How to use downloaded dataset to train Tensorflow image recognition model

Question

I'm trying to build an image recognition model with a dataset I downloaded from this site but I'm not sure how to load it in as a dataset.

I tried this:

dataset_name = 'underwater_dataset'
(train_dataset, test_dataset), dataset_info = tfds.load(name=dataset_name,
                                                       split=['train', 'test'],
                                                       data_dir='C:/Users/tvish/OneDrive/Desktop/Submersible/underwater_dataset',
                                                       shuffle_files=True,
                                                       with_info=True,
                                                       as_supervised=True)

image, label = next(iter(train_dataset.take(1)))
 
plt.imshow(image)
plt.title(label.numpy())
plt.axis('off')
plt.show()

but got this error:

Did you mean: underwater_dataset -> bee_dataset ?

The builder directory C:\Users\tvish\OneDrive\Desktop\Submersible\underwater_dataset\underwater_dataset doesn't contain any versions.
No builder could be found in the directory: C:/Users/tvish/OneDrive/Desktop/Submersible/underwater_dataset/ for the builder: underwater_dataset.
No registered data_dirs were found in:
        - C:/Users/tvish/OneDrive/Desktop/Submersible/underwater_dataset/