Check eval performance when extracting center token for classification tasks#497
Open
Check eval performance when extracting center token for classification tasks#497
Conversation
This means pooling over the center tokens only (across timesteps/modalities) instead of all tokens. Since previously we observed this helps for some tasks. Also add option to use bilinear interpolation followed by linear layer for segmentation tasks.
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
3 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
This adds an option to pool over the center tokens only (across timesteps/modalities) instead of all tokens for classification tasks. The motivation is previous observations that this helps for some tasks like Nandi.
I also added an option to use bilinear interpolation followed by linear layer for segmentation tasks.
It doesn't seem to make a big difference for most tasks: https://wandb.ai/eai-ai2/2026_02_14_eval_changes