app.py

import subprocess
import gradio as gr

MODEL_PATH = "/app/llama.cpp/build/bin/llama-cli"


def generate(prompt, temperature):
    cmd = [
        "/app/llama.cpp/main",
        "-m", MODEL_PATH,
        "-p", prompt,
        "--temp", str(temperature),
        "-n", "512",
        "--ctx-size", "512",
        "--no-mmap",
    ]

    result = subprocess.run(
        cmd,
        stdout=subprocess.PIPE,
        stderr=subprocess.PIPE,
        text=True
    )

    return result.stdout

iface = gr.Interface(
    fn=generate,
    inputs=[
        gr.Textbox(label="Prompt", lines=4),
        gr.Slider(0.01, 2.0, value=0.8, step=0.01, label="Temperature"),
    ],
    outputs=gr.Textbox(label="Output", lines=10),
    title="SLiNeP 7M GGUF",
    description="Tiny 7M parameter GGUF running with llama.cpp"
)

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