Datasets:

ArXiv:
License:
bismarck91 commited on
Commit
d5cc487
1 Parent(s): a54931f

enA-unA files

Browse files
README.md CHANGED
@@ -1,3 +1,144 @@
1
  ---
2
  license: mit
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: mit
3
  ---
4
+ # Dataset Card for Seamless-Align (WIP). Inspired by https://huggingface.co/datasets/allenai/nllb
5
+
6
+ ## Table of Contents
7
+ - [Table of Contents](#table-of-contents)
8
+ - [Dataset Description](#dataset-description)
9
+ - [Dataset Summary](#dataset-summary)
10
+ - [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)
11
+ - [Languages](#languages)
12
+ - [Dataset Structure](#dataset-structure)
13
+ - [Data Instances](#data-instances)
14
+ - [Data Fields](#data-fields)
15
+ - [Data Splits](#data-splits)
16
+ - [Dataset Creation](#dataset-creation)
17
+ - [Curation Rationale](#curation-rationale)
18
+ - [Source Data](#source-data)
19
+ - [Annotations](#annotations)
20
+ - [Personal and Sensitive Information](#personal-and-sensitive-information)
21
+ - [Considerations for Using the Data](#considerations-for-using-the-data)
22
+ - [Social Impact of Dataset](#social-impact-of-dataset)
23
+ - [Discussion of Biases](#discussion-of-biases)
24
+ - [Other Known Limitations](#other-known-limitations)
25
+ - [Additional Information](#additional-information)
26
+ - [Dataset Curators](#dataset-curators)
27
+ - [Licensing Information](#licensing-information)
28
+ - [Citation Information](#citation-information)
29
+ - [Contributions](#contributions)
30
+
31
+ ## Dataset Description
32
+
33
+ - **Homepage:** [Needs More Information]
34
+ - **Repository:** [Needs More Information]
35
+ - **Paper:** [Needs More Information]
36
+ - **Leaderboard:** [Needs More Information]
37
+ - **Point of Contact:** [Needs More Information]
38
+
39
+ ### Dataset Summary
40
+
41
+ This dataset was created based on [metadata](https://github.com/facebookresearch/seamless_communication/blob/main/docs/m4t/seamless_align_README.md) for mined Speech-to-Speech(S2S), Text-to-Speech(TTS) and Speech-to-Text(S2T) released by Meta AI. The S2S contains data for 35 language pairs. The S2S dataset is ~1000GB compressed.
42
+
43
+
44
+ #### How to use the data
45
+ There are two ways to access the data:
46
+ * Via the Hugging Face Python datasets library
47
+
48
+ ```
49
+ Scripts coming soon
50
+ ```
51
+
52
+ * Clone the git repo
53
+ ```
54
+ git lfs install
55
+ git clone https://huggingface.co/datasets/allenai/nllb
56
+ ```
57
+
58
+ ### Supported Tasks and Leaderboards
59
+
60
+ N/A
61
+
62
+ ### Languages
63
+
64
+ Language pairs can be found [here](https://github.com/facebookresearch/seamless_communication/blob/main/docs/m4t/seamless_align_README.md).
65
+
66
+ ## Dataset Structure
67
+
68
+ The S2S dataset contains two gzipped files src.tar.gz annd tgt.tar.gz
69
+
70
+
71
+ ### Data Instances
72
+
73
+ The number of instances for each language pair can be found in the [dataset_infos.json](https://huggingface.co/datasets/allenai/nllb/blob/main/dataset_infos.json) file.
74
+
75
+ ### Data Fields
76
+
77
+ Data Field can be found [here](https://github.com/facebookresearch/seamless_communication/blob/main/docs/m4t/seamless_align_README.md).
78
+
79
+
80
+ ### Data Splits
81
+
82
+ The data is not split.
83
+
84
+
85
+ ## Dataset Creation
86
+
87
+ ### Curation Rationale
88
+
89
+
90
+
91
+ ### Source Data
92
+
93
+ Inspect links in metadata
94
+
95
+ #### Who are the source language producers?
96
+
97
+ Speech and Text was collected from the web many of which are web crawls.
98
+ ### Annotations
99
+
100
+ #### Annotation process
101
+
102
+ Parallel sentences were identified using SONAR encoders. (Duquenne et al., 2023)
103
+
104
+ #### Who are the annotators?
105
+
106
+ The data was not human annotated.
107
+
108
+ ### Personal and Sensitive Information
109
+
110
+ Data may contain personally identifiable information, sensitive content, or toxic content that was publicly shared on the Internet.
111
+
112
+ ## Considerations for Using the Data
113
+
114
+ ### Social Impact of Dataset
115
+
116
+ This dataset provides data for training machine learning systems for many languages.
117
+
118
+ ### Discussion of Biases
119
+
120
+ Biases in the data have not been specifically studied, however as the original source of data is World Wide Web it is likely that the data has biases similar to those prevalent in the Internet. The data may also exhibit biases introduced by language identification and data filtering techniques; lower resource languages generally have lower accuracy.
121
+
122
+ ### Other Known Limitations
123
+
124
+ Some of the translations are in fact machine translations. While some website machine translation tools are identifiable from HTML source, these tools were not filtered out en mass because raw HTML was not available from some sources and CommonCrawl processing started from WET files.
125
+
126
+ ## Additional Information
127
+
128
+ ### Dataset Curators
129
+
130
+ The data was not curated.
131
+
132
+ ### Licensing Information
133
+
134
+ The dataset is released under the terms of [MIT](https://opensource.org/license/mit/). **PLEASE, USE DATA RESPONSIBLY**
135
+
136
+
137
+ ### Citation Information
138
+
139
+ Seamless Communication et al, SeamlessM4T: Massively Multilingual & Multimodal Machine Translation. arXiv https://arxiv.org/abs/2308.11596, 2023. <br>
140
+ Duquenne et al, SONAR: Sentence-Level Multimodal and Language-Agnostic Representations. arXiv https://arxiv.org/abs/2308.11466, 2023
141
+
142
+ ### Contributions
143
+
144
+ We thank the Seamless Communication Meta AI team for open sourcing the meta data and instructions on how to use it with special thanks to Loïc Barrault, Yu-An Chung, Mariano Cora Meglioli, David Dale, Ning Dong, Paul-Ambroise Duquenne, Hady Elsahar, Hongyu Gong, Kevin Heffernan, John Hoffman, Christopher Klaiber, Pengwei Li, Daniel Licht, Jean Maillard, Alice Rakotoarison, Kaushik Ram Sadagopan, Guillaume Wenzek, Ethan Ye, Bapi Akula, Peng-Jen Chen, Naji El Hachem, Brian Ellis, Gabriel Mejia Gonzalez, Justin Haaheim, Prangthip Hansanti, Russ Howes, Bernie Huang, Min-Jae Hwang, Hirofumi Inaguma, Somya Jain, Elahe Kalbassi, Amanda Kallet, Ilia Kulikov, Janice Lam, Daniel Li, Xutai Ma, Ruslan Mavlyutov, Benjamin Peloquin, Mohamed Ramadan, Abinesh Ramakrishnan, Anna Sun, Kevin Tran, Tuan Tran, Igor Tufanov, Vish Vogeti, Carleigh Wood, Yilin Yang, Bokai Yu, Pierre Andrews, Can Balioglu, Marta R. Costa-jussà, Onur Celebi, Maha Elbayad, Cynthia Gao, Francisco Guzmán, Justine Kao, Ann Lee, Alexandre Mourachko, Juan Pino, Sravya Popuri, Christophe Ropers, Safiyyah Saleem, Holger Schwenk, Paden Tomasello, Changhan Wang, Jeff Wang, Skyler Wang. We also thank the Center for Language and Speech Processing(CLSP) for hosting and releasing this data, including Bismarck Bamfo Odoom and Philipp Koehn (for engineering efforts to host the data, and releasing the huggingface dataset), and Alexandre Mourachko (for organizing the connection).
data_s2s/enA-uzA/.gitattributes ADDED
@@ -0,0 +1,2 @@
 
 
 
1
+ src.tar.gz filter=lfs diff=lfs merge=lfs -text
2
+ tgt.tar.gz filter=lfs diff=lfs merge=lfs -text
data_s2s/enA-uzA/src.tar.gz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:91bc66b8c488cdfb7efac6b4657a58fc1e7c02f38b1d4950b477534a81d8d528
3
+ size 175733601
data_s2s/enA-uzA/tgt.tar.gz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a3285107d0979589b79a6d47bb1f6878d8236f225956e12cdb923a30215c68d7
3
+ size 195771845