From the course: Google Cloud Platform for Machine Learning Essential Training

Unlock the full course today

Join today to access over 24,900 courses taught by industry experts.

Design and test speech generative output

Design and test speech generative output - Google Cloud Platform Tutorial

From the course: Google Cloud Platform for Machine Learning Essential Training

Start my 1-month free trial Buy for my team

Design and test speech generative output

“

So we've looked at multimodal, language, vision. So what's left is speech. So this can convert speech into text or synthesized speech using Google's universal speech model. So if we click "OPEN," we can do text to speech. And I would have to download it in this case, so I will probably do this one. And you can do speech to text. Now, this is using Google's Chirp model. So I'm going to go ahead and record myself twice, basically. I am recording so that I can get a translation of what I am saying. I am recording so that I can get a translation of what I am saying. And there I got this. Now, one of the things to notice about this is that it is for small snippets only. There's a different product for a longer input, and you can see, for more advanced features, use Speech Studio. And showing the guidance here, one of the most important things is quick testing. Audio files can be a maximum of 60 seconds or 10 megabytes. So this is quick testing. Files are transcribed with Chirp, and it has…

Contents