Cannot run deepstream-test-1 in deepstream_python_apps: Where is the ../../../../samples/ folder?

Accelerated Computing Intelligent Video Analytics DeepStream SDK
5

  1. According to " Running Sample Applications" in your link (link, you should have cloned the git repo inside the sources directory. As the sample config paths are relative, they assume everything is located there. Cloning it in the suggested path should work right away, without requiring checking and changing every relative path.

  2. If you insist on running it elsewhere, check for the logs:

ERROR: Deserialize engine failed because file path: /home/marcus/dev/deepstream_python_apps/apps/deepstream-test1/../../../../samples/models/Primary_Detector/resnet10.caffemodel_b1_gpu0_int8.engine open error
0:00:03.809189017 12874     0x10ec5430 WARN                 nvinfer gstnvinfer.cpp:616:gst_nvinfer_logger:<primary-inference> NvDsInferContext[UID 1]: Warning from NvDsInferContextImpl::deserializeEngineAndBackend() <nvdsinfer_context_impl.cpp:1690> [UID = 1]: deserialize engine from file :/home/marcus/dev/deepstream_python_apps/apps/deepstream-test1/../../../../samples/models/Primary_Detector/resnet10.caffemodel_b1_gpu0_int8.engine failed
0:00:03.809270010 12874     0x10ec5430 WARN                 nvinfer gstnvinfer.cpp:616:gst_nvinfer_logger:<primary-inference> NvDsInferContext[UID 1]: Warning from NvDsInferContextImpl::generateBackendContext() <nvdsinfer_context_impl.cpp:1797> [UID = 1]: deserialize backend context from engine from file :/home/marcus/dev/deepstream_python_apps/apps/deepstream-test1/../../../../samples/models/Primary_Detector/resnet10.caffemodel_b1_gpu0_int8.engine failed, try rebuild
0:00:03.809302298 12874     0x10ec5430 INFO                 nvinfer gstnvinfer.cpp:619:gst_nvinfer_logger:<primary-inference> NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::buildModel() <nvdsinfer_context_impl.cpp:1715> [UID = 1]: Trying to create engine from model files

Here it says it was not able to find the “serialized” model-engine-file, so it will try to build it using the raw model. For this, it will need the proto-file, model-file, etc (and int8-calib-file if network-mode is for int8)

ERROR: Cannot access prototxt file '/home/marcus/dev/deepstream_python_apps/apps/deepstream-test1/../../../../samples/models/Primary_Detector/resnet10.prototxt'
ERROR: failed to build network since parsing model errors.
ERROR: failed to build network

Here it says it could not find the proto-file

In summary, just make sure to check for every relative path in the config file your using. To do this, you can cd into your folder where you’re running the app and check if files are reachable.

But i strongly suggest you clone it in the intended location:

git clone https://github.com/NVIDIA-AI-IOT/deepstream_python_apps /opt/nvidia/deepstream/deepstream/sources/deepstream_python_apps

(you’ll need probably need sudo for that)

then cd:

cd /opt/nvidia/deepstream/deepstream/sources/deepstream_python_apps/apps/deepstream-test1

then run your test:

python3 deepstream_test_1.py ~/test.264
4 Likes