Datasets:
Update DiaBLa.py
Browse files
DiaBLa.py
CHANGED
|
@@ -67,100 +67,7 @@ class Diabla(datasets.GeneratorBasedBuilder):
|
|
| 67 |
'norm': datasets.Value('string'),
|
| 68 |
'mt': datasets.Value('string'),
|
| 69 |
'ref': datasets.Value('string'),
|
| 70 |
-
|
| 71 |
-
{
|
| 72 |
-
'eval-judgment': ClassLabel(num_classes=3, names=['poor', 'medium', 'perfect']),
|
| 73 |
-
'eval-verbatim': datasets.Value('string'),
|
| 74 |
-
'eval-problems': datasets.features.Sequence(
|
| 75 |
-
[
|
| 76 |
-
ClassLabel(num_classes=6, names=['coherence', 'grammar', 'meaning', 'word choice', 'style', 'other'])
|
| 77 |
-
]
|
| 78 |
-
),
|
| 79 |
-
'lang': ClassLabel(num_classes=2, names=['english', 'french']),
|
| 80 |
-
}
|
| 81 |
-
),
|
| 82 |
-
'dialogue_meta': datasets.features.Sequence(
|
| 83 |
-
{
|
| 84 |
-
'start_time': datasets.Value('string'),
|
| 85 |
-
'end_time' : datasets.Value('string'),
|
| 86 |
-
'translation_model': datasets.Value('string'),
|
| 87 |
-
'final_evaluation_user1': datasets.features.Sequence(
|
| 88 |
-
{
|
| 89 |
-
'style': ClassLabel(num_classes=4, names=['poor', 'average', 'good', 'excellent']),
|
| 90 |
-
'coherence': ClassLabel(num_classes=4, names=['poor', 'average', 'good', 'excellent']),
|
| 91 |
-
'grammaticality': ClassLabel(num_classes=4, names=['poor', 'average', 'good', 'excellent']),
|
| 92 |
-
'meaning': ClassLabel(num_classes=4, names=['poor', 'average', 'good', 'excellent']),
|
| 93 |
-
'word_choice': ClassLabel(num_classes=4, names=['poor', 'average', 'good', 'excellent'])
|
| 94 |
-
}
|
| 95 |
-
),
|
| 96 |
-
'final_evaluation_user2': datasets.features.Sequence(
|
| 97 |
-
{
|
| 98 |
-
'style': ClassLabel(num_classes=4, names=['poor', 'average', 'good', 'excellent']),
|
| 99 |
-
'coherence': ClassLabel(num_classes=4, names=['poor', 'average', 'good', 'excellent']),
|
| 100 |
-
'grammaticality': ClassLabel(num_classes=4, names=['poor', 'average', 'good', 'excellent']),
|
| 101 |
-
'meaning': ClassLabel(num_classes=4, names=['poor', 'average', 'good', 'excellent']),
|
| 102 |
-
'word_choice': ClassLabel(num_classes=4, names=['poor', 'average', 'good', 'excellent'])
|
| 103 |
-
}
|
| 104 |
-
),
|
| 105 |
-
'scenario': datasets.features.Sequence(
|
| 106 |
-
[
|
| 107 |
-
[
|
| 108 |
-
datasets.Value("string")
|
| 109 |
-
]
|
| 110 |
-
]
|
| 111 |
-
),
|
| 112 |
-
'user1': datasets.features.Sequence(
|
| 113 |
-
{
|
| 114 |
-
'rolenum': datasets.Value('int64'),
|
| 115 |
-
'role': datasets.features.Sequence(
|
| 116 |
-
[
|
| 117 |
-
datasets.Value('string')
|
| 118 |
-
]
|
| 119 |
-
),
|
| 120 |
-
'initiated_dialogue': datasets.Value('bool'),
|
| 121 |
-
'turn_number': datasets.Value('int64'),
|
| 122 |
-
'lang': ClassLabel(num_classes=2, names=['english', 'french']),
|
| 123 |
-
}
|
| 124 |
-
),
|
| 125 |
-
'user2': datasets.features.Sequence(
|
| 126 |
-
{
|
| 127 |
-
'rolenum': datasets.Value('int64'),
|
| 128 |
-
'role': datasets.features.Sequence(
|
| 129 |
-
[
|
| 130 |
-
datasets.Value('string')
|
| 131 |
-
]
|
| 132 |
-
),
|
| 133 |
-
'initiated_dialogue': datasets.Value('bool'),
|
| 134 |
-
'turn_number': datasets.Value('int64'),
|
| 135 |
-
'lang': ClassLabel(num_classes=2, names=['english', 'french']),
|
| 136 |
-
}
|
| 137 |
-
)
|
| 138 |
-
}
|
| 139 |
-
),
|
| 140 |
-
'dialogue_history': datasets.features.Sequence(
|
| 141 |
-
[
|
| 142 |
-
datasets.features.Sequence(
|
| 143 |
-
{
|
| 144 |
-
'id': datasets.Value('string'),
|
| 145 |
-
'orig': datasets.Value('string'),
|
| 146 |
-
'norm': datasets.Value('string'),
|
| 147 |
-
'mt': datasets.Value('string'),
|
| 148 |
-
'ref': datasets.Value('string'),
|
| 149 |
-
'utterance_meta': datasets.features.Sequence(
|
| 150 |
-
{
|
| 151 |
-
'eval-judgment': ClassLabel(num_classes=3, names=['poor', 'medium', 'perfect']),
|
| 152 |
-
'eval-verbatim': datasets.Value("string"),
|
| 153 |
-
'eval-problems': datasets.features.Sequence(
|
| 154 |
-
[
|
| 155 |
-
ClassLabel(num_classes=6, names=['coherence', 'grammar', 'meaning', 'word choice', 'style', 'other'])
|
| 156 |
-
]
|
| 157 |
-
),
|
| 158 |
-
'lang': ClassLabel(num_classes=2, names=['english', 'french']),
|
| 159 |
-
}
|
| 160 |
-
),
|
| 161 |
-
}),
|
| 162 |
-
]
|
| 163 |
-
)
|
| 164 |
}
|
| 165 |
),
|
| 166 |
# TODO?
|
|
@@ -225,11 +132,11 @@ class Diabla(datasets.GeneratorBasedBuilder):
|
|
| 225 |
'mt': mt_text,
|
| 226 |
'id': id_,
|
| 227 |
'ref': reference_text,
|
| 228 |
-
'utterance_meta': utterance_info
|
| 229 |
}
|
| 230 |
|
| 231 |
# add to history (without dialogue info and history)
|
| 232 |
dialogue_history.append(utterance_instance.copy())
|
| 233 |
-
utterance_instance['dialogue_meta'] = dialogue_info
|
| 234 |
-
utterance_instance['dialogue_history'] = dialogue_history
|
| 235 |
yield id_, utterance_instance
|
|
|
|
| 67 |
'norm': datasets.Value('string'),
|
| 68 |
'mt': datasets.Value('string'),
|
| 69 |
'ref': datasets.Value('string'),
|
| 70 |
+
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 71 |
}
|
| 72 |
),
|
| 73 |
# TODO?
|
|
|
|
| 132 |
'mt': mt_text,
|
| 133 |
'id': id_,
|
| 134 |
'ref': reference_text,
|
| 135 |
+
#'utterance_meta': utterance_info
|
| 136 |
}
|
| 137 |
|
| 138 |
# add to history (without dialogue info and history)
|
| 139 |
dialogue_history.append(utterance_instance.copy())
|
| 140 |
+
#utterance_instance['dialogue_meta'] = dialogue_info
|
| 141 |
+
#utterance_instance['dialogue_history'] = dialogue_history
|
| 142 |
yield id_, utterance_instance
|