[CVS-94911] Fix difference between train and validation normalization pipeline #1310
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
goodsong81
reviewed
Oct 24, 2022
There was a problem hiding this comment.
Could you replicate this change to feature/otx?
@harimkang Or, do you have plan to move changes to develop to feature/otx?
Commits with (& after) tiling are not included in |
I can directly change otx/feature branch to reflect this PR. However if there are plan to merge develop and otx/feature branch, then I can wait for it. |
|
I checked all of |
sungchul2
previously approved these changes
Oct 25, 2022
|
run ote-test |
eugene123tw
requested changes
Oct 28, 2022
There was a problem hiding this comment.
I'm getting this error running tests. Could you take a look?
Traceback (most recent call last):
File "/tmp/ote_cli/model-preparation-algorithm/venv/bin/ote", line 33, in <module>
sys.exit(load_entry_point('ote-cli', 'console_scripts', 'ote')())
File "/home/yuchunli/git/training_extensions/ote_cli/ote_cli/tools/ote.py", line 67, in main
globals()[f"ote_{name}"]()
File "/home/yuchunli/git/training_extensions/ote_cli/ote_cli/tools/train.py", line 175, in main
task.train(dataset, output_model, train_parameters=TrainParameters())
File "/home/yuchunli/git/training_extensions/external/model-preparation-algorithm/mpa_tasks/apis/detection/task.py", line 528, in train
results = self._run_task(stage_module, mode="train", dataset=dataset, parameters=train_parameters)
File "/home/yuchunli/git/training_extensions/external/model-preparation-algorithm/mpa_tasks/apis/task.py", line 112, in _run_task
output = workflow.run(
File "/home/yuchunli/git/training_extensions/external/model-preparation-algorithm/submodule/mpa/det/trainer.py", line 126, in run
DetectionTrainer.train_worker(
File "/home/yuchunli/git/training_extensions/external/model-preparation-algorithm/submodule/mpa/det/trainer.py", line 161, in train_worker
train_detector(
File "/home/yuchunli/git/training_extensions/external/mmdetection/submodule/mmdet/apis/train.py", line 260, in train_detector
runner.run(data_loaders, cfg.workflow, compression_ctrl=compression_ctrl)
File "/tmp/ote_cli/model-preparation-algorithm/venv/lib/python3.8/site-packages/mmcv/runner/epoch_based_runner.py", line 127, in run
epoch_runner(data_loaders[i], **kwargs)
File "/home/yuchunli/git/training_extensions/ote_sdk/ote_sdk/utils/argument_checks.py", line 250, in validate
return function(**input_parameters_values_map)
File "/home/yuchunli/git/training_extensions/external/model-preparation-algorithm/mpa_tasks/utils/runner.py", line 53, in train
for i, data_batch in enumerate(self.data_loader):
File "/tmp/ote_cli/model-preparation-algorithm/venv/lib/python3.8/site-packages/torch/utils/data/dataloader.py", line 517, in __next__
data = self._next_data()
File "/tmp/ote_cli/model-preparation-algorithm/venv/lib/python3.8/site-packages/torch/utils/data/dataloader.py", line 557, in _next_data
data = self._dataset_fetcher.fetch(index) # may raise StopIteration
File "/tmp/ote_cli/model-preparation-algorithm/venv/lib/python3.8/site-packages/torch/utils/data/_utils/fetch.py", line 44, in fetch
data = [self.dataset[idx] for idx in possibly_batched_index]
File "/tmp/ote_cli/model-preparation-algorithm/venv/lib/python3.8/site-packages/torch/utils/data/_utils/fetch.py", line 44, in <listcomp>
data = [self.dataset[idx] for idx in possibly_batched_index]
File "/home/yuchunli/git/training_extensions/external/mmdetection/submodule/mmdet/datasets/dataset_wrappers.py", line 475, in __getitem__
return self.pipeline(self.tile_dataset[idx])
File "/home/yuchunli/git/training_extensions/external/mmdetection/submodule/mmdet/datasets/pipelines/compose.py", line 40, in __call__
data = t(data)
File "/home/yuchunli/git/training_extensions/external/mmdetection/submodule/mmdet/datasets/pipelines/formating.py", line 322, in __call__
img_meta[key] = results[key]
KeyError: 'ignored_labels'
FAILED
harimkang
reviewed
Oct 31, 2022
Hi @eugene123tw, I talked about this with @jaegukhyun . Currently, other datasets in OTX are able to receive ignored_labels as collection inputs, but the tiling dataset seems to be an issue because it is not handled that way. Once it is tiling, Jaeguk will fix it so that the function is not called. It would be nice to be able to receive |
c97cd7d to
68a74c5
Compare
68a74c5 to
68cef49
Compare
|
run ote-test |
|
http://validationreports.sclab.intel.com:8006/reports/build_number_report?trimReq=on&status=FAIL&test_session_build_number=pr-ote-test-697&environment=iotg-dev-workstation-09 |
eugene123tw
approved these changes
Nov 22, 2022
harimkang
approved these changes
Nov 23, 2022
goodsong81
added a commit
that referenced
this pull request
Dec 28, 2022
* Add tiling module (#1200) * Update submodule branch (#1222) * Enhance training schedule for multi-label classification (#1212) * [CVS-88098] Remove initialize from export functions (#1226) * Train graph added (#1211) * Add @attrs decorator for base configs (#1229) * Pretrained weight download error in MobilenetV3-large-1 of deep-object-reid in SC (#1233) * [Anomaly Task] Revert hpo template (#1230) * 🐞 [Anomaly Task] Fix progress bar (#1223) * [CVS-90555] Fix NaN value in classification (#1244) * update hpo_config.yaml (#1240) * [CVS-90400, CVS-91015] NNCF pruning supported tweaks (#1248) * [Anomaly Task] 🐞 Fix inference when model backbone changes (#1242) * [CVS-91472] Add pruning_supported value (#1263) * Pruning supported tweaks (#1256) * [CVS-90400, CVS-91015] NNCF pruning supported tweaks (#1248) * Revert "[CVS-90400, CVS-91015] NNCF pruning supported tweaks (#1248)" (#1269) * [OTE-TEST] Disable obsolete test cases (#1220) * [OTE-TEST] hot-fix for MPA performance tests (#1273) * [Anomaly Task] ✨ Upgrade anomalib (#1243) * Expose early stopping hyper-parameters for all tasks (#1241) * Resolve pre-commit issues (#1272) * Remove LazyEarlyStopHook in model_multilabel.py (#1281) * Removed xfail (#1239) * Implement IB loss for incremental learning in multi-class classification (#1289) * Edit num_workers and change MPA repo as a latest (#1314) * fix annotation bug (#1320) * Valid POT configs for small HRNet models (#1313) * Disable NNCF optimization for FP16 models (#1312) * fliter object less than 1 pixel (#1305) * Fix some tests (#1322) * [Develop] Move drop_last into MPA (#1357) * Apply changes from releases/v0.3.1-geti1.0.0 (#1337) * anomaly save_model bugfix (#1300) * upgrade networkx module version (#1303) * Forward CVS-94422 size bug fix PR to release branch (#1326) * Valid POT configs for small HRNet models (#1317) * [Release branch] Disable NNCF optimization for FP16 models (#1319) * [RELEASE] CVS-95549 - Hierarchical classification training failed without obvious reason (#1329) * Fix h-label: per-group softmax (#1332) * Fix dataset length bug in mpa task (#1338) * Fix drop_last key issue for det/set (#1340) * Hot-fix for OV inference for iseg output (#1345) * Fix nncf model export bug (#1346) * Fixed merge error (#1359) * Update evaluation iou_thr of ins-seg (#1354) * fix pre-commit test (#1366) * Fix dataset item tests (#1360) * Fix OV Inference issues (tiling tests & detection tests) (#1361) * fix black & add xfail test cases (#1367) * Update check_nncf_graph. (#1330) * [Develop] Hot-fix OV inference issue in rotated detection (#1375) * [Develop] updated documents (#1383) * [CVS-94911] Fix difference between train and validation normalization pipeline (#1310) * Update configs for padim model (#1378) * updated QUICK_START_GUIDE.md (#1397) * Change ote threshold of openvino test for cls (#1401) * Normalize top-1 metrics to [0, 1] (#1394) * Tiling deployment (#1387) * Replace current saliency map generation with Recipro-CAM for cls (#1363) * Class-wise saliency map generation for the detection task (#1402) * Change submodule to develop (#1410) * Send full dataset to POT optimization function (#1379) & Convert NaN to num to make visible in geti UI (#1413) * Add active score evaluation to the classification task * [release/0.4.0][OTX] Enabling GPU execution for exported code (#1416) * [OTE][Release][XAI] Detection fix two stage bbox_head error (#1414) * Update SDK commit for exportable code (#1423) * HRNet-x and HRNe-18--mod2 configs update (#1419) * [Release] Enable tiling oriented detection for v0.4.0/geti1.1.0 (#1427) * [OTE][Releases v0.4.0][XAI] Hot-fix for Detection fix two stage error (#1433) * Temporary MPA branch while dev->otx merge process * Update doc & install for dev->otx changes * Update ote_sdk -> otx.api * Update ote_cli -> otx.cli * Update external/mmsegmentation -> otx/algorithms/segmentation * Align saliency map media instantiation over tasks (#1447) * Update external/d-o-r -> otx/algorithms/classification * Update external/mmdetection -> otx/algorithms/detection * Update external/mpa -> otx/algorithms/* * Fix CLI test run for better error message * Numpy constraint for deprecated np.bool error * Capture stderr only * Align numpy requirement * [OTX/Anomaly] Add changes from external to otx (#1452) * Add changes from external to otx * Address PR comments * Update config files + remove backbone from base * Fix pre-merge checks * Fix pre-commit issues * Update exportable code commit * Fix indent error * Fix flake8 issue * Resolve softmax issue w/ FIXME for future work * Add tiling tests * Revert MPA branch to otx Signed-off-by: Songki Choi <songki.choi@intel.com> Co-authored-by: Eugene Liu <eugene.liu@intel.com> Co-authored-by: Ashwin Vaidya <ashwin.vaidya@intel.com> Co-authored-by: Jaeguk Hyun <jaeguk.hyun@intel.com> Co-authored-by: Nikita Savelyev <nikita.savelyev@intel.com> Co-authored-by: Jihwan Eom <jihwan.eom@intel.com> Co-authored-by: Harim Kang <harim.kang@intel.com> Co-authored-by: Soobee Lee <soobee.lee@intel.com> Co-authored-by: Lee, Soobee <soobeele@intel.com> Co-authored-by: Emily Chun <emily.chun@intel.com> Co-authored-by: ljcornel <ludo.cornelissen@intel.com> Co-authored-by: Eunwoo Shin <eunwoo.shin@intel.com> Co-authored-by: dlyakhov <daniil.lyakhov@intel.com> Co-authored-by: kprokofi <kirill.prokofiev@intel.com> Co-authored-by: Sungman Cho <sungman.cho@intel.com> Co-authored-by: Yunchu Lee <yunchu.lee@intel.com> Co-authored-by: Ashwin Vaidya <ashwinitinvaidya@gmail.com> Co-authored-by: Alexander Dokuchaev <alexander.dokuchaev@intel.com> Co-authored-by: Vladislav Sovrasov <vladislav.sovrasov@intel.com> Co-authored-by: Evgeny Tsykunov <e.tsykunov@gmail.com> Co-authored-by: Galina Zalesskaya <galina.zalesskaya@intel.com> Co-authored-by: dongkwan-kim <dongkwan.kim@intel.com>
sungmanc
added a commit
that referenced
this pull request
Dec 28, 2022
* make logfile saved in save-model-to directory * enable train * make main process train also * bugfix * refactor multi gpu training * make all processs have same output path * prevent child process from being termated by fokred main process * refactor multigpu implementation * refactor multi gpu implementation * modify argument help sentence * add multi gpu test code * align with pre-commit test * separate multi GPU manager class * modify train cli argument 'save-logs-to' to 'output-path' * remove tray excpet during killing child process * apply output_path to all tasks * change print to logger * skip multi gpu test if number of gpu is insufficient * fix typo * multi gpu test bugfix * isort fix * test case bugfix * fix typo and change some variable name * [OTX] Apply changes in develop to feature/otx branch (#1436) * Add tiling module (#1200) * Update submodule branch (#1222) * Enhance training schedule for multi-label classification (#1212) * [CVS-88098] Remove initialize from export functions (#1226) * Train graph added (#1211) * Add @attrs decorator for base configs (#1229) * Pretrained weight download error in MobilenetV3-large-1 of deep-object-reid in SC (#1233) * [Anomaly Task] Revert hpo template (#1230) * 🐞 [Anomaly Task] Fix progress bar (#1223) * [CVS-90555] Fix NaN value in classification (#1244) * update hpo_config.yaml (#1240) * [CVS-90400, CVS-91015] NNCF pruning supported tweaks (#1248) * [Anomaly Task] 🐞 Fix inference when model backbone changes (#1242) * [CVS-91472] Add pruning_supported value (#1263) * Pruning supported tweaks (#1256) * [CVS-90400, CVS-91015] NNCF pruning supported tweaks (#1248) * Revert "[CVS-90400, CVS-91015] NNCF pruning supported tweaks (#1248)" (#1269) * [OTE-TEST] Disable obsolete test cases (#1220) * [OTE-TEST] hot-fix for MPA performance tests (#1273) * [Anomaly Task] ✨ Upgrade anomalib (#1243) * Expose early stopping hyper-parameters for all tasks (#1241) * Resolve pre-commit issues (#1272) * Remove LazyEarlyStopHook in model_multilabel.py (#1281) * Removed xfail (#1239) * Implement IB loss for incremental learning in multi-class classification (#1289) * Edit num_workers and change MPA repo as a latest (#1314) * fix annotation bug (#1320) * Valid POT configs for small HRNet models (#1313) * Disable NNCF optimization for FP16 models (#1312) * fliter object less than 1 pixel (#1305) * Fix some tests (#1322) * [Develop] Move drop_last into MPA (#1357) * Apply changes from releases/v0.3.1-geti1.0.0 (#1337) * anomaly save_model bugfix (#1300) * upgrade networkx module version (#1303) * Forward CVS-94422 size bug fix PR to release branch (#1326) * Valid POT configs for small HRNet models (#1317) * [Release branch] Disable NNCF optimization for FP16 models (#1319) * [RELEASE] CVS-95549 - Hierarchical classification training failed without obvious reason (#1329) * Fix h-label: per-group softmax (#1332) * Fix dataset length bug in mpa task (#1338) * Fix drop_last key issue for det/set (#1340) * Hot-fix for OV inference for iseg output (#1345) * Fix nncf model export bug (#1346) * Fixed merge error (#1359) * Update evaluation iou_thr of ins-seg (#1354) * fix pre-commit test (#1366) * Fix dataset item tests (#1360) * Fix OV Inference issues (tiling tests & detection tests) (#1361) * fix black & add xfail test cases (#1367) * Update check_nncf_graph. (#1330) * [Develop] Hot-fix OV inference issue in rotated detection (#1375) * [Develop] updated documents (#1383) * [CVS-94911] Fix difference between train and validation normalization pipeline (#1310) * Update configs for padim model (#1378) * updated QUICK_START_GUIDE.md (#1397) * Change ote threshold of openvino test for cls (#1401) * Normalize top-1 metrics to [0, 1] (#1394) * Tiling deployment (#1387) * Replace current saliency map generation with Recipro-CAM for cls (#1363) * Class-wise saliency map generation for the detection task (#1402) * Change submodule to develop (#1410) * Send full dataset to POT optimization function (#1379) & Convert NaN to num to make visible in geti UI (#1413) * Add active score evaluation to the classification task * [release/0.4.0][OTX] Enabling GPU execution for exported code (#1416) * [OTE][Release][XAI] Detection fix two stage bbox_head error (#1414) * Update SDK commit for exportable code (#1423) * HRNet-x and HRNe-18--mod2 configs update (#1419) * [Release] Enable tiling oriented detection for v0.4.0/geti1.1.0 (#1427) * [OTE][Releases v0.4.0][XAI] Hot-fix for Detection fix two stage error (#1433) * Temporary MPA branch while dev->otx merge process * Update doc & install for dev->otx changes * Update ote_sdk -> otx.api * Update ote_cli -> otx.cli * Update external/mmsegmentation -> otx/algorithms/segmentation * Align saliency map media instantiation over tasks (#1447) * Update external/d-o-r -> otx/algorithms/classification * Update external/mmdetection -> otx/algorithms/detection * Update external/mpa -> otx/algorithms/* * Fix CLI test run for better error message * Numpy constraint for deprecated np.bool error * Capture stderr only * Align numpy requirement * [OTX/Anomaly] Add changes from external to otx (#1452) * Add changes from external to otx * Address PR comments * Update config files + remove backbone from base * Fix pre-merge checks * Fix pre-commit issues * Update exportable code commit * Fix indent error * Fix flake8 issue * Resolve softmax issue w/ FIXME for future work * Add tiling tests * Revert MPA branch to otx Signed-off-by: Songki Choi <songki.choi@intel.com> Co-authored-by: Eugene Liu <eugene.liu@intel.com> Co-authored-by: Ashwin Vaidya <ashwin.vaidya@intel.com> Co-authored-by: Jaeguk Hyun <jaeguk.hyun@intel.com> Co-authored-by: Nikita Savelyev <nikita.savelyev@intel.com> Co-authored-by: Jihwan Eom <jihwan.eom@intel.com> Co-authored-by: Harim Kang <harim.kang@intel.com> Co-authored-by: Soobee Lee <soobee.lee@intel.com> Co-authored-by: Lee, Soobee <soobeele@intel.com> Co-authored-by: Emily Chun <emily.chun@intel.com> Co-authored-by: ljcornel <ludo.cornelissen@intel.com> Co-authored-by: Eunwoo Shin <eunwoo.shin@intel.com> Co-authored-by: dlyakhov <daniil.lyakhov@intel.com> Co-authored-by: kprokofi <kirill.prokofiev@intel.com> Co-authored-by: Sungman Cho <sungman.cho@intel.com> Co-authored-by: Yunchu Lee <yunchu.lee@intel.com> Co-authored-by: Ashwin Vaidya <ashwinitinvaidya@gmail.com> Co-authored-by: Alexander Dokuchaev <alexander.dokuchaev@intel.com> Co-authored-by: Vladislav Sovrasov <vladislav.sovrasov@intel.com> Co-authored-by: Evgeny Tsykunov <e.tsykunov@gmail.com> Co-authored-by: Galina Zalesskaya <galina.zalesskaya@intel.com> Co-authored-by: dongkwan-kim <dongkwan.kim@intel.com> * Apply latest MPA openvinotoolkit/model_preparation_algorithm#105 Signed-off-by: Songki Choi <songki.choi@intel.com> Signed-off-by: Songki Choi <songki.choi@intel.com> Co-authored-by: eunwoosh <eunwoo.shin@intel.com> Co-authored-by: Eugene Liu <eugene.liu@intel.com> Co-authored-by: Ashwin Vaidya <ashwin.vaidya@intel.com> Co-authored-by: Jaeguk Hyun <jaeguk.hyun@intel.com> Co-authored-by: Nikita Savelyev <nikita.savelyev@intel.com> Co-authored-by: Jihwan Eom <jihwan.eom@intel.com> Co-authored-by: Harim Kang <harim.kang@intel.com> Co-authored-by: Soobee Lee <soobee.lee@intel.com> Co-authored-by: Lee, Soobee <soobeele@intel.com> Co-authored-by: Emily Chun <emily.chun@intel.com> Co-authored-by: ljcornel <ludo.cornelissen@intel.com> Co-authored-by: dlyakhov <daniil.lyakhov@intel.com> Co-authored-by: kprokofi <kirill.prokofiev@intel.com> Co-authored-by: Sungman Cho <sungman.cho@intel.com> Co-authored-by: Yunchu Lee <yunchu.lee@intel.com> Co-authored-by: Ashwin Vaidya <ashwinitinvaidya@gmail.com> Co-authored-by: Alexander Dokuchaev <alexander.dokuchaev@intel.com> Co-authored-by: Vladislav Sovrasov <vladislav.sovrasov@intel.com> Co-authored-by: Evgeny Tsykunov <e.tsykunov@gmail.com> Co-authored-by: Galina Zalesskaya <galina.zalesskaya@intel.com> Co-authored-by: dongkwan-kim <dongkwan.kim@intel.com>
sungmanc
added a commit
that referenced
this pull request
Dec 29, 2022
* make logfile saved in save-model-to directory * enable train * make main process train also * bugfix * refactor multi gpu training * make all processs have same output path * prevent child process from being termated by fokred main process * refactor multigpu implementation * refactor multi gpu implementation * modify argument help sentence * add multi gpu test code * align with pre-commit test * separate multi GPU manager class * modify train cli argument 'save-logs-to' to 'output-path' * remove tray excpet during killing child process * apply output_path to all tasks * change print to logger * skip multi gpu test if number of gpu is insufficient * fix typo * multi gpu test bugfix * isort fix * test case bugfix * fix typo and change some variable name * [OTX] Apply changes in develop to feature/otx branch (#1436) * Add tiling module (#1200) * Update submodule branch (#1222) * Enhance training schedule for multi-label classification (#1212) * [CVS-88098] Remove initialize from export functions (#1226) * Train graph added (#1211) * Add @attrs decorator for base configs (#1229) * Pretrained weight download error in MobilenetV3-large-1 of deep-object-reid in SC (#1233) * [Anomaly Task] Revert hpo template (#1230) * 🐞 [Anomaly Task] Fix progress bar (#1223) * [CVS-90555] Fix NaN value in classification (#1244) * update hpo_config.yaml (#1240) * [CVS-90400, CVS-91015] NNCF pruning supported tweaks (#1248) * [Anomaly Task] 🐞 Fix inference when model backbone changes (#1242) * [CVS-91472] Add pruning_supported value (#1263) * Pruning supported tweaks (#1256) * [CVS-90400, CVS-91015] NNCF pruning supported tweaks (#1248) * Revert "[CVS-90400, CVS-91015] NNCF pruning supported tweaks (#1248)" (#1269) * [OTE-TEST] Disable obsolete test cases (#1220) * [OTE-TEST] hot-fix for MPA performance tests (#1273) * [Anomaly Task] ✨ Upgrade anomalib (#1243) * Expose early stopping hyper-parameters for all tasks (#1241) * Resolve pre-commit issues (#1272) * Remove LazyEarlyStopHook in model_multilabel.py (#1281) * Removed xfail (#1239) * Implement IB loss for incremental learning in multi-class classification (#1289) * Edit num_workers and change MPA repo as a latest (#1314) * fix annotation bug (#1320) * Valid POT configs for small HRNet models (#1313) * Disable NNCF optimization for FP16 models (#1312) * fliter object less than 1 pixel (#1305) * Fix some tests (#1322) * [Develop] Move drop_last into MPA (#1357) * Apply changes from releases/v0.3.1-geti1.0.0 (#1337) * anomaly save_model bugfix (#1300) * upgrade networkx module version (#1303) * Forward CVS-94422 size bug fix PR to release branch (#1326) * Valid POT configs for small HRNet models (#1317) * [Release branch] Disable NNCF optimization for FP16 models (#1319) * [RELEASE] CVS-95549 - Hierarchical classification training failed without obvious reason (#1329) * Fix h-label: per-group softmax (#1332) * Fix dataset length bug in mpa task (#1338) * Fix drop_last key issue for det/set (#1340) * Hot-fix for OV inference for iseg output (#1345) * Fix nncf model export bug (#1346) * Fixed merge error (#1359) * Update evaluation iou_thr of ins-seg (#1354) * fix pre-commit test (#1366) * Fix dataset item tests (#1360) * Fix OV Inference issues (tiling tests & detection tests) (#1361) * fix black & add xfail test cases (#1367) * Update check_nncf_graph. (#1330) * [Develop] Hot-fix OV inference issue in rotated detection (#1375) * [Develop] updated documents (#1383) * [CVS-94911] Fix difference between train and validation normalization pipeline (#1310) * Update configs for padim model (#1378) * updated QUICK_START_GUIDE.md (#1397) * Change ote threshold of openvino test for cls (#1401) * Normalize top-1 metrics to [0, 1] (#1394) * Tiling deployment (#1387) * Replace current saliency map generation with Recipro-CAM for cls (#1363) * Class-wise saliency map generation for the detection task (#1402) * Change submodule to develop (#1410) * Send full dataset to POT optimization function (#1379) & Convert NaN to num to make visible in geti UI (#1413) * Add active score evaluation to the classification task * [release/0.4.0][OTX] Enabling GPU execution for exported code (#1416) * [OTE][Release][XAI] Detection fix two stage bbox_head error (#1414) * Update SDK commit for exportable code (#1423) * HRNet-x and HRNe-18--mod2 configs update (#1419) * [Release] Enable tiling oriented detection for v0.4.0/geti1.1.0 (#1427) * [OTE][Releases v0.4.0][XAI] Hot-fix for Detection fix two stage error (#1433) * Temporary MPA branch while dev->otx merge process * Update doc & install for dev->otx changes * Update ote_sdk -> otx.api * Update ote_cli -> otx.cli * Update external/mmsegmentation -> otx/algorithms/segmentation * Align saliency map media instantiation over tasks (#1447) * Update external/d-o-r -> otx/algorithms/classification * Update external/mmdetection -> otx/algorithms/detection * Update external/mpa -> otx/algorithms/* * Fix CLI test run for better error message * Numpy constraint for deprecated np.bool error * Capture stderr only * Align numpy requirement * [OTX/Anomaly] Add changes from external to otx (#1452) * Add changes from external to otx * Address PR comments * Update config files + remove backbone from base * Fix pre-merge checks * Fix pre-commit issues * Update exportable code commit * Fix indent error * Fix flake8 issue * Resolve softmax issue w/ FIXME for future work * Add tiling tests * Revert MPA branch to otx Signed-off-by: Songki Choi <songki.choi@intel.com> Co-authored-by: Eugene Liu <eugene.liu@intel.com> Co-authored-by: Ashwin Vaidya <ashwin.vaidya@intel.com> Co-authored-by: Jaeguk Hyun <jaeguk.hyun@intel.com> Co-authored-by: Nikita Savelyev <nikita.savelyev@intel.com> Co-authored-by: Jihwan Eom <jihwan.eom@intel.com> Co-authored-by: Harim Kang <harim.kang@intel.com> Co-authored-by: Soobee Lee <soobee.lee@intel.com> Co-authored-by: Lee, Soobee <soobeele@intel.com> Co-authored-by: Emily Chun <emily.chun@intel.com> Co-authored-by: ljcornel <ludo.cornelissen@intel.com> Co-authored-by: Eunwoo Shin <eunwoo.shin@intel.com> Co-authored-by: dlyakhov <daniil.lyakhov@intel.com> Co-authored-by: kprokofi <kirill.prokofiev@intel.com> Co-authored-by: Sungman Cho <sungman.cho@intel.com> Co-authored-by: Yunchu Lee <yunchu.lee@intel.com> Co-authored-by: Ashwin Vaidya <ashwinitinvaidya@gmail.com> Co-authored-by: Alexander Dokuchaev <alexander.dokuchaev@intel.com> Co-authored-by: Vladislav Sovrasov <vladislav.sovrasov@intel.com> Co-authored-by: Evgeny Tsykunov <e.tsykunov@gmail.com> Co-authored-by: Galina Zalesskaya <galina.zalesskaya@intel.com> Co-authored-by: dongkwan-kim <dongkwan.kim@intel.com> * Apply latest MPA openvinotoolkit/model_preparation_algorithm#105 Signed-off-by: Songki Choi <songki.choi@intel.com> Signed-off-by: Songki Choi <songki.choi@intel.com> Co-authored-by: eunwoosh <eunwoo.shin@intel.com> Co-authored-by: Eugene Liu <eugene.liu@intel.com> Co-authored-by: Ashwin Vaidya <ashwin.vaidya@intel.com> Co-authored-by: Jaeguk Hyun <jaeguk.hyun@intel.com> Co-authored-by: Nikita Savelyev <nikita.savelyev@intel.com> Co-authored-by: Jihwan Eom <jihwan.eom@intel.com> Co-authored-by: Harim Kang <harim.kang@intel.com> Co-authored-by: Soobee Lee <soobee.lee@intel.com> Co-authored-by: Lee, Soobee <soobeele@intel.com> Co-authored-by: Emily Chun <emily.chun@intel.com> Co-authored-by: ljcornel <ludo.cornelissen@intel.com> Co-authored-by: dlyakhov <daniil.lyakhov@intel.com> Co-authored-by: kprokofi <kirill.prokofiev@intel.com> Co-authored-by: Sungman Cho <sungman.cho@intel.com> Co-authored-by: Yunchu Lee <yunchu.lee@intel.com> Co-authored-by: Ashwin Vaidya <ashwinitinvaidya@gmail.com> Co-authored-by: Alexander Dokuchaev <alexander.dokuchaev@intel.com> Co-authored-by: Vladislav Sovrasov <vladislav.sovrasov@intel.com> Co-authored-by: Evgeny Tsykunov <e.tsykunov@gmail.com> Co-authored-by: Galina Zalesskaya <galina.zalesskaya@intel.com> Co-authored-by: dongkwan-kim <dongkwan.kim@intel.com>
harimkang
added a commit
that referenced
this pull request
Dec 30, 2022
* Initial commit: integrate MPA intothe OTX * Update Self-SL for seg * Enable pre-commit tests by ignoring imports, remove useless requirements * Edit the requirements * Add dist folder * fix black * [OTX] Rebase latest changes for MPA merge (#1468) * make logfile saved in save-model-to directory * enable train * make main process train also * bugfix * refactor multi gpu training * make all processs have same output path * prevent child process from being termated by fokred main process * refactor multigpu implementation * refactor multi gpu implementation * modify argument help sentence * add multi gpu test code * align with pre-commit test * separate multi GPU manager class * modify train cli argument 'save-logs-to' to 'output-path' * remove tray excpet during killing child process * apply output_path to all tasks * change print to logger * skip multi gpu test if number of gpu is insufficient * fix typo * multi gpu test bugfix * isort fix * test case bugfix * fix typo and change some variable name * [OTX] Apply changes in develop to feature/otx branch (#1436) * Add tiling module (#1200) * Update submodule branch (#1222) * Enhance training schedule for multi-label classification (#1212) * [CVS-88098] Remove initialize from export functions (#1226) * Train graph added (#1211) * Add @attrs decorator for base configs (#1229) * Pretrained weight download error in MobilenetV3-large-1 of deep-object-reid in SC (#1233) * [Anomaly Task] Revert hpo template (#1230) * 🐞 [Anomaly Task] Fix progress bar (#1223) * [CVS-90555] Fix NaN value in classification (#1244) * update hpo_config.yaml (#1240) * [CVS-90400, CVS-91015] NNCF pruning supported tweaks (#1248) * [Anomaly Task] 🐞 Fix inference when model backbone changes (#1242) * [CVS-91472] Add pruning_supported value (#1263) * Pruning supported tweaks (#1256) * [CVS-90400, CVS-91015] NNCF pruning supported tweaks (#1248) * Revert "[CVS-90400, CVS-91015] NNCF pruning supported tweaks (#1248)" (#1269) * [OTE-TEST] Disable obsolete test cases (#1220) * [OTE-TEST] hot-fix for MPA performance tests (#1273) * [Anomaly Task] ✨ Upgrade anomalib (#1243) * Expose early stopping hyper-parameters for all tasks (#1241) * Resolve pre-commit issues (#1272) * Remove LazyEarlyStopHook in model_multilabel.py (#1281) * Removed xfail (#1239) * Implement IB loss for incremental learning in multi-class classification (#1289) * Edit num_workers and change MPA repo as a latest (#1314) * fix annotation bug (#1320) * Valid POT configs for small HRNet models (#1313) * Disable NNCF optimization for FP16 models (#1312) * fliter object less than 1 pixel (#1305) * Fix some tests (#1322) * [Develop] Move drop_last into MPA (#1357) * Apply changes from releases/v0.3.1-geti1.0.0 (#1337) * anomaly save_model bugfix (#1300) * upgrade networkx module version (#1303) * Forward CVS-94422 size bug fix PR to release branch (#1326) * Valid POT configs for small HRNet models (#1317) * [Release branch] Disable NNCF optimization for FP16 models (#1319) * [RELEASE] CVS-95549 - Hierarchical classification training failed without obvious reason (#1329) * Fix h-label: per-group softmax (#1332) * Fix dataset length bug in mpa task (#1338) * Fix drop_last key issue for det/set (#1340) * Hot-fix for OV inference for iseg output (#1345) * Fix nncf model export bug (#1346) * Fixed merge error (#1359) * Update evaluation iou_thr of ins-seg (#1354) * fix pre-commit test (#1366) * Fix dataset item tests (#1360) * Fix OV Inference issues (tiling tests & detection tests) (#1361) * fix black & add xfail test cases (#1367) * Update check_nncf_graph. (#1330) * [Develop] Hot-fix OV inference issue in rotated detection (#1375) * [Develop] updated documents (#1383) * [CVS-94911] Fix difference between train and validation normalization pipeline (#1310) * Update configs for padim model (#1378) * updated QUICK_START_GUIDE.md (#1397) * Change ote threshold of openvino test for cls (#1401) * Normalize top-1 metrics to [0, 1] (#1394) * Tiling deployment (#1387) * Replace current saliency map generation with Recipro-CAM for cls (#1363) * Class-wise saliency map generation for the detection task (#1402) * Change submodule to develop (#1410) * Send full dataset to POT optimization function (#1379) & Convert NaN to num to make visible in geti UI (#1413) * Add active score evaluation to the classification task * [release/0.4.0][OTX] Enabling GPU execution for exported code (#1416) * [OTE][Release][XAI] Detection fix two stage bbox_head error (#1414) * Update SDK commit for exportable code (#1423) * HRNet-x and HRNe-18--mod2 configs update (#1419) * [Release] Enable tiling oriented detection for v0.4.0/geti1.1.0 (#1427) * [OTE][Releases v0.4.0][XAI] Hot-fix for Detection fix two stage error (#1433) * Temporary MPA branch while dev->otx merge process * Update doc & install for dev->otx changes * Update ote_sdk -> otx.api * Update ote_cli -> otx.cli * Update external/mmsegmentation -> otx/algorithms/segmentation * Align saliency map media instantiation over tasks (#1447) * Update external/d-o-r -> otx/algorithms/classification * Update external/mmdetection -> otx/algorithms/detection * Update external/mpa -> otx/algorithms/* * Fix CLI test run for better error message * Numpy constraint for deprecated np.bool error * Capture stderr only * Align numpy requirement * [OTX/Anomaly] Add changes from external to otx (#1452) * Add changes from external to otx * Address PR comments * Update config files + remove backbone from base * Fix pre-merge checks * Fix pre-commit issues * Update exportable code commit * Fix indent error * Fix flake8 issue * Resolve softmax issue w/ FIXME for future work * Add tiling tests * Revert MPA branch to otx Signed-off-by: Songki Choi <songki.choi@intel.com> Co-authored-by: Eugene Liu <eugene.liu@intel.com> Co-authored-by: Ashwin Vaidya <ashwin.vaidya@intel.com> Co-authored-by: Jaeguk Hyun <jaeguk.hyun@intel.com> Co-authored-by: Nikita Savelyev <nikita.savelyev@intel.com> Co-authored-by: Jihwan Eom <jihwan.eom@intel.com> Co-authored-by: Harim Kang <harim.kang@intel.com> Co-authored-by: Soobee Lee <soobee.lee@intel.com> Co-authored-by: Lee, Soobee <soobeele@intel.com> Co-authored-by: Emily Chun <emily.chun@intel.com> Co-authored-by: ljcornel <ludo.cornelissen@intel.com> Co-authored-by: Eunwoo Shin <eunwoo.shin@intel.com> Co-authored-by: dlyakhov <daniil.lyakhov@intel.com> Co-authored-by: kprokofi <kirill.prokofiev@intel.com> Co-authored-by: Sungman Cho <sungman.cho@intel.com> Co-authored-by: Yunchu Lee <yunchu.lee@intel.com> Co-authored-by: Ashwin Vaidya <ashwinitinvaidya@gmail.com> Co-authored-by: Alexander Dokuchaev <alexander.dokuchaev@intel.com> Co-authored-by: Vladislav Sovrasov <vladislav.sovrasov@intel.com> Co-authored-by: Evgeny Tsykunov <e.tsykunov@gmail.com> Co-authored-by: Galina Zalesskaya <galina.zalesskaya@intel.com> Co-authored-by: dongkwan-kim <dongkwan.kim@intel.com> * Apply latest MPA openvinotoolkit/model_preparation_algorithm#105 Signed-off-by: Songki Choi <songki.choi@intel.com> Signed-off-by: Songki Choi <songki.choi@intel.com> Co-authored-by: eunwoosh <eunwoo.shin@intel.com> Co-authored-by: Eugene Liu <eugene.liu@intel.com> Co-authored-by: Ashwin Vaidya <ashwin.vaidya@intel.com> Co-authored-by: Jaeguk Hyun <jaeguk.hyun@intel.com> Co-authored-by: Nikita Savelyev <nikita.savelyev@intel.com> Co-authored-by: Jihwan Eom <jihwan.eom@intel.com> Co-authored-by: Harim Kang <harim.kang@intel.com> Co-authored-by: Soobee Lee <soobee.lee@intel.com> Co-authored-by: Lee, Soobee <soobeele@intel.com> Co-authored-by: Emily Chun <emily.chun@intel.com> Co-authored-by: ljcornel <ludo.cornelissen@intel.com> Co-authored-by: dlyakhov <daniil.lyakhov@intel.com> Co-authored-by: kprokofi <kirill.prokofiev@intel.com> Co-authored-by: Sungman Cho <sungman.cho@intel.com> Co-authored-by: Yunchu Lee <yunchu.lee@intel.com> Co-authored-by: Ashwin Vaidya <ashwinitinvaidya@gmail.com> Co-authored-by: Alexander Dokuchaev <alexander.dokuchaev@intel.com> Co-authored-by: Vladislav Sovrasov <vladislav.sovrasov@intel.com> Co-authored-by: Evgeny Tsykunov <e.tsykunov@gmail.com> Co-authored-by: Galina Zalesskaya <galina.zalesskaya@intel.com> Co-authored-by: dongkwan-kim <dongkwan.kim@intel.com> * Resolve some flake8 issues * Fix semisl trainer import (#1471) Co-authored-by: Lee, Soobee <soobeele@intel.com> * Fixisort * Rebase and fix isort Signed-off-by: Songki Choi <songki.choi@intel.com> Co-authored-by: Songki Choi <songki.choi@intel.com> Co-authored-by: eunwoosh <eunwoo.shin@intel.com> Co-authored-by: Eugene Liu <eugene.liu@intel.com> Co-authored-by: Ashwin Vaidya <ashwin.vaidya@intel.com> Co-authored-by: Jaeguk Hyun <jaeguk.hyun@intel.com> Co-authored-by: Nikita Savelyev <nikita.savelyev@intel.com> Co-authored-by: Jihwan Eom <jihwan.eom@intel.com> Co-authored-by: Harim Kang <harim.kang@intel.com> Co-authored-by: Soobee Lee <soobee.lee@intel.com> Co-authored-by: Lee, Soobee <soobeele@intel.com> Co-authored-by: Emily Chun <emily.chun@intel.com> Co-authored-by: ljcornel <ludo.cornelissen@intel.com> Co-authored-by: dlyakhov <daniil.lyakhov@intel.com> Co-authored-by: kprokofi <kirill.prokofiev@intel.com> Co-authored-by: Yunchu Lee <yunchu.lee@intel.com> Co-authored-by: Ashwin Vaidya <ashwinitinvaidya@gmail.com> Co-authored-by: Alexander Dokuchaev <alexander.dokuchaev@intel.com> Co-authored-by: Vladislav Sovrasov <vladislav.sovrasov@intel.com> Co-authored-by: Evgeny Tsykunov <e.tsykunov@gmail.com> Co-authored-by: Galina Zalesskaya <galina.zalesskaya@intel.com> Co-authored-by: dongkwan-kim <dongkwan.kim@intel.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.
This PR includes solve issue CVS-94911.
Previous implementation only look into pipeline of the most inside dataset
I change the code to also look into outside dataset
I think we should refactor this function recursive way to handle various dataset settings.