Custom semantic segmentation tutorial #2588
Open
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| "cell_type": "markdown", | ||
| "metadata": {}, | ||
| "source": [ | ||
| "flake8: noqa: E501\n", |
| { | ||
| "cells": [ | ||
| { | ||
| "cell_type": "markdown", |
There was a problem hiding this comment.
Let's make this a code cell, otherwise tools like ruff (CPY rule) can't check this file
| "\n", | ||
| "_Written by: Caleb Robinson_\n", | ||
| "\n", | ||
| "In this tutorial, we demonstrate how to extend a TorchGeo [\"trainer class\"](https://torchgeo.readthedocs.io/en/latest/api/trainers.html). In TorchGeo there exist several trainer classes that are pre-made PyTorch Lightning Modules designed to allow for the easy training of models on semantic segmentation, classification, change detection, etc. tasks using TorchGeo's [prebuilt DataModules](https://torchgeo.readthedocs.io/en/latest/api/datamodules.html). While the trainers aim to provide sensible defaults and customization options for common tasks, they will not be able to cover all situations (e.g. researchers will likely want to implement and use their own architectures, loss functions, optimizers, etc. in the training routine). If you run into such a situation, then you can simply extend the trainer class you are interested in, and write custom logic to override the default functionality.\n", |
There was a problem hiding this comment.
Suggested change
| "In this tutorial, we demonstrate how to extend a TorchGeo [\"trainer class\"](https://torchgeo.readthedocs.io/en/latest/api/trainers.html). In TorchGeo there exist several trainer classes that are pre-made PyTorch Lightning Modules designed to allow for the easy training of models on semantic segmentation, classification, change detection, etc. tasks using TorchGeo's [prebuilt DataModules](https://torchgeo.readthedocs.io/en/latest/api/datamodules.html). While the trainers aim to provide sensible defaults and customization options for common tasks, they will not be able to cover all situations (e.g. researchers will likely want to implement and use their own architectures, loss functions, optimizers, etc. in the training routine). If you run into such a situation, then you can simply extend the trainer class you are interested in, and write custom logic to override the default functionality.\n", | |
| "In this tutorial, we demonstrate how to extend a TorchGeo [\"trainer class\"](https://torchgeo.readthedocs.io/en/latest/api/trainers.html). In TorchGeo there exist several trainer classes that are pre-made PyTorch Lightning Modules designed to allow for the easy training of models on semantic segmentation, classification, change detection, etc. tasks using TorchGeo's [prebuilt DataModules](https://torchgeo.readthedocs.io/en/latest/api/datamodules.html). While the trainers aim to provide sensible defaults and customization options for common tasks, they will not be able to cover all situations (e.g., researchers will likely want to implement and use their own architectures, loss functions, optimizers, etc. in the training routine). If you run into such a situation, then you can simply extend the trainer class you are interested in, and write custom logic to override the default functionality.\n", |
There was a problem hiding this comment.
Let's add custom metrics to this list, that's also a common use case.
| "- `configure_callbacks`: We demonstrate how to stack `ModelCheckpoint` callbacks to save the best checkpoint as well as periodic checkpoints\n", | ||
| "- `on_train_epoch_start`: We log the learning rate at the start of each epoch so we can easily see how it decays over a training run\n", | ||
| "\n", | ||
| "Overall these demonstrate how to customize the training routine to investigate specific research questions (e.g. of the scheduler on test performance)." |
There was a problem hiding this comment.
Suggested change
| "Overall these demonstrate how to customize the training routine to investigate specific research questions (e.g. of the scheduler on test performance)." | |
| "Overall these demonstrate how to customize the training routine to investigate specific research questions (e.g., of the scheduler on test performance)." |
| "metadata": {}, | ||
| "outputs": [], | ||
| "source": [ | ||
| "# Get rid of the pesky warnings raised by kornia\n", |
There was a problem hiding this comment.
I would like to not suppress warnings. Can we instead fix them?
| "\n", | ||
| "# You can use the following for actual training runs\n", | ||
| "# from torchgeo.datamodules import LandCoverAIDataModule\n", | ||
| "# dm = LandCoverAIDataModule(root='data', batch_size=64, num_workers=8, download=True)" |
There was a problem hiding this comment.
Not sure if this comment is necessary, but I'll let you decide.
| "name": "stderr", | ||
| "output_type": "stream", | ||
| "text": [ | ||
| "Downloading: \"https://download.pytorch.org/models/resnet50-19c8e357.pth\" to /home/davrob/.cache/torch/hub/checkpoints/resnet50-19c8e357.pth\n", |
There was a problem hiding this comment.
I usually strip the outputs to keep the diffs small. But I know you disagree with this approach.
| " limit_val_batches=1,\n", | ||
| " num_sanity_val_steps=0,\n", | ||
| " max_epochs=1,\n", | ||
| " accelerator='gpu' if torch.cuda.is_available() else 'cpu',\n", |
There was a problem hiding this comment.
We can remove this line, it doesn't cover the other dozen accelerators, and the auto default does
| "source": [ | ||
| "# The following Trainer config is useful just for testing the code in this notebook.\n", | ||
| "trainer = pl.Trainer(\n", | ||
| " limit_train_batches=1,\n", |
There was a problem hiding this comment.
All of this can be replaced by fast_dev_run=True. See the existing Trainers tutorial for how to mock this so that you can do 10+ epochs in the tutorial but 1 step in the tests.
| "name": "stderr", | ||
| "output_type": "stream", | ||
| "text": [ | ||
| "Trainer will use only 1 of 8 GPUs because it is running inside an interactive / notebook environment. You may try to set `Trainer(devices=8)` but please note that multi-GPU inside interactive / notebook environments is considered experimental and unstable. Your mileage may vary.\n", |
There was a problem hiding this comment.
This is one of the reasons I like to strip the output
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Re-upped version of #1897
For review convenience, here's a link to the notebook.