🔊 When you make models open, the community can stretch and extend and get creative with them in ways that you would have never imagined.
@GoogleGemma 4 (and now even EmbeddingGemma) can do speech-to-text, super efficiently!
Gemma 4 and EmbeddingGemma can do efficient speech-to-text
AISummary
Paige Bailey says open models let the community extend them in unexpected ways, and points to Gemma 4 and EmbeddingGemma doing speech-to-text efficiently. A linked post by Shahir reports turning EmbeddingGemma 2 into a speech-to-text model that transcribes an hour of audio in about 13 seconds on a base M3 MacBook Air, or 288× real time.
Post on XView on X
👩💻 Paige BaileyVerified on X
@DynamicWebPaige
I turned @googlegemma EmbeddingGemma 2, an embedding model, into a speech-to-text model, and it turned out to be one of the fastest STT models you can run on device. On a base M3 MacBook Air it transcribes an hour of audio in about 13 seconds: 288× real time. The idea came in March with Gemini Embedding 2. It puts text, images, audio and video into one shared space, so a spoken word lands right next to the written word. If that's true, you should be able to read a transcript straight out of the embedding model. I couldnt tried it then, but last week EmbeddingGemma 2 came out as an open model, so I tried it.
Source: 👩💻 Paige Bailey · x.comPublished