PytorchCNN project construction 5--- mmcv_config

The overall code can be viewed on my github

---

MMCV brief description

mmcv is a basic library function of python, including:

• File IO (provides two general interfaces for loading and dumping different forms of files)
• Image (implemented in opencv, you need to ensure that OpenCV has been installed during use)
• Video (including video reading and conversion interfaces, some methods of video editing, and optical flow reading/writing/bending)
• Here I mainly use the Config module in Utils (this module supports loading configuration from a variety of file formats, and provides dict-like api to get and set values), so I will introduce this part in detail. If you have other requirements, you can download it from Refer to the official website documents in the literature for learning.

---

The Config class is used to process config and config files. It supports loading configuration from multiple file formats (including python, json and yaml). It provides dict-like api to get and set values.

---

Suppose we put the corresponding parameters in the config.py file, and then call this file to call the parameters

from mmcv import Config

cfg = Config.fromfile('./config/config.py')
cfg

Config (path: ./config/config.py): {'a': 1, 'b': {'b1': [0, 1, 2], 'b2': None}, 'c': (1, 2), 'd': 'string'}

cfg.a

1

cfg.b

{'b1': [0, 1, 2], 'b2': None}

cfg.d

'string'

Description:

The usage of this is relatively simple. It is automatically saved in the form of a dictionary. We can also call each parameter separately. The advantage of this is that all the parameters in the project can be written to this file for easy modification.

---

Finally, paste the parameter code you set

PARA = dict(
    train=dict(
        epochs = 100,
        batch_size = 64,
        lr = 0.001,
        momentum=0.9,
        wd = 5e-4,
        num_workers = 2,
        divice_ids = [1],
        gpu_id = 0,
        num_classes=10,
    ),
    test=dict(
        batch_size=100
    ),
    cifar10_paths = dict( #这些路径函数需要替换成自己的，最好写成相对路径，而不是绝对路径
        validation_rate = 0.05,

        root = '../../DATASET/cifar10/',

        original_trainset_path = '../../../DATASET/cifar10/cifar-10-python/',#train_batch_path
        original_testset_path = '../../../DATASET/cifar10/cifar-10-python/',

        after_trainset_path = '../../../DATASET/cifar10/trainset/',
        after_testset_path = '../../../DATASET/cifar10/testset/',
        after_validset_path = '../../../DATASET/cifar10/validset/',

        train_data_txt = '../../../DATASET/cifar10/train.txt',
        test_data_txt = '../../../DATASET/cifar10/test.txt',
        valid_data_txt = '../../../DATASET/cifar10/valid.txt',
    ),
    utils_paths = dict(
        checkpoint_path = './cache/checkpoint/',
        log_path = './cache/log/',
        visual_path = './cache/visual/',
        params_path = './cache/params/',
    ),
)

---

references

1. mmcv official documentation

2. Common functions and function descriptions of mmcv

Finally, I would like to thank my brother, who taught me to build the entire project, and the little friends who studied together in the laboratory~ I hope they will succeed in everything, and they will have a great future!