deep-privacy / sidekit Goto Github PK

Hello!
I tried to reproduce your example code (Downloaded model from https://github.com/deep-privacy/sidekit/releases/download/sidekit_v0.1/best_halp_clr_adam_aam0.2_30_b256_vox12.pt_epoch71)

model_path = './best_halp_clr_adam_aam0.2_30_b256_vox12.pt_epoch71'
device = 'cuda' if torch.cuda.is_available() else 'cpu'

model_config = torch.load(model_path, map_location=device)

model_opts = model_config['model_archi']
if 'embedding_size' not in model_opts:
model_opts['embedding_size'] = 256
xtractor = Xtractor(model_config['speaker_number'],
model_archi=model_opts['model_type'],
loss=model_opts['loss']['type'],
embedding_size=model_opts['embedding_size'])

xtractor.load_state_dict(model_config['model_state_dict'], strict=True)

And I got the following:

RuntimeError Traceback (most recent call last)
in ()
12 embedding_size=model_opts['embedding_size'])
13
---> 14 xtractor.load_state_dict(model_config['model_state_dict'], strict=True)
15 xtractor = xtractor.to(device)
16 xtractor.eval()

/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py in load_state_dict(self, state_dict, strict)
1481 if len(error_msgs) > 0:
1482 raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format(
-> 1483 self.class.name, "\n\t".join(error_msgs)))
1484 return _IncompatibleKeys(missing_keys, unexpected_keys)
1485

RuntimeError: Error(s) in loading state_dict for Xtractor:
size mismatch for stat_pooling.attention.0.weight: copying a param with shape torch.Size([128, 7680, 1]) from checkpoint, the shape in current model is torch.Size([128, 61440, 1]).
size mismatch for stat_pooling.attention.4.weight: copying a param with shape torch.Size([2560, 128, 1]) from checkpoint, the shape in current model is torch.Size([20480, 128, 1]).
size mismatch for stat_pooling.attention.4.bias: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([20480]).