COMPMID-3694 COMPMID-3695 COMPMID-3458: Softmax Axis

* Properly support "axis" in CL and NEON (and GC) SoftmaxLayer and
LogSoftmaxLayer in accord with mainstream frameworks. Axis now defines
the dimension on which softmax is performed, and supports the range
[-rank, rank)

* Extend validation tests to include valid and invalid axes

* Remove unnecessary LogSoftmaxLayer fixture, as it is only a
specialisation of the SoftmaxLayer fixture

* Change the validation fill value range from [-1000, 1000] to [-10,
10], as the former often results in sparse outputs with a single one and
zeros elsewhere

Change-Id: I8a0040453182b04ed88260de3ba434e98258d863
Signed-off-by: Manuel Bottini <manuel.bottini@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/3830
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>

1 files changed, 4 insertions, 4 deletions

diff --git a/src/core/CL/kernels/CLPermuteKernel.cpp b/src/core/CL/kernels/CLPermuteKernel.cpp
index 1636e5a1bc..dc2d6fe4b4 100644
--- a/src/core/CL/kernels/CLPermuteKernel.cpp
+++ b/src/core/CL/kernels/CLPermuteKernel.cpp
@@ -75,16 +75,16 @@ void CLPermuteKernel::configure(const ICLTensor *input, ICLTensor *output, const
 void CLPermuteKernel::configure(const CLCompileContext &compile_context, const ICLTensor *input, ICLTensor *output, const PermutationVector &perm)
 {
     ARM_COMPUTE_ERROR_ON_NULLPTR(input, output);
+    const TensorShape output_shape = get_output_shape(input->info(), perm);
+    // Output auto inizialitation if not yet initialized
+    auto_init_if_empty(*output->info(), input->info()->clone()->set_tensor_shape(output_shape));
+
     ARM_COMPUTE_ERROR_THROW_ON(validate_arguments(input->info(), output->info(), perm));
 
     _input  = input;
     _output = output;
     _perm   = perm;
 
-    const TensorShape output_shape = get_output_shape(input->info(), perm);
-    // Output auto inizialitation if not yet initialized
-    auto_init_if_empty(*output->info(), input->info()->clone()->set_tensor_shape(output_shape));
-
     // Create kernel
     CLBuildOptions build_opts;
     build_opts.add_option("-DDATA_TYPE=" + get_cl_unsigned_type_from_element_size(data_size_from_type(input->info()->data_type())));