Please provide complete information as applicable to your setup.
• Hardware Platform (Jetson / GPU)
NVIDIA RTX A4500
• DeepStream Version
7.1
• JetPack Version (valid for Jetson only)
• TensorRT Version
default tensorrt in nvcr.io/nvidia/deepstream:7.1-triton-multiarch
• NVIDIA GPU Driver Version (valid for GPU only)
Driver Version: 535.161.08
CUDA Version: 12.2
• Issue Type( questions, new requirements, bugs)
questions
• How to reproduce the issue ? (This is for bugs. Including which sample app is using, the configuration files content, the command line used and other details for reproducing)
hello，
I am using the Python backend in Triton Server, and after processing, I return custom JSON data. Now, I need to retrieve the JSON data in the Probe. Is this approach feasible, or do I need to modify the nvinferserver source code to accept custom results? or can I modify the source code of custom_lib to accept custom data?
python backend code：
data = {
“video_has_board”: “ok”,
“video_distance”: self.video_distance,
“video_angle_ok”: self.video_angle_ok,
“focal_length”: self.focal_length,
“normal_vector”: self.normal_vector.tolist(),
“D”: self.D.tolist()
}
result = json.dumps(data, ensure_ascii=False, separators=(‘,’,‘:’))
out = np.array(result)
out_tensor = pb_utils.Tensor(‘results’, out.astype(output_dtype))
inference_response = pb_utils.InferenceResponse(output_tensors=[out_tensor])
responses.append(inference_response)
return responses
How can the custom responses mentioned above be retrieved through the nvinferserver component or other methods?