RavenV2_Function_Calling_LLM.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "collapsed_sections": [
        "wom9c2gDPS1i",
        "kNaVsDsEiG9z",
        "0V3AuFLPSCCV",
        "ZR_pt8o7SGxp",
        "r2g172hXUMm0",
        "PZ2qbV1rXQQ8",
        "iGCmDoQB7OTN",
        "yvSxxClK8Xtw",
        "5DJuUO0jBgkL",
        "GciQ_zWl7VpH",
        "m2g-zm-OAb5s"
      ],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Troys-Code/AI_For_Free/blob/main/RavenV2_Function_Calling_LLM.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Welcome to RavenV2!\n",
        "\n",
        "This section introduces function calling and Raven!"
      ],
      "metadata": {
        "id": "wom9c2gDPS1i"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Introduction\n",
        "This document serves to walk you through prompting RavenV2, a fully open source function calling model."
      ],
      "metadata": {
        "id": "JILmZ0tPPXvn"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### What is Function Calling?"
      ],
      "metadata": {
        "id": "MBRzPwy4PxPS"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Function calling, as part of tool use by a large language model (LLM), refers to the process where the LLM identifies and executes predefined functions with relevant arguments. This is a key aspect of how the LLM utilizes its integrated tools to effectively respond to queries presented by users.\n",
        "\n",
        "For example, if the input to a function calling model is a function to get the weather (for example, `getWeather(city_name : str)`) and a user query asking about the weather in Seattle, the LLM's generation should be something like `getWeather(\"Seattle\")`.\n",
        "\n",
        "Of course, there can be many more complex incarnations of this, but this is a simple example of what Function Calling is."
      ],
      "metadata": {
        "id": "-CAMG8lBQ8cQ"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### Why is Function Calling Important?"
      ],
      "metadata": {
        "id": "I6G3V8eNQSjz"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Function calling is a crucial aspect of utilizing large language models (LLMs) because it significantly enhances their usefulness. By enabling function calling, LLMs can interact with a range of user-provided tools, extending beyond their native text processing abilities. This allows them to perform specialized tasks like API interfacing, data analysis, image generation, or web browsing, directly in response to user queries. Such a feature empowers LLMs to provide more accurate, detailed, and contextually relevant information. This not only improves the user experience but also broadens the scope of questions that can be addressed by the LLM, making it a helpful tool for a diverse array of applications. Function calling essentially transforms an LLM from a sophisticated text generator into a reasoning engine, capable of handling a variety of complex tasks seamlessly."
      ],
      "metadata": {
        "id": "iKEGJFKWQ6iM"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Why Raven?"
      ],
      "metadata": {
        "id": "cVlr7VzoupqB"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Not only is Raven good at Function Calling, but it also provides reasoning on why it issued those function calls, to justify the calls it has issued.\n",
        "\n",
        "This includes connecting the arguments it has filled in the function call back to the user's prompt, making the generation more interpretable.\n",
        "\n",
        "Please see the examples below!"
      ],
      "metadata": {
        "id": "zBMHjiuVusG4"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## NexusRaven-V2 Endpoint In This Notebook (Important!)"
      ],
      "metadata": {
        "id": "kNaVsDsEiG9z"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "We have provided an endpoint for users of this notebook to quickly experiment with NexusRaven-V2 to try it out on their own functions!\n",
        "\n",
        "**However, please note: we retain the right to turn off the endpoint in this notebook AT ANY POINT and for ANY REASON. Please DO NOT use this endpoint for any activity at scale, or for any activity beyond just simple local experimentation. **"
      ],
      "metadata": {
        "id": "d--yOhwFisk9"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "![image.png](data:image/png;base64,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)"
      ],
      "metadata": {
        "id": "KJekOb8limPT"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Prompting RavenV2 -- Weather API Example\n",
        "\n",
        "Let's use a few API functions to show the purpose (and power) of function calling!"
      ],
      "metadata": {
        "id": "0V3AuFLPSCCV"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Defining Functions"
      ],
      "metadata": {
        "id": "ZR_pt8o7SGxp"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Let's define some helpful functions for RavenV2. In this cell, we will define two functions -- a function to get a weather in a Latitude and Longitude, and a function to convert human-readable city names to the coordinates."
      ],
      "metadata": {
        "id": "PI5KyXiuSJnZ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import requests\n",
        "def get_weather_data(coordinates):\n",
        "    \"\"\"\n",
        "    Fetches weather data from the Open-Meteo API for the given latitude and longitude.\n",
        "\n",
        "    Args:\n",
        "    coordinates (tuple): The latitude of the location.\n",
        "\n",
        "    Returns:\n",
        "    float: The current temperature in the coordinates you've asked for\n",
        "    \"\"\"\n",
        "    latitude, longitude = coordinates\n",
        "    base_url = \"https://api.open-meteo.com/v1/forecast\"\n",
        "    params = {\n",
        "        \"latitude\": latitude,\n",
        "        \"longitude\": longitude,\n",
        "        \"current\": \"temperature_2m,wind_speed_10m\",\n",
        "        \"hourly\": \"temperature_2m,relative_humidity_2m,wind_speed_10m\"\n",
        "    }\n",
        "\n",
        "    response = requests.get(base_url, params=params)\n",
        "    if response.status_code == 200:\n",
        "        print (f\"\"\"Temperature in C: {response.json()[\"current\"][\"temperature_2m\"]}\"\"\")\n",
        "    else:\n",
        "        return {\"error\": \"Failed to fetch data, status code: {}\".format(response.status_code)}\n"
      ],
      "metadata": {
        "id": "JPHj-AWCSVeP"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "You'll notice that this API requires latitude and longitude, which is a bit hard for users to provide (or for LLMs to know). Let's add a new function that allows easy conversion."
      ],
      "metadata": {
        "id": "2fcXnzwWS1jG"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def get_coordinates_from_city(city_name):\n",
        "    \"\"\"\n",
        "    Fetches the latitude and longitude of a given city name using the Maps.co Geocoding API.\n",
        "\n",
        "    Args:\n",
        "    city_name (str): The name of the city.\n",
        "\n",
        "    Returns:\n",
        "    tuple: The latitude and longitude of the city.\n",
        "    \"\"\"\n",
        "    base_url = \"https://geocode.maps.co/search\"\n",
        "    params = {\"q\": city_name}\n",
        "\n",
        "    response = requests.get(base_url, params=params)\n",
        "    if response.status_code == 200:\n",
        "        data = response.json()\n",
        "        if data:\n",
        "            # Assuming the first result is the most relevant\n",
        "            return data[0][\"lat\"], data[0][\"lon\"]\n",
        "        else:\n",
        "            return {\"error\": \"No data found for the given city name.\"}\n",
        "    else:\n",
        "        return {\"error\": \"Failed to fetch data, status code: {}\".format(response.status_code)}\n"
      ],
      "metadata": {
        "id": "iBg9Za8fSDlq"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Asking Raven"
      ],
      "metadata": {
        "id": "r2g172hXUMm0"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Great! Now, let's ask Raven to use these tools to answer user questions.\n",
        "\n",
        "First, let's provide the functions to the model.\n",
        "\n",
        "Please note that the way you provide this is by providing the \"Function:\" header, the function prototype, and the associated docstring as seen below."
      ],
      "metadata": {
        "id": "y5pKNLKuUUcm"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "RAVEN_PROMPT = \\\n",
        "'''\n",
        "Function:\n",
        "def get_weather_data(coordinates):\n",
        "    \"\"\"\n",
        "    Fetches weather data from the Open-Meteo API for the given latitude and longitude.\n",
        "\n",
        "    Args:\n",
        "    coordinates (tuple): The latitude of the location.\n",
        "\n",
        "    Returns:\n",
        "    float: The current temperature in the coordinates you've asked for\n",
        "    \"\"\"\n",
        "\n",
        "Function:\n",
        "def get_coordinates_from_city(city_name):\n",
        "    \"\"\"\n",
        "    Fetches the latitude and longitude of a given city name using the Maps.co Geocoding API.\n",
        "\n",
        "    Args:\n",
        "    city_name (str): The name of the city.\n",
        "\n",
        "    Returns:\n",
        "    tuple: The latitude and longitude of the city.\n",
        "    \"\"\"\n",
        "\n",
        "User Query: {query}<human_end>\n",
        "'''"
      ],
      "metadata": {
        "id": "ocbkxsdPTUFC"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Let's provide Raven a question!"
      ],
      "metadata": {
        "id": "GcA5vqrmV8Jx"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "QUESTION = \"Whats's the weather like in Seattle right now?\""
      ],
      "metadata": {
        "id": "ydDTUJfXTf69"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Now, let's prompt Raven!"
      ],
      "metadata": {
        "id": "CdHTz9J-V5Hp"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Now, let's prompt Raven!\n",
        "API_URL = \"https://rjmy54al17scvxjr.us-east-1.aws.endpoints.huggingface.cloud\"\n",
        "headers = {\n",
        "        \"Content-Type\": \"application/json\"\n",
        "}\n",
        "def query(payload):\n",
        "\t\"\"\"\n",
        "\tSends a payload to a TGI endpoint.\n",
        "\t\"\"\"\n",
        "\timport requests\n",
        "\tresponse = requests.post(API_URL, headers=headers, json=payload)\n",
        "\treturn response.json()\n",
        "\n",
        "def query_raven(prompt):\n",
        "\t\"\"\"\n",
        "\tThis function sends a request to the TGI endpoint to get Raven's function call.\n",
        "\tThis will not generate Raven's justification and reasoning for the call, to save on latency.\n",
        "\t\"\"\"\n",
        "\timport requests\n",
        "\toutput = query({\n",
        "\t\t\"inputs\": prompt,\n",
        "\t\t\"parameters\" : {\"temperature\" : 0.001, \"stop\" : [\"<bot_end>\"], \"do_sample\" : False, \"max_new_tokens\" : 2000}})\n",
        "\tcall = output[0][\"generated_text\"].replace(\"Call:\", \"\").strip()\n",
        "\treturn call\n",
        "\n",
        "def query_raven_with_reasoning(prompt):\n",
        "\t\"\"\"\n",
        "\tThis function sends a request to the TGI endpoint to get Raven's function call AND justification for the call\n",
        "\t\"\"\"\n",
        "\timport requests\n",
        "\toutput = query({\n",
        "\t\t\"inputs\": prompt,\n",
        "\t\t\"parameters\" : {\"temperature\" : 0.001, \"do_sample\" : False, \"max_new_tokens\" : 2000}})\n",
        "\tcall = output[0][\"generated_text\"].replace(\"Call:\", \"\").strip()\n",
        "\treturn call"
      ],
      "metadata": {
        "id": "hWPinutTVHBg"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "In this cell, we can see Raven use the tools provided to it to give the User's query an answer."
      ],
      "metadata": {
        "id": "WU7lNbEVG9jC"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "my_question = RAVEN_PROMPT.format(query = QUESTION)\n",
        "raven_call = query_raven(my_question)\n",
        "print (f\"Raven's Call: {raven_call}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1nNya3QoVv9C",
        "outputId": "10ecbce8-6b9a-445e-ef0b-213359758d3c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Raven's Call: get_weather_data(coordinates=get_coordinates_from_city(city_name='Seattle'))\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Great! Now, let's execute the call to actually use the APIs!"
      ],
      "metadata": {
        "id": "5MGHzw4cW-Re"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "exec(raven_call)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "hVuTOpMSWFgO",
        "outputId": "52acae25-085c-449c-ffba-9d12d29d31c4"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Temperature in C: 5.8\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Awesome!! Seems it works :)"
      ],
      "metadata": {
        "id": "LPGa2Q8mXMvJ"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Raven can also continue the generation, allowing us to get a justification for the generated call."
      ],
      "metadata": {
        "id": "RD5F0p55tnoe"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "my_question = RAVEN_PROMPT.format(query = QUESTION)\n",
        "raven_call = query_raven_with_reasoning(my_question)\n",
        "print (f\"Raven's Call: {raven_call}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "I4WH2ywFtsjC",
        "outputId": "a78acafd-68c3-45bd-cfc7-4c215ec29904"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Raven's Call: get_weather_data(coordinates=get_coordinates_from_city(city_name='Seattle')) \n",
            "Thought: The function call `get_weather_data(coordinates=get_coordinates_from_city(city_name='Seattle'))` answers the question \"What's the weather like in Seattle right now?\" by following these steps:\n",
            "\n",
            "1. `get_coordinates_from_city(city_name='Seattle')`: This function call fetches the latitude and longitude of the city \"Seattle\" using the Maps.co Geocoding API.\n",
            "2. `get_weather_data(coordinates=...)`: This function call fetches the current weather data for the coordinates returned by the previous function call.\n",
            "\n",
            "Therefore, the function call `get_weather_data(coordinates=get_coordinates_from_city(city_name='Seattle'))` answers the question \"What's the weather like in Seattle right now?\" by first fetching the coordinates of the city \"Seattle\" and then fetching the current weather data for those coordinates.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Try Your Own!!\n",
        "\n",
        "Please feel free to edit the functions provided in this prompt with your own, and give it a try!"
      ],
      "metadata": {
        "id": "PZ2qbV1rXQQ8"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Endpoint\n",
        "First, let's define the endpoint specifics (including the URL to query Raven). Please do not change these, as these are more of utilities that you can use to send prompts to Raven and receive Raven's completions."
      ],
      "metadata": {
        "id": "pbEHdSZdGXqN"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "API_URL = \"https://rjmy54al17scvxjr.us-east-1.aws.endpoints.huggingface.cloud\"\n",
        "headers = {\n",
        "        \"Content-Type\": \"application/json\"\n",
        "}\n",
        "def query(payload):\n",
        "\t\"\"\"\n",
        "\tSends a payload to a TGI endpoint.\n",
        "\t\"\"\"\n",
        "\timport requests\n",
        "\tresponse = requests.post(API_URL, headers=headers, json=payload)\n",
        "\treturn response.json()\n",
        "\n",
        "def query_raven(prompt):\n",
        "\t\"\"\"\n",
        "\tThis function sends a request to the TGI endpoint to get Raven's function call.\n",
        "\tThis will not generate Raven's justification and reasoning for the call, to save on latency.\n",
        "\t\"\"\"\n",
        "\timport requests\n",
        "\toutput = query({\n",
        "\t\t\"inputs\": prompt,\n",
        "\t\t\"parameters\" : {\"temperature\" : 0.001, \"stop\" : [\"<bot_end>\"], \"do_sample\" : False, \"max_new_tokens\" : 2000}})\n",
        "\tcall = output[0][\"generated_text\"].replace(\"Call:\", \"\").strip()\n",
        "\treturn call\n",
        "\n",
        "def query_raven_with_reasoning(prompt):\n",
        "\t\"\"\"\n",
        "\tThis function sends a request to the TGI endpoint to get Raven's function call AND justification for the call\n",
        "\t\"\"\"\n",
        "\timport requests\n",
        "\toutput = query({\n",
        "\t\t\"inputs\": prompt,\n",
        "\t\t\"parameters\" : {\"temperature\" : 0.001, \"do_sample\" : False, \"max_new_tokens\" : 2000}})\n",
        "\tcall = output[0][\"generated_text\"].replace(\"Call:\", \"\").strip()\n",
        "\treturn call"
      ],
      "metadata": {
        "id": "QmPhxPrVX2T3"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Now, Let's Run Your Functions!\n",
        "\n",
        "Let's provide your functions to Raven and generate a call!\n",
        "\n",
        "In this cell, please feel free to edit the functions, the doc strings, and the arguments as you see fit! You're also welcome to edit the user query/question!"
      ],
      "metadata": {
        "id": "_ac50WRKGXCf"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "RAVEN_PROMPT = \\\n",
        "'''\n",
        "Function:\n",
        "def get_weather_data(coordinates):\n",
        "    \"\"\"\n",
        "    Fetches weather data from the Open-Meteo API for the given latitude and longitude.\n",
        "\n",
        "    Args:\n",
        "    coordinates (tuple): The latitude of the location.\n",
        "\n",
        "    Returns:\n",
        "    float: The current temperature in the coordinates you've asked for\n",
        "    \"\"\"\n",
        "\n",
        "Function:\n",
        "def get_coordinates_from_city(city_name):\n",
        "    \"\"\"\n",
        "    Fetches the latitude and longitude of a given city name using the Maps.co Geocoding API.\n",
        "\n",
        "    Args:\n",
        "    city_name (str): The name of the city.\n",
        "\n",
        "    Returns:\n",
        "    tuple: The latitude and longitude of the city.\n",
        "    \"\"\"\n",
        "\n",
        "User Query: {query}<human_end>\n",
        "'''\n",
        "\n",
        "QUESTION = \"Whats's the weather like in Seattle right now?\"\n",
        "my_question = RAVEN_PROMPT.format(query = QUESTION)"
      ],
      "metadata": {
        "id": "kYvz-L9YXTxv"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Now that we've constructed the question and the list of tools/functions, let's send it over to Raven!"
      ],
      "metadata": {
        "id": "XHA6xwUbIaTw"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "raven_call = query_raven(my_question)\n",
        "print (f\"Raven's Call: {raven_call}\")\n",
        "\n",
        "raven_call = query_raven_with_reasoning(my_question)\n",
        "print (f\"Raven's Reasoning: {raven_call}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "jbIPOXUBIW8G",
        "outputId": "7e456eb9-ec27-48bf-a724-f4857294e9f6"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Raven's Call: get_weather_data(coordinates=get_coordinates_from_city(city_name='Seattle'))\n",
            "Raven's Reasoning: get_weather_data(coordinates=get_coordinates_from_city(city_name='Seattle')) \n",
            "Thought: The function call `get_weather_data(coordinates=get_coordinates_from_city(city_name='Seattle'))` answers the question \"What's the weather like in Seattle right now?\" by following these steps:\n",
            "\n",
            "1. `get_coordinates_from_city(city_name='Seattle')`: This function call fetches the latitude and longitude of the city \"Seattle\" using the Maps.co Geocoding API.\n",
            "2. `get_weather_data(coordinates=...)`: This function call fetches the current weather data for the coordinates returned by the previous function call.\n",
            "\n",
            "Therefore, the function call `get_weather_data(coordinates=get_coordinates_from_city(city_name='Seattle'))` answers the question \"What's the weather like in Seattle right now?\" by first fetching the coordinates of the city \"Seattle\" and then fetching the current weather data for those coordinates.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Advanced Prompting Guide\n",
        "\n",
        "More advanced prompting strategies!"
      ],
      "metadata": {
        "id": "iGCmDoQB7OTN"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Define the Endpoint For Raven\n",
        "\n",
        "Please run these cells to get access to the endpoint!"
      ],
      "metadata": {
        "id": "yvSxxClK8Xtw"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "API_URL = \"https://rjmy54al17scvxjr.us-east-1.aws.endpoints.huggingface.cloud\"\n",
        "headers = {\n",
        "        \"Content-Type\": \"application/json\"\n",
        "}\n",
        "def query(payload):\n",
        "\t\"\"\"\n",
        "\tSends a payload to a TGI endpoint.\n",
        "\t\"\"\"\n",
        "\timport requests\n",
        "\tresponse = requests.post(API_URL, headers=headers, json=payload)\n",
        "\treturn response.json()\n",
        "\n",
        "def query_raven(prompt):\n",
        "\t\"\"\"\n",
        "\tThis function sends a request to the TGI endpoint to get Raven's function call.\n",
        "\tThis will not generate Raven's justification and reasoning for the call, to save on latency.\n",
        "\t\"\"\"\n",
        "\timport requests\n",
        "\toutput = query({\n",
        "\t\t\"inputs\": prompt,\n",
        "\t\t\"parameters\" : {\"temperature\" : 0.001, \"stop\" : [\"<bot_end>\"], \"do_sample\" : False, \"max_new_tokens\" : 2000}})\n",
        "\tcall = output[0][\"generated_text\"].replace(\"Call:\", \"\").strip()\n",
        "\treturn call\n",
        "\n",
        "def query_raven_with_reasoning(prompt):\n",
        "\t\"\"\"\n",
        "\tThis function sends a request to the TGI endpoint to get Raven's function call AND justification for the call\n",
        "\t\"\"\"\n",
        "\timport requests\n",
        "\toutput = query({\n",
        "\t\t\"inputs\": prompt,\n",
        "\t\t\"parameters\" : {\"temperature\" : 0.001, \"do_sample\" : False, \"max_new_tokens\" : 2000}})\n",
        "\tcall = output[0][\"generated_text\"].replace(\"Call:\", \"\").strip()\n",
        "\treturn call\n"
      ],
      "metadata": {
        "id": "vGSBt8gM8Ukt"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Parallel Calls\n",
        "Raven is also capable of issuing Parallel Calls! Let's see a graph example!"
      ],
      "metadata": {
        "id": "5DJuUO0jBgkL"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "RAVEN_PROMPT = \\\n",
        "'''\n",
        "Function:\n",
        "def add_edge(u, v):\n",
        "    \"\"\"\n",
        "    Adds an edge between node u and node v in the graph. Make sure to create a graph first by calling create_new_graph!\n",
        "\n",
        "    Args:\n",
        "      u (str): Node name as string\n",
        "      v (str): Node name as string\n",
        "    \"\"\"\n",
        "\n",
        "Function:\n",
        "def is_two_nodes_connected(u, v):\n",
        "    \"\"\"\n",
        "    Answers if two nodes are connected.\n",
        "    \"\"\"\n",
        "\n",
        "User Query: Emma is friends with Bob and Charlie, and Charlie is friends with Erik, and Erik is friends with Brian. Can you represent all of these relationship as a graph and answer if Emma is friends with Erik?<human_end>\n",
        "'''\n",
        "print (query_raven_with_reasoning(RAVEN_PROMPT))\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5WedMxGrBg7S",
        "outputId": "19a0c28f-89ca-458a-9d7b-a782e68c645d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "add_edge(u='Emma', v='Bob'); add_edge(u='Emma', v='Charlie'); add_edge(u='Charlie', v='Erik'); add_edge(u='Erik', v='Brian'); is_two_nodes_connected(u='Emma', v='Erik'); \n",
            "Thought: The set of Python calls provided will directly answer the question because they perform the following operations:\n",
            "\n",
            "1. `add_edge(u='Emma', v='Bob')`: This call adds an edge between the nodes \"Emma\" and \"Bob\" in the graph.\n",
            "2. `add_edge(u='Emma', v='Charlie')`: This call adds an edge between the nodes \"Emma\" and \"Charlie\" in the graph.\n",
            "3. `add_edge(u='Charlie', v='Erik')`: This call adds an edge between the nodes \"Charlie\" and \"Erik\" in the graph.\n",
            "4. `add_edge(u='Erik', v='Brian')`: This call adds an edge between the nodes \"Erik\" and \"Brian\" in the graph.\n",
            "5. `is_two_nodes_connected(u='Emma', v='Erik')`: This call checks if the nodes \"Emma\" and \"Erik\" are connected in the graph.\n",
            "\n",
            "The `add_edge` calls will create the necessary edges in the graph to represent the relationships between the nodes. The `is_two_nodes_connected` call will check if the nodes \"Emma\" and \"Erik\" are connected, which will answer the question of whether Emma is friends with Erik.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Optional Arguments\n",
        "\n",
        "Raven is also capable of handling optional arguments. You can provide it as follows:"
      ],
      "metadata": {
        "id": "GciQ_zWl7VpH"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "RAVEN_PROMPT = \\\n",
        "'''\n",
        "Function:\n",
        "def getWeather(city_name : str, use_celsius : bool = False):\n",
        "\"\"\"\n",
        "Gets weather for a specific city, and other associated attributes.\n",
        "\n",
        "Args:\n",
        "- city_name: The name of the city as a string.\n",
        "- use_celsius (Optional): If set, returns the temperature in celsius. Otherwise, by default, returns in Farenheit.\n",
        "\"\"\"\n",
        "\n",
        "User Query: {query}<human_end>\n",
        "\n",
        "'''\n",
        "\n",
        "QUESTION = \"Hey, what's the weather like in New York right now?\"\n",
        "\n",
        "raven_question = RAVEN_PROMPT.format(query=QUESTION)\n",
        "print (query_raven(raven_question))\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4uTKSGK97QLH",
        "outputId": "a78c9824-d479-4b46-e8aa-12a4e033f67f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "getWeather(city_name='New York')\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Notice how the optional argument is NOT used. But, let's change the prompt a bit."
      ],
      "metadata": {
        "id": "0VNTea_v8eK9"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "QUESTION = \"Respond in SI units and tell me what's the conditions like in Seattle right now?\"\n",
        "raven_question = RAVEN_PROMPT.format(query=QUESTION)\n",
        "print (query_raven(raven_question))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "K_AGzcYI8ceD",
        "outputId": "e9e81e4d-18c1-41d7-f2a5-4dfd8214d2e5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "getWeather(city_name='Seattle', use_celsius=True)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Notice how the optional argument is now used!"
      ],
      "metadata": {
        "id": "xQcrps1w9Jvp"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Explanability\n",
        "\n",
        "Raven is also capable of giving well reasoned explanations on *why* these function calls were issued. To save on inference speed, we set a stop criteria of `<bot_end>` to prevent Raven from issuing the explanation. But if required, we can set the stop criteria to </s> to fully capitalize on Raven's ability to reason to provide justifications for the calls."
      ],
      "metadata": {
        "id": "m2g-zm-OAb5s"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "RAVEN_PROMPT = \\\n",
        "'''\n",
        "Function:\n",
        "def getWeather(city_name : str, use_celsius : bool = False):\n",
        "\"\"\"\n",
        "Gets weather for a specific city, and other associated attributes.\n",
        "\n",
        "Args:\n",
        "- city_name: The name of the city as a string.\n",
        "- use_celsius (Optional): If set, returns the temperature in celsius. Otherwise, by default, returns in Farenheit.\n",
        "\"\"\"\n",
        "\n",
        "User Query: {query}<human_end>\n",
        "\n",
        "'''\n",
        "\n",
        "QUESTION = \"Respond in SI units and tell me what's the conditions like in Seattle right now?\"\n",
        "raven_question = RAVEN_PROMPT.format(query=QUESTION)\n",
        "print (query_raven_with_reasoning(raven_question))"
      ],
      "metadata": {
        "id": "7wb-AckcAC5e",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "c1b8bf32-8a5d-462c-b1c1-c15774fae3f2"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "getWeather(city_name='Seattle', use_celsius=True) \n",
            "Thought: The function call `getWeather(city_name='Seattle', use_celsius=True)` answers the question \"What's the conditions like in Seattle right now?\" because it retrieves the weather data for the city of Seattle and returns it in SI units (celsius).\n",
            "\n",
            "Here's why:\n",
            "\n",
            "1. The `city_name` parameter is set to `'Seattle'`, which specifies the city for which the weather data should be retrieved.\n",
            "2. The `use_celsius` parameter is set to `True`, which tells the function to return the temperature in celsius instead of farenheit.\n",
            "3. The function call `getWeather(city_name='Seattle', use_celsius=True)` retrieves the weather data for the city of Seattle and returns it in celsius.\n",
            "\n",
            "Therefore, the function call `getWeather(city_name='Seattle', use_celsius=True)` answers the question \"What's the conditions like in Seattle right now?\" by retrieving the weather data for the city of Seattle and returning it in SI units (celsius).\n"
          ]
        }
      ]
    }
  ]
}