# How to sum the result of mapslices on the GPU while tracking gradients?

**URL:** <https://discourse.julialang.org/t/how-to-sum-the-result-of-mapslices-on-the-gpu-while-tracking-gradients/105310>\
**Category:** General Usage\
**Tags:** gpuarrays, zygote, mapslices\
**Created:** [October 23, 2023, 11:52am UTC](https://discourse.julialang.org/t/how-to-sum-the-result-of-mapslices-on-the-gpu-while-tracking-gradients/105310 "2023-10-23T11:52:49Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![taotree](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/taotree/32/31982_2.png) [@taotree](https://discourse.julialang.org/u/taotree)\
**Post date:** [October 23, 2023, 11:52am UTC](https://discourse.julialang.org/t/how-to-sum-the-result-of-mapslices-on-the-gpu-while-tracking-gradients/105310/1 "2023-10-23T11:52:49Z")

</div>

I need to do mapslices but need it to support gradient tracking for Zygote. So, I’m trying to use SliceMap.slicemap. I can’t use SliceMap.mapcols because I have parameters inside the function. SliceMap.slicemap returns a JuliennedArrays.Align. sum works on a CuArray, but on a JuliennedArrays.Align, it gives a scalar indexing error. Is there some solution to this?

Example (this results in scalar index error):

```julia
using CUDA
import SliceMap, Zygote, Flux
CUDA.allowscalar(false)
function func(m, k, data)
    y = SliceMap.slicemap(m; dims=1) do col
        col .* k
    end
    return sum(y) # I think this line gives the scalar indexing error
end
function test_gpu()
    m = fill(1.0, 3, 10) |> Flux.gpu
    data = fill(2.0, 3, 10) |> Flux.gpu
    k = [0.7, 0.5, 0.5] |> Flux.gpu
    Zygote.gradient((m, k) -> func(m, k, data), m, k)
end

```