The default models provided for object detection examples not working. Failed Serialization

Autonomous Vehicles DRIVE AGX Orin DRIVE AGX Orin General
1

Hi Community,

I am trying to run the default object_detector_sample provided with driveworks 5.6 but am running into the below given error.

ProgramArguments: Missing argument 'dwTracePath' requested
[12-09-2022 12:06:43] Platform: Detected Generic x86 Platform
[12-09-2022 12:06:43] TimeSource: monotonic epoch time offset is 1662957855557878
[12-09-2022 12:06:43] Adding variable DW_Base:DW_Version
[12-09-2022 12:06:43] Addded variable DW_Base:DW_Version
[12-09-2022 12:06:43] Platform: number of GPU devices detected 1
[12-09-2022 12:06:43] Platform: currently selected GPU device 0, Resource Data Dir: trt_08_04_11_06, Arch: ga1xx-discrete
[12-09-2022 12:06:43] Platform: currently selected GPU device discrete ID 0
[12-09-2022 12:06:43] Context::mountResourceCandidateDataPath resource FAILED to mount from './resources': VirtualFileSystem: Failed to mount './resources/resources.pak'
[12-09-2022 12:06:43] Context::mountResourceCandidateDataPath resource FAILED to mount from '/home/minuszero/driveworks-copy/samples/build/data': VirtualFileSystem: Failed to mount '/home/asdasdasd/driveworks-copy/samples/build/data/resources.pak'
[12-09-2022 12:06:43] Context::findDataRootInPathWalk data/DATA_ROOT found at: /usr/local/driveworks/data
[12-09-2022 12:06:43] Context::mountResourceCandidateDataPath resource FAILED to mount from '/usr/local/driveworks/data': VirtualFileSystem: Failed to mount '/usr/local/driveworks/data/resources.pak'
[12-09-2022 12:06:43] Context::findDataRootInPathWalk data/DATA_ROOT found at: /usr/local/driveworks-5.6/data
[12-09-2022 12:06:43] Context::mountResourceCandidateDataPath resource FAILED to mount from '/usr/local/driveworks-5.6/data': VirtualFileSystem: Failed to mount '/usr/local/driveworks-5.6/data/resources.pak'
[12-09-2022 12:06:43] Context::findResourcesPackageInPathWalk: Could not find ./resources/resources.pak in upto 7 parent directories from /usr/local/driveworks/targets/x86_64-Linux/lib/libdw_base.so.5.6
[12-09-2022 12:06:43] Context::findResourcesPackageInPathWalk: Could not find ./resources/resources.pak in upto 7 parent directories from /usr/local/driveworks-5.6/targets/x86_64-Linux/lib/libdw_base.so.5.6
[12-09-2022 12:06:43] SDK: No resources(.pak) mounted, some modules will not function properly
[12-09-2022 12:06:43] TimeSource: monotonic epoch time offset is 1662957855557878
[12-09-2022 12:06:43] Initialize DriveWorks SDK v5.6.83
[12-09-2022 12:06:43] Release build with GNU 9.3.0 from buildbrain-branch-0-gdcc2cf22fff
[12-09-2022 12:06:43] Initialize DriveWorks VisualizationSDK v5.6.83
[12-09-2022 12:06:43] Initialize DriveWorksGL SDK v5.6.83
[12-09-2022 12:06:43] GL-SDK: initialize OpenGL
[12-09-2022 12:06:43] SensorFactory::createSensor() -> camera.virtual, offscreen=0,profiling=1,tensorRT_model=,video=/home/asdasdasd/driveworks-copy/data/samples/sfm/triangulation/video_0.h264
[12-09-2022 12:06:43] CameraVirtual: defaulting to non SIPL
[12-09-2022 12:06:43] CameraBase: pool size set to 8
[12-09-2022 12:06:43] CameraNVCUVID: no seek table found at /home/asdasdasd/driveworks-copy/data/samples/sfm/triangulation/video_0.h264.seek, seeking is not available.
SimpleCamera: Camera image: 1280x800
Camera image with 1280x800 at 30 FPS
[12-09-2022 12:06:43] [MemUsageChange] Init CUDA: CPU +364, GPU +0, now: CPU 516, GPU 1115 (MiB)
[12-09-2022 12:06:44] Loaded engine size: 24 MiB
[12-09-2022 12:06:44] 1: [stdArchiveReader.cpp::StdArchiveReader::40] Error Code 1: Serialization (Serialization assertion stdVersionRead == serializationVersion failed.Version tag does not match. Note: Current Version: 213, Serialized Engine Version: 205)
[12-09-2022 12:06:44] 4: [runtime.cpp::deserializeCudaEngine::49] Error Code 4: Internal Error (Engine deserialization failed.)
[12-09-2022 12:06:44] Driveworks exception thrown: DW_DNN_INVALID_MODEL: DNN: Unable to load model.

It seems the model provided with the sample was serialized with some old version of tensorrt optimization tool.

I would like to know the model name and architecture so that I can convert it using the latest tensorrt optimization tool and test it,

My own custom model is working fine as I converted it using new optimization tool.

Thanks.

1 Like
3

Dear @priyam1,
This topic looks related to DRIVE AGX ORIN platform. If so, please post at DRIVE AGX Orin - NVIDIA Developer Forums .
There are some issues with DW 5.6 samples and we recommend to wait until next release.
May I know if this issue block your development?

4

Hi @SivaRamaKrishnaNV , I had tagged this issue as part of AGX Orin but one of the moderators changed it to AGX Xaviers and AGX General.

Anyways, I appreciate your help. I cannot go into detail but yes, this issue is blocking our development as we need a reference/baseline for our model performance as well as a better understanding of the entire DNN pipeline in Driveworks. We plan on getting that using the default model.

8

Dear @priyam1,
Please reinstall the DRIVE OS 6.0.4 + DW 5.6 and test.

10

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.