Training and Evaluating Xception Model
Imports
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import tensorflow as tf
import numpy as np
import os
import matplotlib.pyplot as pltmodels = tf.contrib.keras.models
layers = tf.contrib.keras.layers
utils = tf.contrib.keras.utils
losses = tf.contrib.keras.losses
optimizers = tf.contrib.keras.optimizers
metrics = tf.contrib.keras.metrics
preprocessing_image = tf.contrib.keras.preprocessing.image
applications = tf.contrib.keras.applicationsLoad Pre-Trained Xception Model
# load pre-trained Xception model and exclude top dense layer
base_model = applications.Xception(include_top=False,
weights='imagenet',
input_shape=(299,299,3),
pooling='avg')print("Model input shape: {}\n".format(base_model.input_shape))
print("Model output shape: {}\n".format(base_model.output_shape))
print("Model number of layers: {}\n".format(len(base_model.layers)))Model input shape: (None, 299, 299, 3)
Model output shape: (None, 2048)
Model number of layers: 133Fine-tune Xception Model
def fine_tune_Xception(base_model):
# output of convolutional layers
x = base_model.output
# final Dense layer
outputs = layers.Dense(4, activation='softmax')(x)
# define model with base_model's input
model = models.Model(inputs=base_model.input, outputs=outputs)
# freeze weights of early layers
# to ease training
for layer in model.layers[:40]:
layer.trainable = False
return modelCompile Model
def compile_model(model):
# loss
loss = losses.categorical_crossentropy
# optimizer
optimizer = optimizers.RMSprop(lr=0.0001)
# metrics
metric = [metrics.categorical_accuracy]
# compile model with loss, optimizer, and evaluation metrics
model.compile(optimizer, loss, metric)
return modelInspect Model Architecture
model = fine_tune_Xception(base_model)
model = compile_model(model)
model.summary()Layer (type) Output Shape Param # Connected to
====================================================================================================
input_1 (InputLayer) (None, 299, 299, 3) 0
____________________________________________________________________________________________________
block1_conv1 (Conv2D) (None, 149, 149, 32) 864 input_1[0][0]
____________________________________________________________________________________________________
block1_conv1_bn (BatchNormalizat (None, 149, 149, 32) 128 block1_conv1[0][0]
____________________________________________________________________________________________________
block1_conv1_act (Activation) (None, 149, 149, 32) 0 block1_conv1_bn[0][0]
____________________________________________________________________________________________________
block1_conv2 (Conv2D) (None, 147, 147, 64) 18432 block1_conv1_act[0][0]
____________________________________________________________________________________________________
block1_conv2_bn (BatchNormalizat (None, 147, 147, 64) 256 block1_conv2[0][0]
____________________________________________________________________________________________________
block1_conv2_act (Activation) (None, 147, 147, 64) 0 block1_conv2_bn[0][0]
____________________________________________________________________________________________________
block2_sepconv1 (SeparableConv2D (None, 147, 147, 128) 8768 block1_conv2_act[0][0]
____________________________________________________________________________________________________
block2_sepconv1_bn (BatchNormali (None, 147, 147, 128) 512 block2_sepconv1[0][0]
____________________________________________________________________________________________________
block2_sepconv2_act (Activation) (None, 147, 147, 128) 0 block2_sepconv1_bn[0][0]
____________________________________________________________________________________________________
block2_sepconv2 (SeparableConv2D (None, 147, 147, 128) 17536 block2_sepconv2_act[0][0]
____________________________________________________________________________________________________
block2_sepconv2_bn (BatchNormali (None, 147, 147, 128) 512 block2_sepconv2[0][0]
____________________________________________________________________________________________________
conv2d_1 (Conv2D) (None, 74, 74, 128) 8192 block1_conv2_act[0][0]
____________________________________________________________________________________________________
block2_pool (MaxPooling2D) (None, 74, 74, 128) 0 block2_sepconv2_bn[0][0]
____________________________________________________________________________________________________
batch_normalization_1 (BatchNorm (None, 74, 74, 128) 512 conv2d_1[0][0]
____________________________________________________________________________________________________
add_1 (Add) (None, 74, 74, 128) 0 block2_pool[0][0]
batch_normalization_1[0][0]
____________________________________________________________________________________________________
block3_sepconv1_act (Activation) (None, 74, 74, 128) 0 add_1[0][0]
____________________________________________________________________________________________________
block3_sepconv1 (SeparableConv2D (None, 74, 74, 256) 33920 block3_sepconv1_act[0][0]
____________________________________________________________________________________________________
block3_sepconv1_bn (BatchNormali (None, 74, 74, 256) 1024 block3_sepconv1[0][0]
____________________________________________________________________________________________________
block3_sepconv2_act (Activation) (None, 74, 74, 256) 0 block3_sepconv1_bn[0][0]
____________________________________________________________________________________________________
block3_sepconv2 (SeparableConv2D (None, 74, 74, 256) 67840 block3_sepconv2_act[0][0]
____________________________________________________________________________________________________
block3_sepconv2_bn (BatchNormali (None, 74, 74, 256) 1024 block3_sepconv2[0][0]
____________________________________________________________________________________________________
conv2d_2 (Conv2D) (None, 37, 37, 256) 32768 add_1[0][0]
____________________________________________________________________________________________________
block3_pool (MaxPooling2D) (None, 37, 37, 256) 0 block3_sepconv2_bn[0][0]
____________________________________________________________________________________________________
batch_normalization_2 (BatchNorm (None, 37, 37, 256) 1024 conv2d_2[0][0]
____________________________________________________________________________________________________
add_2 (Add) (None, 37, 37, 256) 0 block3_pool[0][0]
batch_normalization_2[0][0]
____________________________________________________________________________________________________
block4_sepconv1_act (Activation) (None, 37, 37, 256) 0 add_2[0][0]
____________________________________________________________________________________________________
block4_sepconv1 (SeparableConv2D (None, 37, 37, 728) 188672 block4_sepconv1_act[0][0]
____________________________________________________________________________________________________
block4_sepconv1_bn (BatchNormali (None, 37, 37, 728) 2912 block4_sepconv1[0][0]
____________________________________________________________________________________________________
block4_sepconv2_act (Activation) (None, 37, 37, 728) 0 block4_sepconv1_bn[0][0]
____________________________________________________________________________________________________
block4_sepconv2 (SeparableConv2D (None, 37, 37, 728) 536536 block4_sepconv2_act[0][0]
____________________________________________________________________________________________________
block4_sepconv2_bn (BatchNormali (None, 37, 37, 728) 2912 block4_sepconv2[0][0]
____________________________________________________________________________________________________
conv2d_3 (Conv2D) (None, 19, 19, 728) 186368 add_2[0][0]
____________________________________________________________________________________________________
block4_pool (MaxPooling2D) (None, 19, 19, 728) 0 block4_sepconv2_bn[0][0]
____________________________________________________________________________________________________
batch_normalization_3 (BatchNorm (None, 19, 19, 728) 2912 conv2d_3[0][0]
____________________________________________________________________________________________________
add_3 (Add) (None, 19, 19, 728) 0 block4_pool[0][0]
batch_normalization_3[0][0]
____________________________________________________________________________________________________
block5_sepconv1_act (Activation) (None, 19, 19, 728) 0 add_3[0][0]
____________________________________________________________________________________________________
block5_sepconv1 (SeparableConv2D (None, 19, 19, 728) 536536 block5_sepconv1_act[0][0]
____________________________________________________________________________________________________
block5_sepconv1_bn (BatchNormali (None, 19, 19, 728) 2912 block5_sepconv1[0][0]
____________________________________________________________________________________________________
block5_sepconv2_act (Activation) (None, 19, 19, 728) 0 block5_sepconv1_bn[0][0]
____________________________________________________________________________________________________
block5_sepconv2 (SeparableConv2D (None, 19, 19, 728) 536536 block5_sepconv2_act[0][0]
____________________________________________________________________________________________________
block5_sepconv2_bn (BatchNormali (None, 19, 19, 728) 2912 block5_sepconv2[0][0]
____________________________________________________________________________________________________
block5_sepconv3_act (Activation) (None, 19, 19, 728) 0 block5_sepconv2_bn[0][0]
____________________________________________________________________________________________________
block5_sepconv3 (SeparableConv2D (None, 19, 19, 728) 536536 block5_sepconv3_act[0][0]
____________________________________________________________________________________________________
block5_sepconv3_bn (BatchNormali (None, 19, 19, 728) 2912 block5_sepconv3[0][0]
____________________________________________________________________________________________________
add_4 (Add) (None, 19, 19, 728) 0 block5_sepconv3_bn[0][0]
add_3[0][0]
____________________________________________________________________________________________________
block6_sepconv1_act (Activation) (None, 19, 19, 728) 0 add_4[0][0]
____________________________________________________________________________________________________
block6_sepconv1 (SeparableConv2D (None, 19, 19, 728) 536536 block6_sepconv1_act[0][0]
____________________________________________________________________________________________________
block6_sepconv1_bn (BatchNormali (None, 19, 19, 728) 2912 block6_sepconv1[0][0]
____________________________________________________________________________________________________
block6_sepconv2_act (Activation) (None, 19, 19, 728) 0 block6_sepconv1_bn[0][0]
____________________________________________________________________________________________________
block6_sepconv2 (SeparableConv2D (None, 19, 19, 728) 536536 block6_sepconv2_act[0][0]
____________________________________________________________________________________________________
block6_sepconv2_bn (BatchNormali (None, 19, 19, 728) 2912 block6_sepconv2[0][0]
____________________________________________________________________________________________________
block6_sepconv3_act (Activation) (None, 19, 19, 728) 0 block6_sepconv2_bn[0][0]
____________________________________________________________________________________________________
block6_sepconv3 (SeparableConv2D (None, 19, 19, 728) 536536 block6_sepconv3_act[0][0]
____________________________________________________________________________________________________
block6_sepconv3_bn (BatchNormali (None, 19, 19, 728) 2912 block6_sepconv3[0][0]
____________________________________________________________________________________________________
add_5 (Add) (None, 19, 19, 728) 0 block6_sepconv3_bn[0][0]
add_4[0][0]
____________________________________________________________________________________________________
block7_sepconv1_act (Activation) (None, 19, 19, 728) 0 add_5[0][0]
____________________________________________________________________________________________________
block7_sepconv1 (SeparableConv2D (None, 19, 19, 728) 536536 block7_sepconv1_act[0][0]
____________________________________________________________________________________________________
block7_sepconv1_bn (BatchNormali (None, 19, 19, 728) 2912 block7_sepconv1[0][0]
____________________________________________________________________________________________________
block7_sepconv2_act (Activation) (None, 19, 19, 728) 0 block7_sepconv1_bn[0][0]
____________________________________________________________________________________________________
block7_sepconv2 (SeparableConv2D (None, 19, 19, 728) 536536 block7_sepconv2_act[0][0]
____________________________________________________________________________________________________
block7_sepconv2_bn (BatchNormali (None, 19, 19, 728) 2912 block7_sepconv2[0][0]
____________________________________________________________________________________________________
block7_sepconv3_act (Activation) (None, 19, 19, 728) 0 block7_sepconv2_bn[0][0]
____________________________________________________________________________________________________
block7_sepconv3 (SeparableConv2D (None, 19, 19, 728) 536536 block7_sepconv3_act[0][0]
____________________________________________________________________________________________________
block7_sepconv3_bn (BatchNormali (None, 19, 19, 728) 2912 block7_sepconv3[0][0]
____________________________________________________________________________________________________
add_6 (Add) (None, 19, 19, 728) 0 block7_sepconv3_bn[0][0]
add_5[0][0]
____________________________________________________________________________________________________
block8_sepconv1_act (Activation) (None, 19, 19, 728) 0 add_6[0][0]
____________________________________________________________________________________________________
block8_sepconv1 (SeparableConv2D (None, 19, 19, 728) 536536 block8_sepconv1_act[0][0]
____________________________________________________________________________________________________
block8_sepconv1_bn (BatchNormali (None, 19, 19, 728) 2912 block8_sepconv1[0][0]
____________________________________________________________________________________________________
block8_sepconv2_act (Activation) (None, 19, 19, 728) 0 block8_sepconv1_bn[0][0]
____________________________________________________________________________________________________
block8_sepconv2 (SeparableConv2D (None, 19, 19, 728) 536536 block8_sepconv2_act[0][0]
____________________________________________________________________________________________________
block8_sepconv2_bn (BatchNormali (None, 19, 19, 728) 2912 block8_sepconv2[0][0]
____________________________________________________________________________________________________
block8_sepconv3_act (Activation) (None, 19, 19, 728) 0 block8_sepconv2_bn[0][0]
____________________________________________________________________________________________________
block8_sepconv3 (SeparableConv2D (None, 19, 19, 728) 536536 block8_sepconv3_act[0][0]
____________________________________________________________________________________________________
block8_sepconv3_bn (BatchNormali (None, 19, 19, 728) 2912 block8_sepconv3[0][0]
____________________________________________________________________________________________________
add_7 (Add) (None, 19, 19, 728) 0 block8_sepconv3_bn[0][0]
add_6[0][0]
____________________________________________________________________________________________________
block9_sepconv1_act (Activation) (None, 19, 19, 728) 0 add_7[0][0]
____________________________________________________________________________________________________
block9_sepconv1 (SeparableConv2D (None, 19, 19, 728) 536536 block9_sepconv1_act[0][0]
____________________________________________________________________________________________________
block9_sepconv1_bn (BatchNormali (None, 19, 19, 728) 2912 block9_sepconv1[0][0]
____________________________________________________________________________________________________
block9_sepconv2_act (Activation) (None, 19, 19, 728) 0 block9_sepconv1_bn[0][0]
____________________________________________________________________________________________________
block9_sepconv2 (SeparableConv2D (None, 19, 19, 728) 536536 block9_sepconv2_act[0][0]
____________________________________________________________________________________________________
block9_sepconv2_bn (BatchNormali (None, 19, 19, 728) 2912 block9_sepconv2[0][0]
____________________________________________________________________________________________________
block9_sepconv3_act (Activation) (None, 19, 19, 728) 0 block9_sepconv2_bn[0][0]
____________________________________________________________________________________________________
block9_sepconv3 (SeparableConv2D (None, 19, 19, 728) 536536 block9_sepconv3_act[0][0]
____________________________________________________________________________________________________
block9_sepconv3_bn (BatchNormali (None, 19, 19, 728) 2912 block9_sepconv3[0][0]
____________________________________________________________________________________________________
add_8 (Add) (None, 19, 19, 728) 0 block9_sepconv3_bn[0][0]
add_7[0][0]
____________________________________________________________________________________________________
block10_sepconv1_act (Activation (None, 19, 19, 728) 0 add_8[0][0]
____________________________________________________________________________________________________
block10_sepconv1 (SeparableConv2 (None, 19, 19, 728) 536536 block10_sepconv1_act[0][0]
____________________________________________________________________________________________________
block10_sepconv1_bn (BatchNormal (None, 19, 19, 728) 2912 block10_sepconv1[0][0]
____________________________________________________________________________________________________
block10_sepconv2_act (Activation (None, 19, 19, 728) 0 block10_sepconv1_bn[0][0]
____________________________________________________________________________________________________
block10_sepconv2 (SeparableConv2 (None, 19, 19, 728) 536536 block10_sepconv2_act[0][0]
____________________________________________________________________________________________________
block10_sepconv2_bn (BatchNormal (None, 19, 19, 728) 2912 block10_sepconv2[0][0]
____________________________________________________________________________________________________
block10_sepconv3_act (Activation (None, 19, 19, 728) 0 block10_sepconv2_bn[0][0]
____________________________________________________________________________________________________
block10_sepconv3 (SeparableConv2 (None, 19, 19, 728) 536536 block10_sepconv3_act[0][0]
____________________________________________________________________________________________________
block10_sepconv3_bn (BatchNormal (None, 19, 19, 728) 2912 block10_sepconv3[0][0]
____________________________________________________________________________________________________
add_9 (Add) (None, 19, 19, 728) 0 block10_sepconv3_bn[0][0]
add_8[0][0]
____________________________________________________________________________________________________
block11_sepconv1_act (Activation (None, 19, 19, 728) 0 add_9[0][0]
____________________________________________________________________________________________________
block11_sepconv1 (SeparableConv2 (None, 19, 19, 728) 536536 block11_sepconv1_act[0][0]
____________________________________________________________________________________________________
block11_sepconv1_bn (BatchNormal (None, 19, 19, 728) 2912 block11_sepconv1[0][0]
____________________________________________________________________________________________________
block11_sepconv2_act (Activation (None, 19, 19, 728) 0 block11_sepconv1_bn[0][0]
____________________________________________________________________________________________________
block11_sepconv2 (SeparableConv2 (None, 19, 19, 728) 536536 block11_sepconv2_act[0][0]
____________________________________________________________________________________________________
block11_sepconv2_bn (BatchNormal (None, 19, 19, 728) 2912 block11_sepconv2[0][0]
____________________________________________________________________________________________________
block11_sepconv3_act (Activation (None, 19, 19, 728) 0 block11_sepconv2_bn[0][0]
____________________________________________________________________________________________________
block11_sepconv3 (SeparableConv2 (None, 19, 19, 728) 536536 block11_sepconv3_act[0][0]
____________________________________________________________________________________________________
block11_sepconv3_bn (BatchNormal (None, 19, 19, 728) 2912 block11_sepconv3[0][0]
____________________________________________________________________________________________________
add_10 (Add) (None, 19, 19, 728) 0 block11_sepconv3_bn[0][0]
add_9[0][0]
____________________________________________________________________________________________________
block12_sepconv1_act (Activation (None, 19, 19, 728) 0 add_10[0][0]
____________________________________________________________________________________________________
block12_sepconv1 (SeparableConv2 (None, 19, 19, 728) 536536 block12_sepconv1_act[0][0]
____________________________________________________________________________________________________
block12_sepconv1_bn (BatchNormal (None, 19, 19, 728) 2912 block12_sepconv1[0][0]
____________________________________________________________________________________________________
block12_sepconv2_act (Activation (None, 19, 19, 728) 0 block12_sepconv1_bn[0][0]
____________________________________________________________________________________________________
block12_sepconv2 (SeparableConv2 (None, 19, 19, 728) 536536 block12_sepconv2_act[0][0]
____________________________________________________________________________________________________
block12_sepconv2_bn (BatchNormal (None, 19, 19, 728) 2912 block12_sepconv2[0][0]
____________________________________________________________________________________________________
block12_sepconv3_act (Activation (None, 19, 19, 728) 0 block12_sepconv2_bn[0][0]
____________________________________________________________________________________________________
block12_sepconv3 (SeparableConv2 (None, 19, 19, 728) 536536 block12_sepconv3_act[0][0]
____________________________________________________________________________________________________
block12_sepconv3_bn (BatchNormal (None, 19, 19, 728) 2912 block12_sepconv3[0][0]
____________________________________________________________________________________________________
add_11 (Add) (None, 19, 19, 728) 0 block12_sepconv3_bn[0][0]
add_10[0][0]
____________________________________________________________________________________________________
block13_sepconv1_act (Activation (None, 19, 19, 728) 0 add_11[0][0]
____________________________________________________________________________________________________
block13_sepconv1 (SeparableConv2 (None, 19, 19, 728) 536536 block13_sepconv1_act[0][0]
____________________________________________________________________________________________________
block13_sepconv1_bn (BatchNormal (None, 19, 19, 728) 2912 block13_sepconv1[0][0]
____________________________________________________________________________________________________
block13_sepconv2_act (Activation (None, 19, 19, 728) 0 block13_sepconv1_bn[0][0]
____________________________________________________________________________________________________
block13_sepconv2 (SeparableConv2 (None, 19, 19, 1024) 752024 block13_sepconv2_act[0][0]
____________________________________________________________________________________________________
block13_sepconv2_bn (BatchNormal (None, 19, 19, 1024) 4096 block13_sepconv2[0][0]
____________________________________________________________________________________________________
conv2d_4 (Conv2D) (None, 10, 10, 1024) 745472 add_11[0][0]
____________________________________________________________________________________________________
block13_pool (MaxPooling2D) (None, 10, 10, 1024) 0 block13_sepconv2_bn[0][0]
____________________________________________________________________________________________________
batch_normalization_4 (BatchNorm (None, 10, 10, 1024) 4096 conv2d_4[0][0]
____________________________________________________________________________________________________
add_12 (Add) (None, 10, 10, 1024) 0 block13_pool[0][0]
batch_normalization_4[0][0]
____________________________________________________________________________________________________
block14_sepconv1 (SeparableConv2 (None, 10, 10, 1536) 1582080 add_12[0][0]
____________________________________________________________________________________________________
block14_sepconv1_bn (BatchNormal (None, 10, 10, 1536) 6144 block14_sepconv1[0][0]
____________________________________________________________________________________________________
block14_sepconv1_act (Activation (None, 10, 10, 1536) 0 block14_sepconv1_bn[0][0]
____________________________________________________________________________________________________
block14_sepconv2 (SeparableConv2 (None, 10, 10, 2048) 3159552 block14_sepconv1_act[0][0]
____________________________________________________________________________________________________
block14_sepconv2_bn (BatchNormal (None, 10, 10, 2048) 8192 block14_sepconv2[0][0]
____________________________________________________________________________________________________
block14_sepconv2_act (Activation (None, 10, 10, 2048) 0 block14_sepconv2_bn[0][0]
____________________________________________________________________________________________________
global_average_pooling2d_1 (Glob (None, 2048) 0 block14_sepconv2_act[0][0]
____________________________________________________________________________________________________
dense_2 (Dense) (None, 4) 8196 global_average_pooling2d_1[0][0]
====================================================================================================
Total params: 20,869,676
Trainable params: 19,170,396
Non-trainable params: 1,699,280
____________________________________________________________________________________________________Image Preprocessing And Augmentation
def preprocess_image(x):
x /= 255.
x -= 0.5
x *= 2.
# 'RGB'->'BGR'
x = x[..., ::-1]
# Zero-center by mean pixel
x[..., 0] -= 103.939
x[..., 1] -= 116.779
x[..., 2] -= 123.68
return x
train_datagen = preprocessing_image.ImageDataGenerator(
preprocessing_function=preprocess_image,
shear_range=0.2,
zoom_range=0.2,
horizontal_flip=True)
test_datagen = preprocessing_image.ImageDataGenerator(preprocessing_function=preprocess_image)BASE_DIR = "/Users/marvinbertin/Github/marvin/ImageNet_Utils"
train_generator = train_datagen.flow_from_directory(
os.path.join(BASE_DIR, "imageNet_dataset/train"),
target_size=(299, 299),
batch_size=32,
class_mode='categorical',
shuffle=True)
validation_generator = test_datagen.flow_from_directory(
os.path.join(BASE_DIR, "imageNet_dataset/validation"),
target_size=(299, 299),
batch_size=32,
class_mode='categorical',
shuffle=True)Found 2677 images belonging to 4 classes.
Found 668 images belonging to 4 classes.Train Model on ImageNet Dataset
history = model.fit_generator(
train_generator,
steps_per_epoch=80,
epochs=10,
validation_data=validation_generator,
validation_steps=20)Epoch 1/10
80/80 [==============================] - 121s - loss: 0.9725 - categorical_accuracy: 0.5895 - val_loss: 12.2145 - val_categorical_accuracy: 0.2422
Epoch 2/10
80/80 [==============================] - 119s - loss: 0.8422 - categorical_accuracy: 0.6576 - val_loss: 3.3612 - val_categorical_accuracy: 0.2441
Epoch 3/10
80/80 [==============================] - 118s - loss: 0.8019 - categorical_accuracy: 0.6707 - val_loss: 1.3862 - val_categorical_accuracy: 0.2220
Epoch 4/10
80/80 [==============================] - 119s - loss: 0.7753 - categorical_accuracy: 0.6800 - val_loss: 4.0558 - val_categorical_accuracy: 0.1732
Epoch 5/10
80/80 [==============================] - 119s - loss: 0.7529 - categorical_accuracy: 0.6950 - val_loss: 2.9780 - val_categorical_accuracy: 0.2661
Epoch 6/10
80/80 [==============================] - 118s - loss: 0.7216 - categorical_accuracy: 0.7130 - val_loss: 1.6897 - val_categorical_accuracy: 0.5528
Epoch 7/10
80/80 [==============================] - 119s - loss: 0.6669 - categorical_accuracy: 0.7364 - val_loss: 4.1187 - val_categorical_accuracy: 0.1984
Epoch 8/10
80/80 [==============================] - 119s - loss: 0.6815 - categorical_accuracy: 0.7285 - val_loss: 1.1006 - val_categorical_accuracy: 0.6142
Epoch 9/10
80/80 [==============================] - 118s - loss: 0.6641 - categorical_accuracy: 0.7290 - val_loss: 2.0925 - val_categorical_accuracy: 0.5087
Epoch 10/10
80/80 [==============================] - 119s - loss: 0.6518 - categorical_accuracy: 0.7409 - val_loss: 1.2735 - val_categorical_accuracy: 0.5165Plot Accuracy And Loss Over Time
def plot_accuracy_and_loss(history):
plt.figure(1, figsize= (15, 10))
# plot train and test accuracy
plt.subplot(221)
plt.plot(history.history['categorical_accuracy'])
plt.plot(history.history['val_categorical_accuracy'])
plt.title('SqueezeNet accuracy')
plt.ylabel('accuracy')
plt.xlabel('epoch')
plt.legend(['train', 'test'], loc='upper left')
# plot train and test loss
plt.subplot(222)
plt.plot(history.history['loss'])
plt.plot(history.history['val_loss'])
plt.title('SqueezeNet loss')
plt.ylabel('loss')
plt.xlabel('epoch')
plt.legend(['train', 'test'], loc='upper right')
plt.show()plot_accuracy_and_loss(history)Save Model Weights And Configuration
# save model architecture
model_json = model.to_json()
open('xception_model.json', 'w').write(model_json)
# save model's learned weights
model.save_weights('image_classifier_xception.h5', overwrite=True)Next Lesson
CGAN: Conditional Generative Adversarial Networks
Last updated
