新人刚学，用的社区模型一直出来的是杂音，还会经常有以下报错，求大佬解决

Building hifigan
Loading 'vocoder\saved_models\pretrained\g_hifigan.pt'
Complete.
Removing weight norm...
Synthesizer using device: cuda
Trainable Parameters: 0.000M
Traceback (most recent call last):
File "Z:\MockingBird-main\MockingBird-main\toolbox\__init__.py", line 143, in
func = lambda: self.synthesize() or self.vocode()
File "Z:\MockingBird-main\MockingBird-main\toolbox\__init__.py", line 259, in synthesize
specs = self.synthesizer.synthesize_spectrograms(texts, embeds, style_idx=int(self.ui.style_slider.value()), min_stop_token=min_token, steps=int(self.ui.length_slider.value())*200)
File "Z:\MockingBird-main\MockingBird-main\synthesizer\inference.py", line 93, in synthesize_spectrograms
self.load()
File "Z:\MockingBird-main\MockingBird-main\synthesizer\inference.py", line 71, in load
self._model.load(self.model_fpath, self.device)
File "Z:\MockingBird-main\MockingBird-main\synthesizer\models\base.py", line 51, in load
self.load_state_dict(checkpoint["model_state"], strict=False)
File "C:\ProgramData\Anaconda3\lib\site-packages\torch\nn\modules\module.py", line 1482, in load_state_dict
raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format(
RuntimeError: Error(s) in loading state_dict for Tacotron:
size mismatch for encoder.embedding.weight: copying a param with shape torch.Size([75, 512]) from checkpoint, the shape in current model is torch.Size([70, 512]).
size mismatch for gst.stl.attention.W_query.weight: copying a param with shape torch.Size([512, 256]) from checkpoint, the shape in current model is torch.Size([512, 512]).

模型用的是：


Start with synthesizer/inference.py and synthesizer/models/base.py, then trace the call from toolbox/__init__.py. Reproduce the load failure and compare the checkpoint dimensions shown in the traceback with the current Tacotron and GST configuration. Done means the selected community model loads successfully and synthesis no longer produces the reported errors or persistent noise.

Written by the indexing model from the issue text.