python inference.py --tacotron2 --waveglow --wn-channels 256 --cpu -o output/ -i phrases/phrase.txt

i found this

yeah!

for installing rquirements,

pip install -r requirements.txt

From that repo

i did

now

no errors

'usage: inference.py [-h] -i INPUT -o OUTPUT [--suffix SUFFIX] [--tacotron2 TACOTRON2] [--waveglow WAVEGLOW] [-s SIGMA_INFER] [-d DENOISING_STRENGTH] [-sr SAMPLING_RATE] [--fp16 | --cpu] [--log-file LOG_FILE] [--include-warmup] [--stft-hop-length STFT_HOP_LENGTH] inference.py: error: the following arguments are required: -i/--input, -o/--output'

i got this error

thats bad

@zwf did you got this error before?

it's just asking you to pass some additional arguments to the script

check out how it's invoked in the example:

Wait a second

Where can I put the tacotron2 and waveglow models??

I didn't put

should i download these 2 models waveglow and tacotron2

?

@User do i need download checkpoints?

yes

where?

the links are number 7 here: https://github.com/zachwe/DeepLearningExamples/tree/master/PyTorch/SpeechSynthesis/Tacotron2#quick-start-guide

the tacotron model was trained on LJSpeech, so it will sound like her

there is soo many models

haha yea. for this you'll want tacotron and waveglow ... there are some compatibility issues with different tacotron2 checkpoints so you might wanna download the ones linked from the repo

but you may need to experiment a bit.

this file tacotron2_statedict.pt

?