*`-f` forced training (Not mandatory; values can be `y` for a forced training and `n` for a skip (a forced no-training)). By default, the hash value (from the training and test data, the structure of the model (.emadl) and the training parameters (.cnnt) of the model) will be compared. The model is retrained only if the hash changes. This can be used to distribute trained models, by distributing the corresponding `.training_hash` file as well, which will prevent a retraining.
Assuming both the architecture definition `VGG16.emadl` and the corresponding training configuration `VGG16.cnnt` are located in a folder `models` and the target code should be generated in a `target` folder using the `MXNet` backend, an example of a command is then:
