Kasture, Deeksha
Kasture, Deeksha
@vpirogov @jondea Thanks for the clarification.
Hi @jondea Can you share the steps to reproduce the error using `test_benchdnn_modeC_conv_3d_cpu`? For the particular test case `./tests/benchdnn/benchdnn --conv g1ic32id192oc1od1kd192pd0n"3d_conv_large_shape:2"` I'm not getting any errors and am able to...
Memory usage is high for BRGEMM Convolution, and we are analyzing it, but I am not able to identify the error as the tests are passing. # Test output make...
@jondea Yes, this approach can be used to extend the work for 128-bit support, just need to verify that all tests are passing and check if anything more needs to...