Support to specify LR of DCN's conv_offset #344
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 #344 +/- ##
==========================================
+ Coverage 64.76% 64.95% +0.18%
==========================================
Files 133 134 +1
Lines 7597 7632 +35
Branches 1336 1346 +10
==========================================
+ Hits 4920 4957 +37
+ Misses 2432 2425 -7
- Partials 245 250 +5
Flags with carried forward coverage won't be shown. Click here to find out more.
Continue to review full report at Codecov.
|
hellock
approved these changes
Jun 14, 2020
xvjiarui
reviewed
Sep 1, 2020
xvjiarui
reviewed
Sep 1, 2020
hellock
reviewed
Sep 1, 2020
|
Task linked: CU-49u01p Support to specify LR of DCN's conv_offset |
xvjiarui
approved these changes
Sep 14, 2020
hellock
requested changes
Sep 18, 2020
hellock
reviewed
Sep 24, 2020
hellock
reviewed
Sep 24, 2020
hellock
reviewed
Sep 24, 2020
hellock
approved these changes
Sep 25, 2020
wxzs5
pushed a commit
to wxzs5/mmcv
that referenced
this pull request
Nov 20, 2020
* Support to specify LR of DCN's conv_offset * Resolve comments & add unit test * Resolve formats * Fix CI for DCN * Mock DCN when cpu only * Use mock for cpu testing * Fix docstring and support ModulatedDCN * set offset_lr_mult as dcn's arguments, link CU-49u01p * fix lr bug * fall back to set LR in constructor * resolve comments
mattdawkins
added a commit
to VIAME/mmcv
that referenced
this pull request
Dec 3, 2020
* tag 'v1.1.5': (81 commits) bump version to 1.1.5 (open-mmlab#603) Fix wrappers version comparison (open-mmlab#602) revert dcn (open-mmlab#597) lint markdown files (open-mmlab#592) bump version to 1.1.4 (open-mmlab#591) fix lint open-mmlab#580 (open-mmlab#590) Fixed fp16_optimizer state bug (open-mmlab#580) register deconv in CONV_LAYERS (open-mmlab#582) [enhance]: show grad norm in fp16 optimizer hook (open-mmlab#584) [Enhance] Switch to https://download.openmmlab.com (open-mmlab#583) Support to specify LR of DCN's conv_offset (open-mmlab#344) add torchvision roi_align with aligned=True (open-mmlab#581) Add runner builder (open-mmlab#570) [fix]: fix wrapper comparison of pytorch version (open-mmlab#572) Add missing by_epoch arg (open-mmlab#576) merge the calling of train/val_step and batch_processor into run_iter (open-mmlab#553) bump version to 1.1.3 (open-mmlab#569) fix deprecated wrappers exiting bug (open-mmlab#567) add switch for onnx exporter (open-mmlab#564) fix dcon forward and backward bug (open-mmlab#565) ...
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.
This PR add an option
dcn_offset_lr_multinparams_cfg. The modification enables to set the learning rate of DCN's conv offset layer to stabilize training.