i was talking about a talknet model

are you making a talknet singing model for tacotron

nope

is it possible to make singing models for tacotron

i dont know

not yet, some uberduck contributors are working to implement it

anyone had n brio's roger craig smith voice?

well

ok

How long ago did NVIDIA deprecate Talknet?

2 years ago

Damn, and we implemented it like what, a year and a half ago?

Over a year ago

June 2021

Oh

After it was deprecated ig 💀

Hello.

Ahoy

offline version seems to have issues anyone know how to fix it heres the error

Starting TalkNet server. Close this window to shut down the server. Traceback (most recent call last): File "talknet_offline.py", line 3, in from controllable_talknet import * File "C:\Users\kenna\Desktop\cause i had to\ControllableTalkNet\controllable_talknet.py", line 3, in import dash File "C:\Users\kenna\Desktop\cause i had to\miniconda\lib\site-packages\dash\__init__.py", line 5, in from .dash import Dash, no_update # noqa: F401,E402 File "C:\Users\kenna\Desktop\cause i had to\miniconda\lib\site-packages\dash\dash.py", line 22, in from werkzeug.debug.tbtools import get_current_traceback ImportError: cannot import name 'get_current_traceback' from 'werkzeug.debug.tbtools' (C:\Users\kenna\AppData\Roaming\Python\Python38\site-packages\werkzeug\debug\tbtools.py) Press any key to continue . . .

Hi, just wanted to ask if there might be a guide somewhere for the LJSpeech format? I’m currently just making sure my transcriptions are correct… Or let me know if I should have asked this somewhere else…

The LJ Speech format of transcription?

Yeah. According to the guide the transcriptions for a talknet model have to be in LJSpeech if I’m not mistaken?

Not too familiar with the format…

Yes, it does

Is there a summary guide on the format somewhere maybe?

It's simple to understand. Just do it this way in a .txt file:

wavs/001.wav|This is a sentence.

wavs/002.wav|This is another sentence.

wavs/003.wav|An so on.

Ohh that. What about how the sentences are written. I have a bunch of questions… 1. Is it case sensitive? 2. Do they have to be complete sentences and always end with punctuation? 3. Can you use slang words? e.g. fuck, um, uh, oh, ooh, huh 4. Can you use phonetic approximations of non-English words or words with special pronunciation? e.g. ‘mm’, ‘hmm’ (transcribable filler sounds) e.g. ‘hoss sen’ (how to properly pronounce my name) 5. Should you use punctuation based on proper grammar or based on the length of pauses and tone? e.g. “Oh, my god. Wow. What? That was hella amazing.” or “oh. my. god. wow! what??? that was hell uh…amazing!!!” Sorry for the long reply…

1. No. 2. I'm not sure what you mean by "complete sentences" 3. Yes 4. Yes 5. Based on the length of pauses and tone, but you only need one question mark, one period, and/or one exclamation point, not three in a row. e.g.

oh. my. god. wow! what? that was hell uh. amazing!