yentinglin
commited on
Commit
·
7c9f7d3
1
Parent(s):
c2d6eb8
Delete tmlu.py
Browse files
tmlu.py
DELETED
@@ -1,121 +0,0 @@
|
|
1 |
-
# coding=utf-8
|
2 |
-
# Copyright 2020 The HuggingFace Datasets Authors and the current dataset script contributor.
|
3 |
-
#
|
4 |
-
# Licensed under the Apache License, Version 2.0 (the "License");
|
5 |
-
# you may not use this file except in compliance with the License.
|
6 |
-
# You may obtain a copy of the License at
|
7 |
-
#
|
8 |
-
# http://www.apache.org/licenses/LICENSE-2.0
|
9 |
-
#
|
10 |
-
# Unless required by applicable law or agreed to in writing, software
|
11 |
-
# distributed under the License is distributed on an "AS IS" BASIS,
|
12 |
-
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
13 |
-
# See the License for the specific language governing permissions and
|
14 |
-
# limitations under the License.
|
15 |
-
|
16 |
-
|
17 |
-
import json
|
18 |
-
|
19 |
-
import datasets
|
20 |
-
|
21 |
-
|
22 |
-
_CITATION = """\
|
23 |
-
@misc{taiwanllama,
|
24 |
-
author={Lin, Yen-Ting and Chen, Yun-Nung},
|
25 |
-
title={Language Models for Taiwanese Culture},
|
26 |
-
year={2023},
|
27 |
-
url={https://github.com/MiuLab/Taiwan-LLaMa},
|
28 |
-
note={Code and models available at https://github.com/MiuLab/Taiwan-LLaMa},
|
29 |
-
}
|
30 |
-
"""
|
31 |
-
|
32 |
-
_DESCRIPTION = """\
|
33 |
-
This is a Traditional Mandarin multitask test consisting of multiple-choice questions from various branches of knowledge in Taiwan.
|
34 |
-
"""
|
35 |
-
|
36 |
-
_HOMEPAGE = "https://github.com/MiuLab/Taiwan-LLaMa"
|
37 |
-
|
38 |
-
_URL = "https://huggingface.co/datasets/miulab/TMLU/raw/main/{subject}.jsonl"
|
39 |
-
|
40 |
-
# TODO: add "all" subject
|
41 |
-
_SUBJECTS = [
|
42 |
-
"AST_biology",
|
43 |
-
"AST_chemistry",
|
44 |
-
"AST_chinese",
|
45 |
-
"AST_physics",
|
46 |
-
"GSAT_chinese",
|
47 |
-
]
|
48 |
-
|
49 |
-
|
50 |
-
class Tmlu(datasets.GeneratorBasedBuilder):
|
51 |
-
BUILDER_CONFIGS = [
|
52 |
-
datasets.BuilderConfig(
|
53 |
-
name=sub, version=datasets.Version("0.1.0"), description=f"TMLU Subject {sub}"
|
54 |
-
)
|
55 |
-
for sub in _SUBJECTS
|
56 |
-
]
|
57 |
-
|
58 |
-
def _info(self):
|
59 |
-
features = datasets.Features(
|
60 |
-
{
|
61 |
-
"id": datasets.Value("string"),
|
62 |
-
"question": datasets.Value("string"),
|
63 |
-
"correct_choices": datasets.features.Sequence(datasets.Value("string")),
|
64 |
-
"incorrect_choices": datasets.features.Sequence(datasets.Value("string")),
|
65 |
-
"metadata": {
|
66 |
-
"timestamp": datasets.Value("string"),
|
67 |
-
"source": datasets.Value("string"),
|
68 |
-
},
|
69 |
-
"human_evaluation": {
|
70 |
-
"quality": datasets.Value("string"),
|
71 |
-
"comments": datasets.Value("string"),
|
72 |
-
},
|
73 |
-
}
|
74 |
-
)
|
75 |
-
return datasets.DatasetInfo(
|
76 |
-
description=_DESCRIPTION,
|
77 |
-
features=features,
|
78 |
-
homepage=_HOMEPAGE,
|
79 |
-
citation=_CITATION,
|
80 |
-
)
|
81 |
-
|
82 |
-
def _split_generators(self, dl_manager):
|
83 |
-
"""Returns SplitGenerators."""
|
84 |
-
archive = dl_manager.download(
|
85 |
-
{
|
86 |
-
self.config.name: _URL.format(subject=self.config.name),
|
87 |
-
}
|
88 |
-
)
|
89 |
-
return [
|
90 |
-
datasets.SplitGenerator(
|
91 |
-
name=datasets.Split.TEST,
|
92 |
-
gen_kwargs={"path": archive[self.config.name], "split": "test"},
|
93 |
-
),
|
94 |
-
]
|
95 |
-
|
96 |
-
def _generate_examples(self, path, split):
|
97 |
-
with open(path, encoding="utf-8") as f:
|
98 |
-
for id_, row in enumerate(f):
|
99 |
-
data = json.loads(row)
|
100 |
-
yield id_, {
|
101 |
-
"id": data["id"],
|
102 |
-
"question": data["question"],
|
103 |
-
"correct_choices": data["correct_choices"],
|
104 |
-
"incorrect_choices": data["incorrect_choices"],
|
105 |
-
"metadata": data["metadata"],
|
106 |
-
"human_evaluation": data["human_evaluation"],
|
107 |
-
}
|
108 |
-
# def _generate_examples(self, iter_archive, split):
|
109 |
-
# """Yields examples as (key, example) tuples."""
|
110 |
-
# n_yielded_files = 0
|
111 |
-
# for id_file, (path, file) in enumerate(iter_archive):
|
112 |
-
# if f"data/{split}/" in path:
|
113 |
-
# if split == "auxiliary_train" or f"{self.config.name}_{split}.csv" in path or self.config.name == "all":
|
114 |
-
# subset = path.split("/")[-1].rsplit("_",1)[0] if split != "auxiliary_train" else ""
|
115 |
-
# n_yielded_files += 1
|
116 |
-
# lines = (line.decode("utf-8") for line in file)
|
117 |
-
# reader = csv.reader(lines)
|
118 |
-
# for id_line, data in enumerate(reader):
|
119 |
-
# yield f"{id_file}_{id_line}", {"question": data[0], "choices": data[1:5], "answer": data[5], "subject": subset}
|
120 |
-
# if (n_yielded_files == 8 or split != "auxiliary_train") and self.config.name != "all":
|
121 |
-
# break
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|