app.py

import gradio as gr
from transformers import pipeline

def load_model():
    return pipeline(
        "text-generation",
        model="bigcode/santacoder",  # Smaller than DeepSeek for CPU
        device="cpu"
    )

model = load_model()

def generate_code(prompt):
    try:
        response = model(
            f"<|user|>{prompt}</s><|assistant|>",
            max_new_tokens=100,
            temperature=0.7,
            do_sample=True
        )
        return response[0]['generated_text'].split("<|assistant|>")[-1]
    except Exception as e:
        return f"Error: {str(e)}"

demo = gr.Interface(
    fn=generate_code,
    inputs=gr.Textbox(lines=5, placeholder="Ask a coding question..."),
    outputs=gr.Code(language="python"),
    title="CPU-Friendly Coding Assistant",
    description="This simplified version works on free CPU hardware"
)

demo.launch(server_name="0.0.0.0", server_port=7860)