docs: use smaller Gemma 4 Colab model

Author: abetlen

examples/colab/notebook.ipynb

@@ -51,7 +51,7 @@
         "from llama_cpp.llama_chat_format import Gemma4ChatHandler\n",
         "\n",
         "MODEL_REPO = \"ggml-org/gemma-4-12B-it-GGUF\"\n",
-        "MODEL_FILE = \"gemma-4-12B-it-Q8_0.gguf\"\n",
+        "MODEL_FILE = \"gemma-4-12B-it-Q4_K_M.gguf\"\n",
         "MMPROJ_FILE = \"mmproj-gemma-4-12B-it-Q8_0.gguf\"\n",
         "\n",
         "chat_handler = Gemma4ChatHandler.from_pretrained(\n",