Update dataset_viewer_script.py

dataset_viewer_script.py

@@ -2,70 +2,51 @@
 # Lint as: python3
 """Indus Valley Script Character List :: compiled by Iravatham Mahadevan, published by Archaeological Survey of India :: targz dataset."""
 
-
-import json
-
 import datasets
-from datasets.tasks import QuestionAnsweringExtractive
-
 
-logger = datasets.logging.get_logger(__name__)
-
-
-# _CITATION = """
-# author = "Iravatham Mahadevan",
-# title = "The Indus Scripts: Texts, Concordance and Tables",
-# year = 1977
-# publisher = "Archaeological Survey of India"
-# }
-# """
+_CITATION = """\
+@InProceedings{huggingface:dataset,
+title = {IM-417 128x128 Indus Script Dataset},
+author={storesource},
+year={2024}
+}
+"""
 
-_DESCRIPTION = """\
-IM 417
+_DESCRIPTION = """
+IM-417 128x128 Indus Script Dataset.
+Labels derived from Pages 32-34 
+From "The Indus Scripts: Texts, Concordance and Tables" 
+Compiled by Iravatham Mahadevan, 
+Published by Archaeological Survey of India (1977)
 """
+_HOMEPAGE = "https://huggingface.co/datasets/tomeheya/IM-417-128"
 
-_URL = "https://huggingface.co/datasets/tomeheya/IM-417-128/blob/main/IM-417-128x128.tar.gz"
+_LICENSE = ""
 
+_REPO = "https://huggingface.co/datasets/tomeheya/IM-417-128"
 
-class ImageDatasetGenerator(datasets.GeneratorBasedBuilder):
+class ImageSet(datasets.GeneratorBasedBuilder):
+    """Small sample of image-text pairs"""
 
     def _info(self):
-        return datasets.DatasetInfo(
+        def extract_features(example):
+            text = example["__key__"]
+            image = example["png"]
+            return {
+                "text": text,
+                "image": image,
+            }
+
+        dataset = dataset.map(extract_features)
+        dataset_info = DatasetInfo(
             description=_DESCRIPTION,
-            features=datasets.Features(
+            features=Dataset.features(
                 {
-                    "text": datasets.Value("string"),
-                    "image": datasets.Image()
+                    "sign_label": Value("string"),
+                    "image": Image(),
                 }
             ),
-            # No default supervised_keys (as we have to pass both question
-            # and context as input).
             supervised_keys=None,
-            homepage="https://huggingface.co/tomeheya",
+            homepage=_HOMEPAGE,
             citation=_CITATION,
-            task_templates=[
-                QuestionAnsweringExtractive(
-                    question_column="question", context_column="context", answers_column="answers"
-                )
-            ],
         )
-
-    def _split_generators(self, dl_manager):
-        path = dl_manager.download_and_extract(_URL)
-        image_iterable = dl_manager.iter_archive(path) 
-
-        return [
-            datasets.SplitGenerator(
-            name=datasets.Split.TRAIN, 
-            gen_kwargs={"images": image_iterable}),
-        ]
-
-    def _generate_examples(self, images):
-        for file_path, image in images:
-            # Extracting label from file path
-            label = file_path.split("/")[-2]  # Assuming the label is the second last part of the path
-            yield idx, {
-                "image": {"path": file_path, "bytes": image.read()},
-                "label": label
-            }
-            idx += 1