* [Enhance] Support non-scalar type metric value.

* Refactor support.

* Fix non-scalar tags problem during validation

* Update tag processor.

* Separate init_cfgs to pretrained_cfg and other_cfgs

* Fix unit test

* update documentation

* Fix render of initialize.md

* Fix as comment

* rename initialize.md to weight_initialization.md

* add file

* fix ci

* rename weight_initialization.md to initialize.md

* Fix duplicated .md

* [Fix] Fix config cannot parse tmp path like

* Add comments

* Add comments

* Apply suggestions from code review

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* [Enhance] Ensure metrics is not empty when saving best ckpts

* fix warn to warning

* delete a unnecessary method

* Add build_runner_with_tta and PrepareTTAHook

* rename hook file

* support build tta runner with runner type

* add unit test

* Add build_runner_with_tta to index.rst

* minor refine

* Add runner test cast

* Fix unit test

* fix unit test

* tmp save

* pop None if key does not exist

* Fix is_model_wrapper and force register class in test_runner

* [Fix] Fix is_model_wrapper

* destroy group after ut

* register module in testcase

* pass through unit test

* fix as comment

* remove breakpoint

* remove mmengine/testing/runner_test_cast.py

* minor refine

* minor refine

* minor refine

* set default data preprocessor for model

* minor refine

* minor refine

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* fix lint

* Fix unit test

* replace  with  in ImgDataPreprocessor

* Fix as comment

* add inference tutorial in advanced tutorial

* update index.rst

* add tta example

* refine tta tutorial

* Add english tutorial

* add note for build_runner_with_tta

* Fix as comment

* add examples

* remove chinese comment

* Update docs/en/advanced_tutorials/test_time_augmentation.md

Co-authored-by: RangiLyu <lyuchqi@gmail.com>

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>
Co-authored-by: RangiLyu <lyuchqi@gmail.com>

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* [Feature] Add profiler hook functionality

* Apply suggestions from code review

* Update mmengine/hooks/profiler_hook.py

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* only test timer in linux

* Fix merge_stage_test.yml

* remove install ffmpeg

* test_ci

* test_ci

* Fix python -m pip install pip --upgrade to pip install

* don't test unittest with sleep 1 in windows

* debug with tmate

* increase timeout

* increase timeout

* skip test gloo

* skip test gloo

* fix synteax error

* skip test gloo in torch 1.13

* skip test gloo in torch 1.13

* skip testing setup_env in macOS

* skip test test_dist

* skip test setup

* restore triggered during push

* clean the code

* debug macos

* refine macOS CI

* test merge stage test

* trigger CI during push to main

* Fix as comment

* [Feature] Support registry auto import modules.

* update

* rebase and fix ut

* add docstring

* remove count_registered_modules

* update docstring

* resolve comments

* resolve comments

* rename ut

* fix warning

* avoid BC breaking

* update doc

* update doc

* resolve comments

* Update docstring of `structures`

* update docs

* add `import torch` to `examples`

* [Fix] Fix failed unit tests caused by the new released numpy

* minor fix

* fix zero_optimizer error with param groups when pytorch < 1.12.0

* add docstring

* fix docstring

* add unittest

* change ut to use a valid paramwise_cfg

* modify ut

* fix as comments

* [Fix] Fix bias decay mult of depth-wise conv.

* support flatten param weight decay multiplier

* add unit tests

* REMOVE TODO

* update doc

* enhance get_torchvision_model

* remove mmcv

* fix BaseModel `to` method to be consistent with torch.nn.Module

* fix data_preprocessor as well

* fix docstring alignment

* fix docstring alignment

* feat: add conf_path in PetrelBackend

* Update mmengine/fileio/backends/petrel_backend.py

Co-authored-by: Mashiro <57566630+HAOCHENYE@users.noreply.github.com>

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>
Co-authored-by: Mashiro <57566630+HAOCHENYE@users.noreply.github.com>

* remove redundant attribute in dump or print

* optimize efficiency

Co-authored-by: HAOCHENYE <21724054@zju.edu.cn>

* [Enhance] Enhance revert_sync_batchnorm and convert_sync_batchnorm

* [Enhance] Enhance revert_sync_batchnorm and convert_sync_batchnorm

* Fix unit test

* Add coments

* Refine comments

* clean the code

* revert convert_sync_batchnorm

* revert convert_sync_batchnorm

* refine comment

* fix CI

* fix CI

* [Enhance] support eta_min_ratio in CosineAnnealingParamScheduler

* [doc] fix docstring

* [Enhance] add ut for eta_min_ratio

* [doc] update docstring

* avoid bc-breaking of eta_min

* [doc] add docstring in CosineAnnealingParamScheduler and CosineAnnealingMomentum

* Apply suggestions from code review

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* Add test of get_hooks_info()

* Change to use original Runner for get_hook_info() test

* Change to test after_train_iter hooks for get_hook_info()

* Complement the stages list

* Add logging hooks information in Runner.__init__()

* Rearrange the stages list

* Restore the stages to tuple type

* Clean the unnecessary changes

* Replace  statement with TestCase's methods

* add test stages in method_stages_map

* change the hooks info into a f-string

* return list(trigger_stages) directly

* change keys of method_stages_map

* Fix previous changes to method_stages_map.keys

* right allign the log

* fix as comment

* Add comments for magic number 3

* remove max_len_str

* Update mmengine/runner/log_processor.py

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* Add runner test cast

* Fix unit test

* fix unit test

* pop None if key does not exist

* Fix is_model_wrapper and force register class in test_runner

* [Fix] Fix is_model_wrapper

* destroy group after ut

* register module in testcase

* fix as comment

* minor refine

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* fix lint

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* [Fix] Fix offline eval dataset index error.

* update

* update

* fix docstring and documentation used for hub.get_model

* fix config name in test_hub

* fix docstring mentioned faster-rcnn in utils

* [Fix] Fix ConcatDataset error in VOCDataset

* minor fix

* minor fix

* minor fix

* add UT

* minor fix

* minor fix

* minor fix

* Update mmengine/dataset/dataset_wrapper.py

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* minor fix

* Update mmengine/dataset/dataset_wrapper.py

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* [Fix] fix CosineRestart eta_min

* add ut case

* Enhance unit test

Enhance unit test

* remove unused code

Co-authored-by: HAOCHENYE <21724054@zju.edu.cn>

* [Fix] Fix  cound not handle string data

* Minor refine

* Refine type hint

Refine type hint

* fix as comment

* Minor refine

* Update mmengine/model/base_model/data_preprocessor.py

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* [Feature] Support torch ZeRORedundancyOptimizer

Co-authored-by: Junhwa Song <ethan9867@gmail.com>
Signed-off-by: Junhwa Song <ethan9867@gmail.com>
Signed-off-by: Hakjin Lee <nijkah@gmail.com>

* lint

* Fix saving optimizer state_dict

* Fix handling import error

* Add test case

* fix UT

* Revert "fix UT"

This reverts commit dd64538960ff7440c6020f533d43945ffc23f2d2.

* fix handling import in UT

* Fix saving zero checkpoint and delete redundant master_only

* lint

* test unittest

* Fix handling impor error

* Fix UT condition

* Edit docstrings

* Fix typo

* Skip redundant procudure in checkpoint hook

* fix typo again

* Update mmengine/optim/optimizer/zero_optimizer.py

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* Add api info

* lint

* Fix lint

* Handling AmpOptimWrapper case

* handling overlap_with_ddp

* Fix error

Signed-off-by: Junhwa Song <ethan9867@gmail.com>
Signed-off-by: Hakjin Lee <nijkah@gmail.com>
Co-authored-by: Junhwa Song <ethan9867@gmail.com>
Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* call register_all_modules  in Registry.get()

* Fix ci

* fix scope bug (scope_name -> scope), for temp sync

* Fix unit test

* Refine log information

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* Fix typo

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* [Enhancement] Add a deprecated function

* Refine and add ut

* tmp save

* minor refine

* Fix unit test

* Fix unit test

* fix docstring

* Update utils.rst

* init npu

* Update mmengine/optim/optimizer/amp_optimizer_wrapper.py

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* Update mmengine/dist/dist.py

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* change to is_hccl_backend

* Update mmengine/optim/optimizer/amp_optimizer_wrapper.py

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* add comment with AmpOptimWrapper

* Update mmengine/runner/amp.py

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* Update mmengine/runner/amp.py

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* add npu fn in base_model

* Update mmengine/optim/optimizer/amp_optimizer_wrapper.py

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* clean lint

* Update mmengine/optim/optimizer/amp_optimizer_wrapper.py

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* Update mmengine/model/base_model/base_model.py

Co-authored-by: Mashiro <57566630+HAOCHENYE@users.noreply.github.com>

* add is_npu_available

* try to fix

* Add comments

* Refine grammar

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>
Co-authored-by: Mashiro <57566630+HAOCHENYE@users.noreply.github.com>
Co-authored-by: HAOCHENYE <21724054@zju.edu.cn>

* fix add_config func is not called bug
fix wandb add_config bug
add wandb log_code feature

* move log_code_name param to the last one
add log_code_name param to docstring

* add config only when there is a cfg

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* add unit test for log_code_name param of WandbVisBackend

* Update mmengine/visualization/vis_backend.py

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* use log_code_kwargs instead of only log_code_name
add PATH_FILTERS registry

* use log_code_kwargs instead of only log_code_name
add PATH_FILTERS registry

* fix add config unit test

* roll back to log_code_name version

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* Fix error format of log message

* Fix unit test

* remove unnecessary comment

* [Fix] Make autocast compatible with mps

* Enhance unit test

* fix unit test

* clean the code

* fix unit test

* [Enhance] add documents for , and support clip grad by value

* refine docstring

* fix as comment

* Fix as comment

* minor refine

* minor refine

* remove error comment for clip grad

* refine docstring

* add non_blocking feature to BaseDataPreprocessor

* Update mmengine/model/base_model/data_preprocessor.py

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* move new parameters to the last to avoid bc issue

* Update mmengine/model/base_model/data_preprocessor.py

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* Update mmengine/model/base_model/data_preprocessor.py

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>

* First commit

* add BaseTestTimeAugModel

* Add unit test

* move loop logic to test_step

* fix ddp test

* rename model to module

* optim __init__

* Fix as comment

* Fix as comment

* make val_step should not be called

* make tta do not inherit base model

* Fix unit test

* Enhance docstring

* Fix as comment

* Fix as comment

* minor refine

* minor refine

* minor refine

* fix unit test

* minor refine

* minor refine

* minor refine

* minor refine

* minor refine

* minor refine

* fix unit test

* fix dist ut

* fix config & hub ut break due to mmdet faster-rcnn rename