app.py

import streamlit as st
from transformers import AutoTokenizer, AutoModelForCausalLM
from datetime import datetime

# Custom CSS for UI
st.markdown("""
    <style>
    .main { background-color: #f9f9f9; padding: 20px; }
    .stTextArea textarea {
        border: 1px solid #ddd;
        border-radius: 8px;
        padding: 10px;
        font-family: 'Roboto', sans-serif;
        font-size: 16px;
        background-color: #fff;
        box-shadow: 0 2px 4px rgba(0,0,0,0.1);
    }
    .stButton button {
        background-color: #4a90e2;
        color: white;
        border-radius: 8px;
        padding: 10px 20px;
        font-family: 'Roboto', sans-serif;
        font-size: 14px;
    }
    .stButton button:hover {
        background-color: #357abd;
    }
    .code-output {
        background-color: #2b2b2b;
        color: #f0f0f0;
        padding: 15px;
        border-radius: 8px;
        font-family: 'Courier New', monospace;
        font-size: 14px;
        margin-top: 10px;
    }
    .title {
        font-family: 'Roboto', sans-serif;
        font-size: 28px;
        font-weight: bold;
        color: #333;
        margin-bottom: 10px;
    }
    .subtitle {
        font-family: 'Roboto', sans-serif;
        font-size: 16px;
        color: #666;
        margin-bottom: 20px;
    }
    .chat-message {
        font-family: 'Roboto', sans-serif;
        font-size: 16px;
        color: #333;
        margin-bottom: 5px;
    }
    </style>
""", unsafe_allow_html=True)

# Cache model and tokenizer to avoid reloading
@st.cache_resource
def load_model_and_tokenizer():
    checkpoint = "Salesforce/codegen-350M-mono"
    try:
        st.write("Loading tokenizer...")
        tokenizer = AutoTokenizer.from_pretrained(checkpoint)
        st.write("Loading model...")
        model = AutoModelForCausalLM.from_pretrained(checkpoint)
        st.write("Model and tokenizer loaded successfully!")
        return tokenizer, model
    except Exception as e:
        st.error(f"Failed to load model/tokenizer: {e}")
        return None, None

# Load model and tokenizer once
tokenizer, model = load_model_and_tokenizer()
if tokenizer is None or model is None:
    st.stop()

# Function to generate code
def generate_code(description):
    prompt = f"Generate Python code for the following task: {description}\n"
    inputs = tokenizer(prompt, return_tensors="pt")
    try:
        outputs = model.generate(
            **inputs,
            max_length=500,
            num_return_sequences=1,
            pad_token_id=tokenizer.eos_token_id
        )
        code = tokenizer.decode(outputs[0], skip_special_tokens=True)
        return code[len(prompt):].strip()
    except Exception as e:
        st.error(f"Error generating code: {e}")
        return "Error: Could not generate code."

# Initialize chat history
if "chat_history" not in st.session_state:
    st.session_state.chat_history = []

# UI Layout
st.markdown('<div class="title">Code Generation Bot</div>', unsafe_allow_html=True)
st.markdown('<div class="subtitle">Describe your task, and I’ll generate Python code for you!</div>', unsafe_allow_html=True)

with st.container():
    # Input area
    description = st.text_area(
        "Enter your description here",
        placeholder="e.g., Write a function to calculate the factorial of a number",
        height=150
    )
    
    col1, col2 = st.columns([1, 1])
    with col1:
        if st.button("Generate"):
            if description.strip():
                with st.spinner("Thinking..."):
                    generated_code = generate_code(description)
                st.session_state.chat_history.append({
                    "input": description,
                    "output": generated_code,
                    "time": datetime.now().strftime("%H:%M:%S")
                })
            else:
                st.warning("Please enter a description first!")
    
    with col2:
        if st.button("Clear History"):
            st.session_state.chat_history = []
            st.success("Chat history cleared!")

    # Display chat history
    if st.session_state.chat_history:
        st.write("### Chat History")
        for chat in st.session_state.chat_history:
            st.markdown(f'<div class="chat-message"><strong>You ({chat["time"]}):</strong> {chat["input"]}</div>', unsafe_allow_html=True)
            st.markdown(f'<div class="code-output">{chat["output"]}</div>', unsafe_allow_html=True)
            st.markdown("---")

st.info("Tip: Check the generated code for accuracy before using it!")