - 4s - loss: 0.3828 - acc: 0.9364 - val_loss: 0.5546 - val_acc: 0.8956
Epoch 870/1000
2122/2122 [==============================] - 4s - loss: 0.0258 - acc: 0.9934 - val_loss: 0.6866 - val_acc: 0.8923
Epoch 871/1000
2122/2122 [==============================] - 4s - loss: 0.0112 - acc: 0.9972 - val_loss: 0.6983 - val_acc: 0.8923
Epoch 872/1000
2122/2122 [==============================] - 4s - loss: 0.0298 - acc: 0.9934 - val_loss: 0.4717 - val_acc: 0.8560
Epoch 873/1000
2122/2122 [==============================] - 4s - loss: 0.0622 - acc: 0.9849 - val_loss: 0.4472 - val_acc: 0.8956
Epoch 874/1000
2122/2122 [==============================] - 4s - loss: 0.0701 - acc: 0.9840 - val_loss: 0.4839 - val_acc: 0.8703
Epoch 875/1000
2122/2122 [==============================] - 4s - loss: 0.0450 - acc: 0.9901 - val_loss: 0.6757 - val_acc: 0.8967
Epoch 876/1000
2122/2122 [==============================] - 4s - loss: 0.0285 - acc: 0.9929 - val_loss: 0.7175 - val_acc: 0.8978
Epoch 877/1000
2122/2122 [==============================] - 4s - loss: 0.0398 - acc: 0.9943 - val_loss: 0.5458 - val_acc: 0.8879
Epoch 878/1000
2122/2122 [==============================] - 4s - loss: 0.0810 - acc: 0.9844 - val_loss: 0.5469 - val_acc: 0.8791
Epoch 879/1000
2122/2122 [==============================] - 4s - loss: 0.0382 - acc: 0.9920 - val_loss: 0.4629 - val_acc: 0.8901
Epoch 880/1000
2122/2122 [==============================] - 4s - loss: 0.0900 - acc: 0.9849 - val_loss: 0.4487 - val_acc: 0.9000
Epoch 881/1000
2122/2122 [==============================] - 4s - loss: 0.0315 - acc: 0.9939 - val_loss: 0.6829 - val_acc: 0.8923
Epoch 882/1000
2122/2122 [==============================] - 4s - loss: 0.0265 - acc: 0.9939 - val_loss: 1.1904 - val_acc: 0.7286
Epoch 883/1000
2122/2122 [==============================] - 4s - loss: 0.4436 - acc: 0.9288 - val_loss: 0.5935 - val_acc: 0.8802
Epoch 884/1000
2122/2122 [==============================] - 4s - loss: 0.0424 - acc: 0.9892 - val_loss: 0.5641 - val_acc: 0.8824
Epoch 885/1000
2122/2122 [==============================] - 4s - loss: 0.0094 - acc: 0.9958 - val_loss: 0.6861 - val_acc: 0.8835
Epoch 886/1000
2122/2122 [==============================] - 4s - loss: 0.0241 - acc: 0.9967 - val_loss: 0.7099 - val_acc: 0.8923
Epoch 887/1000
2122/2122 [==============================] - 4s - loss: 0.0290 - acc: 0.9939 - val_loss: 0.5694 - val_acc: 0.8846
Epoch 888/1000
2122/2122 [==============================] - 4s - loss: 0.0193 - acc: 0.9962 - val_loss: 0.8021 - val_acc: 0.8626
Epoch 889/1000
2122/2122 [==============================] - 4s - loss: 0.1273 - acc: 0.9797 - val_loss: 0.5462 - val_acc: 0.8813
Epoch 890/1000
2122/2122 [==============================] - 4s - loss: 0.0300 - acc: 0.9953 - val_loss: 0.6055 - val_acc: 0.8330
Epoch 891/1000
2122/2122 [==============================] - 4s - loss: 0.1098 - acc: 0.9722 - val_loss: 0.3964 - val_acc: 0.8890
Epoch 892/1000
2122/2122 [==============================] - 4s - loss: 0.0218 - acc: 0.9958 - val_loss: 0.6590 - val_acc: 0.8934
Epoch 893/1000
2122/2122 [==============================] - 4s - loss: 0.0760 - acc: 0.9873 - val_loss: 0.5975 - val_acc: 0.8407
Epoch 894/1000
2122/2122 [==============================] - 4s - loss: 0.0351 - acc: 0.9873 - val_loss: 0.5533 - val_acc: 0.9044
Epoch 895/1000
2122/2122 [==============================] - 4s - loss: 0.0407 - acc: 0.9906 - val_loss: 0.5426 - val_acc: 0.8813
Epoch 896/1000
2122/2122 [==============================] - 4s - loss: 0.0316 - acc: 0.9925 - val_loss: 0.5936 - val_acc: 0.8835
Epoch 897/1000
2122/2122 [==============================] - 4s - loss: 0.0309 - acc: 0.9910 - val_loss: 0.8346 - val_acc: 0.8736
Epoch 898/1000
2122/2122 [==============================] - 4s - loss: 0.0545 - acc: 0.9910 - val_loss: 0.6681 - val_acc: 0.8956
Epoch 899/1000
2122/2122 [==============================] - 4s - loss: 0.0336 - acc: 0.9925 - val_loss: 0.6555 - val_acc: 0.8890
Epoch 900/1000
2122/2122 [==============================] - 4s - loss: 0.1069 - acc: 0.9840 - val_loss: 0.7005 - val_acc: 0.8725
Epoch 901/1000
2122/2122 [==============================] - 4s - loss: 0.0917 - acc: 0.9887 - val_loss: 0.6014 - val_acc: 0.8934
Epoch 902/1000
2122/2122 [==============================] - 4s - loss: 0.0267 - acc: 0.9934 - val_loss: 0.7259 - val_acc: 0.8934
Epoch 903/1000
2122/2122 [==============================] - 4s - loss: 0.0469 - acc: 0.9910 - val_loss: 0.6909 - val_acc: 0.8703
Epoch 904/1000
2122/2122 [==============================] - 4s - loss: 0.0595 - acc: 0.9873 - val_loss: 1.4284 - val_acc: 0.8319
Epoch 905/1000
2122/2122 [==============================] - 4s - loss: 0.3083 - acc: 0.9500 - val_loss: 0.5488 - val_acc: 0.8989
Epoch 906/1000
2122/2122 [==============================] - 4s - loss: 0.0503 - acc: 0.9910 - val_loss: 0.6562 - val_acc: 0.8330
Epoch 907/1000
2122/2122 [==============================] - 4s - loss: 0.0666 - acc: 0.9830 - val_loss: 0.5298 - val_acc: 0.8824
Epoch 908/1000
2122/2122 [==============================] - 4s - loss: 0.0209 - acc: 0.9958 - val_loss: 0.5276 - val_acc: 0.8978
Epoch 909/1000
2122/2122 [==============================] - 4s - loss: 0.0174 - acc: 0.9958 - val_loss: 0.5694 - val_acc: 0.8813
Epoch 910/1000
2122/2122 [==============================] - 4s - loss: 0.0215 - acc: 0.9953 - val_loss: 0.5930 - val_acc: 0.8769
Epoch 911/1000
2122/2122 [==============================] - 4s - loss: 0.0276 - acc: 0.9939 - val_loss: 0.4892 - val_acc: 0.8978
Epoch 912/1000
2122/2122 [==============================] - 4s - loss: 0.0198 - acc: 0.9948 - val_loss: 0.7104 - val_acc: 0.8934
Epoch 913/1000
2122/2122 [==============================] - 4s - loss: 0.1661 - acc: 0.9675 - val_loss: 0.8175 - val_acc: 0.8582
Epoch 914/1000
2122/2122 [==============================] - 4s - loss: 0.0253 - acc: 0.9939 - val_loss: 0.9863 - val_acc: 0.8560
Epoch 915/1000
2122/2122 [==============================] - 4s - loss: 0.0803 - acc: 0.9821 - val_loss: 0.5990 - val_acc: 0.8945
Epoch 916/1000
2122/2122 [==============================] - 4s - loss: 0.0152 - acc: 0.9976 - val_loss: 0.5032 - val_acc: 0.9066
Epoch 917/1000
2122/2122 [==============================] - 4s - loss: 0.0292 - acc: 0.9958 - val_loss: 0.5484 - val_acc: 0.8978
Epoch 918/1000
2122/2122 [==============================] - 4s - loss: 0.0484 - acc: 0.9892 - val_loss: 0.5377 - val_acc: 0.8736
Epoch 919/1000
2122/2122 [==============================] - 4s - loss: 0.0518 - acc: 0.9849 - val_loss: 0.5975 - val_acc: 0.8912
Epoch 920/1000
2122/2122 [==============================] - 4s - loss: 0.0196 - acc: 0.9953 - val_loss: 0.7143 - val_acc: 0.8989
Epoch 921/1000
2122/2122 [==============================] - 4s - loss: 0.0418 - acc: 0.9906 - val_loss: 0.5696 - val_acc: 0.9000
Epoch 922/1000
2122/2122 [==============================] - 4s - loss: 0.0242 - acc: 0.9953 - val_loss: 0.4930 - val_acc: 0.9088
Epoch 923/1000
2122/2122 [==============================] - 4s - loss: 0.0537 - acc: 0.9873 - val_loss: 1.5268 - val_acc: 0.8440
Epoch 924/1000
2122/2122 [==============================] - 4s - loss: 0.1416 - acc: 0.9835 - val_loss: 0.6686 - val_acc: 0.8868
Epoch 925/1000
2122/2122 [==============================] - 4s - loss: 0.0486 - acc: 0.9939 - val_loss: 0.5902 - val_acc: 0.8978
Epoch 926/1000
2122/2122 [==============================] - 4s - loss: 0.0349 - acc: 0.9934 - val_loss: 0.5608 - val_acc: 0.9044
Epoch 927/1000
2122/2122 [==============================] - 4s - loss: 0.0449 - acc: 0.9939 - val_loss: 0.6568 - val_acc: 0.8934
Epoch 928/1000
2122/2122 [==============================] - 4s - loss: 0.0525 - acc: 0.9896 - val_loss: 0.7186 - val_acc: 0.8560
Epoch 929/1000
2122/2122 [==============================] - 4s - loss: 0.1659 - acc: 0.9675 - val_loss: 0.4972 - val_acc: 0.9088
Epoch 930/1000
2122/2122 [==============================] - 4s - loss: 0.0297 - acc: 0.9934 - val_loss: 0.5379 - val_acc: 0.9000
Epoch 931/1000
2122/2122 [==============================] - 4s - loss: 0.0183 - acc: 0.9962 - val_loss: 0.7892 - val_acc: 0.8725
Epoch 932/1000
2122/2122 [==============================] - 4s - loss: 0.0624 - acc: 0.9849 - val_loss: 0.5817 - val_acc: 0.8989
Epoch 933/1000
2122/2122 [==============================] - 4s - loss: 0.0323 - acc: 0.9953 - val_loss: 0.5854 - val_acc: 0.8516
Epoch 934/1000
2122/2122 [==============================] - 4s - loss: 0.0362 - acc: 0.9910 - val_loss: 0.5328 - val_acc: 0.9000
Epoch 935/1000
2122/2122 [==============================] - 4s - loss: 0.0311 - acc: 0.9962 - val_loss: 0.6864 - val_acc: 0.8857
Epoch 936/1000
2122/2122 [==============================] - 4s - loss: 0.3710 - acc: 0.9359 - val_loss: 0.4627 - val_acc: 0.9055
Epoch 937/1000
2122/2122 [==============================] - 4s - loss: 0.0193 - acc: 0.9939 - val_loss: 0.4820 - val_acc: 0.9110
Epoch 938/1000
2122/2122 [==============================] - 4s - loss: 0.0283 - acc: 0.9934 - val_loss: 0.5543 - val_acc: 0.8747
Epoch 939/1000
2122/2122 [==============================] - 4s - loss: 0.0437 - acc: 0.9934 - val_loss: 0.5732 - val_acc: 0.9132
Epoch 940/1000
2122/2122 [==============================] - 4s - loss: 0.0376 - acc: 0.9943 - val_loss: 0.3760 - val_acc: 0.9165
Epoch 941/1000
2122/2122 [==============================] - 4s - loss: 0.0325 - acc: 0.9948 - val_loss: 0.5063 - val_acc: 0.8989
Epoch 942/1000
2122/2122 [==============================] - 4s - loss: 0.0247 - acc: 0.9948 - val_loss: 0.6299 - val_acc: 0.9066
Epoch 943/1000
2122/2122 [==============================] - 4s - loss: 0.0915 - acc: 0.9835 - val_loss: 0.6448 - val_acc: 0.8692
Epoch 944/1000
2122/2122 [==============================] - 4s - loss: 0.0818 - acc: 0.9802 - val_loss: 0.4176 - val_acc: 0.8967
Epoch 945/1000
2122/2122 [==============================] - 4s - loss: 0.0400 - acc: 0.9939 - val_loss: 0.4556 - val_acc: 0.8967
Epoch 946/1000
2122/2122 [==============================] - 4s - loss: 0.0481 - acc: 0.9892 - val_loss: 0.6394 - val_acc: 0.8956
Epoch 947/1000
2122/2122 [==============================] - 4s - loss: 0.0556 - acc: 0.9873 - val_loss: 0.5067 - val_acc: 0.9055
Epoch 948/1000
2122/2122 [==============================] - 4s - loss: 0.0343 - acc: 0.9920 - val_loss: 1.1750 - val_acc: 0.8418
Epoch 949/1000
2122/2122 [==============================] - 4s - loss: 0.3040 - acc: 0.9524 - val_loss: 0.6977 - val_acc: 0.9022
Epoch 950/1000
2122/2122 [==============================] - 4s - loss: 0.0532 - acc: 0.9906 - val_loss: 0.5251 - val_acc: 0.9044
Epoch 951/1000
2122/2122 [==============================] - 4s - loss: 0.0110 - acc: 0.9967 - val_loss: 0.5594 - val_acc: 0.8978
Epoch 952/1000
2122/2122 [==============================] - 4s - loss: 0.0352 - acc: 0.9925 - val_loss: 0.6471 - val_acc: 0.8890
Epoch 953/1000
2122/2122 [==============================] - 4s - loss: 0.0477 - acc: 0.9901 - val_loss: 0.7115 - val_acc: 0.9132
Epoch 954/1000
2122/2122 [==============================] - 4s - loss: 0.0357 - acc: 0.9906 - val_loss: 0.8106 - val_acc: 0.8549
Epoch 955/1000
2122/2122 [==============================] - 4s - loss: 0.0689 - acc: 0.9873 - val_loss: 0.8060 - val_acc: 0.8967
Epoch 956/1000
2122/2122 [==============================] - 4s - loss: 0.0590 - acc: 0.9906 - val_loss: 0.6837 - val_acc: 0.9077
Epoch 957/1000
2122/2122 [==============================] - 4s - loss: 0.0519 - acc: 0.9910 - val_loss: 0.7239 - val_acc: 0.9033
Epoch 958/1000
2122/2122 [==============================] - 4s - loss: 0.0225 - acc: 0.9972 - val_loss: 0.6051 - val_acc: 0.9077
Epoch 959/1000
2122/2122 [==============================] - 4s - loss: 0.2656 - acc: 0.9656 - val_loss: 0.7105 - val_acc: 0.8670
Epoch 960/1000
2122/2122 [==============================] - 4s - loss: 0.0886 - acc: 0.9826 - val_loss: 0.5121 - val_acc: 0.9088
Epoch 961/1000
2122/2122 [==============================] - 4s - loss: 0.0230 - acc: 0.9943 - val_loss: 0.6195 - val_acc: 0.9011
Epoch 962/1000
2122/2122 [==============================] - 4s - loss: 0.0225 - acc: 0.9953 - val_loss: 0.5565 - val_acc: 0.8989
Epoch 963/1000
2122/2122 [==============================] - 4s - loss: 0.0397 - acc: 0.9925 - val_loss: 0.5939 - val_acc: 0.8901
Epoch 964/1000
2122/2122 [==============================] - 4s - loss: 0.0248 - acc: 0.9939 - val_loss: 0.5026 - val_acc: 0.8813
Epoch 965/1000
2122/2122 [==============================] - 4s - loss: 0.0358 - acc: 0.9925 - val_loss: 0.7535 - val_acc: 0.8923
Epoch 966/1000
2122/2122 [==============================] - 4s - loss: 0.0121 - acc: 0.9967 - val_loss: 0.6982 - val_acc: 0.9022
Epoch 967/1000
2122/2122 [==============================] - 4s - loss: 0.0102 - acc: 0.9976 - val_loss: 0.6622 - val_acc: 0.8967
Epoch 968/1000
2122/2122 [==============================] - 4s - loss: 0.0889 - acc: 0.9830 - val_loss: 2.0099 - val_acc: 0.7747
Epoch 969/1000
2122/2122 [==============================] - 4s - loss: 0.6199 - acc: 0.9222 - val_loss: 0.4532 - val_acc: 0.8835
Epoch 970/1000
2122/2122 [==============================] - 4s - loss: 0.0320 - acc: 0.9915 - val_loss: 0.5462 - val_acc: 0.8978
Epoch 971/1000
2122/2122 [==============================] - 4s - loss: 0.0401 - acc: 0.9920 - val_loss: 0.6195 - val_acc: 0.8681
Epoch 972/1000
2122/2122 [==============================] - 4s - loss: 0.0375 - acc: 0.9906 - val_loss: 0.6856 - val_acc: 0.8802
Epoch 973/1000
2122/2122 [==============================] - 4s - loss: 0.0302 - acc: 0.9943 - val_loss: 0.8180 - val_acc: 0.8846
Epoch 974/1000
2122/2122 [==============================] - 4s - loss: 0.0293 - acc: 0.9929 - val_loss: 1.0361 - val_acc: 0.8857
Epoch 975/1000
2122/2122 [==============================] - 4s - loss: 0.0422 - acc: 0.9929 - val_loss: 0.5849 - val_acc: 0.8593
Epoch 976/1000
2122/2122 [==============================] - 4s - loss: 0.0411 - acc: 0.9915 - val_loss: 0.7616 - val_acc: 0.8879
Epoch 977/1000
2122/2122 [==============================] - 4s - loss: 0.0204 - acc: 0.9943 - val_loss: 1.0723 - val_acc: 0.8714
Epoch 978/1000
2122/2122 [==============================] - 4s - loss: 0.0345 - acc: 0.9929 - val_loss: 0.7063 - val_acc: 0.8769
Epoch 979/1000
2122/2122 [==============================] - 5s - loss: 0.0625 - acc: 0.9892 - val_loss: 0.6148 - val_acc: 0.9066
Epoch 980/1000
2122/2122 [==============================] - 4s - loss: 0.0247 - acc: 0.9939 - val_loss: 0.6866 - val_acc: 0.8945
Epoch 981/1000
2122/2122 [==============================] - 4s - loss: 0.0265 - acc: 0.9943 - val_loss: 0.6597 - val_acc: 0.9055
Epoch 982/1000
2122/2122 [==============================] - 4s - loss: 0.0367 - acc: 0.9934 - val_loss: 0.7127 - val_acc: 0.8835
Epoch 983/1000
2122/2122 [==============================] - 4s - loss: 0.0843 - acc: 0.9859 - val_loss: 0.6169 - val_acc: 0.8835
Epoch 984/1000
2122/2122 [==============================] - 4s - loss: 0.0512 - acc: 0.9896 - val_loss: 0.5646 - val_acc: 0.8989
Epoch 985/1000
2122/2122 [==============================] - 4s - loss: 0.0442 - acc: 0.9920 - val_loss: 0.7267 - val_acc: 0.8824
Epoch 986/1000
2122/2122 [==============================] - 4s - loss: 0.0412 - acc: 0.9920 - val_loss: 0.6500 - val_acc: 0.8835
Epoch 987/1000
2122/2122 [==============================] - 4s - loss: 0.0645 - acc: 0.9868 - val_loss: 0.5561 - val_acc: 0.9055
Epoch 988/1000
2122/2122 [==============================] - 4s - loss: 0.0269 - acc: 0.9934 - val_loss: 0.5650 - val_acc: 0.9099
Epoch 989/1000
2122/2122 [==============================] - 4s - loss: 0.0100 - acc: 0.9976 - val_loss: 0.6112 - val_acc: 0.9121
Epoch 990/1000
2122/2122 [==============================] - 4s - loss: 0.0202 - acc: 0.9943 - val_loss: 0.6758 - val_acc: 0.9022
Epoch 991/1000
2122/2122 [==============================] - 4s - loss: 0.1156 - acc: 0.9750 - val_loss: 1.1665 - val_acc: 0.8462
Epoch 992/1000
2122/2122 [==============================] - 4s - loss: 0.0755 - acc: 0.9877 - val_loss: 0.8427 - val_acc: 0.8802
Epoch 993/1000
2122/2122 [==============================] - 4s - loss: 0.0829 - acc: 0.9877 - val_loss: 0.9999 - val_acc: 0.8945
Epoch 994/1000
2122/2122 [==============================] - 4s - loss: 0.0242 - acc: 0.9953 - val_loss: 0.9255 - val_acc: 0.8989
Epoch 995/1000
2122/2122 [==============================] - 4s - loss: 0.0185 - acc: 0.9967 - val_loss: 0.7295 - val_acc: 0.8593
Epoch 996/1000
2122/2122 [==============================] - 4s - loss: 0.0542 - acc: 0.9877 - val_loss: 0.6018 - val_acc: 0.9055
Epoch 997/1000
2122/2122 [==============================] - 4s - loss: 0.0398 - acc: 0.9910 - val_loss: 0.6653 - val_acc: 0.8901
Epoch 998/1000
2122/2122 [==============================] - 4s - loss: 0.0230 - acc: 0.9943 - val_loss: 0.8072 - val_acc: 0.8846
Epoch 999/1000
2122/2122 [==============================] - 4s - loss: 0.0964 - acc: 0.9873 - val_loss: 0.8299 - val_acc: 0.8956
Epoch 1000/1000
2122/2122 [==============================] - 4s - loss: 0.0747 - acc: 0.9887 - val_loss: 0.8779 - val_acc: 0.8846
2017-07-31T09:24:30.367020

















In [34]:



    


train_loss, train_accuracy = model.evaluate(X_train, y_train, batch_size=500)
train_loss, train_accuracy



    


















    






303/303 [==============================] - 1s










    
Out[34]:








(0.072976924479007721, 0.99009901285171509)

















In [35]:



    


test_loss, test_accuracy = model.evaluate(X_test, y_test, batch_size=BATCH_SIZE)
test_loss, test_accuracy



    


















    






76/76 [==============================] - 0s










    
Out[35]:








(0.12756867706775665, 0.98684209585189819)

















In [36]:



    


evaluation_loss, evaluation_accuracy = model.evaluate(evaluation_X, evaluation_y, batch_size=BATCH_SIZE)
evaluation_loss, evaluation_accuracy



    


















    






3790/3790 [==============================] - 5s     










    
Out[36]:








(11.690376923392504, 0.21820580361700939)

















In [37]:



    


!mkdir models



    


















    






mkdir: cannot create directory ‘models’: File exists

















In [38]:



    


# model.save('models/conv-vgg-augmented.hdf5')
model.save('models/conv-vgg-original.hdf5')



    











In [39]:



    


!ls -lh models



    


















    






total 35M
-rw-rw-r-- 1 ubuntu ubuntu 18M Jul 21 19:56 conv-vgg-augmented.hdf5
-rw-rw-r-- 1 ubuntu ubuntu 18M Jul 31 07:54 conv-vgg-original.hdf5

















In [33]:



    


# !curl --upload-file ./models/conv-vgg-augmented.hdf5 https://transfer.sh/conv-vgg-augmented.hdf5
!curl --upload-file ./models/conv-vgg-original.hdf5 https://transfer.sh/conv-vgg-original.hdf5



    


















    






https://transfer.sh/K5RH8/conv-vgg-augmented.hdf5
















In [ ]:

Content source: DJCordhose/speed-limit-signs

Similar notebooks:

notebook.community | gallery | about