Fix for Blank Images in plot due to Float Tensor Ranges

[menglutao]

Description:
Encountered an issue where ucmerced dataset images would render as blank after using the plot function and matplotlib to show the image.

Issue:
The underlying cause is that Matplotlib expects float values to be normalized between 0 and 1. However, the float tensor representing our image data occasionally contained values in the 0 to 255 range, leading to unexpected behavior in the rendering process.

Resolution:
A conditional check has been introduced. When the maximum value in the image tensor exceeds 1, the tensor is normalized by scaling down to the 0 to 1 range. This ensures compatibility with Matplotlib's expected float value range and resolves the issue of blank images.

Outcome:
With the proposed changes, images are displayed correctly, regardless of whether their data originally spans 0 to 1 or 0 to 255.

I invite the maintainers to review the changes. Please advise if further modifications are required or if the update can be merged.

[adamjstewart]

This relates to #1263. Depending on whether or not normalization is added, all of our plotting breaks. A better solution would be to implement the ideas in #1263, but that's a lot of work.

Looking through the data, it seems like every image is in the range 0–255. Until we implement #1263, it may be simpler just to always divide by 255.

[calebrob6]

Let's merge this as plotting of images on tensorboard is indeed broken

[adamjstewart]

Doesn't tensorboard use the data module that includes normalization which divides all images by 255? If there are images with values greater than 255 that's a whole nother problem.

[adamjstewart]

Just checked and there are none greater than 255. So yes, data module plotting is broken, just like for every other dataset. We really need #1263 if we want both dataset plotting and data module plotting to work.

[adamjstewart]

Let's merge this for now and think about how we want to fix all other datasets in a separate PR. Thanks for the contribution!