app.py

import gradio as gr
from groq import Groq
import os
import json

# Initialize Groq client
client = Groq(api_key=os.environ["GROQ_API_KEY"])

def generate_tutor_output(subject, grade, student_input, model):
    prompt = f"""
    You are an expert tutor in {subject} for a {grade} grade student. 
    The student has provided the following input: "{student_input}"
    
    Please generate:
    1. A fun, engaging lesson (2-3 paragraphs) tailored to a {grade} grader's understanding.
    2. A thought-provoking multiple-choice question (with 4 options: a, b, c, d) to test understanding.
    3. Constructive feedback on the student's input.
    
    Format your response as a JSON object with keys: "lesson", "question", "options", "correct_answer", "feedback"
    """

    completion = client.chat.completions.create(
        messages=[
            {
                "role": "system",
                "content": f"You are a fun, creative AI tutor for {grade} graders, expert in {subject}. You explain concepts in a simple, exciting way with relatable examples (like math problems for their age). Your goal is to spark curiosity and help students practice what they learn!",
            },
            {
                "role": "user",
                "content": prompt,
            }
        ],
        model=model,
        max_tokens=1200,
    )

    return completion.choices[0].message.content

def check_answer(selected_answer, correct_answer):
    if selected_answer == correct_answer:
        return "🎉 Awesome job! You got it right! Keep rocking it!", 10
    else:
        return f"😅 Not quite! The correct answer was '{correct_answer}'. Try again next time!", 0

with gr.Blocks(title="Learn & Practice 🚀") as demo:
    gr.Markdown("# 🚀 Learn & Practice Zone (Grades 5-10)")
    
    # Input Section
    with gr.Row():
        with gr.Column(scale=2):
            subject = gr.Dropdown(
                ["Math", "Science", "History", "Geography", "Economics"], 
                label="Subject", 
                info="Pick your favorite subject!"
            )
            grade = gr.Dropdown(
                ["5th Grade", "6th Grade", "7th Grade", "8th Grade", "9th Grade", "10th Grade"], 
                label="Your Grade", 
                info="Select your grade level"
            )
            model_select = gr.Dropdown(
                [
                    "mixtral-8x7b-32768",
                    "qwen-2.5-coder-32b",
                    "qwen-2.5-32b"
                ],
                label="AI Tutor Model",
                value="mixtral-8x7b-32768",
                info="Choose your AI tutor"
            )
            student_input = gr.Textbox(
                placeholder="What do you want to learn today?", 
                label="Your Question", 
                info="Ask anything about the subject!"
            )
            submit_button = gr.Button("Get Lesson & Practice", variant="primary")
        
        # Output Section
        with gr.Column(scale=3):
            lesson_output = gr.Markdown(label="Your Lesson")
            question_output = gr.Markdown(label="Test Your Skills")
            options_output = gr.Radio(label="Choose an Answer", choices=[], visible=False)
            feedback_output = gr.Markdown(label="Feedback on Your Question")
            answer_feedback = gr.Markdown(label="Answer Feedback")
            points = gr.Number(label="Your Points", value=0)

    # Instructions
    gr.Markdown("""
    ### How to Play & Learn
    1. Pick a subject and your grade.
    2. Choose an AI tutor model.
    3. Ask a question or topic you’re curious about.
    4. Read the fun lesson, then answer the question to test yourself.
    5. Earn points for correct answers and keep learning!
    """)

    def process_output(output):
        try:
            parsed = json.loads(output)
            options = [f"{k}. {v}" for k, v in zip(["a", "b", "c", "d"], parsed["options"])]
            return (
                parsed["lesson"],
                parsed["question"],
                options,
                parsed["correct_answer"],
                parsed["feedback"]
            )
        except:
            return (
                "Error generating lesson",
                "No question available",
                [],
                "",
                "No feedback available"
            )

    def update_interface(subject, grade, student_input, model):
        output = generate_tutor_output(subject, grade, student_input, model)
        lesson, question, options, correct_answer, feedback = process_output(output)
        return (
            lesson,
            question,
            gr.update(choices=options, visible=True),
            feedback,
            "",  # Clear answer feedback
            gr.update(value=0)  # Reset points
        ), correct_answer

    # State to store correct answer
    correct_answer_state = gr.State()

    submit_button.click(
        fn=update_interface,
        inputs=[subject, grade, student_input, model_select],
        outputs=[lesson_output, question_output, options_output, feedback_output, answer_feedback, points]
    ).then(
        fn=lambda x: x,
        inputs=[gr.State()],
        outputs=[correct_answer_state]
    )

    options_output.change(
        fn=check_answer,
        inputs=[options_output, correct_answer_state],
        outputs=[answer_feedback, points]
    )

if __name__ == "__main__":
    demo.launch(server_name="0.0.0.0", server_port=7860)