Could anyone help me understand why calling Flux.params() inside a gradient, changes the gradient? Below is a minimal example where, in get_grad_2, I instantiate a variable prowling_variable = Flux.params(m). The gradients returned in g1 are all 0.0, but g2 outputs all 1.0. It is as if the loss has…

Calling Flux.params() inside gradient changes output?

ToucheSir September 28, 2021, 6:32pm 3

This is almost certainly a bug. I dropped an MWE on the most likely PR culprit, so feel free to follow/comment on there.

1 Like

Topic		Replies	Views
Why calculating gradients from Params is different than doing it directly? General Usage flux	5	751	April 28, 2020
Flux/Zygote: Gradient with respect to inputs and implicit parameters (in 2021) Machine Learning question , flux , zygote	1	986	November 23, 2021
Flux.jl: params() and gradient() ocnfusion Machine Learning	4	682	August 23, 2021
Understanding Flux.jl use of `gradient` and `params` Machine Learning flux	4	3626	October 2, 2021
Weight regularisation which iterates params(m) in Flux -- Mutating arrays is not supported Machine Learning flux , zygote	7	540	July 9, 2021