clonar-voz / TTS /vc /models /__init__.py
Shadhil's picture
voice-clone with single audio sample input
9b2107c
raw
history blame
597 Bytes
import importlib
import re
from typing import Dict, List, Union
def to_camel(text):
text = text.capitalize()
return re.sub(r"(?!^)_([a-zA-Z])", lambda m: m.group(1).upper(), text)
def setup_model(config: "Coqpit", samples: Union[List[List], List[Dict]] = None) -> "BaseVC":
print(" > Using model: {}".format(config.model))
# fetch the right model implementation.
if "model" in config and config["model"].lower() == "freevc":
MyModel = importlib.import_module("TTS.vc.models.freevc").FreeVC
model = MyModel.init_from_config(config, samples)
return model