cpd5.1/notebooks/python_sdk/deployments/foundation_models/prompt_tuning/Use watsonx to tune IBM 'granite-13b-instruct-v2' model with cars4u document.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "![image](https://raw.githubusercontent.com/IBM/watson-machine-learning-samples/master/cloud/notebooks/headers/watsonx-Prompt_Lab-Notebook.png)\n",
        "# Use watsonx to tune IBM `granite-13b-instruct-v2` model with Car Rental Company customer satisfaction document"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "#### Disclaimers\n",
        "\n",
        "- Use only Projects and Spaces that are available in watsonx context.\n",
        "\n",
        "\n",
        "## Notebook content\n",
        "\n",
        "This notebook contains the steps and code to demonstrate support of prompt tuning in watsonx.\n",
        "\n",
        "Some familiarity with Python is helpful. This notebook uses Python 3.11.\n",
        "\n",
        "\n",
        "## Learning goal\n",
        "\n",
        "The goal of this notebook is to demonstrate how to:\n",
        "- upload of dataset with prompts to container (project)\n",
        "- trigger prompt tuning process\n",
        "- review of prompt tuning process details/results\n",
        "- deploy prompt-tuned model/asset\n",
        "- evaluation (via inference) of prompt-tuned model/asset\n",
        "\n",
        "\n",
        "## Contents\n",
        "\n",
        "This notebook contains the following parts:\n",
        "\n",
        "- [Setup](#setup)\n",
        "- [Data loading](#data_loading)\n",
        "- [Initialize experiment](#initialize)\n",
        "- [Run Prompt Tuning](#run_tuning)\n",
        "- [Prompt Tuning details](#run_details)\n",
        "- [Deploy](#deploy)\n",
        "- [Foundation Models Inference](#models_inference)\n",
        "- [Analyze the satisfaction](#predict)\n",
        "- [Summary and next steps](#summary)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "<a id=\"setup\"></a>\n",
        "## Set up the environment\n",
        "\n",
        "Before you use the sample code in this notebook, you must perform the following setup tasks:\n",
        "\n",
        "-  Contact with your Cloud Pack for Data administrator and ask him for your account credentials"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "### Install and import the `ibm-watsonx-ai` and dependecies\n",
        "**Note:** `ibm-watsonx-ai` documentation can be found <a href=\"https://ibm.github.io/watsonx-ai-python-sdk/index.html\" target=\"_blank\" rel=\"noopener no referrer\">here</a>."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [],
      "source": [
        "!pip install \"scikit-learn==1.3.2\" | tail -n 1\n",
        "!pip install -U ibm-watsonx-ai | tail -n 1\n",
        "!pip install matplotlib | tail -n 1\n",
        "!pip install wget | tail -n 1"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "### Connection to WML\n",
        "\n",
        "Authenticate the Watson Machine Learning service on IBM Cloud Pack for Data. You need to provide platform `url`, your `username` and `api_key`."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [],
      "source": [
        "username = 'PASTE YOUR USERNAME HERE'\n",
        "api_key = 'PASTE YOUR API_KEY HERE'\n",
        "url = 'PASTE THE PLATFORM URL HERE'"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "from ibm_watsonx_ai import Credentials\n",
        "\n",
        "credentials = Credentials(\n",
        "    username=username,\n",
        "    api_key=api_key,\n",
        "    url=url,\n",
        "    instance_id=\"openshift\",\n",
        "    version=\"5.1\"\n",
        ")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "Alternatively you can use `username` and `password` to authenticate WML services.\n",
        "\n",
        "```python\n",
        "credentials = Credentials(\n",
        "    username=***,\n",
        "    password=***,\n",
        "    url=***,\n",
        "    instance_id=\"openshift\",\n",
        "    version=\"5.1\"\n",
        ")\n",
        "\n",
        "```"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "### Defining the project id\n",
        "The Foundation Model requires project id that provides the context for the call. We will obtain the id from the project in which this notebook runs. Otherwise, please provide the project id."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [],
      "source": [
        "import os\n",
        "\n",
        "try:\n",
        "    project_id = os.environ[\"PROJECT_ID\"]\n",
        "except KeyError:\n",
        "    project_id = input(\"Please enter your project_id (hit enter): \")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "Create an instance of APIClient with authentication details."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [],
      "source": [
        "from ibm_watsonx_ai import APIClient\n",
        "\n",
        "client = APIClient(credentials)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "To be able to interact with all resources available in Watson Machine Learning, you need to set **project_id** which you will be using."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "'SUCCESS'"
            ]
          },
          "execution_count": 4,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "client.set.default_project(project_id)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "<a id=\"data_loading\"></a>\n",
        "## Data loading"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "This example uses the Consumer Financial Protection Bureau training dataset. "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [],
      "source": [
        "filename = 'car_rental_prompt_tuning_training_data.json'"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "Download training data from git repository and create data assets."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Creating data asset...\n",
            "SUCCESS\n"
          ]
        }
      ],
      "source": [
        "import wget\n",
        "import os\n",
        "\n",
        "url = \"https://raw.github.com/IBM/watson-machine-learning-samples/master/cloud/data/prompt_tuning/car_rental_prompt_tuning_training_data.json\"\n",
        "if not os.path.isfile(filename):\n",
        "    wget.download(url)\n",
        "\n",
        "asset_details = client.data_assets.create(name=filename, file_path=filename)\n",
        "asset_id = client.data_assets.get_id(asset_details)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "collapsed": false,
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "bd60f1d0-1aaf-4753-8ca0-59c288983997\n"
          ]
        }
      ],
      "source": [
        "print(asset_id)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "Define connection information to training data."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [],
      "source": [
        "from ibm_watsonx_ai.helpers import DataConnection\n",
        "\n",
        "data_conn = DataConnection(data_asset_id=asset_id)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "<a id=\"initialize\"></a>\n",
        "## Initialize experiment "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [],
      "source": [
        "from ibm_watsonx_ai.experiment import TuneExperiment\n",
        "\n",
        "experiment = TuneExperiment(credentials, project_id=project_id)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "All avaliable tasks are presented under `Tasks` Enum"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "{'QUESTION_ANSWERING': 'question_answering',\n",
              " 'SUMMARIZATION': 'summarization',\n",
              " 'RETRIEVAL_AUGMENTED_GENERATION': 'retrieval_augmented_generation',\n",
              " 'CLASSIFICATION': 'classification',\n",
              " 'GENERATION': 'generation',\n",
              " 'CODE_GENERATION_AND_CONVERSION': 'code',\n",
              " 'EXTRACTION': 'extraction',\n",
              " 'TRANSLATION': 'translation'}"
            ]
          },
          "execution_count": 10,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "{task.name: task.value for task in experiment.Tasks}"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "Define `prompt_tuner` parameters. \n",
        "\n",
        "**Action:** For parameters details, see <a href=\"https://ibm.github.io/watsonx-ai-python-sdk/tune_experiment.html#ibm_watsonx_ai.experiment.fm_tune.TuneExperiment.prompt_tuner\" target=\"_blank\" rel=\"noopener no referrer\">documentation</a>."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [],
      "source": [
        "prompt_tuner = experiment.prompt_tuner(name=\"granite-13b-instruct-v2 prompt tuning - python SDK\",\n",
        "                                       task_id=experiment.Tasks.CLASSIFICATION,\n",
        "                                       base_model='ibm/granite-13b-instruct-v2',\n",
        "                                       accumulate_steps=16,\n",
        "                                       batch_size=8,\n",
        "                                       learning_rate=0.0002,\n",
        "                                       max_input_tokens=128,\n",
        "                                       max_output_tokens=2,\n",
        "                                       num_epochs=6,\n",
        "                                       tuning_type=experiment.PromptTuningTypes.PT,\n",
        "                                       init_text = \"Extract the satisfaction from the comment. Return simple '1' for satisfied customer or '0' for unsatisfied. Comment:\",\n",
        "                                       init_method = \"text\",\n",
        "                                       verbalizer=\"{{input}}\",\n",
        "                                       auto_update_model=True\n",
        "                                       )"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "You can review previous set parameters"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 12,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "{'base_model': {'model_id': 'ibm/granite-13b-instruct-v2'},\n",
              " 'accumulate_steps': 16,\n",
              " 'batch_size': 8,\n",
              " 'init_method': 'text',\n",
              " 'init_text': \"Extract the satisfaction from the comment. Return simple '1' for satisfied customer or '0' for unsatisfied. Comment:\",\n",
              " 'learning_rate': 0.0002,\n",
              " 'max_input_tokens': 128,\n",
              " 'max_output_tokens': 2,\n",
              " 'num_epochs': 6,\n",
              " 'task_id': 'classification',\n",
              " 'tuning_type': 'prompt_tuning',\n",
              " 'verbalizer': '{{input}}',\n",
              " 'name': 'granite-13b-instruct-v2 prompt tuning - python SDK',\n",
              " 'description': 'Prompt tuning with SDK',\n",
              " 'auto_update_model': True,\n",
              " 'group_by_name': False}"
            ]
          },
          "execution_count": 12,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "prompt_tuner.get_params()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "<a id=\"run_tuning\"></a>\n",
        "## Run Prompt Tuning"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "Run a prompt tuning process of foundation model on top of the training data referenced by DataConnection (tuning may take some time).\n",
        "\n",
        "\n",
        "By changing the `background_mode` parameter to `True`, the prompt tuning process will run in the background."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "\n",
            "##############################################\n",
            "\n",
            "Running 'dd77b3d1-1c08-49e6-921f-03cf3886ac54'\n",
            "\n",
            "##############################################\n",
            "\n",
            "\n",
            "pending...\n",
            "running................................................................\n",
            "completed\n",
            "Training of 'dd77b3d1-1c08-49e6-921f-03cf3886ac54' finished successfully.\n"
          ]
        }
      ],
      "source": [
        "tuning_details = prompt_tuner.run(training_data_references=[data_conn], background_mode=False)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "<a id=\"run_details\"></a>\n",
        "## Prompt Tuning details"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "Check status/state of initialized Prompt Tuning run if ran in background mode or when process finish if background mode is off."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 14,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "'completed'"
            ]
          },
          "execution_count": 14,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "prompt_tuner.get_run_status()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "Get prompt tuning run details."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "{'metadata': {'created_at': '2024-05-06T09:58:06.361Z',\n",
              "  'description': 'Prompt tuning with SDK',\n",
              "  'id': 'dd77b3d1-1c08-49e6-921f-03cf3886ac54',\n",
              "  'modified_at': '2024-05-06T10:04:44.123Z',\n",
              "  'name': 'granite-13b-instruct-v2 prompt tuning - python SDK',\n",
              "  'project_id': '5d52a065-560b-46ce-a7a0-7e0cfd0ee722',\n",
              "  'tags': ['prompt_tuning',\n",
              "   'wx_prompt_tune.0f9ba108-d9f3-4151-b660-22646333272a']},\n",
              " 'entity': {'auto_update_model': True,\n",
              "  'description': 'Prompt tuning with SDK',\n",
              "  'model_id': 'de58d880-0a41-45aa-a031-a4e5fe7cd5ed',\n",
              "  'name': 'granite-13b-instruct-v2 prompt tuning - python SDK',\n",
              "  'project_id': '5d52a065-560b-46ce-a7a0-7e0cfd0ee722',\n",
              "  'prompt_tuning': {'accumulate_steps': 16,\n",
              "   'base_model': {'model_id': 'ibm/granite-13b-instruct-v2'},\n",
              "   'batch_size': 8,\n",
              "   'init_method': 'text',\n",
              "   'init_text': \"Extract the satisfaction from the comment. Return simple '1' for satisfied customer or '0' for unsatisfied. Comment:\",\n",
              "   'learning_rate': 0.0002,\n",
              "   'max_input_tokens': 128,\n",
              "   'max_output_tokens': 2,\n",
              "   'num_epochs': 6,\n",
              "   'num_virtual_tokens': 100,\n",
              "   'task_id': 'classification',\n",
              "   'tuning_type': 'prompt_tuning',\n",
              "   'verbalizer': '{{input}}'},\n",
              "  'results_reference': {'connection': {},\n",
              "   'location': {'path': '/projects/5d52a065-560b-46ce-a7a0-7e0cfd0ee722/assets/wx_prompt_tune',\n",
              "    'training': '/projects/5d52a065-560b-46ce-a7a0-7e0cfd0ee722/assets/wx_prompt_tune/dd77b3d1-1c08-49e6-921f-03cf3886ac54',\n",
              "    'training_status': '/projects/5d52a065-560b-46ce-a7a0-7e0cfd0ee722/assets/wx_prompt_tune/dd77b3d1-1c08-49e6-921f-03cf3886ac54/training-status.json',\n",
              "    'model_request_path': '/projects/5d52a065-560b-46ce-a7a0-7e0cfd0ee722/assets/wx_prompt_tune/dd77b3d1-1c08-49e6-921f-03cf3886ac54/assets/dd77b3d1-1c08-49e6-921f-03cf3886ac54/resources/wml_model/request.json',\n",
              "    'assets_path': '/projects/5d52a065-560b-46ce-a7a0-7e0cfd0ee722/assets/wx_prompt_tune/dd77b3d1-1c08-49e6-921f-03cf3886ac54/assets'},\n",
              "   'type': 'fs'},\n",
              "  'status': {'completed_at': '2024-05-06T10:04:44.108Z',\n",
              "   'running_at': '2024-05-06T09:58:35.270Z',\n",
              "   'state': 'completed'},\n",
              "  'tags': ['prompt_tuning',\n",
              "   'wx_prompt_tune.0f9ba108-d9f3-4151-b660-22646333272a'],\n",
              "  'training_data_references': [{'connection': {},\n",
              "    'location': {'href': '/v2/assets/bd60f1d0-1aaf-4753-8ca0-59c288983997',\n",
              "     'id': 'bd60f1d0-1aaf-4753-8ca0-59c288983997'},\n",
              "    'type': 'data_asset'}]}}"
            ]
          },
          "execution_count": 15,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "prompt_tuner.get_run_details()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "Let's summarize the run."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Enhancements</th>\n",
              "      <th>Base model</th>\n",
              "      <th>Auto store</th>\n",
              "      <th>Epochs</th>\n",
              "      <th>loss</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Model Name</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>model_dd77b3d1-1c08-49e6-921f-03cf3886ac54</th>\n",
              "      <td>[prompt_tuning]</td>\n",
              "      <td>ibm/granite-13b-instruct-v2</td>\n",
              "      <td>True</td>\n",
              "      <td>6</td>\n",
              "      <td>0.1123</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                               Enhancements  \\\n",
              "Model Name                                                    \n",
              "model_dd77b3d1-1c08-49e6-921f-03cf3886ac54  [prompt_tuning]   \n",
              "\n",
              "                                                             Base model  \\\n",
              "Model Name                                                                \n",
              "model_dd77b3d1-1c08-49e6-921f-03cf3886ac54  ibm/granite-13b-instruct-v2   \n",
              "\n",
              "                                            Auto store  Epochs    loss  \n",
              "Model Name                                                              \n",
              "model_dd77b3d1-1c08-49e6-921f-03cf3886ac54        True       6  0.1123  "
            ]
          },
          "execution_count": 16,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "prompt_tuner.summary()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "collapsed": false,
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "Plot learning curves."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {
        "collapsed": false,
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 1800x400 with 3 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "prompt_tuner.plot_learning_curve()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "<a id=\"deploy\"></a>\n",
        "## Deploy"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "You can specify `model_id` from tuning details."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "'de58d880-0a41-45aa-a031-a4e5fe7cd5ed'"
            ]
          },
          "execution_count": 18,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "model_id = None\n",
        "\n",
        "if 'model_id' in tuning_details.get('entity', {}):\n",
        "    model_id = tuning_details['entity']['model_id']\n",
        "model_id"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "Create online deployment for published model."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "\n",
            "######################################################################################\n",
            "\n",
            "Synchronous deployment creation for id: 'de58d880-0a41-45aa-a031-a4e5fe7cd5ed' started\n",
            "\n",
            "######################################################################################\n",
            "\n",
            "\n",
            "initializing.\n",
            "ready\n",
            "\n",
            "\n",
            "-----------------------------------------------------------------------------------------------\n",
            "Successfully finished deployment creation, deployment_id='73913b73-7365-4b66-b147-05df08c1a437'\n",
            "-----------------------------------------------------------------------------------------------\n",
            "\n",
            "\n"
          ]
        }
      ],
      "source": [
        "from datetime import datetime\n",
        "meta_props = {\n",
        "    client.deployments.ConfigurationMetaNames.NAME: \"PT DEPLOYMENT SDK - project\",\n",
        "    client.deployments.ConfigurationMetaNames.ONLINE: {},\n",
        "    client.deployments.ConfigurationMetaNames.SERVING_NAME : f\"pt_sdk_deployment_{datetime.utcnow().strftime('%Y_%m_%d_%H%M%S')}\"\n",
        "}\n",
        "deployment_details = client.deployments.create(model_id, meta_props)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "Additionally you can get deployment details by printing `deployment_details`"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [],
      "source": [
        "deployment_details"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "You can specify `deployment_id` from deployment details."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 20,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "'73913b73-7365-4b66-b147-05df08c1a437'"
            ]
          },
          "execution_count": 20,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "deployment_id = deployment_details['metadata']['id']\n",
        "deployment_id"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "<a id=\"models_inference\"></a>\n",
        "## Foundation Models Inference on `watsonx.ai`"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "Provide a set of model parameters."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 21,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [],
      "source": [
        "from ibm_watsonx_ai.metanames import GenTextParamsMetaNames as GenParams\n",
        "\n",
        "generate_params = {\n",
        "    GenParams.MAX_NEW_TOKENS: 3,\n",
        "    GenParams.STOP_SEQUENCES: [\"\\n\\n\"]\n",
        "}"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "Initialize the `ModelInference` class."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 22,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [],
      "source": [
        "from ibm_watsonx_ai.foundation_models import ModelInference\n",
        "\n",
        "tuned_model = ModelInference(\n",
        "    deployment_id=deployment_id,\n",
        "    params=generate_params,\n",
        "    api_client=client\n",
        ")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "Alternatively you can use `credentials` and `project_id` to initialize the `ModelInference` class.\n",
        "\n",
        "```\n",
        "tuned_model = ModelInference(\n",
        "    deployment_id=deployment_id,\n",
        "    params=generate_params,\n",
        "    credentials=credentials,\n",
        "    project_id=project_id\n",
        ")\n",
        "\n",
        "```"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "Get deployment model inference details."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [],
      "source": [
        "tuned_model.get_details()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "Analyze the product class for a sample prompt."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 24,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "'0'"
            ]
          },
          "execution_count": 24,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "response = tuned_model.generate_text(prompt=\"Extract the satisfaction from the comment. Return simple 1 for satisfied customer or 0 for unsatisfied.\\nComment: I was penalty charged because vehicle was not returned with full tank.\\nSatisfaction:\\n\")\n",
        "response"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "<a id=\"predict\"></a>\n",
        "## Analyze the product classes."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "Download and prepare the `cfpb_test` dataset."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 25,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [],
      "source": [
        "import pandas as pd\n",
        "\n",
        "filename = 'car_rental_prompt_tuning_testing_data.json'\n",
        "\n",
        "url = \"https://raw.github.com/IBM/watson-machine-learning-samples/master/cloud/data/prompt_tuning/car_rental_prompt_tuning_testing_data.json\"\n",
        "if not os.path.isfile(filename):\n",
        "    wget.download(url)\n",
        "\n",
        "data = pd.read_json(filename)\n",
        "prompts = list(data.input)\n",
        "satisfaction = list(data.output)\n",
        "prompts_batch = [\"\\n\".join([prompt]) for prompt in prompts]"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "### Calculate the accuracy of tuned model"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 26,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [],
      "source": [
        "tuned_model_results = tuned_model.generate_text(prompt=prompts_batch)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 27,
      "metadata": {
        "collapsed": false,
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [],
      "source": [
        "tuned_model_results = [int(x) for x in tuned_model_results]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 28,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "accuracy_score: 0.916083916083916\n"
          ]
        }
      ],
      "source": [
        "from sklearn.metrics import accuracy_score\n",
        "\n",
        "print(f'accuracy_score: {accuracy_score(satisfaction, tuned_model_results)}')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "### Calculate the accuracy of base model"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "Initialize the `ModelInference` class with base model."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 29,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [],
      "source": [
        "base_model = ModelInference(\n",
        "    model_id='ibm/granite-13b-instruct-v2',\n",
        "    params=generate_params,\n",
        "    api_client=client\n",
        ")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 30,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [],
      "source": [
        "base_model_results = base_model.generate_text(prompt=prompts_batch)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 31,
      "metadata": {
        "collapsed": false,
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [],
      "source": [
        "base_model_results = [int(x) for x in base_model_results]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 32,
      "metadata": {
        "pycharm": {
          "name": "#%%\n"
        }
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "base model accuracy_score: 0.8181818181818182\n"
          ]
        }
      ],
      "source": [
        "from sklearn.metrics import accuracy_score\n",
        "\n",
        "print(f'base model accuracy_score: {accuracy_score(satisfaction, base_model_results)}')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "<a id=\"summary\"></a>\n",
        "## Summary and next steps\n",
        "\n",
        " You successfully completed this notebook!\n",
        " \n",
        " You learned how to use prompt tuning in watsonx for analyze the product classes for a sample prompts.\n",
        " \n",
        "Check out our _<a href=\"https://ibm.github.io/watson-machine-learning-sdk/samples.html\" target=\"_blank\" rel=\"noopener no referrer\">Online Documentation</a>_ for more samples, tutorials, documentation, how-tos, and blog posts. "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "### Authors: \n",
        "\n",
        "**Lukasz Cmielowski, PhD**, Senior Technical Staff Member at Watson Machine Learning.\n",
        "\n",
        "**Mateusz Szewczyk**, Software Engineer at Watson Machine Learning."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "pycharm": {
          "name": "#%% md\n"
        }
      },
      "source": [
        "Copyright © 2024-2025 IBM. This notebook and its source code are released under the terms of the MIT License."
      ]
    }
  ],
  "metadata": {
    "kernelspec": {
      "display_name": "notebook-samples",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.11.9"
    },
    "orig_nbformat": 4
  },
  "nbformat": 4,
  "nbformat_minor": 2
}