lang: support direct video inference (#9936)

Co-authored-by: Lianmin Zheng <lianminzheng@gmail.com>

docs/references/frontend/frontend_tutorial.ipynb

@@ -33,7 +33,7 @@
     "from sglang import assistant, function, gen, system, user\n",
     "from sglang import image\n",
     "from sglang import RuntimeEndpoint\n",
-    "from sglang.lang.api import set_default_backend\n",
+    "from sglang.lang.api import set_default_backend, video\n",
     "from sglang.srt.utils import load_image\n",
     "from sglang.test.doc_patch import launch_server_cmd\n",
     "from sglang.utils import print_highlight, terminate_process, wait_for_server\n",
@@ -421,7 +421,11 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {},
+   "metadata": {
+    "jupyter": {
+     "is_executing": true
+    }
+   },
    "outputs": [],
    "source": [
     "@function\n",
@@ -436,6 +440,30 @@
     "print_highlight(state[\"answer\"])"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Ask a question about a video"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "@function\n",
+    "def video_qa(s, video_file, question):\n",
+    "    s += user(video(video_file) + question)\n",
+    "    s += assistant(gen(\"answer\", max_tokens=256))\n",
+    "\n",
+    "\n",
+    "video_url = \"https://raw.githubusercontent.com/sgl-project/sgl-test-files/refs/heads/main/videos/jobs_presenting_ipod.mp4\"\n",
+    "state = video_qa(video_url, \"What is in the video?\")\n",
+    "print_highlight(state[\"answer\"])"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,