docs: fix Gemma 4 Colab notebook (#2275)

* docs: fix Gemma 4 Colab notebook

* docs: use smaller Gemma 4 Colab model

* docs: update Gemma 4 Colab CTA

Author: abetlen

README.md

@@ -535,9 +535,11 @@ Below are the supported multi-modal models and their respective chat handlers (P
 | [llama-3-vision-alpha](https://huggingface.co/abetlen/llama-3-vision-alpha-gguf) | `Llama3VisionAlphaChatHandler` | `llama-3-vision-alpha` |
 | [minicpm-v-2.6](https://huggingface.co/openbmb/MiniCPM-V-2_6-gguf) | `MiniCPMv26ChatHandler` | `minicpm-v-2.6` |
 | [qwen2.5-vl](https://huggingface.co/unsloth/Qwen2.5-VL-3B-Instruct-GGUF) | `Qwen25VLChatHandler` | `qwen2.5-vl` |
-| [gemma-4](https://huggingface.co/unsloth/gemma-4-E4B-it-GGUF) [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/abetlen/llama-cpp-python/blob/main/examples/colab/notebook.ipynb) | `Gemma4ChatHandler` | `gemma4` |
+| [gemma-4](https://huggingface.co/unsloth/gemma-4-E4B-it-GGUF) | `Gemma4ChatHandler` | `gemma4` |
 | GGUF models with an mtmd projector and embedded chat template | `MTMDChatHandler` | `mtmd` |
 
+Try Gemma 4 12B in Google Colab -> [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/abetlen/llama-cpp-python/blob/main/examples/colab/notebook.ipynb)
+
 Then you'll need to use a custom chat handler to load the clip model and process the chat messages and images.
 
 ```python

examples/colab/notebook.ipynb

@@ -81,7 +81,7 @@
         "    messages=[\n",
         "        {\n",
         "            \"role\": \"user\",\n",
-        "            \"content\": \"Write the exact string `<stdio.h>` and nothing else.\",\n",
+        "            \"content\": \"What is the capital of France? Answer in one sentence.\",\n",
         "        }\n",
         "    ],\n",
         "    max_tokens=32,\n",
@@ -99,7 +99,7 @@
       "source": [
         "from IPython.display import Image, display\n",
         "\n",
-        "IMAGE_URL = \"https://raw.githubusercontent.com/abetlen/llama-cpp-python/main/vendor/llama.cpp/tools/mtmd/test-1.jpeg\"\n",
+        "IMAGE_URL = \"https://raw.githubusercontent.com/ggml-org/llama.cpp/master/tools/mtmd/test-1.jpeg\"\n",
         "\n",
         "display(Image(url=IMAGE_URL, width=320))\n"
       ]