add_keras_constr

gurobi_ml.keras.keras.add_keras_constr(gp_model, keras_model, input_vars, output_vars=None, **kwargs)

Formulate keras_model into gp_model.

The formulation predicts the values of output_vars using input_vars according to keras_model. See our Users Guide for details on the mip formulation used.

Parameters:

• gp_model (gurobipy model) – The gurobipy model where the predictor should be inserted.

• keras_model (keras.Model <https://keras.io/api/models/model/>) – The keras model to insert as predictor.

• input_vars (mvar_array_like) – Decision variables used as input for Keras model in gp_model.

• output_vars (mvar_array_like, optional) – Decision variables used as output for Keras model in gp_model.

Returns:

Object containing information about what was added to gp_model to formulate keras_model into it

Return type:

KerasNetworkConstr

Raises:

NoModel – If the translation for some of the Keras model structure (layer or activation) is not implemented.

Warning

Only Dense (possibly with relu activation), and ReLU with default settings are supported.

Notes

See add_predictor_constr for acceptable values for input_vars and output_vars