I’m having a hard time figuring out why this works: using Zygote W, b = rand(2, 3), rand(2) predict(x) = W*x .+ b g = gradient(() -> sum(predict([1,2,3])), Params([W, b])) g[W], g[b] but this doesn’t: using Zygote a = 2 x = 2 f(x) = x^a gp = gradient(() -> f(x), Params(a)) gp[a] I get the error: …

gradient((a1,a2)->F(L1,L2; a1=a1,a2=a2), a1,a2) should do the trick, no?

Help with Zygote and parameters

amrods July 1, 2020, 9:05am 6

That works. Then I’m unsure when it is necessary to employ Params. Could you explain a little about that?

Topic		Replies	Views
Allocations in Zygote for a simple derivative Machine Learning performance , zygote	0	682	February 19, 2020
Zygote - parametrize matrix such that gradient is only performed on selected coefficients Machine Learning question , zygote , matrices , natural-gradient , autodiff	1	546	June 5, 2021
Need some help in understanding zygote gradient Machine Learning	2	413	September 7, 2022
Zygote dropgrad for all function numerical arguments that are non-differentiable (e.g., Int) General Usage zygote	1	504	June 25, 2021
Compute gradient of gradient norm using zygote New to Julia zygote	17	2013	August 26, 2022