Triton server getting error

i am using this code for infrencing that is available on ultralyics site:

import contextlib
import tritonclient.http as httpclient #pip install if needed
from ultralytics import YOLO
import subprocess
import time
import cv2
from tritonclient.http import InferenceServerClient
import os

setup triton inference client

triton_client = InferenceServerClient(url=‘localhost:8000’, verbose=False, ssl=False)

Load the Triton Server model

model = YOLO(f’http://localhost:8000/yolov8n’, task=‘detect’)

for filename in os.listdir(directory):

video_path=os.path.join(directory, filename)
cap = cv2.VideoCapture(video_path)

# Loop through the video frames
while cap.isOpened():
    # Read a frame from the video
    success, frame =

    if success:
        # Run YOLOv8 inference on the frame
        results = model(frame)

        # Visualize the results on the frame
        annotated_frame = results[0].plot()

        # Display the annotated frame
        cv2.imshow("YOLOv8 Inference", annotated_frame)

        # Break the loop if 'q' is pressed
        if cv2.waitKey(1) & 0xFF == ord("q"):
        # Break the loop if the end of the video is reached

# Release the video capture object and close the display window

if i export model in default settings i.e. (model.export(onnx)) than the infrencing works fine. i am getting results to any size of input .by default input is =640 (specified by ultraytics)

but i export model specifing size (model.export(onnx, imgsz=800)) and
than run on triton server it gives errors:
tritonclient.utils.InferenceServerException: [400] [request id: <id_unknown>] unexpected shape for input ‘images’ for model ‘yolov8n’. Expected [1,3,800,800], got [1,3,640,640]

for config.pbxt i am using:

name: “yolov8n”
platform: “tensorrt_plan”
max_batch_size: 0
input [
name: “images”
data_type: TYPE_FP32
dims: [1, 3, 800,800]
output [
name: “output0”
data_type: TYPE_FP32
dims: [1, 84, 13125]
instance_group [
kind: KIND_GPU,
count: 1

any solution of this problem?