rename crossentropy loss

docs/src/training/training.md

@@ -30,7 +30,7 @@ loss(x, y) = Flux.mse(m(x), y)
 Flux.train!(loss, data, opt)
 ```
 
-The loss will almost always be defined in terms of some *cost function* that measures the distance of the prediction `m(x)` from the target `y`. Flux has several of these built in, like `mse` for mean squared error or `logloss` for cross entropy loss, but you can calculate it however you want.
+The loss will almost always be defined in terms of some *cost function* that measures the distance of the prediction `m(x)` from the target `y`. Flux has several of these built in, like `mse` for mean squared error or `crossentropy` for cross entropy loss, but you can calculate it however you want.
 
 ## Datasets
 

src/layers/stateless.jl

@@ -2,5 +2,7 @@
 
 mse(ŷ, y) = sum((ŷ .- y).^2)/length(y)
 
-logloss(ŷ::AbstractVecOrMat, y::AbstractVecOrMat) =
+crossentropy(ŷ::AbstractVecOrMat, y::AbstractVecOrMat) =
   -sum(y .* log.(ŷ)) / size(y, 2)
+
+@deprecate logloss(x, y) crossentropy(x, y)

test/tracker.jl

@@ -15,7 +15,7 @@ gradtest(f, dims...) = gradtest(f, rand.(dims)...)
 @test gradtest(x -> softmax(x).*(1:3), (3,5))
 
 @test gradtest(Flux.mse, rand(5,5), rand(5, 5))
-@test gradtest(Flux.logloss, rand(5,5), rand(5, 5))
+@test gradtest(Flux.crossentropy, rand(5,5), rand(5, 5))
 
 @test gradtest(x -> x', rand(5))