Update README.md

README.md

@@ -32,11 +32,12 @@ To use the model for inference:
 from torchvision import transforms
 import torch
 from PIL import Image
-
+from huggingface_hub import hf_hub_download
 device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 
 # Load model
-model = torch.load("PercrptCLIP_Memorability.pth").to(device).eval()
+model_path = hf_hub_download(repo_id="PerceptCLIP/PerceptCLIP_Memorability", filename="perceptCLIP_Memorability.pth")
+model = torch.load(model_path).to(device).eval()
 
 # Load an image
 image = Image.open("image_path.jpg").convert("RGB")