Learning rates are negative, and negative values cannot be drawn using the scipy.stats.triang.rvs method. Instead, positive values are used to define the triangular distribution, and a negative values is applied in the code after the random value is drawn.
Learning rates are negative, and negative values cannot be drawn using the
scipy.stats.triang.rvs
method. Instead, positive values are used to define the triangular distribution, and a negative values is applied in the code after the random value is drawn.