Generation doesn't get finished

[ForxBase]

Very often the whole generation doesn't get finished. The audio input is clear and short. The input text is clear, too:

[mytait]

audio generation is limited to 30 seconds. any longer and your text gets cut.

Also the motions are unstable (it says so in Gradio). If you leave the standard setting its the stablest. Changing the emotions params can produce emotions or introduce errors. dont use

[ForxBase]

audio generation is limited to 30 seconds. any longer and your text gets cut.

Also the motions are unstable (it says so in Gradio). If you leave the standard setting its the stablest. Changing the emotions params can produce emotions or introduce errors. dont use

Thanks. Is getting longer generations something that will come in a later version, and more authentic cloning?

[ForxBase]

audio generation is limited to 30 seconds. any longer and your text gets cut.

Also the motions are unstable (it says so in Gradio). If you leave the standard setting its the stablest. Changing the emotions params can produce emotions or introduce errors. dont use

So what happens is that even in generations under 30 seconds, the generation just has long pauses in between words, or after commas or punctuations.

[soljaragcnc]

audio generation is limited to 30 seconds. any longer and your text gets cut.
Also the motions are unstable (it says so in Gradio). If you leave the standard setting its the stablest. Changing the emotions params can produce emotions or introduce errors. dont use

So what happens is that even in generations under 30 seconds, the generation just has long pauses in between words, or after commas or punctuations.

Yeah, I'm having the exact same issue. seems like the more times I try to use it, the worse it gets. at one point it would just say the first 4 words and the rest would be blank. Even with the Emotions set to unconditional, its doing this.