Custom batch data with torch.utils.data.Dataset and DataLoader

1

Hi~
I am a fresh man in Pytorch, and want to generate mini batch data set.
My data format is below , three fields and no target

feature1, feature2, feature3

But the Dataloader’s input is two fields:

feature , target

So how can I generate batch-size data as above?

2

feature is usually an array, target is usually the label.
You can ignore the target if you dont want it.
See an example for a custom data loader here:
https://github.com/QuantScientist/Deep-Learning-Boot-Camp/blob/master/day%2002%20PyTORCH%20and%20PyCUDA/PyTorch/09%20PyTorch%20Kaggle%20Image%20Data-set%20loading%20with%20CNN.ipynb

https://github.com/QuantScientist/Deep-Learning-Boot-Camp/blob/master/day%2002%20PyTORCH%20and%20PyCUDA/PyTorch/21-PyTorch-CIFAR-10-Custom-data-loader-from-scratch.ipynb

3

Thanks
Get it.
the examples above looks nice.