Temporarily remove reranker

app.py

@@ -48,7 +48,7 @@ index = faiss.IndexFlatL2(corpus_embeddings_np.shape[1])
 index.add(corpus_embeddings_np)
 
 # Reranker model
-reranker = CrossEncoder("cross-encoder/ms-marco-MiniLM-L-6-v2")
+# reranker = CrossEncoder("cross-encoder/ms-marco-MiniLM-L-6-v2")
 
 # Generator (choose one: local HF model or OpenAI)
 tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-Instruct-v0.3")
@@ -68,17 +68,17 @@ def rag_pipeline(query):
     for doc in retrieved_docs:
         print("-", repr(doc))
 
-    # Rerank
-    rerank_pairs = [[str(query), str(doc)] for doc in retrieved_docs if isinstance(doc, str) and doc.strip()]
-    if not rerank_pairs:
-        return "No valid documents found to rerank."
-    scores = reranker.predict(rerank_pairs)
+    # # Rerank
+    # rerank_pairs = [[str(query), str(doc)] for doc in retrieved_docs if isinstance(doc, str) and doc.strip()]
+    # if not rerank_pairs:
+    #     return "No valid documents found to rerank."
+    # scores = reranker.predict(rerank_pairs)
 
-    scores = reranker.predict(rerank_pairs)
-    reranked_docs = [doc for _, doc in sorted(zip(scores, retrieved_docs), reverse=True)]
+    # scores = reranker.predict(rerank_pairs)
+    # reranked_docs = [doc for _, doc in sorted(zip(scores, retrieved_docs), reverse=True)]
 
     # Combine for context
-    context = "\n\n".join(reranked_docs[:2])
+    context = "\n\n".join(retrieved_docs[:2])
     prompt = f"""Answer the following question using the provided context.\n\nContext:\n{context}\n\nQuestion: {query}\nAnswer:"""
 
     # Generate