Update app.py

app.py

@@ -1,8 +1,6 @@
 import streamlit as st
 import fitz  # PyMuPDF
-from transformers import AutoTokenizer, AutoModelForCausalLM
-from langchain.prompts import PromptTemplate
-from langchain.chains import LLMChain
+from transformers import pipeline
 
 # Function to extract text from PDF
 def extract_text_from_pdf(pdf_file):
@@ -13,27 +11,15 @@ def extract_text_from_pdf(pdf_file):
         text += page.get_text()
     return text
 
-# Function to generate MCQs
+# Function to generate MCQs using Hugging Face model
 def generate_mcqs(text, num_questions=5):
-    # Define a prompt template for generating MCQs
-    prompt_template = """
-    Generate {num_questions} multiple-choice questions based on the following text:
-    {text}
-    """
-    
-    prompt = prompt_template.format(text=text, num_questions=num_questions)
-    
-    # Load HuggingFace model
-    tokenizer = AutoTokenizer.from_pretrained("openai-community/gpt2-large")
-    llm = AutoModelForCausalLM.from_pretrained("openai-community/gpt2-large")
-
-    # Create an LLM chain with the prompt
-    chain = LLMChain(prompt_template=PromptTemplate(template=prompt_template), llm=llm)
-
-    # Run the chain to generate MCQs
-    response = chain.run(text=text, num_questions=num_questions)
-    
-    return response
+    # Define a Hugging Face text generation pipeline
+    text_generation = pipeline("text-generation")
+
+    # Generate MCQs by providing the text as a prompt
+    mcqs = text_generation(text, max_length=100, num_return_sequences=num_questions)
+
+    return mcqs
 
 # Streamlit app interface
 st.title("PDF to MCQ Generator")
@@ -50,5 +36,5 @@ if uploaded_file is not None:
     mcqs = generate_mcqs(text, num_questions)
 
     st.write("Generated MCQs:")
-    st.write(mcqs)
-
+    for idx, mcq in enumerate(mcqs):
+        st.write(f"{idx+1}. {mcq['generated_text']}")