Create a Variational Recurrent Neural Network (VRNN) model. Args: param_dict: A :py:class:`ParamVRNN` object to define the hyper-parameters of the network. kwargs: If param_dict is None, you can directly provide keyword argume
(self, param_dict: ParamVRNN = None, **kwargs)
| 208 | """ |
| 209 | |
| 210 | def __init__(self, param_dict: ParamVRNN = None, **kwargs): |
| 211 | """ |
| 212 | Create a Variational Recurrent Neural Network (VRNN) model. |
| 213 | |
| 214 | Args: |
| 215 | param_dict: |
| 216 | A :py:class:`ParamVRNN` object to define the hyper-parameters of the network. |
| 217 | kwargs: |
| 218 | If param_dict is None, you can directly provide keyword arguments of :py:class:`ParamVRNN` here. |
| 219 | """ |
| 220 | super().__init__() |
| 221 | |
| 222 | # read and set configs |
| 223 | if param_dict is not None: |
| 224 | self.config_dict = ParamVRNN(**param_dict) |
| 225 | else: |
| 226 | self.config_dict = ParamVRNN(**kwargs) |
| 227 | |
| 228 | for key in self.config_dict: |
| 229 | setattr(self, key, self.config_dict[key]) |
| 230 | |
| 231 | # construct sub-networks |
| 232 | self._construct_networks() |
| 233 | |
| 234 | def _construct_networks(self): |
| 235 | """Construct the sub-networks. |
nothing calls this directly
no test coverage detected