Update app.py

app.py

@@ -10,22 +10,18 @@ from transformers import RobertaModel, AutoTokenizer
 def load():
     text_encoder = RobertaModel.from_pretrained('SajjadAyoubi/clip-fa-text')
     tokenizer = AutoTokenizer.from_pretrained('SajjadAyoubi/clip-fa-text')
-    link_df = pd.read_csv('links.csv')
+    links = np.load('link.npy', allow_pickle=True)
     image_embeddings = torch.load('embeddings.pt')
-    return text_encoder, tokenizer, link_df, image_embeddings
+    return text_encoder, tokenizer, links, image_embeddings
 
 
-text_encoder, tokenizer, link_df, image_embeddings = load()
+text_encoder, tokenizer, links, image_embeddings = load()
 
 
 def get_html(url_list, height=224):
     html = "<div style='margin-top: 20px; max-width: 1200px; display: flex; flex-wrap: wrap; justify-content: space-evenly'>"
-    for url, link in url_list:
-        html2 = f"<img style='height: {height}px; margin: 5px' src='{escape(url)}'>"
-        if len(link) > 0:
-            html2 = f"<a href='{escape(link)}' target='_blank'>" + \
-                html2 + "</a>"
-                
+    for url in url_list:
+        html2 = f"<img style='height: {height}px; margin: 5px' src='{escape(url)}'>"  
         html = html + html2
     html += "</div>"
     return html
@@ -36,7 +32,7 @@ def image_search(query, top_k=8):
     with torch.no_grad():
         text_embedding = text_encoder(**tokenizer(query, return_tensors='pt')).pooler_output
     values, indices = torch.cosine_similarity(text_embedding, image_embeddings).sort(descending=True)
-    return [(link_df.iloc[i]['path'], link_df.iloc[i]['link']) for i in indices[:top_k]]
+    return [links[i] for i in indices[:top_k]]
 
 
 description = '''