app.py

import os
import subprocess
from huggingface_hub import hf_hub_download

# Rýchle sťahovanie
os.environ["HF_HUB_ENABLE_HF_TRANSFER"] = "1"

def run():
    # Stiahneš model (už máš llama-cpp pripravené v systéme)
    model_path = hf_hub_download(
        repo_id="bartowski/Qwen2.5-Coder-7B-Instruct-GGUF", 
        filename="Qwen2.5-Coder-7B-Instruct-Q4_K_M.gguf",
        local_dir="."
    )

    # Spustíš server
    subprocess.run([
        "python3", "-m", "llama_cpp.server",
        "--model", "Qwen2.5-Coder-7B-Instruct-Q4_K_M.gguf",
        "--host", "0.0.0.0",
        "--port", "7860",
        "--n_ctx", "4096"
    ])

if __name__ == "__main__":
    run()