[Doc] Add English version document of the Training Engine #2797

xiexinch · 2023-03-27T08:45:52Z

Motivation

As title.

Modification

docs/en/advanced_guides/engine.md

csatsurnh · 2023-03-28T08:33:31Z

docs/en/advanced_guides/engine.md

-# Engine
+# Training Engine
+
+MMEngine defines some [basic loop controllers](https://github.com/open-mmlab/mmengine/blob/main/mmengine/runner/loops.py) such as epoch-based training loop (`EpochBasedTrainLoop`), iteration-based training loop (`IterBasedTrainLoop`), standard validation loop (`ValLoop`), and standard testing loop (`TestLoop`).


Suggested change

MMEngine defines some [basic loop controllers](https://github.com/open-mmlab/mmengine/blob/main/mmengine/runner/loops.py) such as epoch-based training loop (`EpochBasedTrainLoop`), iteration-based training loop (`IterBasedTrainLoop`), standard validation loop (`ValLoop`), and standard testing loop (`TestLoop`).

MMEngine defined some [basic loop controllers](https://github.com/open-mmlab/mmengine/blob/main/mmengine/runner/loops.py) such as epoch-based training loop (`EpochBasedTrainLoop`), iteration-based training loop (`IterBasedTrainLoop`), standard validation loop (`ValLoop`), and standard testing loop (`TestLoop`).

csatsurnh · 2023-03-28T08:37:43Z

docs/en/advanced_guides/engine.md

+
+MMEngine defines some [basic loop controllers](https://github.com/open-mmlab/mmengine/blob/main/mmengine/runner/loops.py) such as epoch-based training loop (`EpochBasedTrainLoop`), iteration-based training loop (`IterBasedTrainLoop`), standard validation loop (`ValLoop`), and standard testing loop (`TestLoop`).
+
+OpenMMLab's algorithm libraries like MMSegmentation abstract model training, testing, and inference as executors (`Runner`) to handle. Users can use the default executor in MMEngine directly or modify the executor to meet customized needs. This document mainly introduces how users can configure existing running settings, hooks, and optimizers' basic concepts and usage methods.


Suggested change

OpenMMLab's algorithm libraries like MMSegmentation abstract model training, testing, and inference as executors (`Runner`) to handle. Users can use the default executor in MMEngine directly or modify the executor to meet customized needs. This document mainly introduces how users can configure existing running settings, hooks, and optimizers' basic concepts and usage methods.

OpenMMLab's algorithm libraries like MMSegmentation abstract model training, testing, and inference as `Runner` to handle. Users can use the default runner in MMEngine directly or modify the runner to meet customized needs. This document mainly introduces how users can configure existing running settings, hooks, and optimizers' basic concepts and usage methods.

csatsurnh · 2023-03-28T08:58:23Z

docs/en/advanced_guides/engine.md

+
+### Configuring Training Length
+
+Loop controllers refer to the execution process during training, validation, and testing. They use `train_cfg`, `val_cfg`, and `test_cfg` to build these processes in the configuration file. MMSegmentation sets commonly used training lengths in the configs/_base_/schedules folder's train_cfg.


Suggested change

Loop controllers refer to the execution process during training, validation, and testing. They use `train_cfg`, `val_cfg`, and `test_cfg` to build these processes in the configuration file. MMSegmentation sets commonly used training lengths in the configs/_base_/schedules folder's train_cfg.

Loop controllers refer to the execution process during training, validation, and testing. `train_cfg`, `val_cfg`, and `test_cfg` are used to build these processes in the configuration file. MMSegmentation sets commonly used training iterations in `train_cfg` under the `configs/_base_/schedules` folder.

docs/en/advanced_guides/engine.md

csatsurnh · 2023-03-28T09:53:18Z

docs/en/advanced_guides/engine.md

+
+Custom hooks are defined in the configuration through `custom_hooks`, and `Runner` registers them using the [`register_custom_hooks`](https://github.com/open-mmlab/mmengine/blob/main/mmengine/runner/runner.py#L1820) method.
+
+The priority of custom hooks needs to be set in the configuration file; if not set, it will be set to `NORMAL` by default. The following are some custom hooks implemented in MMEngine:


Suggested change

The priority of custom hooks needs to be set in the configuration file; if not set, it will be set to `NORMAL` by default. The following are some custom hooks implemented in MMEngine:

The priority of custom hooks needs to be set in the configuration file; if not, it will be set to `NORMAL` by default. The following are some custom hooks implemented in MMEngine:

csatsurnh · 2023-03-28T10:00:22Z

docs/en/advanced_guides/engine.md

+
+### SegVisualizationHook
+
+MMSegmentation implements [`SegVisualizationHook`](https://github.com/open-mmlab/mmsegmentation/blob/dev-1.x/mmseg/engine/hooks/visualization_hook.py#L17), which is used to visualize prediction results during validation and testing.


Suggested change

MMSegmentation implements [`SegVisualizationHook`](https://github.com/open-mmlab/mmsegmentation/blob/dev-1.x/mmseg/engine/hooks/visualization_hook.py#L17), which is used to visualize prediction results during validation and testing.

MMSegmentation implemented [`SegVisualizationHook`](https://github.com/open-mmlab/mmsegmentation/blob/dev-1.x/mmseg/engine/hooks/visualization_hook.py#L17), which is used to visualize prediction results during validation and testing.

csatsurnh · 2023-03-28T10:02:33Z

docs/en/advanced_guides/engine.md

+
+OpenMMLab's algorithm libraries like MMSegmentation abstract model training, testing, and inference as executors (`Runner`) to handle. Users can use the default executor in MMEngine directly or modify the executor to meet customized needs. This document mainly introduces how users can configure existing running settings, hooks, and optimizers' basic concepts and usage methods.
+
+## Configuring Running Settings


Suggested change

## Configuring Running Settings

## Configuring Runtime Settings

csatsurnh · 2023-03-28T10:04:14Z

docs/en/advanced_guides/engine.md

+
+## Optimizer
+
+In the previous configuration and runtime settings, we provided a simple example of configuring the training optimizer. This section will further detail how to configure optimizers in MMSegmentation.


Suggested change

In the previous configuration and runtime settings, we provided a simple example of configuring the training optimizer. This section will further detail how to configure optimizers in MMSegmentation.

In the previous configuration and runtime settings, we provided a simple example of configuring the training optimizer. This section will further detailly introduce how to configure optimizers in MMSegmentation.

csatsurnh · 2023-03-28T10:08:35Z

docs/en/advanced_guides/engine.md

+
+### Optimizer Wrapper Constructor
+
+The default optimizer wrapper constructor [`DefaultOptimWrapperConstructor`](https://github.com/open-mmlab/mmengine/blob/main/mmengine/optim/optimizer/default_constructor.py#L19) builds the optimizer used in training based on the input `optim_wrapper` and the `paramwise_cfg` defined in the `optim_wrapper`. When the functionality of [`DefaultOptimWrapperConstructor`](https://github.com/open-mmlab/mmengine/blob/main/mmengine/optim/optimizer/default_constructor.py#L19) does not meet the requirements, you can customize the optimizer wrapper constructor to implement the configuration of hyperparameters.


Suggested change

The default optimizer wrapper constructor [`DefaultOptimWrapperConstructor`](https://github.com/open-mmlab/mmengine/blob/main/mmengine/optim/optimizer/default_constructor.py#L19) builds the optimizer used in training based on the input `optim_wrapper` and the `paramwise_cfg` defined in the `optim_wrapper`. When the functionality of [`DefaultOptimWrapperConstructor`](https://github.com/open-mmlab/mmengine/blob/main/mmengine/optim/optimizer/default_constructor.py#L19) does not meet the requirements, you can customize the optimizer wrapper constructor to implement the configuration of hyperparameters.

The default optimizer wrapper constructor [`DefaultOptimWrapperConstructor`](https://github.com/open-mmlab/mmengine/blob/main/mmengine/optim/optimizer/default_constructor.py#L19) builds the optimizer used in training based on the input `optim_wrapper` and `paramwise_cfg` defined in the `optim_wrapper`. When the functionality of [`DefaultOptimWrapperConstructor`](https://github.com/open-mmlab/mmengine/blob/main/mmengine/optim/optimizer/default_constructor.py#L19) does not meet the requirements, you can customize the optimizer wrapper constructor to implement the configuration of hyperparameters.

Co-authored-by: CSH <[email protected]>

…#2797)

xiexinch added 2 commits March 27, 2023 16:35

english version of engine doc

b64793e

fix

96b5978

xiexinch requested review from csatsurnh and MeowZheng March 27, 2023 08:45

mm-assistant bot assigned csatsurnh Mar 27, 2023

csatsurnh reviewed Mar 28, 2023

View reviewed changes

xiexinch and others added 2 commits March 28, 2023 19:36

fix

9bc7919

Update docs/en/advanced_guides/engine.md

5d58f58

Co-authored-by: CSH <[email protected]>

MeowZheng approved these changes Mar 28, 2023

View reviewed changes

csatsurnh approved these changes Mar 29, 2023

View reviewed changes

xiexinch merged commit 1dd6581 into open-mmlab:dev-1.x Mar 30, 2023

xiexinch deleted the engine_en_doc branch March 30, 2023 06:19

nahidnazifi87 pushed a commit to nahidnazifi87/mmsegmentation_playground that referenced this pull request Apr 5, 2024

[Doc] Add English version document of the Training Engine (open-mmlab…

1035a3a

…#2797)

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[Doc] Add English version document of the Training Engine #2797

[Doc] Add English version document of the Training Engine #2797

xiexinch commented Mar 27, 2023

csatsurnh Mar 28, 2023

csatsurnh Mar 28, 2023

csatsurnh Mar 28, 2023

csatsurnh Mar 28, 2023

csatsurnh Mar 28, 2023

csatsurnh Mar 28, 2023

csatsurnh Mar 28, 2023

csatsurnh Mar 28, 2023

	MMEngine defines some [basic loop controllers](https://github.com/open-mmlab/mmengine/blob/main/mmengine/runner/loops.py) such as epoch-based training loop (`EpochBasedTrainLoop`), iteration-based training loop (`IterBasedTrainLoop`), standard validation loop (`ValLoop`), and standard testing loop (`TestLoop`).
	MMEngine defined some [basic loop controllers](https://github.com/open-mmlab/mmengine/blob/main/mmengine/runner/loops.py) such as epoch-based training loop (`EpochBasedTrainLoop`), iteration-based training loop (`IterBasedTrainLoop`), standard validation loop (`ValLoop`), and standard testing loop (`TestLoop`).


		MMEngine defines some [basic loop controllers](https://github.com/open-mmlab/mmengine/blob/main/mmengine/runner/loops.py) such as epoch-based training loop (`EpochBasedTrainLoop`), iteration-based training loop (`IterBasedTrainLoop`), standard validation loop (`ValLoop`), and standard testing loop (`TestLoop`).

		OpenMMLab's algorithm libraries like MMSegmentation abstract model training, testing, and inference as executors (`Runner`) to handle. Users can use the default executor in MMEngine directly or modify the executor to meet customized needs. This document mainly introduces how users can configure existing running settings, hooks, and optimizers' basic concepts and usage methods.


		### Configuring Training Length

		Loop controllers refer to the execution process during training, validation, and testing. They use `train_cfg`, `val_cfg`, and `test_cfg` to build these processes in the configuration file. MMSegmentation sets commonly used training lengths in the configs/_base_/schedules folder's train_cfg.


		Custom hooks are defined in the configuration through `custom_hooks`, and `Runner` registers them using the [`register_custom_hooks`](https://github.com/open-mmlab/mmengine/blob/main/mmengine/runner/runner.py#L1820) method.

		The priority of custom hooks needs to be set in the configuration file; if not set, it will be set to `NORMAL` by default. The following are some custom hooks implemented in MMEngine:


		### SegVisualizationHook

		MMSegmentation implements [`SegVisualizationHook`](https://github.com/open-mmlab/mmsegmentation/blob/dev-1.x/mmseg/engine/hooks/visualization_hook.py#L17), which is used to visualize prediction results during validation and testing.

	MMSegmentation implements [`SegVisualizationHook`](https://github.com/open-mmlab/mmsegmentation/blob/dev-1.x/mmseg/engine/hooks/visualization_hook.py#L17), which is used to visualize prediction results during validation and testing.
	MMSegmentation implemented [`SegVisualizationHook`](https://github.com/open-mmlab/mmsegmentation/blob/dev-1.x/mmseg/engine/hooks/visualization_hook.py#L17), which is used to visualize prediction results during validation and testing.


		OpenMMLab's algorithm libraries like MMSegmentation abstract model training, testing, and inference as executors (`Runner`) to handle. Users can use the default executor in MMEngine directly or modify the executor to meet customized needs. This document mainly introduces how users can configure existing running settings, hooks, and optimizers' basic concepts and usage methods.

		## Configuring Running Settings

	## Configuring Running Settings
	## Configuring Runtime Settings


		## Optimizer

		In the previous configuration and runtime settings, we provided a simple example of configuring the training optimizer. This section will further detail how to configure optimizers in MMSegmentation.


		### Optimizer Wrapper Constructor

		The default optimizer wrapper constructor [`DefaultOptimWrapperConstructor`](https://github.com/open-mmlab/mmengine/blob/main/mmengine/optim/optimizer/default_constructor.py#L19) builds the optimizer used in training based on the input `optim_wrapper` and the `paramwise_cfg` defined in the `optim_wrapper`. When the functionality of [`DefaultOptimWrapperConstructor`](https://github.com/open-mmlab/mmengine/blob/main/mmengine/optim/optimizer/default_constructor.py#L19) does not meet the requirements, you can customize the optimizer wrapper constructor to implement the configuration of hyperparameters.

[Doc] Add English version document of the Training Engine #2797

[Doc] Add English version document of the Training Engine #2797

Conversation

xiexinch commented Mar 27, 2023

Motivation

Modification

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment