Convert dataset to Parquet
#1
by
mpkato
- opened
README.md
CHANGED
@@ -1,8 +1,27 @@
|
|
1 |
---
|
|
|
|
|
2 |
license: apache-2.0
|
3 |
task_categories:
|
4 |
- text-retrieval
|
5 |
-
language:
|
6 |
-
- ja
|
7 |
pretty_name: a
|
8 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
+
language:
|
3 |
+
- ja
|
4 |
license: apache-2.0
|
5 |
task_categories:
|
6 |
- text-retrieval
|
|
|
|
|
7 |
pretty_name: a
|
8 |
+
dataset_info:
|
9 |
+
features:
|
10 |
+
- name: docid
|
11 |
+
dtype: string
|
12 |
+
- name: title
|
13 |
+
dtype: string
|
14 |
+
- name: text
|
15 |
+
dtype: string
|
16 |
+
splits:
|
17 |
+
- name: train
|
18 |
+
num_bytes: 82583007
|
19 |
+
num_examples: 129260
|
20 |
+
download_size: 44837491
|
21 |
+
dataset_size: 82583007
|
22 |
+
configs:
|
23 |
+
- config_name: default
|
24 |
+
data_files:
|
25 |
+
- split: train
|
26 |
+
path: data/train-*
|
27 |
+
---
|
create_miracl_japanese_small.py
DELETED
@@ -1,30 +0,0 @@
|
|
1 |
-
import pandas as pd
|
2 |
-
import json
|
3 |
-
from tqdm import tqdm
|
4 |
-
from datasets import load_dataset
|
5 |
-
|
6 |
-
DEV_QREL_FILEPATH = "./qrels.miracl-v1.0-ja-dev.tsv"
|
7 |
-
OUTPUT_FILEPATH = "./miracl-japanese-small-docs.jsonl"
|
8 |
-
|
9 |
-
def extract_doc_ids(filepath):
|
10 |
-
dev_qrel = pd.read_csv(filepath, delimiter='\t',
|
11 |
-
names=['query_id', 'ph', 'doc_pas_id', 'rel'])
|
12 |
-
doc_ids = set([int(dp_id.split('#')[0]) for dp_id in dev_qrel.doc_pas_id])
|
13 |
-
return doc_ids
|
14 |
-
|
15 |
-
|
16 |
-
if __name__ == '__main__':
|
17 |
-
dev_doc_ids = extract_doc_ids(DEV_QREL_FILEPATH)
|
18 |
-
doc_ids = dev_doc_ids
|
19 |
-
print("# of docids in dev", len(dev_doc_ids))
|
20 |
-
|
21 |
-
new_dataset = []
|
22 |
-
seen_doc_ids = set()
|
23 |
-
dataset = load_dataset("miracl/miracl-corpus", "ja")
|
24 |
-
for data in tqdm(dataset['train']):
|
25 |
-
docid = int(data["docid"].split("#")[0])
|
26 |
-
if docid in doc_ids:
|
27 |
-
new_dataset.append(data)
|
28 |
-
with open(OUTPUT_FILEPATH, 'w', encoding='utf-8') as f:
|
29 |
-
for data in new_dataset:
|
30 |
-
f.write(json.dumps(data) + '\n')
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
miracl-japanese-small-docs.jsonl.gz → data/train-00000-of-00001.parquet
RENAMED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8b89f5a20e5fd08d4a181728f66eec6810b60387bc87c37b7aeab80b1d10e499
|
3 |
+
size 44837491
|
miracl-japanese-small-corpus.py
DELETED
@@ -1,75 +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 |
-
import json
|
17 |
-
import datasets
|
18 |
-
|
19 |
-
_CITATION = '''
|
20 |
-
'''
|
21 |
-
|
22 |
-
_DESCRIPTION = 'dataset load script for MIRACL Japanese Small Corpus'
|
23 |
-
|
24 |
-
_DATASET_URLS = {
|
25 |
-
'train': 'https://huggingface.co/datasets/mpkato/miracl-japanese-small-corpus/resolve/main/miracl-japanese-small-docs.jsonl.gz'
|
26 |
-
}
|
27 |
-
|
28 |
-
|
29 |
-
class MIRACLJapaneseSmallCorpus(datasets.GeneratorBasedBuilder):
|
30 |
-
BUILDER_CONFIGS = [
|
31 |
-
datasets.BuilderConfig(
|
32 |
-
version=datasets.Version('1.0.0'),
|
33 |
-
description=f'MIRACL Japanese Small dataset.'
|
34 |
-
)
|
35 |
-
]
|
36 |
-
|
37 |
-
def _info(self):
|
38 |
-
features = datasets.Features({
|
39 |
-
'docid': datasets.Value('string'),
|
40 |
-
'title': datasets.Value('string'),
|
41 |
-
'text': datasets.Value('string'),
|
42 |
-
})
|
43 |
-
|
44 |
-
return datasets.DatasetInfo(
|
45 |
-
# This is the description that will appear on the datasets page.
|
46 |
-
description=_DESCRIPTION,
|
47 |
-
# This defines the different columns of the dataset and their types
|
48 |
-
features=features, # Here we define them above because they are different between the two configurations
|
49 |
-
supervised_keys=None,
|
50 |
-
# Homepage of the dataset for documentation
|
51 |
-
homepage='',
|
52 |
-
# License for the dataset if available
|
53 |
-
license='',
|
54 |
-
# Citation for the dataset
|
55 |
-
citation=_CITATION,
|
56 |
-
)
|
57 |
-
|
58 |
-
def _split_generators(self, dl_manager):
|
59 |
-
downloaded_files = dl_manager.download_and_extract(_DATASET_URLS)
|
60 |
-
|
61 |
-
splits = [
|
62 |
-
datasets.SplitGenerator(
|
63 |
-
name='train',
|
64 |
-
gen_kwargs={
|
65 |
-
'filepath': downloaded_files['train'],
|
66 |
-
},
|
67 |
-
),
|
68 |
-
]
|
69 |
-
return splits
|
70 |
-
|
71 |
-
def _generate_examples(self, filepath):
|
72 |
-
with open(filepath, encoding="utf-8") as f:
|
73 |
-
for line in f:
|
74 |
-
data = json.loads(line)
|
75 |
-
yield data['docid'], data
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
qrels.miracl-v1.0-ja-dev.tsv
DELETED
The diff for this file is too large to render.
See raw diff
|
|