NVIDIA-AI-IOT / deepstream_triton_model_deploy

How to deploy open source models using DeepStream and Triton Inference Server
Apache License 2.0
74 stars 15 forks source link

Problem with running centerface model. #4

Closed rahulsharma11 closed 3 years ago

rahulsharma11 commented 3 years ago

Hi, I have installed tritonserver(docker image) and provided centerface model in "model_repository". Docker image successfully loaded the centerface model.

By running - "deepstream-app -c source1_primary_detector.txt"

I am getting error- ERROR: : Failed to create 'primary_gie' ERROR: : create_primary_gie_bin failed ERROR: : create_pipeline failed ERROR: : Failed to create pipeline Quitting App run failed

Please suggest if few other steps are left or not. Thanks

monjha commented 3 years ago

Hi, Could you please run GST_DEBUG=3 and then run the application to get more information on the error?

rahulsharma11 commented 3 years ago

Hi, Thanks for the reply.

With GST_DEBUG=3 , Following is the output-

0:00:00.039526510 22684 0x559877943610 WARN GST_ELEMENT_FACTORY gstelementfactory.c:456:gst_element_factory_make: no such element factory "nvinferserver"! ERROR: : Failed to create 'primary_gie' ERROR: : create_primary_gie_bin failed ERROR: : create_pipeline failed ERROR: : Failed to create pipeline Quitting App run failed

I can see in config-file, there is "nvinferserver" plugin-type in primary-gie. [primary-gie] enable=1

(0): nvinfer; (1): nvdsinfer_server

plugin-type=1

monjha commented 3 years ago

Hi Rahul,

I still don't see a complete log. Could you please run: GST_DEBUG=3 and then run deepstream -c source1_primary_detector.txt Thanks!

rahulsharma11 commented 3 years ago

Hi, I ran with the same. Error is highlighted for your reference-

0:00:00.039526510 22684 0x559877943610 WARN GST_ELEMENT_FACTORY gstelementfactory.c:456:gst_element_factory_make: no such element factory "nvinferserver"!

If you need i can run GST_DEBUG=4 and provide you logs...That is huge.

monjha commented 3 years ago

Thanks Rahul! Are you able to run any other deepstream test?

rahulsharma11 commented 3 years ago

Yes. But without docker environment

mjhuria commented 3 years ago

Hi Rahul,

What do you get when you run: gst-inspect-1.0 nvinferserver

If this has an error then you probably don't have /lib/gst-plugins/libnvdsgst_inferserver.so lib. What DS version are you using?

rahulsharma11 commented 3 years ago

Hi, I got "No such element or plugin 'nvinferserver' ".

I am using DS5 with x86.

mjhuria commented 3 years ago

You don't need to use the triton inference server docker image to run this application. Please use deepstream docker image from https://ngc.nvidia.com/catalog/containers/nvidia:deepstream tag: 5.0.1-20.09-triton

rahulsharma11 commented 3 years ago

ok thanks