MMClassification mainly uses python files as configs. The design of our configuration file system integrates modularity and inheritance, facilitating users to conduct various experiments. All configuration files are placed in the `configs` folder, which mainly contains the primitive configuration folder of `_base_` and many algorithm folders such as `resnet`, `swin_transformer`, `vision_transformer`, etc.
We follow the below convention to name config files. Contributors are advised to follow the same style. The config file names are divided into four parts: algorithm info, module information, training information and data information. Logically, different parts are concatenated by underscores `'_'`, and words in the same part are concatenated by dashes `'-'`.
Training type (mainly seen in the transformer network, such as the `ViT` algorithm, which is usually divided into two training type: pre-training and fine-tuning):
Training recipe. Usually, only the part that is different from the original paper will be marked. These methods will be arranged in the order `{pipeline aug}-{train aug}-{loss trick}-{scheduler}-{epochs}`.
You can easily build your own training config file by inherit some base config files. And the configs that are composed by components from `_base_` are called _primitive_.
For easy understanding, we use [ResNet50 primitive config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnet50_8xb32_in1k.py) as a example and comment the meaning of each line. For more detaile, please refer to the API documentation.
The parameter `"model"` is a python dictionary in the configuration file, which mainly includes information such as network structure and loss function:
-`type`： Classifier name, MMCls supports `ImageClassifier`, refer to [API documentation](https://mmclassification.readthedocs.io/en/latest/api/models.html#classifier).
-`backbone`： Backbone configs, refer to [API documentation](https://mmclassification.readthedocs.io/en/latest/api/models.html#backbones) for available options.
-`head`: Head network name, MMCls supports single-label and multi-label classification head networks, available options refer to [API documentation](https://mmclassification.readthedocs.io/en/latest/api/models.html#heads).
-`loss`: Loss function type, supports `CrossEntropyLoss`, [`LabelSmoothLoss`](https://github.com/open-mmlab/mmclassification/blob/master/configs/_base_/models/resnet50_label_smooth.py) etc., For available options, refer to [API documentation](https://mmclassification.readthedocs.io/en/latest/api/models.html#losses).
-`train_cfg`：Training augment config, MMCls supports [`mixup`](https://github.com/open-mmlab/mmclassification/blob/master/configs/_base_/models/resnet50_mixup.py), [`cutmix`](https://github.com/open-mmlab/mmclassification/blob/master/configs/_base_/models/resnet50_cutmix.py) and other augments.
-`pipeline` : Data processing pipeline, refer to related tutorial [CUSTOM DATA PIPELINES](https://mmclassification.readthedocs.io/en/latest/tutorials/data_pipeline.html)
The parameter `evaluation` is also a dictionary, which is the configuration information of `evaluation hook`, mainly including evaluation interval, evaluation index, etc..
```python
# dataset settings
dataset_type = 'ImageNet' # dataset name，
img_norm_cfg = dict( # Image normalization config to normalize the input images
mean=[123.675, 116.28, 103.53], # Mean values used to pre-training the pre-trained backbone models
std=[58.395, 57.12, 57.375], # Standard variance used to pre-training the pre-trained backbone models
to_rgb=True) # Whether to invert the color channel, rgb2bgr or bgr2rgb.
# train data pipeline
train_pipeline = [
dict(type='LoadImageFromFile'), # First pipeline to load images from file path
dict(type='Collect', keys=['img']) # do not pass gt_label while testing
]
data = dict(
samples_per_gpu=32, # Batch size of a single GPU
workers_per_gpu=2, # Worker to pre-fetch data for each single GPU
train=dict( # Train dataset config
train=dict( # train data config
type=dataset_type, # dataset name
data_prefix='data/imagenet/train', # Dataset root, when ann_file does not exist, the category information is automatically obtained from the root folder
pipeline=train_pipeline), # train data pipeline
val=dict( # val data config
type=dataset_type,
data_prefix='data/imagenet/val',
ann_file='data/imagenet/meta/val.txt', # ann_file existes, the category information is obtained from file
pipeline=test_pipeline),
test=dict( # test data config
type=dataset_type,
data_prefix='data/imagenet/val',
ann_file='data/imagenet/meta/val.txt',
pipeline=test_pipeline))
evaluation = dict( # The config to build the evaluation hook, refer to https://github.com/open-mmlab/mmdetection/blob/master/mmdet/core/evaluation/eval_hooks.py#L7 for more details.
interval=1, # Evaluation interval
metric='accuracy') # Metrics used during evaluation
-`optimizer`: optimizer setting , support all optimizers in `pytorch`, refer to related [mmcv](https://mmcv.readthedocs.io/en/latest/_modules/mmcv/runner/optimizer/default_constructor.html#DefaultOptimizerConstructor) documentation.
-`optimizer_config`: `optimizer hook` configuration file, such as setting gradient limit, refer to related [mmcv](https://github.com/open-mmlab/mmcv/blob/master/mmcv/runner/hooks/optimizer.py#L8) code.
-`lr_config`: Learning rate scheduler, supports "CosineAnnealing", "Step", "Cyclic", etc. refer to related [mmcv](https://mmcv.readthedocs.io/en/latest/_modules/mmcv/runner/hooks/lr_updater.html#LrUpdaterHook) documentation for more options.
-`runner`: For `runner`, please refer to `mmcv` for [`runner`](https://mmcv.readthedocs.io/en/latest/understand_mmcv/runner.html) introduction document.
```python
# he configuration file used to build the optimizer, support all optimizers in PyTorch.
optimizer = dict(type='SGD', # Optimizer type
lr=0.1, # Learning rate of optimizers, see detail usages of the parameters in the documentation of PyTorch
momentum=0.9, # Momentum
weight_decay=0.0001) # Weight decay of SGD
# Config used to build the optimizer hook, refer to https://github.com/open-mmlab/mmcv/blob/master/mmcv/runner/hooks/optimizer.py#L8 for implementation details.
optimizer_config = dict(grad_clip=None) # Most of the methods do not use gradient clip
# Learning rate scheduler config used to register LrUpdater hook
lr_config = dict(policy='step', # The policy of scheduler, also support CosineAnnealing, Cyclic, etc. Refer to details of supported LrUpdater from https://github.com/open-mmlab/mmcv/blob/master/mmcv/runner/hooks/lr_updater.py#L9.
step=[30, 60, 90]) # Steps to decay the learning rate
runner = dict(type='EpochBasedRunner', # Type of runner to use (i.e. IterBasedRunner or EpochBasedRunner)
max_epochs=100) # Runner that runs the workflow in total max_epochs. For IterBasedRunner use `max_iters`
```
### runtime setting
This part mainly includes saving the checkpoint strategy, log configuration, training parameters, breakpoint weight path, working directory, etc..
```python
# Config to set the checkpoint hook, Refer to https://github.com/open-mmlab/mmcv/blob/master/mmcv/runner/hooks/checkpoint.py for implementation.
checkpoint_config = dict(interval=1) # The save interval is 1
# config to register logger hook
log_config = dict(
interval=100, # Interval to print the log
hooks=[
dict(type='TextLoggerHook'), # The Tensorboard logger is also supported
# dict(type='TensorboardLoggerHook')
])
dist_params = dict(backend='nccl') # Parameters to setup distributed training, the port can also be set.
log_level = 'INFO' # The output level of the log.
resume_from = None # Resume checkpoints from a given path, the training will be resumed from the epoch when the checkpoint's is saved.
workflow = [('train', 1)] # Workflow for runner. [('train', 1)] means there is only one workflow and the workflow named 'train' is executed once.
work_dir = 'work_dir' # Directory to save the model checkpoints and logs for the current experiments.
```
## Inherit and Modify Config File
For easy understanding, we recommend contributors to inherit from existing methods.
For all configs under the same folder, it is recommended to have only **one**_primitive_ config. All other configs should inherit from the _primitive_ config. In this way, the maximum of inheritance level is 3.
For example, if your config file is based on ResNet with some other modification, you can first inherit the basic ResNet structure, dataset and other training setting by specifying `_base_ ='./resnet50_8xb32_in1k.py'` (The path relative to your config file), and then modify the necessary parameters in the config file. A more specific example, now we want to use almost all configs in `configs/resnet/resnet50_8xb32_in1k.py`, but change the number of training epochs from 100 to 300, modify when to decay the learning rate, and modify the dataset path, you can create a new config file `configs/resnet/resnet50_8xb32-300e_in1k.py` with content as below:
Some intermediate variables are used in the configuration file. The intermediate variables make the configuration file clearer and easier to modify.
For example, `train_pipeline` / `test_pipeline` is the intermediate variable of the data pipeline. We first need to define `train_pipeline` / `test_pipeline`, and then pass them to `data`. If you want to modify the size of the input image during training and testing, you need to modify the intermediate variables of `train_pipeline` / `test_pipeline`.
Sometimes, you need to set `_delete_=True` to ignore some domain content in the basic configuration file. You can refer to [mmcv](https://mmcv.readthedocs.io/en/latest/understand_mmcv/config.html#inherit-from-base-config-with-ignored-fields) for more instructions.
The following is an example. If you wangt to use cosine schedule in the above ResNet50 case, just using inheritance and directly modify it will report `get unexcepected keyword'step'` error, because the `'step'` field of the basic config in `lr_config` domain information is reserved, and you need to add `_delete_ =True` to ignore the content of `lr_config` related fields in the basic configuration file:
Sometimes, you may refer to some fields in the `_base_` config, so as to avoid duplication of definitions. You can refer to [mmcv](https://mmcv.readthedocs.io/en/latest/understand_mmcv/config.html#reference-variables-from-base) for some more instructions.
The following is an example of using auto augment in the training data preprocessing pipeline， refer to [`configs/_base_/datasets/imagenet_bs64_autoaug.py`](https://github.com/open-mmlab/mmclassification/blob/master/configs/_base_/datasets/imagenet_bs64_autoaug.py). When defining `train_pipeline`, just add the definition file name of auto augment to `_base_`, and then use `{{_base_.auto_increasing_policies}}` to reference the variables:
When users use the script "tools/train.py" or "tools/test.py" to submit tasks or use some other tools, they can directly modify the content of the configuration file used by specifying the `--cfg-options` parameter.
- Update config keys of dict chains.
The config options can be specified following the order of the dict keys in the original config.
For example, `--cfg-options model.backbone.norm_eval=False` changes the all BN modules in model backbones to `train` mode.
- Update keys inside a list of configs.
Some config dicts are composed as a list in your config. For example, the training pipeline `data.train.pipeline` is normally a list
e.g. `[dict(type='LoadImageFromFile'), dict(type='TopDownRandomFlip', flip_prob=0.5), ...]`. If you want to change `'flip_prob=0.5'` to `'flip_prob=0.0'` in the pipeline,
you may specify `--cfg-options data.train.pipeline.1.flip_prob=0.0`.
- Update values of list/tuples.
If the value to be updated is a list or a tuple. For example, the config file normally sets `workflow=[('train', 1)]`. If you want to
support list/tuple data types, and that **NO** white space is allowed inside the quotation marks in the specified value.
## Import user-defined modules
```{note}
This part may only be used when using MMClassification as a third party library to build your own project, and beginners can skip it.
```
After studying the follow-up tutorials [ADDING NEW DATASET](https://mmclassification.readthedocs.io/en/latest/tutorials/new_dataset.html), [CUSTOM DATA PIPELINES](https://mmclassification.readthedocs.io/en/latest/tutorials/data_pipeline.html), [ADDING NEW MODULES](https://mmclassification.readthedocs.io/en/latest/tutorials/new_modules.html). You may use MMClassification to complete your project and create new classes of datasets, models, data enhancements, etc. in the project. In order to streamline the code, you can use MMClassification as a third-party library, you just need to keep your own extra code and import your own custom module in the configuration files. For examples, you may refer to [OpenMMLab Algorithm Competition Project](https://github.com/zhangrui-wolf/openmmlab-competition-2021) .
Add the following code to your own configuration files:
