Since tutorial and documentation doesn’t say much about this. Can someone tell me how to use the model developed in caffe or tensorflow to generate OpenCL code to be executed on FPGA?
Try compiling TVM with OpenCL enabled. Use target=“OpenCL” in build command option.