MCPcopy
hub / github.com/coqui-ai/TTS / eval

Method eval

TTS/tts/models/xtts.py:708–711  ·  view source on GitHub ↗

Sets the model to evaluation mode. Overrides the default eval() method to also set the GPT model to eval mode.

(self)

Source from the content-addressed store, hash-verified

706 return Xtts(config)
707
708 def eval(self): # pylint: disable=redefined-builtin
709 """Sets the model to evaluation mode. Overrides the default eval() method to also set the GPT model to eval mode."""
710 self.gpt.init_gpt_for_inference()
711 super().eval()
712
713 def get_compatible_checkpoint_state_dict(self, model_path):
714 checkpoint = load_fsspec(model_path, map_location=torch.device("cpu"))["model"]

Callers 15

_test_inferenceMethod · 0.45
test_test_runMethod · 0.45
_test_inferenceMethod · 0.45
test_voice_conversionMethod · 0.45
test_test_runMethod · 0.45
get_wavlmFunction · 0.45
load_checkpointMethod · 0.45
load_checkpointFunction · 0.45
load_checkpointMethod · 0.45

Calls 1

Tested by 8

_test_inferenceMethod · 0.36
test_test_runMethod · 0.36
_test_inferenceMethod · 0.36
test_voice_conversionMethod · 0.36
test_test_runMethod · 0.36