use hf_hub_download in place of direct URL download

app.py

@@ -12,7 +12,7 @@ import re
 from sentence_transformers import SentenceTransformer
 from sklearn.metrics.pairwise import cosine_similarity
 from tavily import TavilyClient
-from huggingface_hub import InferenceClient
+from huggingface_hub import InferenceClient, hf_hub_download
 
 class CLIPImageClassifier(nn.Module):
     def __init__(self):
@@ -39,6 +39,7 @@ HF_TOKEN = None
 embed_model = SentenceTransformer("all-MiniLM-L6-v2")
 explain_model = "meta-llama/Llama-3.1-8B-Instruct"
 text_model = "rajyalakshmijampani/fever_finetuned_deberta"
+image_model = "rajyalakshmijampani/finetuned_clip"
 wiki = wikipediaapi.Wikipedia(language='en', user_agent='fact-checker/1.0')
 image_processor = CLIPProcessor.from_pretrained("openai/clip-vit-base-patch32")
 
@@ -51,18 +52,12 @@ def get_text_classifier():
     return text_classifier
 
 def get_image_classifier():
-    global image_classifier
+    global image_classifier, image_model
+    filename = "finetuned_clip.pth"
     if image_classifier is None:
-        url = "https://huggingface.co/rajyalakshmijampani/finetuned_clip/resolve/main/best_clip_finetuned_classifier.pth"
-        path = "best_clip_finetuned_classifier.pth"
-
-        if not os.path.exists(path):
-            r = requests.get(url)
-            with open(path, "wb") as f:
-                f.write(r.content)
-
+        model_path = hf_hub_download(repo_id=image_model, filename=filename)
         image_classifier = CLIPImageClassifier()
-        state = torch.load(path, map_location="cpu",weights_only=False)
+        state = torch.load(model_path, map_location="cpu",weights_only=False)
         clean_state = OrderedDict(
             (k[7:], v) if k.startswith("module.") else (k, v)
             for k, v in state.items()
@@ -276,8 +271,9 @@ with gr.Blocks() as demo:
 
     with gr.Tab("Image Detector"):
         img_input = gr.Image(type="pil", label="Upload Image")
-        img_output = gr.Markdown(label="Model Output", value="Results will appear here...")
         img_button = gr.Button("Classify Image")
+        img_output = gr.Markdown(label="Model Output", value="Results will appear here...")
+        
         img_button.click(classify_image, inputs=img_input, outputs=img_output)
 
 demo.launch()