gistfile1.txt

09:23:52 Running build/test/test_tc_mapper_bugs
09:23:52 [==========] Running 18 tests from 7 test cases.
09:23:52 [----------] Global test environment set-up.
09:23:52 [----------] 9 tests from TensorDot_32_512_8_2_28_28
09:23:52 [ RUN      ] TensorDot_32_512_8_2_28_28.BaseCorrect
09:23:53 [       OK ] Caffe2Test.TcGroupConvolutionOp (4002 ms)
09:23:53 [ RUN      ] Caffe2Test.TcConvolutionOp
09:23:56 [       OK ] TensorDot_32_512_8_2_28_28.BaseCorrect (4525 ms)
09:23:56 [ RUN      ] TensorDot_32_512_8_2_28_28.ReductionUnroll
09:23:59 [       OK ] Caffe2Test.TcConvolutionOp (6669 ms)
09:23:59 [ RUN      ] Caffe2Test.TcBatchMatmul
09:24:03 [       OK ] Caffe2Test.TcBatchMatmul (3346 ms)
09:24:03 [ RUN      ] Caffe2Test.TcFunctions
09:24:05 /var/lib/jenkins/workspace/test/test_harness.h:283: Failure
09:24:05 The difference between Texpected.data<float>()[i + offsetInExpected] and Tactual.data<float>()[i + offsetInActual] is 0.87766313552856445, which exceeds relativePrecision * Texpected.data<float>()[i + offsetInExpected], where
09:24:05 Texpected.data<float>()[i + offsetInExpected] evaluates to 0.87766313552856445,
09:24:05 Tactual.data<float>()[i + offsetInActual] evaluates to 0, and
09:24:05 relativePrecision * Texpected.data<float>()[i + offsetInExpected] evaluates to 8.7766312617532094e-07.