huber_loss(ŷ, y; δ = 1, agg = mean)

Return the mean of the Huber loss given the prediction ŷ and true values y.

			             | 0.5 * |ŷ - y|^2,            for |ŷ - y| <= δ
Huber loss = |
             |  δ * (|ŷ - y| - 0.5 * δ), otherwise

Example

			julia> ŷ = [1.1, 2.1, 3.1];

julia> Flux.huber_loss(ŷ, 1:3)  # default δ = 1 > |ŷ - y|
0.005000000000000009

julia> Flux.huber_loss(ŷ, 1:3, δ=0.05)  # changes behaviour as |ŷ - y| > δ
0.003750000000000005