Secondary classification on triton

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU) GPU
• DeepStream Version 6.1.1
• TensorRT Version NGC docker
• NVIDIA GPU Driver Version (valid for GPU only) 525
• Issue Type( questions, new requirements, bugs) Question

I’m trying to implement secondary gie on triton server and connect from deepstream with nvinferserver but I have not been able to find a working example online. I already have a yolov7 working as a primary gie.
I have come up with the config file below, but I’m getting an error. I’d appreciate any help.


infer_config {
  unique_id: 3
  gpu_ids: [0]
  max_batch_size: 1
  backend {
    triton {
      model_name: "vehicle-class"
      version: -1
      grpc {
        url: "localhost:8001"
        enable_cuda_buffer_sharing: false

  preprocess {
    network_format: IMAGE_FORMAT_RGB
    tensor_order: TENSOR_ORDER_LINEAR
    maintain_aspect_ratio: 1
    frame_scaling_hw: FRAME_SCALING_HW_DEFAULT
    frame_scaling_filter: 1
    normalize {
        scale_factor: 1.0

  postprocess {
    labelfile_path: "/app/model/vehicle-class/labels.txt"

  extra {
    copy_input_to_host_buffers: false
input_control {
  operate_on_gie_id: 1
  operate_on_class_ids: [2,7]
  async_mode: true


2023-02-28 11:00:10.499 | SUCCESS  | Pipeline started
2023-02-28 11:00:10.500 | WARNING  | Warning: gst-library-error-quark: NvInferServer asynchronous mode is applicable for secondaryclassifiers only. Turning off asynchronous mode (5) <<-->> gstnvinferserver_impl.cpp(390): validatePluginConfig (): /GstPipeline:pipeline0/GstBin:infer-1/GstNvInferServer:vehicle-class-infer-1
2023-02-28 11:00:10.500 | ERROR    | Error: gst-resource-error-quark: Failed to initialize InferTrtIsContext (1) <<-->> gstnvinferserver_impl.cpp(547): start (): /GstPipeline:pipeline0/GstBin:infer-1/GstNvInferServer:vehicle-class-infer-1:
Config file path: /app/gies/vehicle-class.txt
2023-02-28 11:00:10.500 | WARNING  | Pipeline stopped

There is no update from you for a period, assuming this is not an issue anymore. Hence we are closing this topic. If need further support, please open a new one. Thanks

please refer to deepstream SDK native sample opt\nvidia\deepstream\deepstream-6.1.1\samples\configs\deepstream-app-triton-grpc. it supports nvinferserver pgie+sgie.

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