IS THIS AI GENERATED? Why it sounds SO CLEAR???

Number 1: No

Gah, damn

Number 2: It's Already Cleared

Can you show "before" version of this audio?

Oops it fucked up again`ModuleNotFoundError: No module named 'emoji'`

Running setup.py install for sklearn ... error
ERROR: Command errored out with exit status 1: /usr/bin/python3 -u -c 'import io, os, sys, setuptools, tokenize; sys.argv[0] = '"'"'/tmp/pip-install-kc8uvsvw/sklearn_a52d76c2056c46dc8cac3ed2d174f1e0/setup.py'"'"'; __file__='"'"'/tmp/pip-install-kc8uvsvw/sklearn_a52d76c2056c46dc8cac3ed2d174f1e0/setup.py'"'"';f = getattr(tokenize, '"'"'open'"'"', open)(__file__) if os.path.exists(__file__) else io.StringIO('"'"'from setuptools import setup; setup()'"'"');code = f.read().replace('"'"'\r\n'"'"', '"'"'\n'"'"');f.close();exec(compile(code, __file__, '"'"'exec'"'"'))' install --record /tmp/pip-record-1wx7eze6/install-record.txt --single-version-externally-managed --compile --install-headers /usr/local/include/python3.8/sklearn Check the logs for full command output.

I Trained The Concept Unification Installation Tape Guy

????

I think my is corrupt. [fakeyou synthesizer]

Hi! Is it somehow possible to download weights for specific voice and run model on my own machine?

no

not off the site at least

😦 Are there any paid plans, where I can download the voice?

no

our models are made by contributors that have differing opinions on whether they want their voice models distributed to anyone

and we don’t offer downloads as voices could be easily stolen and reuploaded with little evidence

Got it! Thank you:)

Do you have pricing plans for commercial use with garantee for latency and rps?

@samkoelle this is the part where i admit i know nothing about the paid plans and offload to the bigger brain

Soft guarantees. The creator plan improves latency to 1-5 seconds depending on whether model is in memory. We can lower further with a dedicated server (for example we have one customer who does this for a single voice to consistently get 1s), but its a fair amount more expensive. We don't offer commercial use of most of the user-contributed voices. Some customers do use them commercially in things like parody content, but the legality of this is use case dependent and their call. We do however have a few (and soon more) commercially licensed voices for more broad use. What sort of voices are you trying to use, and what sort of latency do you require?

I Am Doing The Kirby's Avalanche Female Announcer

This Is The Results

OH BOY, FREE HORROR SOUND AFFECTS! /j

The background sound when the killer is nearby

Hi! Me and @karfly are working in the same team, so I will also write here and discuss our case. First, thanks a lot for your answers! We have tried the creator plan as suggested, and it works very fine in term of latency (5s on average), but we measured RPS limit for short lines of text to be around 1 req/sec (when RPS exceeds this value, latency increases exponentially), which is too low for our purposes. So can we discuss dedicated server pricing plan? If it requires additional information, we are ready to have a call and discuss every detail, or we can just continue here or in a private chat. We also consider an option of fine-tuning the model and hosting it by ourselves, but it would be much more favorable for us to pay for the Uberduck service which seems to provide very good sound quality and convenient API.

Yes, we have a rate limiter in place, but this can be turned off for particular customers. What sort of volume / models are you looking for? I'm tagging in @zwf who handles sales. We'd love to have a call to discuss. Thanks!