Open ysalex3d opened 1 year ago
i also facing the same error: reading text from stdin! when i run the (python scripts/tortoise_tts.py --voice emma --seed 42 --text "$TEXT") on window 10, the file run and then hang the cmd, so result get from the code, why does this, how can we fix this:
anyone find a solution for this? I'm having this problem too
Also experiencing this issue. It looks from the code like this log is intended to print when args.general.text does not exist, but I DID provide that argument
.\scripts\tortoise_tts.py --preset ultra_fast --voice train_mouse --output_dir "E:\Audio\TTS" --candidates 1 --text "If you are hearing this, then it worked"
Cannot figure out how to clone voices in this (fast) version (miniconda, windows 10).
In the original tortoise it was--> python audio.py --voice=
First, when I run any suggested command that calls the script folder, such as (python scripts/tortoise_tts.py --voice emma --seed 42 --text "$TEXT"), I get the error "ModuleNotFoundError: No module named 'tortoise'".
I fixed this by moving all the /scripts .py files up to the main tortoise-tts-fast folder.
So with that said, best I can figure out, the python command in miniconda should be either
python tortoise_tts.py --voice NEWVOICE
or
python tortoise_tts.py --voice NEWVOICE --seed 42 --text "$TEXT"
Problem is with either of these inputs I get something that says "Reading text from stdin!", and then miniconda hangs, and I get nothing. This is the same error I was getting in the other /script .py files, like app.py, before I moved them up to the main folder
Alternatively
Ive watched the google colab version of this tutorial https://www.youtube.com/watch?v=8i4T5v1Fl_M&t=1s&ab_channel=MartinThissen
But when I use "streamlit run app.py", the browser page has no area to create a new voice, as show in the colab video.
Can anyone help me on this, would really appreciate it.