Create dataset.py

dataset.py

@@ -0,0 +1,50 @@
+from datasets import GeneratorBasedBuilder, DatasetInfo, SplitGenerator, Split, BuilderConfig
+import datasets
+import csv
+import pyarrow as pa
+
+class CustomConfig(BuilderConfig):
+    def __init__(self, **kwargs):
+        super().__init__(**kwargs)
+
+class MyDataset(GeneratorBasedBuilder):
+    BUILDER_CONFIGS = [
+        CustomConfig(name="Contrastive Learning", version=datasets.Version("1.0.0"),
+                     description="Loads Arrow files for contrastive learning"),
+        CustomConfig(name="Thresholding", version=datasets.Version("1.0.0"),
+                     description="Loads CSV files for thresholding"),
+    ]
+
+    def _info(self):
+        # No features defined — schema will be inferred
+        return DatasetInfo()
+
+    def _split_generators(self, dl_manager):
+        data_files = self.config.data_files
+
+        def get_path(split_name, fallback=None):
+            for entry in data_files:
+                if entry["split"] == split_name:
+                    return entry["path"]
+            return fallback
+
+        return [
+            SplitGenerator(name=Split.TRAIN, gen_kwargs={"filepath": get_path("train")}),
+            SplitGenerator(name=Split.VALIDATION, gen_kwargs={"filepath": get_path("dev") or get_path("val")}),
+            SplitGenerator(name=Split.TEST, gen_kwargs={"filepath": get_path("test")}),
+        ]
+
+    def _generate_examples(self, filepath):
+        if filepath.endswith(".arrow"):
+            table = pa.ipc.RecordBatchFileReader(filepath).read_all()
+            records = table.to_pydict()
+            keys = list(records.keys())
+            for i in range(len(records[keys[0]])):
+                yield i, {k: records[k][i] for k in keys}
+        elif filepath.endswith(".csv"):
+            with open(filepath, encoding="utf-8") as f:
+                reader = csv.DictReader(f)
+                for i, row in enumerate(reader):
+                    yield i, row
+        else:
+            raise ValueError(f"Unsupported file format for file: {filepath}")