Upload Dockerfile with huggingface_hub

Dockerfile

@@ -9,8 +9,7 @@ ENV LLAMA_CACHE=/tmp/llama-cache \
 
 # HF Spaces routes external traffic to app_port (7860). The GGUF is pulled from
 # the public model repo on first boot and cached in /tmp.
-CMD ["--hf-repo", "cloudunity/stealth-rifle", \
-     "--hf-file", "stealth-rifle-Q4_K_M.gguf", \
+CMD ["-hf", "bartowski/Qwen2.5-3B-Instruct-GGUF:Q4_K_M", \
      "--host", "0.0.0.0", "--port", "7860", \
      "--threads", "2", "--ctx-size", "24576", \
      "--chat-template", "chatml", \