fix: Tries to fix preview

imagenette.py

@@ -45,6 +45,7 @@ _LABEL_MAP = [
     'n03888257',
 ]
 
+_REPO = "https://huggingface.co/datasets/frgfm/imagenette/resolve/main/metadata"
 
 
 class OpenFireConfig(datasets.BuilderConfig):
@@ -68,16 +69,28 @@ class OpenFire(datasets.GeneratorBasedBuilder):
             name="full_size",
             description="All images are in their original size.",
             data_url="https://s3.amazonaws.com/fast-ai-imageclas/imagenette2.tgz",
+            metadata_urls={
+                "train": f"{_REPO}/full_size/train.txt",
+                "validation": f"{_REPO}/imagenette2/val.txt",
+            },
         ),
         OpenFireConfig(
             name="320px",
             description="All images were resized on their shortest side to 320 pixels.",
             data_url="https://s3.amazonaws.com/fast-ai-imageclas/imagenette2-320.tgz",
+            metadata_urls={
+                "train": f"{_REPO}/imagenette2-320/train.txt",
+                "validation": f"{_REPO}/imagenette2-320/val.txt",
+            },
         ),
         OpenFireConfig(
             name="160px",
             description="All images were resized on their shortest side to 160 pixels.",
             data_url="https://s3.amazonaws.com/fast-ai-imageclas/imagenette2-160.tgz",
+            metadata_urls={
+                "train": f"{_REPO}/imagenette2-160/train.txt",
+                "validation": f"{_REPO}/imagenette2-160/val.txt",
+            },
         ),
     ]
 
@@ -110,26 +123,35 @@ class OpenFire(datasets.GeneratorBasedBuilder):
         )
 
     def _split_generators(self, dl_manager):
-        _path = dl_manager.download_and_extract(self.config.data_url)
-        local_extracted_archive = os.path.join(_path, self.config.data_url.rpartition("/")[-1].split('.')[0])
+        archive_path = dl_manager.download(self.config.data_url)
+        metadata_paths = dl_manager.download(self.config.metadata_urls)
+        archive_iter = dl_manager.iter_archive(archive_path)
         return [
             datasets.SplitGenerator(
                 name=datasets.Split.TRAIN,
                 gen_kwargs={
-                    "split_folder": os.path.join(local_extracted_archive, "train"),
+                    "images": archive_iter,
+                    "metadata_path": metadata_paths["train"],
                 },
             ),
             datasets.SplitGenerator(
                 name=datasets.Split.VALIDATION,
                 gen_kwargs={
-                    "split_folder": os.path.join(local_extracted_archive, "val"),
+                    "images": archive_iter,
+                    "metadata_path": metadata_paths["validation"],
                 },
             ),
         ]
 
-    def _generate_examples(self, split_folder):
+    def _generate_examples(self, images, metadata_path):
+        with open(metadata_path, encoding="utf-8") as f:
+            files_to_keep = set(f.read().split("\n"))
         idx = 0
-        for class_idx, class_folder in enumerate(_LABEL_MAP):
-            for path in os.listdir(os.path.join(split_folder, class_folder)):
-                yield idx, {"image": os.path.join(split_folder, class_folder, path), "label": class_idx}
+        for file_path, file_obj in images:
+            if file_path in files_to_keep:
+                label = _LABEL_MAP.index(file_path.split("/")[-1])
+                yield idx, {
+                    "image": {"path": file_path, "bytes": file_obj.read()},
+                    "label": label,
+                }
                 idx += 1