XTTS stream generator/Transformers

[erew123]

I'm sure you have plenty to be on with and I've not looked at this myself (yet), but Im putting this here as a general "to be looked at" one day. (Feel free to close this off if you wish)

When generating streams with XTTS, you can get this message:

_environment\env\Lib\site-packages\TTS\tts\layers\xtts\stream_generator.py:138: UserWarning: You have modified the pretrained model configuration to control generation. This is a deprecated strategy to control generation and will be removed soon, in a future version. Please use a generation configuration file (see https://huggingface.co/docs/transformers/main_classes/text_generation)
  warnings.warn(

This is linked to using SampleOutput method, which I had to ask the Huggingface devs to re-add it to transformers (my post/discussion with them below):

huggingface/transformers#28649

Im sure its probably something that can be resolved/worked around with an update to to the streaming code, its just that, as there was no development going on, there was no way to resolve it.

So Im just adding this discussion as a reference, as something to be looked at some time. If you/someone else never manages to get to it, Ill take a look myself at some point in future when my coding calms down (Im rushing to get my own code out atm).

I just thought it may be handy to put the reference to it all up here.

Again, thanks so much for your work on this!! And I am using 0.24.1 now and all seems good as a quick test. :)

Thanks