[ RUN      ] Test_ONNX_layers.GatherMultiOutput/1, where GetParam() = OCV/OCL_FP16
/build/precommit_linux64-avx2/opencv/modules/dnn/test/test_common.impl.hpp:71: Failure
Expected: (normL1) <= (l1), actual: nan vs 0.004
  |ref| = 0
[  FAILED  ] Test_ONNX_layers.GatherMultiOutput/1, where GetParam() = OCV/OCL_FP16 (1 ms)

[ RUN      ] Test_ONNX_layers.GatherMultiOutput/1, where GetParam() = OCV/OCL_FP16
[ INFO:0] global C:\build\master-win64-vc15\opencv\modules\dnn\src\onnx\onnx_importer.cpp (395) cv::dnn::dnn4_v20201117::ONNXImporter::populateNet DNN/ONNX: loading ONNX v6 model produced by 'pytorch':1.7. Number of nodes = 9, inputs = 1, outputs = 1
c:\build\master-win64-vc15\opencv\modules\dnn\test\test_common.impl.hpp(71): error: Expected: (normL1) <= (l1), actual: nan vs 0.004
  |ref| = 0
[ INFO:0] global C:\build\master-win64-vc15\opencv\modules\ts\src\ts.cpp (847) cvtest::testTearDown Memory_usage (OpenCL): 100 (base=32  current=64)
[  FAILED  ] Test_ONNX_layers.GatherMultiOutput/1, where GetParam() = OCV/OCL_FP16 (2 ms)