KeypointRCNN_vaiq_int8 model support

[AmosLewis]

Failed op:

• [e2e][onnx][model] onnx.Pad SHARK-ModelDev#617
• onnx.Transpose SHARK-ModelDev#651

[PhaneeshB]

After llvm/torch-mlir#3214
Next failure is :

KeypointRCNN_vaiq_int8.default.torch-onnx.mlir:7854:13: error: failed to legalize operation 'torch.operator' that was explicitly marked illegal
    %7310 = torch.operator "onnx.Transpose"(%7309#0) {torch.onnx.perm = [0 : si64, 2 : si64, 1 : si64]} : (!torch.vtensor<[],f32>) -> !torch.vtensor<[],f32> 
            ^

[PhaneeshB]

Model also has - NonMaxSuppressionOp

%7685 = torch.operator "onnx.NonMaxSuppression"(%7678, %7682, %7683, %7684) : (!torch.vtensor<[1,?,4],f32>, !torch.vtensor<[1,1,?],f32>, !torch.vtensor<[1],si64>, !torch.vtensor<[1],f32>) -> !torch.vtensor<[?,3],si64>

[stellaraccident]

NonMaxSuppression will likely require a custom lowering to a specific backend.