File size: 332 Bytes
5fa1a76
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
images = [
         train_data_augmentation(convert_to_tf_tensor(image.convert("RGB"))) for image in example_batch["image"]
     ]
     example_batch["pixel_values"] = [tf.transpose(tf.squeeze(image)) for image in images]
     return example_batch

 def preprocess_val(example_batch):
     """Apply val_transforms across a batch."""