camenduru/ai-voice-cloning
1
0
Fork 1
forked from mrq/ai-voice-cloning
Code Pull Requests Activity

whispercpp actually works now (language loading was weird, slicing needed to divide time by 100), transcribing audio checks for silence and discards them

Browse Source
This commit is contained in:
mrq 2023-03-05 17:54:36 +00:00
parent b8a620e8d7
commit d97639e138
1 changed files with 15 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

22
src/utils.py
View File

@ -39,6 +39,7 @@ from tortoise.utils.device import get_device_name, set_device_name
MODELS['dvae.pth'] = "https://huggingface.co/jbetker/tortoise-tts-v2/resolve/3704aea61678e7e468a06d8eea121dba368a798e/.models/dvae.pth"
WHISPER_MODELS = ["tiny", "base", "small", "medium", "large"]
WHISPER_SPECIALIZED_MODELS = ["tiny.en", "base.en", "small.en", "medium.en"]
EPOCH_SCHEDULE = [ 9, 18, 25, 33 ]
args = None
tts = None