JulesBelveze
commited on
Commit
•
98601fe
1
Parent(s):
ca10947
attempt to format according to datasets library
Browse files- dataset_infos.json +1 -0
- dummy/all/1.1.0/dummy_data.zip +0 -0
- dummy/all/1.1.0/dummy_data/.DS_Store +0 -0
- dummy/all/1.1.0/dummy_data/1.1.0.tar.gz%3Fraw%3Dtrue/.DS_Store +0 -0
- dummy/all/{test.jsonl → 1.1.0/dummy_data/1.1.0.tar.gz%3Fraw%3Dtrue/1.1.0/test.json} +0 -0
- dummy/all/{train.jsonl → 1.1.0/dummy_data/1.1.0.tar.gz%3Fraw%3Dtrue/1.1.0/train.json} +0 -0
- test.jsonl +0 -0
- tldr_news.py +87 -0
- train.jsonl +0 -0
dataset_infos.json
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
{"all": {"description": "The `tldr_news` dataset was constructed by collecting a daily tech newsletter (available at \nhttps://tldr.tech/newsletter). Then for every piece of news, the \"headline\" and its corresponding \"content\" were \ncollected. Such a dataset can be used to train a model to generate a headline from a input piece of text.\n", "citation": "", "homepage": "", "license": "", "features": {"headline": {"dtype": "string", "id": null, "_type": "Value"}, "content": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "tldr_news", "config_name": "all", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 3880430, "num_examples": 7055, "dataset_name": "tldr_news"}, "test": {"name": "test", "num_bytes": 429049, "num_examples": 784, "dataset_name": "tldr_news"}}, "download_checksums": {"https://github.com/JulesBelveze/tldr_news/blob/main/1.1.0.tar.gz?raw=true": {"num_bytes": 1663243, "checksum": "2cdb7b21a2b06af5e1318c0155a20ae652aa418e0d599ff03f73e633b5ada052"}}, "download_size": 1663243, "post_processing_size": null, "dataset_size": 4309479, "size_in_bytes": 5972722}}
|
dummy/all/1.1.0/dummy_data.zip
ADDED
Binary file (4.02 kB). View file
|
|
dummy/all/1.1.0/dummy_data/.DS_Store
ADDED
Binary file (6.15 kB). View file
|
|
dummy/all/1.1.0/dummy_data/1.1.0.tar.gz%3Fraw%3Dtrue/.DS_Store
ADDED
Binary file (6.15 kB). View file
|
|
dummy/all/{test.jsonl → 1.1.0/dummy_data/1.1.0.tar.gz%3Fraw%3Dtrue/1.1.0/test.json}
RENAMED
File without changes
|
dummy/all/{train.jsonl → 1.1.0/dummy_data/1.1.0.tar.gz%3Fraw%3Dtrue/1.1.0/train.json}
RENAMED
File without changes
|
test.jsonl
DELETED
The diff for this file is too large to render.
See raw diff
|
|
tldr_news.py
ADDED
@@ -0,0 +1,87 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# Copyright 2020 The HuggingFace Datasets Authors and the current dataset script contributor.
|
2 |
+
#
|
3 |
+
# Licensed under the Apache License, Version 2.0 (the "License");
|
4 |
+
# you may not use this file except in compliance with the License.
|
5 |
+
# You may obtain a copy of the License at
|
6 |
+
#
|
7 |
+
# http://www.apache.org/licenses/LICENSE-2.0
|
8 |
+
#
|
9 |
+
# Unless required by applicable law or agreed to in writing, software
|
10 |
+
# distributed under the License is distributed on an "AS IS" BASIS,
|
11 |
+
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
12 |
+
# See the License for the specific language governing permissions and
|
13 |
+
# limitations under the License.
|
14 |
+
|
15 |
+
import json
|
16 |
+
import os
|
17 |
+
|
18 |
+
import datasets
|
19 |
+
|
20 |
+
|
21 |
+
_DESCRIPTION = """\
|
22 |
+
The `tldr_news` dataset was constructed by collecting a daily tech newsletter (available at
|
23 |
+
https://tldr.tech/newsletter). Then for every piece of news, the "headline" and its corresponding "content" were
|
24 |
+
collected. Such a dataset can be used to train a model to generate a headline from a input piece of text.
|
25 |
+
"""
|
26 |
+
|
27 |
+
# TODO: Add link to the official dataset URLs here
|
28 |
+
# The HuggingFace Datasets library doesn't host the datasets but only points to the original files.
|
29 |
+
# This can be an arbitrary nested dict/list of URLs (see below in `_split_generators` method)
|
30 |
+
_URLS = {"all": "https://github.com/JulesBelveze/tldr_news/blob/main/1.1.0.tar.gz?raw=true"}
|
31 |
+
|
32 |
+
|
33 |
+
class TLDRNewsConfig(datasets.BuilderConfig):
|
34 |
+
"""BuilderConfig for TLDRNews."""
|
35 |
+
|
36 |
+
def __init__(self, **kwargs):
|
37 |
+
"""BuilderConfig for TLDRNews.
|
38 |
+
Args:
|
39 |
+
**kwargs: keyword arguments forwarded to super.
|
40 |
+
"""
|
41 |
+
super(TLDRNewsConfig, self).__init__(**kwargs)
|
42 |
+
|
43 |
+
|
44 |
+
class TLDRNewsDataset(datasets.GeneratorBasedBuilder):
|
45 |
+
"""Dataset containing headline & content of pieces of news from the tldr tech newsletter."""
|
46 |
+
|
47 |
+
VERSION = datasets.Version("1.1.0")
|
48 |
+
|
49 |
+
BUILDER_CONFIGS = [
|
50 |
+
TLDRNewsConfig(name="all", version=VERSION, description="This contains all the existing newsletter"),
|
51 |
+
]
|
52 |
+
|
53 |
+
DEFAULT_CONFIG_NAME = "all"
|
54 |
+
|
55 |
+
def _info(self):
|
56 |
+
features = datasets.Features(
|
57 |
+
{
|
58 |
+
"headline": datasets.Value("string"),
|
59 |
+
"content": datasets.Value("string"),
|
60 |
+
}
|
61 |
+
)
|
62 |
+
|
63 |
+
return datasets.DatasetInfo(description=_DESCRIPTION, features=features)
|
64 |
+
|
65 |
+
def _split_generators(self, dl_manager):
|
66 |
+
urls = _URLS[self.config.name]
|
67 |
+
data_dir = dl_manager.download_and_extract(urls)
|
68 |
+
data_dir = os.path.join(data_dir, str(self.config.version))
|
69 |
+
return [
|
70 |
+
datasets.SplitGenerator(
|
71 |
+
name=datasets.Split.TRAIN,
|
72 |
+
gen_kwargs={
|
73 |
+
"filepath": os.path.join(data_dir, "train.json"),
|
74 |
+
"split": "train",
|
75 |
+
},
|
76 |
+
),
|
77 |
+
datasets.SplitGenerator(
|
78 |
+
name=datasets.Split.TEST,
|
79 |
+
gen_kwargs={"filepath": os.path.join(data_dir, "test.json"), "split": "test"},
|
80 |
+
),
|
81 |
+
]
|
82 |
+
|
83 |
+
def _generate_examples(self, filepath, split):
|
84 |
+
with open(filepath, encoding="utf-8") as f:
|
85 |
+
data = json.load(f)
|
86 |
+
for key, row in enumerate(data):
|
87 |
+
yield key, {"headline": row["headline"], "content": row["content"]}
|
train.jsonl
DELETED
The diff for this file is too large to render.
See raw diff
|
|