<feat>Support ultralytics classification

[315386775]

Main changes

1. Support classification of ultralytics' yolov11 and yolov8.
2. Render results (such as bounding boxes, masks, keypoints, and probabilities)

Code example

model = xinfer.create_model("ultralytics/yolov11n-cls", device="cuda", dtype="float16")
results = model.infer("https://ultralytics.com/images/bus.jpg")
model.render('./res')
results

@dnth new feature, please check it.

[dnth]

Thank you @315386775!