add train/val split

sv-ident.py

@@ -44,8 +44,9 @@ _HOMEPAGE = "https://github.com/vadis-project/sv-ident"
 # _LICENSE = ""
 
 _URLS = {
-    "train": "https://raw.githubusercontent.com/vadis-project/sv-ident/9962c3274444ce84c59d42e2a6f8c0958ed15a26/data/train/data.tsv",
-    "trial": "https://github.com/vadis-project/sv-ident/tree/9962c3274444ce84c59d42e2a6f8c0958ed15a26/data/trial",
+    "train": "https://raw.githubusercontent.com/vadis-project/sv-ident/a8e71bba570f628c460e2b542d4cc645e4eb7d03/data/train/train.tsv",
+    "val": "https://raw.githubusercontent.com/vadis-project/sv-ident/a8e71bba570f628c460e2b542d4cc645e4eb7d03/data/train/val.tsv",
+#    "trial": "https://github.com/vadis-project/sv-ident/tree/9962c3274444ce84c59d42e2a6f8c0958ed15a26/data/trial",
 }
 
 
@@ -78,12 +79,19 @@ class SVIdent(datasets.GeneratorBasedBuilder):
 
     def _split_generators(self, dl_manager):
         """Returns SplitGenerators."""
-        filepath = dl_manager.download(_URLS["train"])
+        train_data_dir = dl_manager.download(_URLS["train"])
+        val_data_dir = dl_manager.download(_URLS["val"])
         return [
             datasets.SplitGenerator(
                 name=datasets.Split.TRAIN,
                 gen_kwargs={
-                    "filepath": filepath,
+                    "filepath": train_data_dir,
+                },
+            ),
+            datasets.SplitGenerator(
+                name=datasets.Split.VALIDATION,
+                gen_kwargs={
+                    "filepath": val_data_dir,
                 },
             )
         ]