Add eurocrops data module. #1869
Merged
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
It is based on NAIPChesapeakeDataModule which splits bounding box of dataset into 1/2 train, 1/4 val, and 1/4 test. This may not be the best way to train an actual model.
github-actions
bot
added
datasets
|
Hi @favyen2, could you use the new splitting function |
adamjstewart
requested changes
Mar 15, 2024
…ocrops_datamodule
|
I am not able to get this to work without some changes from #1889 like setting Sentinel-2 test data size to 128 instead of 36, so I will wait for that to be merged first since that will make it easier. |
|
#1889 has now been merged, feel free to rebase and copy-n-paste whatever you want from that data module. |
…ocrops_datamodule
adamjstewart
previously approved these changes
Apr 12, 2024
github-actions
bot
added
the
documentation
adamjstewart
approved these changes
Apr 12, 2024
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Outdated suggestions cannot be applied.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.
It is based on NAIPChesapeakeDataModule which splits bounding box of dataset into 1/2 train, 1/4 val, and 1/4 test.
This may not be the best way to train an actual model. I think it is more natural to either split by country, or to randomly assign each large grid cell (e.g. 4096x4096 pixel) to train/val/test and then sample within those grid cells. But I wasn't sure how to split by country since VectorDataset automatically detects all the files, or to assign large grid cell since there's no sampler that can take multiple large bounding boxes and sample patches within them.