barry-jin opened a new issue #20307: URL: https://github.com/apache/incubator-mxnet/issues/20307
## Description Tracking c++ inference issue in master/cpp-package https://github.com/apache/incubator-mxnet/issues/19550#issuecomment-841583103 ### Error Message ``` root@6da7899c2de8:/work/mxnet/cpp-package/example/inference# ./imagenet_inference --symbol_file "./model/Inception-BN-symbol.json" --params_file "./model/Inception-BN-0126.params" --batch_size 16 --num_inference_batches 500 --benchmark --enableTRT [01:51:07] ../cpp-package/example/inference/imagenet_inference.cpp:302: Loading the model from ./model/Inception-BN-symbol.json [01:51:07] ../src/nnvm/legacy_json_util.cc:209: Loading symbol saved by previous version v0.8.0. Attempting to upgrade... [01:51:07] ../src/nnvm/legacy_json_util.cc:217: Symbol successfully upgraded! [01:51:07] ../cpp-package/example/inference/imagenet_inference.cpp:317: Loading the model parameters from ./model/Inception-BN-0126.params [01:51:07] ../src/storage/storage.cc:199: Using Pooled (Naive) StorageManager for CPU [01:51:09] ../src/storage/storage.cc:199: Using Pooled (Naive) StorageManager for GPU terminate called after throwing an instance of 'dmlc::Error' what(): [01:51:09] ../cpp-package/include/mxnet-cpp/symbol.hpp:268: MXNetError: Error in operator TensorRT0: [01:51:09] ../src/operator/subgraph/tensorrt/tensorrt.cc:61: Check failed: idx_g.input_nodes().size() == in_shapes->size() + params_map.size() (55 vs. 1) : Aborted (core dumped) ``` ## To Reproduce $ ./imagenet_inference --symbol_file "./model/Inception-BN-symbol.json" --params_file "./model/Inception-BN-0126.params" --batch_size 16 --num_inference_batches 500 --benchmark --enableTRT -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
