Implementing own "batching" method

[JanoschMenke]

Hi,

I was wondering how do I change the way batches are created by the dataloader. Many tutorials show how I make my own Dataset class which I can then feed into the dataloader. However, I also want to change how batches are created. For example when the matrices have to be combined in a specific way rather than just stacked onto each other.

Is there something like a tutorial out there?

[JanoschMenke]

How to create a dataloader with variable-size input

Hi, I’d like to create a dataloader with different size input images, but don’t know how to do that. I’ve read the official tutorial on loading custum data (http://pytorch.org/tutorials/beginner/data_loading_tutorial.html), however in the tutorial, all the input images are rescaled to 256x256 and randomly cropped to 224*224. I’d like to do another image classification task, and in my task I want to: keep image their original size, and the batch size should be greater than 1. I noticed that th…