This release is a pre-release and may not be stable for production use.
OVOS Wav2Vec2 STT
Description
This is an OVOS speech-to-text (STT) plugin for Wav2Vec2 models. It loads a Wav2Vec2 model from Hugging Face and transcribes audio with it. If you do not set a model, the plugin picks one for you based on the language.
Install
pip install ovos-stt-plugin-wav2vec
Configuration
"stt": {
"module": "ovos-stt-plugin-wav2vec",
"ovos-stt-plugin-wav2vec": {
"model": "proxectonos/Nos_ASR-wav2vec2-large-xlsr-53-gl-with-lm"
}
}
model can be any compatible Wav2Vec2 model from Hugging Face. If you do not set it, the plugin picks a model based on lang.
Models
If your language is not in the list below, use "facebook/mms-1b-all" as the model. In that case, use the dedicated ovos-stt-plugin-mms plugin instead.
Supported languages: ab, ar, as, ba, bas, bg, bn, br, ca, cnh, cv, cy,
cz, da, de, el, en, eo, es, fa, fi, fr, ga, gl, gn, ha, he,
hi, hk, hu, hy, ia, id, it, jp, jv, ka, ko, ku, ky, lb, lg,
lt, ml, mn, mr, mt, ne, nl, or, pa, pl, pt, rm-sursilv, rm-vallader,
ro, ru, sah, sk, sl, su, sv, sw, ta, te, tr, tt, uk, ur, vi, zh
The plugin maps each language to a default model:
LANG2MODEL = {
"gl": "proxectonos/Nos_ASR-wav2vec2-large-xlsr-53-gl-with-lm",
"pt": "jonatasgrosman/wav2vec2-large-xlsr-53-portuguese",
"en": "jonatasgrosman/wav2vec2-large-xlsr-53-english",
"zh": "jonatasgrosman/wav2vec2-large-xlsr-53-chinese-zh-cn",
"ru": "jonatasgrosman/wav2vec2-large-xlsr-53-russian",
"it": "jonatasgrosman/wav2vec2-large-xlsr-53-italian",
"es": "jonatasgrosman/wav2vec2-large-xlsr-53-spanish",
"fr": "jonatasgrosman/wav2vec2-large-xlsr-53-french",
"de": "jonatasgrosman/wav2vec2-large-xlsr-53-german",
"nl": "jonatasgrosman/wav2vec2-large-xlsr-53-dutch",
"jp": "jonatasgrosman/wav2vec2-large-xlsr-53-japanese",
"pl": "jonatasgrosman/wav2vec2-large-xlsr-53-polish",
"hu": "jonatasgrosman/wav2vec2-large-xlsr-53-hungarian",
"ar": "jonatasgrosman/wav2vec2-large-xlsr-53-arabic",
"fi": "jonatasgrosman/wav2vec2-large-xlsr-53-finnish",
"fa": "jonatasgrosman/wav2vec2-large-xlsr-53-persian",
"el": "jonatasgrosman/wav2vec2-large-xlsr-53-greek",
"ia": "infinitejoy/wav2vec2-large-xls-r-300m-interlingua",
"cnh": "infinitejoy/wav2vec2-large-xls-r-300m-hakha-chin",
"ta": "infinitejoy/wav2vec2-large-xls-r-300m-tamil-cv8",
"or": "infinitejoy/wav2vec2-large-xls-r-300m-odia",
"gn": "infinitejoy/wav2vec2-large-xls-r-300m-guarani",
"cy": "infinitejoy/wav2vec2-large-xls-r-300m-welsh",
"ur": "infinitejoy/wav2vec2-large-xls-r-300m-urdu",
"ab": "infinitejoy/wav2vec2-large-xls-r-300m-abkhaz",
"ba": "infinitejoy/wav2vec2-large-xls-r-300m-bashkir",
"bas": "infinitejoy/wav2vec2-large-xls-r-300m-basaa",
"sah": "infinitejoy/wav2vec2-large-xls-r-300m-sakha",
"mr": "infinitejoy/wav2vec2-large-xls-r-300m-marathi-cv8",
"lt": "infinitejoy/wav2vec2-large-xls-r-300m-lithuanian",
"ha": "infinitejoy/wav2vec2-large-xls-r-300m-hausa",
"br": "infinitejoy/wav2vec2-large-xls-r-300m-breton",
"cv": "infinitejoy/wav2vec2-large-xls-r-300m-chuvash",
"hy": "infinitejoy/wav2vec2-large-xls-r-300m-armenian",
"mt": "infinitejoy/wav2vec2-large-xls-r-300m-maltese",
"as": "infinitejoy/wav2vec2-large-xls-r-300m-assamese",
"tt": "infinitejoy/wav2vec2-large-xls-r-300m-tatar",
"ky": "infinitejoy/wav2vec2-large-xls-r-300m-kyrgyz",
"ga": "infinitejoy/wav2vec2-large-xls-r-300m-irish",
"ka": "infinitejoy/wav2vec2-large-xls-r-300m-georgian",
"sk": "infinitejoy/wav2vec2-large-xls-r-300m-slovak",
"sl": "infinitejoy/wav2vec2-large-xls-r-300m-slovenian",
"mn": "infinitejoy/wav2vec2-large-xls-r-300m-mongolian",
"ro": "infinitejoy/wav2vec2-large-xls-r-300m-romanian",
"ku": "infinitejoy/wav2vec2-large-xls-r-300m-kurdish",
"bg": "infinitejoy/wav2vec2-large-xls-r-300m-bulgarian",
"hi": "infinitejoy/wav2vec2-large-xls-r-300m-hindi",
"rm-vallader": "infinitejoy/wav2vec2-large-xls-r-300m-romansh-vallader",
"rm-sursilv": "infinitejoy/wav2vec2-large-xls-r-300m-romansh-sursilvan",
"id": "indonesian-nlp/wav2vec2-large-xlsr-indonesian",
"lg": "indonesian-nlp/wav2vec2-luganda",
"jv": "indonesian-nlp/wav2vec2-indonesian-javanese-sundanese",
"su": "indonesian-nlp/wav2vec2-indonesian-javanese-sundanese",
"da": "vachonni/wav2vec2-large-xls-r-300m-dansk-CV-80",
"ml": "Bluecast/wav2vec2-Malayalam",
"hk": "voidful/wav2vec2-large-xlsr-53-hk",
"ko": "kresnik/wav2vec2-large-xlsr-korean",
"vi": "nguyenvulebinh/wav2vec2-base-vietnamese-250h",
"bn": "arijitx/wav2vec2-large-xlsr-bengali",
"eo": "cpierse/wav2vec2-large-xlsr-53-esperanto",
"lb": "Lemswasabi/wav2vec2-large-xlsr-53-842h-luxembourgish-14h-with-lm",
"te": "anuragshas/wav2vec2-large-xlsr-53-telugu",
"tr": "mpoyraz/wav2vec2-xls-r-300m-cv7-turkish",
"ne": "gagan3012/wav2vec2-xlsr-nepali",
"he": "imvladikon/wav2vec2-large-xlsr-53-hebrew",
"sv": "KBLab/wav2vec2-large-xlsr-53-swedish",
"ca": "PereLluis13/wav2vec2-xls-r-1b-ca-lm",
"cz": "arampacha/wav2vec2-large-xlsr-czech",
"sw": "alokmatta/wav2vec2-large-xlsr-53-sw",
"uk": "arampacha/wav2vec2-xls-r-1b-uk",
"pa": "kingabzpro/wav2vec2-large-xlsr-53-punjabi",
}
Related projects
- OpenVoiceOS/ovos-stt-plugin-mms — STT plugin for
facebook/mms-1b-all, for languages not in this plugin's list - OpenVoiceOS/ovos-plugin-manager — loads and configures STT plugins for OVOS
Credits
TigreGotico built this plugin for OpenVoiceOS under the ILENIA project.
The Ministerio para la Transformación Digital y de la Función Pública and the Plan de Recuperación, Transformación y Resiliencia funded this plugin. Funded by EU, NextGenerationEU, within the framework of the project ILENIA, reference 2022/TL22/00215337.
O Proxecto Nós é un proxecto da Xunta de Galicia. A súa execución foi encomendada á Universidade de Santiago de Compostela, a través de dúas entidades de investigación en intelixencia artificial e tecnoloxías da linguaxe: o ILG (Instituto da Lingua Galega) e o CiTIUS (Centro Singular de Investigación en Tecnoloxías Intelixentes).
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file ovos_stt_plugin_wav2vec-0.3.3a8.tar.gz.
File metadata
- Download URL: ovos_stt_plugin_wav2vec-0.3.3a8.tar.gz
- Upload date:
- Size: 10.4 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via:
twine/7.0.0 CPython/3.13.14
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
7c0cdfccba11b2e54d966a83f9a79cf533f65aa62496d52abe572bb483d92a55
|
|
| MD5 |
0f1fdf42a030fed8d7b667153b90fd16
|
|
| BLAKE2b-256 |
334eb0326bf021b71975f77eec9dd668ebb1e054ff122843e5e9f2f7597e96f3
|
File details
Details for the file ovos_stt_plugin_wav2vec-0.3.3a8-py3-none-any.whl.
File metadata
- Download URL: ovos_stt_plugin_wav2vec-0.3.3a8-py3-none-any.whl
- Upload date:
- Size: 11.2 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via:
twine/7.0.0 CPython/3.13.14
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
ca923a90580033532feda8326f4faf42837e4fb47df6f93b9751d09f21dd0d18
|
|
| MD5 |
de56292630e3a950acfa91eae29f5396
|
|
| BLAKE2b-256 |
16a5c523c0161ca30ee869d2deb90188de823a74f374459f135197827230db50
|