Update custom.py #849
Update custom.py #849
Conversation
skip tmp files to avoid unexspected images
|
|
Codecov Report
@@ Coverage Diff @@
## master #849 +/- ##
==========================================
- Coverage 88.90% 88.87% -0.03%
==========================================
Files 110 110
Lines 5928 5930 +2
Branches 950 951 +1
==========================================
Hits 5270 5270
- Misses 465 466 +1
- Partials 193 194 +1
Flags with carried forward coverage won't be shown. Click here to find out more.
Continue to review full report at Codecov.
|
|
Hi @huster-wgm |
|
Please merge the master branch into your branch, thank you. |
| for img in mmcv.scandir(img_dir, img_suffix, recursive=True): | ||
| if '.ipynb_checkpoints' in img: | ||
| continue # skip tmp files |
There was a problem hiding this comment.
I think img_suffix will filter out the files with '.ipynb_checkpoints'
|
Hi @huster-wgm ！We are grateful for your efforts in helping improve this open-source project during your personal time. |
* test the simple case using tmux to run multiple benchmark regression test tasks * modify and rename the config file and script * Delete config_list.yaml * modify the config and rename the filename * Delete test_benchmark_tmux.py * modify the script and rename the filename * Update setup.cfg * using mmcv.load to avoid introducing the extra dependency on yaml * fix some typo * refactor the config file and modify the script accordingly * modify the config and script * rename the config file
* add dataset info (open-mmlab#663) * Fix import and deprecation issues in unit tests (open-mmlab#871) * fix some bugs in the unit test of smpl model. * reorganize `tests/` to solve importing issue (PEP 420) * fix deprecation warnings in unit tests Co-authored-by: ly015 <[email protected]> * add benchmark regression test script with tmux (open-mmlab#849) * test the simple case using tmux to run multiple benchmark regression test tasks * modify and rename the config file and script * Delete config_list.yaml * modify the config and rename the filename * Delete test_benchmark_tmux.py * modify the script and rename the filename * Update setup.cfg * using mmcv.load to avoid introducing the extra dependency on yaml * fix some typo * refactor the config file and modify the script accordingly * modify the config and script * rename the config file * Correct dataset preparation guide of WFLW (open-mmlab#873) * add pr template (open-mmlab#875) * add CITATION.cff and update setup.py (open-mmlab#876) * Add copyright header and pre-commit hook (open-mmlab#872) * Add pre-commit hook to automatically add copyright file header * update files with copyright header * Limit copyright checking in the first 2 lines of a file * Exclude configs in demo/ * set max-header-lines as 5 * rebase to master and add copyright to new files * move benchmark_regression into .dev_scripts/benchmark * Translate tasks/2d_body_keypoint.md (open-mmlab#842) * 2rd PR remove poseval * fix lint * revise the CN version Co-authored-by: ly015 <[email protected]> * fix some bugs in the unit test of smpl model. * * reorganiz `tests/` to solve importing issue (PEP 420) * add dataset info * fix lint * * fix wrongly modified parts in previous rebase * fix lint * rename datasets/_base_ as datasets/base * resolve compatibility of pose_limb_color * Add dummy dataset base classes with old names for compatibility * * Rewrite relative unittest based on dataset_info * Add bc-breaking test for functions related to dataset_info * Rename DatasetInfo.dataset_info as DatasetInfo._dataset_info * Fix dataset_info of h36m dataset * Handle breaking change pose_limb_color -> pose_link_color * add unittest for old-fashioned dataset initialization without dataset_info * resolve naming conflict in unittests Co-authored-by: zengwang430521 <[email protected]> Co-authored-by: ly015 <[email protected]> * fix typo * fix typo Co-authored-by: Jas <[email protected]> Co-authored-by: zengwang430521 <[email protected]>
Motivation
skip tmp files generated by jupyter notebook to avoid unexpected failture for image loading.
Modification
filtering scan files
BC-breaking (Optional)
No
Use cases (Optional)
No
Checklist