Notebook fixes for prompt completion (#16420)

docs/docs/examples/llm/nvidia_text_completion.ipynb

@@ -99,7 +99,12 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "## Working with NVIDIA API Catalog"
+    "## Working with NVIDIA API Catalog\n",
+    "#### Usage of `use_chat_completions` argument: \n",
+    "Set None (default) to per-invocation decide on using /chat/completions vs /completions endpoints with query keyword arguments\n",
+    "\n",
+    "- set False to universally use /completions endpoint\n",
+    "- set True to universally use /chat/completions endpoint"
    ]
   },
   {
@@ -113,13 +118,22 @@
     "llm = NVIDIA(model=\"bigcode/starcoder2-15b\", use_chat_completions=False)"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Available Models\n",
+    "\n",
+    "`is_chat_model` can be used to get available text completion models"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
-    "llm.available_models"
+    "print([model for model in llm.available_models if model.is_chat_model])"
    ]
   },
   {
@@ -206,6 +220,16 @@
     "x = llm.stream_complete(prompt=\"# Reverse string in python:\", max_tokens=512)"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "for t in x:\n",
+    "    print(t.delta, end=\"\")"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},