vincent-antaki
vincent-antaki
The ExecutionContext's save function uses a method call initialize_if_needed which, in case for a full dump, is called recursively on every step. This functions calls each steps setup function. However,...
See the following [comment](https://github.com/Neuraxio/Neuraxle/pull/448#discussion_r599752835).
Right now, we only measure metrics after each epoch. It could be useful to measure metrics before the first epoch as to gain better insight on the training process. Considerations:...
The save_step/load_step methods of this saver define two cases for each sub-step: one if the should_save method returns True, and one if it returns False. The later crashes on load_step...
Most unit tests are relatively quick to run. However, test_example.py and test_tpe.py are very slow; especially the latter. We should fix this. Seed setting should be considered for test_tpe.
**Is your feature request related to a problem? Please describe.** With the current implementation, the _fit_transform_data_container function calls _fit_data_container() and then _transform_data_container(). With the right parameter, each of them can...
**Is your feature request related to a problem? Please describe.** BaseTransformers instances implement _HasHyperparameters and _HasHyperparameterSpace, both of which assign a RecursiveDict to respectively self.hyperparameters and self.hyperparameter_space. But, since every...