[Feature] Add RepVGG backbone and checkpoints #414
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
|
|
Codecov Report
@@ Coverage Diff @@
## master #414 +/- ##
==========================================
+ Coverage 77.29% 77.97% +0.67%
==========================================
Files 100 101 +1
Lines 5356 5547 +191
Branches 883 915 +32
==========================================
+ Hits 4140 4325 +185
- Misses 1098 1101 +3
- Partials 118 121 +3
Flags with carried forward coverage won't be shown. Click here to find out more.
Continue to review full report at Codecov.
|
Ezra-Yu
reviewed
Aug 26, 2021
| parser.add_argument( | ||
| 'save_path', | ||
| help='The path where the converted checkpoint file is stored.') | ||
| parser.add_argument( |
There was a problem hiding this comment.
"device" here is redundant. we did not need to use GPU.
There was a problem hiding this comment.
Resolved. Unit tests have been included for repvgg reparameterization.
Ezra-Yu
reviewed
Aug 26, 2021
mmcls/models/backbones/repvgg.py
Outdated
Comment on lines
222
to
232
| input_dim = self.in_channels // self.groups | ||
| conv_weight = torch.zeros((self.in_channels, input_dim, 3, 3), | ||
| dtype=branch.weight.dtype) | ||
| for i in range(self.in_channels): | ||
| conv_weight[i, i % input_dim, 1, 1] = 1 | ||
| conv_weight = conv_weight.to(branch.weight.device) | ||
| running_mean = branch.running_mean | ||
| running_var = branch.running_var | ||
| gamma = branch.weight | ||
| beta = branch.bias | ||
| eps = branch.eps |
There was a problem hiding this comment.
Here it is better to extract a function called identity_to_conv3×3，then the function "fuse" and "reparameterize" will be more clear.
some important lines need comments, such line 222-226.
Ezra-Yu
reviewed
Aug 26, 2021
mmcls/models/backbones/repvgg.py
Outdated
| """ | ||
| weight_3x3, bias_3x3 = self._fuse_conv_bn(self.branch_3x3) | ||
| weight_1x1, bias_1x1 = self._fuse_conv_bn(self.branch_1x1) | ||
| weight_1x1 = F.pad(weight_1x1, [1, 1, 1, 1]) |
There was a problem hiding this comment.
here adding "value=0" in F.pad() will help other programers to understand this line.
a comment is needed.
Ezra-Yu
reviewed
Aug 26, 2021
mmcls/models/backbones/repvgg.py
Outdated
Comment on lines
453
to
456
| if len(outs) == 1: | ||
| return outs[0] | ||
| else: | ||
| return tuple(outs) |
There was a problem hiding this comment.
Can I use it like this now?
Ezra-Yu
reviewed
Aug 26, 2021
mmcls/models/backbones/repvgg.py
Outdated
| channels = out_channels | ||
|
|
||
| def _make_stage(self, in_channels, out_channels, num_blocks, stride, | ||
| dilation, next_create_block_idx, se_cfg, init_cfg): |
There was a problem hiding this comment.
"next_create_block_idx" is confused, Is it possible not to use this parameter
Ezra-Yu
reviewed
Aug 26, 2021
Ezra-Yu
reviewed
Aug 26, 2021
mmcls/models/backbones/repvgg.py
Outdated
| dtype=branch.weight.dtype) | ||
| for i in range(self.in_channels): | ||
| conv_weight[i, i % input_dim, 1, 1] = 1 | ||
| conv_weight = conv_weight.to(branch.weight.device) |
There was a problem hiding this comment.
is this line Redundant since line 224?
There was a problem hiding this comment.
I don't think so.
…g.py, and delete setting about device.
Ezra-Yu
reviewed
Aug 26, 2021
mmcls/models/backbones/repvgg.py
Outdated
| """ | ||
|
|
||
| optional_groupwise_layers = [ | ||
| 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26 |
There was a problem hiding this comment.
Are these non-zero even-numbered layers？
mzr1996
reviewed
Sep 23, 2021
mzr1996
reviewed
Sep 23, 2021
mzr1996
requested changes
Sep 23, 2021
configs/repvgg/deploy/repvggA0_4xb64-coslr-120e_in-1k_deploy.py
Outdated
Show resolved
Hide resolved
mzr1996
changed the title
mzr1996
reviewed
Sep 25, 2021
mzr1996
requested changes
Sep 25, 2021
Co-authored-by: Ma Zerun <mzr1996@163.com>
mzr1996
added a commit
to mzr1996/mmpretrain
that referenced
this pull request
Nov 24, 2022
* Add RepVGG code. * Add se_module as plugin. * Add the repvggA0 primitive config * Change repvggA0.py to fit mmcls * Add RepVGG configs * Add repvgg_to_mmcls * Add tools/deployment/convert_repvggblock_param_to_deploy.py * Change configs/repvgg/README.md * Streamlining the number of configuration files. * Fix lints * Delete plugins * Delete code about plugin. * Modify the code for using se module. * Modify config to fit repvgg with se. * Change se_cfg to allow loading of pre-training parameters. * Reduce the complexity of the configuration file. * Finsh unitest for repvgg. * Fix bug about se in repvgg_to_mmcls. * Rename convert_repvggblock_param_to_deploy.py to reparameterize_repvgg.py, and delete setting about device. * test commit * test commit * test commit command * Modify repvgg.py to make the code more readable. * Add value=0 in F.pad() * Add se_cfg to arch_settings. * Fix bug. * modeify some attr name and Update unit tests * rename stage_0 to stem and branch_identity to branch_norm * update unit tests * add m.eval in unit tests * [Enhance] Enhence SE layer to support custom squeeze channels. (open-mmlab#417) * add enhenced SE * Update * rm basechannel * fix docstring * Update se_layer.py fix docstring * [Docs] Add algorithm readme and update meta yml (open-mmlab#418) * Add README.md for models without checkpoints. * Update model-index.yml * Update metafile.yml of seresnet * [Enhance] Add `hparams` argument in `AutoAugment` and `RandAugment` and some other improvement. (open-mmlab#398) * Add hparams argument in `AutoAugment` and `RandAugment`. And `pad_val` supports sequence instead of tuple only. * Add unit tests for `AutoAugment` and `hparams` in `RandAugment`. * Use smaller test image to speed up uni tests. * Use hparams to simplify RandAugment config in swin-transformer. * Rename augment config name from `pipeline` to `pipelines`. * Add some commnet ad docstring. * [Feature] Support classwise weight in losses (open-mmlab#388) * Add classwise weight in losses:CE,BCE,softBCE * Update unit test * rm some extra code * rm some extra code * fix broadcast * fix broadcast * update unit tests * use new_tensor * fix lint * [Enhance] Better result visualization (open-mmlab#419) * Imporve result visualization to support wait time and change the backend to matplotlib. * Add unit test for visualization * Add adaptive dpi function * Rename `imshow_cls_result` to `imshow_infos`. * Support str in `imshow_infos` * Improve docstring. * Bump version to v0.15.0 (open-mmlab#426) * [CI] Add PyTorch 1.9 and Python 3.9 build workflow, and remove some CI. (open-mmlab#422) * Add PyTorch 1.9 build workflow, and remove some CI. * Add Python 3.9 CI * Show Python 3.9 support. * [Enhance] Rename the option `--options` in some tools to `--cfg-options`. (open-mmlab#425) * [Docs] Fix sphinx version (open-mmlab#429) * [Docs] Add `CITATION.cff` (open-mmlab#428) * Add CITATION.cff * Fix typo in setup.py * Change author in setup.py * modeify some attr name and Update unit tests * rename stage_0 to stem and branch_identity to branch_norm * update unit tests * add m.eval in unit tests * Update unit tests * refactor * refactor * Alignment inference accuracy * Update configs, readme and metafile * Update readme * return tuple and fix metafile * fix unit test * rm regnet and classifiers changes * update auto_aug * update metafile & readme * use delattr * rename cfgs * Update checkpoint url * Update readme * Rename config files. * Update readme and metafile * add comment * Update mmcls/models/backbones/repvgg.py Co-authored-by: Ma Zerun <mzr1996@163.com> * Update docstring * Improve docstring. * Update unittest_testblock Co-authored-by: Ezra-Yu <1105212286@qq.com> Co-authored-by: Ma Zerun <mzr1996@163.com>
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.
Thanks for your contribution and we appreciate it a lot. The following instructions would make your pull request more healthy and more easily get feedback. If you do not understand some items, don't worry, just make the pull request and seek help from maintainers.
Motivation
Add RepVGG backbone.
Modification
Add the RepVGG source code, and its accompanying configuration files and related files for deployment..
BC-breaking (Optional)
No
Use cases (Optional)
No
Checklist
Before PR:
After PR: