search

lucidrains / voicebox-pytorch

Implementation of Voicebox, new SOTA Text-to-speech network from MetaAI, in Pytorch
MIT License
589 stars 49 forks source link

speartts model #35

Open happy-machine opened 10 months ago

happy-machine commented 10 months ago

Hi there

Ive tried building and saving the model using the speartts repo but the structure of the output doesnt align. Have also tried matching the parameters in both.

Is it possible to get more information on what text_to_semantic.load('spear/speartest.pt') expects or is there a trained model i can download?

Thanks!

Jon-Zbw commented 8 months ago

the same question

wassimseif commented 7 months ago

Hey @happy-machine.

You can train your own model from this repo: https://github.com/lucidrains/spear-tts-pytorch.

As far as I know, there's no pretrained models for this, or atleast I didn't find any

happy-machine commented 7 months ago

Thanks that’s really helpful!DanDisclaimer: I often send emails at times that suit my family and work commitments but I don’t expect a reply out of office hours. On 18 Jan 2024, at 18:33, Wassim Seifeddine @.***> wrote: Hey @happy-machine. You can train your own model from this repo: https://github.com/lucidrains/spear-tts-pytorch. As far as I know, there's no pretrained models for this, or atleast I didn't find any

—Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you were mentioned.Message ID: @.***>