Load using from_pretrained

This PR simplifies the loading of the checkpoints by leveraging the PyTorchModelHubMixin class.

app.py

@@ -2,12 +2,7 @@ import gradio as gr
 from ultralytics import YOLOv10 
 import supervision as sv
 import spaces
-from huggingface_hub import hf_hub_download
 
-
-def download_models(model_id):
-    hf_hub_download("kadirnar/Yolov10", filename=f"{model_id}", local_dir=f"./")
-    return f"./{model_id}"
     
 box_annotator = sv.BoxAnnotator()
 category_dict = {
@@ -33,7 +28,7 @@ category_dict = {
 @spaces.GPU(duration=200)
 def yolov10_inference(image, model_id, image_size, conf_threshold, iou_threshold):
     model_path = download_models(model_id)
-    model = YOLOv10(model_path)
+    model = YOLOv10.from_pretrained(f"jameslahm/{model_id}")
     results = model(source=image, imgsz=image_size, iou=iou_threshold, conf=conf_threshold, verbose=False)[0]
     detections = sv.Detections.from_ultralytics(results)
     
@@ -54,14 +49,14 @@ def app():
                 model_id = gr.Dropdown(
                     label="Model",
                     choices=[
-                        "yolov10n.pt",
-                        "yolov10s.pt",
-                        "yolov10m.pt",
-                        "yolov10b.pt",
-                        "yolov10l.pt",
-                        "yolov10x.pt",
+                        "yolov10n",
+                        "yolov10s",
+                        "yolov10m",
+                        "yolov10b",
+                        "yolov10l",
+                        "yolov10x",
                     ],
-                    value="yolov10m.pt",
+                    value="yolov10m",
                 )
                 image_size = gr.Slider(
                     label="Image Size",