{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "2d664323-df2d-4f48-a745-195a3e5a8171",
   "metadata": {},
   "source": [
    "### Introduction\n",
    "\n",
    "After we have created a model we will typically want to save it as a binary file so it can be deployed for prediction.\n",
    "\n",
    "Seldon core is a model deployment engine.  Here is an excerpt from the seldon documentation for Scikit-learn deployments\n",
    "\n",
    "> Seldon expects that your model has been saved using joblib, and it named as model.joblib. Note that this is the recommended approach to serialise models by the SKLearn project.\n",
    ">\n",
    "> _Source: https://docs.seldon.io/projects/seldon-core/en/latest/servers/sklearn.html_"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d867015f-dd51-463d-8ae5-afba08444fe4",
   "metadata": {},
   "source": [
    "### Saving a model\n",
    "\n",
    "Here we use the [joblib](https://joblib.readthedocs.io/en/latest/persistence.html) Python library to save a model:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "6fa58627-7b0f-43b7-997e-ec0db3b29705",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.9/site-packages/sklearn/svm/_base.py:1206: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "['iris_model.joblib']"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# set the random state to make the examples repeatable\n",
    "import numpy as np\n",
    "np.random.seed(1)\n",
    "\n",
    "from sklearn.datasets import load_iris\n",
    "from sklearn.svm import LinearSVC\n",
    "\n",
    "iris = load_iris()\n",
    "X, y = iris.data, iris.target\n",
    "\n",
    "classifier = LinearSVC()\n",
    "model = classifier.fit(X, y)\n",
    "\n",
    "from joblib import dump, load\n",
    "dump(model, 'iris_model.joblib') "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5b39f67d-69b3-4a57-93db-1dcb1a936b31",
   "metadata": {},
   "source": [
    "We can execute a shell command with `!`. Let's inspect the saved file:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "4dc28f2b-973a-469f-baf3-9043961473c9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-rw-r--r-- 1 jovyan users 858 Feb 11 15:23 iris_model.joblib\n"
     ]
    }
   ],
   "source": [
    "! ls -l iris_model.joblib"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8b202252-35aa-4a5d-8466-4f0c85a158d8",
   "metadata": {},
   "source": [
    "### Summary\n",
    "\n",
    "More information on model persistence can be found in the Scikit-learn [docs](https://scikit-learn.org/stable/modules/model_persistence.html#)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9f253e47-5b03-4a85-bbd9-e5d3fedec1d8",
   "metadata": {},
   "source": [
    "---\n",
    "\n",
    "### Navigation\n",
    "\n",
    "[Previous](./09_model_performance.ipynb) | [Home](./00-README-FIRST.ipynb) | [Next](./11_model_predictions.ipynb) notebook"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "88691c8f-b8c7-48f1-a139-e12f8d9d821f",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}