Datasets:
Update configs
Browse files
chall.py
CHANGED
@@ -111,9 +111,9 @@ class Chall(GeneratorBasedBuilder):
|
|
111 |
num_to_words=True,
|
112 |
stratify_column="intervention",
|
113 |
folds={
|
114 |
-
"test": ["5"],
|
115 |
-
"eval": ["1", "
|
116 |
-
"train": ["2", "3", "4", "6", "7", "8", "9", "10", "11", "12", "13", "14", "15", "16", "
|
117 |
},
|
118 |
description="Settings used for the paper."
|
119 |
)
|
|
|
111 |
num_to_words=True,
|
112 |
stratify_column="intervention",
|
113 |
folds={
|
114 |
+
"test": ["5", "20", "21"], # add the new data (20 + 21 not in acl) to the test -> This results in about 60 hours of train data.
|
115 |
+
"eval": ["1", "15", "17"],
|
116 |
+
"train": ["2", "3", "4", "6", "7", "8", "9", "10", "11", "12", "13", "14", "15", "16", "18", "19"]
|
117 |
},
|
118 |
description="Settings used for the paper."
|
119 |
)
|