TASK#3 (1.5 Mark): Run the code with various options (critical, atomic, reduction) of solving the race condition and see both the result and the execution time.

parallel code:

import numpy as np

import mnist

import tensorflow as tf

from tensorflow.keras.models import Sequential

from tensorflow.keras.layers import Conv2D, MaxPooling2D, Dense, Flatten

from tensorflow.keras.utils import to_categorical

import multiprocessing

def run_model(train_images, train_labels, test_images, test_labels, process_num):

# Assign GPU to this process.

with tf.device(f'/GPU:{process_num}'):

# Normalize the images.

train_images = (train_images / 255) - 0.5

test_images = (test_images / 255) - 0.5

# Reshape the images.

train_images = np.expand_dims(train_images, axis=3)

test_images = np.expand_dims(test_images, axis=3)

num_filters = 8

filter_size = 3

pool_size = 2

# Build the model.

model = Sequential([

Conv2D(num_filters, filter_size, input_shape=(28, 28, 1)),

MaxPooling2D(pool_size=pool_size),

Flatten(),

Dense(10, activation='softmax'),

])

# Compile the model.

model.compile(

'adam', loss='categorical_crossentropy', metrics=['accuracy'],

)

# Train the model.

model.fit(

train_images, to_categorical(train_labels),

epochs=3,

validation_data=(test_images, to_categorical(test_labels)),

)

# Predict on the first 5 test images.

predictions = model.predict(test_images[:5])

# Print our model's predictions.

print("Process:", process_num, "Predictions:", np.argmax(predictions, axis=1))

# Check our predictions against the ground truths.

print("Process:", process_num, "Ground truth:", test_labels[:5])

if __name__ == '__main__':

with multiprocessing.Manager() as manager:

train_images = mnist.train_images()

train_labels = mnist.train_labels()

test_images = mnist.test_images()

test_labels = mnist.test_labels()

chunk_size = train_images.shape[0] // 4

process_list = []

for i in range(4):

start = i * chunk_size

end = start + chunk_size

process = multiprocessing.Process(

target=run_model,

args=(

train_images[start:end],

train_labels[start:end],

test_images,

test_labels,

i,

)

)

process.start()

process_list.append(process)

for process in process_list:

process.join()