add web demo to Huggingface

[neonbjb]

add web demo to Huggingface

[osanseviero]

You can find a working demo in https://huggingface.co/spaces/osanseviero/tortoisse-tts 🔥

[osanseviero]

Hey there, if it's ok I would love to work on this. I'm actually half-way done through it 🚀 Working on https://huggingface.co/spaces/osanseviero/tortoisse-tts and happy to transfer once done.

[neonbjb]

Awesome, I'd be glad to do this then! Expect something in the next day or two..

[neonbjb]

I'd love to do this, but this model will not run on CPUs (inference will take hours or more). I read that GPU runtimes are for organizations only. Let me know if this understanding is wrong..

[AK391]

@neonbjb great, from the spaces docs https://huggingface.co/docs/hub/spaces

"Each Spaces environment is limited to 16GB RAM and 8 CPU cores. Organization subscribers (Lab, Startup, and Enterprise) can access Spaces with one T4 GPU on a case-by-case basis. Please email us at website at huggingface.co or let us know on Twitter if you need one."

GPU's can be assigned to individual accounts as well if requested

I can also send a request for GPU for a space once its setup

[neonbjb]

Sure, have at it. I'll await your pr.

[neonbjb]

Hmm looks like it is running out of RAM? Did you get it to work without this error?

Tortoise shouldn't need more than 16GB of RAM, but this may be a result of doing the computations in on CPU. I'll have to profile it to be sure.

[osanseviero]

I restarted it and it's up again, but I haven't looked at why we got a out of RAM issue before.

[neonbjb]

Space created: https://huggingface.co/spaces/jbetker/tortoise; I'll contact website@ to see if I can get a GPU instance dedicated to it.

[neonbjb]

Closing this out.