Add random_dataset for testing

codes/data/__init__.py

@@ -49,6 +49,8 @@ def create_dataset(dataset_opt):
         from data.torch_dataset import TorchDataset as D
     elif mode == 'byol_dataset':
         from data.byol_attachment import ByolDatasetWrapper as D
+    elif mode == 'random_dataset':
+        from data.random_dataset import RandomDataset as D
     else:
         raise NotImplementedError('Dataset [{:s}] is not recognized.'.format(mode))
     dataset = D(dataset_opt)

codes/data/random_dataset.py

@@ -0,0 +1,17 @@
+import torch
+from torch.utils.data import Dataset
+
+
+# Dataset that feeds random data into the state. Can be useful for testing or demo purposes without actual data.
+class RandomDataset(Dataset):
+    def __init__(self, opt):
+        self.hq_shape = tuple(opt['hq_shape'])
+        self.lq_shape = tuple(opt['lq_shape'])
+
+    def __getitem__(self, item):
+        return {'lq': torch.rand(self.lq_shape), 'hq': torch.rand(self.hq_shape),
+                'LQ_path': '', 'GT_path': ''}
+
+    def __len__(self):
+        # Arbitrary
+        return 1024 * 1024