Deepstream 6.1 reference_graphs fail with TypeError: 'module' object is not callable

Repro:

  • Install Deepstream 6.1 (I have verified I can run the deepstream-6.1/sources/apps/sample_apps/deepstream-test1)
  • Install graph composer 2.0.0
  • registry repo sync -n ngc-public
  • Install deepstream reference graphs (deepstream-reference-graphs-6.1.deb)
  • Run the deepstream-test1 reference graph (or any graph):

$ /opt/nvidia/graph-composer/execute_graph.sh deepstream-test1.yaml parameters.yaml -d …/common/target_x86_64.yaml
Graphs: deepstream-test1.yaml,parameters.yaml
Target: …/common/target_x86_64.yaml

Running deepstream-test1.yaml

[INFO] Writing manifest to /tmp/ds.deepstream-test1/manifest.yaml
2022-08-29 14:13:17,209 - Registry - INFO - Installing graph …
2022-08-29 14:13:17,236 - Registry - INFO - Extension update found for extension: NvDsVisualizationExt
2022-08-29 14:13:17,236 - Registry - INFO - Old version: 1.0.0 New version: 1.0.1
Traceback (most recent call last):
File “registry/run.py”, line 10, in
File “registry/run.py”, line 7, in main
File “”, line 982, in main
File “”, line 969, in read_entry_and_execute
File “”, line 949, in execute
File “”, line 388, in execute_on
File “”, line 160, in call_subparsers
File “”, line 377, in install
File “”, line 305, in install_graph_with_dir
File “”, line 979, in install_graph_with_dir
File “”, line 907, in install_graph
File “”, line 233, in get_best_variants
File “”, line 272, in _compute_dependency_match
File “”, line 318, in _version_compare
TypeError: ‘module’ object is not callable
[3413807] Failed to execute script ‘run’ due to unhandled exception!


End deepstream-test1.yaml


[INFO] Graph installation directory /tmp/ds.deepstream-test1 and manifest /tmp/ds.deepstream-test1/manifest.yaml retained

Any pointers or help debugging appreciated–thanks!

I just started getting this error yesterday as well… Have they been working for you before this? I have my runs scripted, so I know that I am running them the same as I have been… I did do a registry repo sync between the last successful run and the first one that broke.

This is the topic that I opened:

During your sync, did you get a message like "2022-08-28 13:25:26,442 - Registry - WARNING - Skipping extension NvDsVisualizationExt version 1.0.2
2022-08-28 13:25:26,442 - Registry - WARNING - Failed to sync extension NvDsVisualizationExt version 1.0.2
"?

It did; but 1.0.0 and 1.0.1 seem to install ok.

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU)
• DeepStream Version
• JetPack Version (valid for Jetson only)
• TensorRT Version
• NVIDIA GPU Driver Version (valid for GPU only)

It’s just interesting that everything seemed to work fine until I started getting that message… It may be unrelated, but it’s suspicious.

Worth noting, The error that I am getting is on a dGPU setup on our deep stream/tao/graph composer workstation/server.

The closest common point is the registry update. (It’s nice that you could do it on a Jetson, because it makes it easy to repeat.)

@Fiona.Chen I rechecked the versions, and the Deepstream documentation appeared updated compared to a couple days ago (I was using Deepstream 6.1.0 / CUDA 11.6 whereas I now see Deepstream 6.1.1 / CUDA 11.7)

I reinstalled everything with Deepstream 6.1.1 and now it works. Thank you!

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