| commit | 27400b90a9cb3fe028c5b724b58ce0e82d89b5e8 | [log] [tgz] |
|---|---|---|
| author | Michele Di Giorgio <michele.digiorgio@arm.com> | Thu Nov 01 13:44:05 2018 +0000 |
| committer | Georgios Pinitas <georgios.pinitas@arm.com> | Tue Nov 13 14:49:34 2018 +0000 |
| tree | 4b7dd9d4b29653ada018172dae826fe3e6ef5e08 | |
| parent | bb081cac4f386eb6db6e9927fce27c7027dd7be5 [diff] [blame] |
COMPMID-1707: Create 3 special CLWidthConcatenate kernel to concatenate 2/4 and 8 tensors (Part 1) Creating special cases for concatening 2 and 4 tensors. Change-Id: I6a739a494ae45011acb65369e353f9ef96970b90
diff --git a/tests/validation/fixtures/WidthConcatenateLayerFixture.h b/tests/validation/fixtures/WidthConcatenateLayerFixture.h index caad0fe..1f79210 100644 --- a/tests/validation/fixtures/WidthConcatenateLayerFixture.h +++ b/tests/validation/fixtures/WidthConcatenateLayerFixture.h
@@ -52,7 +52,7 @@ { // Create input shapes std::mt19937 gen(library->seed()); - std::uniform_int_distribution<> num_dis(2, 4); + std::uniform_int_distribution<> num_dis(2, 8); const int num_tensors = num_dis(gen); std::vector<TensorShape> shapes(num_tensors, shape);