Would be nice if anyone could help us benchmark! Our primary focus though is not model performance, but to demonstrate the capability that TVM Unity generates c

Would be nice if anyone could help us benchmark! Our primary focus though is not model performance, but to demonstrate the capability that TVM Unity generates code targeting WebGPU and allows them to run with client GPUs :-)