do random resize on batched images #239
Open
Conversation
This file contains hidden or 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
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Outdated suggestions cannot be applied.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.
Description
Right now we randomly resize each image independently and then pad the batch of images to have the same dimensions as the largest image. This means we waste compute and also that with probability ~1 the model doesn't see images of the target resolution without padding.
This PR moves the random resize to be directly in front of the forward pass. This gives slightly higher accuracy and slightly higher throughput.
It also fixes a bug where the expanded scales argument was ignored for Roboflow datasets. This should give slightly better results but slightly lower throughput.
Type of change
Please delete options that are not relevant.
How has this change been tested, please provide a testcase or example of how you tested the change?
Tested locally via a training script.
Any specific deployment considerations
For example, documentation changes, usability, usage/costs, secrets, etc.
Docs