Update MultiLegalNeg.py

MultiLegalNeg.py

@@ -26,7 +26,15 @@ _URL = {
     'data/'
 }
 _LANGUAGES = [
-    "german", "french", "italian", "swiss", "bioscope_abstracts", "full_papers"
+    "german", "french", "italian", "swiss"
+]
+
+_ENGLISH = [
+    "sherlock", "bioscope", "sfu"
+]
+
+_ENGLISH_SPLITS = [
+    "abstracts", "full_papers", "dev", "test_cardboard_GOLD", "test_circle_GOLD", "training", ""
 ]
 
 
@@ -88,14 +96,23 @@ class MultiLegalNeg(datasets.GeneratorBasedBuilder):
                 try:
                     filepaths.append(dl_manager.download((f'data/{language}.jsonl.xz')))
                 except:
-                    try:
-                        filepaths.append(dl_manager.download((f'data/english/{language}.jsonl.xz')))
-                    except:
-                        break
+                    break
 
             split_generators.append(
                 datasets.SplitGenerator(name=split, gen_kwargs={'filepaths': filepaths})
             )
+
+        for ds in _ENGLISH:
+            filepaths = []
+            for split in _ENGLISH_SPLITS:
+                try:
+                    filepaths.append(dl_manager.download((f'data/english/{ds}_{split}.jsonl.xz')))
+                except:
+                    break
+            
+            split_generators.append(
+                datasets.SplitGenerator(name=ds, gen_kwargs={'filepaths': filepaths})
+        )
             
         return split_generators