app.py

import streamlit as st
from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline
import torch

# Load the fine-tuned MFTCoder model
@st.cache_resource()
def load_model():
    MODEL_NAME = "path-to-your-finetuned-model"  # Replace with your MFTCoder fine-tuned model path
    tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
    model = AutoModelForCausalLM.from_pretrained(
        MODEL_NAME,
        torch_dtype=torch.float16,  # Use float16 for performance optimization
        device_map="auto"  # Automatically allocate to CPU/GPU
    )
    return pipeline("text-generation", model=model, tokenizer=tokenizer)

# Initialize pipeline
code_generator = load_model()

# Streamlit UI
st.title("MFTCoder-powered Code Bot 🚀")
st.subheader("Generate high-quality code snippets with fine-tuned CodeLlama!")

# User input
prompt = st.text_area("Enter a code prompt to generate code:")

# Generate code
if st.button("Generate Code"):
    if prompt.strip():
        st.info("Generating code... Please wait ⏳")
        try:
            # Generate code using the fine-tuned MFTCoder model
            response = code_generator(
                prompt,
                max_length=256,  # Adjust as needed
                temperature=0.3,  # Lower temperature for accurate outputs
                num_return_sequences=1,
                do_sample=True
            )
            generated_code = response[0]['generated_text']
            # Display the code output
            st.code(generated_code, language="python")  # Default to Python for generated output
        except Exception as e:
            st.error(f"Error: {str(e)}")
    else:
        st.warning("Please enter a prompt.")

st.caption("Created by Shamil")