For every dataset, we need to see 100 or so of the images that are being trained on, *as they are being trained on*, so post transforms and more. This task is creating code that does this, and if possible collecting those images.
For every dataset, we need to see 100 or so of the images that are being trained on, as they are being trained on, so post transforms and more.
This task is creating code that does this, and if possible collecting those images.