It sounds like Deep Speech 3 also uses an RNN transducer.
Slides here:
https://sites.grenadine.co/sites/cmu-scs-lti/en/colloquium/items/392
See for example[[1](https://arxiv.org/abs/1707.07413)]
Most helpful comment
It sounds like Deep Speech 3 also uses an RNN transducer.
Slides here:
https://sites.grenadine.co/sites/cmu-scs-lti/en/colloquium/items/392