Python: Added examples of using ChatCompletion models for skill building in Jupyter Notebooks

samples/notebooks/python/03-semantic-function-inline.ipynb

@@ -150,16 +150,60 @@
     "print(summary)"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "id": "1c2c1262",
+   "metadata": {},
+   "source": [
+    "# Using ChatCompletion for Semantic Skills"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "29b59b28",
+   "metadata": {},
+   "source": [
+    "You can also use chat completion models (like `gpt-35-turbo` and `gpt4`) for creating skills. Normally you would have to tweak the API to accommodate for a system and user role, but SK abstracts that away for you by using `kernel.add_chat_service` and `AzureChatCompletion` or `OpenAIChatCompletion`"
+   ]
+  },
   {
    "cell_type": "markdown",
    "id": "4777f447",
    "metadata": {},
    "source": [
-    "Here's one more example of how to write an inline Semantic Function that gives a TLDR for a piece of text.\n",
+    "Here's one more example of how to write an inline Semantic Function that gives a TLDR for a piece of text using a ChatCompletion model\n",
     "\n",
     "\n"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "c5886aeb",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import semantic_kernel as sk\n",
+    "from semantic_kernel.connectors.ai.open_ai import AzureChatCompletion, OpenAIChatCompletion\n",
+    "\n",
+    "kernel = sk.Kernel()\n",
+    "\n",
+    "useAzureOpenAI = False\n",
+    "\n",
+    "# Configure AI service used by the kernel\n",
+    "if useAzureOpenAI:\n",
+    "    deployment, api_key, endpoint = sk.azure_openai_settings_from_dot_env()\n",
+    "    kernel.add_chat_service(\n",
+    "        \"chat_completion\",\n",
+    "        AzureChatCompletion(\"gpt-35-turbo\", endpoint, api_key),\n",
+    "    )\n",
+    "else:\n",
+    "    api_key, org_id = sk.openai_settings_from_dot_env()\n",
+    "    kernel.add_chat_service(\n",
+    "        \"chat-gpt\", OpenAIChatCompletion(\"gpt-3.5-turbo\", api_key, org_id)\n",
+    "    )"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
@@ -170,7 +214,7 @@
     "sk_prompt = \"\"\"\n",
     "{{$input}}\n",
     "\n",
-    "Give me the TLDR in 5 words.\n",
+    "Give me the TLDR in 5 words or less.\n",
     "\"\"\"\n",
     "\n",
     "text = \"\"\"\n",
@@ -208,7 +252,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.10.9"
+   "version": "3.10.11"
   }
  },
  "nbformat": 4,

samples/notebooks/python/05-using-the-planner.ipynb

@@ -231,6 +231,108 @@
    "source": [
     "print(results)"
    ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "1d2f8d04",
+   "metadata": {},
+   "source": [
+    "# Let's see how the planner would work with a ChatCompletion model"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "25aa0e28",
+   "metadata": {},
+   "source": [
+    "Planner works best with more powerful models like `gpt4`, but because that's a ChatCompletion model, we need to modify the kernel to use that instead. Let's see how it works with the cheaper `gpt-35-turbo` model."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "06907360",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import semantic_kernel as sk\n",
+    "from semantic_kernel.connectors.ai.open_ai import AzureChatCompletion, OpenAIChatCompletion\n",
+    "\n",
+    "kernel = sk.Kernel()\n",
+    "\n",
+    "useAzureOpenAI = True\n",
+    "\n",
+    "# Configure AI service used by the kernel\n",
+    "if useAzureOpenAI:\n",
+    "    deployment, api_key, endpoint = sk.azure_openai_settings_from_dot_env()\n",
+    "    kernel.add_chat_service(\n",
+    "        \"chat_completion\",\n",
+    "        AzureChatCompletion(\"gpt-35-turbo\", endpoint, api_key),\n",
+    "    )\n",
+    "else:\n",
+    "    api_key, org_id = sk.openai_settings_from_dot_env()\n",
+    "    kernel.add_chat_service(\n",
+    "        \"chat-gpt\", OpenAIChatCompletion(\"gpt-3.5-turbo\", api_key, org_id)\n",
+    "    )"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "551cabd9",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from semantic_kernel.core_skills.text_skill import TextSkill\n",
+    "\n",
+    "skills_directory = \"../../skills/\"\n",
+    "summarize_skill = kernel.import_semantic_skill_from_directory(skills_directory, \"SummarizeSkill\")\n",
+    "writer_skill = kernel.import_semantic_skill_from_directory(skills_directory, \"WriterSkill\")\n",
+    "text_skill = kernel.import_skill(TextSkill(), \"TextSkill\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "183de28f",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "ask = \"\"\"\n",
+    "I need to plan a birthday party for my best friend. He loves racecars and only speaks Spanish.\n",
+    "Translate the ideas and make the text in all capital letters.\"\"\"\n",
+    "chatgpt_plan = await planner.create_plan_async(ask, kernel)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "a9546ed8",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "print(chatgpt_plan.generated_plan)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "21782c1d",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "results = await planner.execute_plan_async(chatgpt_plan, kernel)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "afe8170c",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "print(results)"
+   ]
   }
  ],
  "metadata": {
@@ -249,7 +351,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.10.9"
+   "version": "3.10.11"
   }
  },
  "nbformat": 4,