Update app.py

app.py

@@ -1,11 +1,11 @@
 import streamlit as st
 import os
-from together import Together
 from PyPDF2 import PdfReader
 from langchain.text_splitter import RecursiveCharacterTextSplitter
 from langchain_community.vectorstores import FAISS
-from langchain_openai import OpenAIEmbeddings
+from langchain_openai import OpenAIEmbeddings  # This might need to be adjusted if `Together` has its own embeddings module
 from langchain.prompts import ChatPromptTemplate
+from together import Together
 
 # Set up API client
 client = Together(api_key=os.environ.get("TOGETHER_API_KEY"))
@@ -27,7 +27,7 @@ def get_text_chunks(text):
 
 # Function to create and save a FAISS vector store from text chunks
 def get_vector_store(text_chunks):
-    embeddings = OpenAIEmbeddings(client=client)
+    embeddings = OpenAIEmbeddings(client=client)  # Adjust this if `Together` has its own method for embeddings
     vector_store = FAISS.from_texts(text_chunks, embedding=embeddings)
     vector_store.save_local("faiss_index")
 
@@ -36,13 +36,12 @@ def get_conversational_chain():
     prompt_template = """Answer the question concisely, focusing on the most relevant and important details from the PDF context.
     If the answer is not found within the PDF, please state 'answer is not available in the context.'"""
     prompt = ChatPromptTemplate.from_template(prompt_template)
-    model = client.create_model("mistralai/Mixtral-8x7B-Instruct-v0.1", temperature=0.2)
-    chain = load_qa_chain(model, chain_type="conversational", prompt=prompt)
+    chain = load_qa_chain(client, "mistralai/Mixtral-8x7B-Instruct-v0.1", prompt=prompt)  # Assuming load_qa_chain can accept Together client and model ID
     return chain
 
 # Function to process user question and provide a response
 def user_input(user_question):
-    embeddings = OpenAIEmbeddings(client=client)
+    embeddings = OpenAIEmbeddings(client=client)  # Adjust this if `Together` has its own method for embeddings
     new_db = FAISS.load_local("faiss_index", embeddings, allow_dangerous_deserialization=True)
     docs = new_db.similarity_search(user_question)
     chain = get_conversational_chain()
@@ -77,4 +76,4 @@ def main():
             st.error(str(e))
 
 if __name__ == "__main__":
-    main()
+    main()