def _info(self):
features = datasets.Features(
{
"source": {
"filename": datasets.Value("string"),
"headline": datasets.Value("string"),
"body": datasets.Value("string"),
"total_number_of_words": datasets.Value("int64"),
"total_number_of_sentences": datasets.Value("int64"),
"number_of_words_with_swr": datasets.Value("int64"),
"newspaper": datasets.Value("string"),
"newsdate": datasets.Value("string"),
"domain": datasets.ClassLabel(
names=[
"business",
"sports",
"national",
"foreign",
"showbiz",
]
),
"classification": datasets.ClassLabel(
names=["wholly_derived", "partially_derived", "not_derived"]
),
},
"derived": {
"filename": datasets.Value("string"),
"headline": datasets.Value("string"),
"body": datasets.Value("string"),
"total_number_of_words": datasets.Value("int64"),
"total_number_of_sentences": datasets.Value("int64"),
"number_of_words_with_swr": datasets.Value("int64"),
"newspaper": datasets.Value("string"),
"newsdate": datasets.Value("string"),
"domain": datasets.ClassLabel(
names=[
"business",
"sports",
"national",
"foreign",
"showbiz",
]
),
"classification": datasets.ClassLabel(
names=["wholly_derived", "partially_derived", "not_derived"]
),
},
}
)
return datasets.DatasetInfo(
description=_DESCRIPTION,
features=features,
supervised_keys=None,
homepage=_HOMEPAGE,
license=_LICENSE,
citation=_CITATION,
)
def _info(self):
if self.config.name == "NER":
features = datasets.Features(
{
"input_ids": datasets.Sequence(feature=datasets.Value("int32")),
"labels": datasets.Sequence(
feature=datasets.ClassLabel(num_classes=len(_NER_LABEL_NAMES), names=_NER_LABEL_NAMES)
),
"tag_mask": datasets.Sequence(feature=datasets.Value("int8")),
}
)
elif self.config.name == "ROLES":
features = datasets.Features(
{
"input_ids": datasets.Sequence(feature=datasets.Value("int32")),
"labels": datasets.Sequence(
feature=datasets.ClassLabel(
num_classes=len(_SEMANTIC_ROLES_LABEL_NAMES), names=_SEMANTIC_ROLES_LABEL_NAMES
)
),
"tag_mask": datasets.Sequence(feature=datasets.Value("int8")),
}
)
elif self.config.name == "BORING":
features = datasets.Features(
{
"input_ids": datasets.Sequence(feature=datasets.Value("int32")),
"labels": datasets.Sequence(
feature=datasets.ClassLabel(num_classes=len(_BORING_LABEL_NAMES), names=_BORING_LABEL_NAMES)
),
}
)
elif self.config.name == "PANELIZATION":
features = datasets.Features(
{
"input_ids": datasets.Sequence(feature=datasets.Value("int32")),
"labels": datasets.Sequence(
feature=datasets.ClassLabel(num_classes=len(_PANEL_START_NAMES), names=_PANEL_START_NAMES)
),
}
)
return datasets.DatasetInfo(
description=_DESCRIPTION,
features=features,
supervised_keys=("input_ids", "labels"),
homepage=_HOMEPAGE,
license=_LICENSE,
citation=_CITATION,
)
def _info(self):
return datasets.DatasetInfo(
description=_DESCRIPTION,
features=datasets.Features(
{
"dialog": datasets.features.Sequence(datasets.Value("string")),
"act": datasets.features.Sequence(datasets.ClassLabel(names=list(act_label.values()))),
"emotion": datasets.features.Sequence(datasets.ClassLabel(names=list(emotion_label.values()))),
}
),
supervised_keys=None,
homepage="http://yanran.li/dailydialog",
citation=_CITATION,
)
