Update multiconer_v2.py

multiconer_v2.py

@@ -287,28 +287,34 @@ class MultiCoNER2(datasets.GeneratorBasedBuilder):
         logger.info("⏳ Generating examples from = %s", filepath)
         with open(filepath, encoding="utf-8") as f:
             guid = 0
+            s_id = None
             tokens = []
             ner_tags = []
             for line in f.read().strip().split('\n'):
-                if line.startswith("# id") or line == "" or line == "\n" or "_ _" not in line:
-                    if tokens:
+                if line.startswith("# id"):
+                    s_id = line.split('\t')[0].split(' ')[-1].strip()
+                    tokens = []
+                    ner_tags = []
+                elif len(line.strip()) == 0:
+                    if s_id and len(tokens) >=1 and len(tokens) == len(ner_tags):
                         yield guid, {
-                            "id": str(guid),
+                            "id": s_id,
                             "tokens": tokens,
                             "ner_tags": ner_tags,
                         }
                         guid += 1
+                        s_id = None
                         tokens = []
                         ner_tags = []
                 else:
                     # Separator is " _ _ "
-                    splits = line.split("_ _")
+                    splits = line.split(" _ _ ")
                     tokens.append(splits[0].strip())
                     ner_tags.append(splits[1].strip())
 
             # last example
             yield guid, {
-                "id": str(guid),
+                "id": s_id,
                 "tokens": tokens,
                 "ner_tags": ner_tags,
             }