HackerTrans
TopNewTrendsCommentsPastAskShowJobs

divamgupta

no profile record

Submissions

Show HN: Kitten TTS – 25MB CPU-Only, Open-Source TTS Model

github.com
1,003 points·by divamgupta·11 maanden geleden·361 comments

NamiGen: Free and Local Alternative to ElevenLabs

namigen.com
10 points·by divamgupta·vorig jaar·2 comments

Stable Diffusion Implemented Using TensorFlow and Keras

twitter.com
3 points·by divamgupta·4 jaar geleden·3 comments

Diffusion Bee: Stable Diffusion GUI App for M1 Mac

github.com
727 points·by divamgupta·4 jaar geleden·190 comments

comments

divamgupta
·11 maanden geleden·discuss
Thanks a lot for the detailed feedback. We are working on some models which do not use a phonemizer
divamgupta
·11 maanden geleden·discuss
Lol
divamgupta
·11 maanden geleden·discuss
Haha interesting pivot!
divamgupta
·11 maanden geleden·discuss
The thing is that the new models keep coming every day. So it’s economically not feasible to make chips for a single model
divamgupta
·11 maanden geleden·discuss
This is our goal too.
divamgupta
·11 maanden geleden·discuss
It was not
divamgupta
·11 maanden geleden·discuss
We are still training the model. We expect the quality to go up in the next release. This is just a preview release :)
divamgupta
·11 maanden geleden·discuss
Thanks for running the benchmarks. Currently the models are not optimized yet. We will optimize loading etc when we release an SDK meant for production :)
divamgupta
·11 maanden geleden·discuss
This is just an early checkpoint. We hope that the quality will improve in the future.
divamgupta
·11 maanden geleden·discuss
Currently we don't have chunking enabled yet. We will add it soon. That will remove the length limitations.
divamgupta
·11 maanden geleden·discuss
That is our vision too!
divamgupta
·11 maanden geleden·discuss
We will try to get rid of dependencies.
divamgupta
·11 maanden geleden·discuss
We are working to fix that. Thanks
divamgupta
·11 maanden geleden·discuss
Hmm, we will look into it.
divamgupta
·11 maanden geleden·discuss
Mostly model size, and input size. Some models which use attention are O(N^2)
divamgupta
·11 maanden geleden·discuss
We have released just a preview of the model. We hope to get the model much better in the future releases.
divamgupta
·11 maanden geleden·discuss
Yup it runs on the web browser. https://clowerweb.github.io/kitten-tts-web-demo/
divamgupta
·11 maanden geleden·discuss
We will release an STT model as well.
divamgupta
·11 maanden geleden·discuss
Thanks
divamgupta
·11 maanden geleden·discuss
Thanks for posting about our project in HN! I am one of the creators of KittenTTS

Here is the link to our repo: https://github.com/KittenML/KittenTTS