Update from sam

Arguement_Mining_CL2017.py

@@ -116,7 +116,7 @@ class ArguementMiningCL2017(datasets.GeneratorBasedBuilder):
                             "tokens": tokens,
                             "chunk_tags": chunk_tags,
                         }
-                        guid += 1
+                        guid = ' '
                         tokens = []
                         chunk_tags = []
                 else:
@@ -124,8 +124,9 @@ class ArguementMiningCL2017(datasets.GeneratorBasedBuilder):
                     line=line.strip('\n')
                     splits = line.split("\t")
                     #print(splits)
-                    tokens.append(splits[0])
-                    chunk_tags.append(splits[1])
+		    guid = splits[0]
+                    tokens.append(splits[1])
+                    chunk_tags.append(splits[2])
                     #print({"id": str(guid),"tokens": tokens,"chunk_tags": chunk_tags,})
             # last example
             yield guid, {