In [1]:
from __future__ import division, print_function
%matplotlib inline
from importlib import reload # Python 3
import utils; reload(utils)
from utils import *
In [2]:
path = "data/dogscats/"
# path = "data/dogscats/sample/"
model_path = path + 'models/'
if not os.path.exists(model_path): os.mkdir(model_path)
batch_size=32
# batch_size=1
In [3]:
batches = get_batches(path+'train', shuffle=False, batch_size=batch_size)
val_batches = get_batches(path+'valid', shuffle=False, batch_size=batch_size)
In [4]:
(val_classes, trn_classes, val_labels, trn_labels,
val_filenames, filenames, test_filenames) = get_classes(path)
In this notebook we're going to create an ensemble of models and use their average as our predictions. For each ensemble, we're going to follow our usual fine-tuning steps:
1) Create a model that retrains just the last layer 2) Add this to a model containing all VGG layers except the last layer 3) Fine-tune just the dense layers of this model (pre-computing the convolutional layers) 4) Add data augmentation, fine-tuning the dense layers without pre-computation.
So first, we need to create our VGG model and pre-compute the output of the conv layers:
In [5]:
model = Vgg16().model
conv_layers,fc_layers = split_at(model, Convolution2D)
In [6]:
conv_model = Sequential(conv_layers)
In [7]:
val_features = conv_model.predict_generator(val_batches, int(np.ceil(val_batches.samples/batch_size)))
trn_features = conv_model.predict_generator(batches, int(np.ceil(batches.samples/batch_size)))
In [8]:
save_array(model_path + 'train_convlayer_features.bc', trn_features)
save_array(model_path + 'valid_convlayer_features.bc', val_features)
In the future we can just load these precomputed features:
In [9]:
trn_features = load_array(model_path+'train_convlayer_features.bc')
val_features = load_array(model_path+'valid_convlayer_features.bc')
We can also save some time by pre-computing the training and validation arrays with the image decoding and resizing already done:
In [10]:
trn = get_data(path+'train')
val = get_data(path+'valid')
In [11]:
save_array(model_path+'train_data.bc', trn)
save_array(model_path+'valid_data.bc', val)
In the future we can just load these resized images:
In [12]:
trn = load_array(model_path+'train_data.bc')
val = load_array(model_path+'valid_data.bc')
Finally, we can precompute the output of all but the last dropout and dense layers, for creating the first stage of the model:
In [13]:
model.pop()
model.pop()
In [14]:
ll_val_feat = model.predict_generator(val_batches, int(np.ceil(val_batches.samples/batch_size)))
ll_feat = model.predict_generator(batches, int(np.ceil(batches.samples/batch_size)))
In [15]:
save_array(model_path + 'train_ll_feat.bc', ll_feat)
save_array(model_path + 'valid_ll_feat.bc', ll_val_feat)
In [16]:
ll_feat = load_array(model_path+ 'train_ll_feat.bc')
ll_val_feat = load_array(model_path + 'valid_ll_feat.bc')
...and let's also grab the test data, for when we need to submit:
In [17]:
test = get_data(path+'test')
save_array(model_path+'test_data.bc', test)
In [18]:
test = load_array(model_path+'test_data.bc')
The functions automate creating a model that trains the last layer from scratch, and then adds those new layers on to the main model.
In [19]:
def get_ll_layers():
return [
BatchNormalization(input_shape=(4096,)),
Dropout(0.5),
Dense(2, activation='softmax')
]
In [20]:
def train_last_layer(i):
ll_layers = get_ll_layers()
ll_model = Sequential(ll_layers)
ll_model.compile(optimizer=Adam(), loss='categorical_crossentropy', metrics=['accuracy'])
ll_model.optimizer.lr=1e-5
ll_model.fit(ll_feat, trn_labels, validation_data=(ll_val_feat, val_labels), epochs=12)
ll_model.optimizer.lr=1e-7
ll_model.fit(ll_feat, trn_labels, validation_data=(ll_val_feat, val_labels), epochs=1)
ll_model.save_weights(model_path+'ll_bn' + i + '.h5')
vgg = Vgg16BN()
model = vgg.model
model.pop(); model.pop(); model.pop()
for layer in model.layers: layer.trainable=False
model.compile(optimizer=Adam(), loss='categorical_crossentropy', metrics=['accuracy'])
ll_layers = get_ll_layers()
for layer in ll_layers: model.add(layer)
for l1,l2 in zip(ll_model.layers, model.layers[-3:]):
l2.set_weights(l1.get_weights())
model.compile(optimizer=Adam(), loss='categorical_crossentropy', metrics=['accuracy'])
model.save_weights(model_path+'bn' + i + '.h5')
return model
In [21]:
def get_conv_model(model):
layers = model.layers
last_conv_idx = [index for index,layer in enumerate(layers)
if type(layer) is Convolution2D][-1]
conv_layers = layers[:last_conv_idx+1]
conv_model = Sequential(conv_layers)
fc_layers = layers[last_conv_idx+1:]
return conv_model, fc_layers, last_conv_idx
In [22]:
def get_fc_layers(p, in_shape):
return [
MaxPooling2D(input_shape=in_shape),
Flatten(),
Dense(4096, activation='relu'),
BatchNormalization(),
Dropout(p),
Dense(4096, activation='relu'),
BatchNormalization(),
Dropout(p),
Dense(2, activation='softmax')
]
In [23]:
def train_dense_layers(i, model):
conv_model, fc_layers, last_conv_idx = get_conv_model(model)
conv_shape = conv_model.output_shape[1:]
fc_model = Sequential(get_fc_layers(0.5, conv_shape))
for l1,l2 in zip(fc_model.layers, fc_layers):
weights = l2.get_weights()
l1.set_weights(weights)
fc_model.compile(optimizer=Adam(1e-5), loss='categorical_crossentropy',
metrics=['accuracy'])
fc_model.fit(trn_features, trn_labels, epochs=2,
batch_size=batch_size, validation_data=(val_features, val_labels))
# width_zoom_range removed from the following because not available in Keras2
gen = image.ImageDataGenerator(rotation_range=10, width_shift_range=0.05, zoom_range=0.05,
channel_shift_range=10, height_shift_range=0.05, shear_range=0.05, horizontal_flip=True)
batches = gen.flow(trn, trn_labels, batch_size=batch_size)
val_batches = image.ImageDataGenerator().flow(val, val_labels,
shuffle=False, batch_size=batch_size)
for layer in conv_model.layers: layer.trainable = False
for layer in get_fc_layers(0.5, conv_shape): conv_model.add(layer)
for l1,l2 in zip(conv_model.layers[last_conv_idx+1:], fc_model.layers):
l1.set_weights(l2.get_weights())
steps_per_epoch = int(np.ceil(batches.n/batch_size))
validation_steps = int(np.ceil(val_batches.n/batch_size))
conv_model.compile(optimizer=Adam(1e-5), loss='categorical_crossentropy',
metrics=['accuracy'])
conv_model.save_weights(model_path+'no_dropout_bn' + i + '.h5')
conv_model.fit_generator(batches, steps_per_epoch=steps_per_epoch, epochs=1,
validation_data=val_batches, validation_steps=validation_steps)
for layer in conv_model.layers[16:]: layer.trainable = True
#- added again the compile instruction in order to avoid a Keras 2.1 warning message
conv_model.compile(optimizer=Adam(1e-5), loss='categorical_crossentropy',
metrics=['accuracy'])
conv_model.fit_generator(batches, steps_per_epoch=steps_per_epoch, epochs=8,
validation_data=val_batches, validation_steps=validation_steps)
conv_model.optimizer.lr = 1e-7
conv_model.fit_generator(batches, steps_per_epoch=steps_per_epoch, epochs=10,
validation_data=val_batches, validation_steps=validation_steps)
conv_model.save_weights(model_path + 'aug' + i + '.h5')
In [24]:
for i in range(5):
i = str(i)
model = train_last_layer(i)
train_dense_layers(i, model)
In [25]:
ens_model = vgg_ft_bn(2)
for layer in ens_model.layers: layer.trainable=True
In [26]:
def get_ens_pred(arr, fname):
ens_pred = []
for i in range(5):
i = str(i)
ens_model.load_weights('{}{}{}.h5'.format(model_path, fname, i))
preds = ens_model.predict(arr, batch_size=batch_size)
ens_pred.append(preds)
return ens_pred
In [27]:
val_pred2 = get_ens_pred(val, 'aug')
In [28]:
val_avg_preds2 = np.stack(val_pred2).mean(axis=0)
In [29]:
categorical_accuracy(val_labels, val_avg_preds2).eval().mean()
Out[29]:
In [ ]: