Skip to content

[Fix] Fix the bug that vit cannot load pretrain properly when using i… #999

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

Merged
merged 9 commits into from
Nov 3, 2021
Merged

[Fix] Fix the bug that vit cannot load pretrain properly when using i… #999

Changes from 1 commit
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
5dfb414
[Fix] Fix the bug that vit cannot load pretrain properly when using i…
Oct 27, 2021
4c8cbf5
[Fix] fix the coverage problem
Oct 27, 2021
433faab
Update mmseg/models/backbones/vit.py
RockeyCoss Oct 27, 2021
00bf9f6
[Fix] make the predicate more concise and clearer
Oct 27, 2021
3764cbc
Merge branch 'fix_one_init' of github.com:RockeyCoss/mmsegmentation i…
Oct 27, 2021
b7aaa6a
[Fix] Modified the judgement logic
Oct 28, 2021
678a57c
Update tests/test_models/test_backbones/test_vit.py
RockeyCoss Nov 3, 2021
fdc9a07
Merge branch 'open-mmlab:master' into fix_one_init
RockeyCoss Nov 3, 2021
1350376
add comments
Nov 3, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Prev Previous commit
Next Next commit
[Fix] Modified the judgement logic
  • Loading branch information
RockeyCoss committed Oct 28, 2021
commit b7aaa6a5972b4e0fb8930047fca0b7be07e2d79c
19 changes: 6 additions & 13 deletions mmseg/models/backbones/vit.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -262,18 +262,11 @@ def norm1(self):
return getattr(self, self.norm1_name)

def init_weights(self):
if (isinstance(self.pretrained, str)
or (isinstance(self.init_cfg, dict)
and self.init_cfg.get('type') == 'Pretrained')):
if (isinstance(self.init_cfg, dict)
and self.init_cfg.get('type') == 'Pretrained'):
logger = get_root_logger()
if self.pretrained:
checkpoint = _load_checkpoint(
self.pretrained, logger=logger, map_location='cpu')
else:
checkpoint = _load_checkpoint(
self.init_cfg['checkpoint'],
logger=logger,
map_location='cpu')
checkpoint = _load_checkpoint(
self.init_cfg['checkpoint'], logger=logger, map_location='cpu')

if 'state_dict' in checkpoint:
state_dict = checkpoint['state_dict']
Expand All @@ -294,9 +287,9 @@ def init_weights(self):
(pos_size, pos_size), self.interpolate_mode)

self.load_state_dict(state_dict, False)

elif self.pretrained is None:
elif self.init_cfg is not None:
super(VisionTransformer, self).init_weights()
else:
# 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_init(self.pos_embed, std=.02)
Expand Down