Update app.py

app.py

@@ -86,12 +86,12 @@ retriever = PineconeHybridSearchRetriever(
 llm = ChatPerplexity(temperature=0, pplx_api_key=GROQ_API_KEY, model="llama-3.1-sonar-large-128k-chat", max_tokens=512, max_retries=2)
 
 # Initialize Reranker
-# model = HuggingFaceCrossEncoder(model_name="BAAI/bge-reranker-base")
-# compressor = CrossEncoderReranker(model=model, top_n=10)
+model = HuggingFaceCrossEncoder(model_name="BAAI/bge-reranker-base")
+compressor = CrossEncoderReranker(model=model, top_n=10)
 
-# compression_retriever = ContextualCompressionRetriever(
-#     base_compressor=compressor, base_retriever=retriever
-# )
+compression_retriever = ContextualCompressionRetriever(
+    base_compressor=compressor, base_retriever=retriever
+)
 # from langchain.retrievers.document_compressors import LLMChainExtractor
 
 # compressor = LLMChainExtractor.from_llm(llm)
@@ -100,9 +100,11 @@ llm = ChatPerplexity(temperature=0, pplx_api_key=GROQ_API_KEY, model="llama-3.1-
 # )
 
 # compressor = FlashrankRerank(top_n=10)
-# compression_retriever = ContextualCompressionRetriever(
-#     base_compressor=compressor, base_retriever=retriever
-# )
+
+
+compression_retriever = ContextualCompressionRetriever(
+    base_compressor=compressor, base_retriever=retriever
+)
 
 # Contextualization prompt and retriever
 contextualize_q_system_prompt = """ Given a chat history and the latest user question \
@@ -117,7 +119,7 @@ contextualize_q_prompt = ChatPromptTemplate.from_messages(
         ("human", "{input}")
     ]
 )
-history_aware_retriever = create_history_aware_retriever(llm, retriever, contextualize_q_prompt)
+history_aware_retriever = create_history_aware_retriever(llm, compression_retriever, contextualize_q_prompt)
 
 # QA system prompt and chain
 qa_system_prompt = """ You are a highly skilled information retrieval assistant. Use the following context to answer questions effectively.