Hi, I notice that onnc just can support nvdla full version.
But the full version of nvdla was too large to mapping into a FPGA Platform, so when I use nvdla backend to generate a loadable but how can I evaluate this loadable can work correctly?
DId ONNC use the QEMU to evaluate this? If so, It is impossile to debug the runtime code, and the runtime was too long.
Hi, I notice that onnc just can support nvdla full version.
But the full version of nvdla was too large to mapping into a FPGA Platform, so when I use nvdla backend to generate a loadable but how can I evaluate this loadable can work correctly?
DId ONNC use the QEMU to evaluate this? If so, It is impossile to debug the runtime code, and the runtime was too long.