SeamlessM4T
https://gyazo.com/d5151b570c29ee640f5020ab4463933b
https://github.com/facebookresearch/seamless_communicationfacebookresearch/seamless_communication
https://seamless.metademolab.com/Demo
https://ai.meta.com/research/publications/seamless-m4t/SeamlessM4T—Massively Multilingual & Multimodal Machine Translation
https://huggingface.co/facebook/seamless-m4t-largeseamless-m4t-large
https://huggingface.co/facebook/seamless-m4t-mediumseamless-m4t-medium
https://huggingface.co/facebook/seamless-m4t-unity-smallseamless-m4t-unity-small
https://huggingface.co/facebook/seamless-m4t-unity-small-s2tseamless-m4t-unity-small-s2t
Demohttps://huggingface.co/spaces/facebook/seamless_m4t
speech2text✕text2speech✕翻訳
これを一つのモデルで行うマルチモーダルモデル
101言語の音声入力
96言語のテキスト入力/出力
35言語の音声出力
#Meta