Upload movingthings.py

movingthings.py

@@ -17,6 +17,7 @@ _DESCRIPTION = (
     "Data was annotated by the script that generates the images with the prompt passed to Stable Diffusion."
 )
 _NAMES = ["roll", "flow", "zigzag", "walk", "linear"]
+
 class MovingThings(datasets.GeneratorBasedBuilder):
 
     VERSION = datasets.Version("1.0.0")  # Update the version as needed
@@ -35,26 +36,11 @@ class MovingThings(datasets.GeneratorBasedBuilder):
             supervised_keys=("image", "labels"),
             homepage=_HOMEPAGE,
             citation=_CITATION,
+            task_templates=[
+                ImageClassification(
+                    image_column="image",
+                    label_column="labels",
+                    label_classes=_NAMES
+                )
+            ],
         )
-
-    def _split_generators(self, dl_manager):
-        return {
-            "train": self._generate_examples(),
-        }
-
-    def _generate_examples(self):
-        # Implement your example generation logic here
-        # Return a list of dictionaries containing the features for each example
-        # You can simulate example data generation here
-        examples = [{"file_name": "soccer-ball-0.png", "text": "there is a man that is standing in the dark with a cell phone", "labels":"roll"},
-{"file_name": "sled-1.png", "text": "there is a red chair that is sitting on the sidewalk", "labels":"roll"},
-{"file_name": "steamroller-2.png", "text": "there is a large yellow pipe laying on the ground next to a yellow pole", "labels":"roll"},
-{"file_name": "hamster-wheel-3.png", "text": "there is a large circular sculpture made of colored plastic", "labels":"roll"},
-{"file_name": "skateboard-4.png", "text": "there is a skateboard that is laying on the ground", "labels":"roll"}]
-        return examples
-
-
-# Task template outside of _info()
-MovingThingsTaskTemplate = ImageClassification(image_column="image", label_column="labels")
-
-