yolact-tvm-cpu-optimization.json

{
    "model_wrapper":
    {
        "type": "kenning.modelwrappers.instance_segmentation.yolact.YOLACT",
        "parameters":
        {
            "model_path": "kenning:///models/instance_segmentation/yolact.onnx"
        }
    },
    "optimizers":
    [
        {
            "type": "kenning.optimizers.tvm.TVMCompiler",
            "parameters":
            {
                "model_framework": "onnx",
                "target": "llvm -mcpu=core-avx2",
                "opt_level": 3,
                "compile_use_vm": false,
                "output_conversion_function": "default",
                "target_host": null,
                "compiled_model_path": "./build/compiled-model.so"
            }
        }
    ]
}