diff --git a/.github/workflows/Documenter.yml b/.github/workflows/Documenter.yml
index 172c0dcff..e3df5d1d8 100644
--- a/.github/workflows/Documenter.yml
+++ b/.github/workflows/Documenter.yml
@@ -12,10 +12,8 @@ jobs:
     name: Documentation
     runs-on: ubuntu-latest
     steps:
-      - uses: actions/checkout@v2
-      - uses: julia-actions/setup-julia@v1
-        with:
-          version: '1'
+      - uses: actions/checkout@v3
+      - uses: julia-actions/setup-julia@latest
       - name: Install dependencies
         run: |
           julia --project=docs -e '
diff --git a/src/GeometricMachineLearning.jl b/src/GeometricMachineLearning.jl
index cc6d70674..8d893d7a0 100644
--- a/src/GeometricMachineLearning.jl
+++ b/src/GeometricMachineLearning.jl
@@ -8,7 +8,6 @@ module GeometricMachineLearning
     using GeometricBase
     using GeometricEquations
     using GeometricIntegrators
-    using GPUArrays
     using KernelAbstractions
     using LinearAlgebra
     using NNlib
@@ -19,8 +18,6 @@ module GeometricMachineLearning
     using InteractiveUtils
     using TimerOutputs
 
-    import CUDA
-
     import AbstractNeuralNetworks: Architecture, Model, AbstractExplicitLayer, AbstractExplicitCell, AbstractNeuralNetwork , NeuralNetwork
     import AbstractNeuralNetworks: Chain, GridCell
     import AbstractNeuralNetworks: Dense, Linear, Recurrent
@@ -227,7 +224,6 @@ module GeometricMachineLearning
     export Batch, optimize_for_one_epoch!
     include("data_loader/tensor_assign.jl")
     include("data_loader/matrix_assign.jl")
-    include("data_loader/data_loader.jl")
     include("data_loader/mnist_utils.jl")
     include("data_loader/batch.jl")
 
diff --git a/src/manifolds/stiefel_manifold.jl b/src/manifolds/stiefel_manifold.jl
index 4149dcd2d..ec1c3fb26 100644
--- a/src/manifolds/stiefel_manifold.jl
+++ b/src/manifolds/stiefel_manifold.jl
@@ -58,10 +58,6 @@ end
 
 Base.:*(Y::StiefelManifold, B::AbstractMatrix) = Y.A*B
 Base.:*(B::AbstractMatrix, Y::StiefelManifold) = B*Y.A
-#this is needed for the implementation of MultiHeadAttention
-function Base.:*(Y::Adjoint{T, StiefelManifold{T, AT}}, B::AbstractMatrix) where {T, AT<:AbstractGPUMatrix{T}}
-    Y.parent.A'*B 
-end
 
 function Base.:*(Y::Adjoint{T, StiefelManifold{T, AT}}, B::AbstractMatrix) where {T, AT<:AbstractMatrix{T}}
     Y.parent.A'*B