Datasets:
Update csatqa.py
Browse files
csatqa.py
CHANGED
@@ -35,12 +35,15 @@ class CSATQA(datasets.GeneratorBasedBuilder):
|
|
35 |
features = datasets.Features(
|
36 |
{
|
37 |
"question": datasets.Value("string"),
|
|
|
38 |
"option#1": datasets.Value("string"),
|
39 |
"option#2": datasets.Value("string"),
|
40 |
"option#3": datasets.Value("string"),
|
41 |
"option#4": datasets.Value("string"),
|
42 |
"option#5": datasets.Value("string"),
|
43 |
"gold": datasets.Value("int8"),
|
|
|
|
|
44 |
}
|
45 |
)
|
46 |
return datasets.DatasetInfo(
|
@@ -67,13 +70,18 @@ class CSATQA(datasets.GeneratorBasedBuilder):
|
|
67 |
buffer = []
|
68 |
for key, row in enumerate(f):
|
69 |
data = json.loads(row)
|
70 |
-
buffer.append({
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
71 |
for idx, dat in enumerate(buffer):
|
72 |
yield idx,dat
|
73 |
|
74 |
-
# def _generate_examples(self, filepath):
|
75 |
-
# with open(filepath, encoding="utf-8") as f:
|
76 |
-
# for key, row in enumerate(f):
|
77 |
-
# data = json.loads(row)
|
78 |
-
# if data["Category"] == self.config.name:
|
79 |
-
# yield key, data
|
|
|
35 |
features = datasets.Features(
|
36 |
{
|
37 |
"question": datasets.Value("string"),
|
38 |
+
"context" : datasets.Value("string"),
|
39 |
"option#1": datasets.Value("string"),
|
40 |
"option#2": datasets.Value("string"),
|
41 |
"option#3": datasets.Value("string"),
|
42 |
"option#4": datasets.Value("string"),
|
43 |
"option#5": datasets.Value("string"),
|
44 |
"gold": datasets.Value("int8"),
|
45 |
+
"category": datasets.Value("string"),
|
46 |
+
"human_peformance": datasets.Value("float16"),
|
47 |
}
|
48 |
)
|
49 |
return datasets.DatasetInfo(
|
|
|
70 |
buffer = []
|
71 |
for key, row in enumerate(f):
|
72 |
data = json.loads(row)
|
73 |
+
buffer.append({
|
74 |
+
"question": data["question"],
|
75 |
+
"context" : data["context"],
|
76 |
+
"option#1": data["option#1"],
|
77 |
+
"option#2": data["option#2"],
|
78 |
+
"option#3": data["option#3"],
|
79 |
+
"option#4": data["option#4"],
|
80 |
+
"option#5": data["option#5"],
|
81 |
+
"gold": data["gold"],
|
82 |
+
"category": data["category"],
|
83 |
+
"human_peformance": data["human_performance"]})
|
84 |
+
|
85 |
for idx, dat in enumerate(buffer):
|
86 |
yield idx,dat
|
87 |
|
|
|
|
|
|
|
|
|
|
|
|