Update app.py

app.py

@@ -1,12 +1,11 @@
 import gradio as gr
-import bs4
-from langchain.embeddings.huggingface import HuggingFaceBgeEmbeddings
-from langchain.document_loaders import WebBaseLoader, PyPDFDirectoryLoader
-from langchain.vectorstores import FAISS
 from langchain.text_splitter import RecursiveCharacterTextSplitter
-from transformers import pipeline
+from langchain_community.document_loaders import WebBaseLoader
+from langchain_community.vectorstores import FAISS
+from langchain_community.embeddings import HuggingFaceBgeEmbeddings
+import ollama
 
-# Function to load, split, and retrieve documents from a URL
+# Function to load, split, and retrieve documents
 def load_and_retrieve_docs(url):
     loader = WebBaseLoader(
         web_paths=(url,),
@@ -21,9 +20,9 @@ def load_and_retrieve_docs(url):
     vectorstore = FAISS.from_documents(documents=splits, embedding=embeddings)
     return vectorstore.as_retriever()
 
-# Function to format documents into a context string
+# Function to format documents
 def format_docs(docs):
-    return "\n\n".join([doc['content'] for doc in docs])
+    return "\n\n".join(doc.page_content for doc in docs)
 
 # Function that defines the RAG chain
 def rag_chain(url, question):
@@ -31,12 +30,8 @@ def rag_chain(url, question):
     retrieved_docs = retriever.invoke(question)
     formatted_context = format_docs(retrieved_docs)
     formatted_prompt = f"Question: {question}\n\nContext: {formatted_context}"
-    
-    # Using HuggingFace transformers for generating response
-    chat_pipeline = pipeline('text-generation', model='gpt-3.5-turbo')  # Use the appropriate model here
-    response = chat_pipeline(formatted_prompt, max_length=512, num_return_sequences=1)
-    
-    return response[0]['generated_text']
+    response = ollama.chat(model='llama3', messages=[{'role': 'user', 'content': formatted_prompt}])
+    return response['message']['content']
 
 # Gradio interface
 iface = gr.Interface(