BenAAndrew / Voice-Cloning-App

A Python/Pytorch app for easily synthesising human voices
BSD 3-Clause "New" or "Revised" License
1.4k stars 233 forks source link

Is it possible for me to extract a runtime/environment from this? #146

Open FlashlightET opened 2 years ago

FlashlightET commented 2 years ago

I can't find a python runtime/environment in the temp folder for this program. This is important as installing a correct environment to run tacotron2 is the biggest pain in the world, and I want to use the voice-cloning-app environment to run tacotron2 in another program, without being restricted to using the app. I could not find a single trace of python.exe, but i did find python related files, so how is it able to run? What is it doing to call the synthesis script?

Avedena commented 2 years ago

It's look like you on Windows. Clone this Repro (or download the Zip) and use Anaconda to create a Python 3.6 Environment. Open a Terminal and install all dependency with pip install -r requirements.txt

Start the App with python main.py and inport your model and Hifi-gan Vocoder. With STRG+C you can exit the python script.

To Synthesis a Text string, execute the following: python synthesis/synthesize.py -m data/models/{YOUR-MODEL-NAME}/checkpoint_36000 -vm "data/hifigan/{VOCODER-MODELL}/model.pt" -hc "data/hifigan/{VOVODER-MODELL}/config.json" -t "My Text to be synthesized" -a audio.wav

You need to change the Path of your Model, the Checkpoint-File and Vocoder. Then you get a WAV-File with the audio in the root of the App.