Upload dataset_script.py

dataset_script.py

@@ -0,0 +1,64 @@
+import os
+import json
+import datasets
+from datasets import Dataset, DatasetInfo, SplitGenerator, Value, Features
+from pathlib import Path
+
+class MyGenerationDataset(datasets.GeneratorBasedBuilder):
+    def _info(self):
+        return DatasetInfo(
+            description="IDEA-Bench.",
+            features=Features({
+                "task name": Value('string'), 
+                "task ID": Value('string'),
+                "case ID": Value('string'),
+                "images": datasets.Sequence(datasets.Image()),
+                "instruction": Value('string'),
+            }),
+            license="cc-by-nc-4.0",
+        )
+
+    def _split_generators(self, dl_manager):
+        data_dir = './'
+        return [
+            SplitGenerator(name=datasets.Split.TEST, gen_kwargs={'data_dir': data_dir}),
+        ]
+
+    def _generate_examples(self, data_dir):
+        for folder_name in os.listdir(data_dir):
+            folder_path = os.path.join(data_dir, folder_name)
+            if os.path.isdir(folder_path):
+                meta_path = os.path.join(folder_path, 'meta.json')
+                if not os.path.exists(meta_path):
+                    continue 
+                
+                with open(meta_path, 'r', encoding='utf-8') as f:
+                    meta_data = json.load(f)
+
+                task_name = meta_data.get("task_name", "")
+                uid = meta_data.get("uid", "")
+                case_id = meta_data.get("case_id", "")
+
+                instruction_path = os.path.join(folder_path, 'instruction.txt')
+                with open(instruction_path, 'r', encoding='utf-8') as f:
+                    instruction = f.read().strip()
+
+                image_files = [f for f in os.listdir(folder_path) if f.endswith('.jpg')]
+                
+                if not image_files:
+                    yield folder_name, {
+                        'task name': task_name, 
+                        'task ID': uid, 
+                        'case ID': case_id, 
+                        'image': [],
+                        'instruction': instruction,
+                    }
+                else:
+                    image_paths = [os.path.join(folder_path, img_file) for img_file in image_files]
+                    yield folder_name, {
+                        'task name': task_name,
+                        'task ID': uid,
+                        'case ID': case_id, 
+                        'images': image_paths, 
+                        'instruction': instruction, 
+                    }