Update app.py

app.py

@@ -1,10 +1,10 @@
 import streamlit as st
 import fitz  # PyMuPDF
-# Load model directly
-from transformers import AutoTokenizer, AutoModelForCausalLM
+from transformers import pipeline, AutoTokenizer, AutoModelForCausalLM
 
-tokenizer = AutoTokenizer.from_pretrained("google/bert_uncased_L-12_H-768_A-12")
-model = AutoModelForCausalLM.from_pretrained("google/bert_uncased_L-12_H-768_A-12")
+# Load the model and tokenizer
+tokenizer = AutoTokenizer.from_pretrained("openai-gpt")
+model = AutoModelForCausalLM.from_pretrained("openai-gpt")
 
 # Function to extract text from PDF
 def extract_text_from_pdf(pdf_file):
@@ -16,7 +16,7 @@ def extract_text_from_pdf(pdf_file):
     return text
 
 # Function to generate MCQs using Hugging Face model
-def generate_mcqs(text, num_questions):
+def generate_mcqs(text, num_questions=5):
     if not text.strip():
         return ["No text extracted from the PDF. Unable to generate MCQs."]
 
@@ -24,18 +24,10 @@ def generate_mcqs(text, num_questions):
     inputs = tokenizer(text, return_tensors="pt", max_length=512, truncation=True)
 
     mcqs = []
+    generator = pipeline("text-generation", model=model, tokenizer=tokenizer)
     for _ in range(num_questions):
         # Generate a single MCQ at a time
-        output = model.generate(
-            **inputs,
-            max_length=512,  # Adjusted to match the tokenized input length
-            max_new_tokens=100,  # Limit the number of new tokens added during generation
-            temperature=0.7,
-            num_return_sequences=1
-        )
-
-        # Decode the generated sequence
-        mcq = tokenizer.decode(output[0], skip_special_tokens=True)
+        mcq = generator(text, max_length=100)[0]["generated_text"]
         mcqs.append(mcq)
 
     return mcqs
@@ -52,8 +44,7 @@ if uploaded_file is not None:
     st.write("Extracted Text:", text)
 
     st.write("Generating MCQs...")
-   # num_questions = st.number_input("Number of MCQs to generate")
-    num_questions = 6
+    num_questions = st.number_input("Number of MCQs to generate", min_value=1, max_value=20, value=5, step=1, format="%d")
     mcqs = generate_mcqs(text, num_questions)
 
     st.write("Generated MCQs:")