Update app.py

app.py

@@ -13,6 +13,7 @@ df = pd.read_csv("data/hadith.csv")
 
 # Load embeddings
 hadith_embeddings = np.load("data/hadith_embeddings.npy")
+print(f"Loaded hadith embeddings: {hadith_embeddings.shape}")
 
 # Load BM25
 with open("data/bm25.pkl", "rb") as f:
@@ -20,6 +21,7 @@ with open("data/bm25.pkl", "rb") as f:
 
 # Load anchor FAISS index
 anchor_index = faiss.read_index("data/faiss_anchor.index")
+print(f"Anchor index dimension: {anchor_index.d}")
 
 # Load anchor mapping
 with open("data/anchor_dict.pkl", "rb") as f:
@@ -32,7 +34,20 @@ with open("data/unique_anchor_texts.pkl", "rb") as f:
 model = SentenceTransformer(
     "sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2"
 )
-model.max_seq_length = 512 
+model.max_seq_length = 512
+
+# Test embedding dimension
+test_emb = model.encode("test", normalize_embeddings=True)
+print(f"Model embedding dimension: {test_emb.shape}")
+
+# Verify dimensions match
+if test_emb.shape[0] != anchor_index.d:
+    raise ValueError(
+        f"Dimension mismatch! Model outputs {test_emb.shape[0]}D but "
+        f"anchor_index expects {anchor_index.d}D. "
+        f"Rebuild your anchor_index with the same model."
+    )
+
 # Import retrieval logic
 from retrieval import hybrid_search_fixed
 from utils import preprocess_query
@@ -44,24 +59,33 @@ from utils import preprocess_query
 def search_hadith(query, top_k):
     if query.strip() == "":
         return pd.DataFrame(columns=["الموضوع", "نص الحديث", "hadith page on Islamweb.net"])
-    results_df, _ = hybrid_search_fixed(
-        query=query,
-        df=df,
-        bm25=bm25, 
-        model=model,
-        preprocess_query=preprocess_query,
-        hadith_embeddings=hadith_embeddings,
-        anchor_index=anchor_index,
-        anchor_dict=anchor_dict,
-        unique_anchor_texts=unique_anchor_texts,
-        top_k=int(top_k)
-    )
+    
+    try:
+        results_df, _ = hybrid_search_fixed(
+            query=query,
+            df=df,
+            bm25=bm25, 
+            model=model,
+            preprocess_query=preprocess_query,
+            hadith_embeddings=hadith_embeddings,
+            anchor_index=anchor_index,
+            anchor_dict=anchor_dict,
+            unique_anchor_texts=unique_anchor_texts,
+            top_k=int(top_k)
+        )
 
-    return results_df[["main_subj", "clean_text", "url"]] \
-        .rename(columns={
-            "main_subj": "الموضوع",
-            "clean_text": "نص الحديث",
-            "url": "hadith page on Islamweb.net"
+        return results_df[["main_subj", "clean_text", "url"]] \
+            .rename(columns={
+                "main_subj": "الموضوع",
+                "clean_text": "نص الحديث",
+                "url": "hadith page on Islamweb.net"
+            })
+    except Exception as e:
+        print(f"Error in search: {e}")
+        return pd.DataFrame({
+            "الموضوع": ["Error"],
+            "نص الحديث": [str(e)],
+            "hadith page on Islamweb.net": [""]
         })
 
 
@@ -87,6 +111,11 @@ interface = gr.Interface(
     ),
     title="Using NLP to search hadith in sahih bukhari",
     description=("AI-powered search engine that understands the **meaning** of queries, not just keyword matches."),
+    examples=[
+        ["أهمية النية وأثرها في قبول الأعمال", 5],
+        ["فضل الصلاة", 5],
+        ["حقوق الجار", 5],
+    ]
 )
 
 # Launch app