Skip to content

Update batchnorm freezing to handle NormAct variants #1633

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Merged
merged 1 commit into from
Jan 13, 2023
Merged

Update batchnorm freezing to handle NormAct variants #1633

merged 1 commit into from
Jan 13, 2023

Conversation

rwightman
Copy link
Collaborator

  • Add GroupNorm1Act
  • Add BatchNormAct2d tracing change from PyTorch

@rwightman
Update batchnorm freezing to handle NormAct variants, Add GroupNorm1A…
e520553 
…ct, update BatchNormAct2d tracing change from PyTorch
@rwightman rwightman mentioned this pull request Jan 13, 2023
Closed
@HuggingFaceDocBuilderDev
Copy link

HuggingFaceDocBuilderDev commented Jan 13, 2023 •
edited
Loading

The documentation is not available anymore as the PR was closed or merged.

@rwightman rwightman merged commit dd60c45 into main Jan 13, 2023
fffffgggg54 added a commit to fffffgggg54/pytorch-image-models that referenced this pull request Jan 15, 2023
@fffffgggg54
Squashed commit of the following:
fb1d4d2 
commit e9aac41
Author: Ross Wightman <[email protected]>
Date:   Sat Jan 14 22:53:56 2023 -0800

    Correct mean/std for CLIP convnexts

commit 42bd8f7
Author: Ross Wightman <[email protected]>
Date:   Sat Jan 14 21:16:29 2023 -0800

    Add convnext_base CLIP image tower weights for fine-tuning / features

commit 65aea97
Author: Ross Wightman <[email protected]>
Date:   Thu Jan 12 21:31:44 2023 -0800

    Update tests.yml

    Attempt to work around flaky azure ubuntu mirrors

commit dd60c45
Merge: a2c14c2 e520553
Author: Ross Wightman <[email protected]>
Date:   Thu Jan 12 21:13:58 2023 -0800

    Merge pull request huggingface#1633 from rwightman/freeze_norm_revisit

    Update batchnorm freezing to handle NormAct variants

commit e520553
Author: Ross Wightman <[email protected]>
Date:   Thu Jan 12 16:55:47 2023 -0800

    Update batchnorm freezing to handle NormAct variants, Add GroupNorm1Act, update BatchNormAct2d tracing change from PyTorch

commit a2c14c2
Author: Ross Wightman <[email protected]>
Date:   Wed Jan 11 14:50:39 2023 -0800

    Add tiny/small in12k pretrained and fine-tuned ConvNeXt models

commit 01aea8c
Author: Ross Wightman <[email protected]>
Date:   Mon Jan 9 13:38:31 2023 -0800

    Version 0.8.6dev0

commit 2e83bba
Author: Ross Wightman <[email protected]>
Date:   Mon Jan 9 13:37:40 2023 -0800

    Revert head norm changes to ConvNeXt as it broke some downstream use, alternate workaround for fcmae weights

commit 2c24cb9
Author: Ikko Eltociear Ashimine <[email protected]>
Date:   Tue Jan 10 01:26:57 2023 +0900

    Fix typo in results/README.md

    occuring -> occurring

commit 1825b5e
Author: Ross Wightman <[email protected]>
Date:   Sun Jan 8 22:42:24 2023 -0800

    maxxvit type

commit 5078b28
Author: Ross Wightman <[email protected]>
Date:   Sun Jan 8 18:17:17 2023 -0800

    More kwarg handling tweaks, maxvit_base_rw def added

commit c0d7388
Author: Ross Wightman <[email protected]>
Date:   Sat Jan 7 16:29:12 2023 -0800

    Improving kwarg merging in more models

commit 94a9159
Author: Ross Wightman <[email protected]>
Date:   Fri Jan 6 21:39:25 2023 -0800

    Update README.md

commit d2ef5a3
Author: Ross Wightman <[email protected]>
Date:   Fri Jan 6 21:38:40 2023 -0800

    Update README.md

commit ae91530
Author: Ross Wightman <[email protected]>
Date:   Fri Jan 6 17:17:35 2023 -0800

    Update version.py

commit 60ebb6c
Author: Ross Wightman <[email protected]>
Date:   Fri Jan 6 14:35:26 2023 -0800

    Re-order vit pretrained entries for more sensible default weights (no .tag specified)

commit e861b74
Author: Ross Wightman <[email protected]>
Date:   Fri Jan 6 12:01:43 2023 -0800

    Pass through --model-kwargs (and --opt-kwargs for train) from command line through to model __init__. Update some models to improve arg overlay. Cleanup along the way.

commit add3fb8
Author: Ross Wightman <[email protected]>
Date:   Thu Jan 5 17:50:11 2023 -0800

    Working on improved model card template for push_to_hf_hub

commit 13c7183
Author: Xa9aX ツ <[email protected]>
Date:   Fri Jan 6 10:33:08 2023 -0500

    Update installation.mdx

commit eb83eb3
Author: Ross Wightman <[email protected]>
Date:   Thu Jan 5 17:27:13 2023 -0800

    Rotate changelogs, add redirects to mkdocs -> equivalent HF docs pages

commit dd0bb32
Author: Ross Wightman <[email protected]>
Date:   Thu Jan 5 07:55:18 2023 -0800

    Update version.py

    Ver 0.8.4dev0

commit 6e5553d
Author: Ross Wightman <[email protected]>
Date:   Thu Jan 5 07:53:32 2023 -0800

    Add ConvNeXt-V2 support (model additions and weights) (huggingface#1614)

    * Add ConvNeXt-V2 support (model additions and weights)

    * ConvNeXt-V2 weights on HF Hub, tweaking some tests

    * Update README, fixing convnextv2 tests

commit 3698e79
Author: nateraw <[email protected]>
Date:   Wed Jan 4 12:33:40 2023 -0500

    :bug: fix github source links in hf docs

commit 9f5bba9
Author: Nathan Raw <[email protected]>
Date:   Tue Jan 3 17:13:53 2023 -0500

    Structure Hugging Face Docs (huggingface#1575)

    * 🎨 structure docs

    * 🚧 wip docs

    * 📝 add installation doc

    * 📝 wip docs

    * 📝 wip docs

    * 📝 wip docs

    * 📝 wip docs

    * 📝 wip docs

    * 📝 add basic reference docs

    * 📝 remove augmentation from toctree

    * 👷 update pr doc builder to bugfix branch

    * 📝 wip docs

    * 🚧 wip

    * 👷 bump CI

    * 🚧 wip

    * 🚧 bump CI

    * 🚧 wip

    * 🚧 wip

    * 🚧 wip

    * 📝 add hf hub tutorial doc

    * 🔥 remove inference tut

    * 🚧 wip

    * 📝 wip docs

    * 📝 wip docs

    * 📝 update docs

    * 📝 move validation script doc up in order

    * 🎨 restructure to remove legacy docs

    * 📝 update index doc

    * 📝 update number of pretrained models

    * Update hfdocs/README.md

    * Update .github/workflows/build_pr_documentation.yml

    * Update build_pr_documentation.yml

    * bump

    * 📌 update gh action to use main branch

    * 🔥 remove comment

commit 960f5f9
Author: Ross Wightman <[email protected]>
Date:   Fri Dec 30 15:42:41 2022 -0800

    Update results csv with latest val/test set runs

commit 6902c48
Author: Ross Wightman <[email protected]>
Date:   Thu Dec 29 16:32:26 2022 -0800

    Fix ResNet based models to work w/ norm layers w/o affine params. Reformat long arg lists into vertical form.

commit d5aa17e
Author: Ross Wightman <[email protected]>
Date:   Wed Dec 28 17:11:35 2022 -0800

    Remove print from auto_augment

commit 7c846d9
Author: Ross Wightman <[email protected]>
Date:   Sat Dec 24 14:36:29 2022 -0800

    Better vmap compat across recent torch versions

commit 1304589
Author: Ross Wightman <[email protected]>
Date:   Fri Dec 23 15:20:43 2022 -0800

    Update README.md

commit d96538f
Author: Ross Wightman <[email protected]>
Date:   Fri Dec 23 15:19:54 2022 -0800

    Update README

commit 4e24f75
Merge: 18ec173 8ece53e
Author: Ross Wightman <[email protected]>
Date:   Fri Dec 23 10:09:08 2022 -0800

    Merge pull request huggingface#1593 from rwightman/multi-weight_effnet_convnext

    Update efficientnet.py and convnext.py to multi-weight, add new 12k pretrained weights

commit 8ece53e
Author: Ross Wightman <[email protected]>
Date:   Thu Dec 22 21:43:04 2022 -0800

    Switch BEiT to HF hub weights

commit d1bfa9a
Author: Ross Wightman <[email protected]>
Date:   Thu Dec 22 21:34:13 2022 -0800

    Support HF datasets and TFSD w/ a sub-path by fixing split, fix huggingface#1598 ... add class mapping support to HF datasets in case class label isn't in info.

commit 35fb00c
Author: Ross Wightman <[email protected]>
Date:   Thu Dec 22 21:32:31 2022 -0800

    Add flexivit to non-std tests list

commit e2fc43b
Author: Ross Wightman <[email protected]>
Date:   Thu Dec 22 17:34:09 2022 -0800

    Version 0.8.2dev0

commit 9a51e4e
Author: Ross Wightman <[email protected]>
Date:   Thu Dec 22 17:19:45 2022 -0800

    Add FlexiViT models and weights, refactoring, push more weights
    * push all vision_transformer*.py weights to HF hub
    * finalize more pretrained tags for pushed weights
    * refactor pos_embed files and module locations, move some pos embed modules to layers
    * tweak hf hub helpers to aid bulk uploading and updating

commit 656e177
Author: Ross Wightman <[email protected]>
Date:   Fri Dec 16 09:29:13 2022 -0800

    Convert mobilenetv3 to multi-weight, tweak PretrainedCfg metadata

commit 6a01101
Author: Ross Wightman <[email protected]>
Date:   Wed Dec 14 20:33:23 2022 -0800

    Update efficientnet.py and convnext.py to multi-weight, add ImageNet-12k pretrained EfficientNet-B5 and ConvNeXt-Nano.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
2 participants
@rwightman @HuggingFaceDocBuilderDev