21–30 of 232 posts
> The MMS code and model weights are released under the CC-BY-NC 4.0 license. Huge bummer. Prevents almost everyone from using this and recouping their costs. I suppose motivated teams could reproduce the paper in a clean room, but that might also be subject to patents.
Id imagine you could use inference from this as training for a commercial model, as that isn't currently protected under copyright.
Code: https://github.com/facebookresearch/fairseq/tree/main/exampl... Blog Post: https://ai.facebook.com/blog/multilingual-model-speech-recog... Paper: https://research.facebook.com/publications/scaling-speech-te... Languages coverage: https://dl.fbaipublicfiles.com/mms/misc/language_coverage_mm...
Based on the availability of STT, TTS, and translation models available to download in that github repo, a real-life babelfish is 'only' some glue code away. Wild times we're living in.
it is still hopeless, and much worse than the dictionary based, at gender/nums/dicleasions in general.
i sometimes use it just to not thing about the grammar in some languages, and most times I'm doing a surprised double take of something that would be completely inappropriate or offensive instead of my simple phrases.
Code: https://github.com/facebookresearch/fairseq/tree/main/exampl... Blog Post: https://ai.facebook.com/blog/multilingual-model-speech-recog... Paper: https://research.facebook.com/publications/scaling-speech-te... Languages coverage: https://dl.fbaipublicfiles.com/mms/misc/language_coverage_mm...
Based on the availability of STT, TTS, and translation models available to download in that github repo, a real-life babelfish is 'only' some glue code away. Wild times we're living in.
So many so-called overnight AI gurus hyping about their snake-oil product and screaming about 'Meta is dying' [0] and 'It is over for Meta' but little of them actually do research in AI and drive the field forward and this once again shows that Meta has always been a consistent contributor to AI research, especially in vision systems. All we can just do is take, take, take the code. But this time, the code's license…
I assume this "competes" directly with https://sites.research.google/usm/ -- would be cool to see side-by-side benchmarks sometime! Maybe I should make those. I requested access to USM but have not been granted any access yet.
I didn’t think so.
Earlier quoted context omitted.
Based on the availability of STT, TTS, and translation models available to download in that github repo, a real-life babelfish is 'only' some glue code away. Wild times we're living in.
hope that glue is better then the duct tapes google use on translate.g.c it is still hopeless, and much worse than the dictionary based, at gender/nums/dicleasions in general. i sometimes use it just to not thing about the grammar in some languages, and most times I'm doing a surprised double take of something that would be completely inappropriate or offensive instead of my simple phrases.
Earlier quoted context omitted.
Based on the availability of STT, TTS, and translation models available to download in that github repo, a real-life babelfish is 'only' some glue code away. Wild times we're living in.
hope that glue is better then the duct tapes google use on translate.g.c it is still hopeless, and much worse than the dictionary based, at gender/nums/dicleasions in general. i sometimes use it just to not thing about the grammar in some languages, and most times I'm doing a surprised double take of something that would be completely inappropriate or offensive instead of my simple phrases.
Modern translation apps and GPS are godsends that make travel a million times easier. And they're free! It blows my mind. Traveling would be so much more incredibly difficult without them.
This looks huge. Anyone know how this compares with Whisper in terms of quality and speed?
according to their blog post[1], MMS achieves ~half the error rate on words, while supporting 11x more languages. pretty impressive. [1] https://ai.facebook.com/blog/multilingual-model-speech-recog...
Edit: Just checked the paper, it seems to be worse[1][2] but feel free to correct me.
I feel like they should've just taken the Whipser architecture, scaled it, and scaled the dataset as they did.
[1] Page: https://i.imgur.com/bq15Tno.png
[2] Paper: https://scontent.fcai19-5.fna.fbcdn.net/v/t39.8562-6/3488279...
meta is doing more for open ai than openai
