Tutorial 1: Learn about Configs Modify a config through script arguments Config file naming convention An example of RotatedRetinaNet FAQ Use intermediate variables in configs Tutorial 2: Customize Datasets Support new data format Reorganize new data formats to existing format 1. Modify the config file for using the customized dataset 2. Check the annotations of the customized dataset Customize datasets by dataset wrappers Repeat dataset Class balanced dataset Concatenate dataset Tutorial 3: Customize Models Develop new components Add a new backbone 1. Define a new backbone (e.g. MobileNet) 2. Import the module 3. Use the backbone in your config file Add new necks 1. Define a neck (e.g. PAFPN) 2. Import the module 3. Modify the config file Add new heads Add new loss Tutorial 4: Customize Runtime Settings Customize optimization settings Customize optimizer supported by Pytorch Customize self-implemented optimizer 1. Define a new optimizer 2. Add the optimizer to registry 3. Specify the optimizer in the config file Customize optimizer constructor Additional settings Customize training schedules Customize workflow Customize hooks Customize self-implemented hooks 1. Implement a new hook 2. Register the new hook 3. Modify the config Use hooks implemented in MMCV 4. Example: NumClassCheckHook Modify default runtime hooks Checkpoint config Log config Evaluation config