Missing "resnet10.caffemodel_b8_fp16.engine" model-engine-file

Autonomous Machines Jetson & Embedded Systems Jetson Nano
1

I am trying to get the “source8_1080p_dec_infer-resnet_tracker_tiled_display_fp16_nano.txt” to run.
However, it hangs with this console output:

testuser@jetson:/opt/nvidia/deepstream/deepstream-4.0/samples/configs/deepstream-app$ deepstream-app -c source8_1080p_dec_infer-resnet_tracker_tiled_display_fp16_nano.txt 
Creating LL OSD context new
gstnvtracker: Loading low-level lib at /opt/nvidia/deepstream/deepstream-4.0/lib/libnvds_mot_klt.so
gstnvtracker: Optional NvMOT_RemoveStreams not implemented
gstnvtracker: Batch processing is OFF
0:00:03.914162582 13069     0x3cc4e4a0 WARN                 nvinfer gstnvinfer.cpp:515:gst_nvinfer_logger:<primary_gie_classifier> NvDsInferContext[UID 1]:useEngineFile(): Failed to read from model engine file
0:00:03.914272839 13069     0x3cc4e4a0 INFO                 nvinfer gstnvinfer.cpp:519:gst_nvinfer_logger:<primary_gie_classifier> NvDsInferContext[UID 1]:initialize(): Trying to create engine from model files

I checked if the required model-engine-file “model-engine-file=…/…/models/Primary_Detector_Nano/resnet10.caffemodel_b8_fp16.engine” exists and it is missing!
I only got these in the respective models directory:

drwxrwxrwx 2 root root    4096 Sep 22 21:03 ./
drwxrwxrwx 8 root root    4096 Sep 22 21:03 ../
-rwxrwxrwx 1 root root      28 Sep 11 12:15 labels.txt*
-rwxrwxrwx 1 root root 6244865 Sep 11 12:15 resnet10.caffemodel*
-rwxrwxrwx 1 root root    7605 Sep 11 12:15 resnet10.prototxt*

Does anybody know what i need to do to get the “resnet10.caffemodel_b8_fp16.engine” file?

2

Hi,

You don’t need a model-engine-file to run the deepstream sample. It’s a harmless warning.

On the first launch, deepstream will try to compile the TensorRT file on your environment .
After that, the engine file can be found on your environment and can be launched directly.

<b>WARN</b> nvinfer gstnvinfer.cpp:515:gst_nvinfer_logger:<primary_gie_classifier> NvDsInferContext[UID 1]:useEngineFile(): Failed to read from model engine file
<b>INFO</b> nvinfer gstnvinfer.cpp:519:gst_nvinfer_logger:<primary_gie_classifier> NvDsInferContext[UID 1]:initialize(): Trying to create engine from model files

Do you meet any error with deepstream-app?
It looks like the deepstream-app run successfully on your environment.

Thanks.

3

Thank you for clarification.
It finally got generated - just needed some more time for compilation.