PydanticModel not recognizing my inputs for an argument __class__

[dafrimer]

I can't get the pydantic model to recognize one of my inputs. Works fine in production but once i tested it out of production its not recognizing the class. I was using this code base but can't undderstand why its not working all of a sudden

https://github.com/aws-samples/amazon-sagemaker-gluonts-entrypoint

for instance:

hyperparameters = {'freq': 'W',
 'use_feat_static_cat': 'True',
 'use_feat_dynamic_real': 'False',
 'prediction_length': 10,
 'distr_output.__class__': 'gluonts.mx.distribution.neg_binomial.NegativeBinomialOutput',
 'trainer.__class__': 'gluonts.trainer.Trainer',
 'trainer.epochs': 100,
 'trainer.learning_rate': 0.001}

from pydoc import locate
klass = locate('gluonts.model.deepar.DeepAREstimator')
klass(**hyperparameters)

~/miniconda3/envs/gluontspy36/lib/python3.6/site-packages/gluonts/core/component.py in init_wrapper(*args, **kwargs)
    437                 self.__class__.__repr__ = validated_repr
    438 
--> 439             return init(self, **all_args)
    440 
    441         # attach the Pydantic model as the attribute of the initializer wrapper

TypeError: __init__() got an unexpected keyword argument 'distr_output.__class__'