Update app.py

app.py

@@ -6,6 +6,7 @@ from langchain import OpenAI
 from langchain.chains import RetrievalQAWithSourcesChain
 from langchain.text_splitter import RecursiveCharacterTextSplitter
 from langchain.document_loaders import UnstructuredURLLoader
+# from langchain.embeddings import OpenAIEmbeddings
 from langchain.embeddings import FakeEmbeddings
 from langchain.llms import HuggingFaceHub
 from langchain.chains import LLMChain
@@ -13,64 +14,65 @@ from langchain.vectorstores import FAISS
 
 from dotenv import load_dotenv
 load_dotenv()  # take environment variables from .env (especially openai api key)
-os.environ["HUGGINGFACEHUB_API_TOKEN"] = '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'
 
+st.title("RockyBot: News Research Tool 📈")
+st.sidebar.title("News Article URLs")
 
-class RockyBot:
-    def __init__(self, llm):
-        self.llm = llm
-        self.vectorstore = None
+urls = []
+for i in range(3):
+    url = st.sidebar.text_input(f"URL {i+1}")
+    urls.append(url)
 
-    def process_urls(self, urls):
-        """Processes the given URLs and saves the FAISS index to a pickle file."""
+process_url_clicked = st.sidebar.button("Process URLs")
+file_path = "faiss_store_openai.pkl"
 
-        # load data
-        loader = UnstructuredURLLoader(urls=urls)
+main_placeholder = st.empty()
+llm = HuggingFaceHub( repo_id="google/flan-t5-xxl", model_kwargs={"temperature": 0.5, "max_length": 64} )
 
-        # split data
-        text_splitter = RecursiveCharacterTextSplitter(
-            separators=['\n\n', '\n', '.', ','],
-            chunk_size=1000
-        )
-        docs = text_splitter.split_documents(loader.load())
+@st.cache
+def process_urls(urls):
+    """Processes the given URLs and saves the FAISS index to a pickle file."""
 
-        # create embeddings and save it to FAISS index
-        embeddings = FakeEmbeddings(size=1352)
-        self.vectorstore = FAISS.from_documents(docs, embeddings)
+    # load data
+    loader = UnstructuredURLLoader(urls=urls)
 
-        # Save the FAISS index to a pickle file
-        with open("faiss_store_openai.pkl", "wb") as f:
-            pickle.dump(self.vectorstore, f)
+    # split data
+    text_splitter = RecursiveCharacterTextSplitter(
+        separators=['\n\n', '\n', '.', ','],
+        chunk_size=1000
+    )
+    docs = text_splitter.split_documents(loader.load())
 
-    def answer_question(self, question):
-        """Answers the given question using the LLM and retriever."""
+    # create embeddings and save it to FAISS index
+    embeddings = FakeEmbeddings(size=1352)
+    vectorstore_openai = FAISS.from_documents(docs, embeddings)
 
-        chain = RetrievalQAWithSourcesChain.from_llm(llm=self.llm, retriever=self.vectorstore.as_retriever())
-        result = chain({"question": question}, return_only_outputs=True)
+    # Save the FAISS index to a pickle file
+    with open(file_path, "wb") as f:
+        pickle.dump(vectorstore_openai, f)
 
-        return result["answer"], result.get("sources", "")
-
-
-if __name__ == '__main__':
-    llm = HuggingFaceHub(repo_id="google/flan-t5-xxl", model_kwargs={"temperature": 0.5, "max_length": 64})
-    rockybot = RockyBot(llm)
-
-    # Process URLs if the button is clicked
-    if st.sidebar.button("Process URLs"):
-        rockybot.process_urls(st.sidebar.text_input("URL 1"), st.sidebar.text_input("URL 2"), st.sidebar.text_input("URL 3"))
+if process_url_clicked:
+    with st.progress(0.0):
+        process_urls(urls)
         st.progress(100.0)
 
-    # Answer the question if it is not empty
-    query = st.text_input("Question: ")
-    if query:
-        answer, sources = rockybot.answer_question(query)
-
-        st.header("Answer")
-        st.write(answer)
-
-        # Display sources, if available
-        if sources:
-            st.subheader("Sources:")
-            for source in sources.split("\n"):
-                st.write(source)
-
+query = main_placeholder.text_input("Question: ")
+if query:
+    try:
+        with open(file_path, "rb") as f:
+            vectorstore = pickle.load(f)
+            chain = RetrievalQAWithSourcesChain.from_llm(llm=llm, retriever=vectorstore.as_retriever())
+            result = chain({"question": query}, return_only_outputs=True)
+            # result will be a dictionary of this format --> {"answer": "", "sources": [] }
+            st.header("Answer")
+            st.write(result["answer"])
+
+            # Display sources, if available
+            sources = result.get("sources", "")
+            if sources:
+                st.subheader("Sources:")
+                sources_list = sources.split("\n")  # Split the sources by newline
+                for source in sources_list:
+                    st.write(source)
+    except Exception as e:
+        st.error(e)