fix bugs in script

qa_align.py

@@ -79,6 +79,16 @@ class QaAlign(datasets.GeneratorBasedBuilder):
     def _info(self):
         # We keep the exact same fields (columns) as the original dataset files (CSV) 
         # See https://github.com/DanielaBWeiss/QA-ALIGN/blob/main/data/official_qa_alignments/readme.md for format description  
+        qa_dict = {
+            "qa_uuid": datasets.Value("string"),
+            "verb": datasets.Value("string"),
+            "verb_idx": datasets.Value("int32"),
+            "question": datasets.Value("string"),
+            "answer": datasets.Value("string"),
+            "answer_range": datasets.Value("string"),
+        }
+        qa_full_dict = dict(qa_dict, qaid_short=datasets.Value("string"))
+        
         features = datasets.Features(
             {
                 "text_1": datasets.Value("string"),
@@ -87,9 +97,12 @@ class QaAlign(datasets.GeneratorBasedBuilder):
                 "prev_text_1": datasets.Value("string"),
                 "abs_sent_id_2": datasets.Value("string"),
                 "abs_sent_id_1": datasets.Value("string"),
-                "qas_1": dict,
-                "qas_2": dict,
-                "alignments": dict,
+                "qas_1": datasets.Sequence(qa_full_dict),
+                "qas_2": datasets.Sequence(qa_full_dict),
+                "alignments": datasets.Sequence({
+                    "sent1": datasets.Sequence(qa_dict),
+                    "sent2": datasets.Sequence(qa_dict),
+                }),
             }
         )
         return datasets.DatasetInfo(
@@ -145,12 +158,12 @@ class QaAlign(datasets.GeneratorBasedBuilder):
         """ Yields QA-Align examples from a csv file. Each instance contains all the alignment for a pair of sentences. """
 
         # merge annotations from sections 
-        df = pd.read_csv(filepath)
+        df = pd.read_csv(filepath, index_col=False)
         for counter, dic in enumerate(df.to_dict('records')):
             columns_to_load_into_object = ["qas_1", "qas_2", "alignments"]
             for key in columns_to_load_into_object:
-                dic[key] = json.loads(dic[key])
-            columns_to_remove = ["worker_id","assignment_id","work_time_in_seconds", "year","topic","key","source","source-data","tokens_1","tokens_2"]
+                dic[key] = eval(dic[key])
+            columns_to_remove = ["Unnamed: 0","worker_id","assignment_id","work_time_in_seconds","abs_scu_id", "year","topic","key","source","source-data","tokens_1","tokens_2"]
             for key in columns_to_remove:
                 del dic[key]
             yield counter, dic