feat: add SimMIM imagenet benchmark - #2004
Open
lorinczszabolcs wants to merge 1 commit into
Open
Conversation
Add benchmarks/imagenet/vitb16/simmim.py and register it in main.py. SimMIM encodes all tokens (masked positions replaced by the mask token), predicts the masked pixel patches with a linear head, and trains with an L1 loss, reusing the shared vitb16 harness (timm ViT-B/16, OnlineLinearClassifier, AdamW + cosine warmup). Script only; the full ImageNet run and results are left to the maintainers' cluster. Addresses the unchecked SimMIM item in lightly-ai#1197.
Contributor
Author
|
This is an older method, so feel free to ignore it if the benchmark gaps for these models no longer matter. |
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Adds a SimMIM benchmark under
benchmarks/imagenet/vitb16/and registers it inmain.py, completing the unchecked "SimMIM ImageNet Benchmark" item in #1197.The benchmark reuses the shared vitb16 utilities (timm ViT-B/16
MaskedVisionTransformerTIMM,OnlineLinearClassifier, AdamW + cosine warmup) and follows the SimMIM paper (https://arxiv.org/abs/2111.09886) settings.Like the other benchmark contributions, this adds the script only. The full ImageNet run, results row, and released checkpoints are left out from this PR.
Two notes for review:
examples/.../simmim.py, which masks single tokens.