GitLab

disable quant_dot test for CPU backend, it requires removing downcast converts in "ref"implementation. Inside the "ref" impelmentation, it downcasting one of the inputs of the GEMM from float to fp8. Which is a lossy conversion and while computing gemm, it upcasts it back to  float.  CPU backend removes converts completely since it is going from "float->fp8->float". due to lossy cast, results are coming out different.

Updated onnx.proto files to be same as onnx==1.14.1 that is used by MIGraphX.   Added more switch cases in onnx parser to handle FP8 dtypes.

Merge branch 'rocblas_mlir_fp8' of github.com:ROCmSoftwarePlatform/AMDMIGraphX into rocblas_mlir_fp8