[Docs] Fix grammar and spelling mistakes in docs (open-mmlab#16)

* fix docs

* fix template

.github/pull_request_template.md

@@ -23,7 +23,7 @@ If this PR introduces a new feature, it is better to list some use cases here an
 
 - [ ] Pre-commit or other linting tools are used to fix the potential lint issues.
 - [ ] Bug fixes are fully covered by unit tests, the case that causes the bug should be added in the unit tests.
-- [ ] The modification is covered by complete unit tests. If not, please add more unit test to ensure the correctness.
+- [ ] The modification is covered by complete unit tests. If not, please add more unit tests to ensure the correctness.
 - [ ] The documentation has been modified accordingly, like docstring or example tutorials.
 
 **After PR**:

README.md

@@ -30,7 +30,7 @@ It is a part of the [OpenMMLab](https://openmmlab.com/) project.
 Major features:
 - **Compatibility**
 
-  MMRazor can be easily applied to various projects in OpenMMLab, due to similar architecture design of OpenMMLab as well as the decoupling of slimming algorithms and vision tasks.
+  MMRazor can be easily applied to various projects in OpenMMLab, due to the similar architecture design of OpenMMLab as well as the decoupling of slimming algorithms and vision tasks.
 
 - **Flexibility**
 

docs/en/test.md

@@ -15,9 +15,9 @@ The usage of optional arguments are the same as corresponding tasks like mmclass
 ## Pruning
 
 ### Split Checkpoint(Optional)
-If you train a slimmable model during retrain, checkpoints of different subnets are
+If you train a slimmable model during retraining, checkpoints of different subnets are
 actually fused in only one checkpoint. You can split this checkpoint to
-multiple independent checkpoints by using following command
+multiple independent checkpoints by using the following command
 
 ```bash
 python tools/model_converters/split_checkpoint.py ${CONFIG_FILE} ${CHECKPOINT_PATH} --channel-cfgs ${CHANNEL_CFG_PATH} [optional arguments]
@@ -39,7 +39,7 @@ python tools/${task}/test_${task}.py ${CONFIG_FILE} ${CHECKPOINT_PATH} --cfg-opt
 
 ## Distillation
 
-To test distillation method, you can use following command
+To test distillation method, you can use the following command
 
 ```bash
 python tools/${task}/test_${task}.py ${CONFIG_FILE} ${CHECKPOINT_PATH} [optional arguments]

docs/en/train.md

@@ -35,7 +35,7 @@ python tools/${task}/train_${task}.py ${CONFIG_FILE} --cfg-options algorithm.mut
 
 ## Pruning
 
-Pruning has the four steps, including **supernet pre-training**, **search for subnet on the trained supernet**, **subnet retraining** and **split checkpoint**. The command of first two steps are similar to NAS, except here we need to use `CONFIG_FILE` of Pruning. Commands of two other steps are as follows.
+Pruning has four steps, including **supernet pre-training**, **search for subnet on the trained supernet**, **subnet retraining** and **split checkpoint**. The commands of the first two steps are similar to NAS, except that we need to use `CONFIG_FILE` of Pruning here. The commands of the two other steps are as follows.
 
 ### Subnet Retraining
 

docs/en/tutorials/Tutorial_1_overview.md

@@ -12,7 +12,7 @@ It is a part of the [OpenMMLab](https://openmmlab.com/) project.
 ## Major features:
 - **Compatibility**
 
-  MMRazor can be easily applied to various projects in OpenMMLab, due to similar architecture design of OpenMMLab as well as the decoupling of slimming algorithms and vision tasks.
+  MMRazor can be easily applied to various projects in OpenMMLab, due to the similar architecture design of OpenMMLab as well as the decoupling of slimming algorithms and vision tasks.
 
 - **Flexibility**
 
@@ -26,11 +26,11 @@ It is a part of the [OpenMMLab](https://openmmlab.com/) project.
 
 ![design_and_implement](../imgs/tutorials/overview/design_and_implement.png)
 
-In terms of overall design, MMRazor mainly includes Component and Algorithm.
+In terms of the overall design, MMRazor mainly includes Component and Algorithm.
 
-Component can be divided into basic component and algorithm component. Basic
+Component can be divided into basic component and algorithm component. The basic
 component consists of searcher, OP, Mutables and other modules in the figure,
-which provide basic function support for algorithm component. Algorithm component
+which provides basic function support for algorithm component. Algorithm component
 consists of Mutator, Pruner, Distiller and other modules in the figure. They
 provide core functionality for implementing various lightweight algorithms.
 The combination of Algorithm and Application can realize the purpose of slimming
@@ -63,7 +63,7 @@ MMRazor consists of 4 main parts: 1) apis, 2) core, 3) models, 4) datasets. mode
 
   - **Mutator**: Core functions provider of different types of NAS, mainly include some functions of changing the structure of architecture.
   - **Pruner**: Core functions provider of different types of pruning, mainly includes some functions of changing the structure of architecture and getting channel group.
-  - **Distiller**: Core functions provider of different types of KD, mainly includes functions of  registering some forward hooks, calculate the kd-loss and so on.
+  - **Distiller**: Core functions provider of different types of KD, mainly includes functions of registering some forward hooks, calculate the kd-loss and so on.
   - **Quantizer**: Core functions provider of different types of quantization. It will come soon.
 
 - **Base components**: Core components of architecture and some algorithm components

docs/en/tutorials/Tutorial_2_learn_about_configs.md

@@ -28,7 +28,7 @@ We follow the below convention to name config files. Contributors are advised to
 
 Same as MMDetection, we incorporate modular and inheritance design into our config system, which is convenient to conduct various experiments.
 
-To help the users have a basic idea of a complete config and the modules in a generation system, we make brief comments on the configs of some examples as the following. For more detailed usage and the corresponding alternative for each modules, please refer to the API documentation.
+To help the users have a basic idea of a complete config and the modules in a generation system, we make brief comments on the configs of some examples as the following. For more detailed usage and the corresponding alternative for each module, please refer to the API documentation.
 
 ### An example of NAS - spos
 

docs/en/tutorials/Tutorial_3_customize_architectures.md

@@ -8,7 +8,7 @@ Here we show how to add a new searchable backbone with an example of searchable_
 
 1. Define a new backbone
 
-    Create a new file `mmrazor/models/architectures/components/backbones/searchable_shufflenet_v2.py`, class `SearchableShuffleNetV2` inherits from `BaseBackBone` of mmcls, which is the codebase that you will to build model.
+    Create a new file `mmrazor/models/architectures/components/backbones/searchable_shufflenet_v2.py`, class `SearchableShuffleNetV2` inherits from `BaseBackBone` of mmcls, which is the codebase that you will to build the model.
 
     ```python
     import torch.nn as nn

docs/en/tutorials/Tutorial_7_customize_mixed_algorithms_with_out_algorithms_components.md

@@ -1,10 +1,10 @@
 # Tutorial 7: Customize mixed algorithms with our algorithm components
 
-Here we show how to customize  mixed algorithms with our algorithm components. We take the slimmable training in autoslim as an example.
+Here we show how to customize mixed algorithms with our algorithm components. We take the slimmable training in autoslim as an example.
 
-The sandwich rule and inplace distillation was introduced to enhance training process. The sandwich rule means that we train the model at smallest width, largest width and (n − 2) random widths, instead of n random widths. By inplace distillation, we use the predicted label of the model at the largest width as the training label for other widths, while for the largest width we use ground truth. So both the KD algorithm and the pruning algorithm are used in slimmable training.
+The sandwich rule and inplace distillation were introduced to enhance the training process. The sandwich rule means that we train the model at smallest width, largest width and (n − 2) random widths, instead of n random widths. By inplace distillation, we use the predicted label of the model at the largest width as the training label for other widths, while for the largest width we use ground truth. So both the KD algorithm and the pruning algorithm are used in slimmable training.
 
-1. In distillation part, we can directly use SelfDistiller in `mmrazor/models/distillers/self_distiller.py`. If distillers provided in MMRazor don't meet your needs, you can develop new algorithm components for your algorithm as step2 in Tutorial 6.
+1. In the distillation part, we can directly use SelfDistiller in `mmrazor/models/distillers/self_distiller.py`. If distillers provided in MMRazor don't meet your needs, you can develop new algorithm components for your algorithm as step2 in Tutorial 6.
 
 2. As the slimmable training is the first step of `Autoslim`, we do not need to register a new algorithm, but rewrite the `train_step`function in AutoSlim as follows: