Please provide complete information as applicable to your setup.
**• Hardware Platform : GPU
**• DeepStream Version : 6.1.0
**• NVIDIA GPU Driver Version : 510.47.03
**• docker image : nvcr.io/nvidia/deepstream:6.1-triton
**• Issue Type : questions,
**• How to reproduce the issue : I use Gst-nvdspreprocess plugin customfill image.when I use Gst-nvinfer or Gst-nvinferserver with model_repo,I can get correct result. But when I use Gst-nvinferserver with grpc,I get wrong result. How do these two modes differ in processing?By the way, where can I find nvds_infer_server source code. For example, I can find and modify nvds_infer.
**• Requirement details: Gst-nvdspreprocess, Gst-nvinfer, Gst-nvinferserver