program_test.py

# Required packages: torch
# Desired packages: numpy, matplotlib
# Add missing imports:

def main():
    training_images = {x for x in range(1, 100001)}
    priority_images = {x * 100 for x in range(1, 11)}
    # Task:
    # 1. Dataloader needs to return a batch of size 10. Each element is a unique element from `training_images`.
    # 2. Elements from `priority_images` must be present in EVERY batch with the ratio 1:1, meaning that batch will have
    #    50% of images from `priority_images` and 50% from `training_images` containers.
    #    NOTE: elements of `priority_images` are part of the `training_images`.
    # 3. The code needs to be multiprocessing safe.
    # 4. Follow comments below to add extra functionality.

    batch_size = 10
    num_epochs = 10

    for epoch in range(num_epochs):
        pass


if __name__ == '__main__':
    main()