A generalizable tensorflow template with TensorBoard integration and inline image viewing.

For this expansion of the generalizable template I’m going to add a function to view images and labels. This is useful for inspecting the data prior to fitting and also assessing the results of your model.

import numpy as np
import tensorflow as tf
import matplotlib.pyplot as plt
import datetime
print(tf.__version__)

2.0.0

from numpy.random import seed
seed(1)
tf.random.set_seed(1234)

%load_ext tensorboard
!rm -rf ./logs/ 

from tensorflow.keras.activations import sigmoid
def swish(x, beta = 1):
    return (x * sigmoid(beta * x))

from tensorflow.keras.utils import get_custom_objects
from tensorflow.keras.layers import Activation
get_custom_objects().update({'swish': Activation(swish)})

from tensorflow.keras.callbacks import Callback
class my_callbacks(Callback):
    def on_train_begin(self, logs=None):
        self.beginning_accuracy = 0.0

    def on_epoch_begin(self, epoch, logs={}):
        current_accuracy=self.beginning_accuracy

    def on_epoch_end(self,epoch,logs={}):
        current_accuracy=logs.get('accuracy')
        if current_accuracy>self.beginning_accuracy:
            test.update({'accuracy':current_accuracy,'epochs':epoch})
        self.beginning_accuracy = current_accuracy

mnist = tf.keras.datasets.fashion_mnist 

Here’s the class names which will be useful for labeling our images.

class_names=['T-shirt/top','Trouser','Pullover','Dress','Coat','Sandal','Shirt','Sneaker','Bag','Ankle_boot']

(training_images, training_labels), (test_images, test_labels) = mnist.load_data() 

training_images  = training_images / 255.0
test_images = test_images / 255.0

test_parameters = ({'epochs':1, 'activation':'swish'},
                    {'epochs':5, 'activation':'swish'},
                    {'epochs':10, 'activation':'swish'},
                    {'epochs':1, 'activation':'relu'},
                    {'epochs':5, 'activation':'relu'},
                    {'epochs':10, 'activation':'relu'})

for test in test_parameters:
    model = tf.keras.models.Sequential([tf.keras.layers.Flatten(),
                                        tf.keras.layers.Dense(128, activation=test["activation"]),
                                        tf.keras.layers.Dense(10, activation=tf.nn.softmax)])
    model.compile(optimizer = 'adam',
                  loss = 'sparse_categorical_crossentropy',
                  metrics=['accuracy'],
                  verbose=0)
    
    log_dir="logs/fit/" + datetime.datetime.now().strftime("%Y%m%d-%H%M%S")
    
    tensorboard_callback = tf.keras.callbacks.TensorBoard(log_dir=log_dir, histogram_freq=1)


    model.fit(training_images,
              training_labels,
              epochs=test["epochs"],
              verbose=0,
              callbacks=[my_callbacks(),tensorboard_callback])

    model.evaluate(test_images,
                   test_labels,
                   verbose=0)
print(test_parameters)

({'epochs': 0, 'activation': 'swish', 'accuracy': 0.82645}, {'epochs': 4, 'activation': 'swish', 'accuracy': 0.8958167}, {'epochs': 9, 'activation': 'swish', 'accuracy': 0.91635}, {'epochs': 0, 'activation': 'relu', 'accuracy': 0.82301664}, {'epochs': 4, 'activation': 'relu', 'accuracy': 0.891}, {'epochs': 9, 'activation': 'relu', 'accuracy': 0.90956664})

!kill 9053

/bin/sh: 1: kill: No such process

%tensorboard --logdir logs/fit

Reusing TensorBoard on port 6006 (pid 12841), started 0:16:53 ago. (Use '!kill 12841' to kill it.)

predictions = model.predict(test_images)
print(predictions)

[[5.7263919e-06 6.6372614e-09 4.0063137e-06 ... 4.1030091e-03
  1.9248000e-06 9.9537212e-01]
 [2.2955917e-05 1.4473169e-11 9.9884146e-01 ... 4.7463745e-15
  4.1615325e-10 9.1840791e-17]
 [1.0521556e-10 1.0000000e+00 2.2429164e-12 ... 2.3259870e-27
  7.9439719e-17 1.8272301e-17]
 ...
 [2.6303707e-03 1.5326654e-10 1.7281596e-03 ... 6.4461265e-06
  9.9467367e-01 7.2260475e-10]
 [7.9042373e-08 9.9999738e-01 9.4543378e-09 ... 8.7363508e-16
  6.1607669e-13 7.4512601e-09]
 [1.9130493e-05 1.4222825e-07 2.7901824e-06 ... 1.4867900e-03
  4.7969297e-04 5.3793042e-06]]

This function prepares and displays whichever images and labels you assign as arguments. Its a good way of assessing your data prior, and after, classification.

def show_images(images, labels):
        plt.figure(figsize=(10,10))
        grid_size = min(25, len(images))
        for i in range(grid_size):
                plt.subplot(5, 5, i+1)
                plt.yticks([])
                plt.xticks([])
                plt.imshow(images[i], cmap=plt.cm.copper)
                plt.xlabel(class_names[labels[i]],size=10,color='orange')

Here we use our prediction values and test images for the show_images function.

show_images(test_images, np.argmax(predictions, axis = 1))
plt.show()