Datasets:
Update megawika-test.py
Browse files- megawika-test.py +37 -37
megawika-test.py
CHANGED
@@ -138,26 +138,26 @@ class MegaWika(datasets.GeneratorBasedBuilder):
|
|
138 |
"source_text": datasets.Value("string"),
|
139 |
|
140 |
# Question/answer pairs
|
141 |
-
|
142 |
-
|
143 |
-
|
144 |
-
|
145 |
-
|
146 |
-
|
147 |
-
|
148 |
-
|
149 |
-
|
150 |
-
|
151 |
-
|
152 |
-
|
153 |
-
|
154 |
-
|
155 |
-
|
156 |
-
|
157 |
-
|
158 |
-
|
159 |
-
|
160 |
-
|
161 |
}
|
162 |
)
|
163 |
}
|
@@ -236,23 +236,23 @@ class MegaWika(datasets.GeneratorBasedBuilder):
|
|
236 |
"source_lang": entry.get("source_lang", ""),
|
237 |
"source_url": entry.get("source_url", ""),
|
238 |
"source_text": entry.get("source_text", ""),
|
239 |
-
|
240 |
-
|
241 |
-
|
242 |
-
|
243 |
-
|
244 |
-
|
245 |
-
|
246 |
-
|
247 |
-
|
248 |
-
|
249 |
-
|
250 |
-
|
251 |
-
|
252 |
-
|
253 |
-
|
254 |
-
|
255 |
-
|
256 |
}
|
257 |
for entry
|
258 |
in example.get("entries", [])
|
|
|
138 |
"source_text": datasets.Value("string"),
|
139 |
|
140 |
# Question/answer pairs
|
141 |
+
"qa_pairs": datasets.Sequence(
|
142 |
+
{
|
143 |
+
"question": datasets.Value("string"),
|
144 |
+
"en_answer": datasets.Value("string"),
|
145 |
+
"lang_answer": datasets.Value("string"),
|
146 |
+
"frames": datasets.Sequence(
|
147 |
+
{
|
148 |
+
"frame": datasets.Value("string"),
|
149 |
+
"argument": datasets.Value("string")
|
150 |
+
}
|
151 |
+
),
|
152 |
+
"en_matches_in_source": [[datasets.Value("int32")]], # list of pair of int indices
|
153 |
+
"en_match_in_passage": [datasets.Value("int32")], # pair of int indices
|
154 |
+
"lang_matches_in_source": [[datasets.Value("int32")]], # list of pair of int indices
|
155 |
+
"lang_match_in_passage": [datasets.Value("int32")], # pair of int indices
|
156 |
+
"passage": [datasets.Value("string")],
|
157 |
+
"en_answer_tokens": [datasets.Value("string")],
|
158 |
+
"match_disambiguated_question": datasets.Value("string"),
|
159 |
+
}
|
160 |
+
)
|
161 |
}
|
162 |
)
|
163 |
}
|
|
|
236 |
"source_lang": entry.get("source_lang", ""),
|
237 |
"source_url": entry.get("source_url", ""),
|
238 |
"source_text": entry.get("source_text", ""),
|
239 |
+
"qa_pairs": [
|
240 |
+
{
|
241 |
+
"question": qa_pair.get('question', ""),
|
242 |
+
"en_answer": qa_pair.get('en_answer', qa_pair.get('answer', "")),
|
243 |
+
'lang_answer': qa_pair.get('lang_answer', ''),
|
244 |
+
'frames': qa_pair.get('frames', []),
|
245 |
+
"en_matches_in_source": self._get_qa_pair_list_features(qa_pair, "en_matches_in_source"),
|
246 |
+
"en_match_in_passage": self._get_qa_pair_list_features(qa_pair, "en_match_in_passage"),
|
247 |
+
"lang_matches_in_source": self._get_qa_pair_list_features(qa_pair, "lang_matches_in_source"),
|
248 |
+
"lang_match_in_passage": self._get_qa_pair_list_features(qa_pair, "lang_match_in_passage"),
|
249 |
+
"passage": qa_pair.get('passage', []),
|
250 |
+
"en_answer_tokens": qa_pair.get('en_answer_tokens', qa_pair.get('answer_tokens', [])),
|
251 |
+
"match_disambiguated_question": qa_pair.get('match_disambiguated_question', ""),
|
252 |
+
}
|
253 |
+
for qa_pair
|
254 |
+
in entry.get('qa_pairs', [])
|
255 |
+
]
|
256 |
}
|
257 |
for entry
|
258 |
in example.get("entries", [])
|