Update app.py

app.py

@@ -58,16 +58,14 @@ def initialize_pinecone(index_name: str):
 ##################################################
 
 # Initialize Pinecone index and BM25 encoder
-pinecone_index = initialize_pinecone("uae-national-library-and-archives-vectorstore")
-bm25 = BM25Encoder().load("./UAE-NLA.json")
+pinecone_index = initialize_pinecone("updated-uae-nla")
+bm25 = BM25Encoder().load("./updated-uae-nla.json")
 
 ##################################################
 ##################################################
 
-# old_embed_model = HuggingFaceEmbeddings(model_name="sentence-transformers/gte-multilingual-base")
-
 # Initialize models and retriever
-embed_model = HuggingFaceEmbeddings(model_name="Alibaba-NLP/gte-multilingual-base", model_kwargs={"trust_remote_code":True})
+embed_model = HuggingFaceEmbeddings(model_name="jinaai/jina-embeddings-v3", model_kwargs={"trust_remote_code":True})
 retriever = PineconeHybridSearchRetriever(
     embeddings=embed_model, 
     sparse_encoder=bm25, 
@@ -95,13 +93,13 @@ contextualize_q_prompt = ChatPromptTemplate.from_messages(
 history_aware_retriever = create_history_aware_retriever(llm, 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. \
-If you don't know the answer, simply state that you don't know. \
+qa_system_prompt = """ You are a highly skilled information retrieval assistant. Use the following context to answer questions effectively. \
+If you don't know the answer, state that you don't know. \
 Your answer should be in {language} language. \
 Provide answers in proper HTML format and keep them concise. \
 When responding to queries, follow these guidelines: \
     1. Provide Clear Answers: \
-        - Based on the language of the question, you have to answer in that language. E.g. if the question is in English language then answer in the English language or if the question is in Arabic language then you should answer in Arabic language. /
+        - Based on the language of the question, you have to answer in that language. E.g. if the question is in the English language then answer in the English language or if the question is in Arabic language then you should answer in Arabic language. /
         - Ensure the response directly addresses the query with accurate and relevant information.\
     2. Include Detailed References: \
         - Links to Sources: Include URLs to credible sources where users can verify information or explore further. \
@@ -110,14 +108,14 @@ When responding to queries, follow these guidelines: \
     
     3. Formatting for Readability: \
         - The answer should be in a proper HTML format with appropriate tags. \
-        - For arabic language response align the text to right and convert numbers also.
-        - Double check if the language of answer is correct or not.
-        - Use bullet points or numbered lists where applicable to present information clearly. \
+        - For Arabic language response align the text to the right and convert numbers.
+        - Double-check if the language of the answer is correct or not.
+        - Use bullet points or numbered lists where applicable to present information. \
         - Highlight key details using bold or italics. \
-        - Provide proper and meaningful abbreviations for urls. Do not include naked urls. \
+        - Provide proper and meaningful abbreviations for URLs. Do not include naked URLs. \
     
     4. Organize Content Logically: \
-        - Structure the content in a logical order, ensuring easy navigation and understanding for the user. \
+        - Structure the content logically, ensuring easy navigation and understanding for the user. \
         
 {context}
 """