Try-on module pre-trained model is not valid

[Lehsuby]

@andrewjong Hi. When I try to load your Try-On model (Unet model) from Google Drive, I get the following error:

  File "test.py", line 10, in <module>
    main(train=False)
  File "/code/train.py", line 44, in main
    model = model_class.load_from_checkpoint(
  File "/root/miniconda3/envs/sams-pt1.6/lib/python3.8/site-packages/pytorch_lightning/core/saving.py", line 153, in load_from_checkpoint
    model = cls._load_model_state(checkpoint, *args, strict=strict, **kwargs)
  File "/root/miniconda3/envs/sams-pt1.6/lib/python3.8/site-packages/pytorch_lightning/core/saving.py", line 192, in _load_model_state
    model.load_state_dict(checkpoint['state_dict'], strict=strict)
  File "/root/miniconda3/envs/sams-pt1.6/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1044, in load_state_dict
    raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format(
RuntimeError: Error(s) in loading state_dict for UnetMaskModel:
        Missing key(s) in state_dict: "unet.model.model.1.model.3.model.1.weight", "unet.model.model.1.model.3.model.1.bias", "unet.model.model.1.model.3.model.3.model.1.weight", "unet.model.model.1.model.3.model.3.model.1.bias", "unet.model.model.1.model.3.model.3.model.3.model.1.weight", "unet.model.model.1.model.3.model.3.model.3.model.1.bias", "unet.model.model.1.model.3.model.3.model.3.model.3.gamma", "unet.model.model.1.model.3.model.3.model.3.model.3.query_conv.weight", "unet.model.model.1.model.3.model.3.model.3.model.3.query_conv.bias", "unet.model.model.1.model.3.model.3.model.3.model.3.key_conv.weight", "unet.model.model.1.model.3.model.3.model.3.model.3.key_conv.bias", "unet.model.model.1.model.3.model.3.model.3.model.3.value_conv.weight", "unet.model.model.1.model.3.model.3.model.3.model.3.value_conv.bias", "unet.model.model.1.model.3.model.3.model.3.model.4.model.1.weight", "unet.model.model.1.model.3.model.3.model.3.model.4.model.1.bias", "unet.model.model.1.model.3.model.3.model.3.model.4.model.2.gamma", "unet.model.model.1.model.3.model.3.model.3.model.4.model.2.query_conv.weight", "unet.model.model.1.model.3.model.3.model.3.model.4.model.2.query_conv.bias", "unet.model.model.1.model.3.model.3.model.3.model.4.model.2.key_conv.weight", "unet.model.model.1.model.3.model.3.model.3.model.4.model.2.key_conv.bias", "unet.model.model.1.model.3.model.3.model.3.model.4.model.2.value_conv.weight", "unet.model.model.1.model.3.model.3.model.3.model.4.model.2.value_conv.bias", "unet.model.model.1.model.3.model.3.model.3.model.4.model.5.weight", "unet.model.model.1.model.3.model.3.model.3.model.4.model.5.bias", "unet.model.model.1.model.3.model.3.model.3.model.4.model.7.gamma", "unet.model.model.1.model.3.model.3.model.3.model.4.model.7.query_conv.weight", "unet.model.model.1.model.3.model.3.model.3.model.4.model.7.query_conv.bias", "unet.model.model.1.model.3.model.3.model.3.model.4.model.7.key_conv.weight", "unet.model.model.1.model.3.model.3.model.3.model.4.model.7.key_conv.bias", "unet.model.model.1.model.3.model.3.model.3.model.4.model.7.value_conv.weight", "unet.model.model.1.model.3.model.3.model.3.model.4.model.7.value_conv.bias", "unet.model.model.1.model.3.model.3.model.3.model.7.weight", "unet.model.model.1.model.3.model.3.model.3.model.7.bias", "unet.model.model.1.model.3.model.3.model.3.model.9.gamma", "unet.model.model.1.model.3.model.3.model.3.model.9.query_conv.weight", "unet.model.model.1.model.3.model.3.model.3.model.9.query_conv.bias", "unet.model.model.1.model.3.model.3.model.3.model.9.key_conv.weight", "unet.model.model.1.model.3.model.3.model.3.model.9.key_conv.bias", "unet.model.model.1.model.3.model.3.model.3.model.9.value_conv.weight", "unet.model.model.1.model.3.model.3.model.3.model.9.value_conv.bias", "unet.model.model.1.model.3.model.3.model.6.weight", "unet.model.model.1.model.3.model.3.model.6.bias", "unet.model.model.1.model.3.model.6.weight", "unet.model.model.1.model.3.model.6.bias", "unet.model.model.1.model.6.weight", "unet.model.model.1.model.6.bias".
        Unexpected key(s) in state_dict: "unet.model.model.1.model.9.gamma", "unet.model.model.1.model.9.query_conv.weight", "unet.model.model.1.model.9.query_conv.bias", "unet.model.model.1.model.9.key_conv.weight", "unet.model.model.1.model.9.key_conv.bias", "unet.model.model.1.model.9.value_conv.weight", "unet.model.model.1.model.9.value_conv.bias", "unet.model.model.1.model.3.gamma", "unet.model.model.1.model.3.query_conv.weight", "unet.model.model.1.model.3.query_conv.bias", "unet.model.model.1.model.3.key_conv.weight", "unet.model.model.1.model.3.key_conv.bias", "unet.model.model.1.model.3.value_conv.weight", "unet.model.model.1.model.3.value_conv.bias", "unet.model.model.1.model.4.model.1.weight", "unet.model.model.1.model.4.model.1.bias", "unet.model.model.1.model.4.model.3.gamma", "unet.model.model.1.model.4.model.3.query_conv.weight", "unet.model.model.1.model.4.model.3.query_conv.bias", "unet.model.model.1.model.4.model.3.key_conv.weight", "unet.model.model.1.model.4.model.3.key_conv.bias", "unet.model.model.1.model.4.model.3.value_conv.weight", "unet.model.model.1.model.4.model.3.value_conv.bias", "unet.model.model.1.model.4.model.4.model.1.weight", "unet.model.model.1.model.4.model.4.model.1.bias", "unet.model.model.1.model.4.model.4.model.3.model.1.weight", "unet.model.model.1.model.4.model.4.model.3.model.1.bias", "unet.model.model.1.model.4.model.4.model.3.model.3.model.1.weight", "unet.model.model.1.model.4.model.4.model.3.model.3.model.1.bias", "unet.model.model.1.model.4.model.4.model.3.model.3.model.4.weight", "unet.model.model.1.model.4.model.4.model.3.model.3.model.4.bias", "unet.model.model.1.model.4.model.4.model.3.model.6.weight", "unet.model.model.1.model.4.model.4.model.3.model.6.bias", "unet.model.model.1.model.4.model.4.model.6.weight", "unet.model.model.1.model.4.model.4.model.6.bias", "unet.model.model.1.model.4.model.7.weight", "unet.model.model.1.model.4.model.7.bias", "unet.model.model.1.model.4.model.9.gamma", "unet.model.model.1.model.4.model.9.query_conv.weight", "unet.model.model.1.model.4.model.9.query_conv.bias", "unet.model.model.1.model.4.model.9.key_conv.weight", "unet.model.model.1.model.4.model.9.key_conv.bias", "unet.model.model.1.model.4.model.9.value_conv.weight", "unet.model.model.1.model.4.model.9.value_conv.bias", "unet.model.model.1.model.7.weight", "unet.model.model.1.model.7.bias".

It seems that the checkpoint was recorded before the last changes were made to the model architecture. Could you help with that?

[Vannaz]

It's a pity that the ShineOn model checkpoint haven't been found now, I don't know if you can be provided again.