Release 2610 from techiaith/kaldi-cy-2610@main
Browse files
LICENSE
ADDED
|
@@ -0,0 +1,202 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
|
| 2 |
+
Apache License
|
| 3 |
+
Version 2.0, January 2004
|
| 4 |
+
http://www.apache.org/licenses/
|
| 5 |
+
|
| 6 |
+
TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
|
| 7 |
+
|
| 8 |
+
1. Definitions.
|
| 9 |
+
|
| 10 |
+
"License" shall mean the terms and conditions for use, reproduction,
|
| 11 |
+
and distribution as defined by Sections 1 through 9 of this document.
|
| 12 |
+
|
| 13 |
+
"Licensor" shall mean the copyright owner or entity authorized by
|
| 14 |
+
the copyright owner that is granting the License.
|
| 15 |
+
|
| 16 |
+
"Legal Entity" shall mean the union of the acting entity and all
|
| 17 |
+
other entities that control, are controlled by, or are under common
|
| 18 |
+
control with that entity. For the purposes of this definition,
|
| 19 |
+
"control" means (i) the power, direct or indirect, to cause the
|
| 20 |
+
direction or management of such entity, whether by contract or
|
| 21 |
+
otherwise, or (ii) ownership of fifty percent (50%) or more of the
|
| 22 |
+
outstanding shares, or (iii) beneficial ownership of such entity.
|
| 23 |
+
|
| 24 |
+
"You" (or "Your") shall mean an individual or Legal Entity
|
| 25 |
+
exercising permissions granted by this License.
|
| 26 |
+
|
| 27 |
+
"Source" form shall mean the preferred form for making modifications,
|
| 28 |
+
including but not limited to software source code, documentation
|
| 29 |
+
source, and configuration files.
|
| 30 |
+
|
| 31 |
+
"Object" form shall mean any form resulting from mechanical
|
| 32 |
+
transformation or translation of a Source form, including but
|
| 33 |
+
not limited to compiled object code, generated documentation,
|
| 34 |
+
and conversions to other media types.
|
| 35 |
+
|
| 36 |
+
"Work" shall mean the work of authorship, whether in Source or
|
| 37 |
+
Object form, made available under the License, as indicated by a
|
| 38 |
+
copyright notice that is included in or attached to the work
|
| 39 |
+
(an example is provided in the Appendix below).
|
| 40 |
+
|
| 41 |
+
"Derivative Works" shall mean any work, whether in Source or Object
|
| 42 |
+
form, that is based on (or derived from) the Work and for which the
|
| 43 |
+
editorial revisions, annotations, elaborations, or other modifications
|
| 44 |
+
represent, as a whole, an original work of authorship. For the purposes
|
| 45 |
+
of this License, Derivative Works shall not include works that remain
|
| 46 |
+
separable from, or merely link (or bind by name) to the interfaces of,
|
| 47 |
+
the Work and Derivative Works thereof.
|
| 48 |
+
|
| 49 |
+
"Contribution" shall mean any work of authorship, including
|
| 50 |
+
the original version of the Work and any modifications or additions
|
| 51 |
+
to that Work or Derivative Works thereof, that is intentionally
|
| 52 |
+
submitted to Licensor for inclusion in the Work by the copyright owner
|
| 53 |
+
or by an individual or Legal Entity authorized to submit on behalf of
|
| 54 |
+
the copyright owner. For the purposes of this definition, "submitted"
|
| 55 |
+
means any form of electronic, verbal, or written communication sent
|
| 56 |
+
to the Licensor or its representatives, including but not limited to
|
| 57 |
+
communication on electronic mailing lists, source code control systems,
|
| 58 |
+
and issue tracking systems that are managed by, or on behalf of, the
|
| 59 |
+
Licensor for the purpose of discussing and improving the Work, but
|
| 60 |
+
excluding communication that is conspicuously marked or otherwise
|
| 61 |
+
designated in writing by the copyright owner as "Not a Contribution."
|
| 62 |
+
|
| 63 |
+
"Contributor" shall mean Licensor and any individual or Legal Entity
|
| 64 |
+
on behalf of whom a Contribution has been received by Licensor and
|
| 65 |
+
subsequently incorporated within the Work.
|
| 66 |
+
|
| 67 |
+
2. Grant of Copyright License. Subject to the terms and conditions of
|
| 68 |
+
this License, each Contributor hereby grants to You a perpetual,
|
| 69 |
+
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
|
| 70 |
+
copyright license to reproduce, prepare Derivative Works of,
|
| 71 |
+
publicly display, publicly perform, sublicense, and distribute the
|
| 72 |
+
Work and such Derivative Works in Source or Object form.
|
| 73 |
+
|
| 74 |
+
3. Grant of Patent License. Subject to the terms and conditions of
|
| 75 |
+
this License, each Contributor hereby grants to You a perpetual,
|
| 76 |
+
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
|
| 77 |
+
(except as stated in this section) patent license to make, have made,
|
| 78 |
+
use, offer to sell, sell, import, and otherwise transfer the Work,
|
| 79 |
+
where such license applies only to those patent claims licensable
|
| 80 |
+
by such Contributor that are necessarily infringed by their
|
| 81 |
+
Contribution(s) alone or by combination of their Contribution(s)
|
| 82 |
+
with the Work to which such Contribution(s) was submitted. If You
|
| 83 |
+
institute patent litigation against any entity (including a
|
| 84 |
+
cross-claim or counterclaim in a lawsuit) alleging that the Work
|
| 85 |
+
or a Contribution incorporated within the Work constitutes direct
|
| 86 |
+
or contributory patent infringement, then any patent licenses
|
| 87 |
+
granted to You under this License for that Work shall terminate
|
| 88 |
+
as of the date such litigation is filed.
|
| 89 |
+
|
| 90 |
+
4. Redistribution. You may reproduce and distribute copies of the
|
| 91 |
+
Work or Derivative Works thereof in any medium, with or without
|
| 92 |
+
modifications, and in Source or Object form, provided that You
|
| 93 |
+
meet the following conditions:
|
| 94 |
+
|
| 95 |
+
(a) You must give any other recipients of the Work or
|
| 96 |
+
Derivative Works a copy of this License; and
|
| 97 |
+
|
| 98 |
+
(b) You must cause any modified files to carry prominent notices
|
| 99 |
+
stating that You changed the files; and
|
| 100 |
+
|
| 101 |
+
(c) You must retain, in the Source form of any Derivative Works
|
| 102 |
+
that You distribute, all copyright, patent, trademark, and
|
| 103 |
+
attribution notices from the Source form of the Work,
|
| 104 |
+
excluding those notices that do not pertain to any part of
|
| 105 |
+
the Derivative Works; and
|
| 106 |
+
|
| 107 |
+
(d) If the Work includes a "NOTICE" text file as part of its
|
| 108 |
+
distribution, then any Derivative Works that You distribute must
|
| 109 |
+
include a readable copy of the attribution notices contained
|
| 110 |
+
within such NOTICE file, excluding those notices that do not
|
| 111 |
+
pertain to any part of the Derivative Works, in at least one
|
| 112 |
+
of the following places: within a NOTICE text file distributed
|
| 113 |
+
as part of the Derivative Works; within the Source form or
|
| 114 |
+
documentation, if provided along with the Derivative Works; or,
|
| 115 |
+
within a display generated by the Derivative Works, if and
|
| 116 |
+
wherever such third-party notices normally appear. The contents
|
| 117 |
+
of the NOTICE file are for informational purposes only and
|
| 118 |
+
do not modify the License. You may add Your own attribution
|
| 119 |
+
notices within Derivative Works that You distribute, alongside
|
| 120 |
+
or as an addendum to the NOTICE text from the Work, provided
|
| 121 |
+
that such additional attribution notices cannot be construed
|
| 122 |
+
as modifying the License.
|
| 123 |
+
|
| 124 |
+
You may add Your own copyright statement to Your modifications and
|
| 125 |
+
may provide additional or different license terms and conditions
|
| 126 |
+
for use, reproduction, or distribution of Your modifications, or
|
| 127 |
+
for any such Derivative Works as a whole, provided Your use,
|
| 128 |
+
reproduction, and distribution of the Work otherwise complies with
|
| 129 |
+
the conditions stated in this License.
|
| 130 |
+
|
| 131 |
+
5. Submission of Contributions. Unless You explicitly state otherwise,
|
| 132 |
+
any Contribution intentionally submitted for inclusion in the Work
|
| 133 |
+
by You to the Licensor shall be under the terms and conditions of
|
| 134 |
+
this License, without any additional terms or conditions.
|
| 135 |
+
Notwithstanding the above, nothing herein shall supersede or modify
|
| 136 |
+
the terms of any separate license agreement you may have executed
|
| 137 |
+
with Licensor regarding such Contributions.
|
| 138 |
+
|
| 139 |
+
6. Trademarks. This License does not grant permission to use the trade
|
| 140 |
+
names, trademarks, service marks, or product names of the Licensor,
|
| 141 |
+
except as required for reasonable and customary use in describing the
|
| 142 |
+
origin of the Work and reproducing the content of the NOTICE file.
|
| 143 |
+
|
| 144 |
+
7. Disclaimer of Warranty. Unless required by applicable law or
|
| 145 |
+
agreed to in writing, Licensor provides the Work (and each
|
| 146 |
+
Contributor provides its Contributions) on an "AS IS" BASIS,
|
| 147 |
+
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
|
| 148 |
+
implied, including, without limitation, any warranties or conditions
|
| 149 |
+
of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
|
| 150 |
+
PARTICULAR PURPOSE. You are solely responsible for determining the
|
| 151 |
+
appropriateness of using or redistributing the Work and assume any
|
| 152 |
+
risks associated with Your exercise of permissions under this License.
|
| 153 |
+
|
| 154 |
+
8. Limitation of Liability. In no event and under no legal theory,
|
| 155 |
+
whether in tort (including negligence), contract, or otherwise,
|
| 156 |
+
unless required by applicable law (such as deliberate and grossly
|
| 157 |
+
negligent acts) or agreed to in writing, shall any Contributor be
|
| 158 |
+
liable to You for damages, including any direct, indirect, special,
|
| 159 |
+
incidental, or consequential damages of any character arising as a
|
| 160 |
+
result of this License or out of the use or inability to use the
|
| 161 |
+
Work (including but not limited to damages for loss of goodwill,
|
| 162 |
+
work stoppage, computer failure or malfunction, or any and all
|
| 163 |
+
other commercial damages or losses), even if such Contributor
|
| 164 |
+
has been advised of the possibility of such damages.
|
| 165 |
+
|
| 166 |
+
9. Accepting Warranty or Additional Liability. While redistributing
|
| 167 |
+
the Work or Derivative Works thereof, You may choose to offer,
|
| 168 |
+
and charge a fee for, acceptance of support, warranty, indemnity,
|
| 169 |
+
or other liability obligations and/or rights consistent with this
|
| 170 |
+
License. However, in accepting such obligations, You may act only
|
| 171 |
+
on Your own behalf and on Your sole responsibility, not on behalf
|
| 172 |
+
of any other Contributor, and only if You agree to indemnify,
|
| 173 |
+
defend, and hold each Contributor harmless for any liability
|
| 174 |
+
incurred by, or claims asserted against, such Contributor by reason
|
| 175 |
+
of your accepting any such warranty or additional liability.
|
| 176 |
+
|
| 177 |
+
END OF TERMS AND CONDITIONS
|
| 178 |
+
|
| 179 |
+
APPENDIX: How to apply the Apache License to your work.
|
| 180 |
+
|
| 181 |
+
To apply the Apache License to your work, attach the following
|
| 182 |
+
boilerplate notice, with the fields enclosed by brackets "[]"
|
| 183 |
+
replaced with your own identifying information. (Don't include
|
| 184 |
+
the brackets!) The text should be enclosed in the appropriate
|
| 185 |
+
comment syntax for the file format. We also recommend that a
|
| 186 |
+
file or class name and description of purpose be included on the
|
| 187 |
+
same "printed page" as the copyright notice for easier
|
| 188 |
+
identification within third-party archives.
|
| 189 |
+
|
| 190 |
+
Copyright [yyyy] [name of copyright owner]
|
| 191 |
+
|
| 192 |
+
Licensed under the Apache License, Version 2.0 (the "License");
|
| 193 |
+
you may not use this file except in compliance with the License.
|
| 194 |
+
You may obtain a copy of the License at
|
| 195 |
+
|
| 196 |
+
http://www.apache.org/licenses/LICENSE-2.0
|
| 197 |
+
|
| 198 |
+
Unless required by applicable law or agreed to in writing, software
|
| 199 |
+
distributed under the License is distributed on an "AS IS" BASIS,
|
| 200 |
+
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
| 201 |
+
See the License for the specific language governing permissions and
|
| 202 |
+
limitations under the License.
|
NOTICE
ADDED
|
@@ -0,0 +1,126 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
kaldi-cy 2610, released October 2026
|
| 2 |
+
Copyright 2026 Bangor University
|
| 3 |
+
|
| 4 |
+
Developed by the Language Technologies Unit (Uned Technolegau Iaith), Bangor University.
|
| 5 |
+
|
| 6 |
+
Licensed under the Apache License, Version 2.0 (the "License"). You may obtain a copy of the
|
| 7 |
+
License at http://www.apache.org/licenses/LICENSE-2.0
|
| 8 |
+
|
| 9 |
+
NAMES
|
| 10 |
+
This licence grants no permission to use the names "Bangor University", "Prifysgol Bangor",
|
| 11 |
+
"Language Technologies Unit", "Uned Technolegau Iaith", "techiaith", the University's logos, or
|
| 12 |
+
the names of its products or staff, except to describe the origin of the model as this NOTICE
|
| 13 |
+
does.
|
| 14 |
+
|
| 15 |
+
WHAT THIS MODEL IS
|
| 16 |
+
This is a research model, trained on the data and evaluated by the methods described in its
|
| 17 |
+
model card. It is not a certified product and is not a substitute for a qualified translator,
|
| 18 |
+
transcriber, clinician or other professional.
|
| 19 |
+
|
| 20 |
+
AUTOMATIC PROCESSING MAKES MISTAKES
|
| 21 |
+
Output from this model, whether a translation, a transcript, generated text or speech, or any
|
| 22 |
+
other result, can be wrong while appearing fluent and confident. Known failure modes include
|
| 23 |
+
omitted or inserted words, incorrect names, numbers, dates and negations, and reduced accuracy
|
| 24 |
+
on dialects, noisy audio or subjects outside the training data.
|
| 25 |
+
|
| 26 |
+
Prifysgol Bangor University therefore advises that output is treated as provisional and is
|
| 27 |
+
checked and corrected by a competent person before it is relied on as a record, used in any
|
| 28 |
+
decision affecting a person, or shared or published. Where the model is used interactively or in
|
| 29 |
+
real time (for example live captions, dictation, suggestions in an editor or read-aloud), users
|
| 30 |
+
should be told that what they see or hear is automatic and may be wrong. The University asks
|
| 31 |
+
anyone who builds on or redistributes this model to pass this warning on to their own users.
|
| 32 |
+
|
| 33 |
+
NOT FIT FOR
|
| 34 |
+
The model has not been evaluated for, and should not be used in, medical, legal, financial,
|
| 35 |
+
safety-critical or emergency settings, or in any decision affecting a person's rights, health,
|
| 36 |
+
liberty or livelihood, unless a qualified person reviews every output.
|
| 37 |
+
|
| 38 |
+
USE IN WALES: EQUAL TREATMENT OF WELSH AND ENGLISH
|
| 39 |
+
This model was developed to help organisations provide services in Welsh that are as good as
|
| 40 |
+
those they provide in English, not to replace Welsh-language provision. Under section 1 of the
|
| 41 |
+
Welsh Language (Wales) Measure 2011 (2011 nawm 1), Welsh has official status in Wales, and that
|
| 42 |
+
status is given legal effect by, among other things, the treatment of Welsh no less favourably
|
| 43 |
+
than English (s. 1(2)(b)) and by standards of conduct relating to the use of Welsh (s. 1(3)(e)).
|
| 44 |
+
These standards are made and enforced under Part 4 of the Measure. Bodies still operating a
|
| 45 |
+
Welsh language scheme under Part II of the Welsh Language Act 1993 (c. 38) must give effect to
|
| 46 |
+
the principle in section 5 of that Act that, in the conduct of public business and the
|
| 47 |
+
administration of justice in Wales, Welsh and English should be treated on a basis of equality.
|
| 48 |
+
Prifysgol Bangor University asks anyone deploying this model in Wales to consider whether the
|
| 49 |
+
intended use meets those requirements. Offering automated Welsh where English is served by
|
| 50 |
+
people, publishing machine translation into Welsh without the review given to English text, or
|
| 51 |
+
otherwise using this technology to replace rather than support Welsh-language provision is
|
| 52 |
+
contrary to the purpose for which the model was released. Which duties apply depends on the body
|
| 53 |
+
and the regulations or scheme that cover it, and responsibility for compliance rests with the
|
| 54 |
+
body deploying the model. See the Welsh Language Commissioner's regulatory policy statement on
|
| 55 |
+
artificial intelligence and the Welsh language, Technology, Websites and Software: Welsh
|
| 56 |
+
Language Considerations, and the advice document Bilingual drafting, translation and using Welsh
|
| 57 |
+
face to face, all linked from the model card. The legislation is at
|
| 58 |
+
legislation.gov.uk/mwa/2011/1 and legislation.gov.uk/ukpga/1993/38.
|
| 59 |
+
|
| 60 |
+
FURTHER INFORMATION AND ERROR REPORTS
|
| 61 |
+
Training data, evaluation results and known limitations: https://huggingface.co/techiaith/kaldi-cy
|
| 62 |
+
Report errors or concerns to: techiaith@bangor.ac.uk
|
| 63 |
+
|
| 64 |
+
PRONUNCIATION DICTIONARY
|
| 65 |
+
The decoding graph (HCLG.fst) and word list (words.txt) in this bundle were built with
|
| 66 |
+
pronunciations from Geiriadur Ynganu Bangor, the Bangor Pronouncing Dictionary,
|
| 67 |
+
https://github.com/techiaith/geiriadur-ynganu-bangor, redistributed here under its own
|
| 68 |
+
licence, which requires the following to be kept with any binary redistribution:
|
| 69 |
+
|
| 70 |
+
Copyright (C) 2025 Prifysgol Bangor University. All rights reserved.
|
| 71 |
+
|
| 72 |
+
Redistribution and use in source and binary forms, with or without modification, are
|
| 73 |
+
permitted provided that the following conditions are met:
|
| 74 |
+
|
| 75 |
+
1. Redistributions of source code must retain the above copyright notice, this list of
|
| 76 |
+
conditions and the following disclaimer. The contents of this file are deemed to be
|
| 77 |
+
source code.
|
| 78 |
+
|
| 79 |
+
2. Redistributions in binary form must reproduce the above copyright notice, this list
|
| 80 |
+
of conditions and the following disclaimer in the documentation and/or other
|
| 81 |
+
materials provided with the distribution.
|
| 82 |
+
|
| 83 |
+
THIS SOFTWARE IS PROVIDED BY PRIFYSGOL BANGOR UNIVERSITY ``AS IS'' AND ANY EXPRESSED OR
|
| 84 |
+
IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
|
| 85 |
+
MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL
|
| 86 |
+
PRIFYSGOL BANGOR UNIVERSITY NOR ITS EMPLOYEES BE LIABLE FOR ANY DIRECT, INDIRECT,
|
| 87 |
+
INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
|
| 88 |
+
PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
|
| 89 |
+
INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
|
| 90 |
+
LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
|
| 91 |
+
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
| 92 |
+
|
| 93 |
+
BUILT WITH
|
| 94 |
+
Trained with Kaldi (Apache License 2.0) and pocolm (Apache License 2.0), and packaged
|
| 95 |
+
for Vosk (Apache License 2.0), following the vosk-cymraeg recipe by Gweltaz
|
| 96 |
+
Duval-Guennoc, Preben Vangberg and Sasha Wanasky (MIT License, Cymru-Breizh Agile Cymru
|
| 97 |
+
Project, https://github.com/Cymru-Breizh-Agile-Cymru-Project/vosk-cymraeg). None of
|
| 98 |
+
their code is included in this bundle.
|
| 99 |
+
|
| 100 |
+
TRAINING DATA ATTRIBUTIONS
|
| 101 |
+
The Apache License applies to the model weights, code and documentation. It does not license the
|
| 102 |
+
data the model was trained on, which is not distributed with the model and remains under the
|
| 103 |
+
terms of its original providers, acknowledged here as those terms require.
|
| 104 |
+
|
| 105 |
+
Dedicated to the public domain under CC0 1.0 Universal
|
| 106 |
+
(https://creativecommons.org/publicdomain/zero/1.0/). CC0 asks for no attribution; the sources
|
| 107 |
+
are listed so that the training data can be traced to its datasheet.
|
| 108 |
+
Banc Trawsgrifiadau Bangor (CC0 1.0), collected by Language Technologies Unit, Bangor
|
| 109 |
+
University. Collection process, consent terms and contact:
|
| 110 |
+
https://huggingface.co/datasets/techiaith/banc-trawsgrifiadau-bangor
|
| 111 |
+
Corpws CLlLC / WLGA (CC0 1.0), collected by Language Technologies Unit, Bangor University.
|
| 112 |
+
Collection process, consent terms and contact:
|
| 113 |
+
https://huggingface.co/datasets/techiaith/corpws-clllc-wlga
|
| 114 |
+
Mozilla Common Voice 23.0, Welsh (CC0 1.0), collected by Mozilla Foundation and the volunteers
|
| 115 |
+
who contributed their voices. Collection process, consent terms and contact:
|
| 116 |
+
https://commonvoice.mozilla.org/cy/datasets. Obtained from Mozilla under their terms for
|
| 117 |
+
Common Voice data and held in a private working copy; it is not redistributed with or by this
|
| 118 |
+
model.
|
| 119 |
+
Lleisiau ARFOR (CC0 1.0), collected by Cymen, with the Language Technologies Unit, Bangor
|
| 120 |
+
University. Collection process, consent terms and contact:
|
| 121 |
+
https://huggingface.co/datasets/cymen-arfor/lleisiau-arfor
|
| 122 |
+
Mozilla Common Voice, Welsh, re-segmented with voice activity detection (CC0 1.0), collected by
|
| 123 |
+
Mozilla Foundation and the volunteers who contributed their voices. Collection process,
|
| 124 |
+
consent terms and contact: https://commonvoice.mozilla.org/cy/datasets. Derived from Common
|
| 125 |
+
Voice data obtained from Mozilla under their terms and held in a private working copy; it is
|
| 126 |
+
not redistributed with or by this model.
|
README.md
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
---
|
| 2 |
-
license:
|
| 3 |
language:
|
| 4 |
- cy
|
| 5 |
pipeline_tag: automatic-speech-recognition
|
|
@@ -13,10 +13,9 @@ datasets:
|
|
| 13 |
- techiaith/banc-trawsgrifiadau-bangor
|
| 14 |
- techiaith/corpws-clllc-wlga
|
| 15 |
- cymen-arfor/lleisiau-arfor
|
| 16 |
-
- techiaith/commonvoice_vad_cy
|
| 17 |
---
|
| 18 |
|
| 19 |
-
# Kaldi / Vosk Welsh Speech Recognition —
|
| 20 |
|
| 21 |
A Kaldi chain (LF-MMI) acoustic model with WFST decoding, packaged for
|
| 22 |
[Vosk](https://alphacephei.com/vosk/). It runs in real time on CPU, which makes it
|
|
@@ -24,32 +23,55 @@ suited to on-device and streaming use where an end-to-end model is too costly.
|
|
| 24 |
|
| 25 |
Welsh transcription only. Output carries no punctuation or casing.
|
| 26 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 27 |
## Evaluation results
|
| 28 |
|
| 29 |
-
WER / CER (lower is better) on the held-out
|
| 30 |
-
|
|
|
|
|
|
|
| 31 |
|
| 32 |
-
Mean across the Welsh test sets: **WER 47.
|
| 33 |
|
| 34 |
-
| Test set | WER | CER | WER (
|
| 35 |
|----------|-----|-----|------------|
|
| 36 |
-
| `cymen-arfor/lleisiau-arfor` (spontaneous, mixed dialects) | 55.
|
| 37 |
-
| `
|
| 38 |
-
| `techiaith/
|
|
|
|
| 39 |
|
| 40 |
## Training data
|
| 41 |
|
| 42 |
Welsh speech corpora used to train this model:
|
| 43 |
|
| 44 |
-
| Dataset | Split | Revision | Access |
|
| 45 |
-
|---------|-------|----------|--------|
|
| 46 |
-
| [techiaith/banc-trawsgrifiadau-bangor](https://huggingface.co/datasets/techiaith/banc-trawsgrifiadau-bangor/tree/26.07) | `train` | `26.07` | public |
|
| 47 |
-
| [techiaith/corpws-clllc-wlga](https://huggingface.co/datasets/techiaith/corpws-clllc-wlga/tree/5d00294c31c78b1d7937bb2c2bc6cc70bc18d410) | `clips` | `5d00294c31` | public |
|
| 48 |
-
| `techiaith/commonvoice_23_0_cy` | `train+dev[50%:]` | `main` | private † |
|
| 49 |
-
| [cymen-arfor/lleisiau-arfor](https://huggingface.co/datasets/cymen-arfor/lleisiau-arfor) | `train` | `main` | public |
|
| 50 |
-
|
|
| 51 |
|
| 52 |
-
† **Common Voice data is not redistributed here.** `techiaith/commonvoice_23_0_cy` (Common Voice 23.0 Welsh)
|
| 53 |
|
| 54 |
The Welsh Common Voice corpus is freely available from Mozilla themselves: [Mozilla Common Voice — Welsh datasets](https://commonvoice.mozilla.org/cy/datasets). Releases are distributed through the Mozilla Data Collective, and the [`datacollective`](https://pypi.org/project/datacollective/) package offers a programmatic download. With that in hand the rest of this pipeline is reproducible — every other dataset in the table is openly available on the Hugging Face Hub. Our thanks to Mozilla and to the Welsh speakers who contributed their voices.
|
| 55 |
|
|
@@ -74,10 +96,10 @@ Welsh transcription audio only, as materialised for training:
|
|
| 74 |
| Epochs | 5 |
|
| 75 |
| Features | 40-dim MFCC (hires) + 100-dim online i-vectors |
|
| 76 |
| SpecAugment | yes (spec-augment-layer in the xconfig) |
|
| 77 |
-
| Language model | 3-gram
|
| 78 |
| Lexicon | Geiriadur Ynganu Bangor pronunciations, with grapheme-to-phoneme fallback |
|
| 79 |
| Decoding vocabulary | 62,392 words |
|
| 80 |
-
| Held-out dev WER | 30.
|
| 81 |
|
| 82 |
## Usage
|
| 83 |
|
|
@@ -117,15 +139,58 @@ returns `True`, as above. A single `AcceptWaveform` call followed by only
|
|
| 117 |
`FinalResult()` discards everything recognised before the last detected
|
| 118 |
end-of-utterance, which silently truncates anything longer than one utterance.
|
| 119 |
|
| 120 |
-
##
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 121 |
|
| 122 |
- **Welsh only.** English audio is not a supported task.
|
| 123 |
- **Closed vocabulary.** WFST decoding can only emit words present in the
|
| 124 |
lexicon (62,392 words), which is built from the training transcripts. Names and
|
| 125 |
-
loanwords outside it cannot be produced at all
|
|
|
|
| 126 |
- **No punctuation or casing.**
|
| 127 |
- The language model is a 3-gram trained on the training transcripts only, so
|
| 128 |
-
the model is weaker on domains and registers
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 129 |
|
| 130 |
## Acknowledgements
|
| 131 |
|
|
|
|
| 1 |
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
language:
|
| 4 |
- cy
|
| 5 |
pipeline_tag: automatic-speech-recognition
|
|
|
|
| 13 |
- techiaith/banc-trawsgrifiadau-bangor
|
| 14 |
- techiaith/corpws-clllc-wlga
|
| 15 |
- cymen-arfor/lleisiau-arfor
|
|
|
|
| 16 |
---
|
| 17 |
|
| 18 |
+
# Kaldi / Vosk Welsh Speech Recognition — 2610
|
| 19 |
|
| 20 |
A Kaldi chain (LF-MMI) acoustic model with WFST decoding, packaged for
|
| 21 |
[Vosk](https://alphacephei.com/vosk/). It runs in real time on CPU, which makes it
|
|
|
|
| 23 |
|
| 24 |
Welsh transcription only. Output carries no punctuation or casing.
|
| 25 |
|
| 26 |
+
## What changed in this release
|
| 27 |
+
|
| 28 |
+
**The language model is estimated with [pocolm](https://github.com/danpovey/pocolm)
|
| 29 |
+
instead of SRILM.** The training data, recipe, lexicon and acoustic model are those of
|
| 30 |
+
the 2607 release; only the 3-gram language model was re-estimated, and the decoding
|
| 31 |
+
graph rebuilt from it. This has two consequences.
|
| 32 |
+
|
| 33 |
+
- **Licence.** SRILM's Research Community License permits non-commercial research use
|
| 34 |
+
only, so every earlier Kaldi/Vosk release from the Unit (2503, 2607) was published
|
| 35 |
+
under CC-BY-NC. pocolm is Apache-2.0, as are Kaldi and Vosk, and the training speech
|
| 36 |
+
is all CC0, so nothing in this model's toolchain or data restricts its use. This is
|
| 37 |
+
the first Kaldi/Vosk model from the Unit released under the Apache License 2.0 (see
|
| 38 |
+
[Licence](#licence) below), and it may be used commercially.
|
| 39 |
+
- **Accuracy.** pocolm tunes its discounting against the held-out dev transcripts
|
| 40 |
+
instead of applying fixed Witten-Bell discounting, and keeps the n-grams that SRILM's
|
| 41 |
+
default count cut-offs discard. Held-out dev WER falls from 30.91% to 30.05%, and the
|
| 42 |
+
benchmark improves on every Welsh test set (table below). The larger language model
|
| 43 |
+
makes the decoding graph (`HCLG.fst`) three times the size, 242 MB against 82 MB, so
|
| 44 |
+
the bundle grows from 70 MB to 121 MB and decoding needs correspondingly more memory.
|
| 45 |
+
|
| 46 |
## Evaluation results
|
| 47 |
|
| 48 |
+
WER / CER (lower is better) on the held-out test sets, scored per utterance against the
|
| 49 |
+
same references for every model compared.
|
| 50 |
+
|
| 51 |
+
The references are the test sets' transcripts as published, with their capitalisation and punctuation intact; only transcription-convention markers (such as `<chwerthin>`) are removed before scoring. This model emits lowercase text without punctuation, so every reference word that carries a capital letter or attached punctuation is scored as an error. The figures are therefore conservative, and higher than a normalised comparison would give, but every model in the table is scored the same way, so the comparison between them holds.
|
| 52 |
|
| 53 |
+
Mean across the Welsh test sets: **WER 47.30**, against **48.20** for the previous release (2607-SRILM) — an improvement of **0.90** points.
|
| 54 |
|
| 55 |
+
| Test set | WER | CER | WER (2607-SRILM) |
|
| 56 |
|----------|-----|-----|------------|
|
| 57 |
+
| `cymen-arfor/lleisiau-arfor` (spontaneous, mixed dialects) | 55.10 | 27.55 | 55.99 |
|
| 58 |
+
| `google/fleurs/cy-gb` | 49.89 | 20.25 | 51.06 |
|
| 59 |
+
| `techiaith/banc-trawsgrifiadau-bangor` (spontaneous, mixed dialects) | 39.69 | 17.25 | 40.23 |
|
| 60 |
+
| `techiaith/commonvoice-23-0-cy` (read) | 44.54 | 15.48 | 45.52 |
|
| 61 |
|
| 62 |
## Training data
|
| 63 |
|
| 64 |
Welsh speech corpora used to train this model:
|
| 65 |
|
| 66 |
+
| Dataset | Split | Revision | Licence | Access | Datasheet |
|
| 67 |
+
|---------|-------|----------|---------|--------|-----------|
|
| 68 |
+
| [techiaith/banc-trawsgrifiadau-bangor](https://huggingface.co/datasets/techiaith/banc-trawsgrifiadau-bangor/tree/26.07) | `train` | `26.07` | [CC0 1.0](https://creativecommons.org/publicdomain/zero/1.0/) | public | [datasheet](https://huggingface.co/datasets/techiaith/banc-trawsgrifiadau-bangor) |
|
| 69 |
+
| [techiaith/corpws-clllc-wlga](https://huggingface.co/datasets/techiaith/corpws-clllc-wlga/tree/5d00294c31c78b1d7937bb2c2bc6cc70bc18d410) | `clips` | `5d00294c31` | [CC0 1.0](https://creativecommons.org/publicdomain/zero/1.0/) | public | [datasheet](https://huggingface.co/datasets/techiaith/corpws-clllc-wlga) |
|
| 70 |
+
| `techiaith/commonvoice_23_0_cy` | `train+dev[50%:]` | `main` | [CC0 1.0](https://creativecommons.org/publicdomain/zero/1.0/) | private † | [datasheet](https://commonvoice.mozilla.org/cy/datasets) |
|
| 71 |
+
| [cymen-arfor/lleisiau-arfor](https://huggingface.co/datasets/cymen-arfor/lleisiau-arfor) | `train` | `main` | [CC0 1.0](https://creativecommons.org/publicdomain/zero/1.0/) | public | [datasheet](https://huggingface.co/datasets/cymen-arfor/lleisiau-arfor) |
|
| 72 |
+
| `techiaith/commonvoice_vad_cy` | `train` | `main` | [CC0 1.0](https://creativecommons.org/publicdomain/zero/1.0/) | private † | [datasheet](https://commonvoice.mozilla.org/cy/datasets) |
|
| 73 |
|
| 74 |
+
† **Common Voice data is not redistributed here.** `techiaith/commonvoice_23_0_cy` (Common Voice 23.0 Welsh), `techiaith/commonvoice_vad_cy` (Common Voice Welsh, re-segmented with voice activity detection) are held in private repositories, in line with the Mozilla Foundation's terms for obtaining and sharing Common Voice data. They are listed above for provenance and reproducibility, not as a download.
|
| 75 |
|
| 76 |
The Welsh Common Voice corpus is freely available from Mozilla themselves: [Mozilla Common Voice — Welsh datasets](https://commonvoice.mozilla.org/cy/datasets). Releases are distributed through the Mozilla Data Collective, and the [`datacollective`](https://pypi.org/project/datacollective/) package offers a programmatic download. With that in hand the rest of this pipeline is reproducible — every other dataset in the table is openly available on the Hugging Face Hub. Our thanks to Mozilla and to the Welsh speakers who contributed their voices.
|
| 77 |
|
|
|
|
| 96 |
| Epochs | 5 |
|
| 97 |
| Features | 40-dim MFCC (hires) + 100-dim online i-vectors |
|
| 98 |
| SpecAugment | yes (spec-augment-layer in the xconfig) |
|
| 99 |
+
| Language model | 3-gram estimated with pocolm on the training transcripts only, discounting tuned on the dev transcripts (dev perplexity 292.8) |
|
| 100 |
| Lexicon | Geiriadur Ynganu Bangor pronunciations, with grapheme-to-phoneme fallback |
|
| 101 |
| Decoding vocabulary | 62,392 words |
|
| 102 |
+
| Held-out dev WER | 30.05% |
|
| 103 |
|
| 104 |
## Usage
|
| 105 |
|
|
|
|
| 139 |
`FinalResult()` discards everything recognised before the last detected
|
| 140 |
end-of-utterance, which silently truncates anything longer than one utterance.
|
| 141 |
|
| 142 |
+
## Intended use and limitations
|
| 143 |
+
|
| 144 |
+
The model transcribes Welsh speech to text. It is a research model, not a certified product,
|
| 145 |
+
and is not a substitute for a qualified transcriber.
|
| 146 |
+
|
| 147 |
+
**Automatic transcription makes mistakes.** Output can be wrong while appearing fluent and
|
| 148 |
+
confident. Known failure modes include omitted, inserted or substituted words; incorrect names, numbers and place names; words outside its fixed vocabulary replaced by similar-sounding ones; and reduced accuracy on dialect, overlapping speakers, noisy or distant audio, and subjects outside the training data. Output carries no punctuation or capitalisation. Treat output as provisional and have it
|
| 149 |
+
checked and corrected by a competent person before it is relied on as a record, used in any
|
| 150 |
+
decision affecting a person, or shared or published. Where the model is used interactively or
|
| 151 |
+
in real time (live captions, dictation), tell users that what they see is an automatic
|
| 152 |
+
transcription and may be wrong.
|
| 153 |
+
|
| 154 |
+
**Not fit for** medical, legal, financial, safety-critical or emergency settings, or any
|
| 155 |
+
decision affecting a person's rights, health, liberty or livelihood, unless a qualified person
|
| 156 |
+
reviews every output.
|
| 157 |
+
|
| 158 |
+
**Use in Wales.** The model was developed to help organisations provide services in Welsh that
|
| 159 |
+
are as good as those they provide in English, not to replace Welsh-language provision. Under
|
| 160 |
+
section 1 of the Welsh Language (Wales) Measure 2011 (2011 nawm 1), Welsh has official status in
|
| 161 |
+
Wales, and that status is given legal effect by, among other things, the treatment of Welsh no
|
| 162 |
+
less favourably than English (s. 1(2)(b)) and by standards of conduct relating to the use of
|
| 163 |
+
Welsh (s. 1(3)(e)). These standards are made and enforced under Part 4 of the Measure. Bodies
|
| 164 |
+
still operating a Welsh language scheme under Part II of the Welsh Language Act 1993 (c. 38)
|
| 165 |
+
must give effect to the principle in section 5 of that Act that, in the conduct of public
|
| 166 |
+
business and the administration of justice in Wales, Welsh and English should be treated on a
|
| 167 |
+
basis of equality. Which duties apply depends on the body and the regulations or scheme that
|
| 168 |
+
cover it, and responsibility for compliance rests with the body deploying the model. See the Welsh Language Commissioner's [regulatory policy statement on artificial intelligence and the Welsh language](https://www.welshlanguagecommissioner.wales/regulation/promoting-compliance/artificial-intelligence-and-the-welsh-language-the-welsh-language-commissioners-regulatory-policy-statement), [Technology, Websites and Software: Welsh Language Considerations](https://www.welshlanguagecommissioner.wales/media/lxechbz0/technoleg-gwefannau-a-meddalwedd-technology-websites-and-software.pdf), and the advice document [Bilingual drafting, translation and using Welsh face to face](https://www.welshlanguagecommissioner.wales/support-resources/translation).
|
| 169 |
+
The legislation is at [legislation.gov.uk/mwa/2011/1](https://www.legislation.gov.uk/mwa/2011/1)
|
| 170 |
+
and [legislation.gov.uk/ukpga/1993/38](https://www.legislation.gov.uk/ukpga/1993/38).
|
| 171 |
+
|
| 172 |
+
Further limitations:
|
| 173 |
|
| 174 |
- **Welsh only.** English audio is not a supported task.
|
| 175 |
- **Closed vocabulary.** WFST decoding can only emit words present in the
|
| 176 |
lexicon (62,392 words), which is built from the training transcripts. Names and
|
| 177 |
+
loanwords outside it cannot be produced at all and this represents a hard floor on
|
| 178 |
+
achievable WER.
|
| 179 |
- **No punctuation or casing.**
|
| 180 |
- The language model is a 3-gram trained on the training transcripts only, so
|
| 181 |
+
the model is weaker on domains and registers that differ from its training data.
|
| 182 |
+
|
| 183 |
+
## Licence
|
| 184 |
+
|
| 185 |
+
Released under the [Apache License 2.0](LICENSE). The accompanying [NOTICE](NOTICE) file
|
| 186 |
+
carries the University's attribution, the warning that automatic transcription makes
|
| 187 |
+
mistakes, and the attributions required by the training-data and pronunciation-dictionary
|
| 188 |
+
licences. Section 4 of the licence requires anyone who redistributes the model or a
|
| 189 |
+
derivative to keep that NOTICE. Both files are also inside `model_cy.tar.gz`, in the `model/`
|
| 190 |
+
directory, so they travel with the unpacked model.
|
| 191 |
+
|
| 192 |
+
One-line attribution: *Includes kaldi-cy 2610, developed by the Language Technologies Unit,
|
| 193 |
+
Bangor University, and released under the Apache License 2.0.*
|
| 194 |
|
| 195 |
## Acknowledgements
|
| 196 |
|
model_cy.tar.gz
CHANGED
|
@@ -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:ddb8ea827262db10191f486a9a021687d95106b9e1c018aa45165ac5ab284919
|
| 3 |
+
size 120503544
|