📄 Add dataset_info.json (train-only, shard_length=200)

dataset_info.json

@@ -0,0 +1,19 @@
+{
+  "total_samples": 5033,
+  "dataset_name": "SCIN",
+  "created_at": "2025-06-20T14:06:16.083768",
+  "image_columns": [
+    "images"
+  ],
+  "audio_columns": [],
+  "subsets": [],
+  "splits": {
+    "default": {
+      "train": {
+        "num_samples": 5033,
+        "num_shards": 26,
+        "shard_length": 200
+      }
+    }
+  }
+}