Update images-demo.py

images-demo.py

@@ -1,90 +1,58 @@
-import json
-
 import datasets
-from datasets.tasks import QuestionAnsweringExtractive
-
-
-logger = datasets.logging.get_logger(__name__)
-
 
 _CITATION = """\
-@article{2016arXiv160605250R,
-       author = {{Rajpurkar}, Pranav and {Zhang}, Jian and {Lopyrev},
-                 Konstantin and {Liang}, Percy},
-        title = "{SQuAD: 100,000+ Questions for Machine Comprehension of Text}"
+@InProceedings{huggingface:dataset,
+title = {Small image-text set},
+author={James Briggs},
+year={2022}
 }
 """
 
 _DESCRIPTION = """\
-Demo...
+Demo dataset for testing or showing image-text capabilities.
 """
+_HOMEPAGE = "https://huggingface.co/datasets/jamescalam/image-text-demo"
 
-# _URL = "https://github.com/jamescalam/hf-datasets/raw/main/01_builder_script/dataset.tar.gz"
-_URL = "https://huggingface.co/datasets/akashsalmuthe/image-demo/resolve/main/images.tar.gz?download=true"
+_LICENSE = ""
 
-descriptions = ["aerial shot of futuristic city with large motorway",
-                "aerial shot of modern city at sunrise",
-                "butterfly landing on the nose of a cat",
-                "cute kitten walking through long grass",
-                "fluffy dog sticking out tongue with yellow background",
-                "futuristic city with led lit tower blocks",
-                "futuristic wet city street after rain with red and blue lights",
-                "ginger striped cat with long whiskers laid on wooden table",
-                "happy dog walking through park area holding ball	",
-                "happy ginger dog sticking out its tongue sat in front of dirt path",
-                "happy small fluffy white dog running across grass",
-                "kitten raising paw to sky with cyan background",
-                "modern city skyline at sunrise with pink to blue sky",
-                "modern neon lit city alleyway",
-                "new york city street view with yellow cabs",
-                "puppy with big ears sat with orange background	",
-                "suburban area with city skyline in distance",
-                "three young dogs on dirt road",
-                "top down shot of black and white cat with yellow background",
-                "two dogs playing in the snow",
-                "two dogs running on dirt path"
-                ]
+_REPO = "https://huggingface.co/datasets/jamescalam/image-text-demo"
 
+class ImageSet(datasets.GeneratorBasedBuilder):
+    """Small sample of image-text pairs"""
 
-
-class ImagesDemo(datasets.GeneratorBasedBuilder):
-    """Dataset creation"""
     def _info(self):
         return datasets.DatasetInfo(
             description=_DESCRIPTION,
             features=datasets.Features(
                 {
-                    "text": datasets.Value("string"),
-                    "image": datasets.Value()
+                    'text': datasets.Value("string"),
+                    'image': datasets.Image(),
                 }
             ),
-            # No default supervised_keys (as we have to pass both question
-            # and context as input).
             supervised_keys=None,
-            homepage="https://github.com/jamescalam/hf-datasets/",
+            homepage=_HOMEPAGE,
             citation=_CITATION,
         )
 
     def _split_generators(self, dl_manager):
-        path = dl_manager.download(_URL)
-        image_iters = dl_manager.iter_archive(path)
+        images_archive = dl_manager.download(f"{_REPO}/resolve/main/images.tgz")
+        image_iters = dl_manager.iter_archive(images_archive)
         return [
             datasets.SplitGenerator(
                 name=datasets.Split.TRAIN,
                 gen_kwargs={
                     "images": image_iters
-                    }
+                }
             ),
         ]
 
     def _generate_examples(self, images):
-        """This function returns the examples in the raw (text) form."""
-        idx = 0
-
-        for filepath, image in images:
-          yield idx, {
-              "image":{"path":filepath, "bytes":image.read()},
-              "text":descriptions[idx]
-          }
-
-          idx += 1
+        """ This function returns the examples in the raw (text) form."""
+        
+        for idx, (filepath, image) in enumerate(images):
+            description = filepath.split('/')[-1][:-4]
+            description = description.replace('_', ' ')
+            yield idx, {
+                "image": {"path": filepath, "bytes": image.read()},
+                "text": description,
+            }