Finish 06

06. cifar-10-convolutional-network-implementation/main.py

+# -*- coding: utf-8 -*-
+
+import tensorflow as tf
+import tensorflow_datasets as tfds
+import numpy as np
+import matplotlib.pyplot as plt
+import datetime as dt
+
+datasets = tfds.load("cifar10")
+train_dataset, test_dataset = datasets["train"], datasets["test"]
+assert isinstance(train_dataset, tf.data.Dataset)
+
+cifar10_builder = tfds.builder("cifar10")
+
+# See the information on the dataset
+# info = cifar10_builder.info
+# print(info)
+
+for batch in train_dataset.batch(50000):
+    x_train = batch['image']
+    y_train = batch['label'].numpy().astype('uint8')
+for batch in test_dataset.batch(10000):
+    x_test = batch['image']
+    y_test = batch['label'].numpy().astype('uint8')
+
+# Normalize pixel values to be between 0 and 1
+x_test = x_test / 255
+x_train = x_train / 255
+
+model = tf.keras.models.Sequential([
+    tf.keras.layers.Conv2D(32, (3, 3), activation='relu', padding='same', input_shape=(32, 32, 3)),
+    tf.keras.layers.Conv2D(32, (3, 3), activation='relu'),
+    tf.keras.layers.MaxPooling2D((2, 2)),
+    tf.keras.layers.Conv2D(64, (3, 3), padding='same', activation='relu'),
+    tf.keras.layers.Conv2D(64, (3, 3), activation='relu'),
+    tf.keras.layers.MaxPooling2D((2, 2)),
+    tf.keras.layers.Conv2D(128, (3, 3), padding='same', activation='relu'),
+    tf.keras.layers.Conv2D(128, (3, 3), activation='relu'),
+    tf.keras.layers.Flatten(),
+    tf.keras.layers.Dense(512, activation='relu'),
+    tf.keras.layers.Dense(10, activation='softmax')  
+])
+
+model.compile(optimizer='adam',
+              loss='sparse_categorical_crossentropy',
+              metrics=['accuracy'])
+
+model.summary()
+
+callbacks = [
+            # Write TensorBoard logs to `./logs` directory
+            tf.keras.callbacks.TensorBoard(log_dir='logs/{}'.format(dt.datetime.now().strftime("%Y-%m-%d-%H-%M-%S")))
+]
+
+history = model.fit(x_train, y_train, epochs=50, validation_data=(x_test, y_test), callbacks=callbacks)
+
+# plt.plot(history.history['accuracy'], label='accuracy')
+# plt.plot(history.history['val_accuracy'], label = 'val_accuracy')
+# plt.xlabel('Epoch')
+# plt.ylabel('Accuracy')
+# plt.ylim([0.5, 1])
+# plt.legend(loc='lower right')
+
+test_loss, test_acc = model.evaluate(x_test, y_test, verbose=0)
+print(test_acc)
\ No newline at end of file