模型的迁移⚓︎

我们在 MMagic 1.x. 版本更新了模型设定，其中重要的改动如下所示：

删除 pretrained 字段.
在模型设定中添加 train_cfg 和 test_cfg 字段.
添加 data_preprocessor 字段. 这里主要是将归一化和颜色空间转换操作从 dataset transform 流程中移动到 data_preprocessor 中. 我们接下来会介绍data_preprocessor.

Original New

model = dict(
    type='BasicRestorer',  # Name of the model
    generator=dict(  # Config of the generator
        type='EDSR',  # Type of the generator
        in_channels=3,  # Channel number of inputs
        out_channels=3,  # Channel number of outputs
        mid_channels=64,  # Channel number of intermediate features
        num_blocks=16,  # Block number in the trunk network
        upscale_factor=scale, # Upsampling factor
        res_scale=1,  # Used to scale the residual in residual block
        rgb_mean=(0.4488, 0.4371, 0.4040),  # Image mean in RGB orders
        rgb_std=(1.0, 1.0, 1.0)),  # Image std in RGB orders
    pretrained=None,
    pixel_loss=dict(type='L1Loss', loss_weight=1.0, reduction='mean'))  # Config for pixel loss model training and testing settings

model = dict(
    type='BaseEditModel',  # Name of the model
    generator=dict(  # Config of the generator
        type='EDSRNet',  # Type of the generator
        in_channels=3,  # Channel number of inputs
        out_channels=3,  # Channel number of outputs
        mid_channels=64,  # Channel number of intermediate features
        num_blocks=16,  # Block number in the trunk network
        upscale_factor=scale, # Upsampling factor
        res_scale=1,  # Used to scale the residual in residual block
        rgb_mean=(0.4488, 0.4371, 0.4040),  # Image mean in RGB orders
        rgb_std=(1.0, 1.0, 1.0)),  # Image std in RGB orders
    pixel_loss=dict(type='L1Loss', loss_weight=1.0, reduction='mean')  # Config for pixel loss
    train_cfg=dict(),  # Config of training model.
    test_cfg=dict(),  # Config of testing model.
    data_preprocessor=dict(  # The Config to build data preprocessor
        type='DataPreprocessor', mean=[0., 0., 0.], std=[255., 255.,
                                                             255.]))

我们在 MMagic 1.x. 版本中对模型进行了重构，其中重要的改动如下所示：

MMagic 1.x 中的 models 被重构为六个部分：archs、base_models、data_preprocessors、editors、diffusion_schedulers 和 losses.
在 models 中添加了 data_preprocessor 模块。这里主要是将归一化和颜色空间转换操作从 dataset transform 流程中移动到 data_preprocessor 中.此时，数据流经过数据预处理后，会先经过 data_preprocessor 模块的转换，然后再输入到模型中.

模型的更多详细信息请参见模型指南.

最后更新: November 27, 2023
创建日期: November 27, 2023