Variable type for NN bounds

[tsaycal]

I am confused about what format of input bounds should be passed with an onnx model to omlt. Trying a list or a dictionary:

lb = np.maximum(0, image - epsilon_infty)
ub = np.minimum(1, image + epsilon_infty)
input_bounds = [(float(l), float(u)) for l, u in zip(lb[0], ub[0])]

or

input_bounds = {}
for i in range(28*28):
    input_bounds[i] = (float(lb[0][i]), float(ub[0][i])) 

Attempting to create the omlt model with either:

write_onnx_model_with_bounds(f.name, None, input_bounds)
network_definition = load_onnx_neural_network_with_bounds(f.name)
formulation = NeuralNetworkFormulation(network_definition)
m = pyo.ConcreteModel()
m.nn = OmltBlock()
m.nn.build_formulation(formulation) 

Results in: ValueError: Variable 'bounds' keyword must be a tuple or function

[fracek]

If you have the onnx model (converted from Torch) you can use load_onnx_neural_network(onnx_model, input_bounds) without first writing to a file. Your input bounds dictionary looks good if the input of the NN is a 1-dimensional vector.

[jalving]

Can you check your version of Pyomo? You have to upgrade to version 6.2 to support dictionary bounds. Regardless, we should update OMLT requirements in the setup file.

[tsaycal]

Using the load_onnx_neural_network(onnx_model, input_bounds) produces the same error. The input is 1-dimensional:

Running input_layer = list(network_definition.input_layers)[0]; print(input_layer.input_indexes) produces:

[(0,), (1,), (2,), (3,), (4,) ..., (780,), (781,), (782,), (783,)]

[tsaycal]

Upgrading to Pyomo 6.2 fixed this issue. Thanks @jalving @fracek !