Missed TensorFlow Dev Summit? Check out the video playlist. Watch recordings

tfl.linear_lib.assert_constraints

View source on GitHub

Asserts that weights satisfy constraints.

tfl.linear_lib.assert_constraints(
    weights, monotonicities, monotonic_dominances, range_dominances, input_min,
    input_max, normalization_order, eps=0.0001
)

Args:

  • weights: Weights of Linear layer.
  • monotonicities: List or tuple of same length as number of elements in 'weights' of {-1, 0, 1} which represent monotonicity constraints per dimension. -1 stands for decreasing, 0 for no constraints, 1 for increasing.
  • monotonic_dominances: List of two-element tuple. First element is the index of the dominant feature. Second element is the index of the weak feature.
  • range_dominances: List of two-element tuples. First element is the index of the dominant feature. Second element is the index of the weak feature.
  • input_min: List or tuple of length same length as number of elements in 'weights' of either None or float which specifies the minimum value to clip by.
  • input_max: List or tuple of length same length as number of elements in 'weights' of either None or float which specifies the maximum value to clip by.
  • normalization_order: Whether weights have to have norm 1. Norm will be computed by: tf.norm(tensor, ord=normalization_order).
  • eps: Allowed constraints violation.

Returns:

List of assetion ops in graph mode or directly executes assertions in eager mode.