Commit
·
ad6d30b
0
Parent(s):
Update files from the datasets library (from 1.0.0)
Browse filesRelease notes: https://github.com/huggingface/datasets/releases/tag/1.0.0
- .gitattributes +27 -0
- dataset_infos.json +1 -0
- dummy/dev_test/2.0.0/dummy_data.zip +3 -0
- dummy/train/2.0.0/dummy_data.zip +3 -0
- ubuntu_dialogs_corpus.py +132 -0
.gitattributes
ADDED
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
*.7z filter=lfs diff=lfs merge=lfs -text
|
2 |
+
*.arrow filter=lfs diff=lfs merge=lfs -text
|
3 |
+
*.bin filter=lfs diff=lfs merge=lfs -text
|
4 |
+
*.bin.* filter=lfs diff=lfs merge=lfs -text
|
5 |
+
*.bz2 filter=lfs diff=lfs merge=lfs -text
|
6 |
+
*.ftz filter=lfs diff=lfs merge=lfs -text
|
7 |
+
*.gz filter=lfs diff=lfs merge=lfs -text
|
8 |
+
*.h5 filter=lfs diff=lfs merge=lfs -text
|
9 |
+
*.joblib filter=lfs diff=lfs merge=lfs -text
|
10 |
+
*.lfs.* filter=lfs diff=lfs merge=lfs -text
|
11 |
+
*.model filter=lfs diff=lfs merge=lfs -text
|
12 |
+
*.msgpack filter=lfs diff=lfs merge=lfs -text
|
13 |
+
*.onnx filter=lfs diff=lfs merge=lfs -text
|
14 |
+
*.ot filter=lfs diff=lfs merge=lfs -text
|
15 |
+
*.parquet filter=lfs diff=lfs merge=lfs -text
|
16 |
+
*.pb filter=lfs diff=lfs merge=lfs -text
|
17 |
+
*.pt filter=lfs diff=lfs merge=lfs -text
|
18 |
+
*.pth filter=lfs diff=lfs merge=lfs -text
|
19 |
+
*.rar filter=lfs diff=lfs merge=lfs -text
|
20 |
+
saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
21 |
+
*.tar.* filter=lfs diff=lfs merge=lfs -text
|
22 |
+
*.tflite filter=lfs diff=lfs merge=lfs -text
|
23 |
+
*.tgz filter=lfs diff=lfs merge=lfs -text
|
24 |
+
*.xz filter=lfs diff=lfs merge=lfs -text
|
25 |
+
*.zip filter=lfs diff=lfs merge=lfs -text
|
26 |
+
*.zstandard filter=lfs diff=lfs merge=lfs -text
|
27 |
+
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
dataset_infos.json
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
{"train": {"description": "Ubuntu Dialogue Corpus, a dataset containing almost 1 million multi-turn dialogues, with a total of over 7 million utterances and 100 million words. This provides a unique resource for research into building dialogue managers based on neural language models that can make use of large amounts of unlabeled data. The dataset has both the multi-turn property of conversations in the Dialog State Tracking Challenge datasets, and the unstructured nature of interactions from microblog services such as Twitter.\n", "citation": "@article{DBLP:journals/corr/LowePSP15,\n author = {Ryan Lowe and\n Nissan Pow and\n Iulian Serban and\n Joelle Pineau},\n title = {The Ubuntu Dialogue Corpus: {A} Large Dataset for Research in Unstructured\n Multi-Turn Dialogue Systems},\n journal = {CoRR},\n volume = {abs/1506.08909},\n year = {2015},\n url = {http://arxiv.org/abs/1506.08909},\n archivePrefix = {arXiv},\n eprint = {1506.08909},\n timestamp = {Mon, 13 Aug 2018 16:48:23 +0200},\n biburl = {https://dblp.org/rec/journals/corr/LowePSP15.bib},\n bibsource = {dblp computer science bibliography, https://dblp.org}\n}\n", "homepage": "https://github.com/rkadlec/ubuntu-ranking-dataset-creator", "license": "", "features": {"Context": {"dtype": "string", "id": null, "_type": "Value"}, "Utterance": {"dtype": "string", "id": null, "_type": "Value"}, "Label": {"dtype": "int32", "id": null, "_type": "Value"}}, "supervised_keys": null, "builder_name": "ubuntu_dialogs_corpus", "config_name": "train", "version": {"version_str": "2.0.0", "description": null, "datasets_version_to_prepare": null, "major": 2, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 65497027, "num_examples": 127422, "dataset_name": "ubuntu_dialogs_corpus"}}, "download_checksums": {}, "download_size": 0, "dataset_size": 65497027, "size_in_bytes": 65497027}}
|
dummy/dev_test/2.0.0/dummy_data.zip
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:cd0d07d667dea3347a6dbe380ad1040e94ef149153909c7b20425cbb048001df
|
3 |
+
size 1824
|
dummy/train/2.0.0/dummy_data.zip
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:14cddad0c84ff0f8a9a8e63f5f38105710bffbd124160334e5fe4362f376e374
|
3 |
+
size 1152
|
ubuntu_dialogs_corpus.py
ADDED
@@ -0,0 +1,132 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
"""TODO(ubuntu_dialogs_corpus): Add a description here."""
|
2 |
+
|
3 |
+
from __future__ import absolute_import, division, print_function
|
4 |
+
|
5 |
+
import csv
|
6 |
+
import os
|
7 |
+
|
8 |
+
import datasets
|
9 |
+
|
10 |
+
|
11 |
+
# TODO(ubuntu_dialogs_corpus): BibTeX citation
|
12 |
+
_CITATION = """\
|
13 |
+
@article{DBLP:journals/corr/LowePSP15,
|
14 |
+
author = {Ryan Lowe and
|
15 |
+
Nissan Pow and
|
16 |
+
Iulian Serban and
|
17 |
+
Joelle Pineau},
|
18 |
+
title = {The Ubuntu Dialogue Corpus: {A} Large Dataset for Research in Unstructured
|
19 |
+
Multi-Turn Dialogue Systems},
|
20 |
+
journal = {CoRR},
|
21 |
+
volume = {abs/1506.08909},
|
22 |
+
year = {2015},
|
23 |
+
url = {http://arxiv.org/abs/1506.08909},
|
24 |
+
archivePrefix = {arXiv},
|
25 |
+
eprint = {1506.08909},
|
26 |
+
timestamp = {Mon, 13 Aug 2018 16:48:23 +0200},
|
27 |
+
biburl = {https://dblp.org/rec/journals/corr/LowePSP15.bib},
|
28 |
+
bibsource = {dblp computer science bibliography, https://dblp.org}
|
29 |
+
}
|
30 |
+
"""
|
31 |
+
|
32 |
+
# TODO(ubuntu_dialogs_corpus):
|
33 |
+
_DESCRIPTION = """\
|
34 |
+
Ubuntu Dialogue Corpus, a dataset containing almost 1 million multi-turn dialogues, with a total of over 7 million utterances and 100 million words. This provides a unique resource for research into building dialogue managers based on neural language models that can make use of large amounts of unlabeled data. The dataset has both the multi-turn property of conversations in the Dialog State Tracking Challenge datasets, and the unstructured nature of interactions from microblog services such as Twitter.
|
35 |
+
"""
|
36 |
+
|
37 |
+
|
38 |
+
class UbuntuDialogsCorpusConfig(datasets.BuilderConfig):
|
39 |
+
"""BuilderConfig for UbuntuDialogsCorpus."""
|
40 |
+
|
41 |
+
def __init__(self, features, **kwargs):
|
42 |
+
"""BuilderConfig for UbuntuDialogsCorpus.
|
43 |
+
|
44 |
+
Args:
|
45 |
+
|
46 |
+
**kwargs: keyword arguments forwarded to super.
|
47 |
+
"""
|
48 |
+
|
49 |
+
super(UbuntuDialogsCorpusConfig, self).__init__(version=datasets.Version("2.0.0"), **kwargs)
|
50 |
+
self.features = features
|
51 |
+
|
52 |
+
|
53 |
+
class UbuntuDialogsCorpus(datasets.GeneratorBasedBuilder):
|
54 |
+
"""TODO(ubuntu_dialogs_corpus): Short description of my dataset."""
|
55 |
+
|
56 |
+
# TODO(ubuntu_dialogs_corpus): Set up version.
|
57 |
+
VERSION = datasets.Version("2.0.0")
|
58 |
+
BUILDER_CONFIGS = [
|
59 |
+
UbuntuDialogsCorpusConfig(
|
60 |
+
name="train", features=["Context", "Utterance", "Label"], description="training features"
|
61 |
+
),
|
62 |
+
UbuntuDialogsCorpusConfig(
|
63 |
+
name="dev_test",
|
64 |
+
features=["Context", "Ground Truth Utterance"] + ["Distractor_" + str(i) for i in range(9)],
|
65 |
+
description="test and dev features",
|
66 |
+
),
|
67 |
+
]
|
68 |
+
|
69 |
+
@property
|
70 |
+
def manual_download_instructions(self):
|
71 |
+
return """\
|
72 |
+
Please download the Ubuntu Dialog Corpus from https://github.com/rkadlec/ubuntu-ranking-dataset-creator. Run ./generate.sh -t -s -l to download the
|
73 |
+
data. Others arguments are left to their default values here. Please save train.csv, test.csv and valid.csv in the same path"""
|
74 |
+
|
75 |
+
def _info(self):
|
76 |
+
# TODO(ubuntu_dialogs_corpus): Specifies the datasets.DatasetInfo object
|
77 |
+
features = {feature: datasets.Value("string") for feature in self.config.features}
|
78 |
+
if self.config.name == "train":
|
79 |
+
features["Label"] = datasets.Value("int32")
|
80 |
+
return datasets.DatasetInfo(
|
81 |
+
# This is the description that will appear on the datasets page.
|
82 |
+
description=_DESCRIPTION,
|
83 |
+
# datasets.features.FeatureConnectors
|
84 |
+
features=datasets.Features(
|
85 |
+
# These are the features of your dataset like images, labels ...
|
86 |
+
features
|
87 |
+
),
|
88 |
+
# If there's a common (input, target) tuple from the features,
|
89 |
+
# specify them here. They'll be used if as_supervised=True in
|
90 |
+
# builder.as_dataset.
|
91 |
+
supervised_keys=None,
|
92 |
+
# Homepage of the dataset for documentation
|
93 |
+
homepage="https://github.com/rkadlec/ubuntu-ranking-dataset-creator",
|
94 |
+
citation=_CITATION,
|
95 |
+
)
|
96 |
+
|
97 |
+
def _split_generators(self, dl_manager):
|
98 |
+
"""Returns SplitGenerators."""
|
99 |
+
# TODO(ubuntu_dialogs_corpus): Downloads the data and defines the splits
|
100 |
+
# dl_manager is a datasets.download.DownloadManager that can be used to
|
101 |
+
# download and extract URLs
|
102 |
+
manual_dir = os.path.abspath(os.path.expanduser(dl_manager.manual_dir))
|
103 |
+
|
104 |
+
if self.config.name == "train":
|
105 |
+
return [
|
106 |
+
datasets.SplitGenerator(
|
107 |
+
name=datasets.Split.TRAIN,
|
108 |
+
# These kwargs will be passed to _generate_examples
|
109 |
+
gen_kwargs={"filepath": os.path.join(manual_dir, "train.csv")},
|
110 |
+
),
|
111 |
+
]
|
112 |
+
else:
|
113 |
+
return [
|
114 |
+
datasets.SplitGenerator(
|
115 |
+
name=datasets.Split.TEST,
|
116 |
+
# These kwargs will be passed to _generate_examples
|
117 |
+
gen_kwargs={"filepath": os.path.join(manual_dir, "test.csv")},
|
118 |
+
),
|
119 |
+
datasets.SplitGenerator(
|
120 |
+
name=datasets.Split.VALIDATION,
|
121 |
+
# These kwargs will be passed to _generate_examples
|
122 |
+
gen_kwargs={"filepath": os.path.join(manual_dir, "valid.csv")},
|
123 |
+
),
|
124 |
+
]
|
125 |
+
|
126 |
+
def _generate_examples(self, filepath):
|
127 |
+
"""Yields examples."""
|
128 |
+
# TODO(ubuntu_dialogs_corpus): Yields (key, example) tuples from the dataset
|
129 |
+
with open(filepath, encoding="utf-8") as f:
|
130 |
+
data = csv.DictReader(f)
|
131 |
+
for id_, row in enumerate(data):
|
132 |
+
yield id_, row
|