ONNX http://onnx.ai/supported-tools page say it is supported by Nvidia TensorRT but I cant find any documentation or tutorial available describing import of ONNX model in TensorRT.
although TensorRT-3 user guide has documentation on importing UFF model in TensorRT but no documentation on importing ONNX model in TensorRT
Hi dusty_nv, can you estimate when the next version of tensorrt with support for onnx would be ready? Our team is trying to put some pytorch model into tensorrt, we are thinking if we should wait for the next version (since pytorch can export to onnx) or write our own parser for pytorch now.