|
Download THIRD_PARTY_NOTICES.md from KaleComputer/kale-tts-1-tiny: direct link, hf CLI and curl.
- Browser
- Download file 7.15 kB
-
https://huggingface.co/KaleComputer/kale-tts-1-tiny/resolve/main/THIRD_PARTY_NOTICES.md
- Command line
-
hf download hf://KaleComputer/kale-tts-1-tiny/THIRD_PARTY_NOTICES.md
-
curl -L -o THIRD_PARTY_NOTICES.md https://huggingface.co/KaleComputer/kale-tts-1-tiny/resolve/main/THIRD_PARTY_NOTICES.md
7.15 kB
| # Third-party notices | |
| kale-tts-1 (weights and code: Apache-2.0; dataset: CC BY 4.0) includes or is derived from the works below. Their licences and notices | |
| travel with every copy. The full Apache License 2.0 text is in `LICENSE` (and in `frontend/dizione/LICENSE`). | |
| ## Model weights | |
| - **Kokoro-82M** (hexgrad, huggingface.co/hexgrad/Kokoro-82M, Apache-2.0). The kale-tts-1 text side starts from Kokoro-82M's weights. | |
| Changes: fine-tuned on the kale-tts-1 voices, distilled to smaller text sides, joined to a new decoder, and exported to ONNX. | |
| The IPA-to-id table (`tokenizer.json`, `frontend/matcha/ipa2kokoro.json`) is Kokoro's. | |
| - **Maya-1** (Maya Research, huggingface.co/maya-research/maya1, Apache-2.0). All training audio is original Maya-1 output from text | |
| voice descriptions (decoded with SNAC, MIT, Copyright (c) 2023 Descript, Copyright (c) 2024 Hubert Siuzdak). Based on Maya. Every voice | |
| is synthetic and is nobody real; follow the Maya-1 use policy (no impersonation, never present the audio as a human recording). | |
| - **StyleTTS2** (github.com/yl4579/StyleTTS2, MIT, Copyright (c) 2023 Aaron (Yinghao) Li) and **kikiri-tts** (github.com/semidark/kikiri-tts, | |
| Apache-2.0): the teacher's training code. | |
| - **Ettin encoder 17M** (jhu-clsp/ettin-encoder-17m, github.com/JHU-CLSP/ettin-encoder-vs-decoder, MIT): kale-tts-1-tiny's text side | |
| contains its weights (frozen) and ships its tokenizer as `ettin-tokenizer.json`; `frontend/ettin.js` reimplements the tokenizer. | |
| ``` | |
| MIT License | |
| Copyright (c) 2025 Johns Hopkins University Center for Language and Speech Processing | |
| Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the | |
| "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, | |
| distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to | |
| the following conditions: | |
| The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. | |
| THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF | |
| MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY | |
| CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE | |
| SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. | |
| ``` | |
| ## Text front end (`frontend/`) | |
| - **CMUdict** (github.com/cmusphinx/cmudict, BSD-2-Clause style). `frontend/matcha/lexicon.tsv`, the en-gb word list | |
| (`frontend/matcha/en-gb/lexicon.tsv`), the unknown-word guessers (`g2p.bin`) and the `kokoro-compatible/lexicon-us-en.txt` files are | |
| made from it. Its licence: | |
| ``` | |
| Copyright (C) 1993-2015 Carnegie Mellon University. All rights reserved. | |
| Redistribution and use in source and binary forms, with or without | |
| modification, are permitted provided that the following conditions | |
| are met: | |
| 1. Redistributions of source code must retain the above copyright | |
| notice, this list of conditions and the following disclaimer. | |
| The contents of this file are deemed to be source code. | |
| 2. Redistributions in binary form must reproduce the above copyright | |
| notice, this list of conditions and the following disclaimer in | |
| the documentation and/or other materials provided with the | |
| distribution. | |
| This work was supported in part by funding from the Defense Advanced | |
| Research Projects Agency, the Office of Naval Research and the National | |
| Science Foundation of the United States of America, and by member | |
| companies of the Carnegie Mellon Sphinx Speech Consortium. We acknowledge | |
| the contributions of many volunteers to the expansion and improvement of | |
| this dictionary. | |
| THIS SOFTWARE IS PROVIDED BY CARNEGIE MELLON UNIVERSITY ``AS IS'' AND | |
| ANY EXPRESSED OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, | |
| THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR | |
| PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL CARNEGIE MELLON UNIVERSITY | |
| NOR ITS EMPLOYEES BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, | |
| SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT | |
| LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, | |
| DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY | |
| THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT | |
| (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE | |
| OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. | |
| ``` | |
| - **Britfone 3.0.1** (github.com/JoseLlarena/Britfone, MIT). The en-gb word list (`frontend/matcha/en-gb/lexicon.tsv`) uses it. | |
| ``` | |
| MIT License | |
| Copyright (c) 2017 Jose Llarena | |
| Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the | |
| "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, | |
| distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to | |
| the following conditions: | |
| The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. | |
| THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF | |
| MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY | |
| CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE | |
| SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. | |
| ``` | |
| - **dizione** (github.com/n4ze3m/dizione, Apache-2.0, `frontend/dizione/LICENSE`): the number, date, time, unit and abbreviation rules in | |
| `frontend/dizione/` are a port of its text rules. The homograph tables (`frontend/dizione/homographs.data.js`) come from | |
| github.com/n4ze3m/tiny-voca (Apache-2.0), trained on the Wikipedia Homograph Data (github.com/google-research-datasets/WikipediaHomographData, | |
| Apache-2.0; Gorman, Mazovetskiy and Nikolaev, LREC 2018); the tables hold word lists and weights only, no Wikipedia text. | |
| ## Training recipe (`recipe/`) | |
| - **Paradee** (github.com/sahilmahendrakar/paradee @ 2de547e45da92840a636c499eb0782e988be1e47, Apache-2.0): `vhead.py`, | |
| `student_decoder.py` and `discriminators.py`, copied unchanged into `recipe/train/third_party/paradee/` with its LICENSE and NOTICE. | |
| No Paradee weights are used. | |
| - StyleTTS2 / AuxiliaryASR (MIT, Copyright (c) 2022-2023 Aaron (Yinghao) Li) and kikiri-tts (Apache-2.0): cloned at pinned revisions by | |
| `recipe/teacher/`; one class copied from AuxiliaryASR is marked in `recipe/teacher/teacher.py`. | |
| ## Browser demo (`space/`) | |
| - **onnxruntime-web** (github.com/microsoft/onnxruntime, MIT, Copyright (c) Microsoft Corporation): `space/ort/`, with its `LICENSE`. | |