Datasets:
Create mtop.py
Browse files
mtop.py
ADDED
@@ -0,0 +1,97 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
"""
|
2 |
+
domain in {'alarm',
|
3 |
+
'calling',
|
4 |
+
'event',
|
5 |
+
'messaging',
|
6 |
+
'music',
|
7 |
+
'news',
|
8 |
+
'people',
|
9 |
+
'recipes',
|
10 |
+
'reminder',
|
11 |
+
'timer',
|
12 |
+
'weather'}
|
13 |
+
"""
|
14 |
+
|
15 |
+
_URL = "https://fb.me/mtop_dataset"
|
16 |
+
|
17 |
+
_CITATION = """@article{li2020mtop,
|
18 |
+
title={MTOP: A comprehensive multilingual task-oriented semantic parsing benchmark},
|
19 |
+
author={Li, Haoran and Arora, Abhinav and Chen, Shuohui and Gupta, Anchit and Gupta, Sonal and Mehdad, Yashar},
|
20 |
+
journal={arXiv preprint arXiv:2008.09335},
|
21 |
+
year={2020}
|
22 |
+
}"""
|
23 |
+
|
24 |
+
_DESCRIPTION = """ """
|
25 |
+
|
26 |
+
|
27 |
+
|
28 |
+
class MtopConfig(datasets.BuilderConfig):
|
29 |
+
"""BuilderConfig for Mtop."""
|
30 |
+
|
31 |
+
def __init__(self, **kwargs):
|
32 |
+
"""BuilderConfig for Mtop.
|
33 |
+
|
34 |
+
Args:
|
35 |
+
**kwargs: keyword arguments forwarded to super.
|
36 |
+
"""
|
37 |
+
super(MtopConfig, self).__init__(**kwargs)
|
38 |
+
|
39 |
+
|
40 |
+
class Mtop(datasets.GeneratorBasedBuilder):
|
41 |
+
|
42 |
+
BUILDER_CONFIGS = [
|
43 |
+
MtopConfig(
|
44 |
+
name="mtop",
|
45 |
+
version=datasets.Version("1.0.0", ""),
|
46 |
+
description="Plain text",
|
47 |
+
),
|
48 |
+
]
|
49 |
+
|
50 |
+
def _info(self):
|
51 |
+
return datasets.DatasetInfo(
|
52 |
+
description=_DESCRIPTION,
|
53 |
+
features=datasets.Features(
|
54 |
+
{
|
55 |
+
"idx": datasets.Value("string"),
|
56 |
+
"intent": datasets.Value("string"),
|
57 |
+
"spans": datasets.Value("string"),
|
58 |
+
"question": datasets.Value("string"),
|
59 |
+
"domain": datasets.Value("string"),
|
60 |
+
"lang": datasets.Value("string"),
|
61 |
+
"logical_form": datasets.Value("string"),
|
62 |
+
"tokenized_question": datasets.Value("string"),
|
63 |
+
}
|
64 |
+
),
|
65 |
+
# No default supervised_keys (as we have to pass both question
|
66 |
+
# and context as input).
|
67 |
+
supervised_keys=None,
|
68 |
+
homepage="",
|
69 |
+
citation=_CITATION,
|
70 |
+
|
71 |
+
)
|
72 |
+
|
73 |
+
def _split_generators(self, dl_manager):
|
74 |
+
filepath = dl_manager.download_and_extract(_URL)
|
75 |
+
|
76 |
+
return [
|
77 |
+
datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"filepath": filepath,"split":"train"}),
|
78 |
+
datasets.SplitGenerator(name=datasets.Split.VALIDATION, gen_kwargs={"filepath": filepath,"split":"eval"}),
|
79 |
+
datasets.SplitGenerator(name=datasets.Split.TEST, gen_kwargs={"filepath": filepath,"split":"test"}),
|
80 |
+
]
|
81 |
+
|
82 |
+
def _generate_examples(self, filepath, split):
|
83 |
+
"""This function returns the examples in the raw (text) form."""
|
84 |
+
key = 0
|
85 |
+
with open(f"{filepath}/mtop/en/{split}.txt", encoding="utf-8") as f:
|
86 |
+
for example in f:
|
87 |
+
example = example.split("\t")
|
88 |
+
dict_example = dict(idx=example[0],
|
89 |
+
intent=example[1],
|
90 |
+
spans=example[2],
|
91 |
+
question=example[3],
|
92 |
+
domain=example[4],
|
93 |
+
lang=example[5],
|
94 |
+
logical_form=example[6],
|
95 |
+
tokenized_question=example[7])
|
96 |
+
yield key, dict_example
|
97 |
+
key += 1
|