sdcubber

test_seq = KagglePlanetSequence('./KagglePlanetMCML.csv',
                       './data/train/',
                       im_size=IM_SIZE,
                       batch_size=32, mode='test')
predictions = model.predict_generator(generator=test_seq, verbose=1)
len(predictions[1])  == len(df_train) # This is True!

sdcubber

len(predictions[1])  == len(df_train) # Total number of images in dataset

sdcubber

featdef = {
           'image': tf.FixedLenFeature(shape=[], dtype=tf.string),
           'weather_labels': tf.FixedLenFeature(shape=[], dtype=tf.string),
           'ground_labels': tf.FixedLenFeature(shape=[], dtype=tf.string)
          }
          
 def _parse_record(example_proto, clip=False):
    """Parse a single record into image, weather labels, ground labels"""
    example = tf.parse_single_example(example_proto, featdef)
    im = tf.decode_raw(example['image'], tf.float32)

sdcubber

with tf.Session() as sess:
    tf.keras.backend.set_session(sess)
    sess.run(ds_tr_init) # initialize the generator
  
    # Rewire network to tie it into the generator
    image_input = tf.keras.Input(tensor=x)

    # Model definition
    ...

sdcubber

import tensorflow as tf

# The export path contains the name and the version of the model
tf.keras.backend.set_learning_phase(0) # Ignore dropout at inference
model = tf.keras.models.load_model('./model.h5')
export_path = './PlanetModel/1'

# Fetch the Keras session and save the model
# The signature definition is defined by the input and output tensors
# And stored with the default serving key

sdcubber

# Serialize images, together with labels, to TF records
def _bytes_feature(value):
    return tf.train.Feature(bytes_list=tf.train.BytesList(value=[value]))

tf_records_filename = './data/KagglePlanetTFRecord_{}'.format(IM_SIZE)
writer = tf.python_io.TFRecordWriter(tf_records_filename)

# List of image paths, np array of labels
im_list = [os.path.join('./data/train', v + '.jpg') for v in df_train['image_name'].tolist()]
w_labels_arr = np.array([ast.literal_eval(l) for l in df_train['weather_labels']])

sdcubber

model = tf.keras.Model(inputs=image_input, outputs=[weather_output, ground_output])

model.compile(optimizer='adam',
              loss={'weather': 'categorical_crossentropy',
                    'ground': 'binary_crossentropy'})

history = model.fit(ds_train, 
                    steps_per_epoch=100, # let's take just a couple of steps
                    epochs=1)

sdcubber

from keras import models
from keras import layers

network = models.Sequential()

network.add(layers.Dense(512, activation='relu', input_shape=(28 * 28, )))

network.add(layers.Dense(256, activation='relu'))
network.add(...)

sdcubber

# download data
from keras.datasets import mnist
(train_images, train_labels), (val_images, val_labels) = mnist.load_data()

# flatten and scale images
train_images = train_images.reshape((-1, 28*28)).astype('float32') / 255
val_images = val_images.reshape((-1, 28*28)).astype('float32') / 255

# one-hot encode the labels
train_labels, val_labels = keras.utils.to_categorical(train_labels), keras.utils.to_categorical(val_labels)

sdcubber

history = network.fit(train_images, train_labels, epochs=5, batch_size=128, validation_data=(val_images, val_labels))