Dataloader Definition

Adding Datasets to the EdgeVolution Framework

To add more datasets to the EdgeVolution framework, follow these steps:

1. Add Data Folder: Place the folder containing your dataset into the data/ directory.

2. Create DataLoader:

   • Add a new file named dataloader_<your_dataset_name>.py to the custom_dataloaders/ directory.
   • Inside this file, create a class that inherits from BaseDataLoader found in src/base_dataloader.py.
   • Implement your dataloader within this class. It should return ds_train, ds_val, and ds_test as TensorFlow datasets. Optionally, you can also return class_weights to be used during training.