1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157 | """All Swedish dataset configurations used in EuroEval."""
from ..data_models import DatasetConfig
from ..languages import SV
from ..tasks import COMMON_SENSE, EUROPEAN_VALUES, KNOW, LA, MCRC, NER, RC, SENT, SUMM
### Official datasets ###
SWEREC_CONFIG = DatasetConfig(
name="swerec",
pretty_name="the truncated version of the Swedish sentiment classification "
"dataset SweReC",
huggingface_id="EuroEval/swerec-mini",
task=SENT,
languages=[SV],
)
SCALA_SV_CONFIG = DatasetConfig(
name="scala-sv",
pretty_name="The Swedish part of the linguistic acceptability dataset ScaLA",
huggingface_id="EuroEval/scala-sv",
task=LA,
languages=[SV],
)
SUC3_CONFIG = DatasetConfig(
name="suc3",
pretty_name="the truncated version of the Swedish named entity recognition "
"dataset SUC 3.0",
huggingface_id="EuroEval/suc3-mini",
task=NER,
languages=[SV],
)
SCANDIQA_SV_CONFIG = DatasetConfig(
name="scandiqa-sv",
pretty_name="the Swedish part of the truncated version of the question answering "
"dataset ScandiQA",
huggingface_id="EuroEval/scandiqa-sv-mini",
task=RC,
languages=[SV],
)
SWEDN_CONFIG = DatasetConfig(
name="swedn",
pretty_name="the truncated version of the Swedish summarisation dataset SweDN",
huggingface_id="EuroEval/swedn-mini",
task=SUMM,
languages=[SV],
)
MMLU_SV_CONFIG = DatasetConfig(
name="mmlu-sv",
pretty_name="the truncated version of the Swedish knowledge dataset MMLU-sv, "
"translated from the English MMLU dataset",
huggingface_id="EuroEval/mmlu-sv-mini",
task=KNOW,
languages=[SV],
)
HELLASWAG_SV_CONFIG = DatasetConfig(
name="hellaswag-sv",
pretty_name="the truncated version of the Swedish common-sense reasoning dataset "
"HellaSwag-sv, translated from the English HellaSwag dataset",
huggingface_id="EuroEval/hellaswag-sv-mini",
task=COMMON_SENSE,
languages=[SV],
)
EUROPEAN_VALUES_SV_CONFIG = DatasetConfig(
name="european-values-sv",
pretty_name="the Swedish version of the European values evaluation dataset",
huggingface_id="EuroEval/european-values-sv",
task=EUROPEAN_VALUES,
languages=[SV],
splits=["test"],
bootstrap_samples=False,
_instruction_prompt="{text}",
)
### Unofficial datasets ###
SCHIBSTED_SV_CONFIG = DatasetConfig(
name="schibsted-sv",
pretty_name="the Swedish summarisation dataset Schibsted-sv",
huggingface_id="EuroEval/schibsted-article-summaries-sv",
task=SUMM,
languages=[SV],
unofficial=True,
)
ARC_SV_CONFIG = DatasetConfig(
name="arc-sv",
pretty_name="the truncated version of the Swedish knowledge dataset ARC-sv, "
"translated from the English ARC dataset",
huggingface_id="EuroEval/arc-sv-mini",
task=KNOW,
languages=[SV],
unofficial=True,
)
BELEBELE_SV_CONFIG = DatasetConfig(
name="belebele-sv",
pretty_name="the Swedish multiple choice reading comprehension dataset "
"BeleBele-sv, translated from the English BeleBele dataset",
huggingface_id="EuroEval/belebele-sv-mini",
task=MCRC,
languages=[SV],
unofficial=True,
)
MULTI_WIKI_QA_SV_CONFIG = DatasetConfig(
name="multi-wiki-qa-sv",
pretty_name="the truncated version of the Swedish part of the reading "
"comprehension dataset MultiWikiQA",
huggingface_id="EuroEval/multi-wiki-qa-sv-mini",
task=RC,
languages=[SV],
unofficial=True,
)
GOLDENSWAG_SV_CONFIG = DatasetConfig(
name="goldenswag-sv",
pretty_name="the truncated version of the Swedish common-sense reasoning "
"dataset GoldenSwag-sv, translated from the English GoldenSwag dataset",
huggingface_id="EuroEval/goldenswag-sv-mini",
task=COMMON_SENSE,
languages=[SV],
unofficial=True,
)
EUROPEAN_VALUES_SITUATIONAL_SV_CONFIG = DatasetConfig(
name="european-values-situational-sv",
pretty_name="the Swedish version of the European values evaluation dataset, where "
"the questions are phrased in a situational way",
huggingface_id="EuroEval/european-values-situational-sv",
task=EUROPEAN_VALUES,
languages=[SV],
splits=["test"],
bootstrap_samples=False,
_instruction_prompt="{text}",
unofficial=True,
)
EUROPEAN_VALUES_COMPLETIONS_SV_CONFIG = DatasetConfig(
name="european-values-completions-sv",
pretty_name="the Swedish version of the European values evaluation dataset, where "
"the questions are phrased as sentence completions",
huggingface_id="EuroEval/european-values-completions-sv",
task=EUROPEAN_VALUES,
languages=[SV],
splits=["test"],
bootstrap_samples=False,
_instruction_prompt="{text}",
unofficial=True,
)
|