[Refactor] Using mmcv transformer bricks to refactor vit. #571
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 #571 +/- ##
==========================================
- Coverage 85.95% 85.45% -0.50%
==========================================
Files 101 101
Lines 5234 5220 -14
Branches 828 840 +12
==========================================
- Hits 4499 4461 -38
- Misses 561 586 +25
+ Partials 174 173 -1
Flags with carried forward coverage won't be shown. Click here to find out more.
Continue to review full report at Codecov.
|
added 3 commits
June 16, 2021 13:07
xvjiarui
reviewed
Jun 17, 2021
| @@ -0,0 +1,53 @@ | |||
| import logging | |||
mmseg/models/utils/helpers.py
Outdated
Comment on lines
1
to
20
| import collections.abc | ||
| from itertools import repeat | ||
|
|
||
|
|
||
| # From PyTorch internals | ||
| def _ntuple(n): | ||
|
|
||
| def parse(x): | ||
| if isinstance(x, collections.abc.Iterable): | ||
| return x | ||
| return tuple(repeat(x, n)) | ||
|
|
||
| return parse | ||
|
|
||
|
|
||
| to_1tuple = _ntuple(1) | ||
| to_2tuple = _ntuple(2) | ||
| to_3tuple = _ntuple(3) | ||
| to_4tuple = _ntuple(4) | ||
| to_ntuple = _ntuple |
There was a problem hiding this comment.
This file is not necessary. Use from torch.nn.modules.utils import _pair as to_2tuple instead.
mmseg/models/backbones/vit.py
Outdated
Comment on lines
347
to
337
| # We only implement the 'jax_impl' initialization implemented at | ||
| # https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/vision_transformer.py#L353 # noqa: E501 | ||
| trunc_normal_(self.pos_embed, std=.02) | ||
| trunc_normal_(self.cls_token, std=.02) | ||
| for n, m in self.named_modules(): | ||
| if isinstance(m, Linear): | ||
| trunc_normal_(m.weight, std=.02) | ||
| if m.bias is not None: | ||
| if 'mlp' in n: | ||
| normal_init(m.bias, std=1e-6) | ||
| else: | ||
| constant_init(m.bias, 0) | ||
| elif isinstance(m, Conv2d): | ||
| kaiming_init(m.weight, mode='fan_in') | ||
| if m.bias is not None: | ||
| constant_init(m.bias, 0) | ||
| elif isinstance(m, (_BatchNorm, nn.GroupNorm, nn.LayerNorm)): | ||
| constant_init(m.bias, 0) | ||
| constant_init(m.weight, 1.0) | ||
| else: | ||
| raise TypeError('pretrained must be a str or None') | ||
| # Modified from ClassyVision | ||
| nn.init.normal_(self.pos_embed, std=0.02) |
There was a problem hiding this comment.
Why remove the initialization?
added 3 commits
June 18, 2021 00:54
1. Use timm style init_weights; 2. Remove to_xtuple and trunc_norm_;
xvjiarui
reviewed
Jun 17, 2021
mmseg/models/backbones/vit.py
Outdated
| @@ -330,10 +299,17 @@ def init_weights(self, pretrained=None): | |||
| else: | |||
| state_dict = checkpoint | |||
|
|
|||
| if 'rwightman/pytorch-image-models' in pretrained: | |||
There was a problem hiding this comment.
If user downloaded the weight from timm and would like to init the model with path, the condition does not hold.
xvjiarui
reviewed
Jun 17, 2021
.github/workflows/build.yml
Outdated
| @@ -37,19 +37,26 @@ jobs: | |||
| include: | |||
| - torch: 1.3.0+cpu | |||
| torchvision: 0.4.1+cpu | |||
| torch_version: 1.3.0 | |||
xvjiarui
reviewed
Jun 17, 2021
| with_cp (bool): Use checkpoint or not. Using checkpoint will save | ||
| some memory while slowing down the training speed. Default: False. | ||
| pretrain_style (str): Choose to use timm or mmcls pretrain weights. | ||
| Default: timm. |
There was a problem hiding this comment.
We should explain what options are supported, and add assert.
xvjiarui
approved these changes
Jun 17, 2021
bowenroom
pushed a commit
to bowenroom/mmsegmentation
that referenced
this pull request
Feb 25, 2022
…#571) * [Refactor] Using mmcv bricks to refactor vit * Follow the vit code structure from mmclassification * Add MMCV install into CI system. * Add to 'Install MMCV' CI item * Add 'Install MMCV_CPU' and 'Install MMCV_GPU CI' items * Fix & Add 1. Fix low code coverage of vit.py; 2. Remove HybirdEmbed; 3. Fix doc string of VisionTransformer; * Add helpers unit test. * Add converter to convert vit pretrain weights from timm style to mmcls style. * Clean some rebundant code and refactor init 1. Use timm style init_weights; 2. Remove to_xtuple and trunc_norm_; * Add comments for VisionTransformer.init_weights() * Add arg: pretrain_style to choose timm or mmcls vit pretrain weights.
wjkim81
pushed a commit
to wjkim81/mmsegmentation
that referenced
this pull request
Dec 3, 2023
* add atrw dataset * add atrw configs * add animal readme * add atrw * update log interval * update readme * update readme * update init
sibozhang
pushed a commit
to sibozhang/mmsegmentation
that referenced
this pull request
Mar 22, 2024
…en-mmlab#571) * polish README * fix typo
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.
The foundation of this PR:
mmcv: open-mmlab/mmcv#978 (merged)
mmclassification: open-mmlab/mmpretrain#295