KeyError: 'ncsnpp' of get_models() in models.utils

[Richardhrz]

hello author,
when I tried to run the checkpoint, an error occured:

KeyError Traceback (most recent call last)
Cell In[12], line 34
30 inverse_scaler = datasets.get_data_inverse_scaler(config)
32 # print(config)
---> 34 score_model = mutils.create_model(config)
36 optimizer = get_optimizer(config, score_model.parameters())
37 ema = ExponentialMovingAverage(score_model.parameters(),
38 decay=config.model.ema_rate)

File /score_sde_pytorch-main/models/utils.py:94, in create_model(config)
92 model_name = config.model.name
93 print(model_name) # ncsnpp
---> 94 score_model = get_model(model_name)(config)
95 score_model = score_model.to(config.device)
96 score_model = torch.nn.DataParallel(score_model)

File /score_sde_pytorch-main/models/utils.py:48, in get_model(name)
46 def get_model(name):
47 print(_MODELS)
---> 48 return _MODELS[name]
52 def get_sigmas(config):

KeyError: 'ncsnpp'

any suggestion would be grateful.

[Richardhrz]

_MODELS={} is empty, so return _MODELS[name] has nothing to do. what form of data should be put int the dict?

[youzhaoyirinide]

Hello, I meet the same issue as yours, could you tell me how to do if you resolve it? Thank you so much.

[Richardhrz]

import all the models in the jupyter file. _MODEL{} is not a global variable.