Update markushgrapher-datasets.py

markushgrapher-datasets.py

@@ -1,5 +1,16 @@
+from datasets import (
+    DatasetInfo,
+    GeneratorBasedBuilder,
+    SplitGenerator,
+    Split,
+    BuilderConfig,
+    Features,
+    Value,
+    Sequence,
+    Image,
+    Dataset
+)
 import os
-from datasets import DatasetInfo, GeneratorBasedBuilder, SplitGenerator, Split, BuilderConfig, Features, Value, Array2D, Array3D, Sequence, Image, DatasetDict
 import datasets
 
 class MarkushGrapherConfig(BuilderConfig):
@@ -16,23 +27,50 @@ class MarkushGrapherDataset(GeneratorBasedBuilder):
     ]
 
     def _info(self):
-        return DatasetInfo(
-            description="MarkushGrapher-Datasets",
-            features=Features({
-                "id": Value("int64"),
-                "image_name": Value("string"),
-                "page_image": Image(mode=None, decode=True, id=None),
-                "description": Value("string"),
-                "annotation": Value("string"),
-                "mol": Value("string"),
-                "cxsmiles_dataset": Value("string"),
-                "cxsmiles": Value("string"),
-                "cxsmiles_opt": Value("string"),
-                "keypoints": Value(dtype='null', id=None),
-                "cells": [{'bbox': Sequence(feature=Value(dtype='float64', id=None), length=-1, id=None), 'text': Value(dtype='string', id=None)}],
-            }),
-            supervised_keys=None,
-        )
+        if self.config.name in ["m2s", "uspto-markush"]:
+            return DatasetInfo(
+                features=Features({
+                    "id": Value("int64"),
+                    "image_name": Value("string"),
+                    "page_image": Image(decode=True),
+                    "description": Value("string"),
+                    "annotation": Value("string"),
+                    "mol": Value("string"),
+                    "cxsmiles_dataset": Value("string"),
+                    "cxsmiles": Value("string"),
+                    "cxsmiles_opt": Value("string"),
+                    "keypoints": Value("null"),
+                    "cells": Sequence({
+                        "bbox": Sequence(Value("float64")),
+                        "text": Value("string"),
+                    }),
+                }),
+                supervised_keys=None,
+            )
+
+        elif self.config.name in ["markushgrapher-synthetic", "markushgrapher-synthetic-training"]:
+            return DatasetInfo(
+                features=Features({
+                    "id": Value("int64"),
+                    "page_image_path": Value("string"),
+                    "description": Value("string"),
+                    "annotation": Value("string"),
+                    "mol": Value("string"),
+                    "cxsmiles_dataset": Value("string"),
+                    "cxsmiles": Value("string"),
+                    "cxsmiles_opt": Value("string"),
+                    "keypoints": Sequence(Sequence(Value("float64"))),
+                    "cells": Sequence({
+                        "bbox": Sequence(Value("float64")),
+                        "text": Value("string"),
+                    }),
+                    "page_image": Image(decode=True),
+                }),
+                supervised_keys=None,
+            )
+
+        else:
+            raise ValueError(f"Unsupported config: {self.config.name}")
 
     def _split_generators(self, dl_manager):
         data_dir = dl_manager.download_and_extract(self.config.name)
@@ -49,9 +87,8 @@ class MarkushGrapherDataset(GeneratorBasedBuilder):
         return splits
 
     def _generate_examples(self, data_dir):
-        # Automatically loads .arrow files
         arrow_files = [os.path.join(data_dir, f) for f in os.listdir(data_dir) if f.endswith(".arrow")]
         for file in arrow_files:
-            dataset = datasets.Dataset.from_file(file)
+            dataset = Dataset.from_file(file)
             for i, row in enumerate(dataset):
                 yield f"{file}_{i}", row