mirror of https://github.com/coqui-ai/TTS.git
small fixes
parent
be51205607
commit
a097541ed4
|
@ -593,7 +593,6 @@ class Xtts(BaseTTS):
|
|||
cond_free_k=2,
|
||||
diffusion_temperature=1.0,
|
||||
decoder_sampler="ddim",
|
||||
use_hifigan=True,
|
||||
**hf_generate_kwargs,
|
||||
):
|
||||
text = f"[{language}]{text.strip().lower()}"
|
||||
|
@ -765,7 +764,7 @@ class Xtts(BaseTTS):
|
|||
checkpoint_dir=None,
|
||||
checkpoint_path=None,
|
||||
vocab_path=None,
|
||||
eval=False,
|
||||
eval=True,
|
||||
strict=True,
|
||||
use_deepspeed=False,
|
||||
):
|
||||
|
@ -777,7 +776,7 @@ class Xtts(BaseTTS):
|
|||
checkpoint_dir (str, optional): The directory where the checkpoint is stored. Defaults to None.
|
||||
checkpoint_path (str, optional): The path to the checkpoint file. Defaults to None.
|
||||
vocab_path (str, optional): The path to the vocabulary file. Defaults to None.
|
||||
eval (bool, optional): Whether to set the model to evaluation mode. Defaults to False.
|
||||
eval (bool, optional): Whether to set the model to evaluation mode. Defaults to True.
|
||||
strict (bool, optional): Whether to strictly enforce that the keys in the checkpoint match the keys in the model. Defaults to True.
|
||||
|
||||
Returns:
|
||||
|
|
|
@ -71,7 +71,7 @@ print("Loading model...")
|
|||
config = XttsConfig()
|
||||
config.load_json("/path/to/xtts/config.json")
|
||||
model = Xtts.init_from_config(config)
|
||||
model.load_checkpoint(config, checkpoint_dir="/path/to/xtts/", eval=True, use_deepspeed=True)
|
||||
model.load_checkpoint(config, checkpoint_dir="/path/to/xtts/", use_deepspeed=True)
|
||||
model.cuda()
|
||||
|
||||
print("Computing speaker latents...")
|
||||
|
@ -108,7 +108,7 @@ print("Loading model...")
|
|||
config = XttsConfig()
|
||||
config.load_json("/path/to/xtts/config.json")
|
||||
model = Xtts.init_from_config(config)
|
||||
model.load_checkpoint(config, checkpoint_dir="/path/to/xtts/", eval=True, use_deepspeed=True)
|
||||
model.load_checkpoint(config, checkpoint_dir="/path/to/xtts/", use_deepspeed=True)
|
||||
model.cuda()
|
||||
|
||||
print("Computing speaker latents...")
|
||||
|
|
Loading…
Reference in New Issue