* update support for linearly learning rate decay
* Fix LinearAnnealingLrUpdaterHook, update LinearAnnealingMomentumUpdaterHook, add unit test
add docstring
add docstring
update linear lr momentum schedule test
fix ci
Fix CI
* add lr and momentum hook to runner and hooks package
add lr and momentum hook to runner and hooks package
* replace multi_optimziers with multi_optimizers
Co-authored-by: HAOCHENYE <21724054@zju.edu.cn>
Co-authored-by: Mashiro <57566630+HAOCHENYE@users.noreply.github.com>
* detect detect_anomalous_params
* fix default value
* merge two case
* fix none case
* add unitest
* fix typo
* change level to error
* fix type
* add more details in docstr
* add windows CI
* clean versions
* only allow pt1.7 on windows
* fix windows install issue
* add win cpu
* fix win command
* clean unnecessary command
* resolve turbojpeg & tempfile on win
* replace os.readlink with os.path.realpath
* fix windows ci
* close file before removing it
* fix windows ci
* fix symlink on windows
* fix windows ci
* fix windows ci
* fix windows ci
* fix windows ci
* fix windows ci
* fix windows ci
* fix windows ci
* fix windows ci
* fix windows ci
* modify according to comment
Co-authored-by: zhouzaida <zhouzaida@163.com>
* Add gradient cumulative optimizer
fixes#190
* Update optimizer.py
* Update optimizer.py
* fix loss scale improperly in last equivalent_iter
* Add `GradientCumulativeOptimizerHook` in `__init__.py`.
* Add docstring of `GradientCumulativeOptimizerHook`.
* Add type check, BN warning and resume warning. And fix typo, lint the
code.
* Add unit test
* Update docstring example.
* Change GradientCumulativeOptimizerHook `__init__` arguments.
* Add GradientCumulativeOptimzierHook unit tests with IterBasedRunner.
* Add GradientCumulativeFp16OptimizerHook.
* Add unit tests of GradientCumulativeFp16OptimizerHook
* Use '!=' instead of '>' to determine resume
Co-authored-by: Zhiyuan Chen <this@zyc.ai>
* add flat cosine lr updater
* add test
* add doc
* update doc
* reformat
* update unittest
* update test flat cos
* remove momentum hook test
* update test
* change assert to ValueError
* fix unittest
* add by_epoch=True unittest
* change to start_percent
* change to start_percent in test
* support print using hooks before running.
* Support to print hook trigger stages.
* Print stage-wise hook infos. And make `stages` as class attribute of
`Hook`
* Add util function `is_method_overriden` and use it in
`Hook.get_trigger_stages`.
* Add unit tests.
* Move `is_method_overriden` to `mmcv/utils/misc.py`
* Improve hook info text.
* Add base_class argument type assertion, and fix some typos.
* Remove `get_trigger_stages` to `get_triggered_stages`
* Use f-string.
* Refine default hooks and custom hooks priority rank.
* Add unit tests for custom hooks with string priority.
* Use priority `ABOVE_NORMAL` and `BELOW_NORMAL` instead of `HIGHER` and
`LOWER`.
And add unit tests for custom hook with the same priority as
default hooks.
* Assign different priority to default hooks, and add custom hook register in base runner.
* Add custom hook register in example train file
* Add unittest of custom hook
* Code format
* support clipping min_lr in StepLrUpdaterHook
* add docstring for StepLrUpdaterHook
* fix small bugs
* add unit test for StepLrUpdaterHook
* fix linting error
* [Fix] OneCycleLrUpdaterHook interface
* revise according to comments
* revise according to comments
* add test
* fix lint
* revise according to comments
* minors
* add pytest param
* fix lint
* ci
* Update lr_updater.py
since epoch/iteration in runner starts with 0, we shouldn't leave the latter iteration to former (12th epoch for example, with first period equal to 12) period.
* Update lr_updater.py
* Update test_hooks.py
* Add build_runner
* Parametrize test_runner
* Add imports to runner __init__
* Refactor max_iters and max_epochs from run to init
* Add assertion error messages
* Add test_builder
* Make change retro-compatible
* Raise ValueError if max_epochs and max_iters
* add ema hook
* add ema hook resume
* add ema hook test
* fix typo
* fix according to comment
* delete logger
* fix according to comment
* fix unitest
* fix typo
* fix according to comment
* change to resume_from
* typo
* fix isort
* feat: add CosineRestartLrUpdaterHook
* style: rename period to periods
* fix: bug in period 0
* feat: rename eta_min to min_lr and add min_lr_ratio
* docs: fix docstring of restart lr updater
* refactor: use annealing_cos
* docs: add docstring to annealing_cos
* feat: cosine restart lr update hook
* refactor: modify code order for unittest
* add a BaseRunner and rename Runner to EpochBasedRunner
* fix the train/val step
* bug fix
* update unit tests
* fix unit tests
* raise an error if both batch_processor and train_step are set
* add a unit test
Artem
Sharpiless
saurbhc
Zaida Zhou
Hakjin Lee
Mashiro
imyhxy
fatih
David de la Iglesia Castro
Shilong Zhang
Wenwei Zhang
Ma Zerun
Gu Wang
Ziyi Wu
gengenkai
Ziyi Wu
Miao Zheng
Ycr
Jerry Jiarui XU
Kuro Latency
Wang Xinjiang
Yawei Li
Kai Chen
Harry