Closed agonzalezd closed 4 years ago
Here is the commit https://github.com/begeekmyfriend/tacotron2/commit/8adb9accf0b4ecb56c821cd80719063706b6df0a Since I do not use docker and I have no plan to install such environmennt recently, I am not sure whether it might work for you. Please let me know if you have any problem. Have fun!
Yes, thank you!
I have to change some things due to hardware compatibility issues but it is a good starting point. Thanks!
Here is a real time synthesiis solution for you https://github.com/begeekmyfriend/tacotron2/issues/13#issuecomment-599992425
This is interesting... Is it compatible with the output of this repository?
Do you plan to make a single repository with both the multispeaker spectrogram generation and the real time synthesis?
Thanks in advance!
This project supports multi-speaker. See scripts/train_tacotron2.sh
. And then throw all the GTA mel spectrograms into vocoder for training.
Yes, I know. I was referring more to the "real time synthesis" part :)
Hello
I was checking your repository for using it with my own data, and I found that I couldn't create a docker image from the Dockerfile since the
requirements.txt
file is missing...Could you provide it or say which are the needed requirements for it to work? I'm having issues with GPU drivers and I would like to check at least those that work for you.
Thank you so much for your response,
Ander