Warm Ollama into VRAM on up and tunnel so Assistent chat is not cold.

A 1-token /api/chat after tags (and again if /api/ps is empty) loads VL weights before the first message. Mid KEEP_ALIVE is 15m so a short image-gen burst does not unload the model.

Co-authored-by: Cursor <cursoragent@cursor.com>

src/gpu_rent/tunnel.py

@@ -299,6 +299,10 @@ def run_tunnel(
             log(f"проверка туннеля: {exc}")
             raise
 
+        from gpu_rent.llm_runtime import maybe_warmup_ollama_local
+
+        maybe_warmup_ollama_local(cfg, log)
+
         from gpu_rent.access_card import print_access_card
 
         print_access_card(cfg, tunneled=True, host=current_host)