💡 [REQUEST] - New recipe tutorial on calculating layer output dimensions

loganthomas commented 3 months ago

🚀 Describe the improvement or the new tutorial

This tutorial will help users understand how to transition from convolutional and pooling layers to linear layers in their models.

Learning objectives:

How to manually calculate the output dimensions after applying a convolution or pooling layer
How to print the shape of internal tensors for inspecting dimensionality changes in a model
How to use the torchinfo package to show output dimensions for all layers in a model

Existing tutorials on this topic

No response

Additional context

I created this draft (https://github.com/pytorch/tutorials/pull/2923) as a part of the PyTorch Docathon H1 2024 effort. I did not realize new tutorials weren't being accepted as part of the sprint and was asked to fill out an issue and convert the PR to a draft.

pratikkorat26 commented 3 months ago

If this is not completed, I would like to contribute to it. Please guide me on how to contribute ?

loganthomas commented 3 months ago

Draft in https://github.com/pytorch/tutorials/pull/2923

pytorch / tutorials