# Copyright (c) Meta Platforms, Inc. and affiliates | |
# All rights reserved. | |
# | |
# This source code is licensed under the license found in the | |
# MIT_LICENSE file in the root directory of this source tree. | |
name: mintox | |
model_name: MinTox | |
etox_dataset: https://dl.fbaipublicfiles.com/nllb/NLLB-200_TWL/nllb-200_twl.zip | |
etox_lang_variants: | |
- kas_Arab | |
- kas_Deva | |
- knc_Arab | |
- knc_Latn | |
- min_Arab | |
- min_Latn | |
- zho_Hans | |
- zho_Hant | |
sp_model: https://huggingface.co/facebook/seamless-m4t-medium/resolve/main/tokenizer.model | |
# For some languages, we use the SentencePiece model. | |
sp_langs: | |
- asm | |
- ben | |
- cmn | |
- guj | |
- mya | |
- hin | |
- gom | |
- ibo | |
- jpn | |
- kan | |
- khm | |
- kor | |
- lao | |
- mai | |
- mal | |
- mar | |
- mni | |
- npi | |
- oan | |
- ory | |
- pan | |
- rwr | |
- sat | |
- tam | |
- tel | |
- tha | |
- wuu | |
- yue | |