i'll show an example

i usually leave about this much

okay thanks my dude!

is 53 seconds of voice data make a good model?

If you use this training notebook, yes: https://colab.research.google.com/drive/1x-sB6K4BDQm9dzEqPND8fQzTcvX4Z3Hh

Thank you!

53 seconds of good quality voice data

I used a pipeline notebook, and the voice I made sucks, I did the finalised one and the 20 one.

well train it until it sounds good

and default hifigan

I need the images aswell, I just focus on the number and image

yep

but what batch size should I use?

I've like 5 wavs

2~4

okay thanks!

I need help with something, can anyone explain this? I cant find the screenshot but heres the text: Generating mels Generating Mels Checking for missing files Checking Training Files wavs/1.npy|Dad, I want to be the piano. [WARNING] does not exist. --------------------------------------------------------------------------- FileNotFoundError Traceback (most recent call last) in 9 get_ipython().system('sed -i -- \'s,.wav|,.npy|,g\' "{hparams.training_files}"; sed -i -- \'s,.wav|,.npy|,g\' "{hparams.validation_files}"') 10 ---> 11 check_dataset(hparams) 2 frames /usr/local/lib/python3.8/dist-packages/numpy/lib/npyio.py in load(file, mmap_mode, allow_pickle, fix_imports, encoding) 415 own_fid = False 416 else: --> 417 fid = stack.enter_context(open(os_fspath(file), "rb")) 418 own_fid = True 419 FileNotFoundError: [Errno 2] No such file or directory: 'wavs/1.npy'

nvm the runtime disconnected

but just incase, tell me what i should do the next time this happens.

the wavs must be in the wavs folder

It doesn't found wav number 1

Hello, does anyone know how to use the Tacotron 2 model in windows python I train the model in https://colab.research.google.com/drive/1x-sB6K4BDQm9dzEqPND8fQzTcvX4Z3Hh

paging @Justin

Im pretty sure I put all the wavs in the folder, I did ctrl + a

When I try to turn the model on taco Tron once it did not work it’s not on the uberduck site anymore will tacotron it’s self get fix eventually

I've 7:08 minutes of voice data and 41 wavs what's the best batch size?

8 or 4

0kay thanks!

Hey guys first time playing with this how much audio is too much and how much should i use to make really good sounding voices ? I am extracting the voices form Final Fantasy and transcribing it all so can have hours worth per voice, want to use it for a mod to make non voiced sections of games a bit more fun so would like to make them as amazing as possible (please tag me if you reply)